diff --git "a/checkpoint-19000/trainer_state.json" "b/checkpoint-19000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-19000/trainer_state.json" @@ -0,0 +1,133041 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 19000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 5.2631578947368424e-05, + "grad_norm": 4.1875, + "learning_rate": 1.0526315789473685e-07, + "loss": 10.9573, + "step": 1 + }, + { + "epoch": 0.00010526315789473685, + "grad_norm": 3.53125, + "learning_rate": 2.105263157894737e-07, + "loss": 10.9581, + "step": 2 + }, + { + "epoch": 0.00015789473684210527, + "grad_norm": 3.734375, + "learning_rate": 3.1578947368421055e-07, + "loss": 10.9436, + "step": 3 + }, + { + "epoch": 0.0002105263157894737, + "grad_norm": 3.6875, + "learning_rate": 4.210526315789474e-07, + "loss": 10.9499, + "step": 4 + }, + { + "epoch": 0.0002631578947368421, + "grad_norm": 3.84375, + "learning_rate": 5.263157894736843e-07, + "loss": 10.948, + "step": 5 + }, + { + "epoch": 0.00031578947368421053, + "grad_norm": 3.8125, + "learning_rate": 6.315789473684211e-07, + "loss": 10.9476, + "step": 6 + }, + { + "epoch": 0.00036842105263157896, + "grad_norm": 3.765625, + "learning_rate": 7.368421052631579e-07, + "loss": 10.9506, + "step": 7 + }, + { + "epoch": 0.0004210526315789474, + "grad_norm": 4.09375, + "learning_rate": 8.421052631578948e-07, + "loss": 10.9611, + "step": 8 + }, + { + "epoch": 0.00047368421052631577, + "grad_norm": 3.703125, + "learning_rate": 9.473684210526317e-07, + "loss": 10.9432, + "step": 9 + }, + { + "epoch": 0.0005263157894736842, + "grad_norm": 3.578125, + "learning_rate": 1.0526315789473685e-06, + "loss": 10.957, + "step": 10 + }, + { + "epoch": 0.0005789473684210527, + "grad_norm": 3.5625, + "learning_rate": 1.1578947368421055e-06, + "loss": 10.9557, + "step": 11 + }, + { + "epoch": 0.0006315789473684211, + "grad_norm": 4.03125, + "learning_rate": 1.2631578947368422e-06, + "loss": 10.9425, + "step": 12 + }, + { + "epoch": 0.0006842105263157894, + "grad_norm": 3.828125, + "learning_rate": 1.3684210526315791e-06, + "loss": 10.9753, + "step": 13 + }, + { + "epoch": 0.0007368421052631579, + "grad_norm": 3.609375, + "learning_rate": 1.4736842105263159e-06, + "loss": 10.9607, + "step": 14 + }, + { + "epoch": 0.0007894736842105263, + "grad_norm": 4.09375, + "learning_rate": 1.5789473684210528e-06, + "loss": 10.9471, + "step": 15 + }, + { + "epoch": 0.0008421052631578948, + "grad_norm": 4.1875, + "learning_rate": 1.6842105263157895e-06, + "loss": 10.9466, + "step": 16 + }, + { + "epoch": 0.0008947368421052632, + "grad_norm": 3.90625, + "learning_rate": 1.7894736842105262e-06, + "loss": 10.9436, + "step": 17 + }, + { + "epoch": 0.0009473684210526315, + "grad_norm": 4.0, + "learning_rate": 1.8947368421052634e-06, + "loss": 10.9496, + "step": 18 + }, + { + "epoch": 0.001, + "grad_norm": 3.859375, + "learning_rate": 2.0000000000000003e-06, + "loss": 10.9522, + "step": 19 + }, + { + "epoch": 0.0010526315789473684, + "grad_norm": 4.5, + "learning_rate": 2.105263157894737e-06, + "loss": 10.9297, + "step": 20 + }, + { + "epoch": 0.0011052631578947368, + "grad_norm": 3.703125, + "learning_rate": 2.2105263157894738e-06, + "loss": 10.9565, + "step": 21 + }, + { + "epoch": 0.0011578947368421054, + "grad_norm": 3.578125, + "learning_rate": 2.315789473684211e-06, + "loss": 10.9493, + "step": 22 + }, + { + "epoch": 0.0012105263157894737, + "grad_norm": 3.65625, + "learning_rate": 2.4210526315789477e-06, + "loss": 10.9445, + "step": 23 + }, + { + "epoch": 0.0012631578947368421, + "grad_norm": 3.40625, + "learning_rate": 2.5263157894736844e-06, + "loss": 10.9413, + "step": 24 + }, + { + "epoch": 0.0013157894736842105, + "grad_norm": 3.703125, + "learning_rate": 2.631578947368421e-06, + "loss": 10.9512, + "step": 25 + }, + { + "epoch": 0.0013684210526315789, + "grad_norm": 3.671875, + "learning_rate": 2.7368421052631583e-06, + "loss": 10.9464, + "step": 26 + }, + { + "epoch": 0.0014210526315789475, + "grad_norm": 3.515625, + "learning_rate": 2.842105263157895e-06, + "loss": 10.952, + "step": 27 + }, + { + "epoch": 0.0014736842105263158, + "grad_norm": 3.734375, + "learning_rate": 2.9473684210526317e-06, + "loss": 10.9405, + "step": 28 + }, + { + "epoch": 0.0015263157894736842, + "grad_norm": 3.90625, + "learning_rate": 3.0526315789473684e-06, + "loss": 10.9396, + "step": 29 + }, + { + "epoch": 0.0015789473684210526, + "grad_norm": 3.453125, + "learning_rate": 3.1578947368421056e-06, + "loss": 10.9331, + "step": 30 + }, + { + "epoch": 0.001631578947368421, + "grad_norm": 3.6875, + "learning_rate": 3.2631578947368423e-06, + "loss": 10.9312, + "step": 31 + }, + { + "epoch": 0.0016842105263157896, + "grad_norm": 3.796875, + "learning_rate": 3.368421052631579e-06, + "loss": 10.9255, + "step": 32 + }, + { + "epoch": 0.001736842105263158, + "grad_norm": 3.515625, + "learning_rate": 3.4736842105263158e-06, + "loss": 10.9277, + "step": 33 + }, + { + "epoch": 0.0017894736842105263, + "grad_norm": 4.0625, + "learning_rate": 3.5789473684210525e-06, + "loss": 10.9309, + "step": 34 + }, + { + "epoch": 0.0018421052631578947, + "grad_norm": 3.875, + "learning_rate": 3.6842105263157892e-06, + "loss": 10.9176, + "step": 35 + }, + { + "epoch": 0.001894736842105263, + "grad_norm": 3.828125, + "learning_rate": 3.789473684210527e-06, + "loss": 10.9115, + "step": 36 + }, + { + "epoch": 0.0019473684210526317, + "grad_norm": 3.703125, + "learning_rate": 3.894736842105264e-06, + "loss": 10.9073, + "step": 37 + }, + { + "epoch": 0.002, + "grad_norm": 3.609375, + "learning_rate": 4.000000000000001e-06, + "loss": 10.8947, + "step": 38 + }, + { + "epoch": 0.0020526315789473684, + "grad_norm": 3.71875, + "learning_rate": 4.105263157894737e-06, + "loss": 10.8998, + "step": 39 + }, + { + "epoch": 0.002105263157894737, + "grad_norm": 4.84375, + "learning_rate": 4.210526315789474e-06, + "loss": 10.9084, + "step": 40 + }, + { + "epoch": 0.002157894736842105, + "grad_norm": 4.1875, + "learning_rate": 4.315789473684211e-06, + "loss": 10.8748, + "step": 41 + }, + { + "epoch": 0.0022105263157894735, + "grad_norm": 3.59375, + "learning_rate": 4.4210526315789476e-06, + "loss": 10.8866, + "step": 42 + }, + { + "epoch": 0.002263157894736842, + "grad_norm": 3.921875, + "learning_rate": 4.526315789473685e-06, + "loss": 10.8705, + "step": 43 + }, + { + "epoch": 0.0023157894736842107, + "grad_norm": 3.53125, + "learning_rate": 4.631578947368422e-06, + "loss": 10.8864, + "step": 44 + }, + { + "epoch": 0.002368421052631579, + "grad_norm": 3.828125, + "learning_rate": 4.736842105263159e-06, + "loss": 10.875, + "step": 45 + }, + { + "epoch": 0.0024210526315789475, + "grad_norm": 3.84375, + "learning_rate": 4.842105263157895e-06, + "loss": 10.8621, + "step": 46 + }, + { + "epoch": 0.002473684210526316, + "grad_norm": 3.890625, + "learning_rate": 4.947368421052632e-06, + "loss": 10.8477, + "step": 47 + }, + { + "epoch": 0.0025263157894736842, + "grad_norm": 3.8125, + "learning_rate": 5.052631578947369e-06, + "loss": 10.8569, + "step": 48 + }, + { + "epoch": 0.0025789473684210526, + "grad_norm": 3.609375, + "learning_rate": 5.1578947368421055e-06, + "loss": 10.8585, + "step": 49 + }, + { + "epoch": 0.002631578947368421, + "grad_norm": 3.75, + "learning_rate": 5.263157894736842e-06, + "loss": 10.8619, + "step": 50 + }, + { + "epoch": 0.0026842105263157894, + "grad_norm": 4.28125, + "learning_rate": 5.36842105263158e-06, + "loss": 10.8261, + "step": 51 + }, + { + "epoch": 0.0027368421052631577, + "grad_norm": 3.390625, + "learning_rate": 5.4736842105263165e-06, + "loss": 10.8544, + "step": 52 + }, + { + "epoch": 0.002789473684210526, + "grad_norm": 3.609375, + "learning_rate": 5.578947368421053e-06, + "loss": 10.8149, + "step": 53 + }, + { + "epoch": 0.002842105263157895, + "grad_norm": 3.859375, + "learning_rate": 5.68421052631579e-06, + "loss": 10.8103, + "step": 54 + }, + { + "epoch": 0.0028947368421052633, + "grad_norm": 3.46875, + "learning_rate": 5.789473684210527e-06, + "loss": 10.8291, + "step": 55 + }, + { + "epoch": 0.0029473684210526317, + "grad_norm": 3.765625, + "learning_rate": 5.8947368421052634e-06, + "loss": 10.7848, + "step": 56 + }, + { + "epoch": 0.003, + "grad_norm": 3.6875, + "learning_rate": 6e-06, + "loss": 10.8035, + "step": 57 + }, + { + "epoch": 0.0030526315789473684, + "grad_norm": 3.609375, + "learning_rate": 6.105263157894737e-06, + "loss": 10.7851, + "step": 58 + }, + { + "epoch": 0.003105263157894737, + "grad_norm": 4.15625, + "learning_rate": 6.2105263157894745e-06, + "loss": 10.743, + "step": 59 + }, + { + "epoch": 0.003157894736842105, + "grad_norm": 3.890625, + "learning_rate": 6.315789473684211e-06, + "loss": 10.7588, + "step": 60 + }, + { + "epoch": 0.0032105263157894736, + "grad_norm": 4.4375, + "learning_rate": 6.421052631578947e-06, + "loss": 10.7042, + "step": 61 + }, + { + "epoch": 0.003263157894736842, + "grad_norm": 3.6875, + "learning_rate": 6.526315789473685e-06, + "loss": 10.7161, + "step": 62 + }, + { + "epoch": 0.0033157894736842103, + "grad_norm": 3.671875, + "learning_rate": 6.631578947368422e-06, + "loss": 10.7265, + "step": 63 + }, + { + "epoch": 0.003368421052631579, + "grad_norm": 3.8125, + "learning_rate": 6.736842105263158e-06, + "loss": 10.7043, + "step": 64 + }, + { + "epoch": 0.0034210526315789475, + "grad_norm": 3.5625, + "learning_rate": 6.842105263157896e-06, + "loss": 10.7256, + "step": 65 + }, + { + "epoch": 0.003473684210526316, + "grad_norm": 3.609375, + "learning_rate": 6.9473684210526315e-06, + "loss": 10.6732, + "step": 66 + }, + { + "epoch": 0.0035263157894736843, + "grad_norm": 3.71875, + "learning_rate": 7.052631578947369e-06, + "loss": 10.6819, + "step": 67 + }, + { + "epoch": 0.0035789473684210526, + "grad_norm": 3.484375, + "learning_rate": 7.157894736842105e-06, + "loss": 10.664, + "step": 68 + }, + { + "epoch": 0.003631578947368421, + "grad_norm": 3.40625, + "learning_rate": 7.2631578947368426e-06, + "loss": 10.6646, + "step": 69 + }, + { + "epoch": 0.0036842105263157894, + "grad_norm": 3.515625, + "learning_rate": 7.3684210526315784e-06, + "loss": 10.6618, + "step": 70 + }, + { + "epoch": 0.0037368421052631578, + "grad_norm": 3.609375, + "learning_rate": 7.473684210526316e-06, + "loss": 10.6203, + "step": 71 + }, + { + "epoch": 0.003789473684210526, + "grad_norm": 3.546875, + "learning_rate": 7.578947368421054e-06, + "loss": 10.6025, + "step": 72 + }, + { + "epoch": 0.0038421052631578945, + "grad_norm": 3.75, + "learning_rate": 7.68421052631579e-06, + "loss": 10.5616, + "step": 73 + }, + { + "epoch": 0.0038947368421052633, + "grad_norm": 3.359375, + "learning_rate": 7.789473684210528e-06, + "loss": 10.6024, + "step": 74 + }, + { + "epoch": 0.003947368421052632, + "grad_norm": 3.484375, + "learning_rate": 7.894736842105263e-06, + "loss": 10.5519, + "step": 75 + }, + { + "epoch": 0.004, + "grad_norm": 3.40625, + "learning_rate": 8.000000000000001e-06, + "loss": 10.546, + "step": 76 + }, + { + "epoch": 0.0040526315789473685, + "grad_norm": 3.25, + "learning_rate": 8.105263157894736e-06, + "loss": 10.557, + "step": 77 + }, + { + "epoch": 0.004105263157894737, + "grad_norm": 3.484375, + "learning_rate": 8.210526315789475e-06, + "loss": 10.4883, + "step": 78 + }, + { + "epoch": 0.004157894736842105, + "grad_norm": 3.21875, + "learning_rate": 8.315789473684212e-06, + "loss": 10.5348, + "step": 79 + }, + { + "epoch": 0.004210526315789474, + "grad_norm": 3.375, + "learning_rate": 8.421052631578948e-06, + "loss": 10.4652, + "step": 80 + }, + { + "epoch": 0.004263157894736842, + "grad_norm": 3.359375, + "learning_rate": 8.526315789473685e-06, + "loss": 10.4416, + "step": 81 + }, + { + "epoch": 0.00431578947368421, + "grad_norm": 3.296875, + "learning_rate": 8.631578947368422e-06, + "loss": 10.4654, + "step": 82 + }, + { + "epoch": 0.004368421052631579, + "grad_norm": 3.203125, + "learning_rate": 8.736842105263158e-06, + "loss": 10.4965, + "step": 83 + }, + { + "epoch": 0.004421052631578947, + "grad_norm": 3.140625, + "learning_rate": 8.842105263157895e-06, + "loss": 10.4513, + "step": 84 + }, + { + "epoch": 0.0044736842105263155, + "grad_norm": 3.171875, + "learning_rate": 8.947368421052632e-06, + "loss": 10.4256, + "step": 85 + }, + { + "epoch": 0.004526315789473684, + "grad_norm": 3.328125, + "learning_rate": 9.05263157894737e-06, + "loss": 10.3631, + "step": 86 + }, + { + "epoch": 0.004578947368421052, + "grad_norm": 2.875, + "learning_rate": 9.157894736842105e-06, + "loss": 10.3985, + "step": 87 + }, + { + "epoch": 0.0046315789473684215, + "grad_norm": 3.078125, + "learning_rate": 9.263157894736844e-06, + "loss": 10.4076, + "step": 88 + }, + { + "epoch": 0.00468421052631579, + "grad_norm": 3.15625, + "learning_rate": 9.368421052631579e-06, + "loss": 10.3709, + "step": 89 + }, + { + "epoch": 0.004736842105263158, + "grad_norm": 3.25, + "learning_rate": 9.473684210526317e-06, + "loss": 10.2645, + "step": 90 + }, + { + "epoch": 0.004789473684210527, + "grad_norm": 2.671875, + "learning_rate": 9.578947368421052e-06, + "loss": 10.3525, + "step": 91 + }, + { + "epoch": 0.004842105263157895, + "grad_norm": 3.015625, + "learning_rate": 9.68421052631579e-06, + "loss": 10.2717, + "step": 92 + }, + { + "epoch": 0.004894736842105263, + "grad_norm": 2.84375, + "learning_rate": 9.789473684210526e-06, + "loss": 10.2835, + "step": 93 + }, + { + "epoch": 0.004947368421052632, + "grad_norm": 2.546875, + "learning_rate": 9.894736842105264e-06, + "loss": 10.3126, + "step": 94 + }, + { + "epoch": 0.005, + "grad_norm": 2.75, + "learning_rate": 1e-05, + "loss": 10.2252, + "step": 95 + }, + { + "epoch": 0.0050526315789473685, + "grad_norm": 2.75, + "learning_rate": 1.0105263157894738e-05, + "loss": 10.159, + "step": 96 + }, + { + "epoch": 0.005105263157894737, + "grad_norm": 2.671875, + "learning_rate": 1.0210526315789474e-05, + "loss": 10.2256, + "step": 97 + }, + { + "epoch": 0.005157894736842105, + "grad_norm": 2.65625, + "learning_rate": 1.0315789473684211e-05, + "loss": 10.1371, + "step": 98 + }, + { + "epoch": 0.005210526315789474, + "grad_norm": 2.734375, + "learning_rate": 1.0421052631578948e-05, + "loss": 10.0691, + "step": 99 + }, + { + "epoch": 0.005263157894736842, + "grad_norm": 2.484375, + "learning_rate": 1.0526315789473684e-05, + "loss": 10.2183, + "step": 100 + }, + { + "epoch": 0.00531578947368421, + "grad_norm": 2.640625, + "learning_rate": 1.0631578947368421e-05, + "loss": 10.1065, + "step": 101 + }, + { + "epoch": 0.005368421052631579, + "grad_norm": 2.75, + "learning_rate": 1.073684210526316e-05, + "loss": 10.1383, + "step": 102 + }, + { + "epoch": 0.005421052631578947, + "grad_norm": 2.78125, + "learning_rate": 1.0842105263157895e-05, + "loss": 10.135, + "step": 103 + }, + { + "epoch": 0.0054736842105263155, + "grad_norm": 2.421875, + "learning_rate": 1.0947368421052633e-05, + "loss": 10.0825, + "step": 104 + }, + { + "epoch": 0.005526315789473684, + "grad_norm": 2.421875, + "learning_rate": 1.1052631578947368e-05, + "loss": 10.0529, + "step": 105 + }, + { + "epoch": 0.005578947368421052, + "grad_norm": 2.859375, + "learning_rate": 1.1157894736842106e-05, + "loss": 9.9912, + "step": 106 + }, + { + "epoch": 0.005631578947368421, + "grad_norm": 2.421875, + "learning_rate": 1.1263157894736842e-05, + "loss": 9.9509, + "step": 107 + }, + { + "epoch": 0.00568421052631579, + "grad_norm": 2.296875, + "learning_rate": 1.136842105263158e-05, + "loss": 10.0711, + "step": 108 + }, + { + "epoch": 0.005736842105263158, + "grad_norm": 2.609375, + "learning_rate": 1.1473684210526315e-05, + "loss": 9.9596, + "step": 109 + }, + { + "epoch": 0.005789473684210527, + "grad_norm": 2.34375, + "learning_rate": 1.1578947368421053e-05, + "loss": 9.9152, + "step": 110 + }, + { + "epoch": 0.005842105263157895, + "grad_norm": 2.21875, + "learning_rate": 1.168421052631579e-05, + "loss": 9.9988, + "step": 111 + }, + { + "epoch": 0.005894736842105263, + "grad_norm": 2.171875, + "learning_rate": 1.1789473684210527e-05, + "loss": 9.9675, + "step": 112 + }, + { + "epoch": 0.005947368421052632, + "grad_norm": 1.9921875, + "learning_rate": 1.1894736842105264e-05, + "loss": 10.0311, + "step": 113 + }, + { + "epoch": 0.006, + "grad_norm": 2.40625, + "learning_rate": 1.2e-05, + "loss": 9.9158, + "step": 114 + }, + { + "epoch": 0.0060526315789473685, + "grad_norm": 2.40625, + "learning_rate": 1.2105263157894737e-05, + "loss": 9.9815, + "step": 115 + }, + { + "epoch": 0.006105263157894737, + "grad_norm": 2.328125, + "learning_rate": 1.2210526315789474e-05, + "loss": 9.8998, + "step": 116 + }, + { + "epoch": 0.006157894736842105, + "grad_norm": 2.21875, + "learning_rate": 1.231578947368421e-05, + "loss": 9.855, + "step": 117 + }, + { + "epoch": 0.006210526315789474, + "grad_norm": 2.375, + "learning_rate": 1.2421052631578949e-05, + "loss": 9.822, + "step": 118 + }, + { + "epoch": 0.006263157894736842, + "grad_norm": 1.9453125, + "learning_rate": 1.2526315789473686e-05, + "loss": 9.9183, + "step": 119 + }, + { + "epoch": 0.00631578947368421, + "grad_norm": 2.109375, + "learning_rate": 1.2631578947368422e-05, + "loss": 9.812, + "step": 120 + }, + { + "epoch": 0.006368421052631579, + "grad_norm": 2.125, + "learning_rate": 1.2736842105263157e-05, + "loss": 9.8318, + "step": 121 + }, + { + "epoch": 0.006421052631578947, + "grad_norm": 2.0, + "learning_rate": 1.2842105263157894e-05, + "loss": 9.8988, + "step": 122 + }, + { + "epoch": 0.0064736842105263155, + "grad_norm": 2.109375, + "learning_rate": 1.2947368421052633e-05, + "loss": 9.7989, + "step": 123 + }, + { + "epoch": 0.006526315789473684, + "grad_norm": 2.125, + "learning_rate": 1.305263157894737e-05, + "loss": 9.76, + "step": 124 + }, + { + "epoch": 0.006578947368421052, + "grad_norm": 2.09375, + "learning_rate": 1.3157894736842106e-05, + "loss": 9.7773, + "step": 125 + }, + { + "epoch": 0.006631578947368421, + "grad_norm": 1.921875, + "learning_rate": 1.3263157894736844e-05, + "loss": 9.84, + "step": 126 + }, + { + "epoch": 0.00668421052631579, + "grad_norm": 2.171875, + "learning_rate": 1.336842105263158e-05, + "loss": 9.8104, + "step": 127 + }, + { + "epoch": 0.006736842105263158, + "grad_norm": 1.90625, + "learning_rate": 1.3473684210526316e-05, + "loss": 9.8066, + "step": 128 + }, + { + "epoch": 0.006789473684210527, + "grad_norm": 1.890625, + "learning_rate": 1.3578947368421053e-05, + "loss": 9.8544, + "step": 129 + }, + { + "epoch": 0.006842105263157895, + "grad_norm": 1.9453125, + "learning_rate": 1.3684210526315791e-05, + "loss": 9.7536, + "step": 130 + }, + { + "epoch": 0.006894736842105263, + "grad_norm": 1.859375, + "learning_rate": 1.3789473684210526e-05, + "loss": 9.8066, + "step": 131 + }, + { + "epoch": 0.006947368421052632, + "grad_norm": 1.8515625, + "learning_rate": 1.3894736842105263e-05, + "loss": 9.7379, + "step": 132 + }, + { + "epoch": 0.007, + "grad_norm": 2.03125, + "learning_rate": 1.4000000000000001e-05, + "loss": 9.7313, + "step": 133 + }, + { + "epoch": 0.0070526315789473685, + "grad_norm": 2.453125, + "learning_rate": 1.4105263157894738e-05, + "loss": 9.6413, + "step": 134 + }, + { + "epoch": 0.007105263157894737, + "grad_norm": 2.03125, + "learning_rate": 1.4210526315789475e-05, + "loss": 9.8383, + "step": 135 + }, + { + "epoch": 0.007157894736842105, + "grad_norm": 1.8828125, + "learning_rate": 1.431578947368421e-05, + "loss": 9.7651, + "step": 136 + }, + { + "epoch": 0.007210526315789474, + "grad_norm": 2.078125, + "learning_rate": 1.4421052631578948e-05, + "loss": 9.6467, + "step": 137 + }, + { + "epoch": 0.007263157894736842, + "grad_norm": 1.9453125, + "learning_rate": 1.4526315789473685e-05, + "loss": 9.7545, + "step": 138 + }, + { + "epoch": 0.00731578947368421, + "grad_norm": 1.7734375, + "learning_rate": 1.4631578947368422e-05, + "loss": 9.8037, + "step": 139 + }, + { + "epoch": 0.007368421052631579, + "grad_norm": 2.265625, + "learning_rate": 1.4736842105263157e-05, + "loss": 9.612, + "step": 140 + }, + { + "epoch": 0.007421052631578947, + "grad_norm": 1.90625, + "learning_rate": 1.4842105263157895e-05, + "loss": 9.7289, + "step": 141 + }, + { + "epoch": 0.0074736842105263155, + "grad_norm": 1.9609375, + "learning_rate": 1.4947368421052632e-05, + "loss": 9.598, + "step": 142 + }, + { + "epoch": 0.007526315789473684, + "grad_norm": 2.28125, + "learning_rate": 1.5052631578947369e-05, + "loss": 9.8352, + "step": 143 + }, + { + "epoch": 0.007578947368421052, + "grad_norm": 1.828125, + "learning_rate": 1.5157894736842107e-05, + "loss": 9.734, + "step": 144 + }, + { + "epoch": 0.007631578947368421, + "grad_norm": 1.921875, + "learning_rate": 1.5263157894736842e-05, + "loss": 9.6372, + "step": 145 + }, + { + "epoch": 0.007684210526315789, + "grad_norm": 2.0, + "learning_rate": 1.536842105263158e-05, + "loss": 9.6481, + "step": 146 + }, + { + "epoch": 0.007736842105263158, + "grad_norm": 1.8984375, + "learning_rate": 1.5473684210526316e-05, + "loss": 9.7081, + "step": 147 + }, + { + "epoch": 0.007789473684210527, + "grad_norm": 1.9453125, + "learning_rate": 1.5578947368421056e-05, + "loss": 9.5962, + "step": 148 + }, + { + "epoch": 0.007842105263157895, + "grad_norm": 1.78125, + "learning_rate": 1.568421052631579e-05, + "loss": 9.7023, + "step": 149 + }, + { + "epoch": 0.007894736842105263, + "grad_norm": 2.09375, + "learning_rate": 1.5789473684210526e-05, + "loss": 9.5248, + "step": 150 + }, + { + "epoch": 0.007947368421052632, + "grad_norm": 1.9765625, + "learning_rate": 1.5894736842105266e-05, + "loss": 9.5968, + "step": 151 + }, + { + "epoch": 0.008, + "grad_norm": 1.9609375, + "learning_rate": 1.6000000000000003e-05, + "loss": 9.6829, + "step": 152 + }, + { + "epoch": 0.008052631578947369, + "grad_norm": 2.015625, + "learning_rate": 1.6105263157894736e-05, + "loss": 9.5834, + "step": 153 + }, + { + "epoch": 0.008105263157894737, + "grad_norm": 2.25, + "learning_rate": 1.6210526315789473e-05, + "loss": 9.4623, + "step": 154 + }, + { + "epoch": 0.008157894736842105, + "grad_norm": 1.90625, + "learning_rate": 1.6315789473684213e-05, + "loss": 9.5766, + "step": 155 + }, + { + "epoch": 0.008210526315789474, + "grad_norm": 1.8671875, + "learning_rate": 1.642105263157895e-05, + "loss": 9.5321, + "step": 156 + }, + { + "epoch": 0.008263157894736842, + "grad_norm": 1.90625, + "learning_rate": 1.6526315789473683e-05, + "loss": 9.505, + "step": 157 + }, + { + "epoch": 0.00831578947368421, + "grad_norm": 2.0625, + "learning_rate": 1.6631578947368423e-05, + "loss": 9.5979, + "step": 158 + }, + { + "epoch": 0.008368421052631579, + "grad_norm": 2.046875, + "learning_rate": 1.673684210526316e-05, + "loss": 9.6172, + "step": 159 + }, + { + "epoch": 0.008421052631578947, + "grad_norm": 2.71875, + "learning_rate": 1.6842105263157896e-05, + "loss": 9.6383, + "step": 160 + }, + { + "epoch": 0.008473684210526316, + "grad_norm": 2.171875, + "learning_rate": 1.694736842105263e-05, + "loss": 9.4054, + "step": 161 + }, + { + "epoch": 0.008526315789473684, + "grad_norm": 1.8984375, + "learning_rate": 1.705263157894737e-05, + "loss": 9.5013, + "step": 162 + }, + { + "epoch": 0.008578947368421052, + "grad_norm": 1.9140625, + "learning_rate": 1.7157894736842107e-05, + "loss": 9.4677, + "step": 163 + }, + { + "epoch": 0.00863157894736842, + "grad_norm": 1.78125, + "learning_rate": 1.7263157894736843e-05, + "loss": 9.5768, + "step": 164 + }, + { + "epoch": 0.008684210526315789, + "grad_norm": 1.71875, + "learning_rate": 1.736842105263158e-05, + "loss": 9.5702, + "step": 165 + }, + { + "epoch": 0.008736842105263157, + "grad_norm": 1.7578125, + "learning_rate": 1.7473684210526317e-05, + "loss": 9.5454, + "step": 166 + }, + { + "epoch": 0.008789473684210526, + "grad_norm": 1.8515625, + "learning_rate": 1.7578947368421054e-05, + "loss": 9.5374, + "step": 167 + }, + { + "epoch": 0.008842105263157894, + "grad_norm": 1.8046875, + "learning_rate": 1.768421052631579e-05, + "loss": 9.509, + "step": 168 + }, + { + "epoch": 0.008894736842105263, + "grad_norm": 1.6953125, + "learning_rate": 1.7789473684210527e-05, + "loss": 9.6149, + "step": 169 + }, + { + "epoch": 0.008947368421052631, + "grad_norm": 1.8046875, + "learning_rate": 1.7894736842105264e-05, + "loss": 9.4539, + "step": 170 + }, + { + "epoch": 0.009, + "grad_norm": 2.03125, + "learning_rate": 1.8e-05, + "loss": 9.4991, + "step": 171 + }, + { + "epoch": 0.009052631578947368, + "grad_norm": 2.0, + "learning_rate": 1.810526315789474e-05, + "loss": 9.5237, + "step": 172 + }, + { + "epoch": 0.009105263157894736, + "grad_norm": 2.015625, + "learning_rate": 1.8210526315789474e-05, + "loss": 9.2873, + "step": 173 + }, + { + "epoch": 0.009157894736842104, + "grad_norm": 1.6953125, + "learning_rate": 1.831578947368421e-05, + "loss": 9.5567, + "step": 174 + }, + { + "epoch": 0.009210526315789473, + "grad_norm": 1.9375, + "learning_rate": 1.8421052631578947e-05, + "loss": 9.5649, + "step": 175 + }, + { + "epoch": 0.009263157894736843, + "grad_norm": 1.828125, + "learning_rate": 1.8526315789473687e-05, + "loss": 9.5634, + "step": 176 + }, + { + "epoch": 0.009315789473684211, + "grad_norm": 2.015625, + "learning_rate": 1.863157894736842e-05, + "loss": 9.5254, + "step": 177 + }, + { + "epoch": 0.00936842105263158, + "grad_norm": 1.8828125, + "learning_rate": 1.8736842105263158e-05, + "loss": 9.4633, + "step": 178 + }, + { + "epoch": 0.009421052631578948, + "grad_norm": 2.171875, + "learning_rate": 1.8842105263157894e-05, + "loss": 9.1744, + "step": 179 + }, + { + "epoch": 0.009473684210526316, + "grad_norm": 1.9296875, + "learning_rate": 1.8947368421052634e-05, + "loss": 9.4562, + "step": 180 + }, + { + "epoch": 0.009526315789473685, + "grad_norm": 1.8828125, + "learning_rate": 1.9052631578947368e-05, + "loss": 9.4247, + "step": 181 + }, + { + "epoch": 0.009578947368421053, + "grad_norm": 2.125, + "learning_rate": 1.9157894736842104e-05, + "loss": 9.4516, + "step": 182 + }, + { + "epoch": 0.009631578947368422, + "grad_norm": 2.09375, + "learning_rate": 1.9263157894736845e-05, + "loss": 9.3927, + "step": 183 + }, + { + "epoch": 0.00968421052631579, + "grad_norm": 1.9921875, + "learning_rate": 1.936842105263158e-05, + "loss": 9.3369, + "step": 184 + }, + { + "epoch": 0.009736842105263158, + "grad_norm": 1.921875, + "learning_rate": 1.9473684210526315e-05, + "loss": 9.2977, + "step": 185 + }, + { + "epoch": 0.009789473684210527, + "grad_norm": 2.078125, + "learning_rate": 1.957894736842105e-05, + "loss": 9.3196, + "step": 186 + }, + { + "epoch": 0.009842105263157895, + "grad_norm": 1.859375, + "learning_rate": 1.968421052631579e-05, + "loss": 9.3424, + "step": 187 + }, + { + "epoch": 0.009894736842105263, + "grad_norm": 2.28125, + "learning_rate": 1.9789473684210528e-05, + "loss": 9.3121, + "step": 188 + }, + { + "epoch": 0.009947368421052632, + "grad_norm": 1.8984375, + "learning_rate": 1.9894736842105265e-05, + "loss": 9.2662, + "step": 189 + }, + { + "epoch": 0.01, + "grad_norm": 2.296875, + "learning_rate": 2e-05, + "loss": 9.2235, + "step": 190 + }, + { + "epoch": 0.010052631578947369, + "grad_norm": 1.765625, + "learning_rate": 2.010526315789474e-05, + "loss": 9.3722, + "step": 191 + }, + { + "epoch": 0.010105263157894737, + "grad_norm": 2.3125, + "learning_rate": 2.0210526315789475e-05, + "loss": 9.1696, + "step": 192 + }, + { + "epoch": 0.010157894736842105, + "grad_norm": 1.734375, + "learning_rate": 2.0315789473684212e-05, + "loss": 9.4174, + "step": 193 + }, + { + "epoch": 0.010210526315789474, + "grad_norm": 3.296875, + "learning_rate": 2.042105263157895e-05, + "loss": 9.5384, + "step": 194 + }, + { + "epoch": 0.010263157894736842, + "grad_norm": 1.5625, + "learning_rate": 2.0526315789473685e-05, + "loss": 9.4179, + "step": 195 + }, + { + "epoch": 0.01031578947368421, + "grad_norm": 1.90625, + "learning_rate": 2.0631578947368422e-05, + "loss": 9.3857, + "step": 196 + }, + { + "epoch": 0.010368421052631579, + "grad_norm": 1.90625, + "learning_rate": 2.073684210526316e-05, + "loss": 9.3737, + "step": 197 + }, + { + "epoch": 0.010421052631578947, + "grad_norm": 1.9921875, + "learning_rate": 2.0842105263157895e-05, + "loss": 9.2431, + "step": 198 + }, + { + "epoch": 0.010473684210526316, + "grad_norm": 1.8515625, + "learning_rate": 2.0947368421052632e-05, + "loss": 9.2809, + "step": 199 + }, + { + "epoch": 0.010526315789473684, + "grad_norm": 1.8828125, + "learning_rate": 2.105263157894737e-05, + "loss": 9.194, + "step": 200 + }, + { + "epoch": 0.010578947368421052, + "grad_norm": 1.9453125, + "learning_rate": 2.1157894736842106e-05, + "loss": 9.3211, + "step": 201 + }, + { + "epoch": 0.01063157894736842, + "grad_norm": 1.9453125, + "learning_rate": 2.1263157894736842e-05, + "loss": 9.1787, + "step": 202 + }, + { + "epoch": 0.010684210526315789, + "grad_norm": 1.8359375, + "learning_rate": 2.136842105263158e-05, + "loss": 9.3412, + "step": 203 + }, + { + "epoch": 0.010736842105263157, + "grad_norm": 1.875, + "learning_rate": 2.147368421052632e-05, + "loss": 9.1615, + "step": 204 + }, + { + "epoch": 0.010789473684210526, + "grad_norm": 2.0, + "learning_rate": 2.1578947368421053e-05, + "loss": 9.1574, + "step": 205 + }, + { + "epoch": 0.010842105263157894, + "grad_norm": 1.7734375, + "learning_rate": 2.168421052631579e-05, + "loss": 9.405, + "step": 206 + }, + { + "epoch": 0.010894736842105263, + "grad_norm": 1.6328125, + "learning_rate": 2.1789473684210526e-05, + "loss": 9.3019, + "step": 207 + }, + { + "epoch": 0.010947368421052631, + "grad_norm": 1.5625, + "learning_rate": 2.1894736842105266e-05, + "loss": 9.2394, + "step": 208 + }, + { + "epoch": 0.011, + "grad_norm": 1.71875, + "learning_rate": 2.2000000000000003e-05, + "loss": 9.1933, + "step": 209 + }, + { + "epoch": 0.011052631578947368, + "grad_norm": 1.828125, + "learning_rate": 2.2105263157894736e-05, + "loss": 9.1869, + "step": 210 + }, + { + "epoch": 0.011105263157894736, + "grad_norm": 2.046875, + "learning_rate": 2.2210526315789476e-05, + "loss": 9.3699, + "step": 211 + }, + { + "epoch": 0.011157894736842104, + "grad_norm": 1.828125, + "learning_rate": 2.2315789473684213e-05, + "loss": 9.0818, + "step": 212 + }, + { + "epoch": 0.011210526315789473, + "grad_norm": 1.7421875, + "learning_rate": 2.242105263157895e-05, + "loss": 9.2358, + "step": 213 + }, + { + "epoch": 0.011263157894736841, + "grad_norm": 1.8125, + "learning_rate": 2.2526315789473683e-05, + "loss": 9.1661, + "step": 214 + }, + { + "epoch": 0.011315789473684211, + "grad_norm": 1.8984375, + "learning_rate": 2.2631578947368423e-05, + "loss": 9.1109, + "step": 215 + }, + { + "epoch": 0.01136842105263158, + "grad_norm": 1.8359375, + "learning_rate": 2.273684210526316e-05, + "loss": 8.9919, + "step": 216 + }, + { + "epoch": 0.011421052631578948, + "grad_norm": 1.7109375, + "learning_rate": 2.2842105263157897e-05, + "loss": 9.2591, + "step": 217 + }, + { + "epoch": 0.011473684210526316, + "grad_norm": 1.8203125, + "learning_rate": 2.294736842105263e-05, + "loss": 9.1485, + "step": 218 + }, + { + "epoch": 0.011526315789473685, + "grad_norm": 1.7734375, + "learning_rate": 2.305263157894737e-05, + "loss": 9.1586, + "step": 219 + }, + { + "epoch": 0.011578947368421053, + "grad_norm": 1.8515625, + "learning_rate": 2.3157894736842107e-05, + "loss": 9.1737, + "step": 220 + }, + { + "epoch": 0.011631578947368422, + "grad_norm": 2.109375, + "learning_rate": 2.3263157894736844e-05, + "loss": 9.12, + "step": 221 + }, + { + "epoch": 0.01168421052631579, + "grad_norm": 2.046875, + "learning_rate": 2.336842105263158e-05, + "loss": 9.0963, + "step": 222 + }, + { + "epoch": 0.011736842105263158, + "grad_norm": 1.9921875, + "learning_rate": 2.3473684210526317e-05, + "loss": 9.0366, + "step": 223 + }, + { + "epoch": 0.011789473684210527, + "grad_norm": 1.9921875, + "learning_rate": 2.3578947368421054e-05, + "loss": 9.2498, + "step": 224 + }, + { + "epoch": 0.011842105263157895, + "grad_norm": 1.9296875, + "learning_rate": 2.368421052631579e-05, + "loss": 9.1162, + "step": 225 + }, + { + "epoch": 0.011894736842105263, + "grad_norm": 1.921875, + "learning_rate": 2.3789473684210527e-05, + "loss": 9.0164, + "step": 226 + }, + { + "epoch": 0.011947368421052632, + "grad_norm": 2.25, + "learning_rate": 2.3894736842105264e-05, + "loss": 9.0108, + "step": 227 + }, + { + "epoch": 0.012, + "grad_norm": 1.625, + "learning_rate": 2.4e-05, + "loss": 9.0273, + "step": 228 + }, + { + "epoch": 0.012052631578947369, + "grad_norm": 1.5625, + "learning_rate": 2.410526315789474e-05, + "loss": 9.1133, + "step": 229 + }, + { + "epoch": 0.012105263157894737, + "grad_norm": 1.6328125, + "learning_rate": 2.4210526315789474e-05, + "loss": 9.0468, + "step": 230 + }, + { + "epoch": 0.012157894736842105, + "grad_norm": 1.59375, + "learning_rate": 2.431578947368421e-05, + "loss": 9.1075, + "step": 231 + }, + { + "epoch": 0.012210526315789474, + "grad_norm": 1.8515625, + "learning_rate": 2.4421052631578948e-05, + "loss": 9.1198, + "step": 232 + }, + { + "epoch": 0.012263157894736842, + "grad_norm": 2.046875, + "learning_rate": 2.4526315789473688e-05, + "loss": 9.0776, + "step": 233 + }, + { + "epoch": 0.01231578947368421, + "grad_norm": 2.1875, + "learning_rate": 2.463157894736842e-05, + "loss": 9.0448, + "step": 234 + }, + { + "epoch": 0.012368421052631579, + "grad_norm": 1.96875, + "learning_rate": 2.4736842105263158e-05, + "loss": 8.9932, + "step": 235 + }, + { + "epoch": 0.012421052631578947, + "grad_norm": 1.875, + "learning_rate": 2.4842105263157898e-05, + "loss": 9.0264, + "step": 236 + }, + { + "epoch": 0.012473684210526316, + "grad_norm": 2.046875, + "learning_rate": 2.4947368421052635e-05, + "loss": 8.9558, + "step": 237 + }, + { + "epoch": 0.012526315789473684, + "grad_norm": 2.140625, + "learning_rate": 2.505263157894737e-05, + "loss": 9.004, + "step": 238 + }, + { + "epoch": 0.012578947368421052, + "grad_norm": 2.375, + "learning_rate": 2.5157894736842108e-05, + "loss": 8.9485, + "step": 239 + }, + { + "epoch": 0.01263157894736842, + "grad_norm": 2.140625, + "learning_rate": 2.5263157894736845e-05, + "loss": 9.0417, + "step": 240 + }, + { + "epoch": 0.01268421052631579, + "grad_norm": 1.71875, + "learning_rate": 2.536842105263158e-05, + "loss": 8.9977, + "step": 241 + }, + { + "epoch": 0.012736842105263158, + "grad_norm": 1.640625, + "learning_rate": 2.5473684210526315e-05, + "loss": 8.9721, + "step": 242 + }, + { + "epoch": 0.012789473684210526, + "grad_norm": 1.5390625, + "learning_rate": 2.557894736842105e-05, + "loss": 9.0329, + "step": 243 + }, + { + "epoch": 0.012842105263157894, + "grad_norm": 1.765625, + "learning_rate": 2.5684210526315788e-05, + "loss": 8.9583, + "step": 244 + }, + { + "epoch": 0.012894736842105263, + "grad_norm": 1.8203125, + "learning_rate": 2.578947368421053e-05, + "loss": 8.9553, + "step": 245 + }, + { + "epoch": 0.012947368421052631, + "grad_norm": 1.96875, + "learning_rate": 2.5894736842105265e-05, + "loss": 8.9724, + "step": 246 + }, + { + "epoch": 0.013, + "grad_norm": 2.578125, + "learning_rate": 2.6000000000000002e-05, + "loss": 8.8389, + "step": 247 + }, + { + "epoch": 0.013052631578947368, + "grad_norm": 1.796875, + "learning_rate": 2.610526315789474e-05, + "loss": 8.9266, + "step": 248 + }, + { + "epoch": 0.013105263157894736, + "grad_norm": 1.796875, + "learning_rate": 2.6210526315789475e-05, + "loss": 8.9503, + "step": 249 + }, + { + "epoch": 0.013157894736842105, + "grad_norm": 1.84375, + "learning_rate": 2.6315789473684212e-05, + "loss": 8.9405, + "step": 250 + }, + { + "epoch": 0.013210526315789473, + "grad_norm": 1.78125, + "learning_rate": 2.6421052631578945e-05, + "loss": 9.0214, + "step": 251 + }, + { + "epoch": 0.013263157894736841, + "grad_norm": 1.6328125, + "learning_rate": 2.652631578947369e-05, + "loss": 9.0138, + "step": 252 + }, + { + "epoch": 0.01331578947368421, + "grad_norm": 1.8203125, + "learning_rate": 2.6631578947368426e-05, + "loss": 8.895, + "step": 253 + }, + { + "epoch": 0.01336842105263158, + "grad_norm": 1.578125, + "learning_rate": 2.673684210526316e-05, + "loss": 8.9189, + "step": 254 + }, + { + "epoch": 0.013421052631578948, + "grad_norm": 1.9375, + "learning_rate": 2.6842105263157896e-05, + "loss": 8.8394, + "step": 255 + }, + { + "epoch": 0.013473684210526317, + "grad_norm": 1.625, + "learning_rate": 2.6947368421052632e-05, + "loss": 8.9703, + "step": 256 + }, + { + "epoch": 0.013526315789473685, + "grad_norm": 1.875, + "learning_rate": 2.705263157894737e-05, + "loss": 8.914, + "step": 257 + }, + { + "epoch": 0.013578947368421053, + "grad_norm": 1.78125, + "learning_rate": 2.7157894736842106e-05, + "loss": 8.8097, + "step": 258 + }, + { + "epoch": 0.013631578947368422, + "grad_norm": 1.8671875, + "learning_rate": 2.7263157894736846e-05, + "loss": 8.8401, + "step": 259 + }, + { + "epoch": 0.01368421052631579, + "grad_norm": 1.6484375, + "learning_rate": 2.7368421052631583e-05, + "loss": 8.9527, + "step": 260 + }, + { + "epoch": 0.013736842105263158, + "grad_norm": 1.9140625, + "learning_rate": 2.747368421052632e-05, + "loss": 8.8624, + "step": 261 + }, + { + "epoch": 0.013789473684210527, + "grad_norm": 2.40625, + "learning_rate": 2.7578947368421053e-05, + "loss": 8.8706, + "step": 262 + }, + { + "epoch": 0.013842105263157895, + "grad_norm": 1.875, + "learning_rate": 2.768421052631579e-05, + "loss": 8.896, + "step": 263 + }, + { + "epoch": 0.013894736842105264, + "grad_norm": 1.921875, + "learning_rate": 2.7789473684210526e-05, + "loss": 8.7473, + "step": 264 + }, + { + "epoch": 0.013947368421052632, + "grad_norm": 1.7421875, + "learning_rate": 2.7894736842105263e-05, + "loss": 8.7699, + "step": 265 + }, + { + "epoch": 0.014, + "grad_norm": 1.75, + "learning_rate": 2.8000000000000003e-05, + "loss": 8.7729, + "step": 266 + }, + { + "epoch": 0.014052631578947369, + "grad_norm": 1.65625, + "learning_rate": 2.810526315789474e-05, + "loss": 8.8068, + "step": 267 + }, + { + "epoch": 0.014105263157894737, + "grad_norm": 1.828125, + "learning_rate": 2.8210526315789476e-05, + "loss": 8.8589, + "step": 268 + }, + { + "epoch": 0.014157894736842105, + "grad_norm": 2.015625, + "learning_rate": 2.8315789473684213e-05, + "loss": 8.7636, + "step": 269 + }, + { + "epoch": 0.014210526315789474, + "grad_norm": 2.0, + "learning_rate": 2.842105263157895e-05, + "loss": 8.805, + "step": 270 + }, + { + "epoch": 0.014263157894736842, + "grad_norm": 1.6640625, + "learning_rate": 2.8526315789473683e-05, + "loss": 8.8148, + "step": 271 + }, + { + "epoch": 0.01431578947368421, + "grad_norm": 1.7578125, + "learning_rate": 2.863157894736842e-05, + "loss": 8.7834, + "step": 272 + }, + { + "epoch": 0.014368421052631579, + "grad_norm": 1.703125, + "learning_rate": 2.8736842105263163e-05, + "loss": 8.6983, + "step": 273 + }, + { + "epoch": 0.014421052631578947, + "grad_norm": 1.71875, + "learning_rate": 2.8842105263157897e-05, + "loss": 8.7276, + "step": 274 + }, + { + "epoch": 0.014473684210526316, + "grad_norm": 2.0625, + "learning_rate": 2.8947368421052634e-05, + "loss": 8.8294, + "step": 275 + }, + { + "epoch": 0.014526315789473684, + "grad_norm": 1.84375, + "learning_rate": 2.905263157894737e-05, + "loss": 8.7651, + "step": 276 + }, + { + "epoch": 0.014578947368421052, + "grad_norm": 2.296875, + "learning_rate": 2.9157894736842107e-05, + "loss": 8.702, + "step": 277 + }, + { + "epoch": 0.01463157894736842, + "grad_norm": 1.953125, + "learning_rate": 2.9263157894736844e-05, + "loss": 8.6747, + "step": 278 + }, + { + "epoch": 0.01468421052631579, + "grad_norm": 2.3125, + "learning_rate": 2.9368421052631577e-05, + "loss": 8.5327, + "step": 279 + }, + { + "epoch": 0.014736842105263158, + "grad_norm": 1.703125, + "learning_rate": 2.9473684210526314e-05, + "loss": 8.7754, + "step": 280 + }, + { + "epoch": 0.014789473684210526, + "grad_norm": 1.96875, + "learning_rate": 2.9578947368421057e-05, + "loss": 8.6362, + "step": 281 + }, + { + "epoch": 0.014842105263157894, + "grad_norm": 1.84375, + "learning_rate": 2.968421052631579e-05, + "loss": 8.6239, + "step": 282 + }, + { + "epoch": 0.014894736842105263, + "grad_norm": 1.8203125, + "learning_rate": 2.9789473684210527e-05, + "loss": 8.6244, + "step": 283 + }, + { + "epoch": 0.014947368421052631, + "grad_norm": 1.921875, + "learning_rate": 2.9894736842105264e-05, + "loss": 8.6772, + "step": 284 + }, + { + "epoch": 0.015, + "grad_norm": 1.8359375, + "learning_rate": 3e-05, + "loss": 8.6122, + "step": 285 + }, + { + "epoch": 0.015052631578947368, + "grad_norm": 1.6484375, + "learning_rate": 3.0105263157894737e-05, + "loss": 8.6227, + "step": 286 + }, + { + "epoch": 0.015105263157894736, + "grad_norm": 1.9609375, + "learning_rate": 3.021052631578947e-05, + "loss": 8.527, + "step": 287 + }, + { + "epoch": 0.015157894736842105, + "grad_norm": 1.6484375, + "learning_rate": 3.0315789473684214e-05, + "loss": 8.6305, + "step": 288 + }, + { + "epoch": 0.015210526315789473, + "grad_norm": 2.09375, + "learning_rate": 3.042105263157895e-05, + "loss": 8.6552, + "step": 289 + }, + { + "epoch": 0.015263157894736841, + "grad_norm": 1.71875, + "learning_rate": 3.0526315789473684e-05, + "loss": 8.6602, + "step": 290 + }, + { + "epoch": 0.01531578947368421, + "grad_norm": 1.8515625, + "learning_rate": 3.0631578947368425e-05, + "loss": 8.683, + "step": 291 + }, + { + "epoch": 0.015368421052631578, + "grad_norm": 1.703125, + "learning_rate": 3.073684210526316e-05, + "loss": 8.6074, + "step": 292 + }, + { + "epoch": 0.015421052631578948, + "grad_norm": 1.6171875, + "learning_rate": 3.084210526315789e-05, + "loss": 8.5625, + "step": 293 + }, + { + "epoch": 0.015473684210526317, + "grad_norm": 1.7421875, + "learning_rate": 3.094736842105263e-05, + "loss": 8.5424, + "step": 294 + }, + { + "epoch": 0.015526315789473685, + "grad_norm": 1.53125, + "learning_rate": 3.105263157894737e-05, + "loss": 8.668, + "step": 295 + }, + { + "epoch": 0.015578947368421053, + "grad_norm": 2.078125, + "learning_rate": 3.115789473684211e-05, + "loss": 8.3626, + "step": 296 + }, + { + "epoch": 0.01563157894736842, + "grad_norm": 1.7109375, + "learning_rate": 3.1263157894736845e-05, + "loss": 8.6605, + "step": 297 + }, + { + "epoch": 0.01568421052631579, + "grad_norm": 1.8671875, + "learning_rate": 3.136842105263158e-05, + "loss": 8.6174, + "step": 298 + }, + { + "epoch": 0.015736842105263157, + "grad_norm": 2.015625, + "learning_rate": 3.147368421052632e-05, + "loss": 8.374, + "step": 299 + }, + { + "epoch": 0.015789473684210527, + "grad_norm": 1.53125, + "learning_rate": 3.157894736842105e-05, + "loss": 8.6408, + "step": 300 + }, + { + "epoch": 0.015842105263157893, + "grad_norm": 1.6328125, + "learning_rate": 3.168421052631579e-05, + "loss": 8.352, + "step": 301 + }, + { + "epoch": 0.015894736842105264, + "grad_norm": 1.59375, + "learning_rate": 3.178947368421053e-05, + "loss": 8.5314, + "step": 302 + }, + { + "epoch": 0.01594736842105263, + "grad_norm": 1.6953125, + "learning_rate": 3.1894736842105265e-05, + "loss": 8.5566, + "step": 303 + }, + { + "epoch": 0.016, + "grad_norm": 1.9140625, + "learning_rate": 3.2000000000000005e-05, + "loss": 8.3328, + "step": 304 + }, + { + "epoch": 0.016052631578947367, + "grad_norm": 1.6484375, + "learning_rate": 3.210526315789474e-05, + "loss": 8.3112, + "step": 305 + }, + { + "epoch": 0.016105263157894737, + "grad_norm": 1.859375, + "learning_rate": 3.221052631578947e-05, + "loss": 8.3805, + "step": 306 + }, + { + "epoch": 0.016157894736842104, + "grad_norm": 2.09375, + "learning_rate": 3.231578947368421e-05, + "loss": 8.4017, + "step": 307 + }, + { + "epoch": 0.016210526315789474, + "grad_norm": 1.8984375, + "learning_rate": 3.2421052631578945e-05, + "loss": 8.4074, + "step": 308 + }, + { + "epoch": 0.01626315789473684, + "grad_norm": 1.515625, + "learning_rate": 3.2526315789473686e-05, + "loss": 8.4459, + "step": 309 + }, + { + "epoch": 0.01631578947368421, + "grad_norm": 1.609375, + "learning_rate": 3.2631578947368426e-05, + "loss": 8.5615, + "step": 310 + }, + { + "epoch": 0.016368421052631577, + "grad_norm": 1.78125, + "learning_rate": 3.273684210526316e-05, + "loss": 8.3835, + "step": 311 + }, + { + "epoch": 0.016421052631578947, + "grad_norm": 1.796875, + "learning_rate": 3.28421052631579e-05, + "loss": 8.2471, + "step": 312 + }, + { + "epoch": 0.016473684210526317, + "grad_norm": 1.5234375, + "learning_rate": 3.294736842105263e-05, + "loss": 8.2518, + "step": 313 + }, + { + "epoch": 0.016526315789473684, + "grad_norm": 1.609375, + "learning_rate": 3.3052631578947366e-05, + "loss": 8.3793, + "step": 314 + }, + { + "epoch": 0.016578947368421054, + "grad_norm": 2.234375, + "learning_rate": 3.3157894736842106e-05, + "loss": 8.1762, + "step": 315 + }, + { + "epoch": 0.01663157894736842, + "grad_norm": 1.65625, + "learning_rate": 3.3263157894736846e-05, + "loss": 8.4103, + "step": 316 + }, + { + "epoch": 0.01668421052631579, + "grad_norm": 1.71875, + "learning_rate": 3.336842105263158e-05, + "loss": 8.4715, + "step": 317 + }, + { + "epoch": 0.016736842105263158, + "grad_norm": 1.6484375, + "learning_rate": 3.347368421052632e-05, + "loss": 8.4896, + "step": 318 + }, + { + "epoch": 0.016789473684210528, + "grad_norm": 1.9609375, + "learning_rate": 3.357894736842105e-05, + "loss": 8.3562, + "step": 319 + }, + { + "epoch": 0.016842105263157894, + "grad_norm": 1.8359375, + "learning_rate": 3.368421052631579e-05, + "loss": 8.4297, + "step": 320 + }, + { + "epoch": 0.016894736842105264, + "grad_norm": 1.96875, + "learning_rate": 3.3789473684210526e-05, + "loss": 8.1998, + "step": 321 + }, + { + "epoch": 0.01694736842105263, + "grad_norm": 1.4140625, + "learning_rate": 3.389473684210526e-05, + "loss": 8.2607, + "step": 322 + }, + { + "epoch": 0.017, + "grad_norm": 1.71875, + "learning_rate": 3.4000000000000007e-05, + "loss": 8.4456, + "step": 323 + }, + { + "epoch": 0.017052631578947368, + "grad_norm": 1.4765625, + "learning_rate": 3.410526315789474e-05, + "loss": 8.3814, + "step": 324 + }, + { + "epoch": 0.017105263157894738, + "grad_norm": 1.421875, + "learning_rate": 3.421052631578947e-05, + "loss": 8.343, + "step": 325 + }, + { + "epoch": 0.017157894736842105, + "grad_norm": 1.578125, + "learning_rate": 3.431578947368421e-05, + "loss": 8.2263, + "step": 326 + }, + { + "epoch": 0.017210526315789475, + "grad_norm": 1.78125, + "learning_rate": 3.442105263157895e-05, + "loss": 8.257, + "step": 327 + }, + { + "epoch": 0.01726315789473684, + "grad_norm": 1.7265625, + "learning_rate": 3.452631578947369e-05, + "loss": 8.1041, + "step": 328 + }, + { + "epoch": 0.01731578947368421, + "grad_norm": 1.7890625, + "learning_rate": 3.463157894736842e-05, + "loss": 8.1556, + "step": 329 + }, + { + "epoch": 0.017368421052631578, + "grad_norm": 1.7734375, + "learning_rate": 3.473684210526316e-05, + "loss": 8.1013, + "step": 330 + }, + { + "epoch": 0.017421052631578948, + "grad_norm": 1.8203125, + "learning_rate": 3.48421052631579e-05, + "loss": 8.156, + "step": 331 + }, + { + "epoch": 0.017473684210526315, + "grad_norm": 1.8515625, + "learning_rate": 3.4947368421052634e-05, + "loss": 8.1753, + "step": 332 + }, + { + "epoch": 0.017526315789473685, + "grad_norm": 2.1875, + "learning_rate": 3.505263157894737e-05, + "loss": 8.3376, + "step": 333 + }, + { + "epoch": 0.01757894736842105, + "grad_norm": 1.5078125, + "learning_rate": 3.515789473684211e-05, + "loss": 8.1885, + "step": 334 + }, + { + "epoch": 0.017631578947368422, + "grad_norm": 1.6953125, + "learning_rate": 3.526315789473684e-05, + "loss": 8.2551, + "step": 335 + }, + { + "epoch": 0.01768421052631579, + "grad_norm": 1.8046875, + "learning_rate": 3.536842105263158e-05, + "loss": 8.0564, + "step": 336 + }, + { + "epoch": 0.01773684210526316, + "grad_norm": 1.5078125, + "learning_rate": 3.547368421052632e-05, + "loss": 8.1111, + "step": 337 + }, + { + "epoch": 0.017789473684210525, + "grad_norm": 1.6015625, + "learning_rate": 3.5578947368421054e-05, + "loss": 8.1987, + "step": 338 + }, + { + "epoch": 0.017842105263157895, + "grad_norm": 2.21875, + "learning_rate": 3.5684210526315794e-05, + "loss": 8.0521, + "step": 339 + }, + { + "epoch": 0.017894736842105262, + "grad_norm": 2.078125, + "learning_rate": 3.578947368421053e-05, + "loss": 8.1031, + "step": 340 + }, + { + "epoch": 0.017947368421052632, + "grad_norm": 2.015625, + "learning_rate": 3.589473684210527e-05, + "loss": 8.0399, + "step": 341 + }, + { + "epoch": 0.018, + "grad_norm": 1.90625, + "learning_rate": 3.6e-05, + "loss": 8.086, + "step": 342 + }, + { + "epoch": 0.01805263157894737, + "grad_norm": 1.984375, + "learning_rate": 3.6105263157894734e-05, + "loss": 8.2486, + "step": 343 + }, + { + "epoch": 0.018105263157894735, + "grad_norm": 1.8125, + "learning_rate": 3.621052631578948e-05, + "loss": 8.0492, + "step": 344 + }, + { + "epoch": 0.018157894736842106, + "grad_norm": 1.859375, + "learning_rate": 3.6315789473684214e-05, + "loss": 8.1344, + "step": 345 + }, + { + "epoch": 0.018210526315789472, + "grad_norm": 2.0, + "learning_rate": 3.642105263157895e-05, + "loss": 8.0447, + "step": 346 + }, + { + "epoch": 0.018263157894736842, + "grad_norm": 1.90625, + "learning_rate": 3.652631578947369e-05, + "loss": 8.0886, + "step": 347 + }, + { + "epoch": 0.01831578947368421, + "grad_norm": 1.53125, + "learning_rate": 3.663157894736842e-05, + "loss": 8.0791, + "step": 348 + }, + { + "epoch": 0.01836842105263158, + "grad_norm": 1.453125, + "learning_rate": 3.673684210526316e-05, + "loss": 8.0758, + "step": 349 + }, + { + "epoch": 0.018421052631578946, + "grad_norm": 1.4765625, + "learning_rate": 3.6842105263157895e-05, + "loss": 7.971, + "step": 350 + }, + { + "epoch": 0.018473684210526316, + "grad_norm": 1.7578125, + "learning_rate": 3.6947368421052635e-05, + "loss": 8.0364, + "step": 351 + }, + { + "epoch": 0.018526315789473686, + "grad_norm": 1.515625, + "learning_rate": 3.7052631578947375e-05, + "loss": 8.0249, + "step": 352 + }, + { + "epoch": 0.018578947368421053, + "grad_norm": 1.40625, + "learning_rate": 3.715789473684211e-05, + "loss": 8.2435, + "step": 353 + }, + { + "epoch": 0.018631578947368423, + "grad_norm": 1.7734375, + "learning_rate": 3.726315789473684e-05, + "loss": 7.9899, + "step": 354 + }, + { + "epoch": 0.01868421052631579, + "grad_norm": 1.6171875, + "learning_rate": 3.736842105263158e-05, + "loss": 7.8852, + "step": 355 + }, + { + "epoch": 0.01873684210526316, + "grad_norm": 2.0625, + "learning_rate": 3.7473684210526315e-05, + "loss": 7.8308, + "step": 356 + }, + { + "epoch": 0.018789473684210526, + "grad_norm": 1.3984375, + "learning_rate": 3.7578947368421055e-05, + "loss": 8.1836, + "step": 357 + }, + { + "epoch": 0.018842105263157896, + "grad_norm": 1.9609375, + "learning_rate": 3.768421052631579e-05, + "loss": 7.9387, + "step": 358 + }, + { + "epoch": 0.018894736842105263, + "grad_norm": 1.5, + "learning_rate": 3.778947368421053e-05, + "loss": 7.9595, + "step": 359 + }, + { + "epoch": 0.018947368421052633, + "grad_norm": 2.1875, + "learning_rate": 3.789473684210527e-05, + "loss": 7.9133, + "step": 360 + }, + { + "epoch": 0.019, + "grad_norm": 1.4921875, + "learning_rate": 3.8e-05, + "loss": 7.9375, + "step": 361 + }, + { + "epoch": 0.01905263157894737, + "grad_norm": 1.453125, + "learning_rate": 3.8105263157894735e-05, + "loss": 7.9921, + "step": 362 + }, + { + "epoch": 0.019105263157894736, + "grad_norm": 1.765625, + "learning_rate": 3.8210526315789476e-05, + "loss": 7.9147, + "step": 363 + }, + { + "epoch": 0.019157894736842106, + "grad_norm": 1.7421875, + "learning_rate": 3.831578947368421e-05, + "loss": 7.9122, + "step": 364 + }, + { + "epoch": 0.019210526315789473, + "grad_norm": 2.046875, + "learning_rate": 3.842105263157895e-05, + "loss": 7.7756, + "step": 365 + }, + { + "epoch": 0.019263157894736843, + "grad_norm": 1.8203125, + "learning_rate": 3.852631578947369e-05, + "loss": 7.8192, + "step": 366 + }, + { + "epoch": 0.01931578947368421, + "grad_norm": 1.5859375, + "learning_rate": 3.863157894736842e-05, + "loss": 8.0632, + "step": 367 + }, + { + "epoch": 0.01936842105263158, + "grad_norm": 1.4921875, + "learning_rate": 3.873684210526316e-05, + "loss": 7.7421, + "step": 368 + }, + { + "epoch": 0.019421052631578947, + "grad_norm": 1.3671875, + "learning_rate": 3.8842105263157896e-05, + "loss": 7.9746, + "step": 369 + }, + { + "epoch": 0.019473684210526317, + "grad_norm": 1.4765625, + "learning_rate": 3.894736842105263e-05, + "loss": 7.8386, + "step": 370 + }, + { + "epoch": 0.019526315789473683, + "grad_norm": 1.3671875, + "learning_rate": 3.905263157894737e-05, + "loss": 7.8791, + "step": 371 + }, + { + "epoch": 0.019578947368421053, + "grad_norm": 1.578125, + "learning_rate": 3.91578947368421e-05, + "loss": 7.7693, + "step": 372 + }, + { + "epoch": 0.01963157894736842, + "grad_norm": 1.5078125, + "learning_rate": 3.926315789473684e-05, + "loss": 7.7841, + "step": 373 + }, + { + "epoch": 0.01968421052631579, + "grad_norm": 1.4921875, + "learning_rate": 3.936842105263158e-05, + "loss": 7.5373, + "step": 374 + }, + { + "epoch": 0.019736842105263157, + "grad_norm": 1.421875, + "learning_rate": 3.9473684210526316e-05, + "loss": 8.0117, + "step": 375 + }, + { + "epoch": 0.019789473684210527, + "grad_norm": 1.703125, + "learning_rate": 3.9578947368421056e-05, + "loss": 7.7446, + "step": 376 + }, + { + "epoch": 0.019842105263157894, + "grad_norm": 2.34375, + "learning_rate": 3.968421052631579e-05, + "loss": 7.7758, + "step": 377 + }, + { + "epoch": 0.019894736842105264, + "grad_norm": 1.4296875, + "learning_rate": 3.978947368421053e-05, + "loss": 7.7345, + "step": 378 + }, + { + "epoch": 0.01994736842105263, + "grad_norm": 1.7421875, + "learning_rate": 3.989473684210526e-05, + "loss": 7.7532, + "step": 379 + }, + { + "epoch": 0.02, + "grad_norm": 1.71875, + "learning_rate": 4e-05, + "loss": 7.9441, + "step": 380 + }, + { + "epoch": 0.020052631578947367, + "grad_norm": 1.671875, + "learning_rate": 4.010526315789474e-05, + "loss": 7.7693, + "step": 381 + }, + { + "epoch": 0.020105263157894737, + "grad_norm": 1.625, + "learning_rate": 4.021052631578948e-05, + "loss": 7.7832, + "step": 382 + }, + { + "epoch": 0.020157894736842104, + "grad_norm": 1.6953125, + "learning_rate": 4.031578947368421e-05, + "loss": 7.86, + "step": 383 + }, + { + "epoch": 0.020210526315789474, + "grad_norm": 3.09375, + "learning_rate": 4.042105263157895e-05, + "loss": 7.5599, + "step": 384 + }, + { + "epoch": 0.02026315789473684, + "grad_norm": 1.421875, + "learning_rate": 4.0526315789473684e-05, + "loss": 7.6648, + "step": 385 + }, + { + "epoch": 0.02031578947368421, + "grad_norm": 1.5703125, + "learning_rate": 4.0631578947368424e-05, + "loss": 7.7642, + "step": 386 + }, + { + "epoch": 0.020368421052631577, + "grad_norm": 1.609375, + "learning_rate": 4.0736842105263164e-05, + "loss": 7.796, + "step": 387 + }, + { + "epoch": 0.020421052631578947, + "grad_norm": 1.4453125, + "learning_rate": 4.08421052631579e-05, + "loss": 7.6522, + "step": 388 + }, + { + "epoch": 0.020473684210526314, + "grad_norm": 1.6484375, + "learning_rate": 4.094736842105264e-05, + "loss": 7.9119, + "step": 389 + }, + { + "epoch": 0.020526315789473684, + "grad_norm": 1.59375, + "learning_rate": 4.105263157894737e-05, + "loss": 7.8309, + "step": 390 + }, + { + "epoch": 0.020578947368421054, + "grad_norm": 1.2578125, + "learning_rate": 4.1157894736842104e-05, + "loss": 7.7205, + "step": 391 + }, + { + "epoch": 0.02063157894736842, + "grad_norm": 1.3359375, + "learning_rate": 4.1263157894736844e-05, + "loss": 7.737, + "step": 392 + }, + { + "epoch": 0.02068421052631579, + "grad_norm": 1.59375, + "learning_rate": 4.136842105263158e-05, + "loss": 7.7448, + "step": 393 + }, + { + "epoch": 0.020736842105263158, + "grad_norm": 1.3203125, + "learning_rate": 4.147368421052632e-05, + "loss": 7.7803, + "step": 394 + }, + { + "epoch": 0.020789473684210528, + "grad_norm": 1.765625, + "learning_rate": 4.157894736842106e-05, + "loss": 7.7504, + "step": 395 + }, + { + "epoch": 0.020842105263157894, + "grad_norm": 1.796875, + "learning_rate": 4.168421052631579e-05, + "loss": 7.5649, + "step": 396 + }, + { + "epoch": 0.020894736842105265, + "grad_norm": 1.5, + "learning_rate": 4.178947368421053e-05, + "loss": 7.5771, + "step": 397 + }, + { + "epoch": 0.02094736842105263, + "grad_norm": 1.5, + "learning_rate": 4.1894736842105264e-05, + "loss": 7.6211, + "step": 398 + }, + { + "epoch": 0.021, + "grad_norm": 1.390625, + "learning_rate": 4.2e-05, + "loss": 7.393, + "step": 399 + }, + { + "epoch": 0.021052631578947368, + "grad_norm": 1.4765625, + "learning_rate": 4.210526315789474e-05, + "loss": 7.5764, + "step": 400 + }, + { + "epoch": 0.021105263157894738, + "grad_norm": 1.265625, + "learning_rate": 4.221052631578948e-05, + "loss": 7.6691, + "step": 401 + }, + { + "epoch": 0.021157894736842105, + "grad_norm": 1.9453125, + "learning_rate": 4.231578947368421e-05, + "loss": 7.424, + "step": 402 + }, + { + "epoch": 0.021210526315789475, + "grad_norm": 1.2265625, + "learning_rate": 4.242105263157895e-05, + "loss": 7.6571, + "step": 403 + }, + { + "epoch": 0.02126315789473684, + "grad_norm": 1.640625, + "learning_rate": 4.2526315789473685e-05, + "loss": 7.5283, + "step": 404 + }, + { + "epoch": 0.02131578947368421, + "grad_norm": 1.2578125, + "learning_rate": 4.2631578947368425e-05, + "loss": 7.6041, + "step": 405 + }, + { + "epoch": 0.021368421052631578, + "grad_norm": 1.3203125, + "learning_rate": 4.273684210526316e-05, + "loss": 7.7277, + "step": 406 + }, + { + "epoch": 0.02142105263157895, + "grad_norm": 1.390625, + "learning_rate": 4.284210526315789e-05, + "loss": 7.7196, + "step": 407 + }, + { + "epoch": 0.021473684210526315, + "grad_norm": 1.609375, + "learning_rate": 4.294736842105264e-05, + "loss": 7.3864, + "step": 408 + }, + { + "epoch": 0.021526315789473685, + "grad_norm": 1.3125, + "learning_rate": 4.305263157894737e-05, + "loss": 7.5906, + "step": 409 + }, + { + "epoch": 0.02157894736842105, + "grad_norm": 1.6953125, + "learning_rate": 4.3157894736842105e-05, + "loss": 7.5789, + "step": 410 + }, + { + "epoch": 0.021631578947368422, + "grad_norm": 1.265625, + "learning_rate": 4.3263157894736845e-05, + "loss": 7.522, + "step": 411 + }, + { + "epoch": 0.02168421052631579, + "grad_norm": 1.515625, + "learning_rate": 4.336842105263158e-05, + "loss": 7.493, + "step": 412 + }, + { + "epoch": 0.02173684210526316, + "grad_norm": 1.4140625, + "learning_rate": 4.347368421052632e-05, + "loss": 7.601, + "step": 413 + }, + { + "epoch": 0.021789473684210525, + "grad_norm": 1.1953125, + "learning_rate": 4.357894736842105e-05, + "loss": 7.5779, + "step": 414 + }, + { + "epoch": 0.021842105263157895, + "grad_norm": 1.40625, + "learning_rate": 4.368421052631579e-05, + "loss": 7.4826, + "step": 415 + }, + { + "epoch": 0.021894736842105262, + "grad_norm": 1.40625, + "learning_rate": 4.378947368421053e-05, + "loss": 7.6257, + "step": 416 + }, + { + "epoch": 0.021947368421052632, + "grad_norm": 1.40625, + "learning_rate": 4.3894736842105266e-05, + "loss": 7.4576, + "step": 417 + }, + { + "epoch": 0.022, + "grad_norm": 1.09375, + "learning_rate": 4.4000000000000006e-05, + "loss": 7.6537, + "step": 418 + }, + { + "epoch": 0.02205263157894737, + "grad_norm": 1.4140625, + "learning_rate": 4.410526315789474e-05, + "loss": 7.2708, + "step": 419 + }, + { + "epoch": 0.022105263157894735, + "grad_norm": 1.390625, + "learning_rate": 4.421052631578947e-05, + "loss": 7.3002, + "step": 420 + }, + { + "epoch": 0.022157894736842106, + "grad_norm": 1.625, + "learning_rate": 4.431578947368421e-05, + "loss": 7.5079, + "step": 421 + }, + { + "epoch": 0.022210526315789472, + "grad_norm": 1.7109375, + "learning_rate": 4.442105263157895e-05, + "loss": 7.1948, + "step": 422 + }, + { + "epoch": 0.022263157894736842, + "grad_norm": 1.1796875, + "learning_rate": 4.4526315789473686e-05, + "loss": 7.5357, + "step": 423 + }, + { + "epoch": 0.02231578947368421, + "grad_norm": 1.65625, + "learning_rate": 4.4631578947368426e-05, + "loss": 7.3673, + "step": 424 + }, + { + "epoch": 0.02236842105263158, + "grad_norm": 1.390625, + "learning_rate": 4.473684210526316e-05, + "loss": 7.238, + "step": 425 + }, + { + "epoch": 0.022421052631578946, + "grad_norm": 1.6015625, + "learning_rate": 4.48421052631579e-05, + "loss": 7.3458, + "step": 426 + }, + { + "epoch": 0.022473684210526316, + "grad_norm": 1.25, + "learning_rate": 4.494736842105263e-05, + "loss": 7.3303, + "step": 427 + }, + { + "epoch": 0.022526315789473682, + "grad_norm": 1.5546875, + "learning_rate": 4.5052631578947366e-05, + "loss": 7.2324, + "step": 428 + }, + { + "epoch": 0.022578947368421053, + "grad_norm": 1.4375, + "learning_rate": 4.515789473684211e-05, + "loss": 7.474, + "step": 429 + }, + { + "epoch": 0.022631578947368423, + "grad_norm": 1.1953125, + "learning_rate": 4.5263157894736846e-05, + "loss": 7.3507, + "step": 430 + }, + { + "epoch": 0.02268421052631579, + "grad_norm": 1.4921875, + "learning_rate": 4.536842105263158e-05, + "loss": 7.5034, + "step": 431 + }, + { + "epoch": 0.02273684210526316, + "grad_norm": 1.75, + "learning_rate": 4.547368421052632e-05, + "loss": 7.4296, + "step": 432 + }, + { + "epoch": 0.022789473684210526, + "grad_norm": 1.5234375, + "learning_rate": 4.557894736842105e-05, + "loss": 7.3397, + "step": 433 + }, + { + "epoch": 0.022842105263157896, + "grad_norm": 1.5703125, + "learning_rate": 4.568421052631579e-05, + "loss": 7.4017, + "step": 434 + }, + { + "epoch": 0.022894736842105263, + "grad_norm": 1.3203125, + "learning_rate": 4.5789473684210527e-05, + "loss": 7.5148, + "step": 435 + }, + { + "epoch": 0.022947368421052633, + "grad_norm": 1.8984375, + "learning_rate": 4.589473684210526e-05, + "loss": 7.206, + "step": 436 + }, + { + "epoch": 0.023, + "grad_norm": 1.2734375, + "learning_rate": 4.600000000000001e-05, + "loss": 7.4591, + "step": 437 + }, + { + "epoch": 0.02305263157894737, + "grad_norm": 1.3984375, + "learning_rate": 4.610526315789474e-05, + "loss": 7.3246, + "step": 438 + }, + { + "epoch": 0.023105263157894736, + "grad_norm": 1.2265625, + "learning_rate": 4.6210526315789473e-05, + "loss": 7.3715, + "step": 439 + }, + { + "epoch": 0.023157894736842106, + "grad_norm": 1.5, + "learning_rate": 4.6315789473684214e-05, + "loss": 7.3916, + "step": 440 + }, + { + "epoch": 0.023210526315789473, + "grad_norm": 1.6484375, + "learning_rate": 4.642105263157895e-05, + "loss": 7.2859, + "step": 441 + }, + { + "epoch": 0.023263157894736843, + "grad_norm": 1.390625, + "learning_rate": 4.652631578947369e-05, + "loss": 7.403, + "step": 442 + }, + { + "epoch": 0.02331578947368421, + "grad_norm": 1.3359375, + "learning_rate": 4.663157894736842e-05, + "loss": 7.2941, + "step": 443 + }, + { + "epoch": 0.02336842105263158, + "grad_norm": 1.3984375, + "learning_rate": 4.673684210526316e-05, + "loss": 7.4918, + "step": 444 + }, + { + "epoch": 0.023421052631578947, + "grad_norm": 1.453125, + "learning_rate": 4.68421052631579e-05, + "loss": 7.1477, + "step": 445 + }, + { + "epoch": 0.023473684210526317, + "grad_norm": 1.546875, + "learning_rate": 4.6947368421052634e-05, + "loss": 7.4694, + "step": 446 + }, + { + "epoch": 0.023526315789473683, + "grad_norm": 1.265625, + "learning_rate": 4.705263157894737e-05, + "loss": 7.1525, + "step": 447 + }, + { + "epoch": 0.023578947368421053, + "grad_norm": 1.40625, + "learning_rate": 4.715789473684211e-05, + "loss": 7.1526, + "step": 448 + }, + { + "epoch": 0.02363157894736842, + "grad_norm": 2.140625, + "learning_rate": 4.726315789473684e-05, + "loss": 7.5369, + "step": 449 + }, + { + "epoch": 0.02368421052631579, + "grad_norm": 1.2109375, + "learning_rate": 4.736842105263158e-05, + "loss": 7.2863, + "step": 450 + }, + { + "epoch": 0.023736842105263157, + "grad_norm": 1.234375, + "learning_rate": 4.747368421052632e-05, + "loss": 7.339, + "step": 451 + }, + { + "epoch": 0.023789473684210527, + "grad_norm": 1.078125, + "learning_rate": 4.7578947368421054e-05, + "loss": 7.1507, + "step": 452 + }, + { + "epoch": 0.023842105263157894, + "grad_norm": 1.9765625, + "learning_rate": 4.7684210526315794e-05, + "loss": 7.1447, + "step": 453 + }, + { + "epoch": 0.023894736842105264, + "grad_norm": 1.359375, + "learning_rate": 4.778947368421053e-05, + "loss": 7.3652, + "step": 454 + }, + { + "epoch": 0.02394736842105263, + "grad_norm": 1.328125, + "learning_rate": 4.789473684210526e-05, + "loss": 7.3232, + "step": 455 + }, + { + "epoch": 0.024, + "grad_norm": 1.359375, + "learning_rate": 4.8e-05, + "loss": 7.2368, + "step": 456 + }, + { + "epoch": 0.024052631578947367, + "grad_norm": 2.015625, + "learning_rate": 4.8105263157894735e-05, + "loss": 7.22, + "step": 457 + }, + { + "epoch": 0.024105263157894737, + "grad_norm": 1.375, + "learning_rate": 4.821052631578948e-05, + "loss": 7.2006, + "step": 458 + }, + { + "epoch": 0.024157894736842104, + "grad_norm": 1.59375, + "learning_rate": 4.8315789473684215e-05, + "loss": 6.8634, + "step": 459 + }, + { + "epoch": 0.024210526315789474, + "grad_norm": 1.4453125, + "learning_rate": 4.842105263157895e-05, + "loss": 7.4333, + "step": 460 + }, + { + "epoch": 0.02426315789473684, + "grad_norm": 1.5078125, + "learning_rate": 4.852631578947369e-05, + "loss": 7.311, + "step": 461 + }, + { + "epoch": 0.02431578947368421, + "grad_norm": 2.03125, + "learning_rate": 4.863157894736842e-05, + "loss": 6.7985, + "step": 462 + }, + { + "epoch": 0.024368421052631577, + "grad_norm": 0.984375, + "learning_rate": 4.873684210526316e-05, + "loss": 7.4754, + "step": 463 + }, + { + "epoch": 0.024421052631578948, + "grad_norm": 1.6171875, + "learning_rate": 4.8842105263157895e-05, + "loss": 7.1957, + "step": 464 + }, + { + "epoch": 0.024473684210526314, + "grad_norm": 1.328125, + "learning_rate": 4.8947368421052635e-05, + "loss": 7.1068, + "step": 465 + }, + { + "epoch": 0.024526315789473684, + "grad_norm": 1.359375, + "learning_rate": 4.9052631578947375e-05, + "loss": 7.201, + "step": 466 + }, + { + "epoch": 0.02457894736842105, + "grad_norm": 1.171875, + "learning_rate": 4.915789473684211e-05, + "loss": 7.3999, + "step": 467 + }, + { + "epoch": 0.02463157894736842, + "grad_norm": 1.4296875, + "learning_rate": 4.926315789473684e-05, + "loss": 7.2408, + "step": 468 + }, + { + "epoch": 0.02468421052631579, + "grad_norm": 1.1328125, + "learning_rate": 4.936842105263158e-05, + "loss": 7.4488, + "step": 469 + }, + { + "epoch": 0.024736842105263158, + "grad_norm": 1.4140625, + "learning_rate": 4.9473684210526315e-05, + "loss": 7.4731, + "step": 470 + }, + { + "epoch": 0.024789473684210528, + "grad_norm": 1.21875, + "learning_rate": 4.9578947368421055e-05, + "loss": 7.0765, + "step": 471 + }, + { + "epoch": 0.024842105263157895, + "grad_norm": 1.4765625, + "learning_rate": 4.9684210526315796e-05, + "loss": 6.7931, + "step": 472 + }, + { + "epoch": 0.024894736842105265, + "grad_norm": 1.3125, + "learning_rate": 4.978947368421053e-05, + "loss": 7.1829, + "step": 473 + }, + { + "epoch": 0.02494736842105263, + "grad_norm": 1.125, + "learning_rate": 4.989473684210527e-05, + "loss": 7.2773, + "step": 474 + }, + { + "epoch": 0.025, + "grad_norm": 1.1484375, + "learning_rate": 5e-05, + "loss": 7.2103, + "step": 475 + }, + { + "epoch": 0.025052631578947368, + "grad_norm": 1.7734375, + "learning_rate": 5.010526315789474e-05, + "loss": 6.9811, + "step": 476 + }, + { + "epoch": 0.025105263157894738, + "grad_norm": 1.1015625, + "learning_rate": 5.0210526315789476e-05, + "loss": 7.1021, + "step": 477 + }, + { + "epoch": 0.025157894736842105, + "grad_norm": 1.4375, + "learning_rate": 5.0315789473684216e-05, + "loss": 7.1445, + "step": 478 + }, + { + "epoch": 0.025210526315789475, + "grad_norm": 1.375, + "learning_rate": 5.042105263157895e-05, + "loss": 7.0559, + "step": 479 + }, + { + "epoch": 0.02526315789473684, + "grad_norm": 1.4765625, + "learning_rate": 5.052631578947369e-05, + "loss": 7.0708, + "step": 480 + }, + { + "epoch": 0.02531578947368421, + "grad_norm": 1.2421875, + "learning_rate": 5.0631578947368416e-05, + "loss": 7.333, + "step": 481 + }, + { + "epoch": 0.02536842105263158, + "grad_norm": 1.6015625, + "learning_rate": 5.073684210526316e-05, + "loss": 7.1137, + "step": 482 + }, + { + "epoch": 0.02542105263157895, + "grad_norm": 1.4296875, + "learning_rate": 5.08421052631579e-05, + "loss": 7.2385, + "step": 483 + }, + { + "epoch": 0.025473684210526315, + "grad_norm": 1.3984375, + "learning_rate": 5.094736842105263e-05, + "loss": 7.0772, + "step": 484 + }, + { + "epoch": 0.025526315789473685, + "grad_norm": 1.0234375, + "learning_rate": 5.1052631578947376e-05, + "loss": 7.2116, + "step": 485 + }, + { + "epoch": 0.025578947368421052, + "grad_norm": 1.3515625, + "learning_rate": 5.11578947368421e-05, + "loss": 6.9401, + "step": 486 + }, + { + "epoch": 0.025631578947368422, + "grad_norm": 1.546875, + "learning_rate": 5.126315789473684e-05, + "loss": 7.1519, + "step": 487 + }, + { + "epoch": 0.02568421052631579, + "grad_norm": 1.0859375, + "learning_rate": 5.1368421052631576e-05, + "loss": 7.0381, + "step": 488 + }, + { + "epoch": 0.02573684210526316, + "grad_norm": 1.71875, + "learning_rate": 5.1473684210526317e-05, + "loss": 6.9751, + "step": 489 + }, + { + "epoch": 0.025789473684210525, + "grad_norm": 1.390625, + "learning_rate": 5.157894736842106e-05, + "loss": 6.8602, + "step": 490 + }, + { + "epoch": 0.025842105263157895, + "grad_norm": 1.265625, + "learning_rate": 5.168421052631579e-05, + "loss": 7.1637, + "step": 491 + }, + { + "epoch": 0.025894736842105262, + "grad_norm": 1.078125, + "learning_rate": 5.178947368421053e-05, + "loss": 6.9194, + "step": 492 + }, + { + "epoch": 0.025947368421052632, + "grad_norm": 1.2578125, + "learning_rate": 5.1894736842105263e-05, + "loss": 7.1639, + "step": 493 + }, + { + "epoch": 0.026, + "grad_norm": 1.03125, + "learning_rate": 5.2000000000000004e-05, + "loss": 7.0881, + "step": 494 + }, + { + "epoch": 0.02605263157894737, + "grad_norm": 1.4921875, + "learning_rate": 5.210526315789474e-05, + "loss": 7.1608, + "step": 495 + }, + { + "epoch": 0.026105263157894736, + "grad_norm": 1.1796875, + "learning_rate": 5.221052631578948e-05, + "loss": 7.0279, + "step": 496 + }, + { + "epoch": 0.026157894736842106, + "grad_norm": 1.2578125, + "learning_rate": 5.231578947368422e-05, + "loss": 7.3157, + "step": 497 + }, + { + "epoch": 0.026210526315789472, + "grad_norm": 1.2265625, + "learning_rate": 5.242105263157895e-05, + "loss": 7.1195, + "step": 498 + }, + { + "epoch": 0.026263157894736842, + "grad_norm": 1.2578125, + "learning_rate": 5.252631578947369e-05, + "loss": 6.9857, + "step": 499 + }, + { + "epoch": 0.02631578947368421, + "grad_norm": 1.203125, + "learning_rate": 5.2631578947368424e-05, + "loss": 6.8336, + "step": 500 + }, + { + "epoch": 0.02636842105263158, + "grad_norm": 1.5078125, + "learning_rate": 5.2736842105263164e-05, + "loss": 6.9307, + "step": 501 + }, + { + "epoch": 0.026421052631578946, + "grad_norm": 1.1640625, + "learning_rate": 5.284210526315789e-05, + "loss": 7.1911, + "step": 502 + }, + { + "epoch": 0.026473684210526316, + "grad_norm": 1.234375, + "learning_rate": 5.294736842105264e-05, + "loss": 7.0338, + "step": 503 + }, + { + "epoch": 0.026526315789473683, + "grad_norm": 1.2421875, + "learning_rate": 5.305263157894738e-05, + "loss": 7.0278, + "step": 504 + }, + { + "epoch": 0.026578947368421053, + "grad_norm": 1.3203125, + "learning_rate": 5.3157894736842104e-05, + "loss": 6.9981, + "step": 505 + }, + { + "epoch": 0.02663157894736842, + "grad_norm": 1.0546875, + "learning_rate": 5.326315789473685e-05, + "loss": 7.2155, + "step": 506 + }, + { + "epoch": 0.02668421052631579, + "grad_norm": 1.1328125, + "learning_rate": 5.336842105263158e-05, + "loss": 6.9476, + "step": 507 + }, + { + "epoch": 0.02673684210526316, + "grad_norm": 1.625, + "learning_rate": 5.347368421052632e-05, + "loss": 6.8187, + "step": 508 + }, + { + "epoch": 0.026789473684210526, + "grad_norm": 1.09375, + "learning_rate": 5.357894736842105e-05, + "loss": 6.9085, + "step": 509 + }, + { + "epoch": 0.026842105263157896, + "grad_norm": 1.34375, + "learning_rate": 5.368421052631579e-05, + "loss": 6.7847, + "step": 510 + }, + { + "epoch": 0.026894736842105263, + "grad_norm": 1.2578125, + "learning_rate": 5.378947368421053e-05, + "loss": 6.9386, + "step": 511 + }, + { + "epoch": 0.026947368421052633, + "grad_norm": 1.40625, + "learning_rate": 5.3894736842105265e-05, + "loss": 6.8521, + "step": 512 + }, + { + "epoch": 0.027, + "grad_norm": 1.0390625, + "learning_rate": 5.4000000000000005e-05, + "loss": 7.0887, + "step": 513 + }, + { + "epoch": 0.02705263157894737, + "grad_norm": 1.609375, + "learning_rate": 5.410526315789474e-05, + "loss": 6.7572, + "step": 514 + }, + { + "epoch": 0.027105263157894736, + "grad_norm": 1.34375, + "learning_rate": 5.421052631578948e-05, + "loss": 7.1054, + "step": 515 + }, + { + "epoch": 0.027157894736842107, + "grad_norm": 1.3359375, + "learning_rate": 5.431578947368421e-05, + "loss": 7.0698, + "step": 516 + }, + { + "epoch": 0.027210526315789473, + "grad_norm": 1.6875, + "learning_rate": 5.442105263157895e-05, + "loss": 6.9722, + "step": 517 + }, + { + "epoch": 0.027263157894736843, + "grad_norm": 0.8984375, + "learning_rate": 5.452631578947369e-05, + "loss": 7.1368, + "step": 518 + }, + { + "epoch": 0.02731578947368421, + "grad_norm": 1.296875, + "learning_rate": 5.4631578947368425e-05, + "loss": 7.0419, + "step": 519 + }, + { + "epoch": 0.02736842105263158, + "grad_norm": 1.109375, + "learning_rate": 5.4736842105263165e-05, + "loss": 7.1524, + "step": 520 + }, + { + "epoch": 0.027421052631578947, + "grad_norm": 1.296875, + "learning_rate": 5.484210526315789e-05, + "loss": 6.6682, + "step": 521 + }, + { + "epoch": 0.027473684210526317, + "grad_norm": 1.0234375, + "learning_rate": 5.494736842105264e-05, + "loss": 7.0072, + "step": 522 + }, + { + "epoch": 0.027526315789473683, + "grad_norm": 1.171875, + "learning_rate": 5.5052631578947365e-05, + "loss": 7.1038, + "step": 523 + }, + { + "epoch": 0.027578947368421054, + "grad_norm": 1.171875, + "learning_rate": 5.5157894736842105e-05, + "loss": 7.0301, + "step": 524 + }, + { + "epoch": 0.02763157894736842, + "grad_norm": 1.2578125, + "learning_rate": 5.526315789473685e-05, + "loss": 7.0223, + "step": 525 + }, + { + "epoch": 0.02768421052631579, + "grad_norm": 1.1328125, + "learning_rate": 5.536842105263158e-05, + "loss": 6.9627, + "step": 526 + }, + { + "epoch": 0.027736842105263157, + "grad_norm": 1.4140625, + "learning_rate": 5.547368421052632e-05, + "loss": 7.3794, + "step": 527 + }, + { + "epoch": 0.027789473684210527, + "grad_norm": 0.921875, + "learning_rate": 5.557894736842105e-05, + "loss": 6.995, + "step": 528 + }, + { + "epoch": 0.027842105263157894, + "grad_norm": 1.0859375, + "learning_rate": 5.568421052631579e-05, + "loss": 6.9893, + "step": 529 + }, + { + "epoch": 0.027894736842105264, + "grad_norm": 1.3828125, + "learning_rate": 5.5789473684210526e-05, + "loss": 7.0297, + "step": 530 + }, + { + "epoch": 0.02794736842105263, + "grad_norm": 1.515625, + "learning_rate": 5.5894736842105266e-05, + "loss": 7.0053, + "step": 531 + }, + { + "epoch": 0.028, + "grad_norm": 1.171875, + "learning_rate": 5.6000000000000006e-05, + "loss": 7.075, + "step": 532 + }, + { + "epoch": 0.028052631578947367, + "grad_norm": 1.3359375, + "learning_rate": 5.610526315789474e-05, + "loss": 6.9526, + "step": 533 + }, + { + "epoch": 0.028105263157894737, + "grad_norm": 0.96875, + "learning_rate": 5.621052631578948e-05, + "loss": 6.939, + "step": 534 + }, + { + "epoch": 0.028157894736842104, + "grad_norm": 1.1875, + "learning_rate": 5.631578947368421e-05, + "loss": 6.839, + "step": 535 + }, + { + "epoch": 0.028210526315789474, + "grad_norm": 1.265625, + "learning_rate": 5.642105263157895e-05, + "loss": 7.0711, + "step": 536 + }, + { + "epoch": 0.02826315789473684, + "grad_norm": 1.2109375, + "learning_rate": 5.652631578947368e-05, + "loss": 6.5658, + "step": 537 + }, + { + "epoch": 0.02831578947368421, + "grad_norm": 1.0859375, + "learning_rate": 5.6631578947368426e-05, + "loss": 6.9596, + "step": 538 + }, + { + "epoch": 0.028368421052631577, + "grad_norm": 1.0546875, + "learning_rate": 5.6736842105263166e-05, + "loss": 6.9004, + "step": 539 + }, + { + "epoch": 0.028421052631578948, + "grad_norm": 1.3359375, + "learning_rate": 5.68421052631579e-05, + "loss": 6.8214, + "step": 540 + }, + { + "epoch": 0.028473684210526314, + "grad_norm": 0.88671875, + "learning_rate": 5.694736842105264e-05, + "loss": 6.9406, + "step": 541 + }, + { + "epoch": 0.028526315789473684, + "grad_norm": 1.2578125, + "learning_rate": 5.7052631578947366e-05, + "loss": 6.8125, + "step": 542 + }, + { + "epoch": 0.02857894736842105, + "grad_norm": 1.7734375, + "learning_rate": 5.715789473684211e-05, + "loss": 6.5691, + "step": 543 + }, + { + "epoch": 0.02863157894736842, + "grad_norm": 1.3515625, + "learning_rate": 5.726315789473684e-05, + "loss": 6.6948, + "step": 544 + }, + { + "epoch": 0.028684210526315788, + "grad_norm": 1.0546875, + "learning_rate": 5.736842105263158e-05, + "loss": 7.0397, + "step": 545 + }, + { + "epoch": 0.028736842105263158, + "grad_norm": 1.1328125, + "learning_rate": 5.747368421052633e-05, + "loss": 7.0625, + "step": 546 + }, + { + "epoch": 0.028789473684210528, + "grad_norm": 0.7890625, + "learning_rate": 5.7578947368421053e-05, + "loss": 7.0955, + "step": 547 + }, + { + "epoch": 0.028842105263157895, + "grad_norm": 1.3828125, + "learning_rate": 5.7684210526315794e-05, + "loss": 6.7998, + "step": 548 + }, + { + "epoch": 0.028894736842105265, + "grad_norm": 1.6171875, + "learning_rate": 5.778947368421053e-05, + "loss": 6.8899, + "step": 549 + }, + { + "epoch": 0.02894736842105263, + "grad_norm": 1.09375, + "learning_rate": 5.789473684210527e-05, + "loss": 6.6558, + "step": 550 + }, + { + "epoch": 0.029, + "grad_norm": 1.0234375, + "learning_rate": 5.8e-05, + "loss": 6.8333, + "step": 551 + }, + { + "epoch": 0.029052631578947368, + "grad_norm": 1.28125, + "learning_rate": 5.810526315789474e-05, + "loss": 6.9781, + "step": 552 + }, + { + "epoch": 0.029105263157894738, + "grad_norm": 1.0234375, + "learning_rate": 5.821052631578948e-05, + "loss": 6.8296, + "step": 553 + }, + { + "epoch": 0.029157894736842105, + "grad_norm": 1.03125, + "learning_rate": 5.8315789473684214e-05, + "loss": 6.9393, + "step": 554 + }, + { + "epoch": 0.029210526315789475, + "grad_norm": 1.3828125, + "learning_rate": 5.8421052631578954e-05, + "loss": 7.187, + "step": 555 + }, + { + "epoch": 0.02926315789473684, + "grad_norm": 0.94921875, + "learning_rate": 5.852631578947369e-05, + "loss": 6.8097, + "step": 556 + }, + { + "epoch": 0.02931578947368421, + "grad_norm": 0.9375, + "learning_rate": 5.863157894736843e-05, + "loss": 6.9098, + "step": 557 + }, + { + "epoch": 0.02936842105263158, + "grad_norm": 1.21875, + "learning_rate": 5.8736842105263154e-05, + "loss": 6.8501, + "step": 558 + }, + { + "epoch": 0.02942105263157895, + "grad_norm": 0.9765625, + "learning_rate": 5.88421052631579e-05, + "loss": 6.9212, + "step": 559 + }, + { + "epoch": 0.029473684210526315, + "grad_norm": 1.3515625, + "learning_rate": 5.894736842105263e-05, + "loss": 6.985, + "step": 560 + }, + { + "epoch": 0.029526315789473685, + "grad_norm": 1.3828125, + "learning_rate": 5.905263157894737e-05, + "loss": 7.0819, + "step": 561 + }, + { + "epoch": 0.029578947368421052, + "grad_norm": 1.484375, + "learning_rate": 5.9157894736842114e-05, + "loss": 6.9483, + "step": 562 + }, + { + "epoch": 0.029631578947368422, + "grad_norm": 1.046875, + "learning_rate": 5.926315789473684e-05, + "loss": 7.1456, + "step": 563 + }, + { + "epoch": 0.02968421052631579, + "grad_norm": 1.2265625, + "learning_rate": 5.936842105263158e-05, + "loss": 6.9142, + "step": 564 + }, + { + "epoch": 0.02973684210526316, + "grad_norm": 1.3671875, + "learning_rate": 5.9473684210526315e-05, + "loss": 6.9206, + "step": 565 + }, + { + "epoch": 0.029789473684210525, + "grad_norm": 1.234375, + "learning_rate": 5.9578947368421055e-05, + "loss": 6.7971, + "step": 566 + }, + { + "epoch": 0.029842105263157895, + "grad_norm": 1.140625, + "learning_rate": 5.968421052631579e-05, + "loss": 6.7485, + "step": 567 + }, + { + "epoch": 0.029894736842105262, + "grad_norm": 1.09375, + "learning_rate": 5.978947368421053e-05, + "loss": 6.83, + "step": 568 + }, + { + "epoch": 0.029947368421052632, + "grad_norm": 1.2421875, + "learning_rate": 5.989473684210527e-05, + "loss": 6.9263, + "step": 569 + }, + { + "epoch": 0.03, + "grad_norm": 1.6015625, + "learning_rate": 6e-05, + "loss": 6.4603, + "step": 570 + }, + { + "epoch": 0.03005263157894737, + "grad_norm": 1.3359375, + "learning_rate": 6.010526315789474e-05, + "loss": 6.757, + "step": 571 + }, + { + "epoch": 0.030105263157894736, + "grad_norm": 1.265625, + "learning_rate": 6.0210526315789475e-05, + "loss": 6.7309, + "step": 572 + }, + { + "epoch": 0.030157894736842106, + "grad_norm": 1.421875, + "learning_rate": 6.0315789473684215e-05, + "loss": 7.0424, + "step": 573 + }, + { + "epoch": 0.030210526315789472, + "grad_norm": 1.6796875, + "learning_rate": 6.042105263157894e-05, + "loss": 6.725, + "step": 574 + }, + { + "epoch": 0.030263157894736843, + "grad_norm": 1.09375, + "learning_rate": 6.052631578947369e-05, + "loss": 6.8007, + "step": 575 + }, + { + "epoch": 0.03031578947368421, + "grad_norm": 1.0625, + "learning_rate": 6.063157894736843e-05, + "loss": 6.9852, + "step": 576 + }, + { + "epoch": 0.03036842105263158, + "grad_norm": 0.8515625, + "learning_rate": 6.073684210526316e-05, + "loss": 6.8625, + "step": 577 + }, + { + "epoch": 0.030421052631578946, + "grad_norm": 1.84375, + "learning_rate": 6.08421052631579e-05, + "loss": 6.9676, + "step": 578 + }, + { + "epoch": 0.030473684210526316, + "grad_norm": 0.84375, + "learning_rate": 6.094736842105263e-05, + "loss": 6.8727, + "step": 579 + }, + { + "epoch": 0.030526315789473683, + "grad_norm": 1.046875, + "learning_rate": 6.105263157894737e-05, + "loss": 6.6164, + "step": 580 + }, + { + "epoch": 0.030578947368421053, + "grad_norm": 1.2890625, + "learning_rate": 6.115789473684211e-05, + "loss": 6.8299, + "step": 581 + }, + { + "epoch": 0.03063157894736842, + "grad_norm": 1.4453125, + "learning_rate": 6.126315789473685e-05, + "loss": 6.7833, + "step": 582 + }, + { + "epoch": 0.03068421052631579, + "grad_norm": 1.2890625, + "learning_rate": 6.136842105263159e-05, + "loss": 6.8275, + "step": 583 + }, + { + "epoch": 0.030736842105263156, + "grad_norm": 1.0, + "learning_rate": 6.147368421052632e-05, + "loss": 6.5518, + "step": 584 + }, + { + "epoch": 0.030789473684210526, + "grad_norm": 1.3046875, + "learning_rate": 6.157894736842106e-05, + "loss": 6.6304, + "step": 585 + }, + { + "epoch": 0.030842105263157896, + "grad_norm": 1.1015625, + "learning_rate": 6.168421052631578e-05, + "loss": 6.7652, + "step": 586 + }, + { + "epoch": 0.030894736842105263, + "grad_norm": 1.46875, + "learning_rate": 6.178947368421054e-05, + "loss": 6.675, + "step": 587 + }, + { + "epoch": 0.030947368421052633, + "grad_norm": 1.3046875, + "learning_rate": 6.189473684210526e-05, + "loss": 6.8262, + "step": 588 + }, + { + "epoch": 0.031, + "grad_norm": 1.2578125, + "learning_rate": 6.2e-05, + "loss": 6.7213, + "step": 589 + }, + { + "epoch": 0.03105263157894737, + "grad_norm": 1.375, + "learning_rate": 6.210526315789474e-05, + "loss": 6.6244, + "step": 590 + }, + { + "epoch": 0.031105263157894737, + "grad_norm": 1.0234375, + "learning_rate": 6.221052631578947e-05, + "loss": 6.8333, + "step": 591 + }, + { + "epoch": 0.031157894736842107, + "grad_norm": 1.140625, + "learning_rate": 6.231578947368422e-05, + "loss": 7.0004, + "step": 592 + }, + { + "epoch": 0.031210526315789473, + "grad_norm": 1.2890625, + "learning_rate": 6.242105263157895e-05, + "loss": 6.8273, + "step": 593 + }, + { + "epoch": 0.03126315789473684, + "grad_norm": 1.671875, + "learning_rate": 6.252631578947369e-05, + "loss": 7.0616, + "step": 594 + }, + { + "epoch": 0.031315789473684214, + "grad_norm": 1.328125, + "learning_rate": 6.263157894736842e-05, + "loss": 6.7137, + "step": 595 + }, + { + "epoch": 0.03136842105263158, + "grad_norm": 1.1640625, + "learning_rate": 6.273684210526316e-05, + "loss": 6.7673, + "step": 596 + }, + { + "epoch": 0.03142105263157895, + "grad_norm": 1.34375, + "learning_rate": 6.28421052631579e-05, + "loss": 6.2884, + "step": 597 + }, + { + "epoch": 0.03147368421052631, + "grad_norm": 0.83984375, + "learning_rate": 6.294736842105264e-05, + "loss": 6.5669, + "step": 598 + }, + { + "epoch": 0.03152631578947369, + "grad_norm": 1.1328125, + "learning_rate": 6.305263157894738e-05, + "loss": 6.5995, + "step": 599 + }, + { + "epoch": 0.031578947368421054, + "grad_norm": 1.21875, + "learning_rate": 6.31578947368421e-05, + "loss": 6.8766, + "step": 600 + }, + { + "epoch": 0.03163157894736842, + "grad_norm": 1.1171875, + "learning_rate": 6.326315789473684e-05, + "loss": 6.9355, + "step": 601 + }, + { + "epoch": 0.03168421052631579, + "grad_norm": 1.34375, + "learning_rate": 6.336842105263158e-05, + "loss": 6.8425, + "step": 602 + }, + { + "epoch": 0.03173684210526316, + "grad_norm": 1.078125, + "learning_rate": 6.347368421052632e-05, + "loss": 6.6703, + "step": 603 + }, + { + "epoch": 0.03178947368421053, + "grad_norm": 1.3984375, + "learning_rate": 6.357894736842106e-05, + "loss": 6.7518, + "step": 604 + }, + { + "epoch": 0.031842105263157894, + "grad_norm": 1.2109375, + "learning_rate": 6.368421052631579e-05, + "loss": 6.6609, + "step": 605 + }, + { + "epoch": 0.03189473684210526, + "grad_norm": 1.078125, + "learning_rate": 6.378947368421053e-05, + "loss": 6.7533, + "step": 606 + }, + { + "epoch": 0.031947368421052634, + "grad_norm": 1.328125, + "learning_rate": 6.389473684210526e-05, + "loss": 6.9184, + "step": 607 + }, + { + "epoch": 0.032, + "grad_norm": 1.1875, + "learning_rate": 6.400000000000001e-05, + "loss": 7.02, + "step": 608 + }, + { + "epoch": 0.03205263157894737, + "grad_norm": 1.0625, + "learning_rate": 6.410526315789474e-05, + "loss": 6.8055, + "step": 609 + }, + { + "epoch": 0.032105263157894734, + "grad_norm": 1.2109375, + "learning_rate": 6.421052631578948e-05, + "loss": 6.8431, + "step": 610 + }, + { + "epoch": 0.03215789473684211, + "grad_norm": 1.140625, + "learning_rate": 6.431578947368422e-05, + "loss": 6.7531, + "step": 611 + }, + { + "epoch": 0.032210526315789474, + "grad_norm": 1.140625, + "learning_rate": 6.442105263157894e-05, + "loss": 6.8617, + "step": 612 + }, + { + "epoch": 0.03226315789473684, + "grad_norm": 1.46875, + "learning_rate": 6.452631578947368e-05, + "loss": 6.9935, + "step": 613 + }, + { + "epoch": 0.03231578947368421, + "grad_norm": 1.296875, + "learning_rate": 6.463157894736842e-05, + "loss": 7.0849, + "step": 614 + }, + { + "epoch": 0.03236842105263158, + "grad_norm": 1.328125, + "learning_rate": 6.473684210526316e-05, + "loss": 6.7118, + "step": 615 + }, + { + "epoch": 0.03242105263157895, + "grad_norm": 1.046875, + "learning_rate": 6.484210526315789e-05, + "loss": 6.7426, + "step": 616 + }, + { + "epoch": 0.032473684210526314, + "grad_norm": 0.984375, + "learning_rate": 6.494736842105263e-05, + "loss": 6.8118, + "step": 617 + }, + { + "epoch": 0.03252631578947368, + "grad_norm": 1.046875, + "learning_rate": 6.505263157894737e-05, + "loss": 6.6473, + "step": 618 + }, + { + "epoch": 0.032578947368421055, + "grad_norm": 1.078125, + "learning_rate": 6.515789473684211e-05, + "loss": 6.8109, + "step": 619 + }, + { + "epoch": 0.03263157894736842, + "grad_norm": 1.2421875, + "learning_rate": 6.526315789473685e-05, + "loss": 6.968, + "step": 620 + }, + { + "epoch": 0.03268421052631579, + "grad_norm": 1.296875, + "learning_rate": 6.536842105263158e-05, + "loss": 6.8707, + "step": 621 + }, + { + "epoch": 0.032736842105263154, + "grad_norm": 1.1796875, + "learning_rate": 6.547368421052632e-05, + "loss": 6.8525, + "step": 622 + }, + { + "epoch": 0.03278947368421053, + "grad_norm": 1.2734375, + "learning_rate": 6.557894736842104e-05, + "loss": 6.8616, + "step": 623 + }, + { + "epoch": 0.032842105263157895, + "grad_norm": 1.59375, + "learning_rate": 6.56842105263158e-05, + "loss": 6.8756, + "step": 624 + }, + { + "epoch": 0.03289473684210526, + "grad_norm": 1.40625, + "learning_rate": 6.578947368421054e-05, + "loss": 6.5405, + "step": 625 + }, + { + "epoch": 0.032947368421052635, + "grad_norm": 0.80859375, + "learning_rate": 6.589473684210526e-05, + "loss": 6.5603, + "step": 626 + }, + { + "epoch": 0.033, + "grad_norm": 1.5546875, + "learning_rate": 6.6e-05, + "loss": 6.6737, + "step": 627 + }, + { + "epoch": 0.03305263157894737, + "grad_norm": 0.9765625, + "learning_rate": 6.610526315789473e-05, + "loss": 6.7987, + "step": 628 + }, + { + "epoch": 0.033105263157894735, + "grad_norm": 1.2421875, + "learning_rate": 6.621052631578947e-05, + "loss": 6.6601, + "step": 629 + }, + { + "epoch": 0.03315789473684211, + "grad_norm": 0.765625, + "learning_rate": 6.631578947368421e-05, + "loss": 6.6248, + "step": 630 + }, + { + "epoch": 0.033210526315789475, + "grad_norm": 1.1640625, + "learning_rate": 6.642105263157895e-05, + "loss": 6.8195, + "step": 631 + }, + { + "epoch": 0.03326315789473684, + "grad_norm": 1.2890625, + "learning_rate": 6.652631578947369e-05, + "loss": 6.4807, + "step": 632 + }, + { + "epoch": 0.03331578947368421, + "grad_norm": 1.09375, + "learning_rate": 6.663157894736842e-05, + "loss": 6.7937, + "step": 633 + }, + { + "epoch": 0.03336842105263158, + "grad_norm": 0.94140625, + "learning_rate": 6.673684210526316e-05, + "loss": 6.593, + "step": 634 + }, + { + "epoch": 0.03342105263157895, + "grad_norm": 1.21875, + "learning_rate": 6.68421052631579e-05, + "loss": 6.7725, + "step": 635 + }, + { + "epoch": 0.033473684210526315, + "grad_norm": 1.140625, + "learning_rate": 6.694736842105264e-05, + "loss": 6.7985, + "step": 636 + }, + { + "epoch": 0.03352631578947368, + "grad_norm": 1.2421875, + "learning_rate": 6.705263157894737e-05, + "loss": 6.7084, + "step": 637 + }, + { + "epoch": 0.033578947368421055, + "grad_norm": 1.0703125, + "learning_rate": 6.71578947368421e-05, + "loss": 6.9715, + "step": 638 + }, + { + "epoch": 0.03363157894736842, + "grad_norm": 1.046875, + "learning_rate": 6.726315789473685e-05, + "loss": 6.5474, + "step": 639 + }, + { + "epoch": 0.03368421052631579, + "grad_norm": 1.125, + "learning_rate": 6.736842105263159e-05, + "loss": 6.7891, + "step": 640 + }, + { + "epoch": 0.033736842105263155, + "grad_norm": 1.1640625, + "learning_rate": 6.747368421052633e-05, + "loss": 6.6465, + "step": 641 + }, + { + "epoch": 0.03378947368421053, + "grad_norm": 1.453125, + "learning_rate": 6.757894736842105e-05, + "loss": 7.0168, + "step": 642 + }, + { + "epoch": 0.033842105263157896, + "grad_norm": 1.265625, + "learning_rate": 6.768421052631579e-05, + "loss": 6.394, + "step": 643 + }, + { + "epoch": 0.03389473684210526, + "grad_norm": 1.3828125, + "learning_rate": 6.778947368421052e-05, + "loss": 6.6288, + "step": 644 + }, + { + "epoch": 0.03394736842105263, + "grad_norm": 0.91796875, + "learning_rate": 6.789473684210527e-05, + "loss": 6.7842, + "step": 645 + }, + { + "epoch": 0.034, + "grad_norm": 1.1640625, + "learning_rate": 6.800000000000001e-05, + "loss": 6.6735, + "step": 646 + }, + { + "epoch": 0.03405263157894737, + "grad_norm": 1.0703125, + "learning_rate": 6.810526315789474e-05, + "loss": 6.7344, + "step": 647 + }, + { + "epoch": 0.034105263157894736, + "grad_norm": 0.80859375, + "learning_rate": 6.821052631578948e-05, + "loss": 6.8532, + "step": 648 + }, + { + "epoch": 0.0341578947368421, + "grad_norm": 1.1328125, + "learning_rate": 6.83157894736842e-05, + "loss": 6.7419, + "step": 649 + }, + { + "epoch": 0.034210526315789476, + "grad_norm": 1.3984375, + "learning_rate": 6.842105263157895e-05, + "loss": 6.5361, + "step": 650 + }, + { + "epoch": 0.03426315789473684, + "grad_norm": 1.125, + "learning_rate": 6.852631578947369e-05, + "loss": 6.6773, + "step": 651 + }, + { + "epoch": 0.03431578947368421, + "grad_norm": 1.2109375, + "learning_rate": 6.863157894736843e-05, + "loss": 6.5243, + "step": 652 + }, + { + "epoch": 0.034368421052631576, + "grad_norm": 1.46875, + "learning_rate": 6.873684210526317e-05, + "loss": 6.8265, + "step": 653 + }, + { + "epoch": 0.03442105263157895, + "grad_norm": 1.0546875, + "learning_rate": 6.88421052631579e-05, + "loss": 6.8278, + "step": 654 + }, + { + "epoch": 0.034473684210526316, + "grad_norm": 1.3046875, + "learning_rate": 6.894736842105263e-05, + "loss": 6.491, + "step": 655 + }, + { + "epoch": 0.03452631578947368, + "grad_norm": 1.28125, + "learning_rate": 6.905263157894737e-05, + "loss": 6.6275, + "step": 656 + }, + { + "epoch": 0.03457894736842105, + "grad_norm": 1.5234375, + "learning_rate": 6.915789473684211e-05, + "loss": 6.7102, + "step": 657 + }, + { + "epoch": 0.03463157894736842, + "grad_norm": 1.3671875, + "learning_rate": 6.926315789473684e-05, + "loss": 6.5517, + "step": 658 + }, + { + "epoch": 0.03468421052631579, + "grad_norm": 1.25, + "learning_rate": 6.936842105263158e-05, + "loss": 6.4749, + "step": 659 + }, + { + "epoch": 0.034736842105263156, + "grad_norm": 0.99609375, + "learning_rate": 6.947368421052632e-05, + "loss": 6.5256, + "step": 660 + }, + { + "epoch": 0.03478947368421052, + "grad_norm": 1.1484375, + "learning_rate": 6.957894736842106e-05, + "loss": 6.5493, + "step": 661 + }, + { + "epoch": 0.034842105263157896, + "grad_norm": 1.0078125, + "learning_rate": 6.96842105263158e-05, + "loss": 6.4834, + "step": 662 + }, + { + "epoch": 0.03489473684210526, + "grad_norm": 0.94921875, + "learning_rate": 6.978947368421053e-05, + "loss": 6.6399, + "step": 663 + }, + { + "epoch": 0.03494736842105263, + "grad_norm": 1.1875, + "learning_rate": 6.989473684210527e-05, + "loss": 6.5919, + "step": 664 + }, + { + "epoch": 0.035, + "grad_norm": 0.90625, + "learning_rate": 7e-05, + "loss": 6.6459, + "step": 665 + }, + { + "epoch": 0.03505263157894737, + "grad_norm": 1.375, + "learning_rate": 7.010526315789473e-05, + "loss": 6.8316, + "step": 666 + }, + { + "epoch": 0.03510526315789474, + "grad_norm": 1.0859375, + "learning_rate": 7.021052631578949e-05, + "loss": 6.7758, + "step": 667 + }, + { + "epoch": 0.0351578947368421, + "grad_norm": 1.40625, + "learning_rate": 7.031578947368421e-05, + "loss": 6.4694, + "step": 668 + }, + { + "epoch": 0.03521052631578948, + "grad_norm": 1.7578125, + "learning_rate": 7.042105263157895e-05, + "loss": 6.5403, + "step": 669 + }, + { + "epoch": 0.035263157894736843, + "grad_norm": 1.3359375, + "learning_rate": 7.052631578947368e-05, + "loss": 6.7183, + "step": 670 + }, + { + "epoch": 0.03531578947368421, + "grad_norm": 0.86328125, + "learning_rate": 7.063157894736842e-05, + "loss": 6.636, + "step": 671 + }, + { + "epoch": 0.03536842105263158, + "grad_norm": 0.91015625, + "learning_rate": 7.073684210526316e-05, + "loss": 6.449, + "step": 672 + }, + { + "epoch": 0.03542105263157895, + "grad_norm": 1.421875, + "learning_rate": 7.08421052631579e-05, + "loss": 6.6972, + "step": 673 + }, + { + "epoch": 0.03547368421052632, + "grad_norm": 1.078125, + "learning_rate": 7.094736842105264e-05, + "loss": 6.448, + "step": 674 + }, + { + "epoch": 0.035526315789473684, + "grad_norm": 1.1015625, + "learning_rate": 7.105263157894737e-05, + "loss": 6.5325, + "step": 675 + }, + { + "epoch": 0.03557894736842105, + "grad_norm": 1.03125, + "learning_rate": 7.115789473684211e-05, + "loss": 6.63, + "step": 676 + }, + { + "epoch": 0.035631578947368424, + "grad_norm": 1.4609375, + "learning_rate": 7.126315789473685e-05, + "loss": 6.6777, + "step": 677 + }, + { + "epoch": 0.03568421052631579, + "grad_norm": 0.984375, + "learning_rate": 7.136842105263159e-05, + "loss": 6.7636, + "step": 678 + }, + { + "epoch": 0.03573684210526316, + "grad_norm": 1.1796875, + "learning_rate": 7.147368421052631e-05, + "loss": 6.8443, + "step": 679 + }, + { + "epoch": 0.035789473684210524, + "grad_norm": 0.984375, + "learning_rate": 7.157894736842105e-05, + "loss": 6.6891, + "step": 680 + }, + { + "epoch": 0.0358421052631579, + "grad_norm": 1.140625, + "learning_rate": 7.16842105263158e-05, + "loss": 6.2291, + "step": 681 + }, + { + "epoch": 0.035894736842105264, + "grad_norm": 0.9375, + "learning_rate": 7.178947368421054e-05, + "loss": 6.524, + "step": 682 + }, + { + "epoch": 0.03594736842105263, + "grad_norm": 1.15625, + "learning_rate": 7.189473684210528e-05, + "loss": 6.4368, + "step": 683 + }, + { + "epoch": 0.036, + "grad_norm": 1.171875, + "learning_rate": 7.2e-05, + "loss": 6.5144, + "step": 684 + }, + { + "epoch": 0.03605263157894737, + "grad_norm": 1.15625, + "learning_rate": 7.210526315789474e-05, + "loss": 6.4634, + "step": 685 + }, + { + "epoch": 0.03610526315789474, + "grad_norm": 1.265625, + "learning_rate": 7.221052631578947e-05, + "loss": 6.4194, + "step": 686 + }, + { + "epoch": 0.036157894736842104, + "grad_norm": 1.265625, + "learning_rate": 7.231578947368421e-05, + "loss": 6.5292, + "step": 687 + }, + { + "epoch": 0.03621052631578947, + "grad_norm": 0.8671875, + "learning_rate": 7.242105263157896e-05, + "loss": 6.851, + "step": 688 + }, + { + "epoch": 0.036263157894736844, + "grad_norm": 1.3515625, + "learning_rate": 7.252631578947369e-05, + "loss": 6.8635, + "step": 689 + }, + { + "epoch": 0.03631578947368421, + "grad_norm": 1.25, + "learning_rate": 7.263157894736843e-05, + "loss": 6.7727, + "step": 690 + }, + { + "epoch": 0.03636842105263158, + "grad_norm": 0.9140625, + "learning_rate": 7.273684210526316e-05, + "loss": 6.57, + "step": 691 + }, + { + "epoch": 0.036421052631578944, + "grad_norm": 1.2421875, + "learning_rate": 7.28421052631579e-05, + "loss": 6.631, + "step": 692 + }, + { + "epoch": 0.03647368421052632, + "grad_norm": 1.0859375, + "learning_rate": 7.294736842105264e-05, + "loss": 6.2733, + "step": 693 + }, + { + "epoch": 0.036526315789473685, + "grad_norm": 1.28125, + "learning_rate": 7.305263157894738e-05, + "loss": 6.1779, + "step": 694 + }, + { + "epoch": 0.03657894736842105, + "grad_norm": 1.1015625, + "learning_rate": 7.315789473684212e-05, + "loss": 6.6416, + "step": 695 + }, + { + "epoch": 0.03663157894736842, + "grad_norm": 1.078125, + "learning_rate": 7.326315789473684e-05, + "loss": 6.5641, + "step": 696 + }, + { + "epoch": 0.03668421052631579, + "grad_norm": 1.5234375, + "learning_rate": 7.336842105263158e-05, + "loss": 6.701, + "step": 697 + }, + { + "epoch": 0.03673684210526316, + "grad_norm": 1.1484375, + "learning_rate": 7.347368421052632e-05, + "loss": 6.3285, + "step": 698 + }, + { + "epoch": 0.036789473684210525, + "grad_norm": 1.1328125, + "learning_rate": 7.357894736842106e-05, + "loss": 6.5902, + "step": 699 + }, + { + "epoch": 0.03684210526315789, + "grad_norm": 1.046875, + "learning_rate": 7.368421052631579e-05, + "loss": 6.5899, + "step": 700 + }, + { + "epoch": 0.036894736842105265, + "grad_norm": 1.125, + "learning_rate": 7.378947368421053e-05, + "loss": 6.551, + "step": 701 + }, + { + "epoch": 0.03694736842105263, + "grad_norm": 1.140625, + "learning_rate": 7.389473684210527e-05, + "loss": 6.6883, + "step": 702 + }, + { + "epoch": 0.037, + "grad_norm": 1.09375, + "learning_rate": 7.4e-05, + "loss": 6.6442, + "step": 703 + }, + { + "epoch": 0.03705263157894737, + "grad_norm": 1.15625, + "learning_rate": 7.410526315789475e-05, + "loss": 6.4402, + "step": 704 + }, + { + "epoch": 0.03710526315789474, + "grad_norm": 0.9296875, + "learning_rate": 7.421052631578948e-05, + "loss": 6.7411, + "step": 705 + }, + { + "epoch": 0.037157894736842105, + "grad_norm": 1.15625, + "learning_rate": 7.431578947368422e-05, + "loss": 6.4159, + "step": 706 + }, + { + "epoch": 0.03721052631578947, + "grad_norm": 0.91015625, + "learning_rate": 7.442105263157894e-05, + "loss": 6.7058, + "step": 707 + }, + { + "epoch": 0.037263157894736845, + "grad_norm": 0.8828125, + "learning_rate": 7.452631578947368e-05, + "loss": 6.5838, + "step": 708 + }, + { + "epoch": 0.03731578947368421, + "grad_norm": 1.0390625, + "learning_rate": 7.463157894736844e-05, + "loss": 6.6839, + "step": 709 + }, + { + "epoch": 0.03736842105263158, + "grad_norm": 1.4140625, + "learning_rate": 7.473684210526316e-05, + "loss": 6.6106, + "step": 710 + }, + { + "epoch": 0.037421052631578945, + "grad_norm": 1.0234375, + "learning_rate": 7.48421052631579e-05, + "loss": 6.5922, + "step": 711 + }, + { + "epoch": 0.03747368421052632, + "grad_norm": 1.03125, + "learning_rate": 7.494736842105263e-05, + "loss": 6.1978, + "step": 712 + }, + { + "epoch": 0.037526315789473685, + "grad_norm": 0.83203125, + "learning_rate": 7.505263157894737e-05, + "loss": 6.4035, + "step": 713 + }, + { + "epoch": 0.03757894736842105, + "grad_norm": 1.265625, + "learning_rate": 7.515789473684211e-05, + "loss": 6.2618, + "step": 714 + }, + { + "epoch": 0.03763157894736842, + "grad_norm": 0.88671875, + "learning_rate": 7.526315789473685e-05, + "loss": 6.5239, + "step": 715 + }, + { + "epoch": 0.03768421052631579, + "grad_norm": 0.98828125, + "learning_rate": 7.536842105263158e-05, + "loss": 6.6604, + "step": 716 + }, + { + "epoch": 0.03773684210526316, + "grad_norm": 1.1953125, + "learning_rate": 7.547368421052632e-05, + "loss": 6.5382, + "step": 717 + }, + { + "epoch": 0.037789473684210526, + "grad_norm": 1.21875, + "learning_rate": 7.557894736842106e-05, + "loss": 6.3057, + "step": 718 + }, + { + "epoch": 0.03784210526315789, + "grad_norm": 0.8515625, + "learning_rate": 7.56842105263158e-05, + "loss": 6.3712, + "step": 719 + }, + { + "epoch": 0.037894736842105266, + "grad_norm": 1.046875, + "learning_rate": 7.578947368421054e-05, + "loss": 6.5491, + "step": 720 + }, + { + "epoch": 0.03794736842105263, + "grad_norm": 1.0703125, + "learning_rate": 7.589473684210526e-05, + "loss": 6.3615, + "step": 721 + }, + { + "epoch": 0.038, + "grad_norm": 1.328125, + "learning_rate": 7.6e-05, + "loss": 6.6763, + "step": 722 + }, + { + "epoch": 0.038052631578947366, + "grad_norm": 0.88671875, + "learning_rate": 7.610526315789473e-05, + "loss": 6.8064, + "step": 723 + }, + { + "epoch": 0.03810526315789474, + "grad_norm": 1.15625, + "learning_rate": 7.621052631578947e-05, + "loss": 6.5439, + "step": 724 + }, + { + "epoch": 0.038157894736842106, + "grad_norm": 1.2265625, + "learning_rate": 7.631578947368422e-05, + "loss": 6.8097, + "step": 725 + }, + { + "epoch": 0.03821052631578947, + "grad_norm": 1.0390625, + "learning_rate": 7.642105263157895e-05, + "loss": 6.4788, + "step": 726 + }, + { + "epoch": 0.03826315789473684, + "grad_norm": 1.0234375, + "learning_rate": 7.652631578947369e-05, + "loss": 6.571, + "step": 727 + }, + { + "epoch": 0.03831578947368421, + "grad_norm": 1.0234375, + "learning_rate": 7.663157894736842e-05, + "loss": 6.8187, + "step": 728 + }, + { + "epoch": 0.03836842105263158, + "grad_norm": 1.1875, + "learning_rate": 7.673684210526316e-05, + "loss": 6.7194, + "step": 729 + }, + { + "epoch": 0.038421052631578946, + "grad_norm": 1.3125, + "learning_rate": 7.68421052631579e-05, + "loss": 6.8424, + "step": 730 + }, + { + "epoch": 0.03847368421052631, + "grad_norm": 1.8046875, + "learning_rate": 7.694736842105264e-05, + "loss": 6.4379, + "step": 731 + }, + { + "epoch": 0.038526315789473686, + "grad_norm": 1.171875, + "learning_rate": 7.705263157894738e-05, + "loss": 6.5743, + "step": 732 + }, + { + "epoch": 0.03857894736842105, + "grad_norm": 1.1796875, + "learning_rate": 7.71578947368421e-05, + "loss": 6.4531, + "step": 733 + }, + { + "epoch": 0.03863157894736842, + "grad_norm": 1.4296875, + "learning_rate": 7.726315789473684e-05, + "loss": 6.4282, + "step": 734 + }, + { + "epoch": 0.038684210526315786, + "grad_norm": 1.1171875, + "learning_rate": 7.736842105263159e-05, + "loss": 6.4274, + "step": 735 + }, + { + "epoch": 0.03873684210526316, + "grad_norm": 1.2265625, + "learning_rate": 7.747368421052633e-05, + "loss": 6.7247, + "step": 736 + }, + { + "epoch": 0.038789473684210526, + "grad_norm": 0.9453125, + "learning_rate": 7.757894736842105e-05, + "loss": 6.6674, + "step": 737 + }, + { + "epoch": 0.03884210526315789, + "grad_norm": 1.0859375, + "learning_rate": 7.768421052631579e-05, + "loss": 6.309, + "step": 738 + }, + { + "epoch": 0.03889473684210526, + "grad_norm": 1.078125, + "learning_rate": 7.778947368421053e-05, + "loss": 6.4261, + "step": 739 + }, + { + "epoch": 0.03894736842105263, + "grad_norm": 1.21875, + "learning_rate": 7.789473684210526e-05, + "loss": 6.5429, + "step": 740 + }, + { + "epoch": 0.039, + "grad_norm": 0.953125, + "learning_rate": 7.800000000000001e-05, + "loss": 6.5424, + "step": 741 + }, + { + "epoch": 0.03905263157894737, + "grad_norm": 1.140625, + "learning_rate": 7.810526315789474e-05, + "loss": 6.7381, + "step": 742 + }, + { + "epoch": 0.03910526315789474, + "grad_norm": 1.25, + "learning_rate": 7.821052631578948e-05, + "loss": 6.4234, + "step": 743 + }, + { + "epoch": 0.03915789473684211, + "grad_norm": 1.1796875, + "learning_rate": 7.83157894736842e-05, + "loss": 6.5553, + "step": 744 + }, + { + "epoch": 0.03921052631578947, + "grad_norm": 0.80859375, + "learning_rate": 7.842105263157895e-05, + "loss": 6.4691, + "step": 745 + }, + { + "epoch": 0.03926315789473684, + "grad_norm": 1.1171875, + "learning_rate": 7.852631578947369e-05, + "loss": 6.3756, + "step": 746 + }, + { + "epoch": 0.039315789473684214, + "grad_norm": 0.87890625, + "learning_rate": 7.863157894736843e-05, + "loss": 6.4865, + "step": 747 + }, + { + "epoch": 0.03936842105263158, + "grad_norm": 1.1875, + "learning_rate": 7.873684210526317e-05, + "loss": 6.4593, + "step": 748 + }, + { + "epoch": 0.03942105263157895, + "grad_norm": 1.1796875, + "learning_rate": 7.884210526315789e-05, + "loss": 6.3955, + "step": 749 + }, + { + "epoch": 0.039473684210526314, + "grad_norm": 1.28125, + "learning_rate": 7.894736842105263e-05, + "loss": 6.1558, + "step": 750 + }, + { + "epoch": 0.03952631578947369, + "grad_norm": 1.078125, + "learning_rate": 7.905263157894737e-05, + "loss": 6.4829, + "step": 751 + }, + { + "epoch": 0.039578947368421054, + "grad_norm": 1.015625, + "learning_rate": 7.915789473684211e-05, + "loss": 6.4231, + "step": 752 + }, + { + "epoch": 0.03963157894736842, + "grad_norm": 1.328125, + "learning_rate": 7.926315789473685e-05, + "loss": 6.3431, + "step": 753 + }, + { + "epoch": 0.03968421052631579, + "grad_norm": 1.0, + "learning_rate": 7.936842105263158e-05, + "loss": 6.386, + "step": 754 + }, + { + "epoch": 0.03973684210526316, + "grad_norm": 1.234375, + "learning_rate": 7.947368421052632e-05, + "loss": 6.4416, + "step": 755 + }, + { + "epoch": 0.03978947368421053, + "grad_norm": 0.88671875, + "learning_rate": 7.957894736842106e-05, + "loss": 6.3795, + "step": 756 + }, + { + "epoch": 0.039842105263157894, + "grad_norm": 1.2109375, + "learning_rate": 7.96842105263158e-05, + "loss": 6.3552, + "step": 757 + }, + { + "epoch": 0.03989473684210526, + "grad_norm": 1.1640625, + "learning_rate": 7.978947368421053e-05, + "loss": 6.3773, + "step": 758 + }, + { + "epoch": 0.039947368421052634, + "grad_norm": 1.2734375, + "learning_rate": 7.989473684210527e-05, + "loss": 6.0715, + "step": 759 + }, + { + "epoch": 0.04, + "grad_norm": 1.0390625, + "learning_rate": 8e-05, + "loss": 6.4522, + "step": 760 + }, + { + "epoch": 0.04005263157894737, + "grad_norm": 1.1015625, + "learning_rate": 8.010526315789473e-05, + "loss": 6.2927, + "step": 761 + }, + { + "epoch": 0.040105263157894734, + "grad_norm": 1.2734375, + "learning_rate": 8.021052631578949e-05, + "loss": 6.1635, + "step": 762 + }, + { + "epoch": 0.04015789473684211, + "grad_norm": 0.89453125, + "learning_rate": 8.031578947368421e-05, + "loss": 6.4677, + "step": 763 + }, + { + "epoch": 0.040210526315789474, + "grad_norm": 1.265625, + "learning_rate": 8.042105263157895e-05, + "loss": 6.5485, + "step": 764 + }, + { + "epoch": 0.04026315789473684, + "grad_norm": 1.0234375, + "learning_rate": 8.052631578947368e-05, + "loss": 6.2118, + "step": 765 + }, + { + "epoch": 0.04031578947368421, + "grad_norm": 0.984375, + "learning_rate": 8.063157894736842e-05, + "loss": 6.8807, + "step": 766 + }, + { + "epoch": 0.04036842105263158, + "grad_norm": 0.99609375, + "learning_rate": 8.073684210526316e-05, + "loss": 6.1885, + "step": 767 + }, + { + "epoch": 0.04042105263157895, + "grad_norm": 1.453125, + "learning_rate": 8.08421052631579e-05, + "loss": 6.1052, + "step": 768 + }, + { + "epoch": 0.040473684210526314, + "grad_norm": 0.82421875, + "learning_rate": 8.094736842105264e-05, + "loss": 6.577, + "step": 769 + }, + { + "epoch": 0.04052631578947368, + "grad_norm": 1.1171875, + "learning_rate": 8.105263157894737e-05, + "loss": 6.4082, + "step": 770 + }, + { + "epoch": 0.040578947368421055, + "grad_norm": 1.0078125, + "learning_rate": 8.115789473684211e-05, + "loss": 6.5778, + "step": 771 + }, + { + "epoch": 0.04063157894736842, + "grad_norm": 1.1640625, + "learning_rate": 8.126315789473685e-05, + "loss": 6.5514, + "step": 772 + }, + { + "epoch": 0.04068421052631579, + "grad_norm": 1.4921875, + "learning_rate": 8.136842105263159e-05, + "loss": 6.3773, + "step": 773 + }, + { + "epoch": 0.040736842105263155, + "grad_norm": 1.0, + "learning_rate": 8.147368421052633e-05, + "loss": 6.8337, + "step": 774 + }, + { + "epoch": 0.04078947368421053, + "grad_norm": 1.671875, + "learning_rate": 8.157894736842105e-05, + "loss": 6.4322, + "step": 775 + }, + { + "epoch": 0.040842105263157895, + "grad_norm": 0.80078125, + "learning_rate": 8.16842105263158e-05, + "loss": 6.3483, + "step": 776 + }, + { + "epoch": 0.04089473684210526, + "grad_norm": 0.9140625, + "learning_rate": 8.178947368421052e-05, + "loss": 6.3701, + "step": 777 + }, + { + "epoch": 0.04094736842105263, + "grad_norm": 1.0703125, + "learning_rate": 8.189473684210527e-05, + "loss": 6.5687, + "step": 778 + }, + { + "epoch": 0.041, + "grad_norm": 1.0859375, + "learning_rate": 8.2e-05, + "loss": 6.6968, + "step": 779 + }, + { + "epoch": 0.04105263157894737, + "grad_norm": 0.94921875, + "learning_rate": 8.210526315789474e-05, + "loss": 6.7713, + "step": 780 + }, + { + "epoch": 0.041105263157894735, + "grad_norm": 1.046875, + "learning_rate": 8.221052631578948e-05, + "loss": 6.4513, + "step": 781 + }, + { + "epoch": 0.04115789473684211, + "grad_norm": 0.796875, + "learning_rate": 8.231578947368421e-05, + "loss": 6.3171, + "step": 782 + }, + { + "epoch": 0.041210526315789475, + "grad_norm": 0.90625, + "learning_rate": 8.242105263157895e-05, + "loss": 6.4306, + "step": 783 + }, + { + "epoch": 0.04126315789473684, + "grad_norm": 1.171875, + "learning_rate": 8.252631578947369e-05, + "loss": 6.2224, + "step": 784 + }, + { + "epoch": 0.04131578947368421, + "grad_norm": 0.8203125, + "learning_rate": 8.263157894736843e-05, + "loss": 6.7259, + "step": 785 + }, + { + "epoch": 0.04136842105263158, + "grad_norm": 0.91796875, + "learning_rate": 8.273684210526315e-05, + "loss": 6.4982, + "step": 786 + }, + { + "epoch": 0.04142105263157895, + "grad_norm": 0.97265625, + "learning_rate": 8.28421052631579e-05, + "loss": 6.3331, + "step": 787 + }, + { + "epoch": 0.041473684210526315, + "grad_norm": 0.859375, + "learning_rate": 8.294736842105263e-05, + "loss": 6.5404, + "step": 788 + }, + { + "epoch": 0.04152631578947368, + "grad_norm": 1.078125, + "learning_rate": 8.305263157894737e-05, + "loss": 6.7766, + "step": 789 + }, + { + "epoch": 0.041578947368421056, + "grad_norm": 0.96484375, + "learning_rate": 8.315789473684212e-05, + "loss": 6.3961, + "step": 790 + }, + { + "epoch": 0.04163157894736842, + "grad_norm": 1.3515625, + "learning_rate": 8.326315789473684e-05, + "loss": 6.4969, + "step": 791 + }, + { + "epoch": 0.04168421052631579, + "grad_norm": 1.28125, + "learning_rate": 8.336842105263158e-05, + "loss": 6.4069, + "step": 792 + }, + { + "epoch": 0.041736842105263156, + "grad_norm": 0.9609375, + "learning_rate": 8.347368421052631e-05, + "loss": 6.3092, + "step": 793 + }, + { + "epoch": 0.04178947368421053, + "grad_norm": 1.0625, + "learning_rate": 8.357894736842106e-05, + "loss": 6.2526, + "step": 794 + }, + { + "epoch": 0.041842105263157896, + "grad_norm": 0.99609375, + "learning_rate": 8.36842105263158e-05, + "loss": 6.326, + "step": 795 + }, + { + "epoch": 0.04189473684210526, + "grad_norm": 1.1484375, + "learning_rate": 8.378947368421053e-05, + "loss": 6.4604, + "step": 796 + }, + { + "epoch": 0.04194736842105263, + "grad_norm": 1.09375, + "learning_rate": 8.389473684210527e-05, + "loss": 6.4629, + "step": 797 + }, + { + "epoch": 0.042, + "grad_norm": 1.0859375, + "learning_rate": 8.4e-05, + "loss": 6.2009, + "step": 798 + }, + { + "epoch": 0.04205263157894737, + "grad_norm": 1.21875, + "learning_rate": 8.410526315789475e-05, + "loss": 6.5999, + "step": 799 + }, + { + "epoch": 0.042105263157894736, + "grad_norm": 1.1640625, + "learning_rate": 8.421052631578948e-05, + "loss": 6.4316, + "step": 800 + }, + { + "epoch": 0.0421578947368421, + "grad_norm": 1.15625, + "learning_rate": 8.431578947368422e-05, + "loss": 6.2384, + "step": 801 + }, + { + "epoch": 0.042210526315789476, + "grad_norm": 1.25, + "learning_rate": 8.442105263157896e-05, + "loss": 6.511, + "step": 802 + }, + { + "epoch": 0.04226315789473684, + "grad_norm": 1.0078125, + "learning_rate": 8.452631578947368e-05, + "loss": 6.3088, + "step": 803 + }, + { + "epoch": 0.04231578947368421, + "grad_norm": 1.328125, + "learning_rate": 8.463157894736842e-05, + "loss": 6.4654, + "step": 804 + }, + { + "epoch": 0.042368421052631576, + "grad_norm": 1.0078125, + "learning_rate": 8.473684210526316e-05, + "loss": 6.5246, + "step": 805 + }, + { + "epoch": 0.04242105263157895, + "grad_norm": 1.125, + "learning_rate": 8.48421052631579e-05, + "loss": 6.466, + "step": 806 + }, + { + "epoch": 0.042473684210526316, + "grad_norm": 0.98828125, + "learning_rate": 8.494736842105263e-05, + "loss": 6.3813, + "step": 807 + }, + { + "epoch": 0.04252631578947368, + "grad_norm": 0.96875, + "learning_rate": 8.505263157894737e-05, + "loss": 6.2182, + "step": 808 + }, + { + "epoch": 0.04257894736842105, + "grad_norm": 1.0234375, + "learning_rate": 8.515789473684211e-05, + "loss": 6.5327, + "step": 809 + }, + { + "epoch": 0.04263157894736842, + "grad_norm": 1.375, + "learning_rate": 8.526315789473685e-05, + "loss": 6.1009, + "step": 810 + }, + { + "epoch": 0.04268421052631579, + "grad_norm": 1.0625, + "learning_rate": 8.536842105263159e-05, + "loss": 6.3407, + "step": 811 + }, + { + "epoch": 0.042736842105263156, + "grad_norm": 1.34375, + "learning_rate": 8.547368421052632e-05, + "loss": 6.1663, + "step": 812 + }, + { + "epoch": 0.04278947368421052, + "grad_norm": 1.125, + "learning_rate": 8.557894736842106e-05, + "loss": 6.2337, + "step": 813 + }, + { + "epoch": 0.0428421052631579, + "grad_norm": 0.8359375, + "learning_rate": 8.568421052631578e-05, + "loss": 6.2135, + "step": 814 + }, + { + "epoch": 0.04289473684210526, + "grad_norm": 1.0703125, + "learning_rate": 8.578947368421054e-05, + "loss": 6.3157, + "step": 815 + }, + { + "epoch": 0.04294736842105263, + "grad_norm": 0.91015625, + "learning_rate": 8.589473684210528e-05, + "loss": 6.1445, + "step": 816 + }, + { + "epoch": 0.043, + "grad_norm": 0.91015625, + "learning_rate": 8.6e-05, + "loss": 6.2215, + "step": 817 + }, + { + "epoch": 0.04305263157894737, + "grad_norm": 1.3515625, + "learning_rate": 8.610526315789474e-05, + "loss": 6.5658, + "step": 818 + }, + { + "epoch": 0.04310526315789474, + "grad_norm": 1.234375, + "learning_rate": 8.621052631578947e-05, + "loss": 6.6626, + "step": 819 + }, + { + "epoch": 0.0431578947368421, + "grad_norm": 1.015625, + "learning_rate": 8.631578947368421e-05, + "loss": 6.5314, + "step": 820 + }, + { + "epoch": 0.04321052631578948, + "grad_norm": 0.91015625, + "learning_rate": 8.642105263157895e-05, + "loss": 6.5715, + "step": 821 + }, + { + "epoch": 0.043263157894736844, + "grad_norm": 1.1953125, + "learning_rate": 8.652631578947369e-05, + "loss": 6.3216, + "step": 822 + }, + { + "epoch": 0.04331578947368421, + "grad_norm": 1.1796875, + "learning_rate": 8.663157894736843e-05, + "loss": 6.3734, + "step": 823 + }, + { + "epoch": 0.04336842105263158, + "grad_norm": 1.0703125, + "learning_rate": 8.673684210526316e-05, + "loss": 6.2693, + "step": 824 + }, + { + "epoch": 0.04342105263157895, + "grad_norm": 1.1796875, + "learning_rate": 8.68421052631579e-05, + "loss": 6.3327, + "step": 825 + }, + { + "epoch": 0.04347368421052632, + "grad_norm": 1.1953125, + "learning_rate": 8.694736842105264e-05, + "loss": 6.1973, + "step": 826 + }, + { + "epoch": 0.043526315789473684, + "grad_norm": 0.890625, + "learning_rate": 8.705263157894738e-05, + "loss": 6.356, + "step": 827 + }, + { + "epoch": 0.04357894736842105, + "grad_norm": 1.4453125, + "learning_rate": 8.71578947368421e-05, + "loss": 6.3657, + "step": 828 + }, + { + "epoch": 0.043631578947368424, + "grad_norm": 1.0078125, + "learning_rate": 8.726315789473684e-05, + "loss": 6.1327, + "step": 829 + }, + { + "epoch": 0.04368421052631579, + "grad_norm": 1.015625, + "learning_rate": 8.736842105263158e-05, + "loss": 6.5449, + "step": 830 + }, + { + "epoch": 0.04373684210526316, + "grad_norm": 1.3203125, + "learning_rate": 8.747368421052632e-05, + "loss": 6.1413, + "step": 831 + }, + { + "epoch": 0.043789473684210524, + "grad_norm": 1.390625, + "learning_rate": 8.757894736842106e-05, + "loss": 6.5375, + "step": 832 + }, + { + "epoch": 0.0438421052631579, + "grad_norm": 1.28125, + "learning_rate": 8.768421052631579e-05, + "loss": 5.9336, + "step": 833 + }, + { + "epoch": 0.043894736842105264, + "grad_norm": 1.015625, + "learning_rate": 8.778947368421053e-05, + "loss": 6.3116, + "step": 834 + }, + { + "epoch": 0.04394736842105263, + "grad_norm": 0.87109375, + "learning_rate": 8.789473684210526e-05, + "loss": 6.3506, + "step": 835 + }, + { + "epoch": 0.044, + "grad_norm": 1.1484375, + "learning_rate": 8.800000000000001e-05, + "loss": 6.5033, + "step": 836 + }, + { + "epoch": 0.04405263157894737, + "grad_norm": 1.1328125, + "learning_rate": 8.810526315789475e-05, + "loss": 6.5652, + "step": 837 + }, + { + "epoch": 0.04410526315789474, + "grad_norm": 1.078125, + "learning_rate": 8.821052631578948e-05, + "loss": 6.139, + "step": 838 + }, + { + "epoch": 0.044157894736842104, + "grad_norm": 1.234375, + "learning_rate": 8.831578947368422e-05, + "loss": 6.4565, + "step": 839 + }, + { + "epoch": 0.04421052631578947, + "grad_norm": 1.125, + "learning_rate": 8.842105263157894e-05, + "loss": 6.4141, + "step": 840 + }, + { + "epoch": 0.044263157894736845, + "grad_norm": 0.8984375, + "learning_rate": 8.852631578947368e-05, + "loss": 6.2788, + "step": 841 + }, + { + "epoch": 0.04431578947368421, + "grad_norm": 1.0625, + "learning_rate": 8.863157894736842e-05, + "loss": 6.5033, + "step": 842 + }, + { + "epoch": 0.04436842105263158, + "grad_norm": 1.0078125, + "learning_rate": 8.873684210526316e-05, + "loss": 6.1942, + "step": 843 + }, + { + "epoch": 0.044421052631578944, + "grad_norm": 1.2578125, + "learning_rate": 8.88421052631579e-05, + "loss": 6.0617, + "step": 844 + }, + { + "epoch": 0.04447368421052632, + "grad_norm": 0.93359375, + "learning_rate": 8.894736842105263e-05, + "loss": 6.5066, + "step": 845 + }, + { + "epoch": 0.044526315789473685, + "grad_norm": 0.95703125, + "learning_rate": 8.905263157894737e-05, + "loss": 6.3333, + "step": 846 + }, + { + "epoch": 0.04457894736842105, + "grad_norm": 1.0390625, + "learning_rate": 8.915789473684211e-05, + "loss": 6.4852, + "step": 847 + }, + { + "epoch": 0.04463157894736842, + "grad_norm": 1.0078125, + "learning_rate": 8.926315789473685e-05, + "loss": 6.4245, + "step": 848 + }, + { + "epoch": 0.04468421052631579, + "grad_norm": 1.03125, + "learning_rate": 8.936842105263158e-05, + "loss": 6.2295, + "step": 849 + }, + { + "epoch": 0.04473684210526316, + "grad_norm": 0.84765625, + "learning_rate": 8.947368421052632e-05, + "loss": 6.305, + "step": 850 + }, + { + "epoch": 0.044789473684210525, + "grad_norm": 1.0078125, + "learning_rate": 8.957894736842106e-05, + "loss": 6.1627, + "step": 851 + }, + { + "epoch": 0.04484210526315789, + "grad_norm": 1.2578125, + "learning_rate": 8.96842105263158e-05, + "loss": 6.0413, + "step": 852 + }, + { + "epoch": 0.044894736842105265, + "grad_norm": 1.3125, + "learning_rate": 8.978947368421054e-05, + "loss": 5.9376, + "step": 853 + }, + { + "epoch": 0.04494736842105263, + "grad_norm": 1.234375, + "learning_rate": 8.989473684210527e-05, + "loss": 6.1253, + "step": 854 + }, + { + "epoch": 0.045, + "grad_norm": 0.82421875, + "learning_rate": 9e-05, + "loss": 6.3335, + "step": 855 + }, + { + "epoch": 0.045052631578947365, + "grad_norm": 0.99609375, + "learning_rate": 9.010526315789473e-05, + "loss": 6.3991, + "step": 856 + }, + { + "epoch": 0.04510526315789474, + "grad_norm": 0.93359375, + "learning_rate": 9.021052631578947e-05, + "loss": 6.2851, + "step": 857 + }, + { + "epoch": 0.045157894736842105, + "grad_norm": 1.0859375, + "learning_rate": 9.031578947368423e-05, + "loss": 6.3521, + "step": 858 + }, + { + "epoch": 0.04521052631578947, + "grad_norm": 0.96484375, + "learning_rate": 9.042105263157895e-05, + "loss": 6.2588, + "step": 859 + }, + { + "epoch": 0.045263157894736845, + "grad_norm": 0.80859375, + "learning_rate": 9.052631578947369e-05, + "loss": 6.5109, + "step": 860 + }, + { + "epoch": 0.04531578947368421, + "grad_norm": 1.0390625, + "learning_rate": 9.063157894736842e-05, + "loss": 6.4478, + "step": 861 + }, + { + "epoch": 0.04536842105263158, + "grad_norm": 1.03125, + "learning_rate": 9.073684210526316e-05, + "loss": 6.4184, + "step": 862 + }, + { + "epoch": 0.045421052631578945, + "grad_norm": 0.94921875, + "learning_rate": 9.08421052631579e-05, + "loss": 6.3009, + "step": 863 + }, + { + "epoch": 0.04547368421052632, + "grad_norm": 1.015625, + "learning_rate": 9.094736842105264e-05, + "loss": 6.3847, + "step": 864 + }, + { + "epoch": 0.045526315789473686, + "grad_norm": 1.0390625, + "learning_rate": 9.105263157894738e-05, + "loss": 6.0883, + "step": 865 + }, + { + "epoch": 0.04557894736842105, + "grad_norm": 1.140625, + "learning_rate": 9.11578947368421e-05, + "loss": 6.2109, + "step": 866 + }, + { + "epoch": 0.04563157894736842, + "grad_norm": 0.96484375, + "learning_rate": 9.126315789473685e-05, + "loss": 6.1742, + "step": 867 + }, + { + "epoch": 0.04568421052631579, + "grad_norm": 1.0, + "learning_rate": 9.136842105263159e-05, + "loss": 6.1872, + "step": 868 + }, + { + "epoch": 0.04573684210526316, + "grad_norm": 0.90625, + "learning_rate": 9.147368421052633e-05, + "loss": 6.3442, + "step": 869 + }, + { + "epoch": 0.045789473684210526, + "grad_norm": 0.8984375, + "learning_rate": 9.157894736842105e-05, + "loss": 6.1095, + "step": 870 + }, + { + "epoch": 0.04584210526315789, + "grad_norm": 0.75, + "learning_rate": 9.16842105263158e-05, + "loss": 6.0219, + "step": 871 + }, + { + "epoch": 0.045894736842105266, + "grad_norm": 1.015625, + "learning_rate": 9.178947368421052e-05, + "loss": 6.0729, + "step": 872 + }, + { + "epoch": 0.04594736842105263, + "grad_norm": 1.1796875, + "learning_rate": 9.189473684210527e-05, + "loss": 6.145, + "step": 873 + }, + { + "epoch": 0.046, + "grad_norm": 1.1171875, + "learning_rate": 9.200000000000001e-05, + "loss": 6.0401, + "step": 874 + }, + { + "epoch": 0.046052631578947366, + "grad_norm": 1.2265625, + "learning_rate": 9.210526315789474e-05, + "loss": 5.7981, + "step": 875 + }, + { + "epoch": 0.04610526315789474, + "grad_norm": 0.95703125, + "learning_rate": 9.221052631578948e-05, + "loss": 6.509, + "step": 876 + }, + { + "epoch": 0.046157894736842106, + "grad_norm": 1.234375, + "learning_rate": 9.231578947368421e-05, + "loss": 6.2705, + "step": 877 + }, + { + "epoch": 0.04621052631578947, + "grad_norm": 1.3125, + "learning_rate": 9.242105263157895e-05, + "loss": 6.5153, + "step": 878 + }, + { + "epoch": 0.04626315789473684, + "grad_norm": 1.25, + "learning_rate": 9.252631578947369e-05, + "loss": 6.246, + "step": 879 + }, + { + "epoch": 0.04631578947368421, + "grad_norm": 0.9296875, + "learning_rate": 9.263157894736843e-05, + "loss": 6.2896, + "step": 880 + }, + { + "epoch": 0.04636842105263158, + "grad_norm": 1.0546875, + "learning_rate": 9.273684210526317e-05, + "loss": 6.5107, + "step": 881 + }, + { + "epoch": 0.046421052631578946, + "grad_norm": 1.0859375, + "learning_rate": 9.28421052631579e-05, + "loss": 6.363, + "step": 882 + }, + { + "epoch": 0.04647368421052631, + "grad_norm": 1.1640625, + "learning_rate": 9.294736842105263e-05, + "loss": 6.4937, + "step": 883 + }, + { + "epoch": 0.046526315789473686, + "grad_norm": 0.953125, + "learning_rate": 9.305263157894737e-05, + "loss": 6.4944, + "step": 884 + }, + { + "epoch": 0.04657894736842105, + "grad_norm": 1.5546875, + "learning_rate": 9.315789473684211e-05, + "loss": 6.6524, + "step": 885 + }, + { + "epoch": 0.04663157894736842, + "grad_norm": 0.984375, + "learning_rate": 9.326315789473684e-05, + "loss": 5.8883, + "step": 886 + }, + { + "epoch": 0.046684210526315786, + "grad_norm": 1.296875, + "learning_rate": 9.336842105263158e-05, + "loss": 6.2053, + "step": 887 + }, + { + "epoch": 0.04673684210526316, + "grad_norm": 0.8515625, + "learning_rate": 9.347368421052632e-05, + "loss": 6.1712, + "step": 888 + }, + { + "epoch": 0.04678947368421053, + "grad_norm": 0.83984375, + "learning_rate": 9.357894736842106e-05, + "loss": 6.3711, + "step": 889 + }, + { + "epoch": 0.04684210526315789, + "grad_norm": 1.140625, + "learning_rate": 9.36842105263158e-05, + "loss": 6.2285, + "step": 890 + }, + { + "epoch": 0.04689473684210526, + "grad_norm": 0.875, + "learning_rate": 9.378947368421053e-05, + "loss": 6.3593, + "step": 891 + }, + { + "epoch": 0.04694736842105263, + "grad_norm": 0.98046875, + "learning_rate": 9.389473684210527e-05, + "loss": 6.3647, + "step": 892 + }, + { + "epoch": 0.047, + "grad_norm": 1.1640625, + "learning_rate": 9.4e-05, + "loss": 5.8563, + "step": 893 + }, + { + "epoch": 0.04705263157894737, + "grad_norm": 0.88671875, + "learning_rate": 9.410526315789473e-05, + "loss": 6.1314, + "step": 894 + }, + { + "epoch": 0.04710526315789473, + "grad_norm": 1.328125, + "learning_rate": 9.421052631578949e-05, + "loss": 6.4728, + "step": 895 + }, + { + "epoch": 0.04715789473684211, + "grad_norm": 1.0, + "learning_rate": 9.431578947368421e-05, + "loss": 6.2365, + "step": 896 + }, + { + "epoch": 0.047210526315789474, + "grad_norm": 1.3359375, + "learning_rate": 9.442105263157895e-05, + "loss": 6.3055, + "step": 897 + }, + { + "epoch": 0.04726315789473684, + "grad_norm": 0.98046875, + "learning_rate": 9.452631578947368e-05, + "loss": 6.3249, + "step": 898 + }, + { + "epoch": 0.047315789473684214, + "grad_norm": 0.91015625, + "learning_rate": 9.463157894736842e-05, + "loss": 6.2401, + "step": 899 + }, + { + "epoch": 0.04736842105263158, + "grad_norm": 0.98828125, + "learning_rate": 9.473684210526316e-05, + "loss": 6.4491, + "step": 900 + }, + { + "epoch": 0.04742105263157895, + "grad_norm": 0.8125, + "learning_rate": 9.48421052631579e-05, + "loss": 6.2155, + "step": 901 + }, + { + "epoch": 0.047473684210526314, + "grad_norm": 0.8828125, + "learning_rate": 9.494736842105264e-05, + "loss": 6.1934, + "step": 902 + }, + { + "epoch": 0.04752631578947369, + "grad_norm": 1.015625, + "learning_rate": 9.505263157894737e-05, + "loss": 6.248, + "step": 903 + }, + { + "epoch": 0.047578947368421054, + "grad_norm": 1.46875, + "learning_rate": 9.515789473684211e-05, + "loss": 6.2465, + "step": 904 + }, + { + "epoch": 0.04763157894736842, + "grad_norm": 0.97265625, + "learning_rate": 9.526315789473685e-05, + "loss": 6.0931, + "step": 905 + }, + { + "epoch": 0.04768421052631579, + "grad_norm": 0.88671875, + "learning_rate": 9.536842105263159e-05, + "loss": 6.4757, + "step": 906 + }, + { + "epoch": 0.04773684210526316, + "grad_norm": 0.8203125, + "learning_rate": 9.547368421052632e-05, + "loss": 6.2212, + "step": 907 + }, + { + "epoch": 0.04778947368421053, + "grad_norm": 0.95703125, + "learning_rate": 9.557894736842106e-05, + "loss": 6.5646, + "step": 908 + }, + { + "epoch": 0.047842105263157894, + "grad_norm": 0.97265625, + "learning_rate": 9.56842105263158e-05, + "loss": 6.2845, + "step": 909 + }, + { + "epoch": 0.04789473684210526, + "grad_norm": 0.90625, + "learning_rate": 9.578947368421052e-05, + "loss": 6.3641, + "step": 910 + }, + { + "epoch": 0.047947368421052634, + "grad_norm": 1.0859375, + "learning_rate": 9.589473684210528e-05, + "loss": 6.119, + "step": 911 + }, + { + "epoch": 0.048, + "grad_norm": 0.9765625, + "learning_rate": 9.6e-05, + "loss": 6.2721, + "step": 912 + }, + { + "epoch": 0.04805263157894737, + "grad_norm": 1.0546875, + "learning_rate": 9.610526315789474e-05, + "loss": 6.463, + "step": 913 + }, + { + "epoch": 0.048105263157894734, + "grad_norm": 1.046875, + "learning_rate": 9.621052631578947e-05, + "loss": 6.4011, + "step": 914 + }, + { + "epoch": 0.04815789473684211, + "grad_norm": 0.85546875, + "learning_rate": 9.631578947368421e-05, + "loss": 6.2156, + "step": 915 + }, + { + "epoch": 0.048210526315789475, + "grad_norm": 0.8203125, + "learning_rate": 9.642105263157896e-05, + "loss": 6.2602, + "step": 916 + }, + { + "epoch": 0.04826315789473684, + "grad_norm": 0.98046875, + "learning_rate": 9.652631578947369e-05, + "loss": 6.3975, + "step": 917 + }, + { + "epoch": 0.04831578947368421, + "grad_norm": 1.0546875, + "learning_rate": 9.663157894736843e-05, + "loss": 6.1695, + "step": 918 + }, + { + "epoch": 0.04836842105263158, + "grad_norm": 0.92578125, + "learning_rate": 9.673684210526316e-05, + "loss": 5.8504, + "step": 919 + }, + { + "epoch": 0.04842105263157895, + "grad_norm": 0.984375, + "learning_rate": 9.68421052631579e-05, + "loss": 6.1437, + "step": 920 + }, + { + "epoch": 0.048473684210526315, + "grad_norm": 0.82421875, + "learning_rate": 9.694736842105264e-05, + "loss": 5.9755, + "step": 921 + }, + { + "epoch": 0.04852631578947368, + "grad_norm": 0.91015625, + "learning_rate": 9.705263157894738e-05, + "loss": 6.2987, + "step": 922 + }, + { + "epoch": 0.048578947368421055, + "grad_norm": 1.0078125, + "learning_rate": 9.715789473684212e-05, + "loss": 6.2352, + "step": 923 + }, + { + "epoch": 0.04863157894736842, + "grad_norm": 1.0625, + "learning_rate": 9.726315789473684e-05, + "loss": 6.2124, + "step": 924 + }, + { + "epoch": 0.04868421052631579, + "grad_norm": 0.93359375, + "learning_rate": 9.736842105263158e-05, + "loss": 6.2916, + "step": 925 + }, + { + "epoch": 0.048736842105263155, + "grad_norm": 0.984375, + "learning_rate": 9.747368421052632e-05, + "loss": 6.2376, + "step": 926 + }, + { + "epoch": 0.04878947368421053, + "grad_norm": 0.98046875, + "learning_rate": 9.757894736842106e-05, + "loss": 6.3217, + "step": 927 + }, + { + "epoch": 0.048842105263157895, + "grad_norm": 1.015625, + "learning_rate": 9.768421052631579e-05, + "loss": 6.0661, + "step": 928 + }, + { + "epoch": 0.04889473684210526, + "grad_norm": 1.1015625, + "learning_rate": 9.778947368421053e-05, + "loss": 6.4374, + "step": 929 + }, + { + "epoch": 0.04894736842105263, + "grad_norm": 0.99609375, + "learning_rate": 9.789473684210527e-05, + "loss": 6.1818, + "step": 930 + }, + { + "epoch": 0.049, + "grad_norm": 0.74609375, + "learning_rate": 9.8e-05, + "loss": 6.3103, + "step": 931 + }, + { + "epoch": 0.04905263157894737, + "grad_norm": 0.8046875, + "learning_rate": 9.810526315789475e-05, + "loss": 6.0478, + "step": 932 + }, + { + "epoch": 0.049105263157894735, + "grad_norm": 0.8671875, + "learning_rate": 9.821052631578948e-05, + "loss": 6.4295, + "step": 933 + }, + { + "epoch": 0.0491578947368421, + "grad_norm": 1.1484375, + "learning_rate": 9.831578947368422e-05, + "loss": 6.5219, + "step": 934 + }, + { + "epoch": 0.049210526315789475, + "grad_norm": 1.1171875, + "learning_rate": 9.842105263157894e-05, + "loss": 6.0292, + "step": 935 + }, + { + "epoch": 0.04926315789473684, + "grad_norm": 1.1328125, + "learning_rate": 9.852631578947368e-05, + "loss": 6.1391, + "step": 936 + }, + { + "epoch": 0.04931578947368421, + "grad_norm": 1.40625, + "learning_rate": 9.863157894736842e-05, + "loss": 6.0392, + "step": 937 + }, + { + "epoch": 0.04936842105263158, + "grad_norm": 0.859375, + "learning_rate": 9.873684210526316e-05, + "loss": 6.3674, + "step": 938 + }, + { + "epoch": 0.04942105263157895, + "grad_norm": 1.0390625, + "learning_rate": 9.88421052631579e-05, + "loss": 6.5431, + "step": 939 + }, + { + "epoch": 0.049473684210526316, + "grad_norm": 1.046875, + "learning_rate": 9.894736842105263e-05, + "loss": 6.0435, + "step": 940 + }, + { + "epoch": 0.04952631578947368, + "grad_norm": 0.96875, + "learning_rate": 9.905263157894737e-05, + "loss": 6.1906, + "step": 941 + }, + { + "epoch": 0.049578947368421056, + "grad_norm": 0.90234375, + "learning_rate": 9.915789473684211e-05, + "loss": 6.2743, + "step": 942 + }, + { + "epoch": 0.04963157894736842, + "grad_norm": 0.86328125, + "learning_rate": 9.926315789473685e-05, + "loss": 6.2985, + "step": 943 + }, + { + "epoch": 0.04968421052631579, + "grad_norm": 0.97265625, + "learning_rate": 9.936842105263159e-05, + "loss": 5.977, + "step": 944 + }, + { + "epoch": 0.049736842105263156, + "grad_norm": 0.91796875, + "learning_rate": 9.947368421052632e-05, + "loss": 6.0319, + "step": 945 + }, + { + "epoch": 0.04978947368421053, + "grad_norm": 0.8671875, + "learning_rate": 9.957894736842106e-05, + "loss": 6.0619, + "step": 946 + }, + { + "epoch": 0.049842105263157896, + "grad_norm": 1.2578125, + "learning_rate": 9.968421052631578e-05, + "loss": 6.1903, + "step": 947 + }, + { + "epoch": 0.04989473684210526, + "grad_norm": 0.77734375, + "learning_rate": 9.978947368421054e-05, + "loss": 6.0686, + "step": 948 + }, + { + "epoch": 0.04994736842105263, + "grad_norm": 0.96875, + "learning_rate": 9.989473684210526e-05, + "loss": 6.081, + "step": 949 + }, + { + "epoch": 0.05, + "grad_norm": 1.125, + "learning_rate": 0.0001, + "loss": 6.3865, + "step": 950 + }, + { + "epoch": 0.05005263157894737, + "grad_norm": 0.9921875, + "learning_rate": 0.00010010526315789474, + "loss": 6.2237, + "step": 951 + }, + { + "epoch": 0.050105263157894736, + "grad_norm": 1.0390625, + "learning_rate": 0.00010021052631578949, + "loss": 6.2769, + "step": 952 + }, + { + "epoch": 0.0501578947368421, + "grad_norm": 1.046875, + "learning_rate": 0.00010031578947368421, + "loss": 6.1141, + "step": 953 + }, + { + "epoch": 0.050210526315789476, + "grad_norm": 0.7578125, + "learning_rate": 0.00010042105263157895, + "loss": 6.1481, + "step": 954 + }, + { + "epoch": 0.05026315789473684, + "grad_norm": 1.0078125, + "learning_rate": 0.00010052631578947369, + "loss": 6.2805, + "step": 955 + }, + { + "epoch": 0.05031578947368421, + "grad_norm": 1.0390625, + "learning_rate": 0.00010063157894736843, + "loss": 5.9349, + "step": 956 + }, + { + "epoch": 0.050368421052631576, + "grad_norm": 0.99609375, + "learning_rate": 0.00010073684210526314, + "loss": 6.049, + "step": 957 + }, + { + "epoch": 0.05042105263157895, + "grad_norm": 1.0234375, + "learning_rate": 0.0001008421052631579, + "loss": 6.3158, + "step": 958 + }, + { + "epoch": 0.050473684210526316, + "grad_norm": 0.953125, + "learning_rate": 0.00010094736842105264, + "loss": 6.48, + "step": 959 + }, + { + "epoch": 0.05052631578947368, + "grad_norm": 1.125, + "learning_rate": 0.00010105263157894738, + "loss": 6.011, + "step": 960 + }, + { + "epoch": 0.05057894736842105, + "grad_norm": 0.89453125, + "learning_rate": 0.00010115789473684212, + "loss": 5.9801, + "step": 961 + }, + { + "epoch": 0.05063157894736842, + "grad_norm": 0.93359375, + "learning_rate": 0.00010126315789473683, + "loss": 6.166, + "step": 962 + }, + { + "epoch": 0.05068421052631579, + "grad_norm": 1.046875, + "learning_rate": 0.00010136842105263159, + "loss": 5.6631, + "step": 963 + }, + { + "epoch": 0.05073684210526316, + "grad_norm": 0.93359375, + "learning_rate": 0.00010147368421052633, + "loss": 6.094, + "step": 964 + }, + { + "epoch": 0.05078947368421052, + "grad_norm": 0.91015625, + "learning_rate": 0.00010157894736842107, + "loss": 6.2138, + "step": 965 + }, + { + "epoch": 0.0508421052631579, + "grad_norm": 1.296875, + "learning_rate": 0.0001016842105263158, + "loss": 6.1788, + "step": 966 + }, + { + "epoch": 0.05089473684210526, + "grad_norm": 1.15625, + "learning_rate": 0.00010178947368421052, + "loss": 5.9861, + "step": 967 + }, + { + "epoch": 0.05094736842105263, + "grad_norm": 1.0078125, + "learning_rate": 0.00010189473684210526, + "loss": 6.0874, + "step": 968 + }, + { + "epoch": 0.051, + "grad_norm": 1.0546875, + "learning_rate": 0.00010200000000000001, + "loss": 6.2731, + "step": 969 + }, + { + "epoch": 0.05105263157894737, + "grad_norm": 1.390625, + "learning_rate": 0.00010210526315789475, + "loss": 6.3605, + "step": 970 + }, + { + "epoch": 0.05110526315789474, + "grad_norm": 1.2890625, + "learning_rate": 0.00010221052631578947, + "loss": 5.772, + "step": 971 + }, + { + "epoch": 0.051157894736842104, + "grad_norm": 0.82421875, + "learning_rate": 0.0001023157894736842, + "loss": 6.0918, + "step": 972 + }, + { + "epoch": 0.05121052631578947, + "grad_norm": 0.80078125, + "learning_rate": 0.00010242105263157895, + "loss": 6.1235, + "step": 973 + }, + { + "epoch": 0.051263157894736844, + "grad_norm": 0.8671875, + "learning_rate": 0.00010252631578947369, + "loss": 6.0478, + "step": 974 + }, + { + "epoch": 0.05131578947368421, + "grad_norm": 0.79296875, + "learning_rate": 0.00010263157894736844, + "loss": 6.2327, + "step": 975 + }, + { + "epoch": 0.05136842105263158, + "grad_norm": 1.0078125, + "learning_rate": 0.00010273684210526315, + "loss": 6.2677, + "step": 976 + }, + { + "epoch": 0.05142105263157895, + "grad_norm": 1.1015625, + "learning_rate": 0.00010284210526315789, + "loss": 6.4946, + "step": 977 + }, + { + "epoch": 0.05147368421052632, + "grad_norm": 1.03125, + "learning_rate": 0.00010294736842105263, + "loss": 6.2424, + "step": 978 + }, + { + "epoch": 0.051526315789473684, + "grad_norm": 0.8203125, + "learning_rate": 0.00010305263157894737, + "loss": 6.289, + "step": 979 + }, + { + "epoch": 0.05157894736842105, + "grad_norm": 1.15625, + "learning_rate": 0.00010315789473684211, + "loss": 5.9867, + "step": 980 + }, + { + "epoch": 0.051631578947368424, + "grad_norm": 0.86328125, + "learning_rate": 0.00010326315789473684, + "loss": 6.1909, + "step": 981 + }, + { + "epoch": 0.05168421052631579, + "grad_norm": 0.8828125, + "learning_rate": 0.00010336842105263158, + "loss": 6.2827, + "step": 982 + }, + { + "epoch": 0.05173684210526316, + "grad_norm": 1.2109375, + "learning_rate": 0.00010347368421052632, + "loss": 6.2896, + "step": 983 + }, + { + "epoch": 0.051789473684210524, + "grad_norm": 0.8515625, + "learning_rate": 0.00010357894736842106, + "loss": 6.2736, + "step": 984 + }, + { + "epoch": 0.0518421052631579, + "grad_norm": 1.0234375, + "learning_rate": 0.00010368421052631579, + "loss": 6.3027, + "step": 985 + }, + { + "epoch": 0.051894736842105264, + "grad_norm": 1.125, + "learning_rate": 0.00010378947368421053, + "loss": 6.1072, + "step": 986 + }, + { + "epoch": 0.05194736842105263, + "grad_norm": 0.91015625, + "learning_rate": 0.00010389473684210527, + "loss": 6.0225, + "step": 987 + }, + { + "epoch": 0.052, + "grad_norm": 1.0625, + "learning_rate": 0.00010400000000000001, + "loss": 6.025, + "step": 988 + }, + { + "epoch": 0.05205263157894737, + "grad_norm": 0.96484375, + "learning_rate": 0.00010410526315789475, + "loss": 5.9979, + "step": 989 + }, + { + "epoch": 0.05210526315789474, + "grad_norm": 1.390625, + "learning_rate": 0.00010421052631578947, + "loss": 6.079, + "step": 990 + }, + { + "epoch": 0.052157894736842104, + "grad_norm": 0.859375, + "learning_rate": 0.00010431578947368421, + "loss": 6.2242, + "step": 991 + }, + { + "epoch": 0.05221052631578947, + "grad_norm": 0.96875, + "learning_rate": 0.00010442105263157895, + "loss": 6.238, + "step": 992 + }, + { + "epoch": 0.052263157894736845, + "grad_norm": 0.84375, + "learning_rate": 0.0001045263157894737, + "loss": 5.9196, + "step": 993 + }, + { + "epoch": 0.05231578947368421, + "grad_norm": 0.8671875, + "learning_rate": 0.00010463157894736843, + "loss": 6.1262, + "step": 994 + }, + { + "epoch": 0.05236842105263158, + "grad_norm": 1.0625, + "learning_rate": 0.00010473684210526316, + "loss": 6.2667, + "step": 995 + }, + { + "epoch": 0.052421052631578945, + "grad_norm": 0.75390625, + "learning_rate": 0.0001048421052631579, + "loss": 6.242, + "step": 996 + }, + { + "epoch": 0.05247368421052632, + "grad_norm": 1.15625, + "learning_rate": 0.00010494736842105264, + "loss": 6.229, + "step": 997 + }, + { + "epoch": 0.052526315789473685, + "grad_norm": 1.0234375, + "learning_rate": 0.00010505263157894738, + "loss": 5.8927, + "step": 998 + }, + { + "epoch": 0.05257894736842105, + "grad_norm": 1.1875, + "learning_rate": 0.0001051578947368421, + "loss": 5.9742, + "step": 999 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 0.72265625, + "learning_rate": 0.00010526315789473685, + "loss": 6.0916, + "step": 1000 + }, + { + "epoch": 0.05268421052631579, + "grad_norm": 0.7734375, + "learning_rate": 0.00010536842105263159, + "loss": 5.8936, + "step": 1001 + }, + { + "epoch": 0.05273684210526316, + "grad_norm": 0.73828125, + "learning_rate": 0.00010547368421052633, + "loss": 6.3795, + "step": 1002 + }, + { + "epoch": 0.052789473684210525, + "grad_norm": 1.0703125, + "learning_rate": 0.00010557894736842107, + "loss": 5.9821, + "step": 1003 + }, + { + "epoch": 0.05284210526315789, + "grad_norm": 1.2734375, + "learning_rate": 0.00010568421052631578, + "loss": 6.1901, + "step": 1004 + }, + { + "epoch": 0.052894736842105265, + "grad_norm": 1.296875, + "learning_rate": 0.00010578947368421052, + "loss": 6.0374, + "step": 1005 + }, + { + "epoch": 0.05294736842105263, + "grad_norm": 0.99609375, + "learning_rate": 0.00010589473684210527, + "loss": 6.1129, + "step": 1006 + }, + { + "epoch": 0.053, + "grad_norm": 0.91015625, + "learning_rate": 0.00010600000000000002, + "loss": 6.0886, + "step": 1007 + }, + { + "epoch": 0.053052631578947365, + "grad_norm": 1.0390625, + "learning_rate": 0.00010610526315789476, + "loss": 6.2423, + "step": 1008 + }, + { + "epoch": 0.05310526315789474, + "grad_norm": 1.1171875, + "learning_rate": 0.00010621052631578947, + "loss": 5.4549, + "step": 1009 + }, + { + "epoch": 0.053157894736842105, + "grad_norm": 0.73828125, + "learning_rate": 0.00010631578947368421, + "loss": 6.0894, + "step": 1010 + }, + { + "epoch": 0.05321052631578947, + "grad_norm": 0.875, + "learning_rate": 0.00010642105263157895, + "loss": 6.2518, + "step": 1011 + }, + { + "epoch": 0.05326315789473684, + "grad_norm": 1.140625, + "learning_rate": 0.0001065263157894737, + "loss": 6.1986, + "step": 1012 + }, + { + "epoch": 0.05331578947368421, + "grad_norm": 0.765625, + "learning_rate": 0.00010663157894736842, + "loss": 6.187, + "step": 1013 + }, + { + "epoch": 0.05336842105263158, + "grad_norm": 1.171875, + "learning_rate": 0.00010673684210526316, + "loss": 5.8875, + "step": 1014 + }, + { + "epoch": 0.053421052631578946, + "grad_norm": 1.09375, + "learning_rate": 0.0001068421052631579, + "loss": 6.1521, + "step": 1015 + }, + { + "epoch": 0.05347368421052632, + "grad_norm": 0.94921875, + "learning_rate": 0.00010694736842105264, + "loss": 6.1439, + "step": 1016 + }, + { + "epoch": 0.053526315789473686, + "grad_norm": 0.95703125, + "learning_rate": 0.00010705263157894738, + "loss": 5.876, + "step": 1017 + }, + { + "epoch": 0.05357894736842105, + "grad_norm": 1.1640625, + "learning_rate": 0.0001071578947368421, + "loss": 6.3901, + "step": 1018 + }, + { + "epoch": 0.05363157894736842, + "grad_norm": 0.8984375, + "learning_rate": 0.00010726315789473684, + "loss": 6.3401, + "step": 1019 + }, + { + "epoch": 0.05368421052631579, + "grad_norm": 0.98046875, + "learning_rate": 0.00010736842105263158, + "loss": 6.1578, + "step": 1020 + }, + { + "epoch": 0.05373684210526316, + "grad_norm": 0.90625, + "learning_rate": 0.00010747368421052632, + "loss": 5.8786, + "step": 1021 + }, + { + "epoch": 0.053789473684210526, + "grad_norm": 1.1640625, + "learning_rate": 0.00010757894736842106, + "loss": 6.0872, + "step": 1022 + }, + { + "epoch": 0.05384210526315789, + "grad_norm": 0.8671875, + "learning_rate": 0.00010768421052631579, + "loss": 6.2519, + "step": 1023 + }, + { + "epoch": 0.053894736842105266, + "grad_norm": 1.296875, + "learning_rate": 0.00010778947368421053, + "loss": 5.9535, + "step": 1024 + }, + { + "epoch": 0.05394736842105263, + "grad_norm": 1.015625, + "learning_rate": 0.00010789473684210527, + "loss": 6.2176, + "step": 1025 + }, + { + "epoch": 0.054, + "grad_norm": 0.69921875, + "learning_rate": 0.00010800000000000001, + "loss": 6.2962, + "step": 1026 + }, + { + "epoch": 0.054052631578947366, + "grad_norm": 1.109375, + "learning_rate": 0.00010810526315789474, + "loss": 5.8289, + "step": 1027 + }, + { + "epoch": 0.05410526315789474, + "grad_norm": 1.09375, + "learning_rate": 0.00010821052631578948, + "loss": 6.2592, + "step": 1028 + }, + { + "epoch": 0.054157894736842106, + "grad_norm": 1.109375, + "learning_rate": 0.00010831578947368422, + "loss": 6.0032, + "step": 1029 + }, + { + "epoch": 0.05421052631578947, + "grad_norm": 1.1484375, + "learning_rate": 0.00010842105263157896, + "loss": 6.1465, + "step": 1030 + }, + { + "epoch": 0.05426315789473684, + "grad_norm": 1.1875, + "learning_rate": 0.0001085263157894737, + "loss": 6.0221, + "step": 1031 + }, + { + "epoch": 0.05431578947368421, + "grad_norm": 0.92578125, + "learning_rate": 0.00010863157894736842, + "loss": 6.1251, + "step": 1032 + }, + { + "epoch": 0.05436842105263158, + "grad_norm": 1.0625, + "learning_rate": 0.00010873684210526316, + "loss": 6.1643, + "step": 1033 + }, + { + "epoch": 0.054421052631578946, + "grad_norm": 1.125, + "learning_rate": 0.0001088421052631579, + "loss": 5.9446, + "step": 1034 + }, + { + "epoch": 0.05447368421052631, + "grad_norm": 0.97265625, + "learning_rate": 0.00010894736842105264, + "loss": 5.7471, + "step": 1035 + }, + { + "epoch": 0.05452631578947369, + "grad_norm": 1.0234375, + "learning_rate": 0.00010905263157894738, + "loss": 5.7865, + "step": 1036 + }, + { + "epoch": 0.05457894736842105, + "grad_norm": 1.03125, + "learning_rate": 0.00010915789473684211, + "loss": 6.1126, + "step": 1037 + }, + { + "epoch": 0.05463157894736842, + "grad_norm": 0.75, + "learning_rate": 0.00010926315789473685, + "loss": 6.181, + "step": 1038 + }, + { + "epoch": 0.05468421052631579, + "grad_norm": 0.8984375, + "learning_rate": 0.00010936842105263159, + "loss": 6.2716, + "step": 1039 + }, + { + "epoch": 0.05473684210526316, + "grad_norm": 1.078125, + "learning_rate": 0.00010947368421052633, + "loss": 5.9006, + "step": 1040 + }, + { + "epoch": 0.05478947368421053, + "grad_norm": 0.83984375, + "learning_rate": 0.00010957894736842104, + "loss": 6.0907, + "step": 1041 + }, + { + "epoch": 0.05484210526315789, + "grad_norm": 0.95703125, + "learning_rate": 0.00010968421052631578, + "loss": 5.7861, + "step": 1042 + }, + { + "epoch": 0.05489473684210526, + "grad_norm": 1.125, + "learning_rate": 0.00010978947368421054, + "loss": 6.0963, + "step": 1043 + }, + { + "epoch": 0.054947368421052634, + "grad_norm": 1.171875, + "learning_rate": 0.00010989473684210528, + "loss": 6.2145, + "step": 1044 + }, + { + "epoch": 0.055, + "grad_norm": 0.83203125, + "learning_rate": 0.00011000000000000002, + "loss": 6.4214, + "step": 1045 + }, + { + "epoch": 0.05505263157894737, + "grad_norm": 1.21875, + "learning_rate": 0.00011010526315789473, + "loss": 5.7887, + "step": 1046 + }, + { + "epoch": 0.055105263157894734, + "grad_norm": 0.8359375, + "learning_rate": 0.00011021052631578947, + "loss": 6.1903, + "step": 1047 + }, + { + "epoch": 0.05515789473684211, + "grad_norm": 1.8125, + "learning_rate": 0.00011031578947368421, + "loss": 5.988, + "step": 1048 + }, + { + "epoch": 0.055210526315789474, + "grad_norm": 0.96484375, + "learning_rate": 0.00011042105263157896, + "loss": 6.0632, + "step": 1049 + }, + { + "epoch": 0.05526315789473684, + "grad_norm": 0.82421875, + "learning_rate": 0.0001105263157894737, + "loss": 6.0995, + "step": 1050 + }, + { + "epoch": 0.05531578947368421, + "grad_norm": 0.9375, + "learning_rate": 0.00011063157894736842, + "loss": 6.1939, + "step": 1051 + }, + { + "epoch": 0.05536842105263158, + "grad_norm": 1.1015625, + "learning_rate": 0.00011073684210526316, + "loss": 5.8412, + "step": 1052 + }, + { + "epoch": 0.05542105263157895, + "grad_norm": 0.86328125, + "learning_rate": 0.0001108421052631579, + "loss": 6.1686, + "step": 1053 + }, + { + "epoch": 0.055473684210526314, + "grad_norm": 1.265625, + "learning_rate": 0.00011094736842105264, + "loss": 5.8319, + "step": 1054 + }, + { + "epoch": 0.05552631578947369, + "grad_norm": 0.765625, + "learning_rate": 0.00011105263157894736, + "loss": 5.9418, + "step": 1055 + }, + { + "epoch": 0.055578947368421054, + "grad_norm": 1.0546875, + "learning_rate": 0.0001111578947368421, + "loss": 6.2675, + "step": 1056 + }, + { + "epoch": 0.05563157894736842, + "grad_norm": 0.98828125, + "learning_rate": 0.00011126315789473684, + "loss": 6.0157, + "step": 1057 + }, + { + "epoch": 0.05568421052631579, + "grad_norm": 1.2578125, + "learning_rate": 0.00011136842105263158, + "loss": 5.7894, + "step": 1058 + }, + { + "epoch": 0.05573684210526316, + "grad_norm": 0.93359375, + "learning_rate": 0.00011147368421052632, + "loss": 5.672, + "step": 1059 + }, + { + "epoch": 0.05578947368421053, + "grad_norm": 1.015625, + "learning_rate": 0.00011157894736842105, + "loss": 6.1491, + "step": 1060 + }, + { + "epoch": 0.055842105263157894, + "grad_norm": 0.9609375, + "learning_rate": 0.00011168421052631579, + "loss": 5.9435, + "step": 1061 + }, + { + "epoch": 0.05589473684210526, + "grad_norm": 0.91015625, + "learning_rate": 0.00011178947368421053, + "loss": 5.9086, + "step": 1062 + }, + { + "epoch": 0.055947368421052635, + "grad_norm": 0.98828125, + "learning_rate": 0.00011189473684210527, + "loss": 6.1632, + "step": 1063 + }, + { + "epoch": 0.056, + "grad_norm": 0.734375, + "learning_rate": 0.00011200000000000001, + "loss": 6.0847, + "step": 1064 + }, + { + "epoch": 0.05605263157894737, + "grad_norm": 1.140625, + "learning_rate": 0.00011210526315789474, + "loss": 5.9541, + "step": 1065 + }, + { + "epoch": 0.056105263157894734, + "grad_norm": 0.9453125, + "learning_rate": 0.00011221052631578948, + "loss": 6.2405, + "step": 1066 + }, + { + "epoch": 0.05615789473684211, + "grad_norm": 1.125, + "learning_rate": 0.00011231578947368422, + "loss": 6.1319, + "step": 1067 + }, + { + "epoch": 0.056210526315789475, + "grad_norm": 0.87890625, + "learning_rate": 0.00011242105263157896, + "loss": 5.9425, + "step": 1068 + }, + { + "epoch": 0.05626315789473684, + "grad_norm": 1.0859375, + "learning_rate": 0.00011252631578947369, + "loss": 5.7247, + "step": 1069 + }, + { + "epoch": 0.05631578947368421, + "grad_norm": 0.81640625, + "learning_rate": 0.00011263157894736843, + "loss": 6.3602, + "step": 1070 + }, + { + "epoch": 0.05636842105263158, + "grad_norm": 1.0625, + "learning_rate": 0.00011273684210526317, + "loss": 5.9757, + "step": 1071 + }, + { + "epoch": 0.05642105263157895, + "grad_norm": 1.1640625, + "learning_rate": 0.0001128421052631579, + "loss": 6.3731, + "step": 1072 + }, + { + "epoch": 0.056473684210526315, + "grad_norm": 1.2265625, + "learning_rate": 0.00011294736842105265, + "loss": 6.1435, + "step": 1073 + }, + { + "epoch": 0.05652631578947368, + "grad_norm": 0.9765625, + "learning_rate": 0.00011305263157894736, + "loss": 5.7906, + "step": 1074 + }, + { + "epoch": 0.056578947368421055, + "grad_norm": 1.125, + "learning_rate": 0.00011315789473684211, + "loss": 5.892, + "step": 1075 + }, + { + "epoch": 0.05663157894736842, + "grad_norm": 0.859375, + "learning_rate": 0.00011326315789473685, + "loss": 6.0686, + "step": 1076 + }, + { + "epoch": 0.05668421052631579, + "grad_norm": 1.0234375, + "learning_rate": 0.00011336842105263159, + "loss": 6.3797, + "step": 1077 + }, + { + "epoch": 0.056736842105263155, + "grad_norm": 1.046875, + "learning_rate": 0.00011347368421052633, + "loss": 5.9231, + "step": 1078 + }, + { + "epoch": 0.05678947368421053, + "grad_norm": 1.0625, + "learning_rate": 0.00011357894736842105, + "loss": 6.1592, + "step": 1079 + }, + { + "epoch": 0.056842105263157895, + "grad_norm": 0.9296875, + "learning_rate": 0.0001136842105263158, + "loss": 5.6817, + "step": 1080 + }, + { + "epoch": 0.05689473684210526, + "grad_norm": 0.8125, + "learning_rate": 0.00011378947368421054, + "loss": 6.0245, + "step": 1081 + }, + { + "epoch": 0.05694736842105263, + "grad_norm": 1.140625, + "learning_rate": 0.00011389473684210528, + "loss": 6.2048, + "step": 1082 + }, + { + "epoch": 0.057, + "grad_norm": 0.8046875, + "learning_rate": 0.00011399999999999999, + "loss": 5.873, + "step": 1083 + }, + { + "epoch": 0.05705263157894737, + "grad_norm": 0.984375, + "learning_rate": 0.00011410526315789473, + "loss": 6.1302, + "step": 1084 + }, + { + "epoch": 0.057105263157894735, + "grad_norm": 0.81640625, + "learning_rate": 0.00011421052631578947, + "loss": 6.0411, + "step": 1085 + }, + { + "epoch": 0.0571578947368421, + "grad_norm": 0.72265625, + "learning_rate": 0.00011431578947368423, + "loss": 5.8797, + "step": 1086 + }, + { + "epoch": 0.057210526315789476, + "grad_norm": 0.8515625, + "learning_rate": 0.00011442105263157897, + "loss": 5.8255, + "step": 1087 + }, + { + "epoch": 0.05726315789473684, + "grad_norm": 0.7578125, + "learning_rate": 0.00011452631578947368, + "loss": 5.9739, + "step": 1088 + }, + { + "epoch": 0.05731578947368421, + "grad_norm": 0.75390625, + "learning_rate": 0.00011463157894736842, + "loss": 6.034, + "step": 1089 + }, + { + "epoch": 0.057368421052631575, + "grad_norm": 0.86328125, + "learning_rate": 0.00011473684210526316, + "loss": 5.8567, + "step": 1090 + }, + { + "epoch": 0.05742105263157895, + "grad_norm": 1.109375, + "learning_rate": 0.0001148421052631579, + "loss": 6.0606, + "step": 1091 + }, + { + "epoch": 0.057473684210526316, + "grad_norm": 1.140625, + "learning_rate": 0.00011494736842105265, + "loss": 6.2965, + "step": 1092 + }, + { + "epoch": 0.05752631578947368, + "grad_norm": 1.0390625, + "learning_rate": 0.00011505263157894737, + "loss": 6.2964, + "step": 1093 + }, + { + "epoch": 0.057578947368421056, + "grad_norm": 1.078125, + "learning_rate": 0.00011515789473684211, + "loss": 5.9438, + "step": 1094 + }, + { + "epoch": 0.05763157894736842, + "grad_norm": 0.859375, + "learning_rate": 0.00011526315789473685, + "loss": 5.9728, + "step": 1095 + }, + { + "epoch": 0.05768421052631579, + "grad_norm": 1.0859375, + "learning_rate": 0.00011536842105263159, + "loss": 5.9805, + "step": 1096 + }, + { + "epoch": 0.057736842105263156, + "grad_norm": 1.0546875, + "learning_rate": 0.00011547368421052631, + "loss": 5.6844, + "step": 1097 + }, + { + "epoch": 0.05778947368421053, + "grad_norm": 0.97265625, + "learning_rate": 0.00011557894736842105, + "loss": 6.1539, + "step": 1098 + }, + { + "epoch": 0.057842105263157896, + "grad_norm": 1.140625, + "learning_rate": 0.0001156842105263158, + "loss": 5.8332, + "step": 1099 + }, + { + "epoch": 0.05789473684210526, + "grad_norm": 0.875, + "learning_rate": 0.00011578947368421053, + "loss": 5.8662, + "step": 1100 + }, + { + "epoch": 0.05794736842105263, + "grad_norm": 1.0078125, + "learning_rate": 0.00011589473684210527, + "loss": 6.0683, + "step": 1101 + }, + { + "epoch": 0.058, + "grad_norm": 1.109375, + "learning_rate": 0.000116, + "loss": 5.9388, + "step": 1102 + }, + { + "epoch": 0.05805263157894737, + "grad_norm": 0.94921875, + "learning_rate": 0.00011610526315789474, + "loss": 5.8895, + "step": 1103 + }, + { + "epoch": 0.058105263157894736, + "grad_norm": 1.203125, + "learning_rate": 0.00011621052631578948, + "loss": 5.929, + "step": 1104 + }, + { + "epoch": 0.0581578947368421, + "grad_norm": 0.921875, + "learning_rate": 0.00011631578947368422, + "loss": 5.9835, + "step": 1105 + }, + { + "epoch": 0.058210526315789476, + "grad_norm": 0.9609375, + "learning_rate": 0.00011642105263157896, + "loss": 6.4084, + "step": 1106 + }, + { + "epoch": 0.05826315789473684, + "grad_norm": 0.9609375, + "learning_rate": 0.00011652631578947369, + "loss": 6.0482, + "step": 1107 + }, + { + "epoch": 0.05831578947368421, + "grad_norm": 0.80859375, + "learning_rate": 0.00011663157894736843, + "loss": 6.1392, + "step": 1108 + }, + { + "epoch": 0.058368421052631576, + "grad_norm": 1.1015625, + "learning_rate": 0.00011673684210526317, + "loss": 5.7017, + "step": 1109 + }, + { + "epoch": 0.05842105263157895, + "grad_norm": 0.82421875, + "learning_rate": 0.00011684210526315791, + "loss": 5.9223, + "step": 1110 + }, + { + "epoch": 0.05847368421052632, + "grad_norm": 0.890625, + "learning_rate": 0.00011694736842105262, + "loss": 6.1614, + "step": 1111 + }, + { + "epoch": 0.05852631578947368, + "grad_norm": 0.95703125, + "learning_rate": 0.00011705263157894737, + "loss": 6.1275, + "step": 1112 + }, + { + "epoch": 0.05857894736842105, + "grad_norm": 1.0625, + "learning_rate": 0.00011715789473684211, + "loss": 5.8907, + "step": 1113 + }, + { + "epoch": 0.05863157894736842, + "grad_norm": 0.9609375, + "learning_rate": 0.00011726315789473685, + "loss": 5.7079, + "step": 1114 + }, + { + "epoch": 0.05868421052631579, + "grad_norm": 0.6875, + "learning_rate": 0.0001173684210526316, + "loss": 6.0685, + "step": 1115 + }, + { + "epoch": 0.05873684210526316, + "grad_norm": 0.86328125, + "learning_rate": 0.00011747368421052631, + "loss": 6.1132, + "step": 1116 + }, + { + "epoch": 0.05878947368421052, + "grad_norm": 1.2109375, + "learning_rate": 0.00011757894736842106, + "loss": 6.2414, + "step": 1117 + }, + { + "epoch": 0.0588421052631579, + "grad_norm": 1.109375, + "learning_rate": 0.0001176842105263158, + "loss": 5.9302, + "step": 1118 + }, + { + "epoch": 0.058894736842105264, + "grad_norm": 0.9921875, + "learning_rate": 0.00011778947368421054, + "loss": 5.5207, + "step": 1119 + }, + { + "epoch": 0.05894736842105263, + "grad_norm": 0.8203125, + "learning_rate": 0.00011789473684210525, + "loss": 6.0358, + "step": 1120 + }, + { + "epoch": 0.059, + "grad_norm": 0.85546875, + "learning_rate": 0.000118, + "loss": 6.1205, + "step": 1121 + }, + { + "epoch": 0.05905263157894737, + "grad_norm": 1.015625, + "learning_rate": 0.00011810526315789474, + "loss": 5.9491, + "step": 1122 + }, + { + "epoch": 0.05910526315789474, + "grad_norm": 0.7578125, + "learning_rate": 0.00011821052631578949, + "loss": 6.1786, + "step": 1123 + }, + { + "epoch": 0.059157894736842104, + "grad_norm": 0.921875, + "learning_rate": 0.00011831578947368423, + "loss": 6.177, + "step": 1124 + }, + { + "epoch": 0.05921052631578947, + "grad_norm": 0.796875, + "learning_rate": 0.00011842105263157894, + "loss": 6.1656, + "step": 1125 + }, + { + "epoch": 0.059263157894736844, + "grad_norm": 0.9609375, + "learning_rate": 0.00011852631578947368, + "loss": 5.917, + "step": 1126 + }, + { + "epoch": 0.05931578947368421, + "grad_norm": 1.078125, + "learning_rate": 0.00011863157894736842, + "loss": 5.946, + "step": 1127 + }, + { + "epoch": 0.05936842105263158, + "grad_norm": 0.72265625, + "learning_rate": 0.00011873684210526316, + "loss": 6.2335, + "step": 1128 + }, + { + "epoch": 0.059421052631578944, + "grad_norm": 0.875, + "learning_rate": 0.00011884210526315792, + "loss": 6.2321, + "step": 1129 + }, + { + "epoch": 0.05947368421052632, + "grad_norm": 1.125, + "learning_rate": 0.00011894736842105263, + "loss": 5.6849, + "step": 1130 + }, + { + "epoch": 0.059526315789473684, + "grad_norm": 0.75, + "learning_rate": 0.00011905263157894737, + "loss": 6.332, + "step": 1131 + }, + { + "epoch": 0.05957894736842105, + "grad_norm": 0.97265625, + "learning_rate": 0.00011915789473684211, + "loss": 5.8335, + "step": 1132 + }, + { + "epoch": 0.059631578947368424, + "grad_norm": 1.3671875, + "learning_rate": 0.00011926315789473685, + "loss": 6.2205, + "step": 1133 + }, + { + "epoch": 0.05968421052631579, + "grad_norm": 0.8359375, + "learning_rate": 0.00011936842105263158, + "loss": 6.1915, + "step": 1134 + }, + { + "epoch": 0.05973684210526316, + "grad_norm": 0.7421875, + "learning_rate": 0.00011947368421052632, + "loss": 6.1426, + "step": 1135 + }, + { + "epoch": 0.059789473684210524, + "grad_norm": 0.875, + "learning_rate": 0.00011957894736842106, + "loss": 6.1404, + "step": 1136 + }, + { + "epoch": 0.0598421052631579, + "grad_norm": 1.3515625, + "learning_rate": 0.0001196842105263158, + "loss": 6.5503, + "step": 1137 + }, + { + "epoch": 0.059894736842105264, + "grad_norm": 0.61328125, + "learning_rate": 0.00011978947368421054, + "loss": 6.111, + "step": 1138 + }, + { + "epoch": 0.05994736842105263, + "grad_norm": 0.72265625, + "learning_rate": 0.00011989473684210526, + "loss": 6.3245, + "step": 1139 + }, + { + "epoch": 0.06, + "grad_norm": 1.171875, + "learning_rate": 0.00012, + "loss": 6.3703, + "step": 1140 + }, + { + "epoch": 0.06005263157894737, + "grad_norm": 0.91796875, + "learning_rate": 0.00012010526315789474, + "loss": 5.7952, + "step": 1141 + }, + { + "epoch": 0.06010526315789474, + "grad_norm": 0.94140625, + "learning_rate": 0.00012021052631578948, + "loss": 6.2602, + "step": 1142 + }, + { + "epoch": 0.060157894736842105, + "grad_norm": 1.171875, + "learning_rate": 0.00012031578947368422, + "loss": 6.2352, + "step": 1143 + }, + { + "epoch": 0.06021052631578947, + "grad_norm": 1.46875, + "learning_rate": 0.00012042105263157895, + "loss": 5.8399, + "step": 1144 + }, + { + "epoch": 0.060263157894736845, + "grad_norm": 1.3828125, + "learning_rate": 0.00012052631578947369, + "loss": 5.4318, + "step": 1145 + }, + { + "epoch": 0.06031578947368421, + "grad_norm": 0.65234375, + "learning_rate": 0.00012063157894736843, + "loss": 5.999, + "step": 1146 + }, + { + "epoch": 0.06036842105263158, + "grad_norm": 0.94921875, + "learning_rate": 0.00012073684210526317, + "loss": 6.0046, + "step": 1147 + }, + { + "epoch": 0.060421052631578945, + "grad_norm": 0.8671875, + "learning_rate": 0.00012084210526315788, + "loss": 5.9908, + "step": 1148 + }, + { + "epoch": 0.06047368421052632, + "grad_norm": 0.828125, + "learning_rate": 0.00012094736842105264, + "loss": 5.7976, + "step": 1149 + }, + { + "epoch": 0.060526315789473685, + "grad_norm": 0.984375, + "learning_rate": 0.00012105263157894738, + "loss": 6.1416, + "step": 1150 + }, + { + "epoch": 0.06057894736842105, + "grad_norm": 1.0625, + "learning_rate": 0.00012115789473684212, + "loss": 6.227, + "step": 1151 + }, + { + "epoch": 0.06063157894736842, + "grad_norm": 1.1953125, + "learning_rate": 0.00012126315789473686, + "loss": 5.9803, + "step": 1152 + }, + { + "epoch": 0.06068421052631579, + "grad_norm": 0.921875, + "learning_rate": 0.00012136842105263157, + "loss": 5.8437, + "step": 1153 + }, + { + "epoch": 0.06073684210526316, + "grad_norm": 0.94921875, + "learning_rate": 0.00012147368421052632, + "loss": 5.9635, + "step": 1154 + }, + { + "epoch": 0.060789473684210525, + "grad_norm": 1.1875, + "learning_rate": 0.00012157894736842106, + "loss": 6.0895, + "step": 1155 + }, + { + "epoch": 0.06084210526315789, + "grad_norm": 1.0078125, + "learning_rate": 0.0001216842105263158, + "loss": 6.0517, + "step": 1156 + }, + { + "epoch": 0.060894736842105265, + "grad_norm": 1.0703125, + "learning_rate": 0.00012178947368421054, + "loss": 6.0795, + "step": 1157 + }, + { + "epoch": 0.06094736842105263, + "grad_norm": 0.89453125, + "learning_rate": 0.00012189473684210526, + "loss": 6.1234, + "step": 1158 + }, + { + "epoch": 0.061, + "grad_norm": 0.84765625, + "learning_rate": 0.000122, + "loss": 6.0547, + "step": 1159 + }, + { + "epoch": 0.061052631578947365, + "grad_norm": 0.87890625, + "learning_rate": 0.00012210526315789474, + "loss": 6.1055, + "step": 1160 + }, + { + "epoch": 0.06110526315789474, + "grad_norm": 0.953125, + "learning_rate": 0.0001222105263157895, + "loss": 6.0545, + "step": 1161 + }, + { + "epoch": 0.061157894736842106, + "grad_norm": 0.8046875, + "learning_rate": 0.00012231578947368422, + "loss": 6.2769, + "step": 1162 + }, + { + "epoch": 0.06121052631578947, + "grad_norm": 0.921875, + "learning_rate": 0.00012242105263157894, + "loss": 6.3541, + "step": 1163 + }, + { + "epoch": 0.06126315789473684, + "grad_norm": 0.81640625, + "learning_rate": 0.0001225263157894737, + "loss": 5.9833, + "step": 1164 + }, + { + "epoch": 0.06131578947368421, + "grad_norm": 0.84375, + "learning_rate": 0.00012263157894736842, + "loss": 5.6965, + "step": 1165 + }, + { + "epoch": 0.06136842105263158, + "grad_norm": 0.97265625, + "learning_rate": 0.00012273684210526318, + "loss": 5.9834, + "step": 1166 + }, + { + "epoch": 0.061421052631578946, + "grad_norm": 0.8359375, + "learning_rate": 0.0001228421052631579, + "loss": 6.0602, + "step": 1167 + }, + { + "epoch": 0.06147368421052631, + "grad_norm": 1.0, + "learning_rate": 0.00012294736842105263, + "loss": 6.1712, + "step": 1168 + }, + { + "epoch": 0.061526315789473686, + "grad_norm": 0.984375, + "learning_rate": 0.00012305263157894739, + "loss": 6.0671, + "step": 1169 + }, + { + "epoch": 0.06157894736842105, + "grad_norm": 1.0390625, + "learning_rate": 0.0001231578947368421, + "loss": 5.9311, + "step": 1170 + }, + { + "epoch": 0.06163157894736842, + "grad_norm": 0.9375, + "learning_rate": 0.00012326315789473687, + "loss": 6.0635, + "step": 1171 + }, + { + "epoch": 0.06168421052631579, + "grad_norm": 1.03125, + "learning_rate": 0.00012336842105263156, + "loss": 5.6215, + "step": 1172 + }, + { + "epoch": 0.06173684210526316, + "grad_norm": 1.1640625, + "learning_rate": 0.00012347368421052632, + "loss": 6.0751, + "step": 1173 + }, + { + "epoch": 0.061789473684210526, + "grad_norm": 0.921875, + "learning_rate": 0.00012357894736842107, + "loss": 5.9024, + "step": 1174 + }, + { + "epoch": 0.06184210526315789, + "grad_norm": 0.9765625, + "learning_rate": 0.0001236842105263158, + "loss": 5.9141, + "step": 1175 + }, + { + "epoch": 0.061894736842105266, + "grad_norm": 1.1171875, + "learning_rate": 0.00012378947368421053, + "loss": 6.1745, + "step": 1176 + }, + { + "epoch": 0.06194736842105263, + "grad_norm": 0.8984375, + "learning_rate": 0.00012389473684210525, + "loss": 6.2833, + "step": 1177 + }, + { + "epoch": 0.062, + "grad_norm": 0.92578125, + "learning_rate": 0.000124, + "loss": 5.697, + "step": 1178 + }, + { + "epoch": 0.062052631578947366, + "grad_norm": 0.91015625, + "learning_rate": 0.00012410526315789476, + "loss": 6.0912, + "step": 1179 + }, + { + "epoch": 0.06210526315789474, + "grad_norm": 0.9453125, + "learning_rate": 0.00012421052631578949, + "loss": 6.0272, + "step": 1180 + }, + { + "epoch": 0.062157894736842106, + "grad_norm": 0.8671875, + "learning_rate": 0.0001243157894736842, + "loss": 5.9588, + "step": 1181 + }, + { + "epoch": 0.06221052631578947, + "grad_norm": 0.984375, + "learning_rate": 0.00012442105263157894, + "loss": 6.069, + "step": 1182 + }, + { + "epoch": 0.06226315789473684, + "grad_norm": 0.9453125, + "learning_rate": 0.0001245263157894737, + "loss": 5.8838, + "step": 1183 + }, + { + "epoch": 0.06231578947368421, + "grad_norm": 1.0234375, + "learning_rate": 0.00012463157894736845, + "loss": 5.8603, + "step": 1184 + }, + { + "epoch": 0.06236842105263158, + "grad_norm": 1.046875, + "learning_rate": 0.00012473684210526317, + "loss": 5.9179, + "step": 1185 + }, + { + "epoch": 0.06242105263157895, + "grad_norm": 0.921875, + "learning_rate": 0.0001248421052631579, + "loss": 6.1559, + "step": 1186 + }, + { + "epoch": 0.06247368421052631, + "grad_norm": 0.98828125, + "learning_rate": 0.00012494736842105263, + "loss": 6.0622, + "step": 1187 + }, + { + "epoch": 0.06252631578947368, + "grad_norm": 1.1484375, + "learning_rate": 0.00012505263157894738, + "loss": 5.7427, + "step": 1188 + }, + { + "epoch": 0.06257894736842105, + "grad_norm": 1.3359375, + "learning_rate": 0.0001251578947368421, + "loss": 5.7808, + "step": 1189 + }, + { + "epoch": 0.06263157894736843, + "grad_norm": 0.79296875, + "learning_rate": 0.00012526315789473683, + "loss": 5.8685, + "step": 1190 + }, + { + "epoch": 0.06268421052631579, + "grad_norm": 0.7578125, + "learning_rate": 0.00012536842105263159, + "loss": 5.7672, + "step": 1191 + }, + { + "epoch": 0.06273684210526316, + "grad_norm": 0.7890625, + "learning_rate": 0.0001254736842105263, + "loss": 6.0562, + "step": 1192 + }, + { + "epoch": 0.06278947368421052, + "grad_norm": 0.70703125, + "learning_rate": 0.00012557894736842107, + "loss": 5.8982, + "step": 1193 + }, + { + "epoch": 0.0628421052631579, + "grad_norm": 0.87109375, + "learning_rate": 0.0001256842105263158, + "loss": 5.9904, + "step": 1194 + }, + { + "epoch": 0.06289473684210527, + "grad_norm": 0.89453125, + "learning_rate": 0.00012578947368421052, + "loss": 5.7381, + "step": 1195 + }, + { + "epoch": 0.06294736842105263, + "grad_norm": 1.046875, + "learning_rate": 0.00012589473684210527, + "loss": 5.982, + "step": 1196 + }, + { + "epoch": 0.063, + "grad_norm": 1.0234375, + "learning_rate": 0.000126, + "loss": 6.0548, + "step": 1197 + }, + { + "epoch": 0.06305263157894737, + "grad_norm": 0.8359375, + "learning_rate": 0.00012610526315789475, + "loss": 5.879, + "step": 1198 + }, + { + "epoch": 0.06310526315789473, + "grad_norm": 0.796875, + "learning_rate": 0.00012621052631578948, + "loss": 5.9796, + "step": 1199 + }, + { + "epoch": 0.06315789473684211, + "grad_norm": 0.8828125, + "learning_rate": 0.0001263157894736842, + "loss": 5.8079, + "step": 1200 + }, + { + "epoch": 0.06321052631578947, + "grad_norm": 0.9296875, + "learning_rate": 0.00012642105263157896, + "loss": 5.8504, + "step": 1201 + }, + { + "epoch": 0.06326315789473684, + "grad_norm": 0.921875, + "learning_rate": 0.0001265263157894737, + "loss": 5.9417, + "step": 1202 + }, + { + "epoch": 0.06331578947368421, + "grad_norm": 1.0, + "learning_rate": 0.00012663157894736844, + "loss": 5.5156, + "step": 1203 + }, + { + "epoch": 0.06336842105263157, + "grad_norm": 0.80859375, + "learning_rate": 0.00012673684210526317, + "loss": 5.8809, + "step": 1204 + }, + { + "epoch": 0.06342105263157895, + "grad_norm": 1.2109375, + "learning_rate": 0.0001268421052631579, + "loss": 5.4398, + "step": 1205 + }, + { + "epoch": 0.06347368421052632, + "grad_norm": 0.65625, + "learning_rate": 0.00012694736842105265, + "loss": 6.1652, + "step": 1206 + }, + { + "epoch": 0.06352631578947368, + "grad_norm": 0.671875, + "learning_rate": 0.00012705263157894737, + "loss": 5.6858, + "step": 1207 + }, + { + "epoch": 0.06357894736842105, + "grad_norm": 0.9140625, + "learning_rate": 0.00012715789473684213, + "loss": 6.0674, + "step": 1208 + }, + { + "epoch": 0.06363157894736843, + "grad_norm": 0.953125, + "learning_rate": 0.00012726315789473683, + "loss": 6.0094, + "step": 1209 + }, + { + "epoch": 0.06368421052631579, + "grad_norm": 0.99609375, + "learning_rate": 0.00012736842105263158, + "loss": 5.9357, + "step": 1210 + }, + { + "epoch": 0.06373684210526316, + "grad_norm": 1.0078125, + "learning_rate": 0.00012747368421052633, + "loss": 6.1089, + "step": 1211 + }, + { + "epoch": 0.06378947368421052, + "grad_norm": 0.85546875, + "learning_rate": 0.00012757894736842106, + "loss": 5.9638, + "step": 1212 + }, + { + "epoch": 0.0638421052631579, + "grad_norm": 1.0703125, + "learning_rate": 0.00012768421052631581, + "loss": 5.9802, + "step": 1213 + }, + { + "epoch": 0.06389473684210527, + "grad_norm": 0.9765625, + "learning_rate": 0.00012778947368421051, + "loss": 5.866, + "step": 1214 + }, + { + "epoch": 0.06394736842105263, + "grad_norm": 1.171875, + "learning_rate": 0.00012789473684210527, + "loss": 5.4016, + "step": 1215 + }, + { + "epoch": 0.064, + "grad_norm": 0.71484375, + "learning_rate": 0.00012800000000000002, + "loss": 5.9172, + "step": 1216 + }, + { + "epoch": 0.06405263157894737, + "grad_norm": 0.98046875, + "learning_rate": 0.00012810526315789475, + "loss": 5.8795, + "step": 1217 + }, + { + "epoch": 0.06410526315789473, + "grad_norm": 0.7265625, + "learning_rate": 0.00012821052631578947, + "loss": 6.0567, + "step": 1218 + }, + { + "epoch": 0.06415789473684211, + "grad_norm": 0.9140625, + "learning_rate": 0.0001283157894736842, + "loss": 5.9023, + "step": 1219 + }, + { + "epoch": 0.06421052631578947, + "grad_norm": 0.9921875, + "learning_rate": 0.00012842105263157895, + "loss": 5.627, + "step": 1220 + }, + { + "epoch": 0.06426315789473684, + "grad_norm": 1.15625, + "learning_rate": 0.00012852631578947368, + "loss": 5.8627, + "step": 1221 + }, + { + "epoch": 0.06431578947368422, + "grad_norm": 0.98046875, + "learning_rate": 0.00012863157894736843, + "loss": 5.8879, + "step": 1222 + }, + { + "epoch": 0.06436842105263157, + "grad_norm": 1.0625, + "learning_rate": 0.00012873684210526316, + "loss": 6.1462, + "step": 1223 + }, + { + "epoch": 0.06442105263157895, + "grad_norm": 0.7109375, + "learning_rate": 0.0001288421052631579, + "loss": 5.9458, + "step": 1224 + }, + { + "epoch": 0.06447368421052632, + "grad_norm": 0.76953125, + "learning_rate": 0.00012894736842105264, + "loss": 5.9956, + "step": 1225 + }, + { + "epoch": 0.06452631578947368, + "grad_norm": 0.875, + "learning_rate": 0.00012905263157894737, + "loss": 5.7111, + "step": 1226 + }, + { + "epoch": 0.06457894736842106, + "grad_norm": 0.83203125, + "learning_rate": 0.00012915789473684212, + "loss": 6.079, + "step": 1227 + }, + { + "epoch": 0.06463157894736841, + "grad_norm": 1.234375, + "learning_rate": 0.00012926315789473685, + "loss": 5.8865, + "step": 1228 + }, + { + "epoch": 0.06468421052631579, + "grad_norm": 0.75390625, + "learning_rate": 0.00012936842105263158, + "loss": 6.235, + "step": 1229 + }, + { + "epoch": 0.06473684210526316, + "grad_norm": 0.90234375, + "learning_rate": 0.00012947368421052633, + "loss": 5.5561, + "step": 1230 + }, + { + "epoch": 0.06478947368421052, + "grad_norm": 1.1953125, + "learning_rate": 0.00012957894736842106, + "loss": 5.9158, + "step": 1231 + }, + { + "epoch": 0.0648421052631579, + "grad_norm": 1.4609375, + "learning_rate": 0.00012968421052631578, + "loss": 5.8116, + "step": 1232 + }, + { + "epoch": 0.06489473684210527, + "grad_norm": 0.57421875, + "learning_rate": 0.00012978947368421054, + "loss": 5.8891, + "step": 1233 + }, + { + "epoch": 0.06494736842105263, + "grad_norm": 0.79296875, + "learning_rate": 0.00012989473684210526, + "loss": 5.8103, + "step": 1234 + }, + { + "epoch": 0.065, + "grad_norm": 0.71484375, + "learning_rate": 0.00013000000000000002, + "loss": 5.8605, + "step": 1235 + }, + { + "epoch": 0.06505263157894736, + "grad_norm": 1.453125, + "learning_rate": 0.00013010526315789474, + "loss": 5.7345, + "step": 1236 + }, + { + "epoch": 0.06510526315789474, + "grad_norm": 0.6875, + "learning_rate": 0.00013021052631578947, + "loss": 6.0024, + "step": 1237 + }, + { + "epoch": 0.06515789473684211, + "grad_norm": 0.953125, + "learning_rate": 0.00013031578947368422, + "loss": 5.7906, + "step": 1238 + }, + { + "epoch": 0.06521052631578947, + "grad_norm": 1.0703125, + "learning_rate": 0.00013042105263157895, + "loss": 5.6746, + "step": 1239 + }, + { + "epoch": 0.06526315789473684, + "grad_norm": 1.15625, + "learning_rate": 0.0001305263157894737, + "loss": 6.0691, + "step": 1240 + }, + { + "epoch": 0.06531578947368422, + "grad_norm": 0.76953125, + "learning_rate": 0.00013063157894736843, + "loss": 5.9348, + "step": 1241 + }, + { + "epoch": 0.06536842105263158, + "grad_norm": 1.078125, + "learning_rate": 0.00013073684210526316, + "loss": 5.9254, + "step": 1242 + }, + { + "epoch": 0.06542105263157895, + "grad_norm": 0.87890625, + "learning_rate": 0.0001308421052631579, + "loss": 5.8584, + "step": 1243 + }, + { + "epoch": 0.06547368421052631, + "grad_norm": 0.796875, + "learning_rate": 0.00013094736842105264, + "loss": 5.8351, + "step": 1244 + }, + { + "epoch": 0.06552631578947368, + "grad_norm": 0.75390625, + "learning_rate": 0.0001310526315789474, + "loss": 6.1832, + "step": 1245 + }, + { + "epoch": 0.06557894736842106, + "grad_norm": 1.125, + "learning_rate": 0.0001311578947368421, + "loss": 5.4019, + "step": 1246 + }, + { + "epoch": 0.06563157894736842, + "grad_norm": 0.84765625, + "learning_rate": 0.00013126315789473684, + "loss": 6.1552, + "step": 1247 + }, + { + "epoch": 0.06568421052631579, + "grad_norm": 0.94140625, + "learning_rate": 0.0001313684210526316, + "loss": 5.9084, + "step": 1248 + }, + { + "epoch": 0.06573684210526316, + "grad_norm": 1.0, + "learning_rate": 0.00013147368421052632, + "loss": 5.9175, + "step": 1249 + }, + { + "epoch": 0.06578947368421052, + "grad_norm": 1.0390625, + "learning_rate": 0.00013157894736842108, + "loss": 5.8553, + "step": 1250 + }, + { + "epoch": 0.0658421052631579, + "grad_norm": 0.99609375, + "learning_rate": 0.00013168421052631578, + "loss": 6.0532, + "step": 1251 + }, + { + "epoch": 0.06589473684210527, + "grad_norm": 0.89453125, + "learning_rate": 0.00013178947368421053, + "loss": 5.799, + "step": 1252 + }, + { + "epoch": 0.06594736842105263, + "grad_norm": 0.63671875, + "learning_rate": 0.00013189473684210528, + "loss": 5.9371, + "step": 1253 + }, + { + "epoch": 0.066, + "grad_norm": 0.6640625, + "learning_rate": 0.000132, + "loss": 5.4702, + "step": 1254 + }, + { + "epoch": 0.06605263157894736, + "grad_norm": 0.76171875, + "learning_rate": 0.00013210526315789476, + "loss": 5.8796, + "step": 1255 + }, + { + "epoch": 0.06610526315789474, + "grad_norm": 1.0390625, + "learning_rate": 0.00013221052631578946, + "loss": 5.6629, + "step": 1256 + }, + { + "epoch": 0.06615789473684211, + "grad_norm": 0.94921875, + "learning_rate": 0.00013231578947368422, + "loss": 5.8421, + "step": 1257 + }, + { + "epoch": 0.06621052631578947, + "grad_norm": 1.0859375, + "learning_rate": 0.00013242105263157894, + "loss": 6.3046, + "step": 1258 + }, + { + "epoch": 0.06626315789473684, + "grad_norm": 0.80078125, + "learning_rate": 0.0001325263157894737, + "loss": 5.7618, + "step": 1259 + }, + { + "epoch": 0.06631578947368422, + "grad_norm": 0.671875, + "learning_rate": 0.00013263157894736842, + "loss": 5.9162, + "step": 1260 + }, + { + "epoch": 0.06636842105263158, + "grad_norm": 0.8671875, + "learning_rate": 0.00013273684210526315, + "loss": 5.5323, + "step": 1261 + }, + { + "epoch": 0.06642105263157895, + "grad_norm": 0.69921875, + "learning_rate": 0.0001328421052631579, + "loss": 6.0268, + "step": 1262 + }, + { + "epoch": 0.06647368421052631, + "grad_norm": 0.9765625, + "learning_rate": 0.00013294736842105263, + "loss": 5.6858, + "step": 1263 + }, + { + "epoch": 0.06652631578947368, + "grad_norm": 1.1015625, + "learning_rate": 0.00013305263157894738, + "loss": 5.7569, + "step": 1264 + }, + { + "epoch": 0.06657894736842106, + "grad_norm": 0.76953125, + "learning_rate": 0.0001331578947368421, + "loss": 5.9232, + "step": 1265 + }, + { + "epoch": 0.06663157894736842, + "grad_norm": 0.859375, + "learning_rate": 0.00013326315789473684, + "loss": 6.045, + "step": 1266 + }, + { + "epoch": 0.06668421052631579, + "grad_norm": 0.84375, + "learning_rate": 0.0001333684210526316, + "loss": 5.8089, + "step": 1267 + }, + { + "epoch": 0.06673684210526316, + "grad_norm": 0.81640625, + "learning_rate": 0.00013347368421052632, + "loss": 5.9585, + "step": 1268 + }, + { + "epoch": 0.06678947368421052, + "grad_norm": 1.0078125, + "learning_rate": 0.00013357894736842104, + "loss": 5.7167, + "step": 1269 + }, + { + "epoch": 0.0668421052631579, + "grad_norm": 0.7890625, + "learning_rate": 0.0001336842105263158, + "loss": 5.7237, + "step": 1270 + }, + { + "epoch": 0.06689473684210526, + "grad_norm": 1.1328125, + "learning_rate": 0.00013378947368421052, + "loss": 5.4536, + "step": 1271 + }, + { + "epoch": 0.06694736842105263, + "grad_norm": 0.625, + "learning_rate": 0.00013389473684210528, + "loss": 5.716, + "step": 1272 + }, + { + "epoch": 0.067, + "grad_norm": 0.88671875, + "learning_rate": 0.000134, + "loss": 5.6375, + "step": 1273 + }, + { + "epoch": 0.06705263157894736, + "grad_norm": 0.74609375, + "learning_rate": 0.00013410526315789473, + "loss": 5.8918, + "step": 1274 + }, + { + "epoch": 0.06710526315789474, + "grad_norm": 0.87109375, + "learning_rate": 0.00013421052631578948, + "loss": 5.7959, + "step": 1275 + }, + { + "epoch": 0.06715789473684211, + "grad_norm": 1.0390625, + "learning_rate": 0.0001343157894736842, + "loss": 5.5365, + "step": 1276 + }, + { + "epoch": 0.06721052631578947, + "grad_norm": 0.9765625, + "learning_rate": 0.00013442105263157896, + "loss": 6.0185, + "step": 1277 + }, + { + "epoch": 0.06726315789473684, + "grad_norm": 1.140625, + "learning_rate": 0.0001345263157894737, + "loss": 5.3255, + "step": 1278 + }, + { + "epoch": 0.0673157894736842, + "grad_norm": 0.765625, + "learning_rate": 0.00013463157894736842, + "loss": 5.8597, + "step": 1279 + }, + { + "epoch": 0.06736842105263158, + "grad_norm": 0.7734375, + "learning_rate": 0.00013473684210526317, + "loss": 5.9404, + "step": 1280 + }, + { + "epoch": 0.06742105263157895, + "grad_norm": 0.8203125, + "learning_rate": 0.0001348421052631579, + "loss": 5.7125, + "step": 1281 + }, + { + "epoch": 0.06747368421052631, + "grad_norm": 0.7109375, + "learning_rate": 0.00013494736842105265, + "loss": 5.6546, + "step": 1282 + }, + { + "epoch": 0.06752631578947368, + "grad_norm": 0.765625, + "learning_rate": 0.00013505263157894735, + "loss": 5.826, + "step": 1283 + }, + { + "epoch": 0.06757894736842106, + "grad_norm": 0.7421875, + "learning_rate": 0.0001351578947368421, + "loss": 6.042, + "step": 1284 + }, + { + "epoch": 0.06763157894736842, + "grad_norm": 0.96484375, + "learning_rate": 0.00013526315789473686, + "loss": 5.7707, + "step": 1285 + }, + { + "epoch": 0.06768421052631579, + "grad_norm": 0.9375, + "learning_rate": 0.00013536842105263159, + "loss": 5.5167, + "step": 1286 + }, + { + "epoch": 0.06773684210526316, + "grad_norm": 0.7890625, + "learning_rate": 0.00013547368421052634, + "loss": 6.0788, + "step": 1287 + }, + { + "epoch": 0.06778947368421052, + "grad_norm": 0.93359375, + "learning_rate": 0.00013557894736842104, + "loss": 5.8472, + "step": 1288 + }, + { + "epoch": 0.0678421052631579, + "grad_norm": 0.91015625, + "learning_rate": 0.0001356842105263158, + "loss": 6.0615, + "step": 1289 + }, + { + "epoch": 0.06789473684210526, + "grad_norm": 0.84765625, + "learning_rate": 0.00013578947368421055, + "loss": 6.1284, + "step": 1290 + }, + { + "epoch": 0.06794736842105263, + "grad_norm": 0.82421875, + "learning_rate": 0.00013589473684210527, + "loss": 5.7931, + "step": 1291 + }, + { + "epoch": 0.068, + "grad_norm": 0.74609375, + "learning_rate": 0.00013600000000000003, + "loss": 5.7802, + "step": 1292 + }, + { + "epoch": 0.06805263157894736, + "grad_norm": 0.80078125, + "learning_rate": 0.00013610526315789473, + "loss": 5.8728, + "step": 1293 + }, + { + "epoch": 0.06810526315789474, + "grad_norm": 0.7734375, + "learning_rate": 0.00013621052631578948, + "loss": 5.7935, + "step": 1294 + }, + { + "epoch": 0.06815789473684211, + "grad_norm": 0.86328125, + "learning_rate": 0.0001363157894736842, + "loss": 6.0133, + "step": 1295 + }, + { + "epoch": 0.06821052631578947, + "grad_norm": 0.796875, + "learning_rate": 0.00013642105263157896, + "loss": 5.4763, + "step": 1296 + }, + { + "epoch": 0.06826315789473685, + "grad_norm": 0.7890625, + "learning_rate": 0.00013652631578947369, + "loss": 5.5164, + "step": 1297 + }, + { + "epoch": 0.0683157894736842, + "grad_norm": 1.015625, + "learning_rate": 0.0001366315789473684, + "loss": 6.1844, + "step": 1298 + }, + { + "epoch": 0.06836842105263158, + "grad_norm": 0.890625, + "learning_rate": 0.00013673684210526317, + "loss": 5.8558, + "step": 1299 + }, + { + "epoch": 0.06842105263157895, + "grad_norm": 1.046875, + "learning_rate": 0.0001368421052631579, + "loss": 5.5221, + "step": 1300 + }, + { + "epoch": 0.06847368421052631, + "grad_norm": 0.8984375, + "learning_rate": 0.00013694736842105265, + "loss": 5.7401, + "step": 1301 + }, + { + "epoch": 0.06852631578947369, + "grad_norm": 0.87109375, + "learning_rate": 0.00013705263157894737, + "loss": 5.9009, + "step": 1302 + }, + { + "epoch": 0.06857894736842106, + "grad_norm": 0.94921875, + "learning_rate": 0.0001371578947368421, + "loss": 6.0604, + "step": 1303 + }, + { + "epoch": 0.06863157894736842, + "grad_norm": 2.671875, + "learning_rate": 0.00013726315789473685, + "loss": 5.64, + "step": 1304 + }, + { + "epoch": 0.06868421052631579, + "grad_norm": 0.82421875, + "learning_rate": 0.00013736842105263158, + "loss": 5.6547, + "step": 1305 + }, + { + "epoch": 0.06873684210526315, + "grad_norm": 0.8046875, + "learning_rate": 0.00013747368421052633, + "loss": 5.8042, + "step": 1306 + }, + { + "epoch": 0.06878947368421053, + "grad_norm": 0.8671875, + "learning_rate": 0.00013757894736842106, + "loss": 6.0139, + "step": 1307 + }, + { + "epoch": 0.0688421052631579, + "grad_norm": 0.7578125, + "learning_rate": 0.0001376842105263158, + "loss": 5.8152, + "step": 1308 + }, + { + "epoch": 0.06889473684210526, + "grad_norm": 1.0234375, + "learning_rate": 0.00013778947368421054, + "loss": 6.0087, + "step": 1309 + }, + { + "epoch": 0.06894736842105263, + "grad_norm": 1.265625, + "learning_rate": 0.00013789473684210527, + "loss": 5.8082, + "step": 1310 + }, + { + "epoch": 0.069, + "grad_norm": 0.82421875, + "learning_rate": 0.000138, + "loss": 5.8438, + "step": 1311 + }, + { + "epoch": 0.06905263157894737, + "grad_norm": 1.046875, + "learning_rate": 0.00013810526315789475, + "loss": 5.4346, + "step": 1312 + }, + { + "epoch": 0.06910526315789474, + "grad_norm": 0.734375, + "learning_rate": 0.00013821052631578947, + "loss": 6.0484, + "step": 1313 + }, + { + "epoch": 0.0691578947368421, + "grad_norm": 1.09375, + "learning_rate": 0.00013831578947368423, + "loss": 5.6892, + "step": 1314 + }, + { + "epoch": 0.06921052631578947, + "grad_norm": 0.86328125, + "learning_rate": 0.00013842105263157895, + "loss": 6.014, + "step": 1315 + }, + { + "epoch": 0.06926315789473685, + "grad_norm": 0.703125, + "learning_rate": 0.00013852631578947368, + "loss": 5.93, + "step": 1316 + }, + { + "epoch": 0.0693157894736842, + "grad_norm": 0.9296875, + "learning_rate": 0.00013863157894736843, + "loss": 6.0458, + "step": 1317 + }, + { + "epoch": 0.06936842105263158, + "grad_norm": 0.94140625, + "learning_rate": 0.00013873684210526316, + "loss": 5.8687, + "step": 1318 + }, + { + "epoch": 0.06942105263157895, + "grad_norm": 1.03125, + "learning_rate": 0.00013884210526315791, + "loss": 5.4546, + "step": 1319 + }, + { + "epoch": 0.06947368421052631, + "grad_norm": 0.7265625, + "learning_rate": 0.00013894736842105264, + "loss": 6.0641, + "step": 1320 + }, + { + "epoch": 0.06952631578947369, + "grad_norm": 0.8515625, + "learning_rate": 0.00013905263157894737, + "loss": 5.3817, + "step": 1321 + }, + { + "epoch": 0.06957894736842105, + "grad_norm": 0.88671875, + "learning_rate": 0.00013915789473684212, + "loss": 6.0213, + "step": 1322 + }, + { + "epoch": 0.06963157894736842, + "grad_norm": 0.89453125, + "learning_rate": 0.00013926315789473685, + "loss": 5.6788, + "step": 1323 + }, + { + "epoch": 0.06968421052631579, + "grad_norm": 0.8828125, + "learning_rate": 0.0001393684210526316, + "loss": 5.7799, + "step": 1324 + }, + { + "epoch": 0.06973684210526315, + "grad_norm": 0.83984375, + "learning_rate": 0.0001394736842105263, + "loss": 5.838, + "step": 1325 + }, + { + "epoch": 0.06978947368421053, + "grad_norm": 0.90234375, + "learning_rate": 0.00013957894736842105, + "loss": 5.6205, + "step": 1326 + }, + { + "epoch": 0.0698421052631579, + "grad_norm": 0.90234375, + "learning_rate": 0.0001396842105263158, + "loss": 5.9609, + "step": 1327 + }, + { + "epoch": 0.06989473684210526, + "grad_norm": 0.80078125, + "learning_rate": 0.00013978947368421053, + "loss": 6.0356, + "step": 1328 + }, + { + "epoch": 0.06994736842105263, + "grad_norm": 0.8125, + "learning_rate": 0.0001398947368421053, + "loss": 5.8849, + "step": 1329 + }, + { + "epoch": 0.07, + "grad_norm": 0.86328125, + "learning_rate": 0.00014, + "loss": 5.7482, + "step": 1330 + }, + { + "epoch": 0.07005263157894737, + "grad_norm": 0.89453125, + "learning_rate": 0.00014010526315789474, + "loss": 5.4925, + "step": 1331 + }, + { + "epoch": 0.07010526315789474, + "grad_norm": 0.9453125, + "learning_rate": 0.00014021052631578947, + "loss": 5.732, + "step": 1332 + }, + { + "epoch": 0.0701578947368421, + "grad_norm": 0.8203125, + "learning_rate": 0.00014031578947368422, + "loss": 5.9218, + "step": 1333 + }, + { + "epoch": 0.07021052631578947, + "grad_norm": 0.91015625, + "learning_rate": 0.00014042105263157898, + "loss": 5.9434, + "step": 1334 + }, + { + "epoch": 0.07026315789473685, + "grad_norm": 0.94140625, + "learning_rate": 0.00014052631578947367, + "loss": 5.6411, + "step": 1335 + }, + { + "epoch": 0.0703157894736842, + "grad_norm": 0.67578125, + "learning_rate": 0.00014063157894736843, + "loss": 5.8385, + "step": 1336 + }, + { + "epoch": 0.07036842105263158, + "grad_norm": 0.67578125, + "learning_rate": 0.00014073684210526315, + "loss": 5.9814, + "step": 1337 + }, + { + "epoch": 0.07042105263157895, + "grad_norm": 0.83203125, + "learning_rate": 0.0001408421052631579, + "loss": 5.9431, + "step": 1338 + }, + { + "epoch": 0.07047368421052631, + "grad_norm": 0.84765625, + "learning_rate": 0.00014094736842105264, + "loss": 5.8575, + "step": 1339 + }, + { + "epoch": 0.07052631578947369, + "grad_norm": 1.046875, + "learning_rate": 0.00014105263157894736, + "loss": 5.9442, + "step": 1340 + }, + { + "epoch": 0.07057894736842105, + "grad_norm": 0.91015625, + "learning_rate": 0.00014115789473684212, + "loss": 5.9606, + "step": 1341 + }, + { + "epoch": 0.07063157894736842, + "grad_norm": 1.0703125, + "learning_rate": 0.00014126315789473684, + "loss": 5.8419, + "step": 1342 + }, + { + "epoch": 0.0706842105263158, + "grad_norm": 0.91796875, + "learning_rate": 0.0001413684210526316, + "loss": 5.8675, + "step": 1343 + }, + { + "epoch": 0.07073684210526315, + "grad_norm": 1.0390625, + "learning_rate": 0.00014147368421052632, + "loss": 5.4465, + "step": 1344 + }, + { + "epoch": 0.07078947368421053, + "grad_norm": 0.8828125, + "learning_rate": 0.00014157894736842105, + "loss": 5.639, + "step": 1345 + }, + { + "epoch": 0.0708421052631579, + "grad_norm": 1.046875, + "learning_rate": 0.0001416842105263158, + "loss": 5.8243, + "step": 1346 + }, + { + "epoch": 0.07089473684210526, + "grad_norm": 0.94140625, + "learning_rate": 0.00014178947368421053, + "loss": 5.7282, + "step": 1347 + }, + { + "epoch": 0.07094736842105263, + "grad_norm": 0.91015625, + "learning_rate": 0.00014189473684210528, + "loss": 5.7445, + "step": 1348 + }, + { + "epoch": 0.071, + "grad_norm": 0.80859375, + "learning_rate": 0.000142, + "loss": 5.5184, + "step": 1349 + }, + { + "epoch": 0.07105263157894737, + "grad_norm": 0.74609375, + "learning_rate": 0.00014210526315789474, + "loss": 5.8466, + "step": 1350 + }, + { + "epoch": 0.07110526315789474, + "grad_norm": 0.78125, + "learning_rate": 0.0001422105263157895, + "loss": 6.1352, + "step": 1351 + }, + { + "epoch": 0.0711578947368421, + "grad_norm": 0.83203125, + "learning_rate": 0.00014231578947368422, + "loss": 5.658, + "step": 1352 + }, + { + "epoch": 0.07121052631578947, + "grad_norm": 0.84765625, + "learning_rate": 0.00014242105263157894, + "loss": 5.7926, + "step": 1353 + }, + { + "epoch": 0.07126315789473685, + "grad_norm": 0.86328125, + "learning_rate": 0.0001425263157894737, + "loss": 5.5354, + "step": 1354 + }, + { + "epoch": 0.07131578947368421, + "grad_norm": 0.84765625, + "learning_rate": 0.00014263157894736842, + "loss": 5.6891, + "step": 1355 + }, + { + "epoch": 0.07136842105263158, + "grad_norm": 0.90234375, + "learning_rate": 0.00014273684210526318, + "loss": 5.6944, + "step": 1356 + }, + { + "epoch": 0.07142105263157894, + "grad_norm": 0.69140625, + "learning_rate": 0.0001428421052631579, + "loss": 5.6281, + "step": 1357 + }, + { + "epoch": 0.07147368421052631, + "grad_norm": 0.84765625, + "learning_rate": 0.00014294736842105263, + "loss": 5.5082, + "step": 1358 + }, + { + "epoch": 0.07152631578947369, + "grad_norm": 0.85546875, + "learning_rate": 0.00014305263157894738, + "loss": 5.4449, + "step": 1359 + }, + { + "epoch": 0.07157894736842105, + "grad_norm": 0.8046875, + "learning_rate": 0.0001431578947368421, + "loss": 5.7863, + "step": 1360 + }, + { + "epoch": 0.07163157894736842, + "grad_norm": 0.90625, + "learning_rate": 0.00014326315789473686, + "loss": 5.7606, + "step": 1361 + }, + { + "epoch": 0.0716842105263158, + "grad_norm": 0.81640625, + "learning_rate": 0.0001433684210526316, + "loss": 5.6329, + "step": 1362 + }, + { + "epoch": 0.07173684210526315, + "grad_norm": 0.80078125, + "learning_rate": 0.00014347368421052632, + "loss": 5.8688, + "step": 1363 + }, + { + "epoch": 0.07178947368421053, + "grad_norm": 0.81640625, + "learning_rate": 0.00014357894736842107, + "loss": 5.6156, + "step": 1364 + }, + { + "epoch": 0.0718421052631579, + "grad_norm": 0.69921875, + "learning_rate": 0.0001436842105263158, + "loss": 5.895, + "step": 1365 + }, + { + "epoch": 0.07189473684210526, + "grad_norm": 0.73828125, + "learning_rate": 0.00014378947368421055, + "loss": 5.7313, + "step": 1366 + }, + { + "epoch": 0.07194736842105263, + "grad_norm": 1.0234375, + "learning_rate": 0.00014389473684210525, + "loss": 5.6473, + "step": 1367 + }, + { + "epoch": 0.072, + "grad_norm": 0.91015625, + "learning_rate": 0.000144, + "loss": 5.8328, + "step": 1368 + }, + { + "epoch": 0.07205263157894737, + "grad_norm": 1.09375, + "learning_rate": 0.00014410526315789473, + "loss": 5.8743, + "step": 1369 + }, + { + "epoch": 0.07210526315789474, + "grad_norm": 1.140625, + "learning_rate": 0.00014421052631578948, + "loss": 5.4902, + "step": 1370 + }, + { + "epoch": 0.0721578947368421, + "grad_norm": 1.2890625, + "learning_rate": 0.00014431578947368424, + "loss": 5.5194, + "step": 1371 + }, + { + "epoch": 0.07221052631578948, + "grad_norm": 0.81640625, + "learning_rate": 0.00014442105263157894, + "loss": 5.6856, + "step": 1372 + }, + { + "epoch": 0.07226315789473685, + "grad_norm": 0.703125, + "learning_rate": 0.0001445263157894737, + "loss": 6.043, + "step": 1373 + }, + { + "epoch": 0.07231578947368421, + "grad_norm": 0.96484375, + "learning_rate": 0.00014463157894736842, + "loss": 5.1395, + "step": 1374 + }, + { + "epoch": 0.07236842105263158, + "grad_norm": 0.94140625, + "learning_rate": 0.00014473684210526317, + "loss": 5.4423, + "step": 1375 + }, + { + "epoch": 0.07242105263157894, + "grad_norm": 0.78125, + "learning_rate": 0.00014484210526315792, + "loss": 5.8361, + "step": 1376 + }, + { + "epoch": 0.07247368421052632, + "grad_norm": 0.87890625, + "learning_rate": 0.00014494736842105262, + "loss": 5.7198, + "step": 1377 + }, + { + "epoch": 0.07252631578947369, + "grad_norm": 1.046875, + "learning_rate": 0.00014505263157894738, + "loss": 5.6862, + "step": 1378 + }, + { + "epoch": 0.07257894736842105, + "grad_norm": 0.67578125, + "learning_rate": 0.0001451578947368421, + "loss": 5.7655, + "step": 1379 + }, + { + "epoch": 0.07263157894736842, + "grad_norm": 0.63671875, + "learning_rate": 0.00014526315789473686, + "loss": 5.7925, + "step": 1380 + }, + { + "epoch": 0.0726842105263158, + "grad_norm": 0.90234375, + "learning_rate": 0.00014536842105263158, + "loss": 5.6566, + "step": 1381 + }, + { + "epoch": 0.07273684210526316, + "grad_norm": 0.6875, + "learning_rate": 0.0001454736842105263, + "loss": 6.0552, + "step": 1382 + }, + { + "epoch": 0.07278947368421053, + "grad_norm": 0.92578125, + "learning_rate": 0.00014557894736842106, + "loss": 5.7086, + "step": 1383 + }, + { + "epoch": 0.07284210526315789, + "grad_norm": 1.0, + "learning_rate": 0.0001456842105263158, + "loss": 5.6257, + "step": 1384 + }, + { + "epoch": 0.07289473684210526, + "grad_norm": 0.76171875, + "learning_rate": 0.00014578947368421054, + "loss": 5.9284, + "step": 1385 + }, + { + "epoch": 0.07294736842105264, + "grad_norm": 0.93359375, + "learning_rate": 0.00014589473684210527, + "loss": 5.6293, + "step": 1386 + }, + { + "epoch": 0.073, + "grad_norm": 0.8203125, + "learning_rate": 0.000146, + "loss": 5.7372, + "step": 1387 + }, + { + "epoch": 0.07305263157894737, + "grad_norm": 0.74609375, + "learning_rate": 0.00014610526315789475, + "loss": 5.8846, + "step": 1388 + }, + { + "epoch": 0.07310526315789474, + "grad_norm": 1.2578125, + "learning_rate": 0.00014621052631578948, + "loss": 5.7523, + "step": 1389 + }, + { + "epoch": 0.0731578947368421, + "grad_norm": 1.0859375, + "learning_rate": 0.00014631578947368423, + "loss": 5.9725, + "step": 1390 + }, + { + "epoch": 0.07321052631578948, + "grad_norm": 0.8671875, + "learning_rate": 0.00014642105263157896, + "loss": 5.7842, + "step": 1391 + }, + { + "epoch": 0.07326315789473684, + "grad_norm": 0.78125, + "learning_rate": 0.00014652631578947369, + "loss": 5.5926, + "step": 1392 + }, + { + "epoch": 0.07331578947368421, + "grad_norm": 0.9296875, + "learning_rate": 0.00014663157894736844, + "loss": 5.7404, + "step": 1393 + }, + { + "epoch": 0.07336842105263158, + "grad_norm": 0.828125, + "learning_rate": 0.00014673684210526317, + "loss": 6.0162, + "step": 1394 + }, + { + "epoch": 0.07342105263157894, + "grad_norm": 1.2421875, + "learning_rate": 0.0001468421052631579, + "loss": 5.8005, + "step": 1395 + }, + { + "epoch": 0.07347368421052632, + "grad_norm": 1.046875, + "learning_rate": 0.00014694736842105265, + "loss": 5.7741, + "step": 1396 + }, + { + "epoch": 0.07352631578947369, + "grad_norm": 0.63671875, + "learning_rate": 0.00014705263157894737, + "loss": 5.658, + "step": 1397 + }, + { + "epoch": 0.07357894736842105, + "grad_norm": 0.83203125, + "learning_rate": 0.00014715789473684213, + "loss": 5.8554, + "step": 1398 + }, + { + "epoch": 0.07363157894736842, + "grad_norm": 1.2109375, + "learning_rate": 0.00014726315789473685, + "loss": 5.848, + "step": 1399 + }, + { + "epoch": 0.07368421052631578, + "grad_norm": 0.7734375, + "learning_rate": 0.00014736842105263158, + "loss": 5.8164, + "step": 1400 + }, + { + "epoch": 0.07373684210526316, + "grad_norm": 1.0546875, + "learning_rate": 0.00014747368421052633, + "loss": 5.6823, + "step": 1401 + }, + { + "epoch": 0.07378947368421053, + "grad_norm": 0.7421875, + "learning_rate": 0.00014757894736842106, + "loss": 5.8139, + "step": 1402 + }, + { + "epoch": 0.07384210526315789, + "grad_norm": 0.76953125, + "learning_rate": 0.0001476842105263158, + "loss": 5.764, + "step": 1403 + }, + { + "epoch": 0.07389473684210526, + "grad_norm": 0.89453125, + "learning_rate": 0.00014778947368421054, + "loss": 5.7544, + "step": 1404 + }, + { + "epoch": 0.07394736842105264, + "grad_norm": 0.6640625, + "learning_rate": 0.00014789473684210527, + "loss": 5.4579, + "step": 1405 + }, + { + "epoch": 0.074, + "grad_norm": 0.70703125, + "learning_rate": 0.000148, + "loss": 5.8994, + "step": 1406 + }, + { + "epoch": 0.07405263157894737, + "grad_norm": 0.91796875, + "learning_rate": 0.00014810526315789475, + "loss": 5.7882, + "step": 1407 + }, + { + "epoch": 0.07410526315789474, + "grad_norm": 0.859375, + "learning_rate": 0.0001482105263157895, + "loss": 5.8055, + "step": 1408 + }, + { + "epoch": 0.0741578947368421, + "grad_norm": 0.7890625, + "learning_rate": 0.0001483157894736842, + "loss": 5.6976, + "step": 1409 + }, + { + "epoch": 0.07421052631578948, + "grad_norm": 0.890625, + "learning_rate": 0.00014842105263157895, + "loss": 5.9837, + "step": 1410 + }, + { + "epoch": 0.07426315789473684, + "grad_norm": 1.125, + "learning_rate": 0.00014852631578947368, + "loss": 5.9509, + "step": 1411 + }, + { + "epoch": 0.07431578947368421, + "grad_norm": 0.78125, + "learning_rate": 0.00014863157894736843, + "loss": 5.9655, + "step": 1412 + }, + { + "epoch": 0.07436842105263158, + "grad_norm": 0.87890625, + "learning_rate": 0.0001487368421052632, + "loss": 5.88, + "step": 1413 + }, + { + "epoch": 0.07442105263157894, + "grad_norm": 0.9921875, + "learning_rate": 0.00014884210526315789, + "loss": 5.796, + "step": 1414 + }, + { + "epoch": 0.07447368421052632, + "grad_norm": 0.79296875, + "learning_rate": 0.00014894736842105264, + "loss": 5.8237, + "step": 1415 + }, + { + "epoch": 0.07452631578947369, + "grad_norm": 0.8046875, + "learning_rate": 0.00014905263157894737, + "loss": 6.0002, + "step": 1416 + }, + { + "epoch": 0.07457894736842105, + "grad_norm": 1.0, + "learning_rate": 0.00014915789473684212, + "loss": 5.8452, + "step": 1417 + }, + { + "epoch": 0.07463157894736842, + "grad_norm": 0.96875, + "learning_rate": 0.00014926315789473687, + "loss": 5.974, + "step": 1418 + }, + { + "epoch": 0.07468421052631578, + "grad_norm": 0.7890625, + "learning_rate": 0.00014936842105263157, + "loss": 5.4242, + "step": 1419 + }, + { + "epoch": 0.07473684210526316, + "grad_norm": 0.828125, + "learning_rate": 0.00014947368421052633, + "loss": 5.7766, + "step": 1420 + }, + { + "epoch": 0.07478947368421053, + "grad_norm": 0.99609375, + "learning_rate": 0.00014957894736842105, + "loss": 5.6084, + "step": 1421 + }, + { + "epoch": 0.07484210526315789, + "grad_norm": 0.96484375, + "learning_rate": 0.0001496842105263158, + "loss": 5.5058, + "step": 1422 + }, + { + "epoch": 0.07489473684210526, + "grad_norm": 0.828125, + "learning_rate": 0.00014978947368421053, + "loss": 5.8309, + "step": 1423 + }, + { + "epoch": 0.07494736842105264, + "grad_norm": 0.765625, + "learning_rate": 0.00014989473684210526, + "loss": 5.5258, + "step": 1424 + }, + { + "epoch": 0.075, + "grad_norm": 0.87109375, + "learning_rate": 0.00015000000000000001, + "loss": 5.916, + "step": 1425 + }, + { + "epoch": 0.07505263157894737, + "grad_norm": 0.80078125, + "learning_rate": 0.00015010526315789474, + "loss": 5.6576, + "step": 1426 + }, + { + "epoch": 0.07510526315789473, + "grad_norm": 0.8671875, + "learning_rate": 0.0001502105263157895, + "loss": 5.6348, + "step": 1427 + }, + { + "epoch": 0.0751578947368421, + "grad_norm": 0.71484375, + "learning_rate": 0.00015031578947368422, + "loss": 5.8216, + "step": 1428 + }, + { + "epoch": 0.07521052631578948, + "grad_norm": 0.6953125, + "learning_rate": 0.00015042105263157895, + "loss": 5.6967, + "step": 1429 + }, + { + "epoch": 0.07526315789473684, + "grad_norm": 0.73828125, + "learning_rate": 0.0001505263157894737, + "loss": 5.6723, + "step": 1430 + }, + { + "epoch": 0.07531578947368421, + "grad_norm": 0.7421875, + "learning_rate": 0.00015063157894736843, + "loss": 5.846, + "step": 1431 + }, + { + "epoch": 0.07536842105263158, + "grad_norm": 0.76171875, + "learning_rate": 0.00015073684210526315, + "loss": 6.092, + "step": 1432 + }, + { + "epoch": 0.07542105263157894, + "grad_norm": 0.83203125, + "learning_rate": 0.0001508421052631579, + "loss": 5.4346, + "step": 1433 + }, + { + "epoch": 0.07547368421052632, + "grad_norm": 0.81640625, + "learning_rate": 0.00015094736842105263, + "loss": 5.4015, + "step": 1434 + }, + { + "epoch": 0.07552631578947368, + "grad_norm": 0.921875, + "learning_rate": 0.0001510526315789474, + "loss": 5.3544, + "step": 1435 + }, + { + "epoch": 0.07557894736842105, + "grad_norm": 0.859375, + "learning_rate": 0.00015115789473684211, + "loss": 5.5941, + "step": 1436 + }, + { + "epoch": 0.07563157894736842, + "grad_norm": 0.6875, + "learning_rate": 0.00015126315789473684, + "loss": 5.4125, + "step": 1437 + }, + { + "epoch": 0.07568421052631578, + "grad_norm": 0.73046875, + "learning_rate": 0.0001513684210526316, + "loss": 5.5538, + "step": 1438 + }, + { + "epoch": 0.07573684210526316, + "grad_norm": 0.7890625, + "learning_rate": 0.00015147368421052632, + "loss": 5.4964, + "step": 1439 + }, + { + "epoch": 0.07578947368421053, + "grad_norm": 0.734375, + "learning_rate": 0.00015157894736842108, + "loss": 5.9429, + "step": 1440 + }, + { + "epoch": 0.07584210526315789, + "grad_norm": 0.78515625, + "learning_rate": 0.0001516842105263158, + "loss": 5.779, + "step": 1441 + }, + { + "epoch": 0.07589473684210526, + "grad_norm": 0.88671875, + "learning_rate": 0.00015178947368421053, + "loss": 5.5769, + "step": 1442 + }, + { + "epoch": 0.07594736842105264, + "grad_norm": 0.74609375, + "learning_rate": 0.00015189473684210525, + "loss": 5.9778, + "step": 1443 + }, + { + "epoch": 0.076, + "grad_norm": 0.796875, + "learning_rate": 0.000152, + "loss": 5.56, + "step": 1444 + }, + { + "epoch": 0.07605263157894737, + "grad_norm": 0.91015625, + "learning_rate": 0.00015210526315789476, + "loss": 5.9068, + "step": 1445 + }, + { + "epoch": 0.07610526315789473, + "grad_norm": 0.78515625, + "learning_rate": 0.00015221052631578946, + "loss": 5.7269, + "step": 1446 + }, + { + "epoch": 0.0761578947368421, + "grad_norm": 0.87109375, + "learning_rate": 0.00015231578947368422, + "loss": 5.4551, + "step": 1447 + }, + { + "epoch": 0.07621052631578948, + "grad_norm": 0.7578125, + "learning_rate": 0.00015242105263157894, + "loss": 5.822, + "step": 1448 + }, + { + "epoch": 0.07626315789473684, + "grad_norm": 0.8046875, + "learning_rate": 0.0001525263157894737, + "loss": 5.6849, + "step": 1449 + }, + { + "epoch": 0.07631578947368421, + "grad_norm": 1.078125, + "learning_rate": 0.00015263157894736845, + "loss": 6.0351, + "step": 1450 + }, + { + "epoch": 0.07636842105263159, + "grad_norm": 0.796875, + "learning_rate": 0.00015273684210526315, + "loss": 5.7523, + "step": 1451 + }, + { + "epoch": 0.07642105263157895, + "grad_norm": 0.8984375, + "learning_rate": 0.0001528421052631579, + "loss": 5.7844, + "step": 1452 + }, + { + "epoch": 0.07647368421052632, + "grad_norm": 1.109375, + "learning_rate": 0.00015294736842105263, + "loss": 5.6375, + "step": 1453 + }, + { + "epoch": 0.07652631578947368, + "grad_norm": 0.65234375, + "learning_rate": 0.00015305263157894738, + "loss": 5.8793, + "step": 1454 + }, + { + "epoch": 0.07657894736842105, + "grad_norm": 0.70703125, + "learning_rate": 0.0001531578947368421, + "loss": 5.7855, + "step": 1455 + }, + { + "epoch": 0.07663157894736843, + "grad_norm": 0.83203125, + "learning_rate": 0.00015326315789473684, + "loss": 5.8171, + "step": 1456 + }, + { + "epoch": 0.07668421052631579, + "grad_norm": 0.80859375, + "learning_rate": 0.0001533684210526316, + "loss": 6.0781, + "step": 1457 + }, + { + "epoch": 0.07673684210526316, + "grad_norm": 0.796875, + "learning_rate": 0.00015347368421052632, + "loss": 5.9799, + "step": 1458 + }, + { + "epoch": 0.07678947368421053, + "grad_norm": 0.8046875, + "learning_rate": 0.00015357894736842107, + "loss": 5.6914, + "step": 1459 + }, + { + "epoch": 0.07684210526315789, + "grad_norm": 0.78515625, + "learning_rate": 0.0001536842105263158, + "loss": 5.8336, + "step": 1460 + }, + { + "epoch": 0.07689473684210527, + "grad_norm": 0.82421875, + "learning_rate": 0.00015378947368421052, + "loss": 5.7134, + "step": 1461 + }, + { + "epoch": 0.07694736842105263, + "grad_norm": 0.65625, + "learning_rate": 0.00015389473684210528, + "loss": 5.5622, + "step": 1462 + }, + { + "epoch": 0.077, + "grad_norm": 0.69140625, + "learning_rate": 0.000154, + "loss": 5.7923, + "step": 1463 + }, + { + "epoch": 0.07705263157894737, + "grad_norm": 0.7890625, + "learning_rate": 0.00015410526315789476, + "loss": 5.5094, + "step": 1464 + }, + { + "epoch": 0.07710526315789473, + "grad_norm": 0.93359375, + "learning_rate": 0.00015421052631578948, + "loss": 5.8516, + "step": 1465 + }, + { + "epoch": 0.0771578947368421, + "grad_norm": 2.0, + "learning_rate": 0.0001543157894736842, + "loss": 5.5756, + "step": 1466 + }, + { + "epoch": 0.07721052631578948, + "grad_norm": 0.78515625, + "learning_rate": 0.00015442105263157896, + "loss": 5.6069, + "step": 1467 + }, + { + "epoch": 0.07726315789473684, + "grad_norm": 0.71875, + "learning_rate": 0.0001545263157894737, + "loss": 5.6348, + "step": 1468 + }, + { + "epoch": 0.07731578947368421, + "grad_norm": 0.84375, + "learning_rate": 0.00015463157894736844, + "loss": 5.4907, + "step": 1469 + }, + { + "epoch": 0.07736842105263157, + "grad_norm": 0.9765625, + "learning_rate": 0.00015473684210526317, + "loss": 5.8202, + "step": 1470 + }, + { + "epoch": 0.07742105263157895, + "grad_norm": 0.890625, + "learning_rate": 0.0001548421052631579, + "loss": 6.0807, + "step": 1471 + }, + { + "epoch": 0.07747368421052632, + "grad_norm": 0.875, + "learning_rate": 0.00015494736842105265, + "loss": 5.8408, + "step": 1472 + }, + { + "epoch": 0.07752631578947368, + "grad_norm": 0.8359375, + "learning_rate": 0.00015505263157894738, + "loss": 5.7343, + "step": 1473 + }, + { + "epoch": 0.07757894736842105, + "grad_norm": 0.9921875, + "learning_rate": 0.0001551578947368421, + "loss": 5.5505, + "step": 1474 + }, + { + "epoch": 0.07763157894736843, + "grad_norm": 0.77734375, + "learning_rate": 0.00015526315789473686, + "loss": 5.8877, + "step": 1475 + }, + { + "epoch": 0.07768421052631579, + "grad_norm": 0.61328125, + "learning_rate": 0.00015536842105263158, + "loss": 5.9007, + "step": 1476 + }, + { + "epoch": 0.07773684210526316, + "grad_norm": 0.93359375, + "learning_rate": 0.00015547368421052634, + "loss": 5.3375, + "step": 1477 + }, + { + "epoch": 0.07778947368421052, + "grad_norm": 0.78125, + "learning_rate": 0.00015557894736842106, + "loss": 5.6854, + "step": 1478 + }, + { + "epoch": 0.07784210526315789, + "grad_norm": 0.75390625, + "learning_rate": 0.0001556842105263158, + "loss": 5.5443, + "step": 1479 + }, + { + "epoch": 0.07789473684210527, + "grad_norm": 0.66796875, + "learning_rate": 0.00015578947368421052, + "loss": 5.6262, + "step": 1480 + }, + { + "epoch": 0.07794736842105263, + "grad_norm": 0.82421875, + "learning_rate": 0.00015589473684210527, + "loss": 5.3292, + "step": 1481 + }, + { + "epoch": 0.078, + "grad_norm": 0.734375, + "learning_rate": 0.00015600000000000002, + "loss": 5.4746, + "step": 1482 + }, + { + "epoch": 0.07805263157894737, + "grad_norm": 1.2265625, + "learning_rate": 0.00015610526315789475, + "loss": 5.8063, + "step": 1483 + }, + { + "epoch": 0.07810526315789473, + "grad_norm": 0.703125, + "learning_rate": 0.00015621052631578948, + "loss": 5.5706, + "step": 1484 + }, + { + "epoch": 0.0781578947368421, + "grad_norm": 0.67578125, + "learning_rate": 0.0001563157894736842, + "loss": 5.3571, + "step": 1485 + }, + { + "epoch": 0.07821052631578948, + "grad_norm": 0.93359375, + "learning_rate": 0.00015642105263157896, + "loss": 5.3521, + "step": 1486 + }, + { + "epoch": 0.07826315789473684, + "grad_norm": 0.765625, + "learning_rate": 0.0001565263157894737, + "loss": 5.6418, + "step": 1487 + }, + { + "epoch": 0.07831578947368421, + "grad_norm": 0.76953125, + "learning_rate": 0.0001566315789473684, + "loss": 5.451, + "step": 1488 + }, + { + "epoch": 0.07836842105263157, + "grad_norm": 0.9921875, + "learning_rate": 0.00015673684210526316, + "loss": 5.4534, + "step": 1489 + }, + { + "epoch": 0.07842105263157895, + "grad_norm": 0.84765625, + "learning_rate": 0.0001568421052631579, + "loss": 5.7575, + "step": 1490 + }, + { + "epoch": 0.07847368421052632, + "grad_norm": 0.83203125, + "learning_rate": 0.00015694736842105264, + "loss": 5.8679, + "step": 1491 + }, + { + "epoch": 0.07852631578947368, + "grad_norm": 0.9375, + "learning_rate": 0.00015705263157894737, + "loss": 5.5609, + "step": 1492 + }, + { + "epoch": 0.07857894736842105, + "grad_norm": 0.83203125, + "learning_rate": 0.0001571578947368421, + "loss": 5.6803, + "step": 1493 + }, + { + "epoch": 0.07863157894736843, + "grad_norm": 0.875, + "learning_rate": 0.00015726315789473685, + "loss": 5.6365, + "step": 1494 + }, + { + "epoch": 0.07868421052631579, + "grad_norm": 1.171875, + "learning_rate": 0.00015736842105263158, + "loss": 5.462, + "step": 1495 + }, + { + "epoch": 0.07873684210526316, + "grad_norm": 1.671875, + "learning_rate": 0.00015747368421052633, + "loss": 5.347, + "step": 1496 + }, + { + "epoch": 0.07878947368421052, + "grad_norm": 0.7578125, + "learning_rate": 0.00015757894736842106, + "loss": 5.7759, + "step": 1497 + }, + { + "epoch": 0.0788421052631579, + "grad_norm": 0.9375, + "learning_rate": 0.00015768421052631578, + "loss": 5.5106, + "step": 1498 + }, + { + "epoch": 0.07889473684210527, + "grad_norm": 0.9453125, + "learning_rate": 0.00015778947368421054, + "loss": 5.6107, + "step": 1499 + }, + { + "epoch": 0.07894736842105263, + "grad_norm": 0.828125, + "learning_rate": 0.00015789473684210527, + "loss": 5.6932, + "step": 1500 + }, + { + "epoch": 0.079, + "grad_norm": 0.9140625, + "learning_rate": 0.00015800000000000002, + "loss": 5.7994, + "step": 1501 + }, + { + "epoch": 0.07905263157894737, + "grad_norm": 0.71484375, + "learning_rate": 0.00015810526315789475, + "loss": 5.6692, + "step": 1502 + }, + { + "epoch": 0.07910526315789473, + "grad_norm": 0.80859375, + "learning_rate": 0.00015821052631578947, + "loss": 5.5734, + "step": 1503 + }, + { + "epoch": 0.07915789473684211, + "grad_norm": 0.72265625, + "learning_rate": 0.00015831578947368423, + "loss": 5.7378, + "step": 1504 + }, + { + "epoch": 0.07921052631578947, + "grad_norm": 0.85546875, + "learning_rate": 0.00015842105263157895, + "loss": 5.4214, + "step": 1505 + }, + { + "epoch": 0.07926315789473684, + "grad_norm": 0.796875, + "learning_rate": 0.0001585263157894737, + "loss": 5.46, + "step": 1506 + }, + { + "epoch": 0.07931578947368421, + "grad_norm": 0.890625, + "learning_rate": 0.00015863157894736843, + "loss": 5.9116, + "step": 1507 + }, + { + "epoch": 0.07936842105263157, + "grad_norm": 0.8671875, + "learning_rate": 0.00015873684210526316, + "loss": 5.6409, + "step": 1508 + }, + { + "epoch": 0.07942105263157895, + "grad_norm": 0.97265625, + "learning_rate": 0.0001588421052631579, + "loss": 5.8221, + "step": 1509 + }, + { + "epoch": 0.07947368421052632, + "grad_norm": 0.78515625, + "learning_rate": 0.00015894736842105264, + "loss": 5.7451, + "step": 1510 + }, + { + "epoch": 0.07952631578947368, + "grad_norm": 0.703125, + "learning_rate": 0.0001590526315789474, + "loss": 5.3767, + "step": 1511 + }, + { + "epoch": 0.07957894736842105, + "grad_norm": 0.78515625, + "learning_rate": 0.00015915789473684212, + "loss": 5.4207, + "step": 1512 + }, + { + "epoch": 0.07963157894736841, + "grad_norm": 0.859375, + "learning_rate": 0.00015926315789473685, + "loss": 5.5169, + "step": 1513 + }, + { + "epoch": 0.07968421052631579, + "grad_norm": 1.15625, + "learning_rate": 0.0001593684210526316, + "loss": 5.1292, + "step": 1514 + }, + { + "epoch": 0.07973684210526316, + "grad_norm": 0.703125, + "learning_rate": 0.00015947368421052633, + "loss": 5.4393, + "step": 1515 + }, + { + "epoch": 0.07978947368421052, + "grad_norm": 0.91015625, + "learning_rate": 0.00015957894736842105, + "loss": 5.6316, + "step": 1516 + }, + { + "epoch": 0.0798421052631579, + "grad_norm": 0.8359375, + "learning_rate": 0.00015968421052631578, + "loss": 5.5821, + "step": 1517 + }, + { + "epoch": 0.07989473684210527, + "grad_norm": 0.95703125, + "learning_rate": 0.00015978947368421053, + "loss": 5.623, + "step": 1518 + }, + { + "epoch": 0.07994736842105263, + "grad_norm": 0.83203125, + "learning_rate": 0.0001598947368421053, + "loss": 5.4249, + "step": 1519 + }, + { + "epoch": 0.08, + "grad_norm": 0.79296875, + "learning_rate": 0.00016, + "loss": 5.6835, + "step": 1520 + }, + { + "epoch": 0.08005263157894738, + "grad_norm": 0.83984375, + "learning_rate": 0.00016010526315789474, + "loss": 6.0452, + "step": 1521 + }, + { + "epoch": 0.08010526315789473, + "grad_norm": 0.66796875, + "learning_rate": 0.00016021052631578947, + "loss": 5.6773, + "step": 1522 + }, + { + "epoch": 0.08015789473684211, + "grad_norm": 0.82421875, + "learning_rate": 0.00016031578947368422, + "loss": 5.9201, + "step": 1523 + }, + { + "epoch": 0.08021052631578947, + "grad_norm": 0.90234375, + "learning_rate": 0.00016042105263157897, + "loss": 5.5565, + "step": 1524 + }, + { + "epoch": 0.08026315789473684, + "grad_norm": 0.77734375, + "learning_rate": 0.0001605263157894737, + "loss": 5.7447, + "step": 1525 + }, + { + "epoch": 0.08031578947368422, + "grad_norm": 1.1484375, + "learning_rate": 0.00016063157894736843, + "loss": 5.5783, + "step": 1526 + }, + { + "epoch": 0.08036842105263158, + "grad_norm": 0.6796875, + "learning_rate": 0.00016073684210526315, + "loss": 5.6109, + "step": 1527 + }, + { + "epoch": 0.08042105263157895, + "grad_norm": 0.7734375, + "learning_rate": 0.0001608421052631579, + "loss": 5.6542, + "step": 1528 + }, + { + "epoch": 0.08047368421052632, + "grad_norm": 1.3515625, + "learning_rate": 0.00016094736842105263, + "loss": 5.6829, + "step": 1529 + }, + { + "epoch": 0.08052631578947368, + "grad_norm": 1.1875, + "learning_rate": 0.00016105263157894736, + "loss": 5.6219, + "step": 1530 + }, + { + "epoch": 0.08057894736842106, + "grad_norm": 0.91796875, + "learning_rate": 0.00016115789473684211, + "loss": 5.6898, + "step": 1531 + }, + { + "epoch": 0.08063157894736842, + "grad_norm": 0.93359375, + "learning_rate": 0.00016126315789473684, + "loss": 5.6901, + "step": 1532 + }, + { + "epoch": 0.08068421052631579, + "grad_norm": 0.671875, + "learning_rate": 0.0001613684210526316, + "loss": 5.774, + "step": 1533 + }, + { + "epoch": 0.08073684210526316, + "grad_norm": 0.90234375, + "learning_rate": 0.00016147368421052632, + "loss": 5.5413, + "step": 1534 + }, + { + "epoch": 0.08078947368421052, + "grad_norm": 1.125, + "learning_rate": 0.00016157894736842105, + "loss": 5.6622, + "step": 1535 + }, + { + "epoch": 0.0808421052631579, + "grad_norm": 0.8984375, + "learning_rate": 0.0001616842105263158, + "loss": 5.7606, + "step": 1536 + }, + { + "epoch": 0.08089473684210527, + "grad_norm": 0.94921875, + "learning_rate": 0.00016178947368421053, + "loss": 5.3792, + "step": 1537 + }, + { + "epoch": 0.08094736842105263, + "grad_norm": 0.87109375, + "learning_rate": 0.00016189473684210528, + "loss": 5.8973, + "step": 1538 + }, + { + "epoch": 0.081, + "grad_norm": 0.76953125, + "learning_rate": 0.000162, + "loss": 5.7326, + "step": 1539 + }, + { + "epoch": 0.08105263157894736, + "grad_norm": 0.69921875, + "learning_rate": 0.00016210526315789473, + "loss": 5.3897, + "step": 1540 + }, + { + "epoch": 0.08110526315789474, + "grad_norm": 0.84375, + "learning_rate": 0.0001622105263157895, + "loss": 5.471, + "step": 1541 + }, + { + "epoch": 0.08115789473684211, + "grad_norm": 1.078125, + "learning_rate": 0.00016231578947368421, + "loss": 5.4582, + "step": 1542 + }, + { + "epoch": 0.08121052631578947, + "grad_norm": 0.83203125, + "learning_rate": 0.00016242105263157897, + "loss": 5.7444, + "step": 1543 + }, + { + "epoch": 0.08126315789473684, + "grad_norm": 0.984375, + "learning_rate": 0.0001625263157894737, + "loss": 5.7477, + "step": 1544 + }, + { + "epoch": 0.08131578947368422, + "grad_norm": 0.8359375, + "learning_rate": 0.00016263157894736842, + "loss": 5.7193, + "step": 1545 + }, + { + "epoch": 0.08136842105263158, + "grad_norm": 0.78125, + "learning_rate": 0.00016273684210526317, + "loss": 5.9726, + "step": 1546 + }, + { + "epoch": 0.08142105263157895, + "grad_norm": 0.88671875, + "learning_rate": 0.0001628421052631579, + "loss": 5.5409, + "step": 1547 + }, + { + "epoch": 0.08147368421052631, + "grad_norm": 0.96875, + "learning_rate": 0.00016294736842105265, + "loss": 5.2953, + "step": 1548 + }, + { + "epoch": 0.08152631578947368, + "grad_norm": 0.9765625, + "learning_rate": 0.00016305263157894735, + "loss": 5.9521, + "step": 1549 + }, + { + "epoch": 0.08157894736842106, + "grad_norm": 1.0234375, + "learning_rate": 0.0001631578947368421, + "loss": 5.7417, + "step": 1550 + }, + { + "epoch": 0.08163157894736842, + "grad_norm": 0.63671875, + "learning_rate": 0.00016326315789473686, + "loss": 5.7722, + "step": 1551 + }, + { + "epoch": 0.08168421052631579, + "grad_norm": 0.80859375, + "learning_rate": 0.0001633684210526316, + "loss": 5.8132, + "step": 1552 + }, + { + "epoch": 0.08173684210526316, + "grad_norm": 0.78125, + "learning_rate": 0.00016347368421052634, + "loss": 5.2263, + "step": 1553 + }, + { + "epoch": 0.08178947368421052, + "grad_norm": 0.703125, + "learning_rate": 0.00016357894736842104, + "loss": 5.9189, + "step": 1554 + }, + { + "epoch": 0.0818421052631579, + "grad_norm": 0.8046875, + "learning_rate": 0.0001636842105263158, + "loss": 5.5277, + "step": 1555 + }, + { + "epoch": 0.08189473684210526, + "grad_norm": 1.03125, + "learning_rate": 0.00016378947368421055, + "loss": 5.254, + "step": 1556 + }, + { + "epoch": 0.08194736842105263, + "grad_norm": 0.953125, + "learning_rate": 0.00016389473684210528, + "loss": 5.5814, + "step": 1557 + }, + { + "epoch": 0.082, + "grad_norm": 0.92578125, + "learning_rate": 0.000164, + "loss": 5.6977, + "step": 1558 + }, + { + "epoch": 0.08205263157894736, + "grad_norm": 0.87890625, + "learning_rate": 0.00016410526315789473, + "loss": 5.6175, + "step": 1559 + }, + { + "epoch": 0.08210526315789474, + "grad_norm": 0.99609375, + "learning_rate": 0.00016421052631578948, + "loss": 5.6784, + "step": 1560 + }, + { + "epoch": 0.08215789473684211, + "grad_norm": 0.94140625, + "learning_rate": 0.00016431578947368424, + "loss": 5.7419, + "step": 1561 + }, + { + "epoch": 0.08221052631578947, + "grad_norm": 0.91796875, + "learning_rate": 0.00016442105263157896, + "loss": 5.618, + "step": 1562 + }, + { + "epoch": 0.08226315789473684, + "grad_norm": 0.828125, + "learning_rate": 0.0001645263157894737, + "loss": 5.5387, + "step": 1563 + }, + { + "epoch": 0.08231578947368422, + "grad_norm": 0.75, + "learning_rate": 0.00016463157894736842, + "loss": 5.5868, + "step": 1564 + }, + { + "epoch": 0.08236842105263158, + "grad_norm": 0.6328125, + "learning_rate": 0.00016473684210526317, + "loss": 5.7021, + "step": 1565 + }, + { + "epoch": 0.08242105263157895, + "grad_norm": 0.734375, + "learning_rate": 0.0001648421052631579, + "loss": 5.6586, + "step": 1566 + }, + { + "epoch": 0.08247368421052631, + "grad_norm": 0.73828125, + "learning_rate": 0.00016494736842105265, + "loss": 5.3282, + "step": 1567 + }, + { + "epoch": 0.08252631578947368, + "grad_norm": 0.875, + "learning_rate": 0.00016505263157894738, + "loss": 5.5844, + "step": 1568 + }, + { + "epoch": 0.08257894736842106, + "grad_norm": 0.859375, + "learning_rate": 0.0001651578947368421, + "loss": 5.7088, + "step": 1569 + }, + { + "epoch": 0.08263157894736842, + "grad_norm": 0.72265625, + "learning_rate": 0.00016526315789473686, + "loss": 5.6972, + "step": 1570 + }, + { + "epoch": 0.08268421052631579, + "grad_norm": 0.9609375, + "learning_rate": 0.00016536842105263158, + "loss": 5.5872, + "step": 1571 + }, + { + "epoch": 0.08273684210526316, + "grad_norm": 0.68359375, + "learning_rate": 0.0001654736842105263, + "loss": 5.5974, + "step": 1572 + }, + { + "epoch": 0.08278947368421052, + "grad_norm": 0.921875, + "learning_rate": 0.00016557894736842106, + "loss": 5.3978, + "step": 1573 + }, + { + "epoch": 0.0828421052631579, + "grad_norm": 0.6796875, + "learning_rate": 0.0001656842105263158, + "loss": 5.7546, + "step": 1574 + }, + { + "epoch": 0.08289473684210526, + "grad_norm": 0.703125, + "learning_rate": 0.00016578947368421054, + "loss": 5.4319, + "step": 1575 + }, + { + "epoch": 0.08294736842105263, + "grad_norm": 0.78515625, + "learning_rate": 0.00016589473684210527, + "loss": 5.6878, + "step": 1576 + }, + { + "epoch": 0.083, + "grad_norm": 0.8046875, + "learning_rate": 0.000166, + "loss": 5.8581, + "step": 1577 + }, + { + "epoch": 0.08305263157894736, + "grad_norm": 1.1328125, + "learning_rate": 0.00016610526315789475, + "loss": 5.8669, + "step": 1578 + }, + { + "epoch": 0.08310526315789474, + "grad_norm": 0.859375, + "learning_rate": 0.00016621052631578948, + "loss": 5.7463, + "step": 1579 + }, + { + "epoch": 0.08315789473684211, + "grad_norm": 0.7265625, + "learning_rate": 0.00016631578947368423, + "loss": 5.7113, + "step": 1580 + }, + { + "epoch": 0.08321052631578947, + "grad_norm": 0.8515625, + "learning_rate": 0.00016642105263157896, + "loss": 5.495, + "step": 1581 + }, + { + "epoch": 0.08326315789473684, + "grad_norm": 0.7421875, + "learning_rate": 0.00016652631578947368, + "loss": 5.5796, + "step": 1582 + }, + { + "epoch": 0.0833157894736842, + "grad_norm": 0.64453125, + "learning_rate": 0.00016663157894736844, + "loss": 5.6909, + "step": 1583 + }, + { + "epoch": 0.08336842105263158, + "grad_norm": 0.75, + "learning_rate": 0.00016673684210526316, + "loss": 5.642, + "step": 1584 + }, + { + "epoch": 0.08342105263157895, + "grad_norm": 0.62109375, + "learning_rate": 0.00016684210526315792, + "loss": 5.6347, + "step": 1585 + }, + { + "epoch": 0.08347368421052631, + "grad_norm": 0.67578125, + "learning_rate": 0.00016694736842105262, + "loss": 5.645, + "step": 1586 + }, + { + "epoch": 0.08352631578947368, + "grad_norm": 0.71875, + "learning_rate": 0.00016705263157894737, + "loss": 5.4759, + "step": 1587 + }, + { + "epoch": 0.08357894736842106, + "grad_norm": 0.6875, + "learning_rate": 0.00016715789473684212, + "loss": 5.82, + "step": 1588 + }, + { + "epoch": 0.08363157894736842, + "grad_norm": 0.7890625, + "learning_rate": 0.00016726315789473685, + "loss": 5.6987, + "step": 1589 + }, + { + "epoch": 0.08368421052631579, + "grad_norm": 0.796875, + "learning_rate": 0.0001673684210526316, + "loss": 5.6396, + "step": 1590 + }, + { + "epoch": 0.08373684210526315, + "grad_norm": 0.91796875, + "learning_rate": 0.0001674736842105263, + "loss": 5.7371, + "step": 1591 + }, + { + "epoch": 0.08378947368421052, + "grad_norm": 0.82421875, + "learning_rate": 0.00016757894736842106, + "loss": 5.5022, + "step": 1592 + }, + { + "epoch": 0.0838421052631579, + "grad_norm": 1.0390625, + "learning_rate": 0.0001676842105263158, + "loss": 5.7168, + "step": 1593 + }, + { + "epoch": 0.08389473684210526, + "grad_norm": 0.87890625, + "learning_rate": 0.00016778947368421054, + "loss": 5.4363, + "step": 1594 + }, + { + "epoch": 0.08394736842105263, + "grad_norm": 1.0078125, + "learning_rate": 0.00016789473684210526, + "loss": 5.4956, + "step": 1595 + }, + { + "epoch": 0.084, + "grad_norm": 1.125, + "learning_rate": 0.000168, + "loss": 5.8613, + "step": 1596 + }, + { + "epoch": 0.08405263157894736, + "grad_norm": 1.0546875, + "learning_rate": 0.00016810526315789474, + "loss": 5.5569, + "step": 1597 + }, + { + "epoch": 0.08410526315789474, + "grad_norm": 0.97265625, + "learning_rate": 0.0001682105263157895, + "loss": 5.5807, + "step": 1598 + }, + { + "epoch": 0.08415789473684211, + "grad_norm": 0.6796875, + "learning_rate": 0.00016831578947368422, + "loss": 5.794, + "step": 1599 + }, + { + "epoch": 0.08421052631578947, + "grad_norm": 0.75, + "learning_rate": 0.00016842105263157895, + "loss": 5.6449, + "step": 1600 + }, + { + "epoch": 0.08426315789473685, + "grad_norm": 0.90234375, + "learning_rate": 0.00016852631578947368, + "loss": 5.5012, + "step": 1601 + }, + { + "epoch": 0.0843157894736842, + "grad_norm": 0.83984375, + "learning_rate": 0.00016863157894736843, + "loss": 5.1441, + "step": 1602 + }, + { + "epoch": 0.08436842105263158, + "grad_norm": 0.92578125, + "learning_rate": 0.00016873684210526316, + "loss": 5.457, + "step": 1603 + }, + { + "epoch": 0.08442105263157895, + "grad_norm": 1.0703125, + "learning_rate": 0.0001688421052631579, + "loss": 5.6292, + "step": 1604 + }, + { + "epoch": 0.08447368421052631, + "grad_norm": 0.69921875, + "learning_rate": 0.00016894736842105264, + "loss": 5.5986, + "step": 1605 + }, + { + "epoch": 0.08452631578947369, + "grad_norm": 0.890625, + "learning_rate": 0.00016905263157894736, + "loss": 5.4978, + "step": 1606 + }, + { + "epoch": 0.08457894736842106, + "grad_norm": 0.72265625, + "learning_rate": 0.00016915789473684212, + "loss": 5.6764, + "step": 1607 + }, + { + "epoch": 0.08463157894736842, + "grad_norm": 0.78515625, + "learning_rate": 0.00016926315789473684, + "loss": 5.6186, + "step": 1608 + }, + { + "epoch": 0.08468421052631579, + "grad_norm": 0.81640625, + "learning_rate": 0.00016936842105263157, + "loss": 5.8602, + "step": 1609 + }, + { + "epoch": 0.08473684210526315, + "grad_norm": 0.8359375, + "learning_rate": 0.00016947368421052633, + "loss": 5.7539, + "step": 1610 + }, + { + "epoch": 0.08478947368421053, + "grad_norm": 0.8984375, + "learning_rate": 0.00016957894736842105, + "loss": 5.7803, + "step": 1611 + }, + { + "epoch": 0.0848421052631579, + "grad_norm": 0.8046875, + "learning_rate": 0.0001696842105263158, + "loss": 5.5246, + "step": 1612 + }, + { + "epoch": 0.08489473684210526, + "grad_norm": 0.8046875, + "learning_rate": 0.00016978947368421053, + "loss": 5.4807, + "step": 1613 + }, + { + "epoch": 0.08494736842105263, + "grad_norm": 0.81640625, + "learning_rate": 0.00016989473684210526, + "loss": 5.5537, + "step": 1614 + }, + { + "epoch": 0.085, + "grad_norm": 0.93359375, + "learning_rate": 0.00017, + "loss": 5.4136, + "step": 1615 + }, + { + "epoch": 0.08505263157894737, + "grad_norm": 0.7578125, + "learning_rate": 0.00017010526315789474, + "loss": 5.4368, + "step": 1616 + }, + { + "epoch": 0.08510526315789474, + "grad_norm": 0.78125, + "learning_rate": 0.0001702105263157895, + "loss": 5.5237, + "step": 1617 + }, + { + "epoch": 0.0851578947368421, + "grad_norm": 0.8203125, + "learning_rate": 0.00017031578947368422, + "loss": 5.1559, + "step": 1618 + }, + { + "epoch": 0.08521052631578947, + "grad_norm": 0.671875, + "learning_rate": 0.00017042105263157895, + "loss": 5.6273, + "step": 1619 + }, + { + "epoch": 0.08526315789473685, + "grad_norm": 0.63671875, + "learning_rate": 0.0001705263157894737, + "loss": 5.4136, + "step": 1620 + }, + { + "epoch": 0.0853157894736842, + "grad_norm": 0.7890625, + "learning_rate": 0.00017063157894736843, + "loss": 5.7641, + "step": 1621 + }, + { + "epoch": 0.08536842105263158, + "grad_norm": 0.671875, + "learning_rate": 0.00017073684210526318, + "loss": 5.5558, + "step": 1622 + }, + { + "epoch": 0.08542105263157895, + "grad_norm": 0.86328125, + "learning_rate": 0.00017084210526315788, + "loss": 5.6348, + "step": 1623 + }, + { + "epoch": 0.08547368421052631, + "grad_norm": 0.71484375, + "learning_rate": 0.00017094736842105263, + "loss": 5.6714, + "step": 1624 + }, + { + "epoch": 0.08552631578947369, + "grad_norm": 0.71875, + "learning_rate": 0.00017105263157894739, + "loss": 5.4099, + "step": 1625 + }, + { + "epoch": 0.08557894736842105, + "grad_norm": 0.890625, + "learning_rate": 0.0001711578947368421, + "loss": 5.0105, + "step": 1626 + }, + { + "epoch": 0.08563157894736842, + "grad_norm": 0.74609375, + "learning_rate": 0.00017126315789473687, + "loss": 5.6407, + "step": 1627 + }, + { + "epoch": 0.0856842105263158, + "grad_norm": 0.79296875, + "learning_rate": 0.00017136842105263157, + "loss": 5.5059, + "step": 1628 + }, + { + "epoch": 0.08573684210526315, + "grad_norm": 0.68359375, + "learning_rate": 0.00017147368421052632, + "loss": 5.7214, + "step": 1629 + }, + { + "epoch": 0.08578947368421053, + "grad_norm": 0.83984375, + "learning_rate": 0.00017157894736842107, + "loss": 5.9169, + "step": 1630 + }, + { + "epoch": 0.0858421052631579, + "grad_norm": 0.7734375, + "learning_rate": 0.0001716842105263158, + "loss": 5.5501, + "step": 1631 + }, + { + "epoch": 0.08589473684210526, + "grad_norm": 0.6796875, + "learning_rate": 0.00017178947368421055, + "loss": 5.8502, + "step": 1632 + }, + { + "epoch": 0.08594736842105263, + "grad_norm": 0.890625, + "learning_rate": 0.00017189473684210525, + "loss": 5.5477, + "step": 1633 + }, + { + "epoch": 0.086, + "grad_norm": 0.91015625, + "learning_rate": 0.000172, + "loss": 5.5143, + "step": 1634 + }, + { + "epoch": 0.08605263157894737, + "grad_norm": 0.78125, + "learning_rate": 0.00017210526315789476, + "loss": 5.4439, + "step": 1635 + }, + { + "epoch": 0.08610526315789474, + "grad_norm": 0.8203125, + "learning_rate": 0.0001722105263157895, + "loss": 5.5142, + "step": 1636 + }, + { + "epoch": 0.0861578947368421, + "grad_norm": 0.83203125, + "learning_rate": 0.0001723157894736842, + "loss": 5.8203, + "step": 1637 + }, + { + "epoch": 0.08621052631578947, + "grad_norm": 0.76953125, + "learning_rate": 0.00017242105263157894, + "loss": 5.5661, + "step": 1638 + }, + { + "epoch": 0.08626315789473685, + "grad_norm": 0.8046875, + "learning_rate": 0.0001725263157894737, + "loss": 5.4703, + "step": 1639 + }, + { + "epoch": 0.0863157894736842, + "grad_norm": 0.8671875, + "learning_rate": 0.00017263157894736842, + "loss": 5.4786, + "step": 1640 + }, + { + "epoch": 0.08636842105263158, + "grad_norm": 0.92578125, + "learning_rate": 0.00017273684210526317, + "loss": 5.6776, + "step": 1641 + }, + { + "epoch": 0.08642105263157895, + "grad_norm": 0.88671875, + "learning_rate": 0.0001728421052631579, + "loss": 5.4898, + "step": 1642 + }, + { + "epoch": 0.08647368421052631, + "grad_norm": 0.8203125, + "learning_rate": 0.00017294736842105263, + "loss": 5.3168, + "step": 1643 + }, + { + "epoch": 0.08652631578947369, + "grad_norm": 0.75390625, + "learning_rate": 0.00017305263157894738, + "loss": 5.8173, + "step": 1644 + }, + { + "epoch": 0.08657894736842105, + "grad_norm": 0.83203125, + "learning_rate": 0.0001731578947368421, + "loss": 5.7464, + "step": 1645 + }, + { + "epoch": 0.08663157894736842, + "grad_norm": 0.69921875, + "learning_rate": 0.00017326315789473686, + "loss": 5.5356, + "step": 1646 + }, + { + "epoch": 0.0866842105263158, + "grad_norm": 1.1875, + "learning_rate": 0.0001733684210526316, + "loss": 5.6604, + "step": 1647 + }, + { + "epoch": 0.08673684210526315, + "grad_norm": 0.8671875, + "learning_rate": 0.00017347368421052631, + "loss": 5.7196, + "step": 1648 + }, + { + "epoch": 0.08678947368421053, + "grad_norm": 0.7734375, + "learning_rate": 0.00017357894736842107, + "loss": 5.7157, + "step": 1649 + }, + { + "epoch": 0.0868421052631579, + "grad_norm": 1.0234375, + "learning_rate": 0.0001736842105263158, + "loss": 4.9407, + "step": 1650 + }, + { + "epoch": 0.08689473684210526, + "grad_norm": 0.73828125, + "learning_rate": 0.00017378947368421052, + "loss": 5.6696, + "step": 1651 + }, + { + "epoch": 0.08694736842105263, + "grad_norm": 0.73828125, + "learning_rate": 0.00017389473684210527, + "loss": 5.6068, + "step": 1652 + }, + { + "epoch": 0.087, + "grad_norm": 0.81640625, + "learning_rate": 0.000174, + "loss": 5.4593, + "step": 1653 + }, + { + "epoch": 0.08705263157894737, + "grad_norm": 1.1953125, + "learning_rate": 0.00017410526315789475, + "loss": 5.607, + "step": 1654 + }, + { + "epoch": 0.08710526315789474, + "grad_norm": 0.74609375, + "learning_rate": 0.00017421052631578948, + "loss": 5.8611, + "step": 1655 + }, + { + "epoch": 0.0871578947368421, + "grad_norm": 0.97265625, + "learning_rate": 0.0001743157894736842, + "loss": 5.6372, + "step": 1656 + }, + { + "epoch": 0.08721052631578947, + "grad_norm": 0.8828125, + "learning_rate": 0.00017442105263157896, + "loss": 5.9104, + "step": 1657 + }, + { + "epoch": 0.08726315789473685, + "grad_norm": 1.5625, + "learning_rate": 0.0001745263157894737, + "loss": 5.4462, + "step": 1658 + }, + { + "epoch": 0.08731578947368421, + "grad_norm": 0.64453125, + "learning_rate": 0.00017463157894736844, + "loss": 5.4893, + "step": 1659 + }, + { + "epoch": 0.08736842105263158, + "grad_norm": 0.63671875, + "learning_rate": 0.00017473684210526317, + "loss": 5.6934, + "step": 1660 + }, + { + "epoch": 0.08742105263157894, + "grad_norm": 0.72265625, + "learning_rate": 0.0001748421052631579, + "loss": 5.7343, + "step": 1661 + }, + { + "epoch": 0.08747368421052631, + "grad_norm": 0.8671875, + "learning_rate": 0.00017494736842105265, + "loss": 5.4929, + "step": 1662 + }, + { + "epoch": 0.08752631578947369, + "grad_norm": 0.69140625, + "learning_rate": 0.00017505263157894738, + "loss": 5.6626, + "step": 1663 + }, + { + "epoch": 0.08757894736842105, + "grad_norm": 0.9296875, + "learning_rate": 0.00017515789473684213, + "loss": 5.5223, + "step": 1664 + }, + { + "epoch": 0.08763157894736842, + "grad_norm": 0.6796875, + "learning_rate": 0.00017526315789473683, + "loss": 5.6476, + "step": 1665 + }, + { + "epoch": 0.0876842105263158, + "grad_norm": 0.68359375, + "learning_rate": 0.00017536842105263158, + "loss": 5.5578, + "step": 1666 + }, + { + "epoch": 0.08773684210526315, + "grad_norm": 0.84765625, + "learning_rate": 0.00017547368421052634, + "loss": 5.5898, + "step": 1667 + }, + { + "epoch": 0.08778947368421053, + "grad_norm": 0.984375, + "learning_rate": 0.00017557894736842106, + "loss": 5.7108, + "step": 1668 + }, + { + "epoch": 0.08784210526315789, + "grad_norm": 0.6640625, + "learning_rate": 0.00017568421052631582, + "loss": 5.5844, + "step": 1669 + }, + { + "epoch": 0.08789473684210526, + "grad_norm": 0.953125, + "learning_rate": 0.00017578947368421052, + "loss": 5.4556, + "step": 1670 + }, + { + "epoch": 0.08794736842105264, + "grad_norm": 0.9765625, + "learning_rate": 0.00017589473684210527, + "loss": 5.435, + "step": 1671 + }, + { + "epoch": 0.088, + "grad_norm": 0.9609375, + "learning_rate": 0.00017600000000000002, + "loss": 5.2845, + "step": 1672 + }, + { + "epoch": 0.08805263157894737, + "grad_norm": 0.671875, + "learning_rate": 0.00017610526315789475, + "loss": 5.5039, + "step": 1673 + }, + { + "epoch": 0.08810526315789474, + "grad_norm": 0.73828125, + "learning_rate": 0.0001762105263157895, + "loss": 5.591, + "step": 1674 + }, + { + "epoch": 0.0881578947368421, + "grad_norm": 0.71484375, + "learning_rate": 0.0001763157894736842, + "loss": 5.5848, + "step": 1675 + }, + { + "epoch": 0.08821052631578948, + "grad_norm": 0.70703125, + "learning_rate": 0.00017642105263157896, + "loss": 5.5449, + "step": 1676 + }, + { + "epoch": 0.08826315789473685, + "grad_norm": 0.78515625, + "learning_rate": 0.00017652631578947368, + "loss": 5.5226, + "step": 1677 + }, + { + "epoch": 0.08831578947368421, + "grad_norm": 0.796875, + "learning_rate": 0.00017663157894736844, + "loss": 5.402, + "step": 1678 + }, + { + "epoch": 0.08836842105263158, + "grad_norm": 0.828125, + "learning_rate": 0.00017673684210526316, + "loss": 5.5666, + "step": 1679 + }, + { + "epoch": 0.08842105263157894, + "grad_norm": 0.91015625, + "learning_rate": 0.0001768421052631579, + "loss": 5.7074, + "step": 1680 + }, + { + "epoch": 0.08847368421052632, + "grad_norm": 0.99609375, + "learning_rate": 0.00017694736842105264, + "loss": 5.5012, + "step": 1681 + }, + { + "epoch": 0.08852631578947369, + "grad_norm": 0.7109375, + "learning_rate": 0.00017705263157894737, + "loss": 5.5614, + "step": 1682 + }, + { + "epoch": 0.08857894736842105, + "grad_norm": 0.6640625, + "learning_rate": 0.00017715789473684212, + "loss": 5.514, + "step": 1683 + }, + { + "epoch": 0.08863157894736842, + "grad_norm": 0.78125, + "learning_rate": 0.00017726315789473685, + "loss": 5.6137, + "step": 1684 + }, + { + "epoch": 0.0886842105263158, + "grad_norm": 0.8828125, + "learning_rate": 0.00017736842105263158, + "loss": 5.4599, + "step": 1685 + }, + { + "epoch": 0.08873684210526316, + "grad_norm": 0.86328125, + "learning_rate": 0.00017747368421052633, + "loss": 5.6457, + "step": 1686 + }, + { + "epoch": 0.08878947368421053, + "grad_norm": 0.79296875, + "learning_rate": 0.00017757894736842106, + "loss": 5.3209, + "step": 1687 + }, + { + "epoch": 0.08884210526315789, + "grad_norm": 0.8828125, + "learning_rate": 0.0001776842105263158, + "loss": 5.3301, + "step": 1688 + }, + { + "epoch": 0.08889473684210526, + "grad_norm": 0.80859375, + "learning_rate": 0.00017778947368421054, + "loss": 5.5133, + "step": 1689 + }, + { + "epoch": 0.08894736842105264, + "grad_norm": 0.734375, + "learning_rate": 0.00017789473684210526, + "loss": 5.6015, + "step": 1690 + }, + { + "epoch": 0.089, + "grad_norm": 0.859375, + "learning_rate": 0.00017800000000000002, + "loss": 5.3407, + "step": 1691 + }, + { + "epoch": 0.08905263157894737, + "grad_norm": 1.0078125, + "learning_rate": 0.00017810526315789474, + "loss": 5.7162, + "step": 1692 + }, + { + "epoch": 0.08910526315789474, + "grad_norm": 0.765625, + "learning_rate": 0.00017821052631578947, + "loss": 5.5275, + "step": 1693 + }, + { + "epoch": 0.0891578947368421, + "grad_norm": 0.8203125, + "learning_rate": 0.00017831578947368422, + "loss": 5.5254, + "step": 1694 + }, + { + "epoch": 0.08921052631578948, + "grad_norm": 0.7421875, + "learning_rate": 0.00017842105263157895, + "loss": 5.4258, + "step": 1695 + }, + { + "epoch": 0.08926315789473684, + "grad_norm": 0.734375, + "learning_rate": 0.0001785263157894737, + "loss": 5.763, + "step": 1696 + }, + { + "epoch": 0.08931578947368421, + "grad_norm": 0.68359375, + "learning_rate": 0.00017863157894736843, + "loss": 5.2304, + "step": 1697 + }, + { + "epoch": 0.08936842105263158, + "grad_norm": 0.796875, + "learning_rate": 0.00017873684210526316, + "loss": 5.564, + "step": 1698 + }, + { + "epoch": 0.08942105263157894, + "grad_norm": 0.8359375, + "learning_rate": 0.0001788421052631579, + "loss": 5.4373, + "step": 1699 + }, + { + "epoch": 0.08947368421052632, + "grad_norm": 0.83984375, + "learning_rate": 0.00017894736842105264, + "loss": 5.521, + "step": 1700 + }, + { + "epoch": 0.08952631578947369, + "grad_norm": 0.921875, + "learning_rate": 0.0001790526315789474, + "loss": 5.5848, + "step": 1701 + }, + { + "epoch": 0.08957894736842105, + "grad_norm": 0.99609375, + "learning_rate": 0.00017915789473684212, + "loss": 5.3061, + "step": 1702 + }, + { + "epoch": 0.08963157894736842, + "grad_norm": 0.97265625, + "learning_rate": 0.00017926315789473684, + "loss": 5.3317, + "step": 1703 + }, + { + "epoch": 0.08968421052631578, + "grad_norm": 0.8359375, + "learning_rate": 0.0001793684210526316, + "loss": 5.4268, + "step": 1704 + }, + { + "epoch": 0.08973684210526316, + "grad_norm": 0.8046875, + "learning_rate": 0.00017947368421052632, + "loss": 5.3299, + "step": 1705 + }, + { + "epoch": 0.08978947368421053, + "grad_norm": 0.58984375, + "learning_rate": 0.00017957894736842108, + "loss": 5.6251, + "step": 1706 + }, + { + "epoch": 0.08984210526315789, + "grad_norm": 0.72265625, + "learning_rate": 0.00017968421052631578, + "loss": 5.4788, + "step": 1707 + }, + { + "epoch": 0.08989473684210526, + "grad_norm": 0.734375, + "learning_rate": 0.00017978947368421053, + "loss": 5.7212, + "step": 1708 + }, + { + "epoch": 0.08994736842105264, + "grad_norm": 0.82421875, + "learning_rate": 0.00017989473684210528, + "loss": 5.625, + "step": 1709 + }, + { + "epoch": 0.09, + "grad_norm": 0.69140625, + "learning_rate": 0.00018, + "loss": 5.3923, + "step": 1710 + }, + { + "epoch": 0.09005263157894737, + "grad_norm": 0.6328125, + "learning_rate": 0.00018010526315789477, + "loss": 5.6889, + "step": 1711 + }, + { + "epoch": 0.09010526315789473, + "grad_norm": 0.6484375, + "learning_rate": 0.00018021052631578946, + "loss": 5.871, + "step": 1712 + }, + { + "epoch": 0.0901578947368421, + "grad_norm": 0.74609375, + "learning_rate": 0.00018031578947368422, + "loss": 5.5637, + "step": 1713 + }, + { + "epoch": 0.09021052631578948, + "grad_norm": 0.98046875, + "learning_rate": 0.00018042105263157894, + "loss": 5.3444, + "step": 1714 + }, + { + "epoch": 0.09026315789473684, + "grad_norm": 0.7890625, + "learning_rate": 0.0001805263157894737, + "loss": 5.7396, + "step": 1715 + }, + { + "epoch": 0.09031578947368421, + "grad_norm": 0.7890625, + "learning_rate": 0.00018063157894736845, + "loss": 5.3234, + "step": 1716 + }, + { + "epoch": 0.09036842105263158, + "grad_norm": 0.90625, + "learning_rate": 0.00018073684210526315, + "loss": 5.4576, + "step": 1717 + }, + { + "epoch": 0.09042105263157894, + "grad_norm": 0.98828125, + "learning_rate": 0.0001808421052631579, + "loss": 5.1353, + "step": 1718 + }, + { + "epoch": 0.09047368421052632, + "grad_norm": 0.95703125, + "learning_rate": 0.00018094736842105263, + "loss": 5.648, + "step": 1719 + }, + { + "epoch": 0.09052631578947369, + "grad_norm": 0.8203125, + "learning_rate": 0.00018105263157894739, + "loss": 5.3508, + "step": 1720 + }, + { + "epoch": 0.09057894736842105, + "grad_norm": 0.65625, + "learning_rate": 0.0001811578947368421, + "loss": 5.5716, + "step": 1721 + }, + { + "epoch": 0.09063157894736842, + "grad_norm": 0.703125, + "learning_rate": 0.00018126315789473684, + "loss": 5.7006, + "step": 1722 + }, + { + "epoch": 0.09068421052631578, + "grad_norm": 0.66015625, + "learning_rate": 0.0001813684210526316, + "loss": 5.5308, + "step": 1723 + }, + { + "epoch": 0.09073684210526316, + "grad_norm": 0.890625, + "learning_rate": 0.00018147368421052632, + "loss": 4.8695, + "step": 1724 + }, + { + "epoch": 0.09078947368421053, + "grad_norm": 0.57421875, + "learning_rate": 0.00018157894736842107, + "loss": 5.5002, + "step": 1725 + }, + { + "epoch": 0.09084210526315789, + "grad_norm": 0.765625, + "learning_rate": 0.0001816842105263158, + "loss": 5.1907, + "step": 1726 + }, + { + "epoch": 0.09089473684210526, + "grad_norm": 0.8125, + "learning_rate": 0.00018178947368421053, + "loss": 5.6182, + "step": 1727 + }, + { + "epoch": 0.09094736842105264, + "grad_norm": 0.64453125, + "learning_rate": 0.00018189473684210528, + "loss": 5.7881, + "step": 1728 + }, + { + "epoch": 0.091, + "grad_norm": 0.76171875, + "learning_rate": 0.000182, + "loss": 5.3998, + "step": 1729 + }, + { + "epoch": 0.09105263157894737, + "grad_norm": 0.8046875, + "learning_rate": 0.00018210526315789476, + "loss": 5.5903, + "step": 1730 + }, + { + "epoch": 0.09110526315789473, + "grad_norm": 0.90625, + "learning_rate": 0.00018221052631578949, + "loss": 5.2496, + "step": 1731 + }, + { + "epoch": 0.0911578947368421, + "grad_norm": 0.84375, + "learning_rate": 0.0001823157894736842, + "loss": 5.6115, + "step": 1732 + }, + { + "epoch": 0.09121052631578948, + "grad_norm": 0.7109375, + "learning_rate": 0.00018242105263157897, + "loss": 5.4887, + "step": 1733 + }, + { + "epoch": 0.09126315789473684, + "grad_norm": 0.7421875, + "learning_rate": 0.0001825263157894737, + "loss": 5.5536, + "step": 1734 + }, + { + "epoch": 0.09131578947368421, + "grad_norm": 0.87109375, + "learning_rate": 0.00018263157894736842, + "loss": 5.5911, + "step": 1735 + }, + { + "epoch": 0.09136842105263158, + "grad_norm": 0.86328125, + "learning_rate": 0.00018273684210526317, + "loss": 5.4876, + "step": 1736 + }, + { + "epoch": 0.09142105263157894, + "grad_norm": 0.79296875, + "learning_rate": 0.0001828421052631579, + "loss": 5.6515, + "step": 1737 + }, + { + "epoch": 0.09147368421052632, + "grad_norm": 0.9609375, + "learning_rate": 0.00018294736842105265, + "loss": 5.457, + "step": 1738 + }, + { + "epoch": 0.09152631578947368, + "grad_norm": 0.6875, + "learning_rate": 0.00018305263157894738, + "loss": 5.538, + "step": 1739 + }, + { + "epoch": 0.09157894736842105, + "grad_norm": 0.60546875, + "learning_rate": 0.0001831578947368421, + "loss": 5.6042, + "step": 1740 + }, + { + "epoch": 0.09163157894736843, + "grad_norm": 0.734375, + "learning_rate": 0.00018326315789473686, + "loss": 5.5869, + "step": 1741 + }, + { + "epoch": 0.09168421052631578, + "grad_norm": 0.78515625, + "learning_rate": 0.0001833684210526316, + "loss": 5.1819, + "step": 1742 + }, + { + "epoch": 0.09173684210526316, + "grad_norm": 0.84375, + "learning_rate": 0.00018347368421052634, + "loss": 5.3732, + "step": 1743 + }, + { + "epoch": 0.09178947368421053, + "grad_norm": 0.74609375, + "learning_rate": 0.00018357894736842104, + "loss": 5.31, + "step": 1744 + }, + { + "epoch": 0.09184210526315789, + "grad_norm": 0.8828125, + "learning_rate": 0.0001836842105263158, + "loss": 5.7988, + "step": 1745 + }, + { + "epoch": 0.09189473684210527, + "grad_norm": 0.9296875, + "learning_rate": 0.00018378947368421055, + "loss": 5.6684, + "step": 1746 + }, + { + "epoch": 0.09194736842105262, + "grad_norm": 0.79296875, + "learning_rate": 0.00018389473684210527, + "loss": 5.5172, + "step": 1747 + }, + { + "epoch": 0.092, + "grad_norm": 0.83203125, + "learning_rate": 0.00018400000000000003, + "loss": 5.4635, + "step": 1748 + }, + { + "epoch": 0.09205263157894737, + "grad_norm": 0.7734375, + "learning_rate": 0.00018410526315789473, + "loss": 5.4515, + "step": 1749 + }, + { + "epoch": 0.09210526315789473, + "grad_norm": 0.80078125, + "learning_rate": 0.00018421052631578948, + "loss": 5.6083, + "step": 1750 + }, + { + "epoch": 0.0921578947368421, + "grad_norm": 0.8984375, + "learning_rate": 0.0001843157894736842, + "loss": 5.4968, + "step": 1751 + }, + { + "epoch": 0.09221052631578948, + "grad_norm": 0.66015625, + "learning_rate": 0.00018442105263157896, + "loss": 5.5576, + "step": 1752 + }, + { + "epoch": 0.09226315789473684, + "grad_norm": 0.7578125, + "learning_rate": 0.00018452631578947371, + "loss": 5.5855, + "step": 1753 + }, + { + "epoch": 0.09231578947368421, + "grad_norm": 0.81640625, + "learning_rate": 0.00018463157894736841, + "loss": 5.684, + "step": 1754 + }, + { + "epoch": 0.09236842105263159, + "grad_norm": 0.91796875, + "learning_rate": 0.00018473684210526317, + "loss": 5.548, + "step": 1755 + }, + { + "epoch": 0.09242105263157895, + "grad_norm": 0.83203125, + "learning_rate": 0.0001848421052631579, + "loss": 5.6469, + "step": 1756 + }, + { + "epoch": 0.09247368421052632, + "grad_norm": 0.76171875, + "learning_rate": 0.00018494736842105265, + "loss": 5.3584, + "step": 1757 + }, + { + "epoch": 0.09252631578947368, + "grad_norm": 0.859375, + "learning_rate": 0.00018505263157894737, + "loss": 5.632, + "step": 1758 + }, + { + "epoch": 0.09257894736842105, + "grad_norm": 1.046875, + "learning_rate": 0.0001851578947368421, + "loss": 5.47, + "step": 1759 + }, + { + "epoch": 0.09263157894736843, + "grad_norm": 0.83203125, + "learning_rate": 0.00018526315789473685, + "loss": 5.1487, + "step": 1760 + }, + { + "epoch": 0.09268421052631579, + "grad_norm": 0.69140625, + "learning_rate": 0.00018536842105263158, + "loss": 5.572, + "step": 1761 + }, + { + "epoch": 0.09273684210526316, + "grad_norm": 1.203125, + "learning_rate": 0.00018547368421052633, + "loss": 5.6715, + "step": 1762 + }, + { + "epoch": 0.09278947368421053, + "grad_norm": 0.75390625, + "learning_rate": 0.00018557894736842106, + "loss": 5.4839, + "step": 1763 + }, + { + "epoch": 0.09284210526315789, + "grad_norm": 0.640625, + "learning_rate": 0.0001856842105263158, + "loss": 5.6358, + "step": 1764 + }, + { + "epoch": 0.09289473684210527, + "grad_norm": 0.875, + "learning_rate": 0.00018578947368421054, + "loss": 5.5321, + "step": 1765 + }, + { + "epoch": 0.09294736842105263, + "grad_norm": 0.8515625, + "learning_rate": 0.00018589473684210527, + "loss": 5.293, + "step": 1766 + }, + { + "epoch": 0.093, + "grad_norm": 0.64453125, + "learning_rate": 0.00018600000000000002, + "loss": 5.5573, + "step": 1767 + }, + { + "epoch": 0.09305263157894737, + "grad_norm": 0.69140625, + "learning_rate": 0.00018610526315789475, + "loss": 5.5786, + "step": 1768 + }, + { + "epoch": 0.09310526315789473, + "grad_norm": 0.78125, + "learning_rate": 0.00018621052631578947, + "loss": 5.4564, + "step": 1769 + }, + { + "epoch": 0.0931578947368421, + "grad_norm": 0.66796875, + "learning_rate": 0.00018631578947368423, + "loss": 5.3343, + "step": 1770 + }, + { + "epoch": 0.09321052631578948, + "grad_norm": 0.93359375, + "learning_rate": 0.00018642105263157896, + "loss": 5.4901, + "step": 1771 + }, + { + "epoch": 0.09326315789473684, + "grad_norm": 0.8125, + "learning_rate": 0.00018652631578947368, + "loss": 5.3685, + "step": 1772 + }, + { + "epoch": 0.09331578947368421, + "grad_norm": 1.0546875, + "learning_rate": 0.00018663157894736844, + "loss": 5.252, + "step": 1773 + }, + { + "epoch": 0.09336842105263157, + "grad_norm": 0.82421875, + "learning_rate": 0.00018673684210526316, + "loss": 5.3395, + "step": 1774 + }, + { + "epoch": 0.09342105263157895, + "grad_norm": 0.71875, + "learning_rate": 0.00018684210526315792, + "loss": 5.3826, + "step": 1775 + }, + { + "epoch": 0.09347368421052632, + "grad_norm": 0.74609375, + "learning_rate": 0.00018694736842105264, + "loss": 5.7904, + "step": 1776 + }, + { + "epoch": 0.09352631578947368, + "grad_norm": 0.828125, + "learning_rate": 0.00018705263157894737, + "loss": 5.498, + "step": 1777 + }, + { + "epoch": 0.09357894736842105, + "grad_norm": 0.6484375, + "learning_rate": 0.00018715789473684212, + "loss": 5.4982, + "step": 1778 + }, + { + "epoch": 0.09363157894736843, + "grad_norm": 0.87890625, + "learning_rate": 0.00018726315789473685, + "loss": 5.4505, + "step": 1779 + }, + { + "epoch": 0.09368421052631579, + "grad_norm": 1.171875, + "learning_rate": 0.0001873684210526316, + "loss": 5.1449, + "step": 1780 + }, + { + "epoch": 0.09373684210526316, + "grad_norm": 0.81640625, + "learning_rate": 0.00018747368421052633, + "loss": 5.6699, + "step": 1781 + }, + { + "epoch": 0.09378947368421052, + "grad_norm": 1.03125, + "learning_rate": 0.00018757894736842106, + "loss": 5.4749, + "step": 1782 + }, + { + "epoch": 0.0938421052631579, + "grad_norm": 0.94921875, + "learning_rate": 0.00018768421052631578, + "loss": 5.7896, + "step": 1783 + }, + { + "epoch": 0.09389473684210527, + "grad_norm": 0.94921875, + "learning_rate": 0.00018778947368421054, + "loss": 5.1023, + "step": 1784 + }, + { + "epoch": 0.09394736842105263, + "grad_norm": 0.6328125, + "learning_rate": 0.0001878947368421053, + "loss": 5.428, + "step": 1785 + }, + { + "epoch": 0.094, + "grad_norm": 0.6953125, + "learning_rate": 0.000188, + "loss": 5.3017, + "step": 1786 + }, + { + "epoch": 0.09405263157894737, + "grad_norm": 0.7265625, + "learning_rate": 0.00018810526315789474, + "loss": 5.6274, + "step": 1787 + }, + { + "epoch": 0.09410526315789473, + "grad_norm": 0.984375, + "learning_rate": 0.00018821052631578947, + "loss": 5.8032, + "step": 1788 + }, + { + "epoch": 0.09415789473684211, + "grad_norm": 0.87109375, + "learning_rate": 0.00018831578947368422, + "loss": 5.3853, + "step": 1789 + }, + { + "epoch": 0.09421052631578947, + "grad_norm": 0.890625, + "learning_rate": 0.00018842105263157898, + "loss": 5.736, + "step": 1790 + }, + { + "epoch": 0.09426315789473684, + "grad_norm": 1.078125, + "learning_rate": 0.00018852631578947368, + "loss": 5.7141, + "step": 1791 + }, + { + "epoch": 0.09431578947368421, + "grad_norm": 0.66015625, + "learning_rate": 0.00018863157894736843, + "loss": 5.4603, + "step": 1792 + }, + { + "epoch": 0.09436842105263157, + "grad_norm": 0.69921875, + "learning_rate": 0.00018873684210526316, + "loss": 5.4111, + "step": 1793 + }, + { + "epoch": 0.09442105263157895, + "grad_norm": 0.64453125, + "learning_rate": 0.0001888421052631579, + "loss": 5.2925, + "step": 1794 + }, + { + "epoch": 0.09447368421052632, + "grad_norm": 0.78515625, + "learning_rate": 0.00018894736842105266, + "loss": 5.4935, + "step": 1795 + }, + { + "epoch": 0.09452631578947368, + "grad_norm": 0.78125, + "learning_rate": 0.00018905263157894736, + "loss": 5.668, + "step": 1796 + }, + { + "epoch": 0.09457894736842105, + "grad_norm": 0.84375, + "learning_rate": 0.00018915789473684212, + "loss": 5.2176, + "step": 1797 + }, + { + "epoch": 0.09463157894736843, + "grad_norm": 0.87890625, + "learning_rate": 0.00018926315789473684, + "loss": 5.58, + "step": 1798 + }, + { + "epoch": 0.09468421052631579, + "grad_norm": 1.015625, + "learning_rate": 0.0001893684210526316, + "loss": 5.338, + "step": 1799 + }, + { + "epoch": 0.09473684210526316, + "grad_norm": 0.9609375, + "learning_rate": 0.00018947368421052632, + "loss": 5.5591, + "step": 1800 + }, + { + "epoch": 0.09478947368421052, + "grad_norm": 0.6796875, + "learning_rate": 0.00018957894736842105, + "loss": 5.6108, + "step": 1801 + }, + { + "epoch": 0.0948421052631579, + "grad_norm": 0.69921875, + "learning_rate": 0.0001896842105263158, + "loss": 5.5023, + "step": 1802 + }, + { + "epoch": 0.09489473684210527, + "grad_norm": 0.953125, + "learning_rate": 0.00018978947368421053, + "loss": 5.1909, + "step": 1803 + }, + { + "epoch": 0.09494736842105263, + "grad_norm": 0.71875, + "learning_rate": 0.00018989473684210528, + "loss": 5.5543, + "step": 1804 + }, + { + "epoch": 0.095, + "grad_norm": 0.7265625, + "learning_rate": 0.00019, + "loss": 5.2812, + "step": 1805 + }, + { + "epoch": 0.09505263157894737, + "grad_norm": 0.67578125, + "learning_rate": 0.00019010526315789474, + "loss": 5.466, + "step": 1806 + }, + { + "epoch": 0.09510526315789473, + "grad_norm": 0.8515625, + "learning_rate": 0.0001902105263157895, + "loss": 5.0739, + "step": 1807 + }, + { + "epoch": 0.09515789473684211, + "grad_norm": 0.81640625, + "learning_rate": 0.00019031578947368422, + "loss": 5.268, + "step": 1808 + }, + { + "epoch": 0.09521052631578947, + "grad_norm": 0.78515625, + "learning_rate": 0.00019042105263157897, + "loss": 5.4298, + "step": 1809 + }, + { + "epoch": 0.09526315789473684, + "grad_norm": 0.9140625, + "learning_rate": 0.0001905263157894737, + "loss": 5.3723, + "step": 1810 + }, + { + "epoch": 0.09531578947368421, + "grad_norm": 0.69921875, + "learning_rate": 0.00019063157894736842, + "loss": 5.5335, + "step": 1811 + }, + { + "epoch": 0.09536842105263157, + "grad_norm": 0.77734375, + "learning_rate": 0.00019073684210526318, + "loss": 5.6167, + "step": 1812 + }, + { + "epoch": 0.09542105263157895, + "grad_norm": 0.80078125, + "learning_rate": 0.0001908421052631579, + "loss": 5.4103, + "step": 1813 + }, + { + "epoch": 0.09547368421052632, + "grad_norm": 0.75390625, + "learning_rate": 0.00019094736842105263, + "loss": 5.5288, + "step": 1814 + }, + { + "epoch": 0.09552631578947368, + "grad_norm": 0.74609375, + "learning_rate": 0.00019105263157894738, + "loss": 5.3815, + "step": 1815 + }, + { + "epoch": 0.09557894736842105, + "grad_norm": 0.85546875, + "learning_rate": 0.0001911578947368421, + "loss": 5.5394, + "step": 1816 + }, + { + "epoch": 0.09563157894736841, + "grad_norm": 0.83203125, + "learning_rate": 0.00019126315789473686, + "loss": 5.5176, + "step": 1817 + }, + { + "epoch": 0.09568421052631579, + "grad_norm": 0.9296875, + "learning_rate": 0.0001913684210526316, + "loss": 5.6454, + "step": 1818 + }, + { + "epoch": 0.09573684210526316, + "grad_norm": 0.84765625, + "learning_rate": 0.00019147368421052632, + "loss": 5.2819, + "step": 1819 + }, + { + "epoch": 0.09578947368421052, + "grad_norm": 0.67578125, + "learning_rate": 0.00019157894736842104, + "loss": 5.3482, + "step": 1820 + }, + { + "epoch": 0.0958421052631579, + "grad_norm": 0.89453125, + "learning_rate": 0.0001916842105263158, + "loss": 5.4891, + "step": 1821 + }, + { + "epoch": 0.09589473684210527, + "grad_norm": 1.0078125, + "learning_rate": 0.00019178947368421055, + "loss": 5.6931, + "step": 1822 + }, + { + "epoch": 0.09594736842105263, + "grad_norm": 0.7890625, + "learning_rate": 0.00019189473684210528, + "loss": 5.1257, + "step": 1823 + }, + { + "epoch": 0.096, + "grad_norm": 0.80078125, + "learning_rate": 0.000192, + "loss": 5.0857, + "step": 1824 + }, + { + "epoch": 0.09605263157894736, + "grad_norm": 0.765625, + "learning_rate": 0.00019210526315789473, + "loss": 5.1843, + "step": 1825 + }, + { + "epoch": 0.09610526315789474, + "grad_norm": 0.67578125, + "learning_rate": 0.00019221052631578949, + "loss": 5.4707, + "step": 1826 + }, + { + "epoch": 0.09615789473684211, + "grad_norm": 0.859375, + "learning_rate": 0.00019231578947368424, + "loss": 5.6354, + "step": 1827 + }, + { + "epoch": 0.09621052631578947, + "grad_norm": 0.828125, + "learning_rate": 0.00019242105263157894, + "loss": 5.6783, + "step": 1828 + }, + { + "epoch": 0.09626315789473684, + "grad_norm": 0.7109375, + "learning_rate": 0.0001925263157894737, + "loss": 5.3738, + "step": 1829 + }, + { + "epoch": 0.09631578947368422, + "grad_norm": 0.7734375, + "learning_rate": 0.00019263157894736842, + "loss": 5.083, + "step": 1830 + }, + { + "epoch": 0.09636842105263158, + "grad_norm": 0.828125, + "learning_rate": 0.00019273684210526317, + "loss": 5.5609, + "step": 1831 + }, + { + "epoch": 0.09642105263157895, + "grad_norm": 1.015625, + "learning_rate": 0.00019284210526315793, + "loss": 5.1928, + "step": 1832 + }, + { + "epoch": 0.09647368421052632, + "grad_norm": 0.875, + "learning_rate": 0.00019294736842105263, + "loss": 5.332, + "step": 1833 + }, + { + "epoch": 0.09652631578947368, + "grad_norm": 0.859375, + "learning_rate": 0.00019305263157894738, + "loss": 5.4549, + "step": 1834 + }, + { + "epoch": 0.09657894736842106, + "grad_norm": 0.85546875, + "learning_rate": 0.0001931578947368421, + "loss": 5.5409, + "step": 1835 + }, + { + "epoch": 0.09663157894736842, + "grad_norm": 0.6953125, + "learning_rate": 0.00019326315789473686, + "loss": 5.6885, + "step": 1836 + }, + { + "epoch": 0.09668421052631579, + "grad_norm": 0.6875, + "learning_rate": 0.00019336842105263159, + "loss": 5.3739, + "step": 1837 + }, + { + "epoch": 0.09673684210526316, + "grad_norm": 0.7734375, + "learning_rate": 0.0001934736842105263, + "loss": 5.5185, + "step": 1838 + }, + { + "epoch": 0.09678947368421052, + "grad_norm": 0.72265625, + "learning_rate": 0.00019357894736842107, + "loss": 5.5817, + "step": 1839 + }, + { + "epoch": 0.0968421052631579, + "grad_norm": 0.6875, + "learning_rate": 0.0001936842105263158, + "loss": 5.2121, + "step": 1840 + }, + { + "epoch": 0.09689473684210527, + "grad_norm": 0.54296875, + "learning_rate": 0.00019378947368421055, + "loss": 5.5437, + "step": 1841 + }, + { + "epoch": 0.09694736842105263, + "grad_norm": 0.58203125, + "learning_rate": 0.00019389473684210527, + "loss": 5.1828, + "step": 1842 + }, + { + "epoch": 0.097, + "grad_norm": 0.6484375, + "learning_rate": 0.000194, + "loss": 5.2301, + "step": 1843 + }, + { + "epoch": 0.09705263157894736, + "grad_norm": 0.58984375, + "learning_rate": 0.00019410526315789475, + "loss": 5.4379, + "step": 1844 + }, + { + "epoch": 0.09710526315789474, + "grad_norm": 0.72265625, + "learning_rate": 0.00019421052631578948, + "loss": 5.7315, + "step": 1845 + }, + { + "epoch": 0.09715789473684211, + "grad_norm": 0.80078125, + "learning_rate": 0.00019431578947368423, + "loss": 5.4423, + "step": 1846 + }, + { + "epoch": 0.09721052631578947, + "grad_norm": 0.64453125, + "learning_rate": 0.00019442105263157896, + "loss": 5.183, + "step": 1847 + }, + { + "epoch": 0.09726315789473684, + "grad_norm": 0.85546875, + "learning_rate": 0.00019452631578947369, + "loss": 5.617, + "step": 1848 + }, + { + "epoch": 0.09731578947368422, + "grad_norm": 0.76953125, + "learning_rate": 0.00019463157894736844, + "loss": 5.2705, + "step": 1849 + }, + { + "epoch": 0.09736842105263158, + "grad_norm": 0.76171875, + "learning_rate": 0.00019473684210526317, + "loss": 5.1209, + "step": 1850 + }, + { + "epoch": 0.09742105263157895, + "grad_norm": 0.7578125, + "learning_rate": 0.00019484210526315792, + "loss": 5.278, + "step": 1851 + }, + { + "epoch": 0.09747368421052631, + "grad_norm": 0.89453125, + "learning_rate": 0.00019494736842105265, + "loss": 5.2348, + "step": 1852 + }, + { + "epoch": 0.09752631578947368, + "grad_norm": 0.703125, + "learning_rate": 0.00019505263157894737, + "loss": 5.5442, + "step": 1853 + }, + { + "epoch": 0.09757894736842106, + "grad_norm": 0.7421875, + "learning_rate": 0.00019515789473684213, + "loss": 5.535, + "step": 1854 + }, + { + "epoch": 0.09763157894736842, + "grad_norm": 0.8359375, + "learning_rate": 0.00019526315789473685, + "loss": 5.3161, + "step": 1855 + }, + { + "epoch": 0.09768421052631579, + "grad_norm": 0.83984375, + "learning_rate": 0.00019536842105263158, + "loss": 5.6684, + "step": 1856 + }, + { + "epoch": 0.09773684210526316, + "grad_norm": 0.65234375, + "learning_rate": 0.0001954736842105263, + "loss": 5.5735, + "step": 1857 + }, + { + "epoch": 0.09778947368421052, + "grad_norm": 0.7109375, + "learning_rate": 0.00019557894736842106, + "loss": 5.5788, + "step": 1858 + }, + { + "epoch": 0.0978421052631579, + "grad_norm": 0.83984375, + "learning_rate": 0.00019568421052631581, + "loss": 5.1445, + "step": 1859 + }, + { + "epoch": 0.09789473684210526, + "grad_norm": 0.77734375, + "learning_rate": 0.00019578947368421054, + "loss": 5.4759, + "step": 1860 + }, + { + "epoch": 0.09794736842105263, + "grad_norm": 0.82421875, + "learning_rate": 0.00019589473684210527, + "loss": 4.9094, + "step": 1861 + }, + { + "epoch": 0.098, + "grad_norm": 0.58984375, + "learning_rate": 0.000196, + "loss": 5.2679, + "step": 1862 + }, + { + "epoch": 0.09805263157894736, + "grad_norm": 0.76171875, + "learning_rate": 0.00019610526315789475, + "loss": 5.6333, + "step": 1863 + }, + { + "epoch": 0.09810526315789474, + "grad_norm": 0.6015625, + "learning_rate": 0.0001962105263157895, + "loss": 5.5442, + "step": 1864 + }, + { + "epoch": 0.09815789473684211, + "grad_norm": 0.671875, + "learning_rate": 0.00019631578947368423, + "loss": 5.1406, + "step": 1865 + }, + { + "epoch": 0.09821052631578947, + "grad_norm": 0.86328125, + "learning_rate": 0.00019642105263157895, + "loss": 5.2837, + "step": 1866 + }, + { + "epoch": 0.09826315789473684, + "grad_norm": 0.65625, + "learning_rate": 0.00019652631578947368, + "loss": 5.6029, + "step": 1867 + }, + { + "epoch": 0.0983157894736842, + "grad_norm": 0.90625, + "learning_rate": 0.00019663157894736843, + "loss": 5.5173, + "step": 1868 + }, + { + "epoch": 0.09836842105263158, + "grad_norm": 0.7890625, + "learning_rate": 0.0001967368421052632, + "loss": 4.9868, + "step": 1869 + }, + { + "epoch": 0.09842105263157895, + "grad_norm": 0.67578125, + "learning_rate": 0.0001968421052631579, + "loss": 5.322, + "step": 1870 + }, + { + "epoch": 0.09847368421052631, + "grad_norm": 0.796875, + "learning_rate": 0.00019694736842105264, + "loss": 5.2963, + "step": 1871 + }, + { + "epoch": 0.09852631578947368, + "grad_norm": 0.66015625, + "learning_rate": 0.00019705263157894737, + "loss": 5.4014, + "step": 1872 + }, + { + "epoch": 0.09857894736842106, + "grad_norm": 0.76953125, + "learning_rate": 0.00019715789473684212, + "loss": 5.5298, + "step": 1873 + }, + { + "epoch": 0.09863157894736842, + "grad_norm": 0.7421875, + "learning_rate": 0.00019726315789473685, + "loss": 5.2861, + "step": 1874 + }, + { + "epoch": 0.09868421052631579, + "grad_norm": 0.75390625, + "learning_rate": 0.00019736842105263157, + "loss": 5.3603, + "step": 1875 + }, + { + "epoch": 0.09873684210526316, + "grad_norm": 0.91796875, + "learning_rate": 0.00019747368421052633, + "loss": 5.1441, + "step": 1876 + }, + { + "epoch": 0.09878947368421052, + "grad_norm": 0.640625, + "learning_rate": 0.00019757894736842105, + "loss": 5.2761, + "step": 1877 + }, + { + "epoch": 0.0988421052631579, + "grad_norm": 0.7421875, + "learning_rate": 0.0001976842105263158, + "loss": 5.2831, + "step": 1878 + }, + { + "epoch": 0.09889473684210526, + "grad_norm": 0.8984375, + "learning_rate": 0.00019778947368421053, + "loss": 5.7421, + "step": 1879 + }, + { + "epoch": 0.09894736842105263, + "grad_norm": 0.90625, + "learning_rate": 0.00019789473684210526, + "loss": 5.2032, + "step": 1880 + }, + { + "epoch": 0.099, + "grad_norm": 0.6875, + "learning_rate": 0.00019800000000000002, + "loss": 5.5139, + "step": 1881 + }, + { + "epoch": 0.09905263157894736, + "grad_norm": 0.72265625, + "learning_rate": 0.00019810526315789474, + "loss": 5.739, + "step": 1882 + }, + { + "epoch": 0.09910526315789474, + "grad_norm": 0.73046875, + "learning_rate": 0.0001982105263157895, + "loss": 5.6562, + "step": 1883 + }, + { + "epoch": 0.09915789473684211, + "grad_norm": 0.7890625, + "learning_rate": 0.00019831578947368422, + "loss": 5.2546, + "step": 1884 + }, + { + "epoch": 0.09921052631578947, + "grad_norm": 0.73046875, + "learning_rate": 0.00019842105263157895, + "loss": 5.4954, + "step": 1885 + }, + { + "epoch": 0.09926315789473684, + "grad_norm": 1.484375, + "learning_rate": 0.0001985263157894737, + "loss": 5.1386, + "step": 1886 + }, + { + "epoch": 0.0993157894736842, + "grad_norm": 0.78515625, + "learning_rate": 0.00019863157894736843, + "loss": 4.9417, + "step": 1887 + }, + { + "epoch": 0.09936842105263158, + "grad_norm": 0.6328125, + "learning_rate": 0.00019873684210526318, + "loss": 5.6506, + "step": 1888 + }, + { + "epoch": 0.09942105263157895, + "grad_norm": 0.67578125, + "learning_rate": 0.0001988421052631579, + "loss": 5.3563, + "step": 1889 + }, + { + "epoch": 0.09947368421052631, + "grad_norm": 0.765625, + "learning_rate": 0.00019894736842105264, + "loss": 5.1909, + "step": 1890 + }, + { + "epoch": 0.09952631578947368, + "grad_norm": 0.63671875, + "learning_rate": 0.0001990526315789474, + "loss": 5.8766, + "step": 1891 + }, + { + "epoch": 0.09957894736842106, + "grad_norm": 0.70703125, + "learning_rate": 0.00019915789473684212, + "loss": 5.5898, + "step": 1892 + }, + { + "epoch": 0.09963157894736842, + "grad_norm": 0.73046875, + "learning_rate": 0.00019926315789473687, + "loss": 5.2133, + "step": 1893 + }, + { + "epoch": 0.09968421052631579, + "grad_norm": 0.58203125, + "learning_rate": 0.00019936842105263157, + "loss": 5.4759, + "step": 1894 + }, + { + "epoch": 0.09973684210526315, + "grad_norm": 0.59765625, + "learning_rate": 0.00019947368421052632, + "loss": 5.1692, + "step": 1895 + }, + { + "epoch": 0.09978947368421053, + "grad_norm": 1.1796875, + "learning_rate": 0.00019957894736842108, + "loss": 5.4111, + "step": 1896 + }, + { + "epoch": 0.0998421052631579, + "grad_norm": 0.796875, + "learning_rate": 0.0001996842105263158, + "loss": 4.9569, + "step": 1897 + }, + { + "epoch": 0.09989473684210526, + "grad_norm": 0.75390625, + "learning_rate": 0.00019978947368421053, + "loss": 5.2601, + "step": 1898 + }, + { + "epoch": 0.09994736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00019989473684210526, + "loss": 5.3862, + "step": 1899 + }, + { + "epoch": 0.1, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 5.3187, + "step": 1900 + }, + { + "epoch": 0.10005263157894737, + "grad_norm": 0.57421875, + "learning_rate": 0.00019998830409356726, + "loss": 5.2348, + "step": 1901 + }, + { + "epoch": 0.10010526315789474, + "grad_norm": 0.640625, + "learning_rate": 0.0001999766081871345, + "loss": 5.3339, + "step": 1902 + }, + { + "epoch": 0.1001578947368421, + "grad_norm": 0.67578125, + "learning_rate": 0.00019996491228070176, + "loss": 5.2268, + "step": 1903 + }, + { + "epoch": 0.10021052631578947, + "grad_norm": 0.6796875, + "learning_rate": 0.00019995321637426904, + "loss": 5.1888, + "step": 1904 + }, + { + "epoch": 0.10026315789473685, + "grad_norm": 0.96484375, + "learning_rate": 0.00019994152046783628, + "loss": 5.5426, + "step": 1905 + }, + { + "epoch": 0.1003157894736842, + "grad_norm": 0.75, + "learning_rate": 0.00019992982456140353, + "loss": 5.1692, + "step": 1906 + }, + { + "epoch": 0.10036842105263158, + "grad_norm": 0.609375, + "learning_rate": 0.00019991812865497078, + "loss": 5.561, + "step": 1907 + }, + { + "epoch": 0.10042105263157895, + "grad_norm": 0.64453125, + "learning_rate": 0.000199906432748538, + "loss": 5.0899, + "step": 1908 + }, + { + "epoch": 0.10047368421052631, + "grad_norm": 0.640625, + "learning_rate": 0.00019989473684210526, + "loss": 5.5854, + "step": 1909 + }, + { + "epoch": 0.10052631578947369, + "grad_norm": 0.703125, + "learning_rate": 0.0001998830409356725, + "loss": 5.5598, + "step": 1910 + }, + { + "epoch": 0.10057894736842106, + "grad_norm": 0.6953125, + "learning_rate": 0.00019987134502923976, + "loss": 5.5401, + "step": 1911 + }, + { + "epoch": 0.10063157894736842, + "grad_norm": 0.64453125, + "learning_rate": 0.00019985964912280703, + "loss": 4.9945, + "step": 1912 + }, + { + "epoch": 0.10068421052631579, + "grad_norm": 0.75390625, + "learning_rate": 0.00019984795321637428, + "loss": 5.3998, + "step": 1913 + }, + { + "epoch": 0.10073684210526315, + "grad_norm": 0.640625, + "learning_rate": 0.00019983625730994153, + "loss": 5.5283, + "step": 1914 + }, + { + "epoch": 0.10078947368421053, + "grad_norm": 0.7734375, + "learning_rate": 0.00019982456140350878, + "loss": 5.5971, + "step": 1915 + }, + { + "epoch": 0.1008421052631579, + "grad_norm": 0.91796875, + "learning_rate": 0.00019981286549707603, + "loss": 5.4363, + "step": 1916 + }, + { + "epoch": 0.10089473684210526, + "grad_norm": 0.7890625, + "learning_rate": 0.00019980116959064328, + "loss": 5.4743, + "step": 1917 + }, + { + "epoch": 0.10094736842105263, + "grad_norm": 0.79296875, + "learning_rate": 0.00019978947368421053, + "loss": 5.4, + "step": 1918 + }, + { + "epoch": 0.101, + "grad_norm": 0.8046875, + "learning_rate": 0.0001997777777777778, + "loss": 5.4439, + "step": 1919 + }, + { + "epoch": 0.10105263157894737, + "grad_norm": 0.87109375, + "learning_rate": 0.00019976608187134506, + "loss": 5.4041, + "step": 1920 + }, + { + "epoch": 0.10110526315789474, + "grad_norm": 1.046875, + "learning_rate": 0.0001997543859649123, + "loss": 5.3884, + "step": 1921 + }, + { + "epoch": 0.1011578947368421, + "grad_norm": 0.61328125, + "learning_rate": 0.00019974269005847955, + "loss": 5.1829, + "step": 1922 + }, + { + "epoch": 0.10121052631578947, + "grad_norm": 0.65234375, + "learning_rate": 0.0001997309941520468, + "loss": 5.5126, + "step": 1923 + }, + { + "epoch": 0.10126315789473685, + "grad_norm": 0.70703125, + "learning_rate": 0.00019971929824561403, + "loss": 5.1721, + "step": 1924 + }, + { + "epoch": 0.1013157894736842, + "grad_norm": 0.77734375, + "learning_rate": 0.00019970760233918128, + "loss": 5.1636, + "step": 1925 + }, + { + "epoch": 0.10136842105263158, + "grad_norm": 0.69140625, + "learning_rate": 0.00019969590643274855, + "loss": 5.675, + "step": 1926 + }, + { + "epoch": 0.10142105263157895, + "grad_norm": 0.76171875, + "learning_rate": 0.0001996842105263158, + "loss": 5.58, + "step": 1927 + }, + { + "epoch": 0.10147368421052631, + "grad_norm": 0.64453125, + "learning_rate": 0.00019967251461988305, + "loss": 5.3761, + "step": 1928 + }, + { + "epoch": 0.10152631578947369, + "grad_norm": 0.65234375, + "learning_rate": 0.0001996608187134503, + "loss": 5.129, + "step": 1929 + }, + { + "epoch": 0.10157894736842105, + "grad_norm": 0.75, + "learning_rate": 0.00019964912280701755, + "loss": 5.6009, + "step": 1930 + }, + { + "epoch": 0.10163157894736842, + "grad_norm": 0.86328125, + "learning_rate": 0.0001996374269005848, + "loss": 5.9491, + "step": 1931 + }, + { + "epoch": 0.1016842105263158, + "grad_norm": 0.7578125, + "learning_rate": 0.00019962573099415205, + "loss": 5.4156, + "step": 1932 + }, + { + "epoch": 0.10173684210526315, + "grad_norm": 0.6796875, + "learning_rate": 0.0001996140350877193, + "loss": 4.8803, + "step": 1933 + }, + { + "epoch": 0.10178947368421053, + "grad_norm": 0.765625, + "learning_rate": 0.00019960233918128658, + "loss": 5.4796, + "step": 1934 + }, + { + "epoch": 0.1018421052631579, + "grad_norm": 0.984375, + "learning_rate": 0.00019959064327485383, + "loss": 5.4607, + "step": 1935 + }, + { + "epoch": 0.10189473684210526, + "grad_norm": 0.82421875, + "learning_rate": 0.00019957894736842108, + "loss": 5.6796, + "step": 1936 + }, + { + "epoch": 0.10194736842105263, + "grad_norm": 0.9765625, + "learning_rate": 0.00019956725146198833, + "loss": 5.0271, + "step": 1937 + }, + { + "epoch": 0.102, + "grad_norm": 1.0078125, + "learning_rate": 0.00019955555555555558, + "loss": 5.5164, + "step": 1938 + }, + { + "epoch": 0.10205263157894737, + "grad_norm": 0.890625, + "learning_rate": 0.0001995438596491228, + "loss": 5.3059, + "step": 1939 + }, + { + "epoch": 0.10210526315789474, + "grad_norm": 0.8046875, + "learning_rate": 0.00019953216374269005, + "loss": 5.2816, + "step": 1940 + }, + { + "epoch": 0.1021578947368421, + "grad_norm": 0.92578125, + "learning_rate": 0.00019952046783625732, + "loss": 5.4597, + "step": 1941 + }, + { + "epoch": 0.10221052631578947, + "grad_norm": 0.6171875, + "learning_rate": 0.00019950877192982457, + "loss": 5.5356, + "step": 1942 + }, + { + "epoch": 0.10226315789473685, + "grad_norm": 0.80078125, + "learning_rate": 0.00019949707602339182, + "loss": 5.5604, + "step": 1943 + }, + { + "epoch": 0.10231578947368421, + "grad_norm": 0.81640625, + "learning_rate": 0.00019948538011695907, + "loss": 5.2788, + "step": 1944 + }, + { + "epoch": 0.10236842105263158, + "grad_norm": 0.921875, + "learning_rate": 0.00019947368421052632, + "loss": 5.5127, + "step": 1945 + }, + { + "epoch": 0.10242105263157894, + "grad_norm": 0.66015625, + "learning_rate": 0.00019946198830409357, + "loss": 5.3567, + "step": 1946 + }, + { + "epoch": 0.10247368421052631, + "grad_norm": 0.76171875, + "learning_rate": 0.00019945029239766082, + "loss": 5.474, + "step": 1947 + }, + { + "epoch": 0.10252631578947369, + "grad_norm": 0.64453125, + "learning_rate": 0.00019943859649122807, + "loss": 5.1056, + "step": 1948 + }, + { + "epoch": 0.10257894736842105, + "grad_norm": 0.85546875, + "learning_rate": 0.00019942690058479535, + "loss": 5.395, + "step": 1949 + }, + { + "epoch": 0.10263157894736842, + "grad_norm": 0.88671875, + "learning_rate": 0.0001994152046783626, + "loss": 4.9541, + "step": 1950 + }, + { + "epoch": 0.1026842105263158, + "grad_norm": 0.55078125, + "learning_rate": 0.00019940350877192985, + "loss": 5.5399, + "step": 1951 + }, + { + "epoch": 0.10273684210526315, + "grad_norm": 0.79296875, + "learning_rate": 0.0001993918128654971, + "loss": 4.9823, + "step": 1952 + }, + { + "epoch": 0.10278947368421053, + "grad_norm": 0.6328125, + "learning_rate": 0.00019938011695906435, + "loss": 5.3608, + "step": 1953 + }, + { + "epoch": 0.1028421052631579, + "grad_norm": 0.734375, + "learning_rate": 0.00019936842105263157, + "loss": 5.4578, + "step": 1954 + }, + { + "epoch": 0.10289473684210526, + "grad_norm": 0.703125, + "learning_rate": 0.00019935672514619882, + "loss": 5.0207, + "step": 1955 + }, + { + "epoch": 0.10294736842105263, + "grad_norm": 0.66796875, + "learning_rate": 0.0001993450292397661, + "loss": 5.364, + "step": 1956 + }, + { + "epoch": 0.103, + "grad_norm": 0.86328125, + "learning_rate": 0.00019933333333333334, + "loss": 5.3158, + "step": 1957 + }, + { + "epoch": 0.10305263157894737, + "grad_norm": 0.8046875, + "learning_rate": 0.0001993216374269006, + "loss": 5.1775, + "step": 1958 + }, + { + "epoch": 0.10310526315789474, + "grad_norm": 0.93359375, + "learning_rate": 0.00019930994152046784, + "loss": 5.7829, + "step": 1959 + }, + { + "epoch": 0.1031578947368421, + "grad_norm": 0.828125, + "learning_rate": 0.0001992982456140351, + "loss": 5.3154, + "step": 1960 + }, + { + "epoch": 0.10321052631578947, + "grad_norm": 0.61328125, + "learning_rate": 0.00019928654970760234, + "loss": 5.1799, + "step": 1961 + }, + { + "epoch": 0.10326315789473685, + "grad_norm": 0.60546875, + "learning_rate": 0.0001992748538011696, + "loss": 5.3731, + "step": 1962 + }, + { + "epoch": 0.10331578947368421, + "grad_norm": 0.640625, + "learning_rate": 0.00019926315789473687, + "loss": 5.4797, + "step": 1963 + }, + { + "epoch": 0.10336842105263158, + "grad_norm": 0.87890625, + "learning_rate": 0.00019925146198830412, + "loss": 5.5439, + "step": 1964 + }, + { + "epoch": 0.10342105263157894, + "grad_norm": 0.7265625, + "learning_rate": 0.00019923976608187137, + "loss": 5.4347, + "step": 1965 + }, + { + "epoch": 0.10347368421052631, + "grad_norm": 0.76171875, + "learning_rate": 0.00019922807017543862, + "loss": 5.1353, + "step": 1966 + }, + { + "epoch": 0.10352631578947369, + "grad_norm": 0.6640625, + "learning_rate": 0.00019921637426900587, + "loss": 5.5993, + "step": 1967 + }, + { + "epoch": 0.10357894736842105, + "grad_norm": 0.828125, + "learning_rate": 0.00019920467836257312, + "loss": 5.4534, + "step": 1968 + }, + { + "epoch": 0.10363157894736842, + "grad_norm": 0.74609375, + "learning_rate": 0.00019919298245614037, + "loss": 5.5027, + "step": 1969 + }, + { + "epoch": 0.1036842105263158, + "grad_norm": 0.90234375, + "learning_rate": 0.0001991812865497076, + "loss": 5.496, + "step": 1970 + }, + { + "epoch": 0.10373684210526316, + "grad_norm": 0.91796875, + "learning_rate": 0.00019916959064327487, + "loss": 5.1866, + "step": 1971 + }, + { + "epoch": 0.10378947368421053, + "grad_norm": 0.6484375, + "learning_rate": 0.00019915789473684212, + "loss": 5.3048, + "step": 1972 + }, + { + "epoch": 0.10384210526315789, + "grad_norm": 0.80859375, + "learning_rate": 0.00019914619883040937, + "loss": 5.3593, + "step": 1973 + }, + { + "epoch": 0.10389473684210526, + "grad_norm": 0.76953125, + "learning_rate": 0.00019913450292397661, + "loss": 5.3414, + "step": 1974 + }, + { + "epoch": 0.10394736842105264, + "grad_norm": 0.68359375, + "learning_rate": 0.00019912280701754386, + "loss": 5.4848, + "step": 1975 + }, + { + "epoch": 0.104, + "grad_norm": 0.87890625, + "learning_rate": 0.00019911111111111111, + "loss": 5.0781, + "step": 1976 + }, + { + "epoch": 0.10405263157894737, + "grad_norm": 0.67578125, + "learning_rate": 0.00019909941520467836, + "loss": 5.3659, + "step": 1977 + }, + { + "epoch": 0.10410526315789474, + "grad_norm": 0.74609375, + "learning_rate": 0.00019908771929824564, + "loss": 5.4414, + "step": 1978 + }, + { + "epoch": 0.1041578947368421, + "grad_norm": 0.6796875, + "learning_rate": 0.0001990760233918129, + "loss": 5.3991, + "step": 1979 + }, + { + "epoch": 0.10421052631578948, + "grad_norm": 0.80078125, + "learning_rate": 0.00019906432748538014, + "loss": 5.4747, + "step": 1980 + }, + { + "epoch": 0.10426315789473684, + "grad_norm": 0.8515625, + "learning_rate": 0.0001990526315789474, + "loss": 5.3346, + "step": 1981 + }, + { + "epoch": 0.10431578947368421, + "grad_norm": 0.67578125, + "learning_rate": 0.00019904093567251464, + "loss": 4.9582, + "step": 1982 + }, + { + "epoch": 0.10436842105263158, + "grad_norm": 0.89453125, + "learning_rate": 0.0001990292397660819, + "loss": 5.2771, + "step": 1983 + }, + { + "epoch": 0.10442105263157894, + "grad_norm": 0.71484375, + "learning_rate": 0.00019901754385964914, + "loss": 5.1456, + "step": 1984 + }, + { + "epoch": 0.10447368421052632, + "grad_norm": 0.7265625, + "learning_rate": 0.00019900584795321636, + "loss": 5.2965, + "step": 1985 + }, + { + "epoch": 0.10452631578947369, + "grad_norm": 0.6875, + "learning_rate": 0.00019899415204678364, + "loss": 5.5809, + "step": 1986 + }, + { + "epoch": 0.10457894736842105, + "grad_norm": 0.734375, + "learning_rate": 0.0001989824561403509, + "loss": 5.456, + "step": 1987 + }, + { + "epoch": 0.10463157894736842, + "grad_norm": 0.64453125, + "learning_rate": 0.00019897076023391814, + "loss": 5.3018, + "step": 1988 + }, + { + "epoch": 0.1046842105263158, + "grad_norm": 0.640625, + "learning_rate": 0.00019895906432748539, + "loss": 5.3277, + "step": 1989 + }, + { + "epoch": 0.10473684210526316, + "grad_norm": 0.8359375, + "learning_rate": 0.00019894736842105264, + "loss": 5.4233, + "step": 1990 + }, + { + "epoch": 0.10478947368421053, + "grad_norm": 0.734375, + "learning_rate": 0.00019893567251461989, + "loss": 5.1915, + "step": 1991 + }, + { + "epoch": 0.10484210526315789, + "grad_norm": 0.8671875, + "learning_rate": 0.00019892397660818713, + "loss": 5.5762, + "step": 1992 + }, + { + "epoch": 0.10489473684210526, + "grad_norm": 0.90625, + "learning_rate": 0.0001989122807017544, + "loss": 4.9546, + "step": 1993 + }, + { + "epoch": 0.10494736842105264, + "grad_norm": 0.88671875, + "learning_rate": 0.00019890058479532166, + "loss": 5.333, + "step": 1994 + }, + { + "epoch": 0.105, + "grad_norm": 0.609375, + "learning_rate": 0.0001988888888888889, + "loss": 5.3479, + "step": 1995 + }, + { + "epoch": 0.10505263157894737, + "grad_norm": 0.65625, + "learning_rate": 0.00019887719298245616, + "loss": 5.4655, + "step": 1996 + }, + { + "epoch": 0.10510526315789474, + "grad_norm": 0.8125, + "learning_rate": 0.0001988654970760234, + "loss": 5.0026, + "step": 1997 + }, + { + "epoch": 0.1051578947368421, + "grad_norm": 0.92578125, + "learning_rate": 0.00019885380116959066, + "loss": 5.168, + "step": 1998 + }, + { + "epoch": 0.10521052631578948, + "grad_norm": 0.61328125, + "learning_rate": 0.0001988421052631579, + "loss": 5.197, + "step": 1999 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 0.875, + "learning_rate": 0.00019883040935672513, + "loss": 5.3386, + "step": 2000 + }, + { + "epoch": 0.10531578947368421, + "grad_norm": 0.84375, + "learning_rate": 0.0001988187134502924, + "loss": 5.3575, + "step": 2001 + }, + { + "epoch": 0.10536842105263158, + "grad_norm": 0.84375, + "learning_rate": 0.00019880701754385966, + "loss": 5.1642, + "step": 2002 + }, + { + "epoch": 0.10542105263157894, + "grad_norm": 0.61328125, + "learning_rate": 0.0001987953216374269, + "loss": 5.2176, + "step": 2003 + }, + { + "epoch": 0.10547368421052632, + "grad_norm": 0.66015625, + "learning_rate": 0.00019878362573099416, + "loss": 5.3651, + "step": 2004 + }, + { + "epoch": 0.10552631578947369, + "grad_norm": 0.9140625, + "learning_rate": 0.0001987719298245614, + "loss": 4.8548, + "step": 2005 + }, + { + "epoch": 0.10557894736842105, + "grad_norm": 0.6953125, + "learning_rate": 0.00019876023391812866, + "loss": 4.9025, + "step": 2006 + }, + { + "epoch": 0.10563157894736842, + "grad_norm": 0.64453125, + "learning_rate": 0.0001987485380116959, + "loss": 5.5182, + "step": 2007 + }, + { + "epoch": 0.10568421052631578, + "grad_norm": 0.84765625, + "learning_rate": 0.00019873684210526318, + "loss": 5.0638, + "step": 2008 + }, + { + "epoch": 0.10573684210526316, + "grad_norm": 0.97265625, + "learning_rate": 0.00019872514619883043, + "loss": 5.2019, + "step": 2009 + }, + { + "epoch": 0.10578947368421053, + "grad_norm": 0.875, + "learning_rate": 0.00019871345029239768, + "loss": 5.3143, + "step": 2010 + }, + { + "epoch": 0.10584210526315789, + "grad_norm": 0.90625, + "learning_rate": 0.00019870175438596493, + "loss": 5.2235, + "step": 2011 + }, + { + "epoch": 0.10589473684210526, + "grad_norm": 0.6484375, + "learning_rate": 0.00019869005847953218, + "loss": 5.2487, + "step": 2012 + }, + { + "epoch": 0.10594736842105264, + "grad_norm": 0.86328125, + "learning_rate": 0.00019867836257309943, + "loss": 5.3595, + "step": 2013 + }, + { + "epoch": 0.106, + "grad_norm": 0.59375, + "learning_rate": 0.00019866666666666668, + "loss": 5.3331, + "step": 2014 + }, + { + "epoch": 0.10605263157894737, + "grad_norm": 0.68359375, + "learning_rate": 0.00019865497076023393, + "loss": 5.6177, + "step": 2015 + }, + { + "epoch": 0.10610526315789473, + "grad_norm": 0.8671875, + "learning_rate": 0.00019864327485380118, + "loss": 5.4372, + "step": 2016 + }, + { + "epoch": 0.1061578947368421, + "grad_norm": 1.0, + "learning_rate": 0.00019863157894736843, + "loss": 4.9193, + "step": 2017 + }, + { + "epoch": 0.10621052631578948, + "grad_norm": 0.72265625, + "learning_rate": 0.00019861988304093568, + "loss": 5.4679, + "step": 2018 + }, + { + "epoch": 0.10626315789473684, + "grad_norm": 0.8046875, + "learning_rate": 0.00019860818713450293, + "loss": 5.2831, + "step": 2019 + }, + { + "epoch": 0.10631578947368421, + "grad_norm": 0.6640625, + "learning_rate": 0.00019859649122807018, + "loss": 5.4883, + "step": 2020 + }, + { + "epoch": 0.10636842105263158, + "grad_norm": 0.95703125, + "learning_rate": 0.00019858479532163743, + "loss": 5.1102, + "step": 2021 + }, + { + "epoch": 0.10642105263157894, + "grad_norm": 0.7421875, + "learning_rate": 0.00019857309941520468, + "loss": 5.0899, + "step": 2022 + }, + { + "epoch": 0.10647368421052632, + "grad_norm": 0.67578125, + "learning_rate": 0.00019856140350877195, + "loss": 5.4248, + "step": 2023 + }, + { + "epoch": 0.10652631578947368, + "grad_norm": 0.72265625, + "learning_rate": 0.0001985497076023392, + "loss": 5.2828, + "step": 2024 + }, + { + "epoch": 0.10657894736842105, + "grad_norm": 0.6640625, + "learning_rate": 0.00019853801169590645, + "loss": 5.6264, + "step": 2025 + }, + { + "epoch": 0.10663157894736842, + "grad_norm": 0.76171875, + "learning_rate": 0.0001985263157894737, + "loss": 4.9321, + "step": 2026 + }, + { + "epoch": 0.10668421052631578, + "grad_norm": 0.75390625, + "learning_rate": 0.00019851461988304095, + "loss": 5.0225, + "step": 2027 + }, + { + "epoch": 0.10673684210526316, + "grad_norm": 0.60546875, + "learning_rate": 0.0001985029239766082, + "loss": 5.4788, + "step": 2028 + }, + { + "epoch": 0.10678947368421053, + "grad_norm": 0.62109375, + "learning_rate": 0.00019849122807017545, + "loss": 5.4784, + "step": 2029 + }, + { + "epoch": 0.10684210526315789, + "grad_norm": 0.84375, + "learning_rate": 0.0001984795321637427, + "loss": 5.3978, + "step": 2030 + }, + { + "epoch": 0.10689473684210526, + "grad_norm": 0.625, + "learning_rate": 0.00019846783625730995, + "loss": 5.3057, + "step": 2031 + }, + { + "epoch": 0.10694736842105264, + "grad_norm": 0.6015625, + "learning_rate": 0.0001984561403508772, + "loss": 5.4717, + "step": 2032 + }, + { + "epoch": 0.107, + "grad_norm": 0.7578125, + "learning_rate": 0.00019844444444444445, + "loss": 5.3104, + "step": 2033 + }, + { + "epoch": 0.10705263157894737, + "grad_norm": 0.84375, + "learning_rate": 0.0001984327485380117, + "loss": 5.3596, + "step": 2034 + }, + { + "epoch": 0.10710526315789473, + "grad_norm": 0.6875, + "learning_rate": 0.00019842105263157895, + "loss": 5.4299, + "step": 2035 + }, + { + "epoch": 0.1071578947368421, + "grad_norm": 0.66015625, + "learning_rate": 0.0001984093567251462, + "loss": 5.1283, + "step": 2036 + }, + { + "epoch": 0.10721052631578948, + "grad_norm": 0.72265625, + "learning_rate": 0.00019839766081871345, + "loss": 5.2139, + "step": 2037 + }, + { + "epoch": 0.10726315789473684, + "grad_norm": 0.859375, + "learning_rate": 0.00019838596491228072, + "loss": 5.4605, + "step": 2038 + }, + { + "epoch": 0.10731578947368421, + "grad_norm": 0.79296875, + "learning_rate": 0.00019837426900584797, + "loss": 5.3093, + "step": 2039 + }, + { + "epoch": 0.10736842105263159, + "grad_norm": 0.8203125, + "learning_rate": 0.00019836257309941522, + "loss": 5.1713, + "step": 2040 + }, + { + "epoch": 0.10742105263157894, + "grad_norm": 0.75, + "learning_rate": 0.00019835087719298247, + "loss": 5.2657, + "step": 2041 + }, + { + "epoch": 0.10747368421052632, + "grad_norm": 0.69140625, + "learning_rate": 0.00019833918128654972, + "loss": 5.3142, + "step": 2042 + }, + { + "epoch": 0.10752631578947368, + "grad_norm": 0.65625, + "learning_rate": 0.00019832748538011697, + "loss": 4.9997, + "step": 2043 + }, + { + "epoch": 0.10757894736842105, + "grad_norm": 0.69140625, + "learning_rate": 0.00019831578947368422, + "loss": 5.118, + "step": 2044 + }, + { + "epoch": 0.10763157894736843, + "grad_norm": 1.59375, + "learning_rate": 0.00019830409356725147, + "loss": 5.476, + "step": 2045 + }, + { + "epoch": 0.10768421052631579, + "grad_norm": 0.97265625, + "learning_rate": 0.00019829239766081872, + "loss": 5.4726, + "step": 2046 + }, + { + "epoch": 0.10773684210526316, + "grad_norm": 0.63671875, + "learning_rate": 0.00019828070175438597, + "loss": 5.2921, + "step": 2047 + }, + { + "epoch": 0.10778947368421053, + "grad_norm": 0.7578125, + "learning_rate": 0.00019826900584795322, + "loss": 5.3054, + "step": 2048 + }, + { + "epoch": 0.10784210526315789, + "grad_norm": 0.75390625, + "learning_rate": 0.00019825730994152047, + "loss": 5.697, + "step": 2049 + }, + { + "epoch": 0.10789473684210527, + "grad_norm": 0.66015625, + "learning_rate": 0.00019824561403508772, + "loss": 5.623, + "step": 2050 + }, + { + "epoch": 0.10794736842105263, + "grad_norm": 0.71484375, + "learning_rate": 0.00019823391812865497, + "loss": 5.1156, + "step": 2051 + }, + { + "epoch": 0.108, + "grad_norm": 0.87109375, + "learning_rate": 0.00019822222222222225, + "loss": 5.3449, + "step": 2052 + }, + { + "epoch": 0.10805263157894737, + "grad_norm": 0.75, + "learning_rate": 0.0001982105263157895, + "loss": 5.5343, + "step": 2053 + }, + { + "epoch": 0.10810526315789473, + "grad_norm": 0.8203125, + "learning_rate": 0.00019819883040935675, + "loss": 5.2413, + "step": 2054 + }, + { + "epoch": 0.1081578947368421, + "grad_norm": 0.63671875, + "learning_rate": 0.000198187134502924, + "loss": 4.9966, + "step": 2055 + }, + { + "epoch": 0.10821052631578948, + "grad_norm": 0.65625, + "learning_rate": 0.00019817543859649124, + "loss": 5.2876, + "step": 2056 + }, + { + "epoch": 0.10826315789473684, + "grad_norm": 0.62890625, + "learning_rate": 0.0001981637426900585, + "loss": 5.3615, + "step": 2057 + }, + { + "epoch": 0.10831578947368421, + "grad_norm": 0.6796875, + "learning_rate": 0.00019815204678362574, + "loss": 5.5834, + "step": 2058 + }, + { + "epoch": 0.10836842105263157, + "grad_norm": 0.73828125, + "learning_rate": 0.000198140350877193, + "loss": 5.6302, + "step": 2059 + }, + { + "epoch": 0.10842105263157895, + "grad_norm": 0.7421875, + "learning_rate": 0.00019812865497076024, + "loss": 4.8244, + "step": 2060 + }, + { + "epoch": 0.10847368421052632, + "grad_norm": 0.74609375, + "learning_rate": 0.0001981169590643275, + "loss": 5.3123, + "step": 2061 + }, + { + "epoch": 0.10852631578947368, + "grad_norm": 0.6796875, + "learning_rate": 0.00019810526315789474, + "loss": 5.0814, + "step": 2062 + }, + { + "epoch": 0.10857894736842105, + "grad_norm": 0.77734375, + "learning_rate": 0.000198093567251462, + "loss": 5.4031, + "step": 2063 + }, + { + "epoch": 0.10863157894736843, + "grad_norm": 0.6953125, + "learning_rate": 0.00019808187134502924, + "loss": 5.5171, + "step": 2064 + }, + { + "epoch": 0.10868421052631579, + "grad_norm": 0.73046875, + "learning_rate": 0.0001980701754385965, + "loss": 5.5872, + "step": 2065 + }, + { + "epoch": 0.10873684210526316, + "grad_norm": 0.71875, + "learning_rate": 0.00019805847953216374, + "loss": 5.3466, + "step": 2066 + }, + { + "epoch": 0.10878947368421053, + "grad_norm": 0.7265625, + "learning_rate": 0.00019804678362573102, + "loss": 5.5201, + "step": 2067 + }, + { + "epoch": 0.10884210526315789, + "grad_norm": 0.70703125, + "learning_rate": 0.00019803508771929827, + "loss": 5.0485, + "step": 2068 + }, + { + "epoch": 0.10889473684210527, + "grad_norm": 0.80078125, + "learning_rate": 0.00019802339181286552, + "loss": 5.4261, + "step": 2069 + }, + { + "epoch": 0.10894736842105263, + "grad_norm": 0.6953125, + "learning_rate": 0.00019801169590643277, + "loss": 5.2091, + "step": 2070 + }, + { + "epoch": 0.109, + "grad_norm": 0.703125, + "learning_rate": 0.00019800000000000002, + "loss": 5.3311, + "step": 2071 + }, + { + "epoch": 0.10905263157894737, + "grad_norm": 0.66015625, + "learning_rate": 0.00019798830409356726, + "loss": 5.2789, + "step": 2072 + }, + { + "epoch": 0.10910526315789473, + "grad_norm": 0.6171875, + "learning_rate": 0.00019797660818713451, + "loss": 5.1561, + "step": 2073 + }, + { + "epoch": 0.1091578947368421, + "grad_norm": 0.56640625, + "learning_rate": 0.00019796491228070176, + "loss": 5.1664, + "step": 2074 + }, + { + "epoch": 0.10921052631578948, + "grad_norm": 0.59375, + "learning_rate": 0.000197953216374269, + "loss": 4.9987, + "step": 2075 + }, + { + "epoch": 0.10926315789473684, + "grad_norm": 0.6015625, + "learning_rate": 0.00019794152046783626, + "loss": 5.2044, + "step": 2076 + }, + { + "epoch": 0.10931578947368421, + "grad_norm": 0.67578125, + "learning_rate": 0.0001979298245614035, + "loss": 5.2274, + "step": 2077 + }, + { + "epoch": 0.10936842105263157, + "grad_norm": 0.54296875, + "learning_rate": 0.00019791812865497076, + "loss": 5.2673, + "step": 2078 + }, + { + "epoch": 0.10942105263157895, + "grad_norm": 0.6875, + "learning_rate": 0.000197906432748538, + "loss": 5.2982, + "step": 2079 + }, + { + "epoch": 0.10947368421052632, + "grad_norm": 0.72265625, + "learning_rate": 0.00019789473684210526, + "loss": 4.9071, + "step": 2080 + }, + { + "epoch": 0.10952631578947368, + "grad_norm": 0.6328125, + "learning_rate": 0.0001978830409356725, + "loss": 5.3157, + "step": 2081 + }, + { + "epoch": 0.10957894736842105, + "grad_norm": 0.765625, + "learning_rate": 0.0001978713450292398, + "loss": 5.2345, + "step": 2082 + }, + { + "epoch": 0.10963157894736843, + "grad_norm": 0.8125, + "learning_rate": 0.00019785964912280704, + "loss": 5.4157, + "step": 2083 + }, + { + "epoch": 0.10968421052631579, + "grad_norm": 0.8359375, + "learning_rate": 0.0001978479532163743, + "loss": 5.1182, + "step": 2084 + }, + { + "epoch": 0.10973684210526316, + "grad_norm": 0.765625, + "learning_rate": 0.00019783625730994154, + "loss": 5.0096, + "step": 2085 + }, + { + "epoch": 0.10978947368421052, + "grad_norm": 0.69921875, + "learning_rate": 0.00019782456140350879, + "loss": 5.1773, + "step": 2086 + }, + { + "epoch": 0.1098421052631579, + "grad_norm": 0.609375, + "learning_rate": 0.00019781286549707604, + "loss": 5.1736, + "step": 2087 + }, + { + "epoch": 0.10989473684210527, + "grad_norm": 0.65234375, + "learning_rate": 0.00019780116959064329, + "loss": 5.3787, + "step": 2088 + }, + { + "epoch": 0.10994736842105263, + "grad_norm": 0.84765625, + "learning_rate": 0.00019778947368421053, + "loss": 5.3246, + "step": 2089 + }, + { + "epoch": 0.11, + "grad_norm": 0.75390625, + "learning_rate": 0.00019777777777777778, + "loss": 5.1811, + "step": 2090 + }, + { + "epoch": 0.11005263157894737, + "grad_norm": 0.69140625, + "learning_rate": 0.00019776608187134503, + "loss": 5.2075, + "step": 2091 + }, + { + "epoch": 0.11010526315789473, + "grad_norm": 0.7265625, + "learning_rate": 0.00019775438596491228, + "loss": 5.1179, + "step": 2092 + }, + { + "epoch": 0.11015789473684211, + "grad_norm": 0.91796875, + "learning_rate": 0.00019774269005847953, + "loss": 4.9771, + "step": 2093 + }, + { + "epoch": 0.11021052631578947, + "grad_norm": 0.70703125, + "learning_rate": 0.00019773099415204678, + "loss": 5.4201, + "step": 2094 + }, + { + "epoch": 0.11026315789473684, + "grad_norm": 0.75390625, + "learning_rate": 0.00019771929824561403, + "loss": 5.4435, + "step": 2095 + }, + { + "epoch": 0.11031578947368421, + "grad_norm": 0.69921875, + "learning_rate": 0.00019770760233918128, + "loss": 5.2725, + "step": 2096 + }, + { + "epoch": 0.11036842105263157, + "grad_norm": 0.67578125, + "learning_rate": 0.00019769590643274856, + "loss": 5.2865, + "step": 2097 + }, + { + "epoch": 0.11042105263157895, + "grad_norm": 0.77734375, + "learning_rate": 0.0001976842105263158, + "loss": 5.1659, + "step": 2098 + }, + { + "epoch": 0.11047368421052632, + "grad_norm": 1.0625, + "learning_rate": 0.00019767251461988306, + "loss": 5.2392, + "step": 2099 + }, + { + "epoch": 0.11052631578947368, + "grad_norm": 0.66796875, + "learning_rate": 0.0001976608187134503, + "loss": 5.3785, + "step": 2100 + }, + { + "epoch": 0.11057894736842105, + "grad_norm": 0.8203125, + "learning_rate": 0.00019764912280701756, + "loss": 4.9587, + "step": 2101 + }, + { + "epoch": 0.11063157894736841, + "grad_norm": 0.66015625, + "learning_rate": 0.0001976374269005848, + "loss": 5.3968, + "step": 2102 + }, + { + "epoch": 0.11068421052631579, + "grad_norm": 0.78125, + "learning_rate": 0.00019762573099415206, + "loss": 5.2609, + "step": 2103 + }, + { + "epoch": 0.11073684210526316, + "grad_norm": 0.8203125, + "learning_rate": 0.0001976140350877193, + "loss": 5.4337, + "step": 2104 + }, + { + "epoch": 0.11078947368421052, + "grad_norm": 0.71484375, + "learning_rate": 0.00019760233918128656, + "loss": 5.4049, + "step": 2105 + }, + { + "epoch": 0.1108421052631579, + "grad_norm": 0.6953125, + "learning_rate": 0.0001975906432748538, + "loss": 5.2403, + "step": 2106 + }, + { + "epoch": 0.11089473684210527, + "grad_norm": 0.94921875, + "learning_rate": 0.00019757894736842105, + "loss": 5.3891, + "step": 2107 + }, + { + "epoch": 0.11094736842105263, + "grad_norm": 0.59765625, + "learning_rate": 0.0001975672514619883, + "loss": 5.5286, + "step": 2108 + }, + { + "epoch": 0.111, + "grad_norm": 0.69140625, + "learning_rate": 0.00019755555555555555, + "loss": 5.3337, + "step": 2109 + }, + { + "epoch": 0.11105263157894738, + "grad_norm": 0.83203125, + "learning_rate": 0.0001975438596491228, + "loss": 4.936, + "step": 2110 + }, + { + "epoch": 0.11110526315789473, + "grad_norm": 0.59375, + "learning_rate": 0.00019753216374269005, + "loss": 5.2715, + "step": 2111 + }, + { + "epoch": 0.11115789473684211, + "grad_norm": 0.87890625, + "learning_rate": 0.00019752046783625733, + "loss": 5.1439, + "step": 2112 + }, + { + "epoch": 0.11121052631578947, + "grad_norm": 0.734375, + "learning_rate": 0.00019750877192982458, + "loss": 5.303, + "step": 2113 + }, + { + "epoch": 0.11126315789473684, + "grad_norm": 0.71875, + "learning_rate": 0.00019749707602339183, + "loss": 4.9712, + "step": 2114 + }, + { + "epoch": 0.11131578947368422, + "grad_norm": 0.6015625, + "learning_rate": 0.00019748538011695908, + "loss": 5.3725, + "step": 2115 + }, + { + "epoch": 0.11136842105263157, + "grad_norm": 0.71484375, + "learning_rate": 0.00019747368421052633, + "loss": 4.9532, + "step": 2116 + }, + { + "epoch": 0.11142105263157895, + "grad_norm": 0.59375, + "learning_rate": 0.00019746198830409358, + "loss": 5.1157, + "step": 2117 + }, + { + "epoch": 0.11147368421052632, + "grad_norm": 0.65625, + "learning_rate": 0.00019745029239766083, + "loss": 5.4002, + "step": 2118 + }, + { + "epoch": 0.11152631578947368, + "grad_norm": 0.6484375, + "learning_rate": 0.00019743859649122808, + "loss": 5.1181, + "step": 2119 + }, + { + "epoch": 0.11157894736842106, + "grad_norm": 0.67578125, + "learning_rate": 0.00019742690058479533, + "loss": 5.2147, + "step": 2120 + }, + { + "epoch": 0.11163157894736842, + "grad_norm": 0.8203125, + "learning_rate": 0.00019741520467836258, + "loss": 5.2549, + "step": 2121 + }, + { + "epoch": 0.11168421052631579, + "grad_norm": 0.6171875, + "learning_rate": 0.00019740350877192983, + "loss": 5.3532, + "step": 2122 + }, + { + "epoch": 0.11173684210526316, + "grad_norm": 0.62109375, + "learning_rate": 0.00019739181286549708, + "loss": 5.207, + "step": 2123 + }, + { + "epoch": 0.11178947368421052, + "grad_norm": 0.68359375, + "learning_rate": 0.00019738011695906432, + "loss": 5.3163, + "step": 2124 + }, + { + "epoch": 0.1118421052631579, + "grad_norm": 0.8515625, + "learning_rate": 0.00019736842105263157, + "loss": 5.3513, + "step": 2125 + }, + { + "epoch": 0.11189473684210527, + "grad_norm": 0.8359375, + "learning_rate": 0.00019735672514619885, + "loss": 5.0495, + "step": 2126 + }, + { + "epoch": 0.11194736842105263, + "grad_norm": 0.703125, + "learning_rate": 0.0001973450292397661, + "loss": 5.3069, + "step": 2127 + }, + { + "epoch": 0.112, + "grad_norm": 0.72265625, + "learning_rate": 0.00019733333333333335, + "loss": 5.1379, + "step": 2128 + }, + { + "epoch": 0.11205263157894736, + "grad_norm": 0.5859375, + "learning_rate": 0.0001973216374269006, + "loss": 5.3568, + "step": 2129 + }, + { + "epoch": 0.11210526315789474, + "grad_norm": 0.83203125, + "learning_rate": 0.00019730994152046785, + "loss": 5.3195, + "step": 2130 + }, + { + "epoch": 0.11215789473684211, + "grad_norm": 0.64453125, + "learning_rate": 0.0001972982456140351, + "loss": 5.4859, + "step": 2131 + }, + { + "epoch": 0.11221052631578947, + "grad_norm": 1.0078125, + "learning_rate": 0.00019728654970760235, + "loss": 5.0326, + "step": 2132 + }, + { + "epoch": 0.11226315789473684, + "grad_norm": 0.99609375, + "learning_rate": 0.0001972748538011696, + "loss": 5.1119, + "step": 2133 + }, + { + "epoch": 0.11231578947368422, + "grad_norm": 0.82421875, + "learning_rate": 0.00019726315789473685, + "loss": 5.1707, + "step": 2134 + }, + { + "epoch": 0.11236842105263158, + "grad_norm": 0.7109375, + "learning_rate": 0.0001972514619883041, + "loss": 5.1856, + "step": 2135 + }, + { + "epoch": 0.11242105263157895, + "grad_norm": 0.77734375, + "learning_rate": 0.00019723976608187135, + "loss": 5.3928, + "step": 2136 + }, + { + "epoch": 0.11247368421052631, + "grad_norm": 1.0234375, + "learning_rate": 0.0001972280701754386, + "loss": 5.2767, + "step": 2137 + }, + { + "epoch": 0.11252631578947368, + "grad_norm": 0.66796875, + "learning_rate": 0.00019721637426900585, + "loss": 5.0121, + "step": 2138 + }, + { + "epoch": 0.11257894736842106, + "grad_norm": 0.578125, + "learning_rate": 0.0001972046783625731, + "loss": 4.8904, + "step": 2139 + }, + { + "epoch": 0.11263157894736842, + "grad_norm": 0.671875, + "learning_rate": 0.00019719298245614035, + "loss": 5.5302, + "step": 2140 + }, + { + "epoch": 0.11268421052631579, + "grad_norm": 0.6875, + "learning_rate": 0.00019718128654970762, + "loss": 5.4213, + "step": 2141 + }, + { + "epoch": 0.11273684210526316, + "grad_norm": 0.6796875, + "learning_rate": 0.00019716959064327487, + "loss": 4.8864, + "step": 2142 + }, + { + "epoch": 0.11278947368421052, + "grad_norm": 0.625, + "learning_rate": 0.00019715789473684212, + "loss": 5.2269, + "step": 2143 + }, + { + "epoch": 0.1128421052631579, + "grad_norm": 0.76953125, + "learning_rate": 0.00019714619883040937, + "loss": 5.1297, + "step": 2144 + }, + { + "epoch": 0.11289473684210527, + "grad_norm": 0.94140625, + "learning_rate": 0.00019713450292397662, + "loss": 5.0137, + "step": 2145 + }, + { + "epoch": 0.11294736842105263, + "grad_norm": 0.6328125, + "learning_rate": 0.00019712280701754387, + "loss": 5.3375, + "step": 2146 + }, + { + "epoch": 0.113, + "grad_norm": 0.66796875, + "learning_rate": 0.00019711111111111112, + "loss": 5.2678, + "step": 2147 + }, + { + "epoch": 0.11305263157894736, + "grad_norm": 0.79296875, + "learning_rate": 0.00019709941520467837, + "loss": 5.0624, + "step": 2148 + }, + { + "epoch": 0.11310526315789474, + "grad_norm": 0.61328125, + "learning_rate": 0.00019708771929824562, + "loss": 5.2896, + "step": 2149 + }, + { + "epoch": 0.11315789473684211, + "grad_norm": 0.64453125, + "learning_rate": 0.00019707602339181287, + "loss": 5.4087, + "step": 2150 + }, + { + "epoch": 0.11321052631578947, + "grad_norm": 0.98046875, + "learning_rate": 0.00019706432748538012, + "loss": 5.1665, + "step": 2151 + }, + { + "epoch": 0.11326315789473684, + "grad_norm": 0.9609375, + "learning_rate": 0.00019705263157894737, + "loss": 5.1803, + "step": 2152 + }, + { + "epoch": 0.11331578947368422, + "grad_norm": 0.57421875, + "learning_rate": 0.00019704093567251462, + "loss": 5.351, + "step": 2153 + }, + { + "epoch": 0.11336842105263158, + "grad_norm": 0.5546875, + "learning_rate": 0.00019702923976608187, + "loss": 5.1497, + "step": 2154 + }, + { + "epoch": 0.11342105263157895, + "grad_norm": 0.65625, + "learning_rate": 0.00019701754385964912, + "loss": 5.0336, + "step": 2155 + }, + { + "epoch": 0.11347368421052631, + "grad_norm": 0.62890625, + "learning_rate": 0.0001970058479532164, + "loss": 5.2585, + "step": 2156 + }, + { + "epoch": 0.11352631578947368, + "grad_norm": 0.6875, + "learning_rate": 0.00019699415204678364, + "loss": 5.1398, + "step": 2157 + }, + { + "epoch": 0.11357894736842106, + "grad_norm": 0.62890625, + "learning_rate": 0.0001969824561403509, + "loss": 5.4769, + "step": 2158 + }, + { + "epoch": 0.11363157894736842, + "grad_norm": 0.8359375, + "learning_rate": 0.00019697076023391814, + "loss": 5.3436, + "step": 2159 + }, + { + "epoch": 0.11368421052631579, + "grad_norm": 0.62109375, + "learning_rate": 0.0001969590643274854, + "loss": 5.4122, + "step": 2160 + }, + { + "epoch": 0.11373684210526316, + "grad_norm": 0.625, + "learning_rate": 0.00019694736842105264, + "loss": 4.9557, + "step": 2161 + }, + { + "epoch": 0.11378947368421052, + "grad_norm": 0.74609375, + "learning_rate": 0.0001969356725146199, + "loss": 5.0336, + "step": 2162 + }, + { + "epoch": 0.1138421052631579, + "grad_norm": 0.73046875, + "learning_rate": 0.00019692397660818714, + "loss": 5.2157, + "step": 2163 + }, + { + "epoch": 0.11389473684210526, + "grad_norm": 0.625, + "learning_rate": 0.00019691228070175442, + "loss": 5.3264, + "step": 2164 + }, + { + "epoch": 0.11394736842105263, + "grad_norm": 0.75, + "learning_rate": 0.00019690058479532164, + "loss": 5.3899, + "step": 2165 + }, + { + "epoch": 0.114, + "grad_norm": 0.73828125, + "learning_rate": 0.0001968888888888889, + "loss": 4.973, + "step": 2166 + }, + { + "epoch": 0.11405263157894736, + "grad_norm": 0.671875, + "learning_rate": 0.00019687719298245614, + "loss": 5.3937, + "step": 2167 + }, + { + "epoch": 0.11410526315789474, + "grad_norm": 0.6796875, + "learning_rate": 0.0001968654970760234, + "loss": 5.1176, + "step": 2168 + }, + { + "epoch": 0.11415789473684211, + "grad_norm": 0.7265625, + "learning_rate": 0.00019685380116959064, + "loss": 5.2361, + "step": 2169 + }, + { + "epoch": 0.11421052631578947, + "grad_norm": 0.62890625, + "learning_rate": 0.0001968421052631579, + "loss": 5.2205, + "step": 2170 + }, + { + "epoch": 0.11426315789473684, + "grad_norm": 0.74609375, + "learning_rate": 0.00019683040935672516, + "loss": 5.051, + "step": 2171 + }, + { + "epoch": 0.1143157894736842, + "grad_norm": 0.79296875, + "learning_rate": 0.00019681871345029241, + "loss": 5.3756, + "step": 2172 + }, + { + "epoch": 0.11436842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.00019680701754385966, + "loss": 5.1397, + "step": 2173 + }, + { + "epoch": 0.11442105263157895, + "grad_norm": 0.6875, + "learning_rate": 0.0001967953216374269, + "loss": 5.3323, + "step": 2174 + }, + { + "epoch": 0.11447368421052631, + "grad_norm": 0.67578125, + "learning_rate": 0.00019678362573099416, + "loss": 5.3068, + "step": 2175 + }, + { + "epoch": 0.11452631578947368, + "grad_norm": 0.640625, + "learning_rate": 0.0001967719298245614, + "loss": 5.3615, + "step": 2176 + }, + { + "epoch": 0.11457894736842106, + "grad_norm": 0.66015625, + "learning_rate": 0.00019676023391812866, + "loss": 5.261, + "step": 2177 + }, + { + "epoch": 0.11463157894736842, + "grad_norm": 0.66796875, + "learning_rate": 0.00019674853801169594, + "loss": 5.5216, + "step": 2178 + }, + { + "epoch": 0.11468421052631579, + "grad_norm": 0.74609375, + "learning_rate": 0.0001967368421052632, + "loss": 4.985, + "step": 2179 + }, + { + "epoch": 0.11473684210526315, + "grad_norm": 0.5703125, + "learning_rate": 0.0001967251461988304, + "loss": 5.1215, + "step": 2180 + }, + { + "epoch": 0.11478947368421052, + "grad_norm": 1.15625, + "learning_rate": 0.00019671345029239766, + "loss": 5.2651, + "step": 2181 + }, + { + "epoch": 0.1148421052631579, + "grad_norm": 0.80078125, + "learning_rate": 0.0001967017543859649, + "loss": 4.8301, + "step": 2182 + }, + { + "epoch": 0.11489473684210526, + "grad_norm": 0.5625, + "learning_rate": 0.00019669005847953216, + "loss": 5.2958, + "step": 2183 + }, + { + "epoch": 0.11494736842105263, + "grad_norm": 0.71484375, + "learning_rate": 0.0001966783625730994, + "loss": 5.2789, + "step": 2184 + }, + { + "epoch": 0.115, + "grad_norm": 0.71484375, + "learning_rate": 0.00019666666666666666, + "loss": 5.2543, + "step": 2185 + }, + { + "epoch": 0.11505263157894736, + "grad_norm": 0.73828125, + "learning_rate": 0.00019665497076023394, + "loss": 5.4705, + "step": 2186 + }, + { + "epoch": 0.11510526315789474, + "grad_norm": 0.765625, + "learning_rate": 0.00019664327485380118, + "loss": 5.2304, + "step": 2187 + }, + { + "epoch": 0.11515789473684211, + "grad_norm": 0.6171875, + "learning_rate": 0.00019663157894736843, + "loss": 5.2241, + "step": 2188 + }, + { + "epoch": 0.11521052631578947, + "grad_norm": 0.765625, + "learning_rate": 0.00019661988304093568, + "loss": 5.3363, + "step": 2189 + }, + { + "epoch": 0.11526315789473685, + "grad_norm": 0.79296875, + "learning_rate": 0.00019660818713450293, + "loss": 4.9017, + "step": 2190 + }, + { + "epoch": 0.1153157894736842, + "grad_norm": 0.74609375, + "learning_rate": 0.00019659649122807018, + "loss": 5.4105, + "step": 2191 + }, + { + "epoch": 0.11536842105263158, + "grad_norm": 0.78125, + "learning_rate": 0.00019658479532163743, + "loss": 5.0143, + "step": 2192 + }, + { + "epoch": 0.11542105263157895, + "grad_norm": 0.71484375, + "learning_rate": 0.0001965730994152047, + "loss": 5.0973, + "step": 2193 + }, + { + "epoch": 0.11547368421052631, + "grad_norm": 0.8046875, + "learning_rate": 0.00019656140350877196, + "loss": 5.1798, + "step": 2194 + }, + { + "epoch": 0.11552631578947369, + "grad_norm": 0.7578125, + "learning_rate": 0.00019654970760233918, + "loss": 5.1248, + "step": 2195 + }, + { + "epoch": 0.11557894736842106, + "grad_norm": 0.6875, + "learning_rate": 0.00019653801169590643, + "loss": 5.4165, + "step": 2196 + }, + { + "epoch": 0.11563157894736842, + "grad_norm": 0.65625, + "learning_rate": 0.00019652631578947368, + "loss": 5.2367, + "step": 2197 + }, + { + "epoch": 0.11568421052631579, + "grad_norm": 0.86328125, + "learning_rate": 0.00019651461988304093, + "loss": 5.3908, + "step": 2198 + }, + { + "epoch": 0.11573684210526315, + "grad_norm": 0.65625, + "learning_rate": 0.00019650292397660818, + "loss": 5.269, + "step": 2199 + }, + { + "epoch": 0.11578947368421053, + "grad_norm": 0.796875, + "learning_rate": 0.00019649122807017543, + "loss": 4.9807, + "step": 2200 + }, + { + "epoch": 0.1158421052631579, + "grad_norm": 0.703125, + "learning_rate": 0.0001964795321637427, + "loss": 5.5706, + "step": 2201 + }, + { + "epoch": 0.11589473684210526, + "grad_norm": 0.75, + "learning_rate": 0.00019646783625730996, + "loss": 5.2532, + "step": 2202 + }, + { + "epoch": 0.11594736842105263, + "grad_norm": 0.6875, + "learning_rate": 0.0001964561403508772, + "loss": 5.1947, + "step": 2203 + }, + { + "epoch": 0.116, + "grad_norm": 0.75390625, + "learning_rate": 0.00019644444444444445, + "loss": 5.0227, + "step": 2204 + }, + { + "epoch": 0.11605263157894737, + "grad_norm": 0.74609375, + "learning_rate": 0.0001964327485380117, + "loss": 4.9852, + "step": 2205 + }, + { + "epoch": 0.11610526315789474, + "grad_norm": 0.85546875, + "learning_rate": 0.00019642105263157895, + "loss": 5.3639, + "step": 2206 + }, + { + "epoch": 0.1161578947368421, + "grad_norm": 0.75, + "learning_rate": 0.0001964093567251462, + "loss": 5.0212, + "step": 2207 + }, + { + "epoch": 0.11621052631578947, + "grad_norm": 0.71484375, + "learning_rate": 0.00019639766081871348, + "loss": 4.8767, + "step": 2208 + }, + { + "epoch": 0.11626315789473685, + "grad_norm": 0.69140625, + "learning_rate": 0.00019638596491228073, + "loss": 5.2239, + "step": 2209 + }, + { + "epoch": 0.1163157894736842, + "grad_norm": 0.734375, + "learning_rate": 0.00019637426900584798, + "loss": 4.9579, + "step": 2210 + }, + { + "epoch": 0.11636842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.0001963625730994152, + "loss": 5.1199, + "step": 2211 + }, + { + "epoch": 0.11642105263157895, + "grad_norm": 0.75, + "learning_rate": 0.00019635087719298245, + "loss": 5.2978, + "step": 2212 + }, + { + "epoch": 0.11647368421052631, + "grad_norm": 1.0078125, + "learning_rate": 0.0001963391812865497, + "loss": 5.3084, + "step": 2213 + }, + { + "epoch": 0.11652631578947369, + "grad_norm": 0.57421875, + "learning_rate": 0.00019632748538011695, + "loss": 5.3259, + "step": 2214 + }, + { + "epoch": 0.11657894736842105, + "grad_norm": 0.76171875, + "learning_rate": 0.00019631578947368423, + "loss": 5.239, + "step": 2215 + }, + { + "epoch": 0.11663157894736842, + "grad_norm": 0.58984375, + "learning_rate": 0.00019630409356725148, + "loss": 5.2814, + "step": 2216 + }, + { + "epoch": 0.11668421052631579, + "grad_norm": 0.84375, + "learning_rate": 0.00019629239766081873, + "loss": 5.2026, + "step": 2217 + }, + { + "epoch": 0.11673684210526315, + "grad_norm": 1.0625, + "learning_rate": 0.00019628070175438598, + "loss": 5.3972, + "step": 2218 + }, + { + "epoch": 0.11678947368421053, + "grad_norm": 0.671875, + "learning_rate": 0.00019626900584795323, + "loss": 5.3539, + "step": 2219 + }, + { + "epoch": 0.1168421052631579, + "grad_norm": 0.9453125, + "learning_rate": 0.00019625730994152048, + "loss": 5.1173, + "step": 2220 + }, + { + "epoch": 0.11689473684210526, + "grad_norm": 0.73046875, + "learning_rate": 0.00019624561403508773, + "loss": 5.4106, + "step": 2221 + }, + { + "epoch": 0.11694736842105263, + "grad_norm": 0.6796875, + "learning_rate": 0.00019623391812865497, + "loss": 5.3149, + "step": 2222 + }, + { + "epoch": 0.117, + "grad_norm": 0.6640625, + "learning_rate": 0.00019622222222222225, + "loss": 5.4413, + "step": 2223 + }, + { + "epoch": 0.11705263157894737, + "grad_norm": 0.79296875, + "learning_rate": 0.0001962105263157895, + "loss": 5.4404, + "step": 2224 + }, + { + "epoch": 0.11710526315789474, + "grad_norm": 0.734375, + "learning_rate": 0.00019619883040935675, + "loss": 4.6299, + "step": 2225 + }, + { + "epoch": 0.1171578947368421, + "grad_norm": 0.953125, + "learning_rate": 0.00019618713450292397, + "loss": 5.2151, + "step": 2226 + }, + { + "epoch": 0.11721052631578947, + "grad_norm": 0.6015625, + "learning_rate": 0.00019617543859649122, + "loss": 5.1842, + "step": 2227 + }, + { + "epoch": 0.11726315789473685, + "grad_norm": 0.73828125, + "learning_rate": 0.00019616374269005847, + "loss": 5.4594, + "step": 2228 + }, + { + "epoch": 0.1173157894736842, + "grad_norm": 0.984375, + "learning_rate": 0.00019615204678362572, + "loss": 5.2332, + "step": 2229 + }, + { + "epoch": 0.11736842105263158, + "grad_norm": 0.5859375, + "learning_rate": 0.000196140350877193, + "loss": 5.0852, + "step": 2230 + }, + { + "epoch": 0.11742105263157895, + "grad_norm": 0.5859375, + "learning_rate": 0.00019612865497076025, + "loss": 5.0435, + "step": 2231 + }, + { + "epoch": 0.11747368421052631, + "grad_norm": 0.70703125, + "learning_rate": 0.0001961169590643275, + "loss": 5.161, + "step": 2232 + }, + { + "epoch": 0.11752631578947369, + "grad_norm": 0.68359375, + "learning_rate": 0.00019610526315789475, + "loss": 5.0615, + "step": 2233 + }, + { + "epoch": 0.11757894736842105, + "grad_norm": 0.625, + "learning_rate": 0.000196093567251462, + "loss": 5.2988, + "step": 2234 + }, + { + "epoch": 0.11763157894736842, + "grad_norm": 0.7890625, + "learning_rate": 0.00019608187134502925, + "loss": 4.9453, + "step": 2235 + }, + { + "epoch": 0.1176842105263158, + "grad_norm": 0.828125, + "learning_rate": 0.0001960701754385965, + "loss": 4.7842, + "step": 2236 + }, + { + "epoch": 0.11773684210526315, + "grad_norm": 0.75390625, + "learning_rate": 0.00019605847953216375, + "loss": 4.776, + "step": 2237 + }, + { + "epoch": 0.11778947368421053, + "grad_norm": 0.71484375, + "learning_rate": 0.00019604678362573102, + "loss": 4.9194, + "step": 2238 + }, + { + "epoch": 0.1178421052631579, + "grad_norm": 0.79296875, + "learning_rate": 0.00019603508771929827, + "loss": 5.1149, + "step": 2239 + }, + { + "epoch": 0.11789473684210526, + "grad_norm": 0.83984375, + "learning_rate": 0.00019602339181286552, + "loss": 5.1394, + "step": 2240 + }, + { + "epoch": 0.11794736842105263, + "grad_norm": 0.63671875, + "learning_rate": 0.00019601169590643274, + "loss": 5.0731, + "step": 2241 + }, + { + "epoch": 0.118, + "grad_norm": 0.62109375, + "learning_rate": 0.000196, + "loss": 4.9852, + "step": 2242 + }, + { + "epoch": 0.11805263157894737, + "grad_norm": 0.703125, + "learning_rate": 0.00019598830409356724, + "loss": 5.2573, + "step": 2243 + }, + { + "epoch": 0.11810526315789474, + "grad_norm": 0.75390625, + "learning_rate": 0.0001959766081871345, + "loss": 5.0353, + "step": 2244 + }, + { + "epoch": 0.1181578947368421, + "grad_norm": 0.66015625, + "learning_rate": 0.00019596491228070177, + "loss": 5.0671, + "step": 2245 + }, + { + "epoch": 0.11821052631578947, + "grad_norm": 0.6796875, + "learning_rate": 0.00019595321637426902, + "loss": 5.4635, + "step": 2246 + }, + { + "epoch": 0.11826315789473685, + "grad_norm": 0.6875, + "learning_rate": 0.00019594152046783627, + "loss": 5.4165, + "step": 2247 + }, + { + "epoch": 0.11831578947368421, + "grad_norm": 0.9296875, + "learning_rate": 0.00019592982456140352, + "loss": 4.8574, + "step": 2248 + }, + { + "epoch": 0.11836842105263158, + "grad_norm": 0.69140625, + "learning_rate": 0.00019591812865497077, + "loss": 5.2216, + "step": 2249 + }, + { + "epoch": 0.11842105263157894, + "grad_norm": 0.609375, + "learning_rate": 0.00019590643274853802, + "loss": 5.0686, + "step": 2250 + }, + { + "epoch": 0.11847368421052631, + "grad_norm": 0.66015625, + "learning_rate": 0.00019589473684210527, + "loss": 5.1851, + "step": 2251 + }, + { + "epoch": 0.11852631578947369, + "grad_norm": 0.671875, + "learning_rate": 0.00019588304093567254, + "loss": 5.3315, + "step": 2252 + }, + { + "epoch": 0.11857894736842105, + "grad_norm": 0.79296875, + "learning_rate": 0.0001958713450292398, + "loss": 5.0684, + "step": 2253 + }, + { + "epoch": 0.11863157894736842, + "grad_norm": 0.61328125, + "learning_rate": 0.00019585964912280704, + "loss": 5.3341, + "step": 2254 + }, + { + "epoch": 0.1186842105263158, + "grad_norm": 0.66015625, + "learning_rate": 0.0001958479532163743, + "loss": 4.9823, + "step": 2255 + }, + { + "epoch": 0.11873684210526315, + "grad_norm": 0.72265625, + "learning_rate": 0.00019583625730994154, + "loss": 4.9083, + "step": 2256 + }, + { + "epoch": 0.11878947368421053, + "grad_norm": 0.796875, + "learning_rate": 0.00019582456140350876, + "loss": 4.8508, + "step": 2257 + }, + { + "epoch": 0.11884210526315789, + "grad_norm": 0.5625, + "learning_rate": 0.00019581286549707601, + "loss": 5.3643, + "step": 2258 + }, + { + "epoch": 0.11889473684210526, + "grad_norm": 0.6171875, + "learning_rate": 0.00019580116959064326, + "loss": 4.9386, + "step": 2259 + }, + { + "epoch": 0.11894736842105263, + "grad_norm": 0.98046875, + "learning_rate": 0.00019578947368421054, + "loss": 5.1839, + "step": 2260 + }, + { + "epoch": 0.119, + "grad_norm": 0.765625, + "learning_rate": 0.0001957777777777778, + "loss": 4.9986, + "step": 2261 + }, + { + "epoch": 0.11905263157894737, + "grad_norm": 0.6953125, + "learning_rate": 0.00019576608187134504, + "loss": 4.9076, + "step": 2262 + }, + { + "epoch": 0.11910526315789474, + "grad_norm": 0.71484375, + "learning_rate": 0.0001957543859649123, + "loss": 5.0333, + "step": 2263 + }, + { + "epoch": 0.1191578947368421, + "grad_norm": 0.83203125, + "learning_rate": 0.00019574269005847954, + "loss": 5.0367, + "step": 2264 + }, + { + "epoch": 0.11921052631578948, + "grad_norm": 0.6484375, + "learning_rate": 0.0001957309941520468, + "loss": 5.1933, + "step": 2265 + }, + { + "epoch": 0.11926315789473685, + "grad_norm": 0.65234375, + "learning_rate": 0.00019571929824561404, + "loss": 5.1795, + "step": 2266 + }, + { + "epoch": 0.11931578947368421, + "grad_norm": 0.625, + "learning_rate": 0.00019570760233918131, + "loss": 5.3055, + "step": 2267 + }, + { + "epoch": 0.11936842105263158, + "grad_norm": 0.6015625, + "learning_rate": 0.00019569590643274856, + "loss": 5.2283, + "step": 2268 + }, + { + "epoch": 0.11942105263157894, + "grad_norm": 0.69921875, + "learning_rate": 0.00019568421052631581, + "loss": 5.0748, + "step": 2269 + }, + { + "epoch": 0.11947368421052632, + "grad_norm": 0.64453125, + "learning_rate": 0.00019567251461988306, + "loss": 5.4106, + "step": 2270 + }, + { + "epoch": 0.11952631578947369, + "grad_norm": 0.6171875, + "learning_rate": 0.0001956608187134503, + "loss": 5.326, + "step": 2271 + }, + { + "epoch": 0.11957894736842105, + "grad_norm": 0.765625, + "learning_rate": 0.00019564912280701754, + "loss": 4.8746, + "step": 2272 + }, + { + "epoch": 0.11963157894736842, + "grad_norm": 0.9140625, + "learning_rate": 0.00019563742690058479, + "loss": 5.1729, + "step": 2273 + }, + { + "epoch": 0.1196842105263158, + "grad_norm": 0.62890625, + "learning_rate": 0.00019562573099415203, + "loss": 5.0531, + "step": 2274 + }, + { + "epoch": 0.11973684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.0001956140350877193, + "loss": 5.3854, + "step": 2275 + }, + { + "epoch": 0.11978947368421053, + "grad_norm": 0.62109375, + "learning_rate": 0.00019560233918128656, + "loss": 5.1739, + "step": 2276 + }, + { + "epoch": 0.11984210526315789, + "grad_norm": 0.921875, + "learning_rate": 0.0001955906432748538, + "loss": 4.9726, + "step": 2277 + }, + { + "epoch": 0.11989473684210526, + "grad_norm": 0.6875, + "learning_rate": 0.00019557894736842106, + "loss": 4.9326, + "step": 2278 + }, + { + "epoch": 0.11994736842105264, + "grad_norm": 1.03125, + "learning_rate": 0.0001955672514619883, + "loss": 5.3223, + "step": 2279 + }, + { + "epoch": 0.12, + "grad_norm": 0.66796875, + "learning_rate": 0.00019555555555555556, + "loss": 5.118, + "step": 2280 + }, + { + "epoch": 0.12005263157894737, + "grad_norm": 0.63671875, + "learning_rate": 0.0001955438596491228, + "loss": 5.0095, + "step": 2281 + }, + { + "epoch": 0.12010526315789474, + "grad_norm": 0.640625, + "learning_rate": 0.00019553216374269009, + "loss": 4.9418, + "step": 2282 + }, + { + "epoch": 0.1201578947368421, + "grad_norm": 0.640625, + "learning_rate": 0.00019552046783625734, + "loss": 5.1198, + "step": 2283 + }, + { + "epoch": 0.12021052631578948, + "grad_norm": 0.8984375, + "learning_rate": 0.00019550877192982458, + "loss": 5.3403, + "step": 2284 + }, + { + "epoch": 0.12026315789473684, + "grad_norm": 0.74609375, + "learning_rate": 0.00019549707602339183, + "loss": 5.0352, + "step": 2285 + }, + { + "epoch": 0.12031578947368421, + "grad_norm": 0.73046875, + "learning_rate": 0.00019548538011695908, + "loss": 5.0949, + "step": 2286 + }, + { + "epoch": 0.12036842105263158, + "grad_norm": 0.92578125, + "learning_rate": 0.0001954736842105263, + "loss": 4.9364, + "step": 2287 + }, + { + "epoch": 0.12042105263157894, + "grad_norm": 0.796875, + "learning_rate": 0.00019546198830409356, + "loss": 5.0516, + "step": 2288 + }, + { + "epoch": 0.12047368421052632, + "grad_norm": 0.6796875, + "learning_rate": 0.00019545029239766083, + "loss": 5.1307, + "step": 2289 + }, + { + "epoch": 0.12052631578947369, + "grad_norm": 0.70703125, + "learning_rate": 0.00019543859649122808, + "loss": 5.1789, + "step": 2290 + }, + { + "epoch": 0.12057894736842105, + "grad_norm": 0.58984375, + "learning_rate": 0.00019542690058479533, + "loss": 5.2282, + "step": 2291 + }, + { + "epoch": 0.12063157894736842, + "grad_norm": 0.6875, + "learning_rate": 0.00019541520467836258, + "loss": 5.4607, + "step": 2292 + }, + { + "epoch": 0.12068421052631578, + "grad_norm": 0.76953125, + "learning_rate": 0.00019540350877192983, + "loss": 5.1581, + "step": 2293 + }, + { + "epoch": 0.12073684210526316, + "grad_norm": 0.859375, + "learning_rate": 0.00019539181286549708, + "loss": 5.0334, + "step": 2294 + }, + { + "epoch": 0.12078947368421053, + "grad_norm": 0.625, + "learning_rate": 0.00019538011695906433, + "loss": 5.6691, + "step": 2295 + }, + { + "epoch": 0.12084210526315789, + "grad_norm": 0.70703125, + "learning_rate": 0.00019536842105263158, + "loss": 5.2761, + "step": 2296 + }, + { + "epoch": 0.12089473684210526, + "grad_norm": 0.72265625, + "learning_rate": 0.00019535672514619886, + "loss": 4.7108, + "step": 2297 + }, + { + "epoch": 0.12094736842105264, + "grad_norm": 0.98046875, + "learning_rate": 0.0001953450292397661, + "loss": 4.9053, + "step": 2298 + }, + { + "epoch": 0.121, + "grad_norm": 0.63671875, + "learning_rate": 0.00019533333333333336, + "loss": 5.1995, + "step": 2299 + }, + { + "epoch": 0.12105263157894737, + "grad_norm": 0.6484375, + "learning_rate": 0.0001953216374269006, + "loss": 4.5747, + "step": 2300 + }, + { + "epoch": 0.12110526315789474, + "grad_norm": 0.6484375, + "learning_rate": 0.00019530994152046786, + "loss": 5.0607, + "step": 2301 + }, + { + "epoch": 0.1211578947368421, + "grad_norm": 0.70703125, + "learning_rate": 0.0001952982456140351, + "loss": 4.8728, + "step": 2302 + }, + { + "epoch": 0.12121052631578948, + "grad_norm": 0.69921875, + "learning_rate": 0.00019528654970760233, + "loss": 5.2048, + "step": 2303 + }, + { + "epoch": 0.12126315789473684, + "grad_norm": 0.8984375, + "learning_rate": 0.0001952748538011696, + "loss": 5.3109, + "step": 2304 + }, + { + "epoch": 0.12131578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00019526315789473685, + "loss": 5.1703, + "step": 2305 + }, + { + "epoch": 0.12136842105263158, + "grad_norm": 0.625, + "learning_rate": 0.0001952514619883041, + "loss": 4.9505, + "step": 2306 + }, + { + "epoch": 0.12142105263157894, + "grad_norm": 0.66796875, + "learning_rate": 0.00019523976608187135, + "loss": 5.257, + "step": 2307 + }, + { + "epoch": 0.12147368421052632, + "grad_norm": 0.703125, + "learning_rate": 0.0001952280701754386, + "loss": 5.0451, + "step": 2308 + }, + { + "epoch": 0.12152631578947369, + "grad_norm": 0.69140625, + "learning_rate": 0.00019521637426900585, + "loss": 5.1531, + "step": 2309 + }, + { + "epoch": 0.12157894736842105, + "grad_norm": 0.62890625, + "learning_rate": 0.0001952046783625731, + "loss": 5.3138, + "step": 2310 + }, + { + "epoch": 0.12163157894736842, + "grad_norm": 0.65234375, + "learning_rate": 0.00019519298245614035, + "loss": 5.0788, + "step": 2311 + }, + { + "epoch": 0.12168421052631578, + "grad_norm": 0.734375, + "learning_rate": 0.00019518128654970763, + "loss": 5.3256, + "step": 2312 + }, + { + "epoch": 0.12173684210526316, + "grad_norm": 0.63671875, + "learning_rate": 0.00019516959064327488, + "loss": 5.2997, + "step": 2313 + }, + { + "epoch": 0.12178947368421053, + "grad_norm": 0.6796875, + "learning_rate": 0.00019515789473684213, + "loss": 5.3302, + "step": 2314 + }, + { + "epoch": 0.12184210526315789, + "grad_norm": 0.73046875, + "learning_rate": 0.00019514619883040938, + "loss": 4.9241, + "step": 2315 + }, + { + "epoch": 0.12189473684210526, + "grad_norm": 0.83203125, + "learning_rate": 0.00019513450292397663, + "loss": 4.9665, + "step": 2316 + }, + { + "epoch": 0.12194736842105264, + "grad_norm": 0.7265625, + "learning_rate": 0.00019512280701754388, + "loss": 4.9967, + "step": 2317 + }, + { + "epoch": 0.122, + "grad_norm": 0.64453125, + "learning_rate": 0.0001951111111111111, + "loss": 4.814, + "step": 2318 + }, + { + "epoch": 0.12205263157894737, + "grad_norm": 0.69921875, + "learning_rate": 0.00019509941520467837, + "loss": 5.308, + "step": 2319 + }, + { + "epoch": 0.12210526315789473, + "grad_norm": 0.6796875, + "learning_rate": 0.00019508771929824562, + "loss": 5.2257, + "step": 2320 + }, + { + "epoch": 0.1221578947368421, + "grad_norm": 0.6875, + "learning_rate": 0.00019507602339181287, + "loss": 5.2393, + "step": 2321 + }, + { + "epoch": 0.12221052631578948, + "grad_norm": 0.99609375, + "learning_rate": 0.00019506432748538012, + "loss": 4.6521, + "step": 2322 + }, + { + "epoch": 0.12226315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.00019505263157894737, + "loss": 5.3016, + "step": 2323 + }, + { + "epoch": 0.12231578947368421, + "grad_norm": 0.62109375, + "learning_rate": 0.00019504093567251462, + "loss": 5.0342, + "step": 2324 + }, + { + "epoch": 0.12236842105263158, + "grad_norm": 0.62109375, + "learning_rate": 0.00019502923976608187, + "loss": 4.7746, + "step": 2325 + }, + { + "epoch": 0.12242105263157894, + "grad_norm": 0.66796875, + "learning_rate": 0.00019501754385964912, + "loss": 4.8801, + "step": 2326 + }, + { + "epoch": 0.12247368421052632, + "grad_norm": 0.53125, + "learning_rate": 0.0001950058479532164, + "loss": 5.2247, + "step": 2327 + }, + { + "epoch": 0.12252631578947368, + "grad_norm": 0.64453125, + "learning_rate": 0.00019499415204678365, + "loss": 5.212, + "step": 2328 + }, + { + "epoch": 0.12257894736842105, + "grad_norm": 0.67578125, + "learning_rate": 0.0001949824561403509, + "loss": 5.4555, + "step": 2329 + }, + { + "epoch": 0.12263157894736842, + "grad_norm": 0.875, + "learning_rate": 0.00019497076023391815, + "loss": 5.0176, + "step": 2330 + }, + { + "epoch": 0.12268421052631578, + "grad_norm": 0.63671875, + "learning_rate": 0.0001949590643274854, + "loss": 4.9535, + "step": 2331 + }, + { + "epoch": 0.12273684210526316, + "grad_norm": 0.765625, + "learning_rate": 0.00019494736842105265, + "loss": 4.6845, + "step": 2332 + }, + { + "epoch": 0.12278947368421053, + "grad_norm": 0.55859375, + "learning_rate": 0.00019493567251461987, + "loss": 5.3827, + "step": 2333 + }, + { + "epoch": 0.12284210526315789, + "grad_norm": 0.67578125, + "learning_rate": 0.00019492397660818715, + "loss": 5.0692, + "step": 2334 + }, + { + "epoch": 0.12289473684210526, + "grad_norm": 0.63671875, + "learning_rate": 0.0001949122807017544, + "loss": 5.1762, + "step": 2335 + }, + { + "epoch": 0.12294736842105262, + "grad_norm": 0.64453125, + "learning_rate": 0.00019490058479532165, + "loss": 4.8147, + "step": 2336 + }, + { + "epoch": 0.123, + "grad_norm": 0.6328125, + "learning_rate": 0.0001948888888888889, + "loss": 4.6483, + "step": 2337 + }, + { + "epoch": 0.12305263157894737, + "grad_norm": 0.640625, + "learning_rate": 0.00019487719298245614, + "loss": 4.8528, + "step": 2338 + }, + { + "epoch": 0.12310526315789473, + "grad_norm": 0.69921875, + "learning_rate": 0.0001948654970760234, + "loss": 5.3065, + "step": 2339 + }, + { + "epoch": 0.1231578947368421, + "grad_norm": 0.9921875, + "learning_rate": 0.00019485380116959064, + "loss": 5.3075, + "step": 2340 + }, + { + "epoch": 0.12321052631578948, + "grad_norm": 0.62890625, + "learning_rate": 0.00019484210526315792, + "loss": 5.1533, + "step": 2341 + }, + { + "epoch": 0.12326315789473684, + "grad_norm": 0.609375, + "learning_rate": 0.00019483040935672517, + "loss": 5.2281, + "step": 2342 + }, + { + "epoch": 0.12331578947368421, + "grad_norm": 1.6484375, + "learning_rate": 0.00019481871345029242, + "loss": 5.5381, + "step": 2343 + }, + { + "epoch": 0.12336842105263159, + "grad_norm": 0.6171875, + "learning_rate": 0.00019480701754385967, + "loss": 5.1961, + "step": 2344 + }, + { + "epoch": 0.12342105263157895, + "grad_norm": 0.75390625, + "learning_rate": 0.00019479532163742692, + "loss": 5.2055, + "step": 2345 + }, + { + "epoch": 0.12347368421052632, + "grad_norm": 0.71484375, + "learning_rate": 0.00019478362573099417, + "loss": 5.0414, + "step": 2346 + }, + { + "epoch": 0.12352631578947368, + "grad_norm": 0.67578125, + "learning_rate": 0.00019477192982456142, + "loss": 4.936, + "step": 2347 + }, + { + "epoch": 0.12357894736842105, + "grad_norm": 0.65625, + "learning_rate": 0.00019476023391812867, + "loss": 4.992, + "step": 2348 + }, + { + "epoch": 0.12363157894736843, + "grad_norm": 0.64453125, + "learning_rate": 0.00019474853801169592, + "loss": 4.9923, + "step": 2349 + }, + { + "epoch": 0.12368421052631579, + "grad_norm": 0.6328125, + "learning_rate": 0.00019473684210526317, + "loss": 5.1768, + "step": 2350 + }, + { + "epoch": 0.12373684210526316, + "grad_norm": 0.98828125, + "learning_rate": 0.00019472514619883042, + "loss": 5.2238, + "step": 2351 + }, + { + "epoch": 0.12378947368421053, + "grad_norm": 0.59375, + "learning_rate": 0.00019471345029239767, + "loss": 5.0653, + "step": 2352 + }, + { + "epoch": 0.12384210526315789, + "grad_norm": 0.59765625, + "learning_rate": 0.00019470175438596492, + "loss": 5.1955, + "step": 2353 + }, + { + "epoch": 0.12389473684210527, + "grad_norm": 0.6640625, + "learning_rate": 0.00019469005847953216, + "loss": 5.0932, + "step": 2354 + }, + { + "epoch": 0.12394736842105263, + "grad_norm": 0.7109375, + "learning_rate": 0.00019467836257309941, + "loss": 5.3607, + "step": 2355 + }, + { + "epoch": 0.124, + "grad_norm": 0.703125, + "learning_rate": 0.0001946666666666667, + "loss": 5.3178, + "step": 2356 + }, + { + "epoch": 0.12405263157894737, + "grad_norm": 0.6796875, + "learning_rate": 0.00019465497076023394, + "loss": 5.0334, + "step": 2357 + }, + { + "epoch": 0.12410526315789473, + "grad_norm": 0.79296875, + "learning_rate": 0.0001946432748538012, + "loss": 5.191, + "step": 2358 + }, + { + "epoch": 0.1241578947368421, + "grad_norm": 0.84765625, + "learning_rate": 0.00019463157894736844, + "loss": 4.9839, + "step": 2359 + }, + { + "epoch": 0.12421052631578948, + "grad_norm": 0.671875, + "learning_rate": 0.0001946198830409357, + "loss": 5.0581, + "step": 2360 + }, + { + "epoch": 0.12426315789473684, + "grad_norm": 0.69140625, + "learning_rate": 0.00019460818713450294, + "loss": 5.0102, + "step": 2361 + }, + { + "epoch": 0.12431578947368421, + "grad_norm": 0.890625, + "learning_rate": 0.0001945964912280702, + "loss": 5.008, + "step": 2362 + }, + { + "epoch": 0.12436842105263157, + "grad_norm": 0.79296875, + "learning_rate": 0.00019458479532163744, + "loss": 5.0867, + "step": 2363 + }, + { + "epoch": 0.12442105263157895, + "grad_norm": 0.6953125, + "learning_rate": 0.0001945730994152047, + "loss": 5.1115, + "step": 2364 + }, + { + "epoch": 0.12447368421052632, + "grad_norm": 0.83203125, + "learning_rate": 0.00019456140350877194, + "loss": 5.1272, + "step": 2365 + }, + { + "epoch": 0.12452631578947368, + "grad_norm": 1.09375, + "learning_rate": 0.0001945497076023392, + "loss": 4.6986, + "step": 2366 + }, + { + "epoch": 0.12457894736842105, + "grad_norm": 0.77734375, + "learning_rate": 0.00019453801169590644, + "loss": 5.0736, + "step": 2367 + }, + { + "epoch": 0.12463157894736843, + "grad_norm": 0.8359375, + "learning_rate": 0.00019452631578947369, + "loss": 4.7161, + "step": 2368 + }, + { + "epoch": 0.12468421052631579, + "grad_norm": 0.72265625, + "learning_rate": 0.00019451461988304094, + "loss": 4.9505, + "step": 2369 + }, + { + "epoch": 0.12473684210526316, + "grad_norm": 0.734375, + "learning_rate": 0.00019450292397660819, + "loss": 5.1587, + "step": 2370 + }, + { + "epoch": 0.12478947368421052, + "grad_norm": 0.640625, + "learning_rate": 0.00019449122807017546, + "loss": 4.7971, + "step": 2371 + }, + { + "epoch": 0.1248421052631579, + "grad_norm": 0.59375, + "learning_rate": 0.0001944795321637427, + "loss": 4.9983, + "step": 2372 + }, + { + "epoch": 0.12489473684210527, + "grad_norm": 0.66015625, + "learning_rate": 0.00019446783625730996, + "loss": 4.7944, + "step": 2373 + }, + { + "epoch": 0.12494736842105263, + "grad_norm": 0.6484375, + "learning_rate": 0.0001944561403508772, + "loss": 4.9912, + "step": 2374 + }, + { + "epoch": 0.125, + "grad_norm": 0.671875, + "learning_rate": 0.00019444444444444446, + "loss": 4.8803, + "step": 2375 + }, + { + "epoch": 0.12505263157894736, + "grad_norm": 0.62890625, + "learning_rate": 0.0001944327485380117, + "loss": 5.1856, + "step": 2376 + }, + { + "epoch": 0.12510526315789475, + "grad_norm": 0.65234375, + "learning_rate": 0.00019442105263157896, + "loss": 5.0788, + "step": 2377 + }, + { + "epoch": 0.1251578947368421, + "grad_norm": 0.640625, + "learning_rate": 0.0001944093567251462, + "loss": 5.063, + "step": 2378 + }, + { + "epoch": 0.12521052631578947, + "grad_norm": 0.87109375, + "learning_rate": 0.00019439766081871346, + "loss": 5.0238, + "step": 2379 + }, + { + "epoch": 0.12526315789473685, + "grad_norm": 0.67578125, + "learning_rate": 0.0001943859649122807, + "loss": 4.9676, + "step": 2380 + }, + { + "epoch": 0.1253157894736842, + "grad_norm": 0.67578125, + "learning_rate": 0.00019437426900584796, + "loss": 5.0031, + "step": 2381 + }, + { + "epoch": 0.12536842105263157, + "grad_norm": 0.6640625, + "learning_rate": 0.0001943625730994152, + "loss": 5.2855, + "step": 2382 + }, + { + "epoch": 0.12542105263157896, + "grad_norm": 0.65234375, + "learning_rate": 0.00019435087719298246, + "loss": 4.8462, + "step": 2383 + }, + { + "epoch": 0.12547368421052632, + "grad_norm": 0.5859375, + "learning_rate": 0.0001943391812865497, + "loss": 5.0014, + "step": 2384 + }, + { + "epoch": 0.12552631578947368, + "grad_norm": 0.640625, + "learning_rate": 0.00019432748538011696, + "loss": 5.0329, + "step": 2385 + }, + { + "epoch": 0.12557894736842104, + "grad_norm": 0.625, + "learning_rate": 0.00019431578947368423, + "loss": 5.2529, + "step": 2386 + }, + { + "epoch": 0.12563157894736843, + "grad_norm": 0.59765625, + "learning_rate": 0.00019430409356725148, + "loss": 4.9002, + "step": 2387 + }, + { + "epoch": 0.1256842105263158, + "grad_norm": 0.734375, + "learning_rate": 0.00019429239766081873, + "loss": 5.0113, + "step": 2388 + }, + { + "epoch": 0.12573684210526315, + "grad_norm": 0.57421875, + "learning_rate": 0.00019428070175438598, + "loss": 4.774, + "step": 2389 + }, + { + "epoch": 0.12578947368421053, + "grad_norm": 0.640625, + "learning_rate": 0.00019426900584795323, + "loss": 5.0354, + "step": 2390 + }, + { + "epoch": 0.1258421052631579, + "grad_norm": 0.7578125, + "learning_rate": 0.00019425730994152048, + "loss": 5.164, + "step": 2391 + }, + { + "epoch": 0.12589473684210525, + "grad_norm": 0.953125, + "learning_rate": 0.00019424561403508773, + "loss": 5.0743, + "step": 2392 + }, + { + "epoch": 0.12594736842105264, + "grad_norm": 0.72265625, + "learning_rate": 0.00019423391812865498, + "loss": 4.7617, + "step": 2393 + }, + { + "epoch": 0.126, + "grad_norm": 0.5625, + "learning_rate": 0.00019422222222222223, + "loss": 5.0892, + "step": 2394 + }, + { + "epoch": 0.12605263157894736, + "grad_norm": 0.6640625, + "learning_rate": 0.00019421052631578948, + "loss": 5.0172, + "step": 2395 + }, + { + "epoch": 0.12610526315789475, + "grad_norm": 0.70703125, + "learning_rate": 0.00019419883040935673, + "loss": 4.9566, + "step": 2396 + }, + { + "epoch": 0.1261578947368421, + "grad_norm": 0.69140625, + "learning_rate": 0.00019418713450292398, + "loss": 5.0714, + "step": 2397 + }, + { + "epoch": 0.12621052631578947, + "grad_norm": 0.72265625, + "learning_rate": 0.00019417543859649123, + "loss": 5.2632, + "step": 2398 + }, + { + "epoch": 0.12626315789473685, + "grad_norm": 0.65234375, + "learning_rate": 0.00019416374269005848, + "loss": 4.9159, + "step": 2399 + }, + { + "epoch": 0.12631578947368421, + "grad_norm": 0.76171875, + "learning_rate": 0.00019415204678362573, + "loss": 5.0556, + "step": 2400 + }, + { + "epoch": 0.12636842105263157, + "grad_norm": 0.68359375, + "learning_rate": 0.000194140350877193, + "loss": 5.0174, + "step": 2401 + }, + { + "epoch": 0.12642105263157893, + "grad_norm": 0.57421875, + "learning_rate": 0.00019412865497076025, + "loss": 5.1647, + "step": 2402 + }, + { + "epoch": 0.12647368421052632, + "grad_norm": 0.67578125, + "learning_rate": 0.0001941169590643275, + "loss": 4.8902, + "step": 2403 + }, + { + "epoch": 0.12652631578947368, + "grad_norm": 0.546875, + "learning_rate": 0.00019410526315789475, + "loss": 5.0784, + "step": 2404 + }, + { + "epoch": 0.12657894736842104, + "grad_norm": 0.5546875, + "learning_rate": 0.000194093567251462, + "loss": 5.0691, + "step": 2405 + }, + { + "epoch": 0.12663157894736843, + "grad_norm": 0.7890625, + "learning_rate": 0.00019408187134502925, + "loss": 4.8549, + "step": 2406 + }, + { + "epoch": 0.1266842105263158, + "grad_norm": 0.6171875, + "learning_rate": 0.0001940701754385965, + "loss": 5.0816, + "step": 2407 + }, + { + "epoch": 0.12673684210526315, + "grad_norm": 0.60546875, + "learning_rate": 0.00019405847953216375, + "loss": 4.8458, + "step": 2408 + }, + { + "epoch": 0.12678947368421054, + "grad_norm": 0.53125, + "learning_rate": 0.000194046783625731, + "loss": 5.098, + "step": 2409 + }, + { + "epoch": 0.1268421052631579, + "grad_norm": 0.6171875, + "learning_rate": 0.00019403508771929825, + "loss": 4.9926, + "step": 2410 + }, + { + "epoch": 0.12689473684210525, + "grad_norm": 0.640625, + "learning_rate": 0.0001940233918128655, + "loss": 5.102, + "step": 2411 + }, + { + "epoch": 0.12694736842105264, + "grad_norm": 0.76171875, + "learning_rate": 0.00019401169590643275, + "loss": 4.8715, + "step": 2412 + }, + { + "epoch": 0.127, + "grad_norm": 0.734375, + "learning_rate": 0.000194, + "loss": 5.3035, + "step": 2413 + }, + { + "epoch": 0.12705263157894736, + "grad_norm": 0.85546875, + "learning_rate": 0.00019398830409356725, + "loss": 4.9086, + "step": 2414 + }, + { + "epoch": 0.12710526315789475, + "grad_norm": 0.8359375, + "learning_rate": 0.00019397660818713453, + "loss": 5.0509, + "step": 2415 + }, + { + "epoch": 0.1271578947368421, + "grad_norm": 0.65234375, + "learning_rate": 0.00019396491228070178, + "loss": 5.1182, + "step": 2416 + }, + { + "epoch": 0.12721052631578947, + "grad_norm": 0.69140625, + "learning_rate": 0.00019395321637426902, + "loss": 5.1281, + "step": 2417 + }, + { + "epoch": 0.12726315789473686, + "grad_norm": 0.703125, + "learning_rate": 0.00019394152046783627, + "loss": 4.9315, + "step": 2418 + }, + { + "epoch": 0.12731578947368422, + "grad_norm": 0.71484375, + "learning_rate": 0.00019392982456140352, + "loss": 4.6423, + "step": 2419 + }, + { + "epoch": 0.12736842105263158, + "grad_norm": 0.67578125, + "learning_rate": 0.00019391812865497077, + "loss": 4.973, + "step": 2420 + }, + { + "epoch": 0.12742105263157893, + "grad_norm": 0.7890625, + "learning_rate": 0.00019390643274853802, + "loss": 5.3098, + "step": 2421 + }, + { + "epoch": 0.12747368421052632, + "grad_norm": 0.80859375, + "learning_rate": 0.00019389473684210527, + "loss": 4.8353, + "step": 2422 + }, + { + "epoch": 0.12752631578947368, + "grad_norm": 0.54296875, + "learning_rate": 0.00019388304093567252, + "loss": 5.3817, + "step": 2423 + }, + { + "epoch": 0.12757894736842104, + "grad_norm": 0.71875, + "learning_rate": 0.00019387134502923977, + "loss": 4.5891, + "step": 2424 + }, + { + "epoch": 0.12763157894736843, + "grad_norm": 0.5703125, + "learning_rate": 0.00019385964912280702, + "loss": 5.1191, + "step": 2425 + }, + { + "epoch": 0.1276842105263158, + "grad_norm": 0.6171875, + "learning_rate": 0.00019384795321637427, + "loss": 4.8847, + "step": 2426 + }, + { + "epoch": 0.12773684210526315, + "grad_norm": 0.58984375, + "learning_rate": 0.00019383625730994152, + "loss": 4.9106, + "step": 2427 + }, + { + "epoch": 0.12778947368421054, + "grad_norm": 0.609375, + "learning_rate": 0.00019382456140350877, + "loss": 4.7795, + "step": 2428 + }, + { + "epoch": 0.1278421052631579, + "grad_norm": 0.62890625, + "learning_rate": 0.00019381286549707602, + "loss": 5.2717, + "step": 2429 + }, + { + "epoch": 0.12789473684210526, + "grad_norm": 0.6171875, + "learning_rate": 0.0001938011695906433, + "loss": 5.1624, + "step": 2430 + }, + { + "epoch": 0.12794736842105264, + "grad_norm": 1.015625, + "learning_rate": 0.00019378947368421055, + "loss": 4.9018, + "step": 2431 + }, + { + "epoch": 0.128, + "grad_norm": 0.7578125, + "learning_rate": 0.0001937777777777778, + "loss": 5.1228, + "step": 2432 + }, + { + "epoch": 0.12805263157894736, + "grad_norm": 0.5234375, + "learning_rate": 0.00019376608187134505, + "loss": 5.26, + "step": 2433 + }, + { + "epoch": 0.12810526315789475, + "grad_norm": 0.75390625, + "learning_rate": 0.0001937543859649123, + "loss": 5.0822, + "step": 2434 + }, + { + "epoch": 0.1281578947368421, + "grad_norm": 0.69140625, + "learning_rate": 0.00019374269005847954, + "loss": 5.1, + "step": 2435 + }, + { + "epoch": 0.12821052631578947, + "grad_norm": 0.6875, + "learning_rate": 0.0001937309941520468, + "loss": 4.8652, + "step": 2436 + }, + { + "epoch": 0.12826315789473683, + "grad_norm": 0.82421875, + "learning_rate": 0.00019371929824561404, + "loss": 4.9719, + "step": 2437 + }, + { + "epoch": 0.12831578947368422, + "grad_norm": 0.796875, + "learning_rate": 0.0001937076023391813, + "loss": 5.0175, + "step": 2438 + }, + { + "epoch": 0.12836842105263158, + "grad_norm": 0.67578125, + "learning_rate": 0.00019369590643274854, + "loss": 5.213, + "step": 2439 + }, + { + "epoch": 0.12842105263157894, + "grad_norm": 0.82421875, + "learning_rate": 0.0001936842105263158, + "loss": 5.0449, + "step": 2440 + }, + { + "epoch": 0.12847368421052632, + "grad_norm": 0.6171875, + "learning_rate": 0.00019367251461988304, + "loss": 5.1137, + "step": 2441 + }, + { + "epoch": 0.12852631578947368, + "grad_norm": 0.609375, + "learning_rate": 0.0001936608187134503, + "loss": 5.0049, + "step": 2442 + }, + { + "epoch": 0.12857894736842104, + "grad_norm": 0.578125, + "learning_rate": 0.00019364912280701754, + "loss": 4.8665, + "step": 2443 + }, + { + "epoch": 0.12863157894736843, + "grad_norm": 0.7109375, + "learning_rate": 0.0001936374269005848, + "loss": 5.2895, + "step": 2444 + }, + { + "epoch": 0.1286842105263158, + "grad_norm": 0.6484375, + "learning_rate": 0.00019362573099415207, + "loss": 5.0859, + "step": 2445 + }, + { + "epoch": 0.12873684210526315, + "grad_norm": 0.921875, + "learning_rate": 0.00019361403508771932, + "loss": 4.4901, + "step": 2446 + }, + { + "epoch": 0.12878947368421054, + "grad_norm": 0.5625, + "learning_rate": 0.00019360233918128657, + "loss": 5.0828, + "step": 2447 + }, + { + "epoch": 0.1288421052631579, + "grad_norm": 0.63671875, + "learning_rate": 0.00019359064327485382, + "loss": 5.0248, + "step": 2448 + }, + { + "epoch": 0.12889473684210526, + "grad_norm": 0.8515625, + "learning_rate": 0.00019357894736842107, + "loss": 4.9117, + "step": 2449 + }, + { + "epoch": 0.12894736842105264, + "grad_norm": 0.5390625, + "learning_rate": 0.00019356725146198832, + "loss": 4.8877, + "step": 2450 + }, + { + "epoch": 0.129, + "grad_norm": 0.70703125, + "learning_rate": 0.00019355555555555557, + "loss": 4.908, + "step": 2451 + }, + { + "epoch": 0.12905263157894736, + "grad_norm": 0.6953125, + "learning_rate": 0.00019354385964912281, + "loss": 5.1422, + "step": 2452 + }, + { + "epoch": 0.12910526315789475, + "grad_norm": 0.875, + "learning_rate": 0.00019353216374269006, + "loss": 4.958, + "step": 2453 + }, + { + "epoch": 0.1291578947368421, + "grad_norm": 0.6015625, + "learning_rate": 0.00019352046783625731, + "loss": 4.9821, + "step": 2454 + }, + { + "epoch": 0.12921052631578947, + "grad_norm": 0.64453125, + "learning_rate": 0.00019350877192982456, + "loss": 4.8059, + "step": 2455 + }, + { + "epoch": 0.12926315789473683, + "grad_norm": 0.84375, + "learning_rate": 0.0001934970760233918, + "loss": 4.4789, + "step": 2456 + }, + { + "epoch": 0.12931578947368422, + "grad_norm": 0.6015625, + "learning_rate": 0.00019348538011695906, + "loss": 5.1054, + "step": 2457 + }, + { + "epoch": 0.12936842105263158, + "grad_norm": 0.76171875, + "learning_rate": 0.0001934736842105263, + "loss": 4.7771, + "step": 2458 + }, + { + "epoch": 0.12942105263157894, + "grad_norm": 0.78125, + "learning_rate": 0.00019346198830409356, + "loss": 5.0105, + "step": 2459 + }, + { + "epoch": 0.12947368421052632, + "grad_norm": 0.859375, + "learning_rate": 0.00019345029239766084, + "loss": 4.8195, + "step": 2460 + }, + { + "epoch": 0.12952631578947368, + "grad_norm": 0.55078125, + "learning_rate": 0.0001934385964912281, + "loss": 5.2069, + "step": 2461 + }, + { + "epoch": 0.12957894736842104, + "grad_norm": 0.64453125, + "learning_rate": 0.00019342690058479534, + "loss": 5.1893, + "step": 2462 + }, + { + "epoch": 0.12963157894736843, + "grad_norm": 0.7578125, + "learning_rate": 0.0001934152046783626, + "loss": 5.058, + "step": 2463 + }, + { + "epoch": 0.1296842105263158, + "grad_norm": 0.65625, + "learning_rate": 0.00019340350877192984, + "loss": 5.4534, + "step": 2464 + }, + { + "epoch": 0.12973684210526315, + "grad_norm": 0.734375, + "learning_rate": 0.0001933918128654971, + "loss": 4.5618, + "step": 2465 + }, + { + "epoch": 0.12978947368421054, + "grad_norm": 0.69921875, + "learning_rate": 0.00019338011695906434, + "loss": 5.4299, + "step": 2466 + }, + { + "epoch": 0.1298421052631579, + "grad_norm": 0.7265625, + "learning_rate": 0.00019336842105263159, + "loss": 5.2914, + "step": 2467 + }, + { + "epoch": 0.12989473684210526, + "grad_norm": 0.77734375, + "learning_rate": 0.00019335672514619884, + "loss": 4.9809, + "step": 2468 + }, + { + "epoch": 0.12994736842105264, + "grad_norm": 0.703125, + "learning_rate": 0.00019334502923976608, + "loss": 4.912, + "step": 2469 + }, + { + "epoch": 0.13, + "grad_norm": 0.671875, + "learning_rate": 0.00019333333333333333, + "loss": 4.9305, + "step": 2470 + }, + { + "epoch": 0.13005263157894736, + "grad_norm": 0.609375, + "learning_rate": 0.00019332163742690058, + "loss": 5.065, + "step": 2471 + }, + { + "epoch": 0.13010526315789472, + "grad_norm": 0.58203125, + "learning_rate": 0.00019330994152046783, + "loss": 4.8423, + "step": 2472 + }, + { + "epoch": 0.1301578947368421, + "grad_norm": 0.703125, + "learning_rate": 0.00019329824561403508, + "loss": 4.7724, + "step": 2473 + }, + { + "epoch": 0.13021052631578947, + "grad_norm": 0.5859375, + "learning_rate": 0.00019328654970760233, + "loss": 5.083, + "step": 2474 + }, + { + "epoch": 0.13026315789473683, + "grad_norm": 0.81640625, + "learning_rate": 0.0001932748538011696, + "loss": 4.8683, + "step": 2475 + }, + { + "epoch": 0.13031578947368422, + "grad_norm": 0.6875, + "learning_rate": 0.00019326315789473686, + "loss": 4.8649, + "step": 2476 + }, + { + "epoch": 0.13036842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.0001932514619883041, + "loss": 4.8503, + "step": 2477 + }, + { + "epoch": 0.13042105263157894, + "grad_norm": 0.55859375, + "learning_rate": 0.00019323976608187136, + "loss": 4.9395, + "step": 2478 + }, + { + "epoch": 0.13047368421052633, + "grad_norm": 0.91015625, + "learning_rate": 0.0001932280701754386, + "loss": 4.9142, + "step": 2479 + }, + { + "epoch": 0.13052631578947368, + "grad_norm": 0.5703125, + "learning_rate": 0.00019321637426900586, + "loss": 4.8942, + "step": 2480 + }, + { + "epoch": 0.13057894736842104, + "grad_norm": 0.58984375, + "learning_rate": 0.0001932046783625731, + "loss": 5.0309, + "step": 2481 + }, + { + "epoch": 0.13063157894736843, + "grad_norm": 0.6953125, + "learning_rate": 0.00019319298245614036, + "loss": 4.9306, + "step": 2482 + }, + { + "epoch": 0.1306842105263158, + "grad_norm": 0.75390625, + "learning_rate": 0.0001931812865497076, + "loss": 4.8939, + "step": 2483 + }, + { + "epoch": 0.13073684210526315, + "grad_norm": 1.0390625, + "learning_rate": 0.00019316959064327486, + "loss": 4.7576, + "step": 2484 + }, + { + "epoch": 0.13078947368421054, + "grad_norm": 0.59375, + "learning_rate": 0.0001931578947368421, + "loss": 5.0496, + "step": 2485 + }, + { + "epoch": 0.1308421052631579, + "grad_norm": 0.66015625, + "learning_rate": 0.00019314619883040936, + "loss": 5.3095, + "step": 2486 + }, + { + "epoch": 0.13089473684210526, + "grad_norm": 0.6875, + "learning_rate": 0.0001931345029239766, + "loss": 4.8546, + "step": 2487 + }, + { + "epoch": 0.13094736842105262, + "grad_norm": 0.625, + "learning_rate": 0.00019312280701754385, + "loss": 4.8966, + "step": 2488 + }, + { + "epoch": 0.131, + "grad_norm": 0.66015625, + "learning_rate": 0.0001931111111111111, + "loss": 5.2977, + "step": 2489 + }, + { + "epoch": 0.13105263157894737, + "grad_norm": 0.60546875, + "learning_rate": 0.00019309941520467838, + "loss": 5.126, + "step": 2490 + }, + { + "epoch": 0.13110526315789472, + "grad_norm": 0.6796875, + "learning_rate": 0.00019308771929824563, + "loss": 5.2441, + "step": 2491 + }, + { + "epoch": 0.1311578947368421, + "grad_norm": 0.66796875, + "learning_rate": 0.00019307602339181288, + "loss": 5.139, + "step": 2492 + }, + { + "epoch": 0.13121052631578947, + "grad_norm": 0.77734375, + "learning_rate": 0.00019306432748538013, + "loss": 4.8562, + "step": 2493 + }, + { + "epoch": 0.13126315789473683, + "grad_norm": 0.71484375, + "learning_rate": 0.00019305263157894738, + "loss": 4.96, + "step": 2494 + }, + { + "epoch": 0.13131578947368422, + "grad_norm": 0.72265625, + "learning_rate": 0.00019304093567251463, + "loss": 4.9731, + "step": 2495 + }, + { + "epoch": 0.13136842105263158, + "grad_norm": 1.046875, + "learning_rate": 0.00019302923976608188, + "loss": 4.7407, + "step": 2496 + }, + { + "epoch": 0.13142105263157894, + "grad_norm": 0.62890625, + "learning_rate": 0.00019301754385964915, + "loss": 4.777, + "step": 2497 + }, + { + "epoch": 0.13147368421052633, + "grad_norm": 0.625, + "learning_rate": 0.00019300584795321638, + "loss": 4.9406, + "step": 2498 + }, + { + "epoch": 0.13152631578947369, + "grad_norm": 0.59765625, + "learning_rate": 0.00019299415204678363, + "loss": 5.0825, + "step": 2499 + }, + { + "epoch": 0.13157894736842105, + "grad_norm": 0.8828125, + "learning_rate": 0.00019298245614035088, + "loss": 5.1799, + "step": 2500 + }, + { + "epoch": 0.13163157894736843, + "grad_norm": 0.6796875, + "learning_rate": 0.00019297076023391813, + "loss": 4.9929, + "step": 2501 + }, + { + "epoch": 0.1316842105263158, + "grad_norm": 0.85546875, + "learning_rate": 0.00019295906432748538, + "loss": 4.9771, + "step": 2502 + }, + { + "epoch": 0.13173684210526315, + "grad_norm": 0.63671875, + "learning_rate": 0.00019294736842105263, + "loss": 4.8971, + "step": 2503 + }, + { + "epoch": 0.13178947368421054, + "grad_norm": 0.72265625, + "learning_rate": 0.0001929356725146199, + "loss": 5.236, + "step": 2504 + }, + { + "epoch": 0.1318421052631579, + "grad_norm": 0.90234375, + "learning_rate": 0.00019292397660818715, + "loss": 4.8495, + "step": 2505 + }, + { + "epoch": 0.13189473684210526, + "grad_norm": 0.5703125, + "learning_rate": 0.0001929122807017544, + "loss": 5.1403, + "step": 2506 + }, + { + "epoch": 0.13194736842105262, + "grad_norm": 0.71875, + "learning_rate": 0.00019290058479532165, + "loss": 5.0338, + "step": 2507 + }, + { + "epoch": 0.132, + "grad_norm": 0.703125, + "learning_rate": 0.0001928888888888889, + "loss": 5.2184, + "step": 2508 + }, + { + "epoch": 0.13205263157894737, + "grad_norm": 0.859375, + "learning_rate": 0.00019287719298245615, + "loss": 4.953, + "step": 2509 + }, + { + "epoch": 0.13210526315789473, + "grad_norm": 0.984375, + "learning_rate": 0.0001928654970760234, + "loss": 5.1267, + "step": 2510 + }, + { + "epoch": 0.1321578947368421, + "grad_norm": 0.62109375, + "learning_rate": 0.00019285380116959065, + "loss": 5.1551, + "step": 2511 + }, + { + "epoch": 0.13221052631578947, + "grad_norm": 0.61328125, + "learning_rate": 0.00019284210526315793, + "loss": 4.8941, + "step": 2512 + }, + { + "epoch": 0.13226315789473683, + "grad_norm": 0.55859375, + "learning_rate": 0.00019283040935672515, + "loss": 5.0138, + "step": 2513 + }, + { + "epoch": 0.13231578947368422, + "grad_norm": 0.640625, + "learning_rate": 0.0001928187134502924, + "loss": 5.0445, + "step": 2514 + }, + { + "epoch": 0.13236842105263158, + "grad_norm": 0.58203125, + "learning_rate": 0.00019280701754385965, + "loss": 5.2949, + "step": 2515 + }, + { + "epoch": 0.13242105263157894, + "grad_norm": 0.57421875, + "learning_rate": 0.0001927953216374269, + "loss": 4.7601, + "step": 2516 + }, + { + "epoch": 0.13247368421052633, + "grad_norm": 0.609375, + "learning_rate": 0.00019278362573099415, + "loss": 4.9896, + "step": 2517 + }, + { + "epoch": 0.1325263157894737, + "grad_norm": 0.66796875, + "learning_rate": 0.0001927719298245614, + "loss": 4.8732, + "step": 2518 + }, + { + "epoch": 0.13257894736842105, + "grad_norm": 0.7109375, + "learning_rate": 0.00019276023391812867, + "loss": 5.1138, + "step": 2519 + }, + { + "epoch": 0.13263157894736843, + "grad_norm": 0.72265625, + "learning_rate": 0.00019274853801169592, + "loss": 4.7042, + "step": 2520 + }, + { + "epoch": 0.1326842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.00019273684210526317, + "loss": 5.2786, + "step": 2521 + }, + { + "epoch": 0.13273684210526315, + "grad_norm": 0.7578125, + "learning_rate": 0.00019272514619883042, + "loss": 4.8441, + "step": 2522 + }, + { + "epoch": 0.1327894736842105, + "grad_norm": 0.91015625, + "learning_rate": 0.00019271345029239767, + "loss": 5.0817, + "step": 2523 + }, + { + "epoch": 0.1328421052631579, + "grad_norm": 0.74609375, + "learning_rate": 0.00019270175438596492, + "loss": 4.7367, + "step": 2524 + }, + { + "epoch": 0.13289473684210526, + "grad_norm": 0.89453125, + "learning_rate": 0.00019269005847953217, + "loss": 4.5537, + "step": 2525 + }, + { + "epoch": 0.13294736842105262, + "grad_norm": 0.59375, + "learning_rate": 0.00019267836257309942, + "loss": 4.8975, + "step": 2526 + }, + { + "epoch": 0.133, + "grad_norm": 0.7421875, + "learning_rate": 0.0001926666666666667, + "loss": 5.0918, + "step": 2527 + }, + { + "epoch": 0.13305263157894737, + "grad_norm": 0.63671875, + "learning_rate": 0.00019265497076023392, + "loss": 4.8143, + "step": 2528 + }, + { + "epoch": 0.13310526315789473, + "grad_norm": 0.8359375, + "learning_rate": 0.00019264327485380117, + "loss": 4.9272, + "step": 2529 + }, + { + "epoch": 0.13315789473684211, + "grad_norm": 0.92578125, + "learning_rate": 0.00019263157894736842, + "loss": 4.972, + "step": 2530 + }, + { + "epoch": 0.13321052631578947, + "grad_norm": 0.88671875, + "learning_rate": 0.00019261988304093567, + "loss": 4.9683, + "step": 2531 + }, + { + "epoch": 0.13326315789473683, + "grad_norm": 0.82421875, + "learning_rate": 0.00019260818713450292, + "loss": 4.8268, + "step": 2532 + }, + { + "epoch": 0.13331578947368422, + "grad_norm": 0.609375, + "learning_rate": 0.00019259649122807017, + "loss": 4.9799, + "step": 2533 + }, + { + "epoch": 0.13336842105263158, + "grad_norm": 0.6953125, + "learning_rate": 0.00019258479532163744, + "loss": 5.045, + "step": 2534 + }, + { + "epoch": 0.13342105263157894, + "grad_norm": 0.96484375, + "learning_rate": 0.0001925730994152047, + "loss": 4.9397, + "step": 2535 + }, + { + "epoch": 0.13347368421052633, + "grad_norm": 0.6328125, + "learning_rate": 0.00019256140350877194, + "loss": 5.0084, + "step": 2536 + }, + { + "epoch": 0.1335263157894737, + "grad_norm": 0.68359375, + "learning_rate": 0.0001925497076023392, + "loss": 4.9159, + "step": 2537 + }, + { + "epoch": 0.13357894736842105, + "grad_norm": 0.671875, + "learning_rate": 0.00019253801169590644, + "loss": 5.1703, + "step": 2538 + }, + { + "epoch": 0.13363157894736843, + "grad_norm": 0.8203125, + "learning_rate": 0.0001925263157894737, + "loss": 5.0475, + "step": 2539 + }, + { + "epoch": 0.1336842105263158, + "grad_norm": 0.69140625, + "learning_rate": 0.00019251461988304094, + "loss": 4.8679, + "step": 2540 + }, + { + "epoch": 0.13373684210526315, + "grad_norm": 0.6640625, + "learning_rate": 0.00019250292397660822, + "loss": 5.1631, + "step": 2541 + }, + { + "epoch": 0.1337894736842105, + "grad_norm": 0.72265625, + "learning_rate": 0.00019249122807017547, + "loss": 5.1649, + "step": 2542 + }, + { + "epoch": 0.1338421052631579, + "grad_norm": 0.65625, + "learning_rate": 0.00019247953216374272, + "loss": 4.8684, + "step": 2543 + }, + { + "epoch": 0.13389473684210526, + "grad_norm": 0.6875, + "learning_rate": 0.00019246783625730994, + "loss": 5.3473, + "step": 2544 + }, + { + "epoch": 0.13394736842105262, + "grad_norm": 0.62890625, + "learning_rate": 0.0001924561403508772, + "loss": 4.8891, + "step": 2545 + }, + { + "epoch": 0.134, + "grad_norm": 0.6875, + "learning_rate": 0.00019244444444444444, + "loss": 4.551, + "step": 2546 + }, + { + "epoch": 0.13405263157894737, + "grad_norm": 0.68359375, + "learning_rate": 0.0001924327485380117, + "loss": 4.8636, + "step": 2547 + }, + { + "epoch": 0.13410526315789473, + "grad_norm": 0.515625, + "learning_rate": 0.00019242105263157894, + "loss": 5.2359, + "step": 2548 + }, + { + "epoch": 0.13415789473684211, + "grad_norm": 0.62109375, + "learning_rate": 0.00019240935672514621, + "loss": 4.8043, + "step": 2549 + }, + { + "epoch": 0.13421052631578947, + "grad_norm": 0.57421875, + "learning_rate": 0.00019239766081871346, + "loss": 4.9737, + "step": 2550 + }, + { + "epoch": 0.13426315789473683, + "grad_norm": 0.67578125, + "learning_rate": 0.00019238596491228071, + "loss": 4.702, + "step": 2551 + }, + { + "epoch": 0.13431578947368422, + "grad_norm": 0.625, + "learning_rate": 0.00019237426900584796, + "loss": 4.5455, + "step": 2552 + }, + { + "epoch": 0.13436842105263158, + "grad_norm": 0.6484375, + "learning_rate": 0.0001923625730994152, + "loss": 5.1601, + "step": 2553 + }, + { + "epoch": 0.13442105263157894, + "grad_norm": 0.61328125, + "learning_rate": 0.00019235087719298246, + "loss": 5.2422, + "step": 2554 + }, + { + "epoch": 0.13447368421052633, + "grad_norm": 0.6328125, + "learning_rate": 0.0001923391812865497, + "loss": 4.7276, + "step": 2555 + }, + { + "epoch": 0.1345263157894737, + "grad_norm": 0.52734375, + "learning_rate": 0.000192327485380117, + "loss": 5.1078, + "step": 2556 + }, + { + "epoch": 0.13457894736842105, + "grad_norm": 0.6328125, + "learning_rate": 0.00019231578947368424, + "loss": 4.7071, + "step": 2557 + }, + { + "epoch": 0.1346315789473684, + "grad_norm": 0.76171875, + "learning_rate": 0.0001923040935672515, + "loss": 5.0036, + "step": 2558 + }, + { + "epoch": 0.1346842105263158, + "grad_norm": 0.73828125, + "learning_rate": 0.0001922923976608187, + "loss": 5.0446, + "step": 2559 + }, + { + "epoch": 0.13473684210526315, + "grad_norm": 0.65625, + "learning_rate": 0.00019228070175438596, + "loss": 4.5899, + "step": 2560 + }, + { + "epoch": 0.13478947368421051, + "grad_norm": 0.64453125, + "learning_rate": 0.0001922690058479532, + "loss": 4.7638, + "step": 2561 + }, + { + "epoch": 0.1348421052631579, + "grad_norm": 0.66015625, + "learning_rate": 0.00019225730994152046, + "loss": 4.9315, + "step": 2562 + }, + { + "epoch": 0.13489473684210526, + "grad_norm": 0.6875, + "learning_rate": 0.0001922456140350877, + "loss": 4.8292, + "step": 2563 + }, + { + "epoch": 0.13494736842105262, + "grad_norm": 0.92578125, + "learning_rate": 0.00019223391812865499, + "loss": 4.728, + "step": 2564 + }, + { + "epoch": 0.135, + "grad_norm": 0.6328125, + "learning_rate": 0.00019222222222222224, + "loss": 5.199, + "step": 2565 + }, + { + "epoch": 0.13505263157894737, + "grad_norm": 0.6953125, + "learning_rate": 0.00019221052631578949, + "loss": 5.2383, + "step": 2566 + }, + { + "epoch": 0.13510526315789473, + "grad_norm": 0.73046875, + "learning_rate": 0.00019219883040935673, + "loss": 5.1242, + "step": 2567 + }, + { + "epoch": 0.13515789473684212, + "grad_norm": 0.59765625, + "learning_rate": 0.00019218713450292398, + "loss": 4.8991, + "step": 2568 + }, + { + "epoch": 0.13521052631578948, + "grad_norm": 0.640625, + "learning_rate": 0.00019217543859649123, + "loss": 4.6708, + "step": 2569 + }, + { + "epoch": 0.13526315789473684, + "grad_norm": 0.55078125, + "learning_rate": 0.00019216374269005848, + "loss": 5.2744, + "step": 2570 + }, + { + "epoch": 0.13531578947368422, + "grad_norm": 0.76953125, + "learning_rate": 0.00019215204678362576, + "loss": 4.6058, + "step": 2571 + }, + { + "epoch": 0.13536842105263158, + "grad_norm": 0.6640625, + "learning_rate": 0.000192140350877193, + "loss": 4.7221, + "step": 2572 + }, + { + "epoch": 0.13542105263157894, + "grad_norm": 0.71484375, + "learning_rate": 0.00019212865497076026, + "loss": 5.0032, + "step": 2573 + }, + { + "epoch": 0.13547368421052633, + "grad_norm": 0.87109375, + "learning_rate": 0.00019211695906432748, + "loss": 4.8963, + "step": 2574 + }, + { + "epoch": 0.1355263157894737, + "grad_norm": 0.59765625, + "learning_rate": 0.00019210526315789473, + "loss": 4.7288, + "step": 2575 + }, + { + "epoch": 0.13557894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.00019209356725146198, + "loss": 5.0213, + "step": 2576 + }, + { + "epoch": 0.1356315789473684, + "grad_norm": 0.7734375, + "learning_rate": 0.00019208187134502923, + "loss": 5.1376, + "step": 2577 + }, + { + "epoch": 0.1356842105263158, + "grad_norm": 0.66796875, + "learning_rate": 0.0001920701754385965, + "loss": 4.697, + "step": 2578 + }, + { + "epoch": 0.13573684210526316, + "grad_norm": 0.921875, + "learning_rate": 0.00019205847953216376, + "loss": 4.5023, + "step": 2579 + }, + { + "epoch": 0.13578947368421052, + "grad_norm": 0.62109375, + "learning_rate": 0.000192046783625731, + "loss": 4.6405, + "step": 2580 + }, + { + "epoch": 0.1358421052631579, + "grad_norm": 0.66015625, + "learning_rate": 0.00019203508771929826, + "loss": 4.6753, + "step": 2581 + }, + { + "epoch": 0.13589473684210526, + "grad_norm": 0.69140625, + "learning_rate": 0.0001920233918128655, + "loss": 4.5603, + "step": 2582 + }, + { + "epoch": 0.13594736842105262, + "grad_norm": 0.7109375, + "learning_rate": 0.00019201169590643276, + "loss": 4.8514, + "step": 2583 + }, + { + "epoch": 0.136, + "grad_norm": 0.62890625, + "learning_rate": 0.000192, + "loss": 5.1876, + "step": 2584 + }, + { + "epoch": 0.13605263157894737, + "grad_norm": 1.0546875, + "learning_rate": 0.00019198830409356725, + "loss": 4.6501, + "step": 2585 + }, + { + "epoch": 0.13610526315789473, + "grad_norm": 0.94921875, + "learning_rate": 0.00019197660818713453, + "loss": 4.7952, + "step": 2586 + }, + { + "epoch": 0.13615789473684212, + "grad_norm": 0.57421875, + "learning_rate": 0.00019196491228070178, + "loss": 5.034, + "step": 2587 + }, + { + "epoch": 0.13621052631578948, + "grad_norm": 0.62109375, + "learning_rate": 0.00019195321637426903, + "loss": 4.7803, + "step": 2588 + }, + { + "epoch": 0.13626315789473684, + "grad_norm": 0.67578125, + "learning_rate": 0.00019194152046783628, + "loss": 4.9898, + "step": 2589 + }, + { + "epoch": 0.13631578947368422, + "grad_norm": 0.70703125, + "learning_rate": 0.0001919298245614035, + "loss": 5.0944, + "step": 2590 + }, + { + "epoch": 0.13636842105263158, + "grad_norm": 2.046875, + "learning_rate": 0.00019191812865497075, + "loss": 4.9325, + "step": 2591 + }, + { + "epoch": 0.13642105263157894, + "grad_norm": 0.69921875, + "learning_rate": 0.000191906432748538, + "loss": 4.9435, + "step": 2592 + }, + { + "epoch": 0.1364736842105263, + "grad_norm": 0.78515625, + "learning_rate": 0.00019189473684210528, + "loss": 4.9345, + "step": 2593 + }, + { + "epoch": 0.1365263157894737, + "grad_norm": 0.74609375, + "learning_rate": 0.00019188304093567253, + "loss": 4.9033, + "step": 2594 + }, + { + "epoch": 0.13657894736842105, + "grad_norm": 0.625, + "learning_rate": 0.00019187134502923978, + "loss": 4.7257, + "step": 2595 + }, + { + "epoch": 0.1366315789473684, + "grad_norm": 0.796875, + "learning_rate": 0.00019185964912280703, + "loss": 4.9478, + "step": 2596 + }, + { + "epoch": 0.1366842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.00019184795321637428, + "loss": 5.2966, + "step": 2597 + }, + { + "epoch": 0.13673684210526316, + "grad_norm": 0.65625, + "learning_rate": 0.00019183625730994153, + "loss": 5.1954, + "step": 2598 + }, + { + "epoch": 0.13678947368421052, + "grad_norm": 0.80859375, + "learning_rate": 0.00019182456140350878, + "loss": 5.0349, + "step": 2599 + }, + { + "epoch": 0.1368421052631579, + "grad_norm": 0.80078125, + "learning_rate": 0.00019181286549707603, + "loss": 4.7078, + "step": 2600 + }, + { + "epoch": 0.13689473684210526, + "grad_norm": 0.7109375, + "learning_rate": 0.0001918011695906433, + "loss": 5.0814, + "step": 2601 + }, + { + "epoch": 0.13694736842105262, + "grad_norm": 0.83203125, + "learning_rate": 0.00019178947368421055, + "loss": 4.7688, + "step": 2602 + }, + { + "epoch": 0.137, + "grad_norm": 0.6640625, + "learning_rate": 0.0001917777777777778, + "loss": 5.0977, + "step": 2603 + }, + { + "epoch": 0.13705263157894737, + "grad_norm": 0.66796875, + "learning_rate": 0.00019176608187134505, + "loss": 4.6394, + "step": 2604 + }, + { + "epoch": 0.13710526315789473, + "grad_norm": 0.67578125, + "learning_rate": 0.00019175438596491227, + "loss": 4.9538, + "step": 2605 + }, + { + "epoch": 0.13715789473684212, + "grad_norm": 0.796875, + "learning_rate": 0.00019174269005847952, + "loss": 4.4711, + "step": 2606 + }, + { + "epoch": 0.13721052631578948, + "grad_norm": 0.57421875, + "learning_rate": 0.00019173099415204677, + "loss": 4.616, + "step": 2607 + }, + { + "epoch": 0.13726315789473684, + "grad_norm": 0.60546875, + "learning_rate": 0.00019171929824561405, + "loss": 4.5763, + "step": 2608 + }, + { + "epoch": 0.13731578947368422, + "grad_norm": 0.55078125, + "learning_rate": 0.0001917076023391813, + "loss": 4.9942, + "step": 2609 + }, + { + "epoch": 0.13736842105263158, + "grad_norm": 0.58203125, + "learning_rate": 0.00019169590643274855, + "loss": 5.1545, + "step": 2610 + }, + { + "epoch": 0.13742105263157894, + "grad_norm": 0.71484375, + "learning_rate": 0.0001916842105263158, + "loss": 4.73, + "step": 2611 + }, + { + "epoch": 0.1374736842105263, + "grad_norm": 0.62890625, + "learning_rate": 0.00019167251461988305, + "loss": 4.8655, + "step": 2612 + }, + { + "epoch": 0.1375263157894737, + "grad_norm": 0.578125, + "learning_rate": 0.0001916608187134503, + "loss": 5.1121, + "step": 2613 + }, + { + "epoch": 0.13757894736842105, + "grad_norm": 0.6328125, + "learning_rate": 0.00019164912280701755, + "loss": 5.1277, + "step": 2614 + }, + { + "epoch": 0.1376315789473684, + "grad_norm": 0.97265625, + "learning_rate": 0.00019163742690058482, + "loss": 4.8368, + "step": 2615 + }, + { + "epoch": 0.1376842105263158, + "grad_norm": 0.61328125, + "learning_rate": 0.00019162573099415207, + "loss": 4.9854, + "step": 2616 + }, + { + "epoch": 0.13773684210526316, + "grad_norm": 0.6640625, + "learning_rate": 0.00019161403508771932, + "loss": 4.9371, + "step": 2617 + }, + { + "epoch": 0.13778947368421052, + "grad_norm": 0.58984375, + "learning_rate": 0.00019160233918128657, + "loss": 4.8627, + "step": 2618 + }, + { + "epoch": 0.1378421052631579, + "grad_norm": 0.8046875, + "learning_rate": 0.00019159064327485382, + "loss": 4.7382, + "step": 2619 + }, + { + "epoch": 0.13789473684210526, + "grad_norm": 0.83203125, + "learning_rate": 0.00019157894736842104, + "loss": 5.194, + "step": 2620 + }, + { + "epoch": 0.13794736842105262, + "grad_norm": 0.71484375, + "learning_rate": 0.0001915672514619883, + "loss": 4.9549, + "step": 2621 + }, + { + "epoch": 0.138, + "grad_norm": 1.1875, + "learning_rate": 0.00019155555555555554, + "loss": 4.8825, + "step": 2622 + }, + { + "epoch": 0.13805263157894737, + "grad_norm": 0.64453125, + "learning_rate": 0.00019154385964912282, + "loss": 4.9828, + "step": 2623 + }, + { + "epoch": 0.13810526315789473, + "grad_norm": 0.62890625, + "learning_rate": 0.00019153216374269007, + "loss": 5.0092, + "step": 2624 + }, + { + "epoch": 0.13815789473684212, + "grad_norm": 0.96484375, + "learning_rate": 0.00019152046783625732, + "loss": 4.531, + "step": 2625 + }, + { + "epoch": 0.13821052631578948, + "grad_norm": 0.5625, + "learning_rate": 0.00019150877192982457, + "loss": 5.3532, + "step": 2626 + }, + { + "epoch": 0.13826315789473684, + "grad_norm": 0.66015625, + "learning_rate": 0.00019149707602339182, + "loss": 5.0915, + "step": 2627 + }, + { + "epoch": 0.1383157894736842, + "grad_norm": 0.61328125, + "learning_rate": 0.00019148538011695907, + "loss": 5.0554, + "step": 2628 + }, + { + "epoch": 0.13836842105263158, + "grad_norm": 0.62109375, + "learning_rate": 0.00019147368421052632, + "loss": 4.925, + "step": 2629 + }, + { + "epoch": 0.13842105263157894, + "grad_norm": 0.66015625, + "learning_rate": 0.0001914619883040936, + "loss": 4.6852, + "step": 2630 + }, + { + "epoch": 0.1384736842105263, + "grad_norm": 0.734375, + "learning_rate": 0.00019145029239766084, + "loss": 4.7733, + "step": 2631 + }, + { + "epoch": 0.1385263157894737, + "grad_norm": 0.625, + "learning_rate": 0.0001914385964912281, + "loss": 4.9247, + "step": 2632 + }, + { + "epoch": 0.13857894736842105, + "grad_norm": 0.80078125, + "learning_rate": 0.00019142690058479534, + "loss": 5.0461, + "step": 2633 + }, + { + "epoch": 0.1386315789473684, + "grad_norm": 0.78125, + "learning_rate": 0.0001914152046783626, + "loss": 5.1327, + "step": 2634 + }, + { + "epoch": 0.1386842105263158, + "grad_norm": 0.8125, + "learning_rate": 0.00019140350877192984, + "loss": 4.6851, + "step": 2635 + }, + { + "epoch": 0.13873684210526316, + "grad_norm": 0.79296875, + "learning_rate": 0.00019139181286549706, + "loss": 5.0176, + "step": 2636 + }, + { + "epoch": 0.13878947368421052, + "grad_norm": 0.73046875, + "learning_rate": 0.00019138011695906431, + "loss": 5.0724, + "step": 2637 + }, + { + "epoch": 0.1388421052631579, + "grad_norm": 0.7109375, + "learning_rate": 0.0001913684210526316, + "loss": 5.2817, + "step": 2638 + }, + { + "epoch": 0.13889473684210527, + "grad_norm": 0.66015625, + "learning_rate": 0.00019135672514619884, + "loss": 4.9744, + "step": 2639 + }, + { + "epoch": 0.13894736842105262, + "grad_norm": 0.7265625, + "learning_rate": 0.0001913450292397661, + "loss": 4.9935, + "step": 2640 + }, + { + "epoch": 0.139, + "grad_norm": 0.93359375, + "learning_rate": 0.00019133333333333334, + "loss": 5.1011, + "step": 2641 + }, + { + "epoch": 0.13905263157894737, + "grad_norm": 0.62109375, + "learning_rate": 0.0001913216374269006, + "loss": 4.8985, + "step": 2642 + }, + { + "epoch": 0.13910526315789473, + "grad_norm": 0.578125, + "learning_rate": 0.00019130994152046784, + "loss": 4.6485, + "step": 2643 + }, + { + "epoch": 0.1391578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.0001912982456140351, + "loss": 4.8397, + "step": 2644 + }, + { + "epoch": 0.13921052631578948, + "grad_norm": 0.65234375, + "learning_rate": 0.00019128654970760237, + "loss": 4.4137, + "step": 2645 + }, + { + "epoch": 0.13926315789473684, + "grad_norm": 0.5703125, + "learning_rate": 0.00019127485380116962, + "loss": 5.1858, + "step": 2646 + }, + { + "epoch": 0.1393157894736842, + "grad_norm": 0.64453125, + "learning_rate": 0.00019126315789473686, + "loss": 5.1326, + "step": 2647 + }, + { + "epoch": 0.13936842105263159, + "grad_norm": 0.95703125, + "learning_rate": 0.00019125146198830411, + "loss": 4.9173, + "step": 2648 + }, + { + "epoch": 0.13942105263157895, + "grad_norm": 0.625, + "learning_rate": 0.00019123976608187136, + "loss": 4.8462, + "step": 2649 + }, + { + "epoch": 0.1394736842105263, + "grad_norm": 0.76171875, + "learning_rate": 0.0001912280701754386, + "loss": 4.8983, + "step": 2650 + }, + { + "epoch": 0.1395263157894737, + "grad_norm": 0.56640625, + "learning_rate": 0.00019121637426900584, + "loss": 4.9019, + "step": 2651 + }, + { + "epoch": 0.13957894736842105, + "grad_norm": 0.7421875, + "learning_rate": 0.00019120467836257309, + "loss": 5.0044, + "step": 2652 + }, + { + "epoch": 0.1396315789473684, + "grad_norm": 0.640625, + "learning_rate": 0.00019119298245614036, + "loss": 4.9704, + "step": 2653 + }, + { + "epoch": 0.1396842105263158, + "grad_norm": 0.6796875, + "learning_rate": 0.0001911812865497076, + "loss": 4.9, + "step": 2654 + }, + { + "epoch": 0.13973684210526316, + "grad_norm": 0.671875, + "learning_rate": 0.00019116959064327486, + "loss": 4.7629, + "step": 2655 + }, + { + "epoch": 0.13978947368421052, + "grad_norm": 0.62890625, + "learning_rate": 0.0001911578947368421, + "loss": 4.7342, + "step": 2656 + }, + { + "epoch": 0.1398421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.00019114619883040936, + "loss": 5.0389, + "step": 2657 + }, + { + "epoch": 0.13989473684210527, + "grad_norm": 1.96875, + "learning_rate": 0.0001911345029239766, + "loss": 5.3783, + "step": 2658 + }, + { + "epoch": 0.13994736842105263, + "grad_norm": 0.70703125, + "learning_rate": 0.00019112280701754386, + "loss": 5.1447, + "step": 2659 + }, + { + "epoch": 0.14, + "grad_norm": 0.54296875, + "learning_rate": 0.00019111111111111114, + "loss": 4.9925, + "step": 2660 + }, + { + "epoch": 0.14005263157894737, + "grad_norm": 0.69140625, + "learning_rate": 0.00019109941520467839, + "loss": 4.766, + "step": 2661 + }, + { + "epoch": 0.14010526315789473, + "grad_norm": 0.6015625, + "learning_rate": 0.00019108771929824564, + "loss": 5.2329, + "step": 2662 + }, + { + "epoch": 0.1401578947368421, + "grad_norm": 0.72265625, + "learning_rate": 0.00019107602339181289, + "loss": 5.1068, + "step": 2663 + }, + { + "epoch": 0.14021052631578948, + "grad_norm": 0.6640625, + "learning_rate": 0.00019106432748538013, + "loss": 4.7243, + "step": 2664 + }, + { + "epoch": 0.14026315789473684, + "grad_norm": 0.58203125, + "learning_rate": 0.00019105263157894738, + "loss": 5.1442, + "step": 2665 + }, + { + "epoch": 0.1403157894736842, + "grad_norm": 0.7109375, + "learning_rate": 0.0001910409356725146, + "loss": 4.7585, + "step": 2666 + }, + { + "epoch": 0.1403684210526316, + "grad_norm": 0.69140625, + "learning_rate": 0.00019102923976608188, + "loss": 4.8595, + "step": 2667 + }, + { + "epoch": 0.14042105263157895, + "grad_norm": 0.6328125, + "learning_rate": 0.00019101754385964913, + "loss": 4.5239, + "step": 2668 + }, + { + "epoch": 0.1404736842105263, + "grad_norm": 0.609375, + "learning_rate": 0.00019100584795321638, + "loss": 4.7818, + "step": 2669 + }, + { + "epoch": 0.1405263157894737, + "grad_norm": 0.61328125, + "learning_rate": 0.00019099415204678363, + "loss": 4.6702, + "step": 2670 + }, + { + "epoch": 0.14057894736842105, + "grad_norm": 0.640625, + "learning_rate": 0.00019098245614035088, + "loss": 4.8748, + "step": 2671 + }, + { + "epoch": 0.1406315789473684, + "grad_norm": 0.78515625, + "learning_rate": 0.00019097076023391813, + "loss": 5.0491, + "step": 2672 + }, + { + "epoch": 0.1406842105263158, + "grad_norm": 0.74609375, + "learning_rate": 0.00019095906432748538, + "loss": 5.0105, + "step": 2673 + }, + { + "epoch": 0.14073684210526316, + "grad_norm": 0.88671875, + "learning_rate": 0.00019094736842105263, + "loss": 4.5673, + "step": 2674 + }, + { + "epoch": 0.14078947368421052, + "grad_norm": 0.5546875, + "learning_rate": 0.0001909356725146199, + "loss": 5.0075, + "step": 2675 + }, + { + "epoch": 0.1408421052631579, + "grad_norm": 0.62109375, + "learning_rate": 0.00019092397660818716, + "loss": 4.3642, + "step": 2676 + }, + { + "epoch": 0.14089473684210527, + "grad_norm": 0.6953125, + "learning_rate": 0.0001909122807017544, + "loss": 4.4272, + "step": 2677 + }, + { + "epoch": 0.14094736842105263, + "grad_norm": 0.73828125, + "learning_rate": 0.00019090058479532166, + "loss": 4.7146, + "step": 2678 + }, + { + "epoch": 0.141, + "grad_norm": 0.8515625, + "learning_rate": 0.0001908888888888889, + "loss": 4.7356, + "step": 2679 + }, + { + "epoch": 0.14105263157894737, + "grad_norm": 0.76953125, + "learning_rate": 0.00019087719298245616, + "loss": 4.7069, + "step": 2680 + }, + { + "epoch": 0.14110526315789473, + "grad_norm": 0.58984375, + "learning_rate": 0.0001908654970760234, + "loss": 4.8569, + "step": 2681 + }, + { + "epoch": 0.1411578947368421, + "grad_norm": 0.80859375, + "learning_rate": 0.00019085380116959065, + "loss": 5.1344, + "step": 2682 + }, + { + "epoch": 0.14121052631578948, + "grad_norm": 0.8359375, + "learning_rate": 0.0001908421052631579, + "loss": 5.019, + "step": 2683 + }, + { + "epoch": 0.14126315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.00019083040935672515, + "loss": 4.9857, + "step": 2684 + }, + { + "epoch": 0.1413157894736842, + "grad_norm": 0.71484375, + "learning_rate": 0.0001908187134502924, + "loss": 4.7413, + "step": 2685 + }, + { + "epoch": 0.1413684210526316, + "grad_norm": 0.609375, + "learning_rate": 0.00019080701754385965, + "loss": 4.8928, + "step": 2686 + }, + { + "epoch": 0.14142105263157895, + "grad_norm": 0.6484375, + "learning_rate": 0.0001907953216374269, + "loss": 4.9885, + "step": 2687 + }, + { + "epoch": 0.1414736842105263, + "grad_norm": 0.88671875, + "learning_rate": 0.00019078362573099415, + "loss": 4.8307, + "step": 2688 + }, + { + "epoch": 0.1415263157894737, + "grad_norm": 0.87890625, + "learning_rate": 0.0001907719298245614, + "loss": 4.7182, + "step": 2689 + }, + { + "epoch": 0.14157894736842105, + "grad_norm": 0.71484375, + "learning_rate": 0.00019076023391812868, + "loss": 4.8866, + "step": 2690 + }, + { + "epoch": 0.14163157894736841, + "grad_norm": 0.734375, + "learning_rate": 0.00019074853801169593, + "loss": 5.016, + "step": 2691 + }, + { + "epoch": 0.1416842105263158, + "grad_norm": 0.63671875, + "learning_rate": 0.00019073684210526318, + "loss": 4.5994, + "step": 2692 + }, + { + "epoch": 0.14173684210526316, + "grad_norm": 0.66015625, + "learning_rate": 0.00019072514619883043, + "loss": 4.5386, + "step": 2693 + }, + { + "epoch": 0.14178947368421052, + "grad_norm": 0.640625, + "learning_rate": 0.00019071345029239768, + "loss": 4.596, + "step": 2694 + }, + { + "epoch": 0.1418421052631579, + "grad_norm": 0.6796875, + "learning_rate": 0.00019070175438596493, + "loss": 4.8268, + "step": 2695 + }, + { + "epoch": 0.14189473684210527, + "grad_norm": 0.64453125, + "learning_rate": 0.00019069005847953218, + "loss": 4.9089, + "step": 2696 + }, + { + "epoch": 0.14194736842105263, + "grad_norm": 0.703125, + "learning_rate": 0.00019067836257309943, + "loss": 5.1627, + "step": 2697 + }, + { + "epoch": 0.142, + "grad_norm": 0.76171875, + "learning_rate": 0.00019066666666666668, + "loss": 5.1438, + "step": 2698 + }, + { + "epoch": 0.14205263157894737, + "grad_norm": 0.6953125, + "learning_rate": 0.00019065497076023392, + "loss": 4.8054, + "step": 2699 + }, + { + "epoch": 0.14210526315789473, + "grad_norm": 0.6796875, + "learning_rate": 0.00019064327485380117, + "loss": 4.78, + "step": 2700 + }, + { + "epoch": 0.1421578947368421, + "grad_norm": 0.66015625, + "learning_rate": 0.00019063157894736842, + "loss": 4.5964, + "step": 2701 + }, + { + "epoch": 0.14221052631578948, + "grad_norm": 0.56640625, + "learning_rate": 0.00019061988304093567, + "loss": 4.8921, + "step": 2702 + }, + { + "epoch": 0.14226315789473684, + "grad_norm": 0.61328125, + "learning_rate": 0.00019060818713450292, + "loss": 4.7922, + "step": 2703 + }, + { + "epoch": 0.1423157894736842, + "grad_norm": 0.68359375, + "learning_rate": 0.0001905964912280702, + "loss": 4.6226, + "step": 2704 + }, + { + "epoch": 0.1423684210526316, + "grad_norm": 0.55078125, + "learning_rate": 0.00019058479532163745, + "loss": 5.1992, + "step": 2705 + }, + { + "epoch": 0.14242105263157895, + "grad_norm": 0.67578125, + "learning_rate": 0.0001905730994152047, + "loss": 4.6499, + "step": 2706 + }, + { + "epoch": 0.1424736842105263, + "grad_norm": 0.5859375, + "learning_rate": 0.00019056140350877195, + "loss": 4.9778, + "step": 2707 + }, + { + "epoch": 0.1425263157894737, + "grad_norm": 0.640625, + "learning_rate": 0.0001905497076023392, + "loss": 4.5286, + "step": 2708 + }, + { + "epoch": 0.14257894736842106, + "grad_norm": 0.5625, + "learning_rate": 0.00019053801169590645, + "loss": 4.7374, + "step": 2709 + }, + { + "epoch": 0.14263157894736841, + "grad_norm": 0.5703125, + "learning_rate": 0.0001905263157894737, + "loss": 4.9348, + "step": 2710 + }, + { + "epoch": 0.1426842105263158, + "grad_norm": 0.7421875, + "learning_rate": 0.00019051461988304095, + "loss": 5.2328, + "step": 2711 + }, + { + "epoch": 0.14273684210526316, + "grad_norm": 0.8359375, + "learning_rate": 0.0001905029239766082, + "loss": 4.3049, + "step": 2712 + }, + { + "epoch": 0.14278947368421052, + "grad_norm": 0.65234375, + "learning_rate": 0.00019049122807017545, + "loss": 4.7064, + "step": 2713 + }, + { + "epoch": 0.14284210526315788, + "grad_norm": 0.59375, + "learning_rate": 0.0001904795321637427, + "loss": 4.9541, + "step": 2714 + }, + { + "epoch": 0.14289473684210527, + "grad_norm": 0.6796875, + "learning_rate": 0.00019046783625730995, + "loss": 4.9353, + "step": 2715 + }, + { + "epoch": 0.14294736842105263, + "grad_norm": 0.6484375, + "learning_rate": 0.0001904561403508772, + "loss": 4.9396, + "step": 2716 + }, + { + "epoch": 0.143, + "grad_norm": 0.6953125, + "learning_rate": 0.00019044444444444444, + "loss": 4.6429, + "step": 2717 + }, + { + "epoch": 0.14305263157894738, + "grad_norm": 0.63671875, + "learning_rate": 0.0001904327485380117, + "loss": 4.8877, + "step": 2718 + }, + { + "epoch": 0.14310526315789474, + "grad_norm": 0.70703125, + "learning_rate": 0.00019042105263157897, + "loss": 4.8906, + "step": 2719 + }, + { + "epoch": 0.1431578947368421, + "grad_norm": 0.74609375, + "learning_rate": 0.00019040935672514622, + "loss": 4.8609, + "step": 2720 + }, + { + "epoch": 0.14321052631578948, + "grad_norm": 0.671875, + "learning_rate": 0.00019039766081871347, + "loss": 4.8706, + "step": 2721 + }, + { + "epoch": 0.14326315789473684, + "grad_norm": 0.6171875, + "learning_rate": 0.00019038596491228072, + "loss": 4.6693, + "step": 2722 + }, + { + "epoch": 0.1433157894736842, + "grad_norm": 0.59375, + "learning_rate": 0.00019037426900584797, + "loss": 4.7187, + "step": 2723 + }, + { + "epoch": 0.1433684210526316, + "grad_norm": 0.59765625, + "learning_rate": 0.00019036257309941522, + "loss": 4.7606, + "step": 2724 + }, + { + "epoch": 0.14342105263157895, + "grad_norm": 0.61328125, + "learning_rate": 0.00019035087719298247, + "loss": 4.9651, + "step": 2725 + }, + { + "epoch": 0.1434736842105263, + "grad_norm": 0.59765625, + "learning_rate": 0.00019033918128654972, + "loss": 4.6924, + "step": 2726 + }, + { + "epoch": 0.1435263157894737, + "grad_norm": 0.6875, + "learning_rate": 0.00019032748538011697, + "loss": 4.6496, + "step": 2727 + }, + { + "epoch": 0.14357894736842106, + "grad_norm": 0.73046875, + "learning_rate": 0.00019031578947368422, + "loss": 4.991, + "step": 2728 + }, + { + "epoch": 0.14363157894736842, + "grad_norm": 0.625, + "learning_rate": 0.00019030409356725147, + "loss": 4.9135, + "step": 2729 + }, + { + "epoch": 0.1436842105263158, + "grad_norm": 0.859375, + "learning_rate": 0.00019029239766081872, + "loss": 4.8957, + "step": 2730 + }, + { + "epoch": 0.14373684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00019028070175438597, + "loss": 4.8315, + "step": 2731 + }, + { + "epoch": 0.14378947368421052, + "grad_norm": 0.625, + "learning_rate": 0.00019026900584795322, + "loss": 4.6955, + "step": 2732 + }, + { + "epoch": 0.14384210526315788, + "grad_norm": 0.5390625, + "learning_rate": 0.00019025730994152047, + "loss": 4.6713, + "step": 2733 + }, + { + "epoch": 0.14389473684210527, + "grad_norm": 0.60546875, + "learning_rate": 0.00019024561403508774, + "loss": 4.6923, + "step": 2734 + }, + { + "epoch": 0.14394736842105263, + "grad_norm": 0.5625, + "learning_rate": 0.000190233918128655, + "loss": 4.6419, + "step": 2735 + }, + { + "epoch": 0.144, + "grad_norm": 0.609375, + "learning_rate": 0.00019022222222222224, + "loss": 4.772, + "step": 2736 + }, + { + "epoch": 0.14405263157894738, + "grad_norm": 0.5859375, + "learning_rate": 0.0001902105263157895, + "loss": 4.6985, + "step": 2737 + }, + { + "epoch": 0.14410526315789474, + "grad_norm": 0.71875, + "learning_rate": 0.00019019883040935674, + "loss": 4.6458, + "step": 2738 + }, + { + "epoch": 0.1441578947368421, + "grad_norm": 0.64453125, + "learning_rate": 0.000190187134502924, + "loss": 4.9054, + "step": 2739 + }, + { + "epoch": 0.14421052631578948, + "grad_norm": 0.60546875, + "learning_rate": 0.00019017543859649124, + "loss": 4.7755, + "step": 2740 + }, + { + "epoch": 0.14426315789473684, + "grad_norm": 0.65234375, + "learning_rate": 0.0001901637426900585, + "loss": 4.8858, + "step": 2741 + }, + { + "epoch": 0.1443157894736842, + "grad_norm": 0.80078125, + "learning_rate": 0.00019015204678362574, + "loss": 4.7368, + "step": 2742 + }, + { + "epoch": 0.1443684210526316, + "grad_norm": 0.71484375, + "learning_rate": 0.000190140350877193, + "loss": 4.6043, + "step": 2743 + }, + { + "epoch": 0.14442105263157895, + "grad_norm": 0.56640625, + "learning_rate": 0.00019012865497076024, + "loss": 4.8222, + "step": 2744 + }, + { + "epoch": 0.1444736842105263, + "grad_norm": 0.65625, + "learning_rate": 0.0001901169590643275, + "loss": 4.6638, + "step": 2745 + }, + { + "epoch": 0.1445263157894737, + "grad_norm": 0.75390625, + "learning_rate": 0.00019010526315789474, + "loss": 4.6759, + "step": 2746 + }, + { + "epoch": 0.14457894736842106, + "grad_norm": 0.640625, + "learning_rate": 0.000190093567251462, + "loss": 4.6633, + "step": 2747 + }, + { + "epoch": 0.14463157894736842, + "grad_norm": 0.71484375, + "learning_rate": 0.00019008187134502924, + "loss": 4.7087, + "step": 2748 + }, + { + "epoch": 0.14468421052631578, + "grad_norm": 0.69921875, + "learning_rate": 0.0001900701754385965, + "loss": 4.7962, + "step": 2749 + }, + { + "epoch": 0.14473684210526316, + "grad_norm": 0.6328125, + "learning_rate": 0.00019005847953216376, + "loss": 4.8272, + "step": 2750 + }, + { + "epoch": 0.14478947368421052, + "grad_norm": 0.73828125, + "learning_rate": 0.000190046783625731, + "loss": 4.6728, + "step": 2751 + }, + { + "epoch": 0.14484210526315788, + "grad_norm": 0.5546875, + "learning_rate": 0.00019003508771929826, + "loss": 5.2065, + "step": 2752 + }, + { + "epoch": 0.14489473684210527, + "grad_norm": 0.65625, + "learning_rate": 0.0001900233918128655, + "loss": 4.9007, + "step": 2753 + }, + { + "epoch": 0.14494736842105263, + "grad_norm": 0.734375, + "learning_rate": 0.00019001169590643276, + "loss": 4.6036, + "step": 2754 + }, + { + "epoch": 0.145, + "grad_norm": 0.64453125, + "learning_rate": 0.00019, + "loss": 4.4339, + "step": 2755 + }, + { + "epoch": 0.14505263157894738, + "grad_norm": 0.72265625, + "learning_rate": 0.00018998830409356726, + "loss": 4.8886, + "step": 2756 + }, + { + "epoch": 0.14510526315789474, + "grad_norm": 0.6796875, + "learning_rate": 0.0001899766081871345, + "loss": 4.7075, + "step": 2757 + }, + { + "epoch": 0.1451578947368421, + "grad_norm": 0.6328125, + "learning_rate": 0.00018996491228070176, + "loss": 5.0232, + "step": 2758 + }, + { + "epoch": 0.14521052631578948, + "grad_norm": 0.75390625, + "learning_rate": 0.000189953216374269, + "loss": 4.76, + "step": 2759 + }, + { + "epoch": 0.14526315789473684, + "grad_norm": 0.58203125, + "learning_rate": 0.00018994152046783626, + "loss": 4.7758, + "step": 2760 + }, + { + "epoch": 0.1453157894736842, + "grad_norm": 0.79296875, + "learning_rate": 0.0001899298245614035, + "loss": 4.5642, + "step": 2761 + }, + { + "epoch": 0.1453684210526316, + "grad_norm": 0.6015625, + "learning_rate": 0.00018991812865497076, + "loss": 4.8467, + "step": 2762 + }, + { + "epoch": 0.14542105263157895, + "grad_norm": 0.6484375, + "learning_rate": 0.000189906432748538, + "loss": 5.0752, + "step": 2763 + }, + { + "epoch": 0.1454736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00018989473684210528, + "loss": 4.9051, + "step": 2764 + }, + { + "epoch": 0.1455263157894737, + "grad_norm": 0.62109375, + "learning_rate": 0.00018988304093567253, + "loss": 4.6158, + "step": 2765 + }, + { + "epoch": 0.14557894736842106, + "grad_norm": 0.57421875, + "learning_rate": 0.00018987134502923978, + "loss": 4.6559, + "step": 2766 + }, + { + "epoch": 0.14563157894736842, + "grad_norm": 0.59765625, + "learning_rate": 0.00018985964912280703, + "loss": 4.6818, + "step": 2767 + }, + { + "epoch": 0.14568421052631578, + "grad_norm": 0.76953125, + "learning_rate": 0.00018984795321637428, + "loss": 4.6748, + "step": 2768 + }, + { + "epoch": 0.14573684210526316, + "grad_norm": 0.6953125, + "learning_rate": 0.00018983625730994153, + "loss": 4.8977, + "step": 2769 + }, + { + "epoch": 0.14578947368421052, + "grad_norm": 0.8046875, + "learning_rate": 0.00018982456140350878, + "loss": 4.6828, + "step": 2770 + }, + { + "epoch": 0.14584210526315788, + "grad_norm": 0.703125, + "learning_rate": 0.00018981286549707603, + "loss": 4.564, + "step": 2771 + }, + { + "epoch": 0.14589473684210527, + "grad_norm": 0.69921875, + "learning_rate": 0.00018980116959064328, + "loss": 4.8538, + "step": 2772 + }, + { + "epoch": 0.14594736842105263, + "grad_norm": 0.66015625, + "learning_rate": 0.00018978947368421053, + "loss": 4.9435, + "step": 2773 + }, + { + "epoch": 0.146, + "grad_norm": 0.63671875, + "learning_rate": 0.00018977777777777778, + "loss": 5.1913, + "step": 2774 + }, + { + "epoch": 0.14605263157894738, + "grad_norm": 0.63671875, + "learning_rate": 0.00018976608187134503, + "loss": 4.7288, + "step": 2775 + }, + { + "epoch": 0.14610526315789474, + "grad_norm": 0.796875, + "learning_rate": 0.00018975438596491228, + "loss": 4.7422, + "step": 2776 + }, + { + "epoch": 0.1461578947368421, + "grad_norm": 0.62109375, + "learning_rate": 0.00018974269005847953, + "loss": 4.5232, + "step": 2777 + }, + { + "epoch": 0.14621052631578949, + "grad_norm": 0.5234375, + "learning_rate": 0.0001897309941520468, + "loss": 5.0346, + "step": 2778 + }, + { + "epoch": 0.14626315789473684, + "grad_norm": 0.7109375, + "learning_rate": 0.00018971929824561405, + "loss": 4.6652, + "step": 2779 + }, + { + "epoch": 0.1463157894736842, + "grad_norm": 0.65234375, + "learning_rate": 0.0001897076023391813, + "loss": 4.8234, + "step": 2780 + }, + { + "epoch": 0.1463684210526316, + "grad_norm": 0.70703125, + "learning_rate": 0.00018969590643274855, + "loss": 4.5951, + "step": 2781 + }, + { + "epoch": 0.14642105263157895, + "grad_norm": 0.6015625, + "learning_rate": 0.0001896842105263158, + "loss": 4.667, + "step": 2782 + }, + { + "epoch": 0.1464736842105263, + "grad_norm": 0.640625, + "learning_rate": 0.00018967251461988305, + "loss": 4.7977, + "step": 2783 + }, + { + "epoch": 0.14652631578947367, + "grad_norm": 0.6171875, + "learning_rate": 0.0001896608187134503, + "loss": 4.7126, + "step": 2784 + }, + { + "epoch": 0.14657894736842106, + "grad_norm": 0.69921875, + "learning_rate": 0.00018964912280701755, + "loss": 5.174, + "step": 2785 + }, + { + "epoch": 0.14663157894736842, + "grad_norm": 0.71875, + "learning_rate": 0.0001896374269005848, + "loss": 4.5575, + "step": 2786 + }, + { + "epoch": 0.14668421052631578, + "grad_norm": 0.625, + "learning_rate": 0.00018962573099415205, + "loss": 4.9464, + "step": 2787 + }, + { + "epoch": 0.14673684210526317, + "grad_norm": 0.71484375, + "learning_rate": 0.0001896140350877193, + "loss": 4.6569, + "step": 2788 + }, + { + "epoch": 0.14678947368421053, + "grad_norm": 0.77734375, + "learning_rate": 0.00018960233918128655, + "loss": 4.7139, + "step": 2789 + }, + { + "epoch": 0.14684210526315788, + "grad_norm": 0.73046875, + "learning_rate": 0.0001895906432748538, + "loss": 4.3826, + "step": 2790 + }, + { + "epoch": 0.14689473684210527, + "grad_norm": 1.0625, + "learning_rate": 0.00018957894736842105, + "loss": 4.8365, + "step": 2791 + }, + { + "epoch": 0.14694736842105263, + "grad_norm": 0.609375, + "learning_rate": 0.0001895672514619883, + "loss": 4.6411, + "step": 2792 + }, + { + "epoch": 0.147, + "grad_norm": 0.640625, + "learning_rate": 0.00018955555555555558, + "loss": 4.6025, + "step": 2793 + }, + { + "epoch": 0.14705263157894738, + "grad_norm": 0.69140625, + "learning_rate": 0.00018954385964912283, + "loss": 4.7755, + "step": 2794 + }, + { + "epoch": 0.14710526315789474, + "grad_norm": 0.5859375, + "learning_rate": 0.00018953216374269008, + "loss": 4.5669, + "step": 2795 + }, + { + "epoch": 0.1471578947368421, + "grad_norm": 0.65625, + "learning_rate": 0.00018952046783625733, + "loss": 4.7897, + "step": 2796 + }, + { + "epoch": 0.14721052631578949, + "grad_norm": 0.8046875, + "learning_rate": 0.00018950877192982457, + "loss": 4.7308, + "step": 2797 + }, + { + "epoch": 0.14726315789473685, + "grad_norm": 0.99609375, + "learning_rate": 0.00018949707602339182, + "loss": 4.7797, + "step": 2798 + }, + { + "epoch": 0.1473157894736842, + "grad_norm": 0.58984375, + "learning_rate": 0.00018948538011695907, + "loss": 4.6225, + "step": 2799 + }, + { + "epoch": 0.14736842105263157, + "grad_norm": 0.6328125, + "learning_rate": 0.00018947368421052632, + "loss": 4.461, + "step": 2800 + }, + { + "epoch": 0.14742105263157895, + "grad_norm": 0.640625, + "learning_rate": 0.00018946198830409357, + "loss": 4.9319, + "step": 2801 + }, + { + "epoch": 0.1474736842105263, + "grad_norm": 0.62109375, + "learning_rate": 0.00018945029239766082, + "loss": 4.893, + "step": 2802 + }, + { + "epoch": 0.14752631578947367, + "grad_norm": 0.6328125, + "learning_rate": 0.00018943859649122807, + "loss": 4.8114, + "step": 2803 + }, + { + "epoch": 0.14757894736842106, + "grad_norm": 0.77734375, + "learning_rate": 0.00018942690058479532, + "loss": 4.7662, + "step": 2804 + }, + { + "epoch": 0.14763157894736842, + "grad_norm": 0.609375, + "learning_rate": 0.00018941520467836257, + "loss": 4.8031, + "step": 2805 + }, + { + "epoch": 0.14768421052631578, + "grad_norm": 0.69921875, + "learning_rate": 0.00018940350877192982, + "loss": 4.8051, + "step": 2806 + }, + { + "epoch": 0.14773684210526317, + "grad_norm": 0.56640625, + "learning_rate": 0.00018939181286549707, + "loss": 4.5434, + "step": 2807 + }, + { + "epoch": 0.14778947368421053, + "grad_norm": 0.6015625, + "learning_rate": 0.00018938011695906435, + "loss": 4.7176, + "step": 2808 + }, + { + "epoch": 0.14784210526315789, + "grad_norm": 0.609375, + "learning_rate": 0.0001893684210526316, + "loss": 5.059, + "step": 2809 + }, + { + "epoch": 0.14789473684210527, + "grad_norm": 0.703125, + "learning_rate": 0.00018935672514619885, + "loss": 4.4975, + "step": 2810 + }, + { + "epoch": 0.14794736842105263, + "grad_norm": 0.609375, + "learning_rate": 0.0001893450292397661, + "loss": 4.7804, + "step": 2811 + }, + { + "epoch": 0.148, + "grad_norm": 0.67578125, + "learning_rate": 0.00018933333333333335, + "loss": 4.9917, + "step": 2812 + }, + { + "epoch": 0.14805263157894738, + "grad_norm": 1.5234375, + "learning_rate": 0.0001893216374269006, + "loss": 4.7518, + "step": 2813 + }, + { + "epoch": 0.14810526315789474, + "grad_norm": 0.6875, + "learning_rate": 0.00018930994152046784, + "loss": 4.666, + "step": 2814 + }, + { + "epoch": 0.1481578947368421, + "grad_norm": 0.75390625, + "learning_rate": 0.0001892982456140351, + "loss": 4.7626, + "step": 2815 + }, + { + "epoch": 0.1482105263157895, + "grad_norm": 0.63671875, + "learning_rate": 0.00018928654970760234, + "loss": 4.7035, + "step": 2816 + }, + { + "epoch": 0.14826315789473685, + "grad_norm": 0.71875, + "learning_rate": 0.0001892748538011696, + "loss": 4.8079, + "step": 2817 + }, + { + "epoch": 0.1483157894736842, + "grad_norm": 0.88671875, + "learning_rate": 0.00018926315789473684, + "loss": 4.929, + "step": 2818 + }, + { + "epoch": 0.14836842105263157, + "grad_norm": 0.72265625, + "learning_rate": 0.0001892514619883041, + "loss": 5.076, + "step": 2819 + }, + { + "epoch": 0.14842105263157895, + "grad_norm": 0.69140625, + "learning_rate": 0.00018923976608187134, + "loss": 5.1407, + "step": 2820 + }, + { + "epoch": 0.1484736842105263, + "grad_norm": 0.6640625, + "learning_rate": 0.0001892280701754386, + "loss": 4.3178, + "step": 2821 + }, + { + "epoch": 0.14852631578947367, + "grad_norm": 0.828125, + "learning_rate": 0.00018921637426900584, + "loss": 4.6816, + "step": 2822 + }, + { + "epoch": 0.14857894736842106, + "grad_norm": 0.5859375, + "learning_rate": 0.00018920467836257312, + "loss": 4.6343, + "step": 2823 + }, + { + "epoch": 0.14863157894736842, + "grad_norm": 0.60546875, + "learning_rate": 0.00018919298245614037, + "loss": 4.6136, + "step": 2824 + }, + { + "epoch": 0.14868421052631578, + "grad_norm": 0.66015625, + "learning_rate": 0.00018918128654970762, + "loss": 4.452, + "step": 2825 + }, + { + "epoch": 0.14873684210526317, + "grad_norm": 0.6640625, + "learning_rate": 0.00018916959064327487, + "loss": 4.5364, + "step": 2826 + }, + { + "epoch": 0.14878947368421053, + "grad_norm": 0.72265625, + "learning_rate": 0.00018915789473684212, + "loss": 4.915, + "step": 2827 + }, + { + "epoch": 0.1488421052631579, + "grad_norm": 0.7109375, + "learning_rate": 0.00018914619883040937, + "loss": 4.5404, + "step": 2828 + }, + { + "epoch": 0.14889473684210527, + "grad_norm": 0.75, + "learning_rate": 0.00018913450292397662, + "loss": 4.8525, + "step": 2829 + }, + { + "epoch": 0.14894736842105263, + "grad_norm": 0.703125, + "learning_rate": 0.0001891228070175439, + "loss": 4.3852, + "step": 2830 + }, + { + "epoch": 0.149, + "grad_norm": 0.7421875, + "learning_rate": 0.00018911111111111112, + "loss": 4.7649, + "step": 2831 + }, + { + "epoch": 0.14905263157894738, + "grad_norm": 0.61328125, + "learning_rate": 0.00018909941520467836, + "loss": 4.8907, + "step": 2832 + }, + { + "epoch": 0.14910526315789474, + "grad_norm": 0.68359375, + "learning_rate": 0.00018908771929824561, + "loss": 5.0815, + "step": 2833 + }, + { + "epoch": 0.1491578947368421, + "grad_norm": 0.6875, + "learning_rate": 0.00018907602339181286, + "loss": 4.9763, + "step": 2834 + }, + { + "epoch": 0.14921052631578946, + "grad_norm": 0.6171875, + "learning_rate": 0.0001890643274853801, + "loss": 4.6639, + "step": 2835 + }, + { + "epoch": 0.14926315789473685, + "grad_norm": 0.70703125, + "learning_rate": 0.00018905263157894736, + "loss": 5.0244, + "step": 2836 + }, + { + "epoch": 0.1493157894736842, + "grad_norm": 0.7109375, + "learning_rate": 0.0001890409356725146, + "loss": 4.9665, + "step": 2837 + }, + { + "epoch": 0.14936842105263157, + "grad_norm": 0.59765625, + "learning_rate": 0.0001890292397660819, + "loss": 4.9514, + "step": 2838 + }, + { + "epoch": 0.14942105263157895, + "grad_norm": 0.62890625, + "learning_rate": 0.00018901754385964914, + "loss": 4.7003, + "step": 2839 + }, + { + "epoch": 0.14947368421052631, + "grad_norm": 0.60546875, + "learning_rate": 0.0001890058479532164, + "loss": 4.8371, + "step": 2840 + }, + { + "epoch": 0.14952631578947367, + "grad_norm": 0.60546875, + "learning_rate": 0.00018899415204678364, + "loss": 4.699, + "step": 2841 + }, + { + "epoch": 0.14957894736842106, + "grad_norm": 0.6171875, + "learning_rate": 0.0001889824561403509, + "loss": 4.7449, + "step": 2842 + }, + { + "epoch": 0.14963157894736842, + "grad_norm": 0.5859375, + "learning_rate": 0.00018897076023391814, + "loss": 4.6808, + "step": 2843 + }, + { + "epoch": 0.14968421052631578, + "grad_norm": 0.640625, + "learning_rate": 0.0001889590643274854, + "loss": 4.6412, + "step": 2844 + }, + { + "epoch": 0.14973684210526317, + "grad_norm": 0.66015625, + "learning_rate": 0.00018894736842105266, + "loss": 4.3993, + "step": 2845 + }, + { + "epoch": 0.14978947368421053, + "grad_norm": 0.703125, + "learning_rate": 0.00018893567251461989, + "loss": 4.808, + "step": 2846 + }, + { + "epoch": 0.1498421052631579, + "grad_norm": 0.7421875, + "learning_rate": 0.00018892397660818714, + "loss": 4.6085, + "step": 2847 + }, + { + "epoch": 0.14989473684210528, + "grad_norm": 0.65234375, + "learning_rate": 0.00018891228070175439, + "loss": 4.8995, + "step": 2848 + }, + { + "epoch": 0.14994736842105263, + "grad_norm": 0.64453125, + "learning_rate": 0.00018890058479532163, + "loss": 5.0647, + "step": 2849 + }, + { + "epoch": 0.15, + "grad_norm": 0.7265625, + "learning_rate": 0.00018888888888888888, + "loss": 4.6795, + "step": 2850 + }, + { + "epoch": 0.15005263157894738, + "grad_norm": 0.74609375, + "learning_rate": 0.00018887719298245613, + "loss": 4.3932, + "step": 2851 + }, + { + "epoch": 0.15010526315789474, + "grad_norm": 0.83203125, + "learning_rate": 0.00018886549707602338, + "loss": 4.5728, + "step": 2852 + }, + { + "epoch": 0.1501578947368421, + "grad_norm": 0.78515625, + "learning_rate": 0.00018885380116959066, + "loss": 4.7049, + "step": 2853 + }, + { + "epoch": 0.15021052631578946, + "grad_norm": 0.80859375, + "learning_rate": 0.0001888421052631579, + "loss": 4.5987, + "step": 2854 + }, + { + "epoch": 0.15026315789473685, + "grad_norm": 0.73046875, + "learning_rate": 0.00018883040935672516, + "loss": 4.44, + "step": 2855 + }, + { + "epoch": 0.1503157894736842, + "grad_norm": 0.640625, + "learning_rate": 0.0001888187134502924, + "loss": 4.6068, + "step": 2856 + }, + { + "epoch": 0.15036842105263157, + "grad_norm": 0.58203125, + "learning_rate": 0.00018880701754385966, + "loss": 4.8788, + "step": 2857 + }, + { + "epoch": 0.15042105263157896, + "grad_norm": 0.87890625, + "learning_rate": 0.0001887953216374269, + "loss": 4.299, + "step": 2858 + }, + { + "epoch": 0.15047368421052632, + "grad_norm": 0.66015625, + "learning_rate": 0.00018878362573099416, + "loss": 4.7732, + "step": 2859 + }, + { + "epoch": 0.15052631578947367, + "grad_norm": 0.8515625, + "learning_rate": 0.00018877192982456143, + "loss": 4.3698, + "step": 2860 + }, + { + "epoch": 0.15057894736842106, + "grad_norm": 0.61328125, + "learning_rate": 0.00018876023391812866, + "loss": 4.673, + "step": 2861 + }, + { + "epoch": 0.15063157894736842, + "grad_norm": 0.83203125, + "learning_rate": 0.0001887485380116959, + "loss": 4.7992, + "step": 2862 + }, + { + "epoch": 0.15068421052631578, + "grad_norm": 0.6484375, + "learning_rate": 0.00018873684210526316, + "loss": 4.9989, + "step": 2863 + }, + { + "epoch": 0.15073684210526317, + "grad_norm": 0.578125, + "learning_rate": 0.0001887251461988304, + "loss": 4.7262, + "step": 2864 + }, + { + "epoch": 0.15078947368421053, + "grad_norm": 0.66796875, + "learning_rate": 0.00018871345029239766, + "loss": 5.035, + "step": 2865 + }, + { + "epoch": 0.1508421052631579, + "grad_norm": 0.82421875, + "learning_rate": 0.0001887017543859649, + "loss": 4.7491, + "step": 2866 + }, + { + "epoch": 0.15089473684210528, + "grad_norm": 0.734375, + "learning_rate": 0.00018869005847953218, + "loss": 4.7866, + "step": 2867 + }, + { + "epoch": 0.15094736842105264, + "grad_norm": 0.55859375, + "learning_rate": 0.00018867836257309943, + "loss": 4.7738, + "step": 2868 + }, + { + "epoch": 0.151, + "grad_norm": 0.75390625, + "learning_rate": 0.00018866666666666668, + "loss": 4.6952, + "step": 2869 + }, + { + "epoch": 0.15105263157894736, + "grad_norm": 0.79296875, + "learning_rate": 0.00018865497076023393, + "loss": 4.3179, + "step": 2870 + }, + { + "epoch": 0.15110526315789474, + "grad_norm": 0.58984375, + "learning_rate": 0.00018864327485380118, + "loss": 4.5504, + "step": 2871 + }, + { + "epoch": 0.1511578947368421, + "grad_norm": 0.6015625, + "learning_rate": 0.00018863157894736843, + "loss": 4.663, + "step": 2872 + }, + { + "epoch": 0.15121052631578946, + "grad_norm": 0.66015625, + "learning_rate": 0.00018861988304093568, + "loss": 4.7591, + "step": 2873 + }, + { + "epoch": 0.15126315789473685, + "grad_norm": 0.71484375, + "learning_rate": 0.00018860818713450293, + "loss": 4.6862, + "step": 2874 + }, + { + "epoch": 0.1513157894736842, + "grad_norm": 0.70703125, + "learning_rate": 0.0001885964912280702, + "loss": 4.7628, + "step": 2875 + }, + { + "epoch": 0.15136842105263157, + "grad_norm": 0.65234375, + "learning_rate": 0.00018858479532163746, + "loss": 4.6239, + "step": 2876 + }, + { + "epoch": 0.15142105263157896, + "grad_norm": 0.703125, + "learning_rate": 0.00018857309941520468, + "loss": 4.5312, + "step": 2877 + }, + { + "epoch": 0.15147368421052632, + "grad_norm": 0.75, + "learning_rate": 0.00018856140350877193, + "loss": 4.6754, + "step": 2878 + }, + { + "epoch": 0.15152631578947368, + "grad_norm": 0.5625, + "learning_rate": 0.00018854970760233918, + "loss": 4.6662, + "step": 2879 + }, + { + "epoch": 0.15157894736842106, + "grad_norm": 0.5703125, + "learning_rate": 0.00018853801169590643, + "loss": 4.7446, + "step": 2880 + }, + { + "epoch": 0.15163157894736842, + "grad_norm": 0.85546875, + "learning_rate": 0.00018852631578947368, + "loss": 4.4877, + "step": 2881 + }, + { + "epoch": 0.15168421052631578, + "grad_norm": 0.58203125, + "learning_rate": 0.00018851461988304095, + "loss": 4.7966, + "step": 2882 + }, + { + "epoch": 0.15173684210526317, + "grad_norm": 0.92578125, + "learning_rate": 0.0001885029239766082, + "loss": 4.7432, + "step": 2883 + }, + { + "epoch": 0.15178947368421053, + "grad_norm": 0.5859375, + "learning_rate": 0.00018849122807017545, + "loss": 4.7099, + "step": 2884 + }, + { + "epoch": 0.1518421052631579, + "grad_norm": 0.640625, + "learning_rate": 0.0001884795321637427, + "loss": 4.6458, + "step": 2885 + }, + { + "epoch": 0.15189473684210528, + "grad_norm": 0.67578125, + "learning_rate": 0.00018846783625730995, + "loss": 4.5525, + "step": 2886 + }, + { + "epoch": 0.15194736842105264, + "grad_norm": 0.69140625, + "learning_rate": 0.0001884561403508772, + "loss": 4.5754, + "step": 2887 + }, + { + "epoch": 0.152, + "grad_norm": 0.59765625, + "learning_rate": 0.00018844444444444445, + "loss": 4.5037, + "step": 2888 + }, + { + "epoch": 0.15205263157894736, + "grad_norm": 0.58203125, + "learning_rate": 0.0001884327485380117, + "loss": 4.3135, + "step": 2889 + }, + { + "epoch": 0.15210526315789474, + "grad_norm": 0.62109375, + "learning_rate": 0.00018842105263157898, + "loss": 4.6556, + "step": 2890 + }, + { + "epoch": 0.1521578947368421, + "grad_norm": 0.58203125, + "learning_rate": 0.00018840935672514623, + "loss": 4.4596, + "step": 2891 + }, + { + "epoch": 0.15221052631578946, + "grad_norm": 0.6171875, + "learning_rate": 0.00018839766081871345, + "loss": 4.3577, + "step": 2892 + }, + { + "epoch": 0.15226315789473685, + "grad_norm": 0.796875, + "learning_rate": 0.0001883859649122807, + "loss": 4.4351, + "step": 2893 + }, + { + "epoch": 0.1523157894736842, + "grad_norm": 0.78125, + "learning_rate": 0.00018837426900584795, + "loss": 4.5758, + "step": 2894 + }, + { + "epoch": 0.15236842105263157, + "grad_norm": 0.66015625, + "learning_rate": 0.0001883625730994152, + "loss": 4.3707, + "step": 2895 + }, + { + "epoch": 0.15242105263157896, + "grad_norm": 0.6875, + "learning_rate": 0.00018835087719298245, + "loss": 4.7608, + "step": 2896 + }, + { + "epoch": 0.15247368421052632, + "grad_norm": 0.6640625, + "learning_rate": 0.00018833918128654972, + "loss": 4.995, + "step": 2897 + }, + { + "epoch": 0.15252631578947368, + "grad_norm": 0.69921875, + "learning_rate": 0.00018832748538011697, + "loss": 4.6923, + "step": 2898 + }, + { + "epoch": 0.15257894736842106, + "grad_norm": 0.74609375, + "learning_rate": 0.00018831578947368422, + "loss": 4.4172, + "step": 2899 + }, + { + "epoch": 0.15263157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00018830409356725147, + "loss": 5.0632, + "step": 2900 + }, + { + "epoch": 0.15268421052631578, + "grad_norm": 0.73046875, + "learning_rate": 0.00018829239766081872, + "loss": 4.879, + "step": 2901 + }, + { + "epoch": 0.15273684210526317, + "grad_norm": 0.62109375, + "learning_rate": 0.00018828070175438597, + "loss": 5.0834, + "step": 2902 + }, + { + "epoch": 0.15278947368421053, + "grad_norm": 1.03125, + "learning_rate": 0.00018826900584795322, + "loss": 4.6693, + "step": 2903 + }, + { + "epoch": 0.1528421052631579, + "grad_norm": 0.7578125, + "learning_rate": 0.0001882573099415205, + "loss": 4.2586, + "step": 2904 + }, + { + "epoch": 0.15289473684210525, + "grad_norm": 0.6171875, + "learning_rate": 0.00018824561403508775, + "loss": 4.7998, + "step": 2905 + }, + { + "epoch": 0.15294736842105264, + "grad_norm": 0.6171875, + "learning_rate": 0.000188233918128655, + "loss": 4.5737, + "step": 2906 + }, + { + "epoch": 0.153, + "grad_norm": 0.56640625, + "learning_rate": 0.00018822222222222222, + "loss": 4.5124, + "step": 2907 + }, + { + "epoch": 0.15305263157894736, + "grad_norm": 0.65625, + "learning_rate": 0.00018821052631578947, + "loss": 4.4331, + "step": 2908 + }, + { + "epoch": 0.15310526315789474, + "grad_norm": 0.59765625, + "learning_rate": 0.00018819883040935672, + "loss": 4.6327, + "step": 2909 + }, + { + "epoch": 0.1531578947368421, + "grad_norm": 0.5546875, + "learning_rate": 0.00018818713450292397, + "loss": 4.8903, + "step": 2910 + }, + { + "epoch": 0.15321052631578946, + "grad_norm": 0.69140625, + "learning_rate": 0.00018817543859649122, + "loss": 4.4827, + "step": 2911 + }, + { + "epoch": 0.15326315789473685, + "grad_norm": 0.52734375, + "learning_rate": 0.0001881637426900585, + "loss": 4.7011, + "step": 2912 + }, + { + "epoch": 0.1533157894736842, + "grad_norm": 0.7265625, + "learning_rate": 0.00018815204678362574, + "loss": 4.6144, + "step": 2913 + }, + { + "epoch": 0.15336842105263157, + "grad_norm": 0.66796875, + "learning_rate": 0.000188140350877193, + "loss": 4.9508, + "step": 2914 + }, + { + "epoch": 0.15342105263157896, + "grad_norm": 0.69140625, + "learning_rate": 0.00018812865497076024, + "loss": 4.6697, + "step": 2915 + }, + { + "epoch": 0.15347368421052632, + "grad_norm": 0.70703125, + "learning_rate": 0.0001881169590643275, + "loss": 4.8981, + "step": 2916 + }, + { + "epoch": 0.15352631578947368, + "grad_norm": 0.69921875, + "learning_rate": 0.00018810526315789474, + "loss": 4.649, + "step": 2917 + }, + { + "epoch": 0.15357894736842106, + "grad_norm": 0.64453125, + "learning_rate": 0.000188093567251462, + "loss": 4.8556, + "step": 2918 + }, + { + "epoch": 0.15363157894736842, + "grad_norm": 0.73046875, + "learning_rate": 0.00018808187134502927, + "loss": 4.6079, + "step": 2919 + }, + { + "epoch": 0.15368421052631578, + "grad_norm": 0.828125, + "learning_rate": 0.00018807017543859652, + "loss": 4.6484, + "step": 2920 + }, + { + "epoch": 0.15373684210526317, + "grad_norm": 0.6953125, + "learning_rate": 0.00018805847953216377, + "loss": 4.7064, + "step": 2921 + }, + { + "epoch": 0.15378947368421053, + "grad_norm": 1.0390625, + "learning_rate": 0.00018804678362573102, + "loss": 4.5138, + "step": 2922 + }, + { + "epoch": 0.1538421052631579, + "grad_norm": 0.59375, + "learning_rate": 0.00018803508771929824, + "loss": 4.7977, + "step": 2923 + }, + { + "epoch": 0.15389473684210525, + "grad_norm": 0.55859375, + "learning_rate": 0.0001880233918128655, + "loss": 4.8821, + "step": 2924 + }, + { + "epoch": 0.15394736842105264, + "grad_norm": 0.625, + "learning_rate": 0.00018801169590643274, + "loss": 4.6709, + "step": 2925 + }, + { + "epoch": 0.154, + "grad_norm": 0.6015625, + "learning_rate": 0.000188, + "loss": 4.4098, + "step": 2926 + }, + { + "epoch": 0.15405263157894736, + "grad_norm": 0.6953125, + "learning_rate": 0.00018798830409356727, + "loss": 4.3328, + "step": 2927 + }, + { + "epoch": 0.15410526315789475, + "grad_norm": 0.57421875, + "learning_rate": 0.00018797660818713452, + "loss": 4.4745, + "step": 2928 + }, + { + "epoch": 0.1541578947368421, + "grad_norm": 0.6015625, + "learning_rate": 0.00018796491228070176, + "loss": 4.7573, + "step": 2929 + }, + { + "epoch": 0.15421052631578946, + "grad_norm": 0.66796875, + "learning_rate": 0.00018795321637426901, + "loss": 4.5, + "step": 2930 + }, + { + "epoch": 0.15426315789473685, + "grad_norm": 0.77734375, + "learning_rate": 0.00018794152046783626, + "loss": 4.389, + "step": 2931 + }, + { + "epoch": 0.1543157894736842, + "grad_norm": 0.8671875, + "learning_rate": 0.00018792982456140351, + "loss": 4.8041, + "step": 2932 + }, + { + "epoch": 0.15436842105263157, + "grad_norm": 0.67578125, + "learning_rate": 0.00018791812865497076, + "loss": 4.6951, + "step": 2933 + }, + { + "epoch": 0.15442105263157896, + "grad_norm": 0.64453125, + "learning_rate": 0.00018790643274853804, + "loss": 5.0467, + "step": 2934 + }, + { + "epoch": 0.15447368421052632, + "grad_norm": 0.796875, + "learning_rate": 0.0001878947368421053, + "loss": 4.4283, + "step": 2935 + }, + { + "epoch": 0.15452631578947368, + "grad_norm": 0.72265625, + "learning_rate": 0.00018788304093567254, + "loss": 5.0127, + "step": 2936 + }, + { + "epoch": 0.15457894736842107, + "grad_norm": 0.87890625, + "learning_rate": 0.0001878713450292398, + "loss": 4.5532, + "step": 2937 + }, + { + "epoch": 0.15463157894736843, + "grad_norm": 0.671875, + "learning_rate": 0.000187859649122807, + "loss": 4.9008, + "step": 2938 + }, + { + "epoch": 0.15468421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00018784795321637426, + "loss": 4.6853, + "step": 2939 + }, + { + "epoch": 0.15473684210526314, + "grad_norm": 0.55078125, + "learning_rate": 0.0001878362573099415, + "loss": 4.8225, + "step": 2940 + }, + { + "epoch": 0.15478947368421053, + "grad_norm": 0.66015625, + "learning_rate": 0.0001878245614035088, + "loss": 4.707, + "step": 2941 + }, + { + "epoch": 0.1548421052631579, + "grad_norm": 0.6640625, + "learning_rate": 0.00018781286549707604, + "loss": 4.2723, + "step": 2942 + }, + { + "epoch": 0.15489473684210525, + "grad_norm": 0.74609375, + "learning_rate": 0.00018780116959064329, + "loss": 5.0401, + "step": 2943 + }, + { + "epoch": 0.15494736842105264, + "grad_norm": 0.6015625, + "learning_rate": 0.00018778947368421054, + "loss": 4.47, + "step": 2944 + }, + { + "epoch": 0.155, + "grad_norm": 0.62890625, + "learning_rate": 0.00018777777777777779, + "loss": 4.5377, + "step": 2945 + }, + { + "epoch": 0.15505263157894736, + "grad_norm": 0.66796875, + "learning_rate": 0.00018776608187134504, + "loss": 4.7295, + "step": 2946 + }, + { + "epoch": 0.15510526315789475, + "grad_norm": 0.82421875, + "learning_rate": 0.00018775438596491228, + "loss": 4.507, + "step": 2947 + }, + { + "epoch": 0.1551578947368421, + "grad_norm": 1.6171875, + "learning_rate": 0.00018774269005847953, + "loss": 4.5021, + "step": 2948 + }, + { + "epoch": 0.15521052631578947, + "grad_norm": 0.5234375, + "learning_rate": 0.0001877309941520468, + "loss": 4.5229, + "step": 2949 + }, + { + "epoch": 0.15526315789473685, + "grad_norm": 0.5859375, + "learning_rate": 0.00018771929824561406, + "loss": 4.5241, + "step": 2950 + }, + { + "epoch": 0.1553157894736842, + "grad_norm": 0.5703125, + "learning_rate": 0.0001877076023391813, + "loss": 4.6547, + "step": 2951 + }, + { + "epoch": 0.15536842105263157, + "grad_norm": 0.61328125, + "learning_rate": 0.00018769590643274856, + "loss": 4.9633, + "step": 2952 + }, + { + "epoch": 0.15542105263157896, + "grad_norm": 0.6953125, + "learning_rate": 0.00018768421052631578, + "loss": 4.3546, + "step": 2953 + }, + { + "epoch": 0.15547368421052632, + "grad_norm": 0.60546875, + "learning_rate": 0.00018767251461988303, + "loss": 4.2713, + "step": 2954 + }, + { + "epoch": 0.15552631578947368, + "grad_norm": 0.61328125, + "learning_rate": 0.00018766081871345028, + "loss": 4.664, + "step": 2955 + }, + { + "epoch": 0.15557894736842104, + "grad_norm": 0.7109375, + "learning_rate": 0.00018764912280701756, + "loss": 4.3035, + "step": 2956 + }, + { + "epoch": 0.15563157894736843, + "grad_norm": 0.765625, + "learning_rate": 0.0001876374269005848, + "loss": 4.5346, + "step": 2957 + }, + { + "epoch": 0.15568421052631579, + "grad_norm": 0.609375, + "learning_rate": 0.00018762573099415206, + "loss": 4.7949, + "step": 2958 + }, + { + "epoch": 0.15573684210526315, + "grad_norm": 0.60546875, + "learning_rate": 0.0001876140350877193, + "loss": 4.6771, + "step": 2959 + }, + { + "epoch": 0.15578947368421053, + "grad_norm": 0.6328125, + "learning_rate": 0.00018760233918128656, + "loss": 4.4951, + "step": 2960 + }, + { + "epoch": 0.1558421052631579, + "grad_norm": 0.58203125, + "learning_rate": 0.0001875906432748538, + "loss": 4.8393, + "step": 2961 + }, + { + "epoch": 0.15589473684210525, + "grad_norm": 0.5546875, + "learning_rate": 0.00018757894736842106, + "loss": 4.602, + "step": 2962 + }, + { + "epoch": 0.15594736842105264, + "grad_norm": 0.59375, + "learning_rate": 0.0001875672514619883, + "loss": 4.7798, + "step": 2963 + }, + { + "epoch": 0.156, + "grad_norm": 0.671875, + "learning_rate": 0.00018755555555555558, + "loss": 4.4372, + "step": 2964 + }, + { + "epoch": 0.15605263157894736, + "grad_norm": 0.59765625, + "learning_rate": 0.00018754385964912283, + "loss": 4.7312, + "step": 2965 + }, + { + "epoch": 0.15610526315789475, + "grad_norm": 0.7734375, + "learning_rate": 0.00018753216374269008, + "loss": 4.6247, + "step": 2966 + }, + { + "epoch": 0.1561578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.00018752046783625733, + "loss": 4.9146, + "step": 2967 + }, + { + "epoch": 0.15621052631578947, + "grad_norm": 0.66015625, + "learning_rate": 0.00018750877192982458, + "loss": 5.0524, + "step": 2968 + }, + { + "epoch": 0.15626315789473685, + "grad_norm": 0.640625, + "learning_rate": 0.0001874970760233918, + "loss": 4.5635, + "step": 2969 + }, + { + "epoch": 0.1563157894736842, + "grad_norm": 0.66015625, + "learning_rate": 0.00018748538011695905, + "loss": 4.7236, + "step": 2970 + }, + { + "epoch": 0.15636842105263157, + "grad_norm": 0.609375, + "learning_rate": 0.00018747368421052633, + "loss": 4.4996, + "step": 2971 + }, + { + "epoch": 0.15642105263157896, + "grad_norm": 0.640625, + "learning_rate": 0.00018746198830409358, + "loss": 4.7106, + "step": 2972 + }, + { + "epoch": 0.15647368421052632, + "grad_norm": 0.66796875, + "learning_rate": 0.00018745029239766083, + "loss": 4.4818, + "step": 2973 + }, + { + "epoch": 0.15652631578947368, + "grad_norm": 0.58984375, + "learning_rate": 0.00018743859649122808, + "loss": 4.7841, + "step": 2974 + }, + { + "epoch": 0.15657894736842104, + "grad_norm": 0.6640625, + "learning_rate": 0.00018742690058479533, + "loss": 4.7314, + "step": 2975 + }, + { + "epoch": 0.15663157894736843, + "grad_norm": 0.74609375, + "learning_rate": 0.00018741520467836258, + "loss": 4.5212, + "step": 2976 + }, + { + "epoch": 0.1566842105263158, + "grad_norm": 0.7265625, + "learning_rate": 0.00018740350877192983, + "loss": 4.8124, + "step": 2977 + }, + { + "epoch": 0.15673684210526315, + "grad_norm": 0.671875, + "learning_rate": 0.00018739181286549708, + "loss": 4.4781, + "step": 2978 + }, + { + "epoch": 0.15678947368421053, + "grad_norm": 0.6328125, + "learning_rate": 0.00018738011695906435, + "loss": 4.6379, + "step": 2979 + }, + { + "epoch": 0.1568421052631579, + "grad_norm": 0.59375, + "learning_rate": 0.0001873684210526316, + "loss": 4.7982, + "step": 2980 + }, + { + "epoch": 0.15689473684210525, + "grad_norm": 0.765625, + "learning_rate": 0.00018735672514619885, + "loss": 4.4027, + "step": 2981 + }, + { + "epoch": 0.15694736842105264, + "grad_norm": 0.55078125, + "learning_rate": 0.0001873450292397661, + "loss": 4.6635, + "step": 2982 + }, + { + "epoch": 0.157, + "grad_norm": 0.61328125, + "learning_rate": 0.00018733333333333335, + "loss": 4.9101, + "step": 2983 + }, + { + "epoch": 0.15705263157894736, + "grad_norm": 0.515625, + "learning_rate": 0.00018732163742690057, + "loss": 4.8215, + "step": 2984 + }, + { + "epoch": 0.15710526315789475, + "grad_norm": 0.67578125, + "learning_rate": 0.00018730994152046782, + "loss": 4.7306, + "step": 2985 + }, + { + "epoch": 0.1571578947368421, + "grad_norm": 0.66796875, + "learning_rate": 0.0001872982456140351, + "loss": 4.5771, + "step": 2986 + }, + { + "epoch": 0.15721052631578947, + "grad_norm": 0.65234375, + "learning_rate": 0.00018728654970760235, + "loss": 4.6755, + "step": 2987 + }, + { + "epoch": 0.15726315789473685, + "grad_norm": 0.7265625, + "learning_rate": 0.0001872748538011696, + "loss": 4.383, + "step": 2988 + }, + { + "epoch": 0.15731578947368421, + "grad_norm": 0.75, + "learning_rate": 0.00018726315789473685, + "loss": 4.7561, + "step": 2989 + }, + { + "epoch": 0.15736842105263157, + "grad_norm": 0.6328125, + "learning_rate": 0.0001872514619883041, + "loss": 4.5141, + "step": 2990 + }, + { + "epoch": 0.15742105263157893, + "grad_norm": 0.6171875, + "learning_rate": 0.00018723976608187135, + "loss": 4.4329, + "step": 2991 + }, + { + "epoch": 0.15747368421052632, + "grad_norm": 0.58984375, + "learning_rate": 0.0001872280701754386, + "loss": 4.6408, + "step": 2992 + }, + { + "epoch": 0.15752631578947368, + "grad_norm": 0.56640625, + "learning_rate": 0.00018721637426900587, + "loss": 4.742, + "step": 2993 + }, + { + "epoch": 0.15757894736842104, + "grad_norm": 0.6484375, + "learning_rate": 0.00018720467836257312, + "loss": 4.3389, + "step": 2994 + }, + { + "epoch": 0.15763157894736843, + "grad_norm": 0.5625, + "learning_rate": 0.00018719298245614037, + "loss": 4.6757, + "step": 2995 + }, + { + "epoch": 0.1576842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.00018718128654970762, + "loss": 4.9437, + "step": 2996 + }, + { + "epoch": 0.15773684210526315, + "grad_norm": 0.80859375, + "learning_rate": 0.00018716959064327487, + "loss": 4.682, + "step": 2997 + }, + { + "epoch": 0.15778947368421053, + "grad_norm": 0.53125, + "learning_rate": 0.00018715789473684212, + "loss": 4.3534, + "step": 2998 + }, + { + "epoch": 0.1578421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00018714619883040937, + "loss": 4.803, + "step": 2999 + }, + { + "epoch": 0.15789473684210525, + "grad_norm": 0.61328125, + "learning_rate": 0.0001871345029239766, + "loss": 4.4249, + "step": 3000 + }, + { + "epoch": 0.15794736842105264, + "grad_norm": 0.80859375, + "learning_rate": 0.00018712280701754387, + "loss": 4.5907, + "step": 3001 + }, + { + "epoch": 0.158, + "grad_norm": 0.72265625, + "learning_rate": 0.00018711111111111112, + "loss": 4.5512, + "step": 3002 + }, + { + "epoch": 0.15805263157894736, + "grad_norm": 0.80078125, + "learning_rate": 0.00018709941520467837, + "loss": 4.6287, + "step": 3003 + }, + { + "epoch": 0.15810526315789475, + "grad_norm": 0.79296875, + "learning_rate": 0.00018708771929824562, + "loss": 4.6929, + "step": 3004 + }, + { + "epoch": 0.1581578947368421, + "grad_norm": 0.640625, + "learning_rate": 0.00018707602339181287, + "loss": 4.3069, + "step": 3005 + }, + { + "epoch": 0.15821052631578947, + "grad_norm": 0.57421875, + "learning_rate": 0.00018706432748538012, + "loss": 4.6152, + "step": 3006 + }, + { + "epoch": 0.15826315789473686, + "grad_norm": 0.671875, + "learning_rate": 0.00018705263157894737, + "loss": 4.2959, + "step": 3007 + }, + { + "epoch": 0.15831578947368422, + "grad_norm": 0.6171875, + "learning_rate": 0.00018704093567251465, + "loss": 4.578, + "step": 3008 + }, + { + "epoch": 0.15836842105263157, + "grad_norm": 0.7578125, + "learning_rate": 0.0001870292397660819, + "loss": 4.4104, + "step": 3009 + }, + { + "epoch": 0.15842105263157893, + "grad_norm": 0.65234375, + "learning_rate": 0.00018701754385964914, + "loss": 4.3336, + "step": 3010 + }, + { + "epoch": 0.15847368421052632, + "grad_norm": 0.671875, + "learning_rate": 0.0001870058479532164, + "loss": 4.3281, + "step": 3011 + }, + { + "epoch": 0.15852631578947368, + "grad_norm": 0.70703125, + "learning_rate": 0.00018699415204678364, + "loss": 4.9683, + "step": 3012 + }, + { + "epoch": 0.15857894736842104, + "grad_norm": 0.984375, + "learning_rate": 0.0001869824561403509, + "loss": 4.6813, + "step": 3013 + }, + { + "epoch": 0.15863157894736843, + "grad_norm": 0.6875, + "learning_rate": 0.00018697076023391814, + "loss": 4.7058, + "step": 3014 + }, + { + "epoch": 0.1586842105263158, + "grad_norm": 0.64453125, + "learning_rate": 0.00018695906432748537, + "loss": 4.5558, + "step": 3015 + }, + { + "epoch": 0.15873684210526315, + "grad_norm": 0.625, + "learning_rate": 0.00018694736842105264, + "loss": 4.6386, + "step": 3016 + }, + { + "epoch": 0.15878947368421054, + "grad_norm": 0.67578125, + "learning_rate": 0.0001869356725146199, + "loss": 4.8854, + "step": 3017 + }, + { + "epoch": 0.1588421052631579, + "grad_norm": 0.68359375, + "learning_rate": 0.00018692397660818714, + "loss": 4.9262, + "step": 3018 + }, + { + "epoch": 0.15889473684210526, + "grad_norm": 0.58203125, + "learning_rate": 0.0001869122807017544, + "loss": 4.3815, + "step": 3019 + }, + { + "epoch": 0.15894736842105264, + "grad_norm": 0.59765625, + "learning_rate": 0.00018690058479532164, + "loss": 4.6009, + "step": 3020 + }, + { + "epoch": 0.159, + "grad_norm": 0.55859375, + "learning_rate": 0.0001868888888888889, + "loss": 4.5561, + "step": 3021 + }, + { + "epoch": 0.15905263157894736, + "grad_norm": 0.58984375, + "learning_rate": 0.00018687719298245614, + "loss": 4.5355, + "step": 3022 + }, + { + "epoch": 0.15910526315789475, + "grad_norm": 0.57421875, + "learning_rate": 0.00018686549707602342, + "loss": 4.5707, + "step": 3023 + }, + { + "epoch": 0.1591578947368421, + "grad_norm": 0.640625, + "learning_rate": 0.00018685380116959067, + "loss": 4.7807, + "step": 3024 + }, + { + "epoch": 0.15921052631578947, + "grad_norm": 0.578125, + "learning_rate": 0.00018684210526315792, + "loss": 4.473, + "step": 3025 + }, + { + "epoch": 0.15926315789473683, + "grad_norm": 0.7265625, + "learning_rate": 0.00018683040935672517, + "loss": 4.3273, + "step": 3026 + }, + { + "epoch": 0.15931578947368422, + "grad_norm": 0.55859375, + "learning_rate": 0.00018681871345029241, + "loss": 4.4485, + "step": 3027 + }, + { + "epoch": 0.15936842105263158, + "grad_norm": 0.609375, + "learning_rate": 0.00018680701754385966, + "loss": 4.5302, + "step": 3028 + }, + { + "epoch": 0.15942105263157894, + "grad_norm": 0.67578125, + "learning_rate": 0.00018679532163742691, + "loss": 4.526, + "step": 3029 + }, + { + "epoch": 0.15947368421052632, + "grad_norm": 0.58984375, + "learning_rate": 0.00018678362573099416, + "loss": 4.7894, + "step": 3030 + }, + { + "epoch": 0.15952631578947368, + "grad_norm": 0.66796875, + "learning_rate": 0.0001867719298245614, + "loss": 4.782, + "step": 3031 + }, + { + "epoch": 0.15957894736842104, + "grad_norm": 0.66015625, + "learning_rate": 0.00018676023391812866, + "loss": 4.4145, + "step": 3032 + }, + { + "epoch": 0.15963157894736843, + "grad_norm": 0.63671875, + "learning_rate": 0.0001867485380116959, + "loss": 4.3977, + "step": 3033 + }, + { + "epoch": 0.1596842105263158, + "grad_norm": 0.625, + "learning_rate": 0.00018673684210526316, + "loss": 4.5475, + "step": 3034 + }, + { + "epoch": 0.15973684210526315, + "grad_norm": 0.609375, + "learning_rate": 0.0001867251461988304, + "loss": 4.3925, + "step": 3035 + }, + { + "epoch": 0.15978947368421054, + "grad_norm": 0.71484375, + "learning_rate": 0.00018671345029239766, + "loss": 4.1424, + "step": 3036 + }, + { + "epoch": 0.1598421052631579, + "grad_norm": 0.66796875, + "learning_rate": 0.0001867017543859649, + "loss": 4.4489, + "step": 3037 + }, + { + "epoch": 0.15989473684210526, + "grad_norm": 0.70703125, + "learning_rate": 0.0001866900584795322, + "loss": 4.6603, + "step": 3038 + }, + { + "epoch": 0.15994736842105264, + "grad_norm": 0.6875, + "learning_rate": 0.00018667836257309944, + "loss": 4.2788, + "step": 3039 + }, + { + "epoch": 0.16, + "grad_norm": 0.74609375, + "learning_rate": 0.0001866666666666667, + "loss": 4.5781, + "step": 3040 + }, + { + "epoch": 0.16005263157894736, + "grad_norm": 0.625, + "learning_rate": 0.00018665497076023394, + "loss": 4.6206, + "step": 3041 + }, + { + "epoch": 0.16010526315789475, + "grad_norm": 0.703125, + "learning_rate": 0.00018664327485380119, + "loss": 4.2304, + "step": 3042 + }, + { + "epoch": 0.1601578947368421, + "grad_norm": 0.609375, + "learning_rate": 0.00018663157894736844, + "loss": 4.5359, + "step": 3043 + }, + { + "epoch": 0.16021052631578947, + "grad_norm": 0.5546875, + "learning_rate": 0.00018661988304093568, + "loss": 4.6475, + "step": 3044 + }, + { + "epoch": 0.16026315789473683, + "grad_norm": 0.6953125, + "learning_rate": 0.00018660818713450293, + "loss": 4.3728, + "step": 3045 + }, + { + "epoch": 0.16031578947368422, + "grad_norm": 0.5859375, + "learning_rate": 0.00018659649122807018, + "loss": 4.6176, + "step": 3046 + }, + { + "epoch": 0.16036842105263158, + "grad_norm": 0.640625, + "learning_rate": 0.00018658479532163743, + "loss": 4.3651, + "step": 3047 + }, + { + "epoch": 0.16042105263157894, + "grad_norm": 0.61328125, + "learning_rate": 0.00018657309941520468, + "loss": 5.077, + "step": 3048 + }, + { + "epoch": 0.16047368421052632, + "grad_norm": 1.3359375, + "learning_rate": 0.00018656140350877193, + "loss": 4.4948, + "step": 3049 + }, + { + "epoch": 0.16052631578947368, + "grad_norm": 1.9921875, + "learning_rate": 0.00018654970760233918, + "loss": 4.4311, + "step": 3050 + }, + { + "epoch": 0.16057894736842104, + "grad_norm": 0.5390625, + "learning_rate": 0.00018653801169590643, + "loss": 4.5295, + "step": 3051 + }, + { + "epoch": 0.16063157894736843, + "grad_norm": 0.62109375, + "learning_rate": 0.00018652631578947368, + "loss": 4.5114, + "step": 3052 + }, + { + "epoch": 0.1606842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00018651461988304096, + "loss": 4.5866, + "step": 3053 + }, + { + "epoch": 0.16073684210526315, + "grad_norm": 0.55859375, + "learning_rate": 0.0001865029239766082, + "loss": 5.0396, + "step": 3054 + }, + { + "epoch": 0.16078947368421054, + "grad_norm": 0.62109375, + "learning_rate": 0.00018649122807017546, + "loss": 4.3396, + "step": 3055 + }, + { + "epoch": 0.1608421052631579, + "grad_norm": 0.7890625, + "learning_rate": 0.0001864795321637427, + "loss": 4.1598, + "step": 3056 + }, + { + "epoch": 0.16089473684210526, + "grad_norm": 0.609375, + "learning_rate": 0.00018646783625730996, + "loss": 4.5176, + "step": 3057 + }, + { + "epoch": 0.16094736842105264, + "grad_norm": 0.640625, + "learning_rate": 0.0001864561403508772, + "loss": 4.7742, + "step": 3058 + }, + { + "epoch": 0.161, + "grad_norm": 0.640625, + "learning_rate": 0.00018644444444444446, + "loss": 4.4831, + "step": 3059 + }, + { + "epoch": 0.16105263157894736, + "grad_norm": 0.734375, + "learning_rate": 0.0001864327485380117, + "loss": 4.96, + "step": 3060 + }, + { + "epoch": 0.16110526315789472, + "grad_norm": 0.86328125, + "learning_rate": 0.00018642105263157896, + "loss": 4.6956, + "step": 3061 + }, + { + "epoch": 0.1611578947368421, + "grad_norm": 0.640625, + "learning_rate": 0.0001864093567251462, + "loss": 4.5444, + "step": 3062 + }, + { + "epoch": 0.16121052631578947, + "grad_norm": 0.62109375, + "learning_rate": 0.00018639766081871345, + "loss": 4.4299, + "step": 3063 + }, + { + "epoch": 0.16126315789473683, + "grad_norm": 0.69921875, + "learning_rate": 0.0001863859649122807, + "loss": 4.1984, + "step": 3064 + }, + { + "epoch": 0.16131578947368422, + "grad_norm": 0.94140625, + "learning_rate": 0.00018637426900584795, + "loss": 4.5744, + "step": 3065 + }, + { + "epoch": 0.16136842105263158, + "grad_norm": 0.53515625, + "learning_rate": 0.0001863625730994152, + "loss": 4.9981, + "step": 3066 + }, + { + "epoch": 0.16142105263157894, + "grad_norm": 0.7265625, + "learning_rate": 0.00018635087719298248, + "loss": 4.5203, + "step": 3067 + }, + { + "epoch": 0.16147368421052632, + "grad_norm": 0.5625, + "learning_rate": 0.00018633918128654973, + "loss": 4.5822, + "step": 3068 + }, + { + "epoch": 0.16152631578947368, + "grad_norm": 0.59765625, + "learning_rate": 0.00018632748538011698, + "loss": 4.5372, + "step": 3069 + }, + { + "epoch": 0.16157894736842104, + "grad_norm": 0.59765625, + "learning_rate": 0.00018631578947368423, + "loss": 4.7848, + "step": 3070 + }, + { + "epoch": 0.16163157894736843, + "grad_norm": 0.67578125, + "learning_rate": 0.00018630409356725148, + "loss": 4.4135, + "step": 3071 + }, + { + "epoch": 0.1616842105263158, + "grad_norm": 0.58203125, + "learning_rate": 0.00018629239766081873, + "loss": 4.4906, + "step": 3072 + }, + { + "epoch": 0.16173684210526315, + "grad_norm": 0.671875, + "learning_rate": 0.00018628070175438598, + "loss": 4.4279, + "step": 3073 + }, + { + "epoch": 0.16178947368421054, + "grad_norm": 1.0546875, + "learning_rate": 0.00018626900584795323, + "loss": 4.728, + "step": 3074 + }, + { + "epoch": 0.1618421052631579, + "grad_norm": 0.56640625, + "learning_rate": 0.00018625730994152048, + "loss": 4.7198, + "step": 3075 + }, + { + "epoch": 0.16189473684210526, + "grad_norm": 0.5859375, + "learning_rate": 0.00018624561403508773, + "loss": 4.5669, + "step": 3076 + }, + { + "epoch": 0.16194736842105265, + "grad_norm": 0.62109375, + "learning_rate": 0.00018623391812865498, + "loss": 4.8796, + "step": 3077 + }, + { + "epoch": 0.162, + "grad_norm": 0.8671875, + "learning_rate": 0.00018622222222222223, + "loss": 4.486, + "step": 3078 + }, + { + "epoch": 0.16205263157894736, + "grad_norm": 0.67578125, + "learning_rate": 0.00018621052631578947, + "loss": 4.7997, + "step": 3079 + }, + { + "epoch": 0.16210526315789472, + "grad_norm": 0.6015625, + "learning_rate": 0.00018619883040935672, + "loss": 4.4496, + "step": 3080 + }, + { + "epoch": 0.1621578947368421, + "grad_norm": 0.59765625, + "learning_rate": 0.00018618713450292397, + "loss": 4.1883, + "step": 3081 + }, + { + "epoch": 0.16221052631578947, + "grad_norm": 0.578125, + "learning_rate": 0.00018617543859649125, + "loss": 4.1452, + "step": 3082 + }, + { + "epoch": 0.16226315789473683, + "grad_norm": 0.65234375, + "learning_rate": 0.0001861637426900585, + "loss": 4.344, + "step": 3083 + }, + { + "epoch": 0.16231578947368422, + "grad_norm": 0.68359375, + "learning_rate": 0.00018615204678362575, + "loss": 4.7063, + "step": 3084 + }, + { + "epoch": 0.16236842105263158, + "grad_norm": 0.62890625, + "learning_rate": 0.000186140350877193, + "loss": 4.15, + "step": 3085 + }, + { + "epoch": 0.16242105263157894, + "grad_norm": 0.62109375, + "learning_rate": 0.00018612865497076025, + "loss": 4.7595, + "step": 3086 + }, + { + "epoch": 0.16247368421052633, + "grad_norm": 0.734375, + "learning_rate": 0.0001861169590643275, + "loss": 4.3673, + "step": 3087 + }, + { + "epoch": 0.16252631578947369, + "grad_norm": 0.64453125, + "learning_rate": 0.00018610526315789475, + "loss": 4.5822, + "step": 3088 + }, + { + "epoch": 0.16257894736842105, + "grad_norm": 0.66015625, + "learning_rate": 0.000186093567251462, + "loss": 4.562, + "step": 3089 + }, + { + "epoch": 0.16263157894736843, + "grad_norm": 0.7265625, + "learning_rate": 0.00018608187134502925, + "loss": 4.8047, + "step": 3090 + }, + { + "epoch": 0.1626842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.0001860701754385965, + "loss": 4.7913, + "step": 3091 + }, + { + "epoch": 0.16273684210526315, + "grad_norm": 0.7421875, + "learning_rate": 0.00018605847953216375, + "loss": 4.4475, + "step": 3092 + }, + { + "epoch": 0.16278947368421054, + "grad_norm": 0.66796875, + "learning_rate": 0.000186046783625731, + "loss": 4.3758, + "step": 3093 + }, + { + "epoch": 0.1628421052631579, + "grad_norm": 0.78125, + "learning_rate": 0.00018603508771929825, + "loss": 4.0479, + "step": 3094 + }, + { + "epoch": 0.16289473684210526, + "grad_norm": 0.97265625, + "learning_rate": 0.0001860233918128655, + "loss": 4.3239, + "step": 3095 + }, + { + "epoch": 0.16294736842105262, + "grad_norm": 0.5703125, + "learning_rate": 0.00018601169590643274, + "loss": 4.446, + "step": 3096 + }, + { + "epoch": 0.163, + "grad_norm": 0.5390625, + "learning_rate": 0.00018600000000000002, + "loss": 4.4625, + "step": 3097 + }, + { + "epoch": 0.16305263157894737, + "grad_norm": 0.69921875, + "learning_rate": 0.00018598830409356727, + "loss": 4.3342, + "step": 3098 + }, + { + "epoch": 0.16310526315789473, + "grad_norm": 0.62109375, + "learning_rate": 0.00018597660818713452, + "loss": 4.699, + "step": 3099 + }, + { + "epoch": 0.1631578947368421, + "grad_norm": 0.921875, + "learning_rate": 0.00018596491228070177, + "loss": 4.49, + "step": 3100 + }, + { + "epoch": 0.16321052631578947, + "grad_norm": 1.1875, + "learning_rate": 0.00018595321637426902, + "loss": 4.3088, + "step": 3101 + }, + { + "epoch": 0.16326315789473683, + "grad_norm": 0.60546875, + "learning_rate": 0.00018594152046783627, + "loss": 4.4924, + "step": 3102 + }, + { + "epoch": 0.16331578947368422, + "grad_norm": 0.6015625, + "learning_rate": 0.00018592982456140352, + "loss": 4.3599, + "step": 3103 + }, + { + "epoch": 0.16336842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.00018591812865497077, + "loss": 4.8586, + "step": 3104 + }, + { + "epoch": 0.16342105263157894, + "grad_norm": 0.56640625, + "learning_rate": 0.00018590643274853802, + "loss": 4.6683, + "step": 3105 + }, + { + "epoch": 0.16347368421052633, + "grad_norm": 0.6640625, + "learning_rate": 0.00018589473684210527, + "loss": 4.4111, + "step": 3106 + }, + { + "epoch": 0.1635263157894737, + "grad_norm": 0.58203125, + "learning_rate": 0.00018588304093567252, + "loss": 4.5187, + "step": 3107 + }, + { + "epoch": 0.16357894736842105, + "grad_norm": 0.5859375, + "learning_rate": 0.00018587134502923977, + "loss": 4.7377, + "step": 3108 + }, + { + "epoch": 0.16363157894736843, + "grad_norm": 0.76953125, + "learning_rate": 0.00018585964912280702, + "loss": 4.6351, + "step": 3109 + }, + { + "epoch": 0.1636842105263158, + "grad_norm": 0.7265625, + "learning_rate": 0.00018584795321637427, + "loss": 4.3945, + "step": 3110 + }, + { + "epoch": 0.16373684210526315, + "grad_norm": 0.6875, + "learning_rate": 0.00018583625730994152, + "loss": 4.4891, + "step": 3111 + }, + { + "epoch": 0.1637894736842105, + "grad_norm": 0.60546875, + "learning_rate": 0.0001858245614035088, + "loss": 4.572, + "step": 3112 + }, + { + "epoch": 0.1638421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.00018581286549707604, + "loss": 4.1256, + "step": 3113 + }, + { + "epoch": 0.16389473684210526, + "grad_norm": 0.625, + "learning_rate": 0.0001858011695906433, + "loss": 4.2287, + "step": 3114 + }, + { + "epoch": 0.16394736842105262, + "grad_norm": 0.5546875, + "learning_rate": 0.00018578947368421054, + "loss": 4.5899, + "step": 3115 + }, + { + "epoch": 0.164, + "grad_norm": 0.66015625, + "learning_rate": 0.0001857777777777778, + "loss": 4.6944, + "step": 3116 + }, + { + "epoch": 0.16405263157894737, + "grad_norm": 0.67578125, + "learning_rate": 0.00018576608187134504, + "loss": 4.8325, + "step": 3117 + }, + { + "epoch": 0.16410526315789473, + "grad_norm": 0.69921875, + "learning_rate": 0.0001857543859649123, + "loss": 4.7502, + "step": 3118 + }, + { + "epoch": 0.1641578947368421, + "grad_norm": 0.67578125, + "learning_rate": 0.00018574269005847954, + "loss": 4.6014, + "step": 3119 + }, + { + "epoch": 0.16421052631578947, + "grad_norm": 0.66015625, + "learning_rate": 0.0001857309941520468, + "loss": 3.9951, + "step": 3120 + }, + { + "epoch": 0.16426315789473683, + "grad_norm": 0.890625, + "learning_rate": 0.00018571929824561404, + "loss": 4.3606, + "step": 3121 + }, + { + "epoch": 0.16431578947368422, + "grad_norm": 0.5859375, + "learning_rate": 0.0001857076023391813, + "loss": 4.8984, + "step": 3122 + }, + { + "epoch": 0.16436842105263158, + "grad_norm": 0.7265625, + "learning_rate": 0.00018569590643274854, + "loss": 4.3073, + "step": 3123 + }, + { + "epoch": 0.16442105263157894, + "grad_norm": 0.66796875, + "learning_rate": 0.0001856842105263158, + "loss": 4.7236, + "step": 3124 + }, + { + "epoch": 0.16447368421052633, + "grad_norm": 0.98828125, + "learning_rate": 0.00018567251461988304, + "loss": 4.0479, + "step": 3125 + }, + { + "epoch": 0.1645263157894737, + "grad_norm": 0.6328125, + "learning_rate": 0.0001856608187134503, + "loss": 4.0203, + "step": 3126 + }, + { + "epoch": 0.16457894736842105, + "grad_norm": 0.65625, + "learning_rate": 0.00018564912280701756, + "loss": 4.3736, + "step": 3127 + }, + { + "epoch": 0.16463157894736843, + "grad_norm": 0.6015625, + "learning_rate": 0.0001856374269005848, + "loss": 4.5103, + "step": 3128 + }, + { + "epoch": 0.1646842105263158, + "grad_norm": 0.75390625, + "learning_rate": 0.00018562573099415206, + "loss": 4.7338, + "step": 3129 + }, + { + "epoch": 0.16473684210526315, + "grad_norm": 0.640625, + "learning_rate": 0.0001856140350877193, + "loss": 4.4818, + "step": 3130 + }, + { + "epoch": 0.1647894736842105, + "grad_norm": 0.73828125, + "learning_rate": 0.00018560233918128656, + "loss": 4.3238, + "step": 3131 + }, + { + "epoch": 0.1648421052631579, + "grad_norm": 0.70703125, + "learning_rate": 0.0001855906432748538, + "loss": 4.542, + "step": 3132 + }, + { + "epoch": 0.16489473684210526, + "grad_norm": 0.7734375, + "learning_rate": 0.00018557894736842106, + "loss": 4.6776, + "step": 3133 + }, + { + "epoch": 0.16494736842105262, + "grad_norm": 0.83984375, + "learning_rate": 0.0001855672514619883, + "loss": 4.5094, + "step": 3134 + }, + { + "epoch": 0.165, + "grad_norm": 0.5625, + "learning_rate": 0.00018555555555555556, + "loss": 4.1997, + "step": 3135 + }, + { + "epoch": 0.16505263157894737, + "grad_norm": 0.58203125, + "learning_rate": 0.0001855438596491228, + "loss": 4.7109, + "step": 3136 + }, + { + "epoch": 0.16510526315789473, + "grad_norm": 0.59375, + "learning_rate": 0.00018553216374269006, + "loss": 4.5396, + "step": 3137 + }, + { + "epoch": 0.16515789473684211, + "grad_norm": 0.6328125, + "learning_rate": 0.0001855204678362573, + "loss": 4.2907, + "step": 3138 + }, + { + "epoch": 0.16521052631578947, + "grad_norm": 0.5703125, + "learning_rate": 0.00018550877192982456, + "loss": 4.65, + "step": 3139 + }, + { + "epoch": 0.16526315789473683, + "grad_norm": 0.63671875, + "learning_rate": 0.0001854970760233918, + "loss": 4.2579, + "step": 3140 + }, + { + "epoch": 0.16531578947368422, + "grad_norm": 0.55078125, + "learning_rate": 0.00018548538011695906, + "loss": 4.6547, + "step": 3141 + }, + { + "epoch": 0.16536842105263158, + "grad_norm": 0.57421875, + "learning_rate": 0.00018547368421052633, + "loss": 4.3095, + "step": 3142 + }, + { + "epoch": 0.16542105263157894, + "grad_norm": 0.60546875, + "learning_rate": 0.00018546198830409358, + "loss": 4.4826, + "step": 3143 + }, + { + "epoch": 0.16547368421052633, + "grad_norm": 0.61328125, + "learning_rate": 0.00018545029239766083, + "loss": 4.6681, + "step": 3144 + }, + { + "epoch": 0.1655263157894737, + "grad_norm": 0.74609375, + "learning_rate": 0.00018543859649122808, + "loss": 4.3961, + "step": 3145 + }, + { + "epoch": 0.16557894736842105, + "grad_norm": 0.6328125, + "learning_rate": 0.00018542690058479533, + "loss": 4.4994, + "step": 3146 + }, + { + "epoch": 0.1656315789473684, + "grad_norm": 0.7265625, + "learning_rate": 0.00018541520467836258, + "loss": 4.4159, + "step": 3147 + }, + { + "epoch": 0.1656842105263158, + "grad_norm": 0.6171875, + "learning_rate": 0.00018540350877192983, + "loss": 4.6467, + "step": 3148 + }, + { + "epoch": 0.16573684210526315, + "grad_norm": 0.578125, + "learning_rate": 0.00018539181286549708, + "loss": 4.5602, + "step": 3149 + }, + { + "epoch": 0.16578947368421051, + "grad_norm": 0.59375, + "learning_rate": 0.00018538011695906433, + "loss": 4.5857, + "step": 3150 + }, + { + "epoch": 0.1658421052631579, + "grad_norm": 0.69921875, + "learning_rate": 0.00018536842105263158, + "loss": 4.4622, + "step": 3151 + }, + { + "epoch": 0.16589473684210526, + "grad_norm": 0.64453125, + "learning_rate": 0.00018535672514619883, + "loss": 4.3708, + "step": 3152 + }, + { + "epoch": 0.16594736842105262, + "grad_norm": 0.6171875, + "learning_rate": 0.00018534502923976608, + "loss": 4.2998, + "step": 3153 + }, + { + "epoch": 0.166, + "grad_norm": 0.71484375, + "learning_rate": 0.00018533333333333333, + "loss": 4.4505, + "step": 3154 + }, + { + "epoch": 0.16605263157894737, + "grad_norm": 0.546875, + "learning_rate": 0.00018532163742690058, + "loss": 4.765, + "step": 3155 + }, + { + "epoch": 0.16610526315789473, + "grad_norm": 0.63671875, + "learning_rate": 0.00018530994152046786, + "loss": 4.4735, + "step": 3156 + }, + { + "epoch": 0.16615789473684212, + "grad_norm": 0.62109375, + "learning_rate": 0.0001852982456140351, + "loss": 4.5708, + "step": 3157 + }, + { + "epoch": 0.16621052631578948, + "grad_norm": 0.625, + "learning_rate": 0.00018528654970760236, + "loss": 4.4413, + "step": 3158 + }, + { + "epoch": 0.16626315789473683, + "grad_norm": 0.58984375, + "learning_rate": 0.0001852748538011696, + "loss": 4.6954, + "step": 3159 + }, + { + "epoch": 0.16631578947368422, + "grad_norm": 0.6875, + "learning_rate": 0.00018526315789473685, + "loss": 4.812, + "step": 3160 + }, + { + "epoch": 0.16636842105263158, + "grad_norm": 0.7578125, + "learning_rate": 0.0001852514619883041, + "loss": 4.2388, + "step": 3161 + }, + { + "epoch": 0.16642105263157894, + "grad_norm": 0.58984375, + "learning_rate": 0.00018523976608187135, + "loss": 4.3439, + "step": 3162 + }, + { + "epoch": 0.16647368421052633, + "grad_norm": 0.5859375, + "learning_rate": 0.0001852280701754386, + "loss": 4.8557, + "step": 3163 + }, + { + "epoch": 0.1665263157894737, + "grad_norm": 0.734375, + "learning_rate": 0.00018521637426900585, + "loss": 4.1583, + "step": 3164 + }, + { + "epoch": 0.16657894736842105, + "grad_norm": 0.6640625, + "learning_rate": 0.0001852046783625731, + "loss": 4.4744, + "step": 3165 + }, + { + "epoch": 0.1666315789473684, + "grad_norm": 0.69921875, + "learning_rate": 0.00018519298245614035, + "loss": 4.3191, + "step": 3166 + }, + { + "epoch": 0.1666842105263158, + "grad_norm": 0.609375, + "learning_rate": 0.0001851812865497076, + "loss": 4.5143, + "step": 3167 + }, + { + "epoch": 0.16673684210526316, + "grad_norm": 0.60546875, + "learning_rate": 0.00018516959064327485, + "loss": 4.6153, + "step": 3168 + }, + { + "epoch": 0.16678947368421052, + "grad_norm": 0.609375, + "learning_rate": 0.0001851578947368421, + "loss": 4.4727, + "step": 3169 + }, + { + "epoch": 0.1668421052631579, + "grad_norm": 0.63671875, + "learning_rate": 0.00018514619883040935, + "loss": 4.5322, + "step": 3170 + }, + { + "epoch": 0.16689473684210526, + "grad_norm": 0.6796875, + "learning_rate": 0.00018513450292397663, + "loss": 4.4042, + "step": 3171 + }, + { + "epoch": 0.16694736842105262, + "grad_norm": 0.62890625, + "learning_rate": 0.00018512280701754388, + "loss": 4.8642, + "step": 3172 + }, + { + "epoch": 0.167, + "grad_norm": 0.70703125, + "learning_rate": 0.00018511111111111113, + "loss": 4.6126, + "step": 3173 + }, + { + "epoch": 0.16705263157894737, + "grad_norm": 0.76171875, + "learning_rate": 0.00018509941520467838, + "loss": 4.4634, + "step": 3174 + }, + { + "epoch": 0.16710526315789473, + "grad_norm": 0.5546875, + "learning_rate": 0.00018508771929824563, + "loss": 3.9215, + "step": 3175 + }, + { + "epoch": 0.16715789473684212, + "grad_norm": 0.6328125, + "learning_rate": 0.00018507602339181287, + "loss": 4.5088, + "step": 3176 + }, + { + "epoch": 0.16721052631578948, + "grad_norm": 0.75390625, + "learning_rate": 0.00018506432748538012, + "loss": 4.108, + "step": 3177 + }, + { + "epoch": 0.16726315789473684, + "grad_norm": 0.609375, + "learning_rate": 0.00018505263157894737, + "loss": 4.2065, + "step": 3178 + }, + { + "epoch": 0.16731578947368422, + "grad_norm": 0.57421875, + "learning_rate": 0.00018504093567251462, + "loss": 4.6732, + "step": 3179 + }, + { + "epoch": 0.16736842105263158, + "grad_norm": 0.671875, + "learning_rate": 0.00018502923976608187, + "loss": 4.4309, + "step": 3180 + }, + { + "epoch": 0.16742105263157894, + "grad_norm": 0.71484375, + "learning_rate": 0.00018501754385964912, + "loss": 4.1911, + "step": 3181 + }, + { + "epoch": 0.1674736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00018500584795321637, + "loss": 4.4974, + "step": 3182 + }, + { + "epoch": 0.1675263157894737, + "grad_norm": 0.5703125, + "learning_rate": 0.00018499415204678362, + "loss": 4.3423, + "step": 3183 + }, + { + "epoch": 0.16757894736842105, + "grad_norm": 0.65625, + "learning_rate": 0.00018498245614035087, + "loss": 3.8931, + "step": 3184 + }, + { + "epoch": 0.1676315789473684, + "grad_norm": 0.7421875, + "learning_rate": 0.00018497076023391812, + "loss": 4.639, + "step": 3185 + }, + { + "epoch": 0.1676842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.0001849590643274854, + "loss": 4.7214, + "step": 3186 + }, + { + "epoch": 0.16773684210526316, + "grad_norm": 0.6640625, + "learning_rate": 0.00018494736842105265, + "loss": 4.552, + "step": 3187 + }, + { + "epoch": 0.16778947368421052, + "grad_norm": 0.9921875, + "learning_rate": 0.0001849356725146199, + "loss": 4.7849, + "step": 3188 + }, + { + "epoch": 0.1678421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.00018492397660818715, + "loss": 4.5546, + "step": 3189 + }, + { + "epoch": 0.16789473684210526, + "grad_norm": 0.625, + "learning_rate": 0.0001849122807017544, + "loss": 4.5958, + "step": 3190 + }, + { + "epoch": 0.16794736842105262, + "grad_norm": 0.69921875, + "learning_rate": 0.00018490058479532165, + "loss": 4.2944, + "step": 3191 + }, + { + "epoch": 0.168, + "grad_norm": 0.56640625, + "learning_rate": 0.0001848888888888889, + "loss": 4.5746, + "step": 3192 + }, + { + "epoch": 0.16805263157894737, + "grad_norm": 0.57421875, + "learning_rate": 0.00018487719298245617, + "loss": 4.3284, + "step": 3193 + }, + { + "epoch": 0.16810526315789473, + "grad_norm": 0.82421875, + "learning_rate": 0.00018486549707602342, + "loss": 4.3488, + "step": 3194 + }, + { + "epoch": 0.16815789473684212, + "grad_norm": 0.5546875, + "learning_rate": 0.00018485380116959064, + "loss": 4.5592, + "step": 3195 + }, + { + "epoch": 0.16821052631578948, + "grad_norm": 0.625, + "learning_rate": 0.0001848421052631579, + "loss": 4.4948, + "step": 3196 + }, + { + "epoch": 0.16826315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.00018483040935672514, + "loss": 4.406, + "step": 3197 + }, + { + "epoch": 0.16831578947368422, + "grad_norm": 0.53125, + "learning_rate": 0.0001848187134502924, + "loss": 4.7213, + "step": 3198 + }, + { + "epoch": 0.16836842105263158, + "grad_norm": 0.7265625, + "learning_rate": 0.00018480701754385964, + "loss": 3.9688, + "step": 3199 + }, + { + "epoch": 0.16842105263157894, + "grad_norm": 0.57421875, + "learning_rate": 0.0001847953216374269, + "loss": 4.4726, + "step": 3200 + }, + { + "epoch": 0.1684736842105263, + "grad_norm": 0.5859375, + "learning_rate": 0.00018478362573099417, + "loss": 4.3994, + "step": 3201 + }, + { + "epoch": 0.1685263157894737, + "grad_norm": 0.6328125, + "learning_rate": 0.00018477192982456142, + "loss": 4.4548, + "step": 3202 + }, + { + "epoch": 0.16857894736842105, + "grad_norm": 0.6328125, + "learning_rate": 0.00018476023391812867, + "loss": 4.2268, + "step": 3203 + }, + { + "epoch": 0.1686315789473684, + "grad_norm": 0.65625, + "learning_rate": 0.00018474853801169592, + "loss": 4.5034, + "step": 3204 + }, + { + "epoch": 0.1686842105263158, + "grad_norm": 0.68359375, + "learning_rate": 0.00018473684210526317, + "loss": 4.4312, + "step": 3205 + }, + { + "epoch": 0.16873684210526316, + "grad_norm": 0.67578125, + "learning_rate": 0.00018472514619883042, + "loss": 4.3658, + "step": 3206 + }, + { + "epoch": 0.16878947368421052, + "grad_norm": 0.5546875, + "learning_rate": 0.00018471345029239767, + "loss": 4.6078, + "step": 3207 + }, + { + "epoch": 0.1688421052631579, + "grad_norm": 0.59765625, + "learning_rate": 0.00018470175438596494, + "loss": 4.5059, + "step": 3208 + }, + { + "epoch": 0.16889473684210526, + "grad_norm": 0.609375, + "learning_rate": 0.0001846900584795322, + "loss": 4.5656, + "step": 3209 + }, + { + "epoch": 0.16894736842105262, + "grad_norm": 0.75, + "learning_rate": 0.00018467836257309942, + "loss": 4.4472, + "step": 3210 + }, + { + "epoch": 0.169, + "grad_norm": 0.68359375, + "learning_rate": 0.00018466666666666666, + "loss": 4.4173, + "step": 3211 + }, + { + "epoch": 0.16905263157894737, + "grad_norm": 0.84765625, + "learning_rate": 0.00018465497076023391, + "loss": 4.5965, + "step": 3212 + }, + { + "epoch": 0.16910526315789473, + "grad_norm": 0.6484375, + "learning_rate": 0.00018464327485380116, + "loss": 3.9, + "step": 3213 + }, + { + "epoch": 0.16915789473684212, + "grad_norm": 0.75390625, + "learning_rate": 0.00018463157894736841, + "loss": 4.5134, + "step": 3214 + }, + { + "epoch": 0.16921052631578948, + "grad_norm": 0.578125, + "learning_rate": 0.00018461988304093566, + "loss": 4.2197, + "step": 3215 + }, + { + "epoch": 0.16926315789473684, + "grad_norm": 1.0, + "learning_rate": 0.00018460818713450294, + "loss": 4.4218, + "step": 3216 + }, + { + "epoch": 0.1693157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.0001845964912280702, + "loss": 4.3505, + "step": 3217 + }, + { + "epoch": 0.16936842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.00018458479532163744, + "loss": 4.6674, + "step": 3218 + }, + { + "epoch": 0.16942105263157894, + "grad_norm": 0.57421875, + "learning_rate": 0.0001845730994152047, + "loss": 4.4094, + "step": 3219 + }, + { + "epoch": 0.1694736842105263, + "grad_norm": 0.6484375, + "learning_rate": 0.00018456140350877194, + "loss": 4.3454, + "step": 3220 + }, + { + "epoch": 0.1695263157894737, + "grad_norm": 0.5859375, + "learning_rate": 0.0001845497076023392, + "loss": 4.3847, + "step": 3221 + }, + { + "epoch": 0.16957894736842105, + "grad_norm": 0.62109375, + "learning_rate": 0.00018453801169590644, + "loss": 4.4661, + "step": 3222 + }, + { + "epoch": 0.1696315789473684, + "grad_norm": 0.57421875, + "learning_rate": 0.00018452631578947371, + "loss": 4.097, + "step": 3223 + }, + { + "epoch": 0.1696842105263158, + "grad_norm": 0.5546875, + "learning_rate": 0.00018451461988304096, + "loss": 4.8514, + "step": 3224 + }, + { + "epoch": 0.16973684210526316, + "grad_norm": 0.8046875, + "learning_rate": 0.00018450292397660819, + "loss": 4.5257, + "step": 3225 + }, + { + "epoch": 0.16978947368421052, + "grad_norm": 0.6171875, + "learning_rate": 0.00018449122807017544, + "loss": 4.6985, + "step": 3226 + }, + { + "epoch": 0.1698421052631579, + "grad_norm": 0.6171875, + "learning_rate": 0.00018447953216374269, + "loss": 4.6965, + "step": 3227 + }, + { + "epoch": 0.16989473684210527, + "grad_norm": 0.68359375, + "learning_rate": 0.00018446783625730994, + "loss": 4.6855, + "step": 3228 + }, + { + "epoch": 0.16994736842105262, + "grad_norm": 0.671875, + "learning_rate": 0.00018445614035087718, + "loss": 4.3973, + "step": 3229 + }, + { + "epoch": 0.17, + "grad_norm": 0.62109375, + "learning_rate": 0.00018444444444444446, + "loss": 3.9864, + "step": 3230 + }, + { + "epoch": 0.17005263157894737, + "grad_norm": 0.51953125, + "learning_rate": 0.0001844327485380117, + "loss": 4.3778, + "step": 3231 + }, + { + "epoch": 0.17010526315789473, + "grad_norm": 0.546875, + "learning_rate": 0.00018442105263157896, + "loss": 4.5801, + "step": 3232 + }, + { + "epoch": 0.17015789473684212, + "grad_norm": 0.58984375, + "learning_rate": 0.0001844093567251462, + "loss": 4.3195, + "step": 3233 + }, + { + "epoch": 0.17021052631578948, + "grad_norm": 0.62109375, + "learning_rate": 0.00018439766081871346, + "loss": 4.1166, + "step": 3234 + }, + { + "epoch": 0.17026315789473684, + "grad_norm": 0.6328125, + "learning_rate": 0.0001843859649122807, + "loss": 4.1585, + "step": 3235 + }, + { + "epoch": 0.1703157894736842, + "grad_norm": 0.71875, + "learning_rate": 0.00018437426900584796, + "loss": 4.6251, + "step": 3236 + }, + { + "epoch": 0.17036842105263159, + "grad_norm": 0.62109375, + "learning_rate": 0.0001843625730994152, + "loss": 4.4381, + "step": 3237 + }, + { + "epoch": 0.17042105263157895, + "grad_norm": 0.61328125, + "learning_rate": 0.00018435087719298249, + "loss": 4.6852, + "step": 3238 + }, + { + "epoch": 0.1704736842105263, + "grad_norm": 0.73828125, + "learning_rate": 0.00018433918128654973, + "loss": 4.3493, + "step": 3239 + }, + { + "epoch": 0.1705263157894737, + "grad_norm": 0.66015625, + "learning_rate": 0.00018432748538011698, + "loss": 4.4627, + "step": 3240 + }, + { + "epoch": 0.17057894736842105, + "grad_norm": 0.6328125, + "learning_rate": 0.0001843157894736842, + "loss": 4.885, + "step": 3241 + }, + { + "epoch": 0.1706315789473684, + "grad_norm": 0.60546875, + "learning_rate": 0.00018430409356725146, + "loss": 4.4098, + "step": 3242 + }, + { + "epoch": 0.1706842105263158, + "grad_norm": 0.6171875, + "learning_rate": 0.0001842923976608187, + "loss": 4.3726, + "step": 3243 + }, + { + "epoch": 0.17073684210526316, + "grad_norm": 0.5703125, + "learning_rate": 0.00018428070175438596, + "loss": 4.1455, + "step": 3244 + }, + { + "epoch": 0.17078947368421052, + "grad_norm": 0.61328125, + "learning_rate": 0.00018426900584795323, + "loss": 4.56, + "step": 3245 + }, + { + "epoch": 0.1708421052631579, + "grad_norm": 0.59375, + "learning_rate": 0.00018425730994152048, + "loss": 3.9257, + "step": 3246 + }, + { + "epoch": 0.17089473684210527, + "grad_norm": 0.578125, + "learning_rate": 0.00018424561403508773, + "loss": 4.255, + "step": 3247 + }, + { + "epoch": 0.17094736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00018423391812865498, + "loss": 4.7764, + "step": 3248 + }, + { + "epoch": 0.171, + "grad_norm": 0.65625, + "learning_rate": 0.00018422222222222223, + "loss": 4.2049, + "step": 3249 + }, + { + "epoch": 0.17105263157894737, + "grad_norm": 0.61328125, + "learning_rate": 0.00018421052631578948, + "loss": 4.4159, + "step": 3250 + }, + { + "epoch": 0.17110526315789473, + "grad_norm": 1.2890625, + "learning_rate": 0.00018419883040935673, + "loss": 4.6758, + "step": 3251 + }, + { + "epoch": 0.1711578947368421, + "grad_norm": 0.5859375, + "learning_rate": 0.00018418713450292398, + "loss": 4.4296, + "step": 3252 + }, + { + "epoch": 0.17121052631578948, + "grad_norm": 0.79296875, + "learning_rate": 0.00018417543859649126, + "loss": 4.0448, + "step": 3253 + }, + { + "epoch": 0.17126315789473684, + "grad_norm": 0.6640625, + "learning_rate": 0.0001841637426900585, + "loss": 4.5001, + "step": 3254 + }, + { + "epoch": 0.1713157894736842, + "grad_norm": 0.62890625, + "learning_rate": 0.00018415204678362576, + "loss": 4.304, + "step": 3255 + }, + { + "epoch": 0.1713684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00018414035087719298, + "loss": 4.3505, + "step": 3256 + }, + { + "epoch": 0.17142105263157895, + "grad_norm": 0.58203125, + "learning_rate": 0.00018412865497076023, + "loss": 4.2645, + "step": 3257 + }, + { + "epoch": 0.1714736842105263, + "grad_norm": 0.5625, + "learning_rate": 0.00018411695906432748, + "loss": 4.4916, + "step": 3258 + }, + { + "epoch": 0.1715263157894737, + "grad_norm": 0.59375, + "learning_rate": 0.00018410526315789473, + "loss": 4.4138, + "step": 3259 + }, + { + "epoch": 0.17157894736842105, + "grad_norm": 0.640625, + "learning_rate": 0.000184093567251462, + "loss": 4.6611, + "step": 3260 + }, + { + "epoch": 0.1716315789473684, + "grad_norm": 0.68359375, + "learning_rate": 0.00018408187134502925, + "loss": 4.4746, + "step": 3261 + }, + { + "epoch": 0.1716842105263158, + "grad_norm": 0.796875, + "learning_rate": 0.0001840701754385965, + "loss": 4.0964, + "step": 3262 + }, + { + "epoch": 0.17173684210526316, + "grad_norm": 0.5859375, + "learning_rate": 0.00018405847953216375, + "loss": 4.2936, + "step": 3263 + }, + { + "epoch": 0.17178947368421052, + "grad_norm": 0.6484375, + "learning_rate": 0.000184046783625731, + "loss": 4.5731, + "step": 3264 + }, + { + "epoch": 0.1718421052631579, + "grad_norm": 0.64453125, + "learning_rate": 0.00018403508771929825, + "loss": 4.5545, + "step": 3265 + }, + { + "epoch": 0.17189473684210527, + "grad_norm": 0.8125, + "learning_rate": 0.0001840233918128655, + "loss": 4.4576, + "step": 3266 + }, + { + "epoch": 0.17194736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.00018401169590643278, + "loss": 4.4462, + "step": 3267 + }, + { + "epoch": 0.172, + "grad_norm": 0.9609375, + "learning_rate": 0.00018400000000000003, + "loss": 3.7605, + "step": 3268 + }, + { + "epoch": 0.17205263157894737, + "grad_norm": 0.62109375, + "learning_rate": 0.00018398830409356728, + "loss": 4.4611, + "step": 3269 + }, + { + "epoch": 0.17210526315789473, + "grad_norm": 0.69140625, + "learning_rate": 0.00018397660818713453, + "loss": 4.5365, + "step": 3270 + }, + { + "epoch": 0.1721578947368421, + "grad_norm": 0.70703125, + "learning_rate": 0.00018396491228070175, + "loss": 4.4002, + "step": 3271 + }, + { + "epoch": 0.17221052631578948, + "grad_norm": 0.70703125, + "learning_rate": 0.000183953216374269, + "loss": 4.6798, + "step": 3272 + }, + { + "epoch": 0.17226315789473684, + "grad_norm": 0.63671875, + "learning_rate": 0.00018394152046783625, + "loss": 4.4712, + "step": 3273 + }, + { + "epoch": 0.1723157894736842, + "grad_norm": 0.6640625, + "learning_rate": 0.0001839298245614035, + "loss": 4.2855, + "step": 3274 + }, + { + "epoch": 0.1723684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00018391812865497077, + "loss": 4.6396, + "step": 3275 + }, + { + "epoch": 0.17242105263157895, + "grad_norm": 0.609375, + "learning_rate": 0.00018390643274853802, + "loss": 4.4199, + "step": 3276 + }, + { + "epoch": 0.1724736842105263, + "grad_norm": 0.64453125, + "learning_rate": 0.00018389473684210527, + "loss": 4.3224, + "step": 3277 + }, + { + "epoch": 0.1725263157894737, + "grad_norm": 0.6484375, + "learning_rate": 0.00018388304093567252, + "loss": 4.2579, + "step": 3278 + }, + { + "epoch": 0.17257894736842105, + "grad_norm": 0.53125, + "learning_rate": 0.00018387134502923977, + "loss": 4.657, + "step": 3279 + }, + { + "epoch": 0.1726315789473684, + "grad_norm": 0.65234375, + "learning_rate": 0.00018385964912280702, + "loss": 4.2604, + "step": 3280 + }, + { + "epoch": 0.1726842105263158, + "grad_norm": 0.625, + "learning_rate": 0.00018384795321637427, + "loss": 4.689, + "step": 3281 + }, + { + "epoch": 0.17273684210526316, + "grad_norm": 0.671875, + "learning_rate": 0.00018383625730994155, + "loss": 4.744, + "step": 3282 + }, + { + "epoch": 0.17278947368421052, + "grad_norm": 0.60546875, + "learning_rate": 0.0001838245614035088, + "loss": 4.4251, + "step": 3283 + }, + { + "epoch": 0.1728421052631579, + "grad_norm": 0.6484375, + "learning_rate": 0.00018381286549707605, + "loss": 3.9766, + "step": 3284 + }, + { + "epoch": 0.17289473684210527, + "grad_norm": 0.7890625, + "learning_rate": 0.0001838011695906433, + "loss": 4.492, + "step": 3285 + }, + { + "epoch": 0.17294736842105263, + "grad_norm": 1.296875, + "learning_rate": 0.00018378947368421055, + "loss": 4.6691, + "step": 3286 + }, + { + "epoch": 0.173, + "grad_norm": 0.6328125, + "learning_rate": 0.00018377777777777777, + "loss": 4.0814, + "step": 3287 + }, + { + "epoch": 0.17305263157894737, + "grad_norm": 0.61328125, + "learning_rate": 0.00018376608187134502, + "loss": 4.4277, + "step": 3288 + }, + { + "epoch": 0.17310526315789473, + "grad_norm": 0.58984375, + "learning_rate": 0.00018375438596491227, + "loss": 4.4421, + "step": 3289 + }, + { + "epoch": 0.1731578947368421, + "grad_norm": 0.5859375, + "learning_rate": 0.00018374269005847955, + "loss": 4.6216, + "step": 3290 + }, + { + "epoch": 0.17321052631578948, + "grad_norm": 0.71484375, + "learning_rate": 0.0001837309941520468, + "loss": 4.1368, + "step": 3291 + }, + { + "epoch": 0.17326315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.00018371929824561404, + "loss": 4.515, + "step": 3292 + }, + { + "epoch": 0.1733157894736842, + "grad_norm": 0.62109375, + "learning_rate": 0.0001837076023391813, + "loss": 4.3324, + "step": 3293 + }, + { + "epoch": 0.1733684210526316, + "grad_norm": 0.6171875, + "learning_rate": 0.00018369590643274854, + "loss": 4.7558, + "step": 3294 + }, + { + "epoch": 0.17342105263157895, + "grad_norm": 0.74609375, + "learning_rate": 0.0001836842105263158, + "loss": 4.0724, + "step": 3295 + }, + { + "epoch": 0.1734736842105263, + "grad_norm": 0.5703125, + "learning_rate": 0.00018367251461988304, + "loss": 4.3679, + "step": 3296 + }, + { + "epoch": 0.1735263157894737, + "grad_norm": 0.5859375, + "learning_rate": 0.00018366081871345032, + "loss": 4.5928, + "step": 3297 + }, + { + "epoch": 0.17357894736842105, + "grad_norm": 0.5703125, + "learning_rate": 0.00018364912280701757, + "loss": 4.4627, + "step": 3298 + }, + { + "epoch": 0.17363157894736841, + "grad_norm": 0.609375, + "learning_rate": 0.00018363742690058482, + "loss": 4.3025, + "step": 3299 + }, + { + "epoch": 0.1736842105263158, + "grad_norm": 0.75390625, + "learning_rate": 0.00018362573099415207, + "loss": 4.0381, + "step": 3300 + }, + { + "epoch": 0.17373684210526316, + "grad_norm": 0.484375, + "learning_rate": 0.00018361403508771932, + "loss": 4.484, + "step": 3301 + }, + { + "epoch": 0.17378947368421052, + "grad_norm": 0.57421875, + "learning_rate": 0.00018360233918128654, + "loss": 4.643, + "step": 3302 + }, + { + "epoch": 0.17384210526315788, + "grad_norm": 0.6796875, + "learning_rate": 0.0001835906432748538, + "loss": 4.2906, + "step": 3303 + }, + { + "epoch": 0.17389473684210527, + "grad_norm": 0.62890625, + "learning_rate": 0.00018357894736842104, + "loss": 4.6164, + "step": 3304 + }, + { + "epoch": 0.17394736842105263, + "grad_norm": 0.63671875, + "learning_rate": 0.00018356725146198832, + "loss": 4.2923, + "step": 3305 + }, + { + "epoch": 0.174, + "grad_norm": 0.76953125, + "learning_rate": 0.00018355555555555557, + "loss": 4.3214, + "step": 3306 + }, + { + "epoch": 0.17405263157894738, + "grad_norm": 0.5859375, + "learning_rate": 0.00018354385964912282, + "loss": 4.5221, + "step": 3307 + }, + { + "epoch": 0.17410526315789474, + "grad_norm": 0.6171875, + "learning_rate": 0.00018353216374269007, + "loss": 4.363, + "step": 3308 + }, + { + "epoch": 0.1741578947368421, + "grad_norm": 0.55078125, + "learning_rate": 0.00018352046783625731, + "loss": 4.2648, + "step": 3309 + }, + { + "epoch": 0.17421052631578948, + "grad_norm": 0.4921875, + "learning_rate": 0.00018350877192982456, + "loss": 4.5243, + "step": 3310 + }, + { + "epoch": 0.17426315789473684, + "grad_norm": 0.62890625, + "learning_rate": 0.00018349707602339181, + "loss": 4.295, + "step": 3311 + }, + { + "epoch": 0.1743157894736842, + "grad_norm": 0.61328125, + "learning_rate": 0.0001834853801169591, + "loss": 3.9364, + "step": 3312 + }, + { + "epoch": 0.1743684210526316, + "grad_norm": 0.6328125, + "learning_rate": 0.00018347368421052634, + "loss": 3.9849, + "step": 3313 + }, + { + "epoch": 0.17442105263157895, + "grad_norm": 0.640625, + "learning_rate": 0.0001834619883040936, + "loss": 4.1035, + "step": 3314 + }, + { + "epoch": 0.1744736842105263, + "grad_norm": 0.58203125, + "learning_rate": 0.00018345029239766084, + "loss": 4.0059, + "step": 3315 + }, + { + "epoch": 0.1745263157894737, + "grad_norm": 0.60546875, + "learning_rate": 0.0001834385964912281, + "loss": 4.6224, + "step": 3316 + }, + { + "epoch": 0.17457894736842106, + "grad_norm": 0.62890625, + "learning_rate": 0.0001834269005847953, + "loss": 4.3144, + "step": 3317 + }, + { + "epoch": 0.17463157894736842, + "grad_norm": 0.703125, + "learning_rate": 0.00018341520467836256, + "loss": 4.3266, + "step": 3318 + }, + { + "epoch": 0.1746842105263158, + "grad_norm": 0.59765625, + "learning_rate": 0.00018340350877192984, + "loss": 4.5963, + "step": 3319 + }, + { + "epoch": 0.17473684210526316, + "grad_norm": 0.60546875, + "learning_rate": 0.0001833918128654971, + "loss": 3.8689, + "step": 3320 + }, + { + "epoch": 0.17478947368421052, + "grad_norm": 0.5625, + "learning_rate": 0.00018338011695906434, + "loss": 4.3872, + "step": 3321 + }, + { + "epoch": 0.17484210526315788, + "grad_norm": 0.60546875, + "learning_rate": 0.0001833684210526316, + "loss": 3.7499, + "step": 3322 + }, + { + "epoch": 0.17489473684210527, + "grad_norm": 0.5625, + "learning_rate": 0.00018335672514619884, + "loss": 4.2532, + "step": 3323 + }, + { + "epoch": 0.17494736842105263, + "grad_norm": 0.64453125, + "learning_rate": 0.00018334502923976609, + "loss": 4.5075, + "step": 3324 + }, + { + "epoch": 0.175, + "grad_norm": 0.60546875, + "learning_rate": 0.00018333333333333334, + "loss": 4.2751, + "step": 3325 + }, + { + "epoch": 0.17505263157894738, + "grad_norm": 0.5390625, + "learning_rate": 0.00018332163742690058, + "loss": 4.3275, + "step": 3326 + }, + { + "epoch": 0.17510526315789474, + "grad_norm": 0.546875, + "learning_rate": 0.00018330994152046786, + "loss": 4.5261, + "step": 3327 + }, + { + "epoch": 0.1751578947368421, + "grad_norm": 0.625, + "learning_rate": 0.0001832982456140351, + "loss": 3.971, + "step": 3328 + }, + { + "epoch": 0.17521052631578948, + "grad_norm": 0.640625, + "learning_rate": 0.00018328654970760236, + "loss": 4.6164, + "step": 3329 + }, + { + "epoch": 0.17526315789473684, + "grad_norm": 0.59765625, + "learning_rate": 0.0001832748538011696, + "loss": 4.1524, + "step": 3330 + }, + { + "epoch": 0.1753157894736842, + "grad_norm": 0.61328125, + "learning_rate": 0.00018326315789473686, + "loss": 4.5037, + "step": 3331 + }, + { + "epoch": 0.1753684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.0001832514619883041, + "loss": 4.3101, + "step": 3332 + }, + { + "epoch": 0.17542105263157895, + "grad_norm": 0.609375, + "learning_rate": 0.00018323976608187133, + "loss": 3.7995, + "step": 3333 + }, + { + "epoch": 0.1754736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.0001832280701754386, + "loss": 4.0759, + "step": 3334 + }, + { + "epoch": 0.1755263157894737, + "grad_norm": 0.56640625, + "learning_rate": 0.00018321637426900586, + "loss": 4.3777, + "step": 3335 + }, + { + "epoch": 0.17557894736842106, + "grad_norm": 0.59765625, + "learning_rate": 0.0001832046783625731, + "loss": 4.3008, + "step": 3336 + }, + { + "epoch": 0.17563157894736842, + "grad_norm": 0.5625, + "learning_rate": 0.00018319298245614036, + "loss": 4.3, + "step": 3337 + }, + { + "epoch": 0.17568421052631578, + "grad_norm": 0.54296875, + "learning_rate": 0.0001831812865497076, + "loss": 4.1927, + "step": 3338 + }, + { + "epoch": 0.17573684210526316, + "grad_norm": 0.54296875, + "learning_rate": 0.00018316959064327486, + "loss": 4.4374, + "step": 3339 + }, + { + "epoch": 0.17578947368421052, + "grad_norm": 0.56640625, + "learning_rate": 0.0001831578947368421, + "loss": 4.3643, + "step": 3340 + }, + { + "epoch": 0.17584210526315788, + "grad_norm": 0.53515625, + "learning_rate": 0.00018314619883040936, + "loss": 4.5746, + "step": 3341 + }, + { + "epoch": 0.17589473684210527, + "grad_norm": 0.6328125, + "learning_rate": 0.00018313450292397663, + "loss": 4.4484, + "step": 3342 + }, + { + "epoch": 0.17594736842105263, + "grad_norm": 0.64453125, + "learning_rate": 0.00018312280701754388, + "loss": 4.3474, + "step": 3343 + }, + { + "epoch": 0.176, + "grad_norm": 0.58203125, + "learning_rate": 0.00018311111111111113, + "loss": 4.5491, + "step": 3344 + }, + { + "epoch": 0.17605263157894738, + "grad_norm": 0.55859375, + "learning_rate": 0.00018309941520467838, + "loss": 4.244, + "step": 3345 + }, + { + "epoch": 0.17610526315789474, + "grad_norm": 0.51171875, + "learning_rate": 0.00018308771929824563, + "loss": 4.7363, + "step": 3346 + }, + { + "epoch": 0.1761578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00018307602339181288, + "loss": 4.6478, + "step": 3347 + }, + { + "epoch": 0.17621052631578948, + "grad_norm": 0.578125, + "learning_rate": 0.0001830643274853801, + "loss": 4.5112, + "step": 3348 + }, + { + "epoch": 0.17626315789473684, + "grad_norm": 0.7265625, + "learning_rate": 0.00018305263157894738, + "loss": 4.2482, + "step": 3349 + }, + { + "epoch": 0.1763157894736842, + "grad_norm": 0.5703125, + "learning_rate": 0.00018304093567251463, + "loss": 4.2869, + "step": 3350 + }, + { + "epoch": 0.1763684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00018302923976608188, + "loss": 3.8997, + "step": 3351 + }, + { + "epoch": 0.17642105263157895, + "grad_norm": 0.5, + "learning_rate": 0.00018301754385964913, + "loss": 4.5878, + "step": 3352 + }, + { + "epoch": 0.1764736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00018300584795321638, + "loss": 4.339, + "step": 3353 + }, + { + "epoch": 0.1765263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00018299415204678363, + "loss": 4.2922, + "step": 3354 + }, + { + "epoch": 0.17657894736842106, + "grad_norm": 0.5546875, + "learning_rate": 0.00018298245614035088, + "loss": 4.4805, + "step": 3355 + }, + { + "epoch": 0.17663157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00018297076023391815, + "loss": 4.1191, + "step": 3356 + }, + { + "epoch": 0.17668421052631578, + "grad_norm": 0.57421875, + "learning_rate": 0.0001829590643274854, + "loss": 4.0618, + "step": 3357 + }, + { + "epoch": 0.17673684210526316, + "grad_norm": 0.5859375, + "learning_rate": 0.00018294736842105265, + "loss": 4.4158, + "step": 3358 + }, + { + "epoch": 0.17678947368421052, + "grad_norm": 0.62890625, + "learning_rate": 0.0001829356725146199, + "loss": 4.6104, + "step": 3359 + }, + { + "epoch": 0.17684210526315788, + "grad_norm": 0.8671875, + "learning_rate": 0.00018292397660818715, + "loss": 4.3763, + "step": 3360 + }, + { + "epoch": 0.17689473684210527, + "grad_norm": 0.6953125, + "learning_rate": 0.0001829122807017544, + "loss": 4.1432, + "step": 3361 + }, + { + "epoch": 0.17694736842105263, + "grad_norm": 0.69921875, + "learning_rate": 0.00018290058479532165, + "loss": 4.3725, + "step": 3362 + }, + { + "epoch": 0.177, + "grad_norm": 0.73828125, + "learning_rate": 0.00018288888888888887, + "loss": 4.6002, + "step": 3363 + }, + { + "epoch": 0.17705263157894738, + "grad_norm": 0.59765625, + "learning_rate": 0.00018287719298245615, + "loss": 4.4966, + "step": 3364 + }, + { + "epoch": 0.17710526315789474, + "grad_norm": 0.6796875, + "learning_rate": 0.0001828654970760234, + "loss": 4.1431, + "step": 3365 + }, + { + "epoch": 0.1771578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00018285380116959065, + "loss": 4.396, + "step": 3366 + }, + { + "epoch": 0.17721052631578948, + "grad_norm": 0.94140625, + "learning_rate": 0.0001828421052631579, + "loss": 4.3798, + "step": 3367 + }, + { + "epoch": 0.17726315789473684, + "grad_norm": 0.83203125, + "learning_rate": 0.00018283040935672515, + "loss": 4.3564, + "step": 3368 + }, + { + "epoch": 0.1773157894736842, + "grad_norm": 0.8125, + "learning_rate": 0.0001828187134502924, + "loss": 4.1733, + "step": 3369 + }, + { + "epoch": 0.1773684210526316, + "grad_norm": 0.5625, + "learning_rate": 0.00018280701754385965, + "loss": 4.1811, + "step": 3370 + }, + { + "epoch": 0.17742105263157895, + "grad_norm": 0.609375, + "learning_rate": 0.00018279532163742693, + "loss": 4.137, + "step": 3371 + }, + { + "epoch": 0.1774736842105263, + "grad_norm": 0.62109375, + "learning_rate": 0.00018278362573099417, + "loss": 4.2384, + "step": 3372 + }, + { + "epoch": 0.17752631578947367, + "grad_norm": 0.6328125, + "learning_rate": 0.00018277192982456142, + "loss": 3.9399, + "step": 3373 + }, + { + "epoch": 0.17757894736842106, + "grad_norm": 0.5859375, + "learning_rate": 0.00018276023391812867, + "loss": 4.0685, + "step": 3374 + }, + { + "epoch": 0.17763157894736842, + "grad_norm": 0.60546875, + "learning_rate": 0.00018274853801169592, + "loss": 4.5155, + "step": 3375 + }, + { + "epoch": 0.17768421052631578, + "grad_norm": 0.6484375, + "learning_rate": 0.00018273684210526317, + "loss": 4.2546, + "step": 3376 + }, + { + "epoch": 0.17773684210526317, + "grad_norm": 0.70703125, + "learning_rate": 0.00018272514619883042, + "loss": 4.1923, + "step": 3377 + }, + { + "epoch": 0.17778947368421052, + "grad_norm": 0.55859375, + "learning_rate": 0.00018271345029239767, + "loss": 4.2648, + "step": 3378 + }, + { + "epoch": 0.17784210526315788, + "grad_norm": 0.5703125, + "learning_rate": 0.00018270175438596492, + "loss": 4.3922, + "step": 3379 + }, + { + "epoch": 0.17789473684210527, + "grad_norm": 0.6875, + "learning_rate": 0.00018269005847953217, + "loss": 4.4737, + "step": 3380 + }, + { + "epoch": 0.17794736842105263, + "grad_norm": 0.73046875, + "learning_rate": 0.00018267836257309942, + "loss": 4.2896, + "step": 3381 + }, + { + "epoch": 0.178, + "grad_norm": 0.62890625, + "learning_rate": 0.00018266666666666667, + "loss": 4.2503, + "step": 3382 + }, + { + "epoch": 0.17805263157894738, + "grad_norm": 0.65625, + "learning_rate": 0.00018265497076023392, + "loss": 4.0236, + "step": 3383 + }, + { + "epoch": 0.17810526315789474, + "grad_norm": 0.67578125, + "learning_rate": 0.00018264327485380117, + "loss": 4.6046, + "step": 3384 + }, + { + "epoch": 0.1781578947368421, + "grad_norm": 0.921875, + "learning_rate": 0.00018263157894736842, + "loss": 4.4088, + "step": 3385 + }, + { + "epoch": 0.17821052631578949, + "grad_norm": 0.53125, + "learning_rate": 0.0001826198830409357, + "loss": 4.0342, + "step": 3386 + }, + { + "epoch": 0.17826315789473685, + "grad_norm": 0.546875, + "learning_rate": 0.00018260818713450295, + "loss": 4.3996, + "step": 3387 + }, + { + "epoch": 0.1783157894736842, + "grad_norm": 0.640625, + "learning_rate": 0.0001825964912280702, + "loss": 4.7065, + "step": 3388 + }, + { + "epoch": 0.1783684210526316, + "grad_norm": 0.71875, + "learning_rate": 0.00018258479532163744, + "loss": 4.4174, + "step": 3389 + }, + { + "epoch": 0.17842105263157895, + "grad_norm": 0.609375, + "learning_rate": 0.0001825730994152047, + "loss": 4.4112, + "step": 3390 + }, + { + "epoch": 0.1784736842105263, + "grad_norm": 0.6171875, + "learning_rate": 0.00018256140350877194, + "loss": 4.3488, + "step": 3391 + }, + { + "epoch": 0.17852631578947367, + "grad_norm": 0.61328125, + "learning_rate": 0.0001825497076023392, + "loss": 3.7203, + "step": 3392 + }, + { + "epoch": 0.17857894736842106, + "grad_norm": 0.58203125, + "learning_rate": 0.00018253801169590644, + "loss": 4.2792, + "step": 3393 + }, + { + "epoch": 0.17863157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.0001825263157894737, + "loss": 4.3322, + "step": 3394 + }, + { + "epoch": 0.17868421052631578, + "grad_norm": 0.59765625, + "learning_rate": 0.00018251461988304094, + "loss": 4.3957, + "step": 3395 + }, + { + "epoch": 0.17873684210526317, + "grad_norm": 0.57421875, + "learning_rate": 0.0001825029239766082, + "loss": 4.3906, + "step": 3396 + }, + { + "epoch": 0.17878947368421053, + "grad_norm": 0.56640625, + "learning_rate": 0.00018249122807017544, + "loss": 4.2308, + "step": 3397 + }, + { + "epoch": 0.17884210526315789, + "grad_norm": 0.6640625, + "learning_rate": 0.0001824795321637427, + "loss": 4.2435, + "step": 3398 + }, + { + "epoch": 0.17889473684210527, + "grad_norm": 0.609375, + "learning_rate": 0.00018246783625730994, + "loss": 4.2973, + "step": 3399 + }, + { + "epoch": 0.17894736842105263, + "grad_norm": 0.671875, + "learning_rate": 0.0001824561403508772, + "loss": 4.3496, + "step": 3400 + }, + { + "epoch": 0.179, + "grad_norm": 0.68359375, + "learning_rate": 0.00018244444444444447, + "loss": 4.5982, + "step": 3401 + }, + { + "epoch": 0.17905263157894738, + "grad_norm": 0.59375, + "learning_rate": 0.00018243274853801172, + "loss": 4.2744, + "step": 3402 + }, + { + "epoch": 0.17910526315789474, + "grad_norm": 0.61328125, + "learning_rate": 0.00018242105263157897, + "loss": 4.2424, + "step": 3403 + }, + { + "epoch": 0.1791578947368421, + "grad_norm": 0.609375, + "learning_rate": 0.00018240935672514622, + "loss": 4.8392, + "step": 3404 + }, + { + "epoch": 0.1792105263157895, + "grad_norm": 0.734375, + "learning_rate": 0.00018239766081871347, + "loss": 4.4762, + "step": 3405 + }, + { + "epoch": 0.17926315789473685, + "grad_norm": 0.6484375, + "learning_rate": 0.00018238596491228071, + "loss": 4.1502, + "step": 3406 + }, + { + "epoch": 0.1793157894736842, + "grad_norm": 0.62890625, + "learning_rate": 0.00018237426900584796, + "loss": 4.417, + "step": 3407 + }, + { + "epoch": 0.17936842105263157, + "grad_norm": 0.5703125, + "learning_rate": 0.00018236257309941521, + "loss": 4.2409, + "step": 3408 + }, + { + "epoch": 0.17942105263157895, + "grad_norm": 0.7109375, + "learning_rate": 0.00018235087719298246, + "loss": 4.7019, + "step": 3409 + }, + { + "epoch": 0.1794736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.0001823391812865497, + "loss": 4.0448, + "step": 3410 + }, + { + "epoch": 0.17952631578947367, + "grad_norm": 0.609375, + "learning_rate": 0.00018232748538011696, + "loss": 4.4258, + "step": 3411 + }, + { + "epoch": 0.17957894736842106, + "grad_norm": 0.58984375, + "learning_rate": 0.0001823157894736842, + "loss": 4.166, + "step": 3412 + }, + { + "epoch": 0.17963157894736842, + "grad_norm": 0.578125, + "learning_rate": 0.00018230409356725146, + "loss": 4.0094, + "step": 3413 + }, + { + "epoch": 0.17968421052631578, + "grad_norm": 0.546875, + "learning_rate": 0.0001822923976608187, + "loss": 4.3152, + "step": 3414 + }, + { + "epoch": 0.17973684210526317, + "grad_norm": 0.53125, + "learning_rate": 0.00018228070175438596, + "loss": 4.4376, + "step": 3415 + }, + { + "epoch": 0.17978947368421053, + "grad_norm": 0.56640625, + "learning_rate": 0.00018226900584795324, + "loss": 4.5919, + "step": 3416 + }, + { + "epoch": 0.1798421052631579, + "grad_norm": 0.59375, + "learning_rate": 0.0001822573099415205, + "loss": 4.0692, + "step": 3417 + }, + { + "epoch": 0.17989473684210527, + "grad_norm": 0.54296875, + "learning_rate": 0.00018224561403508774, + "loss": 4.5097, + "step": 3418 + }, + { + "epoch": 0.17994736842105263, + "grad_norm": 0.5703125, + "learning_rate": 0.000182233918128655, + "loss": 4.2411, + "step": 3419 + }, + { + "epoch": 0.18, + "grad_norm": 0.6171875, + "learning_rate": 0.00018222222222222224, + "loss": 4.2547, + "step": 3420 + }, + { + "epoch": 0.18005263157894738, + "grad_norm": 0.6796875, + "learning_rate": 0.00018221052631578949, + "loss": 4.1086, + "step": 3421 + }, + { + "epoch": 0.18010526315789474, + "grad_norm": 0.5625, + "learning_rate": 0.00018219883040935674, + "loss": 4.1163, + "step": 3422 + }, + { + "epoch": 0.1801578947368421, + "grad_norm": 0.609375, + "learning_rate": 0.00018218713450292399, + "loss": 3.974, + "step": 3423 + }, + { + "epoch": 0.18021052631578946, + "grad_norm": 0.52734375, + "learning_rate": 0.00018217543859649123, + "loss": 4.4683, + "step": 3424 + }, + { + "epoch": 0.18026315789473685, + "grad_norm": 0.62890625, + "learning_rate": 0.00018216374269005848, + "loss": 4.1216, + "step": 3425 + }, + { + "epoch": 0.1803157894736842, + "grad_norm": 0.55078125, + "learning_rate": 0.00018215204678362573, + "loss": 4.318, + "step": 3426 + }, + { + "epoch": 0.18036842105263157, + "grad_norm": 0.8671875, + "learning_rate": 0.00018214035087719298, + "loss": 4.1249, + "step": 3427 + }, + { + "epoch": 0.18042105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.00018212865497076023, + "loss": 4.5579, + "step": 3428 + }, + { + "epoch": 0.18047368421052631, + "grad_norm": 0.5703125, + "learning_rate": 0.00018211695906432748, + "loss": 4.0994, + "step": 3429 + }, + { + "epoch": 0.18052631578947367, + "grad_norm": 0.63671875, + "learning_rate": 0.00018210526315789476, + "loss": 4.1748, + "step": 3430 + }, + { + "epoch": 0.18057894736842106, + "grad_norm": 0.609375, + "learning_rate": 0.000182093567251462, + "loss": 4.2773, + "step": 3431 + }, + { + "epoch": 0.18063157894736842, + "grad_norm": 0.76953125, + "learning_rate": 0.00018208187134502926, + "loss": 4.556, + "step": 3432 + }, + { + "epoch": 0.18068421052631578, + "grad_norm": 0.84375, + "learning_rate": 0.0001820701754385965, + "loss": 4.8474, + "step": 3433 + }, + { + "epoch": 0.18073684210526317, + "grad_norm": 0.81640625, + "learning_rate": 0.00018205847953216376, + "loss": 4.6861, + "step": 3434 + }, + { + "epoch": 0.18078947368421053, + "grad_norm": 0.6953125, + "learning_rate": 0.000182046783625731, + "loss": 4.4484, + "step": 3435 + }, + { + "epoch": 0.1808421052631579, + "grad_norm": 0.75390625, + "learning_rate": 0.00018203508771929826, + "loss": 4.0177, + "step": 3436 + }, + { + "epoch": 0.18089473684210527, + "grad_norm": 0.478515625, + "learning_rate": 0.0001820233918128655, + "loss": 4.7023, + "step": 3437 + }, + { + "epoch": 0.18094736842105263, + "grad_norm": 0.62890625, + "learning_rate": 0.00018201169590643276, + "loss": 4.2406, + "step": 3438 + }, + { + "epoch": 0.181, + "grad_norm": 0.65625, + "learning_rate": 0.000182, + "loss": 4.1871, + "step": 3439 + }, + { + "epoch": 0.18105263157894738, + "grad_norm": 0.53515625, + "learning_rate": 0.00018198830409356726, + "loss": 4.5257, + "step": 3440 + }, + { + "epoch": 0.18110526315789474, + "grad_norm": 0.6015625, + "learning_rate": 0.0001819766081871345, + "loss": 4.1513, + "step": 3441 + }, + { + "epoch": 0.1811578947368421, + "grad_norm": 0.625, + "learning_rate": 0.00018196491228070175, + "loss": 4.0653, + "step": 3442 + }, + { + "epoch": 0.18121052631578946, + "grad_norm": 0.640625, + "learning_rate": 0.000181953216374269, + "loss": 4.3968, + "step": 3443 + }, + { + "epoch": 0.18126315789473685, + "grad_norm": 0.609375, + "learning_rate": 0.00018194152046783625, + "loss": 4.2967, + "step": 3444 + }, + { + "epoch": 0.1813157894736842, + "grad_norm": 0.67578125, + "learning_rate": 0.00018192982456140353, + "loss": 3.8499, + "step": 3445 + }, + { + "epoch": 0.18136842105263157, + "grad_norm": 0.6015625, + "learning_rate": 0.00018191812865497078, + "loss": 4.43, + "step": 3446 + }, + { + "epoch": 0.18142105263157896, + "grad_norm": 0.68359375, + "learning_rate": 0.00018190643274853803, + "loss": 4.5863, + "step": 3447 + }, + { + "epoch": 0.18147368421052631, + "grad_norm": 0.6875, + "learning_rate": 0.00018189473684210528, + "loss": 4.09, + "step": 3448 + }, + { + "epoch": 0.18152631578947367, + "grad_norm": 0.7109375, + "learning_rate": 0.00018188304093567253, + "loss": 3.9229, + "step": 3449 + }, + { + "epoch": 0.18157894736842106, + "grad_norm": 0.7109375, + "learning_rate": 0.00018187134502923978, + "loss": 3.98, + "step": 3450 + }, + { + "epoch": 0.18163157894736842, + "grad_norm": 0.48828125, + "learning_rate": 0.00018185964912280703, + "loss": 4.7412, + "step": 3451 + }, + { + "epoch": 0.18168421052631578, + "grad_norm": 0.5625, + "learning_rate": 0.00018184795321637428, + "loss": 4.3452, + "step": 3452 + }, + { + "epoch": 0.18173684210526317, + "grad_norm": 0.609375, + "learning_rate": 0.00018183625730994153, + "loss": 4.8593, + "step": 3453 + }, + { + "epoch": 0.18178947368421053, + "grad_norm": 0.6328125, + "learning_rate": 0.00018182456140350878, + "loss": 4.6142, + "step": 3454 + }, + { + "epoch": 0.1818421052631579, + "grad_norm": 0.6328125, + "learning_rate": 0.00018181286549707603, + "loss": 4.6307, + "step": 3455 + }, + { + "epoch": 0.18189473684210528, + "grad_norm": 0.65234375, + "learning_rate": 0.00018180116959064328, + "loss": 4.0778, + "step": 3456 + }, + { + "epoch": 0.18194736842105264, + "grad_norm": 0.58203125, + "learning_rate": 0.00018178947368421053, + "loss": 3.9926, + "step": 3457 + }, + { + "epoch": 0.182, + "grad_norm": 0.57421875, + "learning_rate": 0.00018177777777777778, + "loss": 4.5432, + "step": 3458 + }, + { + "epoch": 0.18205263157894735, + "grad_norm": 0.71875, + "learning_rate": 0.00018176608187134502, + "loss": 4.1173, + "step": 3459 + }, + { + "epoch": 0.18210526315789474, + "grad_norm": 0.59375, + "learning_rate": 0.0001817543859649123, + "loss": 4.0536, + "step": 3460 + }, + { + "epoch": 0.1821578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.00018174269005847955, + "loss": 4.8059, + "step": 3461 + }, + { + "epoch": 0.18221052631578946, + "grad_norm": 0.5625, + "learning_rate": 0.0001817309941520468, + "loss": 4.0414, + "step": 3462 + }, + { + "epoch": 0.18226315789473685, + "grad_norm": 0.6796875, + "learning_rate": 0.00018171929824561405, + "loss": 4.1214, + "step": 3463 + }, + { + "epoch": 0.1823157894736842, + "grad_norm": 0.59375, + "learning_rate": 0.0001817076023391813, + "loss": 4.3551, + "step": 3464 + }, + { + "epoch": 0.18236842105263157, + "grad_norm": 0.58203125, + "learning_rate": 0.00018169590643274855, + "loss": 4.2125, + "step": 3465 + }, + { + "epoch": 0.18242105263157896, + "grad_norm": 0.58984375, + "learning_rate": 0.0001816842105263158, + "loss": 4.4052, + "step": 3466 + }, + { + "epoch": 0.18247368421052632, + "grad_norm": 0.57421875, + "learning_rate": 0.00018167251461988305, + "loss": 4.0724, + "step": 3467 + }, + { + "epoch": 0.18252631578947368, + "grad_norm": 1.3203125, + "learning_rate": 0.0001816608187134503, + "loss": 4.3214, + "step": 3468 + }, + { + "epoch": 0.18257894736842106, + "grad_norm": 0.484375, + "learning_rate": 0.00018164912280701755, + "loss": 4.3253, + "step": 3469 + }, + { + "epoch": 0.18263157894736842, + "grad_norm": 0.609375, + "learning_rate": 0.0001816374269005848, + "loss": 4.5372, + "step": 3470 + }, + { + "epoch": 0.18268421052631578, + "grad_norm": 0.53515625, + "learning_rate": 0.00018162573099415205, + "loss": 4.4423, + "step": 3471 + }, + { + "epoch": 0.18273684210526317, + "grad_norm": 0.74609375, + "learning_rate": 0.0001816140350877193, + "loss": 3.8748, + "step": 3472 + }, + { + "epoch": 0.18278947368421053, + "grad_norm": 0.55859375, + "learning_rate": 0.00018160233918128655, + "loss": 4.4557, + "step": 3473 + }, + { + "epoch": 0.1828421052631579, + "grad_norm": 0.55859375, + "learning_rate": 0.0001815906432748538, + "loss": 4.2283, + "step": 3474 + }, + { + "epoch": 0.18289473684210528, + "grad_norm": 0.52734375, + "learning_rate": 0.00018157894736842107, + "loss": 4.1742, + "step": 3475 + }, + { + "epoch": 0.18294736842105264, + "grad_norm": 0.6015625, + "learning_rate": 0.00018156725146198832, + "loss": 4.3092, + "step": 3476 + }, + { + "epoch": 0.183, + "grad_norm": 0.55078125, + "learning_rate": 0.00018155555555555557, + "loss": 4.2195, + "step": 3477 + }, + { + "epoch": 0.18305263157894736, + "grad_norm": 0.55859375, + "learning_rate": 0.00018154385964912282, + "loss": 3.8861, + "step": 3478 + }, + { + "epoch": 0.18310526315789474, + "grad_norm": 0.55859375, + "learning_rate": 0.00018153216374269007, + "loss": 4.2014, + "step": 3479 + }, + { + "epoch": 0.1831578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.00018152046783625732, + "loss": 4.0863, + "step": 3480 + }, + { + "epoch": 0.18321052631578946, + "grad_norm": 1.6171875, + "learning_rate": 0.00018150877192982457, + "loss": 4.1035, + "step": 3481 + }, + { + "epoch": 0.18326315789473685, + "grad_norm": 0.5546875, + "learning_rate": 0.00018149707602339182, + "loss": 4.1081, + "step": 3482 + }, + { + "epoch": 0.1833157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00018148538011695907, + "loss": 4.2867, + "step": 3483 + }, + { + "epoch": 0.18336842105263157, + "grad_norm": 0.57421875, + "learning_rate": 0.00018147368421052632, + "loss": 4.2708, + "step": 3484 + }, + { + "epoch": 0.18342105263157896, + "grad_norm": 0.609375, + "learning_rate": 0.00018146198830409357, + "loss": 4.1099, + "step": 3485 + }, + { + "epoch": 0.18347368421052632, + "grad_norm": 0.5703125, + "learning_rate": 0.00018145029239766082, + "loss": 4.4528, + "step": 3486 + }, + { + "epoch": 0.18352631578947368, + "grad_norm": 0.625, + "learning_rate": 0.00018143859649122807, + "loss": 4.2697, + "step": 3487 + }, + { + "epoch": 0.18357894736842106, + "grad_norm": 0.6328125, + "learning_rate": 0.00018142690058479532, + "loss": 4.1924, + "step": 3488 + }, + { + "epoch": 0.18363157894736842, + "grad_norm": 0.59765625, + "learning_rate": 0.00018141520467836257, + "loss": 4.1861, + "step": 3489 + }, + { + "epoch": 0.18368421052631578, + "grad_norm": 0.6640625, + "learning_rate": 0.00018140350877192984, + "loss": 4.7469, + "step": 3490 + }, + { + "epoch": 0.18373684210526317, + "grad_norm": 0.73828125, + "learning_rate": 0.0001813918128654971, + "loss": 4.0214, + "step": 3491 + }, + { + "epoch": 0.18378947368421053, + "grad_norm": 0.63671875, + "learning_rate": 0.00018138011695906434, + "loss": 4.2665, + "step": 3492 + }, + { + "epoch": 0.1838421052631579, + "grad_norm": 0.67578125, + "learning_rate": 0.0001813684210526316, + "loss": 4.2594, + "step": 3493 + }, + { + "epoch": 0.18389473684210525, + "grad_norm": 0.57421875, + "learning_rate": 0.00018135672514619884, + "loss": 3.8956, + "step": 3494 + }, + { + "epoch": 0.18394736842105264, + "grad_norm": 0.578125, + "learning_rate": 0.0001813450292397661, + "loss": 4.5716, + "step": 3495 + }, + { + "epoch": 0.184, + "grad_norm": 0.68359375, + "learning_rate": 0.00018133333333333334, + "loss": 4.7741, + "step": 3496 + }, + { + "epoch": 0.18405263157894736, + "grad_norm": 0.66796875, + "learning_rate": 0.0001813216374269006, + "loss": 4.1339, + "step": 3497 + }, + { + "epoch": 0.18410526315789474, + "grad_norm": 0.63671875, + "learning_rate": 0.00018130994152046784, + "loss": 4.1026, + "step": 3498 + }, + { + "epoch": 0.1841578947368421, + "grad_norm": 0.609375, + "learning_rate": 0.0001812982456140351, + "loss": 4.5044, + "step": 3499 + }, + { + "epoch": 0.18421052631578946, + "grad_norm": 0.609375, + "learning_rate": 0.00018128654970760234, + "loss": 4.4003, + "step": 3500 + }, + { + "epoch": 0.18426315789473685, + "grad_norm": 0.60546875, + "learning_rate": 0.0001812748538011696, + "loss": 4.3512, + "step": 3501 + }, + { + "epoch": 0.1843157894736842, + "grad_norm": 0.7109375, + "learning_rate": 0.00018126315789473684, + "loss": 4.2298, + "step": 3502 + }, + { + "epoch": 0.18436842105263157, + "grad_norm": 0.65234375, + "learning_rate": 0.0001812514619883041, + "loss": 3.7726, + "step": 3503 + }, + { + "epoch": 0.18442105263157896, + "grad_norm": 0.55078125, + "learning_rate": 0.00018123976608187134, + "loss": 4.1444, + "step": 3504 + }, + { + "epoch": 0.18447368421052632, + "grad_norm": 0.65625, + "learning_rate": 0.00018122807017543861, + "loss": 4.2762, + "step": 3505 + }, + { + "epoch": 0.18452631578947368, + "grad_norm": 0.6015625, + "learning_rate": 0.00018121637426900586, + "loss": 4.3896, + "step": 3506 + }, + { + "epoch": 0.18457894736842106, + "grad_norm": 0.546875, + "learning_rate": 0.00018120467836257311, + "loss": 4.2187, + "step": 3507 + }, + { + "epoch": 0.18463157894736842, + "grad_norm": 0.59765625, + "learning_rate": 0.00018119298245614036, + "loss": 4.2388, + "step": 3508 + }, + { + "epoch": 0.18468421052631578, + "grad_norm": 0.78125, + "learning_rate": 0.0001811812865497076, + "loss": 4.331, + "step": 3509 + }, + { + "epoch": 0.18473684210526317, + "grad_norm": 0.65625, + "learning_rate": 0.00018116959064327486, + "loss": 4.0679, + "step": 3510 + }, + { + "epoch": 0.18478947368421053, + "grad_norm": 0.671875, + "learning_rate": 0.0001811578947368421, + "loss": 4.4364, + "step": 3511 + }, + { + "epoch": 0.1848421052631579, + "grad_norm": 1.0078125, + "learning_rate": 0.00018114619883040936, + "loss": 4.1049, + "step": 3512 + }, + { + "epoch": 0.18489473684210525, + "grad_norm": 0.546875, + "learning_rate": 0.0001811345029239766, + "loss": 4.1588, + "step": 3513 + }, + { + "epoch": 0.18494736842105264, + "grad_norm": 0.54296875, + "learning_rate": 0.00018112280701754386, + "loss": 4.1997, + "step": 3514 + }, + { + "epoch": 0.185, + "grad_norm": 0.59765625, + "learning_rate": 0.0001811111111111111, + "loss": 3.8215, + "step": 3515 + }, + { + "epoch": 0.18505263157894736, + "grad_norm": 0.59375, + "learning_rate": 0.00018109941520467836, + "loss": 4.5784, + "step": 3516 + }, + { + "epoch": 0.18510526315789474, + "grad_norm": 0.5859375, + "learning_rate": 0.0001810877192982456, + "loss": 4.1644, + "step": 3517 + }, + { + "epoch": 0.1851578947368421, + "grad_norm": 0.75390625, + "learning_rate": 0.00018107602339181286, + "loss": 4.1816, + "step": 3518 + }, + { + "epoch": 0.18521052631578946, + "grad_norm": 0.6015625, + "learning_rate": 0.00018106432748538014, + "loss": 4.2574, + "step": 3519 + }, + { + "epoch": 0.18526315789473685, + "grad_norm": 0.60546875, + "learning_rate": 0.00018105263157894739, + "loss": 4.8147, + "step": 3520 + }, + { + "epoch": 0.1853157894736842, + "grad_norm": 0.78515625, + "learning_rate": 0.00018104093567251463, + "loss": 4.1448, + "step": 3521 + }, + { + "epoch": 0.18536842105263157, + "grad_norm": 0.5625, + "learning_rate": 0.00018102923976608188, + "loss": 4.2058, + "step": 3522 + }, + { + "epoch": 0.18542105263157896, + "grad_norm": 0.609375, + "learning_rate": 0.00018101754385964913, + "loss": 4.2974, + "step": 3523 + }, + { + "epoch": 0.18547368421052632, + "grad_norm": 0.59375, + "learning_rate": 0.00018100584795321638, + "loss": 4.2, + "step": 3524 + }, + { + "epoch": 0.18552631578947368, + "grad_norm": 0.5703125, + "learning_rate": 0.00018099415204678363, + "loss": 4.227, + "step": 3525 + }, + { + "epoch": 0.18557894736842107, + "grad_norm": 0.60546875, + "learning_rate": 0.00018098245614035088, + "loss": 3.9063, + "step": 3526 + }, + { + "epoch": 0.18563157894736843, + "grad_norm": 0.5625, + "learning_rate": 0.00018097076023391816, + "loss": 4.1287, + "step": 3527 + }, + { + "epoch": 0.18568421052631578, + "grad_norm": 0.5625, + "learning_rate": 0.00018095906432748538, + "loss": 4.2813, + "step": 3528 + }, + { + "epoch": 0.18573684210526314, + "grad_norm": 0.625, + "learning_rate": 0.00018094736842105263, + "loss": 4.3243, + "step": 3529 + }, + { + "epoch": 0.18578947368421053, + "grad_norm": 0.60546875, + "learning_rate": 0.00018093567251461988, + "loss": 4.1905, + "step": 3530 + }, + { + "epoch": 0.1858421052631579, + "grad_norm": 0.578125, + "learning_rate": 0.00018092397660818713, + "loss": 4.3892, + "step": 3531 + }, + { + "epoch": 0.18589473684210525, + "grad_norm": 0.5625, + "learning_rate": 0.00018091228070175438, + "loss": 4.0465, + "step": 3532 + }, + { + "epoch": 0.18594736842105264, + "grad_norm": 0.5859375, + "learning_rate": 0.00018090058479532163, + "loss": 4.3219, + "step": 3533 + }, + { + "epoch": 0.186, + "grad_norm": 0.67578125, + "learning_rate": 0.0001808888888888889, + "loss": 4.1631, + "step": 3534 + }, + { + "epoch": 0.18605263157894736, + "grad_norm": 0.53515625, + "learning_rate": 0.00018087719298245616, + "loss": 4.5104, + "step": 3535 + }, + { + "epoch": 0.18610526315789475, + "grad_norm": 0.625, + "learning_rate": 0.0001808654970760234, + "loss": 4.2124, + "step": 3536 + }, + { + "epoch": 0.1861578947368421, + "grad_norm": 0.60546875, + "learning_rate": 0.00018085380116959066, + "loss": 4.1141, + "step": 3537 + }, + { + "epoch": 0.18621052631578947, + "grad_norm": 0.7578125, + "learning_rate": 0.0001808421052631579, + "loss": 3.6059, + "step": 3538 + }, + { + "epoch": 0.18626315789473685, + "grad_norm": 0.5078125, + "learning_rate": 0.00018083040935672515, + "loss": 4.258, + "step": 3539 + }, + { + "epoch": 0.1863157894736842, + "grad_norm": 0.5859375, + "learning_rate": 0.0001808187134502924, + "loss": 3.9485, + "step": 3540 + }, + { + "epoch": 0.18636842105263157, + "grad_norm": 0.5703125, + "learning_rate": 0.00018080701754385965, + "loss": 4.354, + "step": 3541 + }, + { + "epoch": 0.18642105263157896, + "grad_norm": 0.59765625, + "learning_rate": 0.00018079532163742693, + "loss": 4.2654, + "step": 3542 + }, + { + "epoch": 0.18647368421052632, + "grad_norm": 0.83984375, + "learning_rate": 0.00018078362573099415, + "loss": 4.7846, + "step": 3543 + }, + { + "epoch": 0.18652631578947368, + "grad_norm": 0.8125, + "learning_rate": 0.0001807719298245614, + "loss": 4.2412, + "step": 3544 + }, + { + "epoch": 0.18657894736842107, + "grad_norm": 0.6953125, + "learning_rate": 0.00018076023391812865, + "loss": 3.9822, + "step": 3545 + }, + { + "epoch": 0.18663157894736843, + "grad_norm": 0.609375, + "learning_rate": 0.0001807485380116959, + "loss": 4.1944, + "step": 3546 + }, + { + "epoch": 0.18668421052631579, + "grad_norm": 0.5859375, + "learning_rate": 0.00018073684210526315, + "loss": 4.1993, + "step": 3547 + }, + { + "epoch": 0.18673684210526315, + "grad_norm": 0.59765625, + "learning_rate": 0.0001807251461988304, + "loss": 4.3116, + "step": 3548 + }, + { + "epoch": 0.18678947368421053, + "grad_norm": 0.6171875, + "learning_rate": 0.00018071345029239768, + "loss": 3.9556, + "step": 3549 + }, + { + "epoch": 0.1868421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.00018070175438596493, + "loss": 4.3095, + "step": 3550 + }, + { + "epoch": 0.18689473684210525, + "grad_norm": 0.625, + "learning_rate": 0.00018069005847953218, + "loss": 4.2966, + "step": 3551 + }, + { + "epoch": 0.18694736842105264, + "grad_norm": 0.6328125, + "learning_rate": 0.00018067836257309943, + "loss": 4.6382, + "step": 3552 + }, + { + "epoch": 0.187, + "grad_norm": 0.7734375, + "learning_rate": 0.00018066666666666668, + "loss": 4.6306, + "step": 3553 + }, + { + "epoch": 0.18705263157894736, + "grad_norm": 0.76953125, + "learning_rate": 0.00018065497076023393, + "loss": 4.1071, + "step": 3554 + }, + { + "epoch": 0.18710526315789475, + "grad_norm": 0.56640625, + "learning_rate": 0.00018064327485380118, + "loss": 4.1696, + "step": 3555 + }, + { + "epoch": 0.1871578947368421, + "grad_norm": 0.65625, + "learning_rate": 0.00018063157894736845, + "loss": 3.9961, + "step": 3556 + }, + { + "epoch": 0.18721052631578947, + "grad_norm": 0.52734375, + "learning_rate": 0.0001806198830409357, + "loss": 4.1396, + "step": 3557 + }, + { + "epoch": 0.18726315789473685, + "grad_norm": 0.5859375, + "learning_rate": 0.00018060818713450292, + "loss": 4.1229, + "step": 3558 + }, + { + "epoch": 0.1873157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.00018059649122807017, + "loss": 4.101, + "step": 3559 + }, + { + "epoch": 0.18736842105263157, + "grad_norm": 0.6953125, + "learning_rate": 0.00018058479532163742, + "loss": 4.2252, + "step": 3560 + }, + { + "epoch": 0.18742105263157896, + "grad_norm": 0.5625, + "learning_rate": 0.00018057309941520467, + "loss": 3.9533, + "step": 3561 + }, + { + "epoch": 0.18747368421052632, + "grad_norm": 0.609375, + "learning_rate": 0.00018056140350877192, + "loss": 4.4454, + "step": 3562 + }, + { + "epoch": 0.18752631578947368, + "grad_norm": 0.62890625, + "learning_rate": 0.00018054970760233917, + "loss": 4.1032, + "step": 3563 + }, + { + "epoch": 0.18757894736842104, + "grad_norm": 0.60546875, + "learning_rate": 0.00018053801169590645, + "loss": 4.0204, + "step": 3564 + }, + { + "epoch": 0.18763157894736843, + "grad_norm": 0.62890625, + "learning_rate": 0.0001805263157894737, + "loss": 4.0684, + "step": 3565 + }, + { + "epoch": 0.1876842105263158, + "grad_norm": 0.53125, + "learning_rate": 0.00018051461988304095, + "loss": 4.4902, + "step": 3566 + }, + { + "epoch": 0.18773684210526315, + "grad_norm": 0.5625, + "learning_rate": 0.0001805029239766082, + "loss": 4.4984, + "step": 3567 + }, + { + "epoch": 0.18778947368421053, + "grad_norm": 0.5625, + "learning_rate": 0.00018049122807017545, + "loss": 3.7638, + "step": 3568 + }, + { + "epoch": 0.1878421052631579, + "grad_norm": 0.5625, + "learning_rate": 0.0001804795321637427, + "loss": 4.4986, + "step": 3569 + }, + { + "epoch": 0.18789473684210525, + "grad_norm": 0.60546875, + "learning_rate": 0.00018046783625730995, + "loss": 4.2203, + "step": 3570 + }, + { + "epoch": 0.18794736842105264, + "grad_norm": 0.60546875, + "learning_rate": 0.00018045614035087722, + "loss": 3.9373, + "step": 3571 + }, + { + "epoch": 0.188, + "grad_norm": 0.6484375, + "learning_rate": 0.00018044444444444447, + "loss": 4.2919, + "step": 3572 + }, + { + "epoch": 0.18805263157894736, + "grad_norm": 0.5859375, + "learning_rate": 0.00018043274853801172, + "loss": 4.0414, + "step": 3573 + }, + { + "epoch": 0.18810526315789475, + "grad_norm": 0.515625, + "learning_rate": 0.00018042105263157894, + "loss": 4.0844, + "step": 3574 + }, + { + "epoch": 0.1881578947368421, + "grad_norm": 0.62109375, + "learning_rate": 0.0001804093567251462, + "loss": 4.0205, + "step": 3575 + }, + { + "epoch": 0.18821052631578947, + "grad_norm": 0.55859375, + "learning_rate": 0.00018039766081871344, + "loss": 4.0519, + "step": 3576 + }, + { + "epoch": 0.18826315789473685, + "grad_norm": 0.6796875, + "learning_rate": 0.0001803859649122807, + "loss": 3.9864, + "step": 3577 + }, + { + "epoch": 0.18831578947368421, + "grad_norm": 0.69921875, + "learning_rate": 0.00018037426900584794, + "loss": 4.141, + "step": 3578 + }, + { + "epoch": 0.18836842105263157, + "grad_norm": 0.5, + "learning_rate": 0.00018036257309941522, + "loss": 4.0598, + "step": 3579 + }, + { + "epoch": 0.18842105263157893, + "grad_norm": 0.640625, + "learning_rate": 0.00018035087719298247, + "loss": 3.8929, + "step": 3580 + }, + { + "epoch": 0.18847368421052632, + "grad_norm": 0.5859375, + "learning_rate": 0.00018033918128654972, + "loss": 4.5195, + "step": 3581 + }, + { + "epoch": 0.18852631578947368, + "grad_norm": 0.53125, + "learning_rate": 0.00018032748538011697, + "loss": 4.1302, + "step": 3582 + }, + { + "epoch": 0.18857894736842104, + "grad_norm": 0.54296875, + "learning_rate": 0.00018031578947368422, + "loss": 4.344, + "step": 3583 + }, + { + "epoch": 0.18863157894736843, + "grad_norm": 0.58984375, + "learning_rate": 0.00018030409356725147, + "loss": 4.0605, + "step": 3584 + }, + { + "epoch": 0.1886842105263158, + "grad_norm": 0.625, + "learning_rate": 0.00018029239766081872, + "loss": 4.353, + "step": 3585 + }, + { + "epoch": 0.18873684210526315, + "grad_norm": 0.59375, + "learning_rate": 0.000180280701754386, + "loss": 4.4714, + "step": 3586 + }, + { + "epoch": 0.18878947368421053, + "grad_norm": 0.6328125, + "learning_rate": 0.00018026900584795324, + "loss": 4.5286, + "step": 3587 + }, + { + "epoch": 0.1888421052631579, + "grad_norm": 0.66015625, + "learning_rate": 0.0001802573099415205, + "loss": 4.0639, + "step": 3588 + }, + { + "epoch": 0.18889473684210525, + "grad_norm": 0.55859375, + "learning_rate": 0.00018024561403508772, + "loss": 4.4876, + "step": 3589 + }, + { + "epoch": 0.18894736842105264, + "grad_norm": 0.546875, + "learning_rate": 0.00018023391812865497, + "loss": 4.4143, + "step": 3590 + }, + { + "epoch": 0.189, + "grad_norm": 0.5234375, + "learning_rate": 0.00018022222222222221, + "loss": 4.5101, + "step": 3591 + }, + { + "epoch": 0.18905263157894736, + "grad_norm": 0.6875, + "learning_rate": 0.00018021052631578946, + "loss": 4.2279, + "step": 3592 + }, + { + "epoch": 0.18910526315789475, + "grad_norm": 0.71875, + "learning_rate": 0.00018019883040935674, + "loss": 4.3468, + "step": 3593 + }, + { + "epoch": 0.1891578947368421, + "grad_norm": 0.58984375, + "learning_rate": 0.000180187134502924, + "loss": 4.4279, + "step": 3594 + }, + { + "epoch": 0.18921052631578947, + "grad_norm": 0.69140625, + "learning_rate": 0.00018017543859649124, + "loss": 4.381, + "step": 3595 + }, + { + "epoch": 0.18926315789473686, + "grad_norm": 0.68359375, + "learning_rate": 0.0001801637426900585, + "loss": 4.2869, + "step": 3596 + }, + { + "epoch": 0.18931578947368422, + "grad_norm": 0.6796875, + "learning_rate": 0.00018015204678362574, + "loss": 4.316, + "step": 3597 + }, + { + "epoch": 0.18936842105263157, + "grad_norm": 0.52734375, + "learning_rate": 0.000180140350877193, + "loss": 4.3583, + "step": 3598 + }, + { + "epoch": 0.18942105263157893, + "grad_norm": 0.54296875, + "learning_rate": 0.00018012865497076024, + "loss": 4.3787, + "step": 3599 + }, + { + "epoch": 0.18947368421052632, + "grad_norm": 0.6015625, + "learning_rate": 0.0001801169590643275, + "loss": 4.0279, + "step": 3600 + }, + { + "epoch": 0.18952631578947368, + "grad_norm": 0.546875, + "learning_rate": 0.00018010526315789477, + "loss": 4.1935, + "step": 3601 + }, + { + "epoch": 0.18957894736842104, + "grad_norm": 0.59765625, + "learning_rate": 0.00018009356725146201, + "loss": 4.3304, + "step": 3602 + }, + { + "epoch": 0.18963157894736843, + "grad_norm": 0.59765625, + "learning_rate": 0.00018008187134502926, + "loss": 4.1859, + "step": 3603 + }, + { + "epoch": 0.1896842105263158, + "grad_norm": 0.609375, + "learning_rate": 0.0001800701754385965, + "loss": 4.128, + "step": 3604 + }, + { + "epoch": 0.18973684210526315, + "grad_norm": 0.546875, + "learning_rate": 0.00018005847953216374, + "loss": 4.1827, + "step": 3605 + }, + { + "epoch": 0.18978947368421054, + "grad_norm": 0.5703125, + "learning_rate": 0.00018004678362573099, + "loss": 4.3231, + "step": 3606 + }, + { + "epoch": 0.1898421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00018003508771929824, + "loss": 4.6835, + "step": 3607 + }, + { + "epoch": 0.18989473684210526, + "grad_norm": 0.71484375, + "learning_rate": 0.0001800233918128655, + "loss": 4.1395, + "step": 3608 + }, + { + "epoch": 0.18994736842105264, + "grad_norm": 0.5546875, + "learning_rate": 0.00018001169590643276, + "loss": 4.6156, + "step": 3609 + }, + { + "epoch": 0.19, + "grad_norm": 0.56640625, + "learning_rate": 0.00018, + "loss": 4.147, + "step": 3610 + }, + { + "epoch": 0.19005263157894736, + "grad_norm": 0.578125, + "learning_rate": 0.00017998830409356726, + "loss": 4.4005, + "step": 3611 + }, + { + "epoch": 0.19010526315789475, + "grad_norm": 0.60546875, + "learning_rate": 0.0001799766081871345, + "loss": 4.1769, + "step": 3612 + }, + { + "epoch": 0.1901578947368421, + "grad_norm": 0.65625, + "learning_rate": 0.00017996491228070176, + "loss": 4.468, + "step": 3613 + }, + { + "epoch": 0.19021052631578947, + "grad_norm": 0.7265625, + "learning_rate": 0.000179953216374269, + "loss": 4.2177, + "step": 3614 + }, + { + "epoch": 0.19026315789473683, + "grad_norm": 0.62109375, + "learning_rate": 0.00017994152046783626, + "loss": 3.9871, + "step": 3615 + }, + { + "epoch": 0.19031578947368422, + "grad_norm": 0.5859375, + "learning_rate": 0.00017992982456140354, + "loss": 4.4779, + "step": 3616 + }, + { + "epoch": 0.19036842105263158, + "grad_norm": 0.66015625, + "learning_rate": 0.00017991812865497079, + "loss": 4.6341, + "step": 3617 + }, + { + "epoch": 0.19042105263157894, + "grad_norm": 0.6328125, + "learning_rate": 0.00017990643274853804, + "loss": 4.264, + "step": 3618 + }, + { + "epoch": 0.19047368421052632, + "grad_norm": 0.640625, + "learning_rate": 0.00017989473684210528, + "loss": 4.1811, + "step": 3619 + }, + { + "epoch": 0.19052631578947368, + "grad_norm": 0.8203125, + "learning_rate": 0.0001798830409356725, + "loss": 4.0618, + "step": 3620 + }, + { + "epoch": 0.19057894736842104, + "grad_norm": 0.546875, + "learning_rate": 0.00017987134502923976, + "loss": 4.4605, + "step": 3621 + }, + { + "epoch": 0.19063157894736843, + "grad_norm": 0.65625, + "learning_rate": 0.000179859649122807, + "loss": 4.4355, + "step": 3622 + }, + { + "epoch": 0.1906842105263158, + "grad_norm": 0.5859375, + "learning_rate": 0.00017984795321637428, + "loss": 4.5126, + "step": 3623 + }, + { + "epoch": 0.19073684210526315, + "grad_norm": 0.6328125, + "learning_rate": 0.00017983625730994153, + "loss": 4.1148, + "step": 3624 + }, + { + "epoch": 0.19078947368421054, + "grad_norm": 0.59765625, + "learning_rate": 0.00017982456140350878, + "loss": 4.1946, + "step": 3625 + }, + { + "epoch": 0.1908421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.00017981286549707603, + "loss": 4.0581, + "step": 3626 + }, + { + "epoch": 0.19089473684210526, + "grad_norm": 0.54296875, + "learning_rate": 0.00017980116959064328, + "loss": 4.2392, + "step": 3627 + }, + { + "epoch": 0.19094736842105264, + "grad_norm": 0.5859375, + "learning_rate": 0.00017978947368421053, + "loss": 4.4565, + "step": 3628 + }, + { + "epoch": 0.191, + "grad_norm": 0.57421875, + "learning_rate": 0.00017977777777777778, + "loss": 3.9659, + "step": 3629 + }, + { + "epoch": 0.19105263157894736, + "grad_norm": 0.71484375, + "learning_rate": 0.00017976608187134503, + "loss": 4.3325, + "step": 3630 + }, + { + "epoch": 0.19110526315789475, + "grad_norm": 0.60546875, + "learning_rate": 0.0001797543859649123, + "loss": 4.3775, + "step": 3631 + }, + { + "epoch": 0.1911578947368421, + "grad_norm": 0.70703125, + "learning_rate": 0.00017974269005847956, + "loss": 4.0727, + "step": 3632 + }, + { + "epoch": 0.19121052631578947, + "grad_norm": 0.62109375, + "learning_rate": 0.0001797309941520468, + "loss": 4.5187, + "step": 3633 + }, + { + "epoch": 0.19126315789473683, + "grad_norm": 0.5703125, + "learning_rate": 0.00017971929824561406, + "loss": 4.3419, + "step": 3634 + }, + { + "epoch": 0.19131578947368422, + "grad_norm": 0.55859375, + "learning_rate": 0.00017970760233918128, + "loss": 4.4819, + "step": 3635 + }, + { + "epoch": 0.19136842105263158, + "grad_norm": 0.5703125, + "learning_rate": 0.00017969590643274853, + "loss": 4.0626, + "step": 3636 + }, + { + "epoch": 0.19142105263157894, + "grad_norm": 0.625, + "learning_rate": 0.00017968421052631578, + "loss": 4.2783, + "step": 3637 + }, + { + "epoch": 0.19147368421052632, + "grad_norm": 0.609375, + "learning_rate": 0.00017967251461988305, + "loss": 4.3584, + "step": 3638 + }, + { + "epoch": 0.19152631578947368, + "grad_norm": 0.68359375, + "learning_rate": 0.0001796608187134503, + "loss": 3.8956, + "step": 3639 + }, + { + "epoch": 0.19157894736842104, + "grad_norm": 0.55859375, + "learning_rate": 0.00017964912280701755, + "loss": 4.0336, + "step": 3640 + }, + { + "epoch": 0.19163157894736843, + "grad_norm": 0.546875, + "learning_rate": 0.0001796374269005848, + "loss": 4.1139, + "step": 3641 + }, + { + "epoch": 0.1916842105263158, + "grad_norm": 0.71875, + "learning_rate": 0.00017962573099415205, + "loss": 4.2294, + "step": 3642 + }, + { + "epoch": 0.19173684210526315, + "grad_norm": 0.69140625, + "learning_rate": 0.0001796140350877193, + "loss": 3.8433, + "step": 3643 + }, + { + "epoch": 0.19178947368421054, + "grad_norm": 0.58984375, + "learning_rate": 0.00017960233918128655, + "loss": 4.0442, + "step": 3644 + }, + { + "epoch": 0.1918421052631579, + "grad_norm": 0.734375, + "learning_rate": 0.00017959064327485383, + "loss": 4.278, + "step": 3645 + }, + { + "epoch": 0.19189473684210526, + "grad_norm": 0.625, + "learning_rate": 0.00017957894736842108, + "loss": 4.0739, + "step": 3646 + }, + { + "epoch": 0.19194736842105264, + "grad_norm": 0.65625, + "learning_rate": 0.00017956725146198833, + "loss": 3.7219, + "step": 3647 + }, + { + "epoch": 0.192, + "grad_norm": 0.63671875, + "learning_rate": 0.00017955555555555558, + "loss": 4.172, + "step": 3648 + }, + { + "epoch": 0.19205263157894736, + "grad_norm": 0.546875, + "learning_rate": 0.00017954385964912283, + "loss": 4.3963, + "step": 3649 + }, + { + "epoch": 0.19210526315789472, + "grad_norm": 0.53125, + "learning_rate": 0.00017953216374269005, + "loss": 4.6039, + "step": 3650 + }, + { + "epoch": 0.1921578947368421, + "grad_norm": 0.6328125, + "learning_rate": 0.0001795204678362573, + "loss": 4.2272, + "step": 3651 + }, + { + "epoch": 0.19221052631578947, + "grad_norm": 0.5625, + "learning_rate": 0.00017950877192982455, + "loss": 4.3411, + "step": 3652 + }, + { + "epoch": 0.19226315789473683, + "grad_norm": 0.6171875, + "learning_rate": 0.00017949707602339183, + "loss": 3.6401, + "step": 3653 + }, + { + "epoch": 0.19231578947368422, + "grad_norm": 0.515625, + "learning_rate": 0.00017948538011695907, + "loss": 4.4033, + "step": 3654 + }, + { + "epoch": 0.19236842105263158, + "grad_norm": 0.66015625, + "learning_rate": 0.00017947368421052632, + "loss": 3.9067, + "step": 3655 + }, + { + "epoch": 0.19242105263157894, + "grad_norm": 0.59375, + "learning_rate": 0.00017946198830409357, + "loss": 4.1122, + "step": 3656 + }, + { + "epoch": 0.19247368421052632, + "grad_norm": 0.61328125, + "learning_rate": 0.00017945029239766082, + "loss": 4.4353, + "step": 3657 + }, + { + "epoch": 0.19252631578947368, + "grad_norm": 0.54296875, + "learning_rate": 0.00017943859649122807, + "loss": 4.6124, + "step": 3658 + }, + { + "epoch": 0.19257894736842104, + "grad_norm": 0.59375, + "learning_rate": 0.00017942690058479532, + "loss": 4.4246, + "step": 3659 + }, + { + "epoch": 0.19263157894736843, + "grad_norm": 0.59375, + "learning_rate": 0.0001794152046783626, + "loss": 4.3475, + "step": 3660 + }, + { + "epoch": 0.1926842105263158, + "grad_norm": 0.82421875, + "learning_rate": 0.00017940350877192985, + "loss": 4.1154, + "step": 3661 + }, + { + "epoch": 0.19273684210526315, + "grad_norm": 0.671875, + "learning_rate": 0.0001793918128654971, + "loss": 3.7341, + "step": 3662 + }, + { + "epoch": 0.19278947368421054, + "grad_norm": 0.53515625, + "learning_rate": 0.00017938011695906435, + "loss": 4.3342, + "step": 3663 + }, + { + "epoch": 0.1928421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.0001793684210526316, + "loss": 3.9139, + "step": 3664 + }, + { + "epoch": 0.19289473684210526, + "grad_norm": 0.498046875, + "learning_rate": 0.00017935672514619885, + "loss": 4.1141, + "step": 3665 + }, + { + "epoch": 0.19294736842105265, + "grad_norm": 0.625, + "learning_rate": 0.00017934502923976607, + "loss": 3.9852, + "step": 3666 + }, + { + "epoch": 0.193, + "grad_norm": 0.62890625, + "learning_rate": 0.00017933333333333332, + "loss": 4.0002, + "step": 3667 + }, + { + "epoch": 0.19305263157894736, + "grad_norm": 0.5, + "learning_rate": 0.0001793216374269006, + "loss": 4.4017, + "step": 3668 + }, + { + "epoch": 0.19310526315789472, + "grad_norm": 0.609375, + "learning_rate": 0.00017930994152046785, + "loss": 4.3318, + "step": 3669 + }, + { + "epoch": 0.1931578947368421, + "grad_norm": 0.71875, + "learning_rate": 0.0001792982456140351, + "loss": 3.7896, + "step": 3670 + }, + { + "epoch": 0.19321052631578947, + "grad_norm": 0.5625, + "learning_rate": 0.00017928654970760234, + "loss": 4.0221, + "step": 3671 + }, + { + "epoch": 0.19326315789473683, + "grad_norm": 0.53515625, + "learning_rate": 0.0001792748538011696, + "loss": 3.8382, + "step": 3672 + }, + { + "epoch": 0.19331578947368422, + "grad_norm": 0.75390625, + "learning_rate": 0.00017926315789473684, + "loss": 4.2087, + "step": 3673 + }, + { + "epoch": 0.19336842105263158, + "grad_norm": 0.62890625, + "learning_rate": 0.0001792514619883041, + "loss": 4.297, + "step": 3674 + }, + { + "epoch": 0.19342105263157894, + "grad_norm": 0.6171875, + "learning_rate": 0.00017923976608187137, + "loss": 4.166, + "step": 3675 + }, + { + "epoch": 0.19347368421052633, + "grad_norm": 0.66015625, + "learning_rate": 0.00017922807017543862, + "loss": 4.4926, + "step": 3676 + }, + { + "epoch": 0.19352631578947369, + "grad_norm": 0.5390625, + "learning_rate": 0.00017921637426900587, + "loss": 4.5617, + "step": 3677 + }, + { + "epoch": 0.19357894736842104, + "grad_norm": 0.5234375, + "learning_rate": 0.00017920467836257312, + "loss": 4.4014, + "step": 3678 + }, + { + "epoch": 0.19363157894736843, + "grad_norm": 0.6328125, + "learning_rate": 0.00017919298245614037, + "loss": 3.9847, + "step": 3679 + }, + { + "epoch": 0.1936842105263158, + "grad_norm": 0.5859375, + "learning_rate": 0.00017918128654970762, + "loss": 4.6326, + "step": 3680 + }, + { + "epoch": 0.19373684210526315, + "grad_norm": 0.625, + "learning_rate": 0.00017916959064327484, + "loss": 4.3388, + "step": 3681 + }, + { + "epoch": 0.19378947368421054, + "grad_norm": 0.5859375, + "learning_rate": 0.00017915789473684212, + "loss": 4.0231, + "step": 3682 + }, + { + "epoch": 0.1938421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00017914619883040937, + "loss": 4.2666, + "step": 3683 + }, + { + "epoch": 0.19389473684210526, + "grad_norm": 0.74609375, + "learning_rate": 0.00017913450292397662, + "loss": 4.344, + "step": 3684 + }, + { + "epoch": 0.19394736842105262, + "grad_norm": 0.7890625, + "learning_rate": 0.00017912280701754387, + "loss": 4.1754, + "step": 3685 + }, + { + "epoch": 0.194, + "grad_norm": 0.81640625, + "learning_rate": 0.00017911111111111112, + "loss": 4.0002, + "step": 3686 + }, + { + "epoch": 0.19405263157894737, + "grad_norm": 0.4921875, + "learning_rate": 0.00017909941520467837, + "loss": 4.1602, + "step": 3687 + }, + { + "epoch": 0.19410526315789473, + "grad_norm": 0.55859375, + "learning_rate": 0.00017908771929824562, + "loss": 3.7285, + "step": 3688 + }, + { + "epoch": 0.1941578947368421, + "grad_norm": 0.6171875, + "learning_rate": 0.00017907602339181286, + "loss": 4.0462, + "step": 3689 + }, + { + "epoch": 0.19421052631578947, + "grad_norm": 0.56640625, + "learning_rate": 0.00017906432748538014, + "loss": 4.0057, + "step": 3690 + }, + { + "epoch": 0.19426315789473683, + "grad_norm": 0.6171875, + "learning_rate": 0.0001790526315789474, + "loss": 3.8107, + "step": 3691 + }, + { + "epoch": 0.19431578947368422, + "grad_norm": 0.546875, + "learning_rate": 0.00017904093567251464, + "loss": 4.1497, + "step": 3692 + }, + { + "epoch": 0.19436842105263158, + "grad_norm": 0.578125, + "learning_rate": 0.0001790292397660819, + "loss": 4.1618, + "step": 3693 + }, + { + "epoch": 0.19442105263157894, + "grad_norm": 0.60546875, + "learning_rate": 0.00017901754385964914, + "loss": 4.2034, + "step": 3694 + }, + { + "epoch": 0.19447368421052633, + "grad_norm": 0.5390625, + "learning_rate": 0.0001790058479532164, + "loss": 3.8892, + "step": 3695 + }, + { + "epoch": 0.19452631578947369, + "grad_norm": 0.55859375, + "learning_rate": 0.0001789941520467836, + "loss": 4.0073, + "step": 3696 + }, + { + "epoch": 0.19457894736842105, + "grad_norm": 0.5625, + "learning_rate": 0.0001789824561403509, + "loss": 3.949, + "step": 3697 + }, + { + "epoch": 0.19463157894736843, + "grad_norm": 0.60546875, + "learning_rate": 0.00017897076023391814, + "loss": 4.098, + "step": 3698 + }, + { + "epoch": 0.1946842105263158, + "grad_norm": 0.5703125, + "learning_rate": 0.0001789590643274854, + "loss": 4.259, + "step": 3699 + }, + { + "epoch": 0.19473684210526315, + "grad_norm": 0.5703125, + "learning_rate": 0.00017894736842105264, + "loss": 4.1725, + "step": 3700 + }, + { + "epoch": 0.19478947368421054, + "grad_norm": 0.5703125, + "learning_rate": 0.0001789356725146199, + "loss": 4.3684, + "step": 3701 + }, + { + "epoch": 0.1948421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00017892397660818714, + "loss": 4.5872, + "step": 3702 + }, + { + "epoch": 0.19489473684210526, + "grad_norm": 0.546875, + "learning_rate": 0.00017891228070175439, + "loss": 3.8087, + "step": 3703 + }, + { + "epoch": 0.19494736842105262, + "grad_norm": 0.54296875, + "learning_rate": 0.00017890058479532164, + "loss": 3.8881, + "step": 3704 + }, + { + "epoch": 0.195, + "grad_norm": 0.5703125, + "learning_rate": 0.0001788888888888889, + "loss": 4.257, + "step": 3705 + }, + { + "epoch": 0.19505263157894737, + "grad_norm": 0.5625, + "learning_rate": 0.00017887719298245616, + "loss": 4.2867, + "step": 3706 + }, + { + "epoch": 0.19510526315789473, + "grad_norm": 0.62890625, + "learning_rate": 0.0001788654970760234, + "loss": 3.8501, + "step": 3707 + }, + { + "epoch": 0.1951578947368421, + "grad_norm": 0.62890625, + "learning_rate": 0.00017885380116959066, + "loss": 4.3072, + "step": 3708 + }, + { + "epoch": 0.19521052631578947, + "grad_norm": 0.6640625, + "learning_rate": 0.0001788421052631579, + "loss": 4.3747, + "step": 3709 + }, + { + "epoch": 0.19526315789473683, + "grad_norm": 0.75, + "learning_rate": 0.00017883040935672516, + "loss": 4.1571, + "step": 3710 + }, + { + "epoch": 0.19531578947368422, + "grad_norm": 0.56640625, + "learning_rate": 0.0001788187134502924, + "loss": 4.1263, + "step": 3711 + }, + { + "epoch": 0.19536842105263158, + "grad_norm": 0.5625, + "learning_rate": 0.00017880701754385966, + "loss": 4.0662, + "step": 3712 + }, + { + "epoch": 0.19542105263157894, + "grad_norm": 0.54296875, + "learning_rate": 0.0001787953216374269, + "loss": 4.3722, + "step": 3713 + }, + { + "epoch": 0.19547368421052633, + "grad_norm": 0.55078125, + "learning_rate": 0.00017878362573099416, + "loss": 4.2897, + "step": 3714 + }, + { + "epoch": 0.1955263157894737, + "grad_norm": 0.65625, + "learning_rate": 0.0001787719298245614, + "loss": 4.0004, + "step": 3715 + }, + { + "epoch": 0.19557894736842105, + "grad_norm": 1.96875, + "learning_rate": 0.00017876023391812866, + "loss": 4.0105, + "step": 3716 + }, + { + "epoch": 0.19563157894736843, + "grad_norm": 0.7265625, + "learning_rate": 0.0001787485380116959, + "loss": 4.1982, + "step": 3717 + }, + { + "epoch": 0.1956842105263158, + "grad_norm": 0.6640625, + "learning_rate": 0.00017873684210526316, + "loss": 4.1796, + "step": 3718 + }, + { + "epoch": 0.19573684210526315, + "grad_norm": 0.625, + "learning_rate": 0.00017872514619883043, + "loss": 4.617, + "step": 3719 + }, + { + "epoch": 0.1957894736842105, + "grad_norm": 0.64453125, + "learning_rate": 0.00017871345029239768, + "loss": 4.4054, + "step": 3720 + }, + { + "epoch": 0.1958421052631579, + "grad_norm": 0.70703125, + "learning_rate": 0.00017870175438596493, + "loss": 4.2627, + "step": 3721 + }, + { + "epoch": 0.19589473684210526, + "grad_norm": 0.578125, + "learning_rate": 0.00017869005847953218, + "loss": 4.0648, + "step": 3722 + }, + { + "epoch": 0.19594736842105262, + "grad_norm": 0.51953125, + "learning_rate": 0.00017867836257309943, + "loss": 4.2066, + "step": 3723 + }, + { + "epoch": 0.196, + "grad_norm": 0.62890625, + "learning_rate": 0.00017866666666666668, + "loss": 4.3194, + "step": 3724 + }, + { + "epoch": 0.19605263157894737, + "grad_norm": 0.5625, + "learning_rate": 0.00017865497076023393, + "loss": 4.2704, + "step": 3725 + }, + { + "epoch": 0.19610526315789473, + "grad_norm": 0.60546875, + "learning_rate": 0.00017864327485380118, + "loss": 4.1926, + "step": 3726 + }, + { + "epoch": 0.19615789473684211, + "grad_norm": 0.55078125, + "learning_rate": 0.00017863157894736843, + "loss": 4.0293, + "step": 3727 + }, + { + "epoch": 0.19621052631578947, + "grad_norm": 0.625, + "learning_rate": 0.00017861988304093568, + "loss": 4.2465, + "step": 3728 + }, + { + "epoch": 0.19626315789473683, + "grad_norm": 0.640625, + "learning_rate": 0.00017860818713450293, + "loss": 4.208, + "step": 3729 + }, + { + "epoch": 0.19631578947368422, + "grad_norm": 0.63671875, + "learning_rate": 0.00017859649122807018, + "loss": 4.4006, + "step": 3730 + }, + { + "epoch": 0.19636842105263158, + "grad_norm": 0.7265625, + "learning_rate": 0.00017858479532163743, + "loss": 4.2094, + "step": 3731 + }, + { + "epoch": 0.19642105263157894, + "grad_norm": 0.62890625, + "learning_rate": 0.00017857309941520468, + "loss": 4.2234, + "step": 3732 + }, + { + "epoch": 0.19647368421052633, + "grad_norm": 0.625, + "learning_rate": 0.00017856140350877193, + "loss": 4.1221, + "step": 3733 + }, + { + "epoch": 0.1965263157894737, + "grad_norm": 0.62890625, + "learning_rate": 0.0001785497076023392, + "loss": 4.2741, + "step": 3734 + }, + { + "epoch": 0.19657894736842105, + "grad_norm": 0.65234375, + "learning_rate": 0.00017853801169590645, + "loss": 4.3779, + "step": 3735 + }, + { + "epoch": 0.1966315789473684, + "grad_norm": 0.828125, + "learning_rate": 0.0001785263157894737, + "loss": 4.082, + "step": 3736 + }, + { + "epoch": 0.1966842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.00017851461988304095, + "loss": 4.3126, + "step": 3737 + }, + { + "epoch": 0.19673684210526315, + "grad_norm": 0.54296875, + "learning_rate": 0.0001785029239766082, + "loss": 3.9585, + "step": 3738 + }, + { + "epoch": 0.19678947368421051, + "grad_norm": 0.65625, + "learning_rate": 0.00017849122807017545, + "loss": 4.3214, + "step": 3739 + }, + { + "epoch": 0.1968421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.0001784795321637427, + "loss": 4.0866, + "step": 3740 + }, + { + "epoch": 0.19689473684210526, + "grad_norm": 0.58984375, + "learning_rate": 0.00017846783625730995, + "loss": 3.7803, + "step": 3741 + }, + { + "epoch": 0.19694736842105262, + "grad_norm": 0.67578125, + "learning_rate": 0.0001784561403508772, + "loss": 3.751, + "step": 3742 + }, + { + "epoch": 0.197, + "grad_norm": 0.490234375, + "learning_rate": 0.00017844444444444445, + "loss": 4.2435, + "step": 3743 + }, + { + "epoch": 0.19705263157894737, + "grad_norm": 0.6640625, + "learning_rate": 0.0001784327485380117, + "loss": 3.9828, + "step": 3744 + }, + { + "epoch": 0.19710526315789473, + "grad_norm": 0.74609375, + "learning_rate": 0.00017842105263157895, + "loss": 3.949, + "step": 3745 + }, + { + "epoch": 0.19715789473684212, + "grad_norm": 0.7109375, + "learning_rate": 0.0001784093567251462, + "loss": 3.729, + "step": 3746 + }, + { + "epoch": 0.19721052631578948, + "grad_norm": 0.51953125, + "learning_rate": 0.00017839766081871345, + "loss": 3.9338, + "step": 3747 + }, + { + "epoch": 0.19726315789473683, + "grad_norm": 0.51171875, + "learning_rate": 0.0001783859649122807, + "loss": 4.0959, + "step": 3748 + }, + { + "epoch": 0.19731578947368422, + "grad_norm": 0.4921875, + "learning_rate": 0.00017837426900584798, + "loss": 4.1586, + "step": 3749 + }, + { + "epoch": 0.19736842105263158, + "grad_norm": 0.58203125, + "learning_rate": 0.00017836257309941523, + "loss": 4.6796, + "step": 3750 + }, + { + "epoch": 0.19742105263157894, + "grad_norm": 0.58203125, + "learning_rate": 0.00017835087719298247, + "loss": 4.4961, + "step": 3751 + }, + { + "epoch": 0.19747368421052633, + "grad_norm": 0.65625, + "learning_rate": 0.00017833918128654972, + "loss": 4.3747, + "step": 3752 + }, + { + "epoch": 0.1975263157894737, + "grad_norm": 0.64453125, + "learning_rate": 0.00017832748538011697, + "loss": 3.7726, + "step": 3753 + }, + { + "epoch": 0.19757894736842105, + "grad_norm": 0.59375, + "learning_rate": 0.00017831578947368422, + "loss": 4.0057, + "step": 3754 + }, + { + "epoch": 0.1976315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.00017830409356725147, + "loss": 4.3845, + "step": 3755 + }, + { + "epoch": 0.1976842105263158, + "grad_norm": 0.578125, + "learning_rate": 0.00017829239766081872, + "loss": 4.0559, + "step": 3756 + }, + { + "epoch": 0.19773684210526316, + "grad_norm": 0.5390625, + "learning_rate": 0.00017828070175438597, + "loss": 3.7287, + "step": 3757 + }, + { + "epoch": 0.19778947368421052, + "grad_norm": 0.6015625, + "learning_rate": 0.00017826900584795322, + "loss": 4.0163, + "step": 3758 + }, + { + "epoch": 0.1978421052631579, + "grad_norm": 0.4765625, + "learning_rate": 0.00017825730994152047, + "loss": 4.4255, + "step": 3759 + }, + { + "epoch": 0.19789473684210526, + "grad_norm": 0.57421875, + "learning_rate": 0.00017824561403508772, + "loss": 4.2944, + "step": 3760 + }, + { + "epoch": 0.19794736842105262, + "grad_norm": 0.80078125, + "learning_rate": 0.00017823391812865497, + "loss": 3.8442, + "step": 3761 + }, + { + "epoch": 0.198, + "grad_norm": 0.7734375, + "learning_rate": 0.00017822222222222222, + "loss": 3.8099, + "step": 3762 + }, + { + "epoch": 0.19805263157894737, + "grad_norm": 0.57421875, + "learning_rate": 0.00017821052631578947, + "loss": 4.0282, + "step": 3763 + }, + { + "epoch": 0.19810526315789473, + "grad_norm": 0.53515625, + "learning_rate": 0.00017819883040935675, + "loss": 4.1583, + "step": 3764 + }, + { + "epoch": 0.19815789473684212, + "grad_norm": 0.5390625, + "learning_rate": 0.000178187134502924, + "loss": 4.4665, + "step": 3765 + }, + { + "epoch": 0.19821052631578948, + "grad_norm": 0.6171875, + "learning_rate": 0.00017817543859649125, + "loss": 3.8485, + "step": 3766 + }, + { + "epoch": 0.19826315789473684, + "grad_norm": 0.51953125, + "learning_rate": 0.0001781637426900585, + "loss": 4.073, + "step": 3767 + }, + { + "epoch": 0.19831578947368422, + "grad_norm": 0.67578125, + "learning_rate": 0.00017815204678362575, + "loss": 4.264, + "step": 3768 + }, + { + "epoch": 0.19836842105263158, + "grad_norm": 0.490234375, + "learning_rate": 0.000178140350877193, + "loss": 4.2266, + "step": 3769 + }, + { + "epoch": 0.19842105263157894, + "grad_norm": 0.74609375, + "learning_rate": 0.00017812865497076024, + "loss": 4.0507, + "step": 3770 + }, + { + "epoch": 0.1984736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.0001781169590643275, + "loss": 4.358, + "step": 3771 + }, + { + "epoch": 0.1985263157894737, + "grad_norm": 0.53515625, + "learning_rate": 0.00017810526315789474, + "loss": 4.0994, + "step": 3772 + }, + { + "epoch": 0.19857894736842105, + "grad_norm": 0.64453125, + "learning_rate": 0.000178093567251462, + "loss": 3.6683, + "step": 3773 + }, + { + "epoch": 0.1986315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.00017808187134502924, + "loss": 4.3948, + "step": 3774 + }, + { + "epoch": 0.1986842105263158, + "grad_norm": 0.63671875, + "learning_rate": 0.0001780701754385965, + "loss": 4.0698, + "step": 3775 + }, + { + "epoch": 0.19873684210526316, + "grad_norm": 0.515625, + "learning_rate": 0.00017805847953216374, + "loss": 4.3094, + "step": 3776 + }, + { + "epoch": 0.19878947368421052, + "grad_norm": 0.52734375, + "learning_rate": 0.000178046783625731, + "loss": 4.5311, + "step": 3777 + }, + { + "epoch": 0.1988421052631579, + "grad_norm": 0.58984375, + "learning_rate": 0.00017803508771929824, + "loss": 4.174, + "step": 3778 + }, + { + "epoch": 0.19889473684210526, + "grad_norm": 0.625, + "learning_rate": 0.00017802339181286552, + "loss": 4.5208, + "step": 3779 + }, + { + "epoch": 0.19894736842105262, + "grad_norm": 0.5859375, + "learning_rate": 0.00017801169590643277, + "loss": 4.0142, + "step": 3780 + }, + { + "epoch": 0.199, + "grad_norm": 0.6875, + "learning_rate": 0.00017800000000000002, + "loss": 4.4815, + "step": 3781 + }, + { + "epoch": 0.19905263157894737, + "grad_norm": 0.609375, + "learning_rate": 0.00017798830409356727, + "loss": 3.9563, + "step": 3782 + }, + { + "epoch": 0.19910526315789473, + "grad_norm": 0.58203125, + "learning_rate": 0.00017797660818713452, + "loss": 3.9757, + "step": 3783 + }, + { + "epoch": 0.19915789473684212, + "grad_norm": 0.52734375, + "learning_rate": 0.00017796491228070177, + "loss": 4.0844, + "step": 3784 + }, + { + "epoch": 0.19921052631578948, + "grad_norm": 0.5859375, + "learning_rate": 0.00017795321637426902, + "loss": 4.1302, + "step": 3785 + }, + { + "epoch": 0.19926315789473684, + "grad_norm": 0.7421875, + "learning_rate": 0.00017794152046783626, + "loss": 4.1537, + "step": 3786 + }, + { + "epoch": 0.19931578947368422, + "grad_norm": 0.5234375, + "learning_rate": 0.00017792982456140351, + "loss": 4.2175, + "step": 3787 + }, + { + "epoch": 0.19936842105263158, + "grad_norm": 0.60546875, + "learning_rate": 0.00017791812865497076, + "loss": 4.0349, + "step": 3788 + }, + { + "epoch": 0.19942105263157894, + "grad_norm": 0.6171875, + "learning_rate": 0.00017790643274853801, + "loss": 4.0334, + "step": 3789 + }, + { + "epoch": 0.1994736842105263, + "grad_norm": 0.5625, + "learning_rate": 0.00017789473684210526, + "loss": 4.2964, + "step": 3790 + }, + { + "epoch": 0.1995263157894737, + "grad_norm": 0.62890625, + "learning_rate": 0.0001778830409356725, + "loss": 4.1697, + "step": 3791 + }, + { + "epoch": 0.19957894736842105, + "grad_norm": 0.63671875, + "learning_rate": 0.00017787134502923976, + "loss": 4.029, + "step": 3792 + }, + { + "epoch": 0.1996315789473684, + "grad_norm": 0.52734375, + "learning_rate": 0.000177859649122807, + "loss": 4.4606, + "step": 3793 + }, + { + "epoch": 0.1996842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.0001778479532163743, + "loss": 4.2194, + "step": 3794 + }, + { + "epoch": 0.19973684210526316, + "grad_norm": 0.60546875, + "learning_rate": 0.00017783625730994154, + "loss": 4.1256, + "step": 3795 + }, + { + "epoch": 0.19978947368421052, + "grad_norm": 0.578125, + "learning_rate": 0.0001778245614035088, + "loss": 4.2334, + "step": 3796 + }, + { + "epoch": 0.1998421052631579, + "grad_norm": 0.6796875, + "learning_rate": 0.00017781286549707604, + "loss": 4.123, + "step": 3797 + }, + { + "epoch": 0.19989473684210526, + "grad_norm": 0.65234375, + "learning_rate": 0.0001778011695906433, + "loss": 4.2962, + "step": 3798 + }, + { + "epoch": 0.19994736842105262, + "grad_norm": 0.60546875, + "learning_rate": 0.00017778947368421054, + "loss": 4.5104, + "step": 3799 + }, + { + "epoch": 0.2, + "grad_norm": 0.8671875, + "learning_rate": 0.00017777777777777779, + "loss": 4.2254, + "step": 3800 + }, + { + "epoch": 0.20005263157894737, + "grad_norm": 0.57421875, + "learning_rate": 0.00017776608187134504, + "loss": 3.7591, + "step": 3801 + }, + { + "epoch": 0.20010526315789473, + "grad_norm": 0.55078125, + "learning_rate": 0.00017775438596491229, + "loss": 3.9366, + "step": 3802 + }, + { + "epoch": 0.20015789473684212, + "grad_norm": 0.482421875, + "learning_rate": 0.00017774269005847954, + "loss": 4.1937, + "step": 3803 + }, + { + "epoch": 0.20021052631578948, + "grad_norm": 0.4921875, + "learning_rate": 0.00017773099415204678, + "loss": 4.3724, + "step": 3804 + }, + { + "epoch": 0.20026315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.00017771929824561403, + "loss": 4.082, + "step": 3805 + }, + { + "epoch": 0.2003157894736842, + "grad_norm": 0.58984375, + "learning_rate": 0.00017770760233918128, + "loss": 4.0364, + "step": 3806 + }, + { + "epoch": 0.20036842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00017769590643274853, + "loss": 4.1414, + "step": 3807 + }, + { + "epoch": 0.20042105263157894, + "grad_norm": 0.51171875, + "learning_rate": 0.0001776842105263158, + "loss": 4.1304, + "step": 3808 + }, + { + "epoch": 0.2004736842105263, + "grad_norm": 0.6171875, + "learning_rate": 0.00017767251461988306, + "loss": 4.513, + "step": 3809 + }, + { + "epoch": 0.2005263157894737, + "grad_norm": 0.5859375, + "learning_rate": 0.0001776608187134503, + "loss": 3.9869, + "step": 3810 + }, + { + "epoch": 0.20057894736842105, + "grad_norm": 0.578125, + "learning_rate": 0.00017764912280701756, + "loss": 3.9913, + "step": 3811 + }, + { + "epoch": 0.2006315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.0001776374269005848, + "loss": 4.3184, + "step": 3812 + }, + { + "epoch": 0.2006842105263158, + "grad_norm": 0.61328125, + "learning_rate": 0.00017762573099415206, + "loss": 4.2315, + "step": 3813 + }, + { + "epoch": 0.20073684210526316, + "grad_norm": 0.6015625, + "learning_rate": 0.0001776140350877193, + "loss": 4.3139, + "step": 3814 + }, + { + "epoch": 0.20078947368421052, + "grad_norm": 0.5859375, + "learning_rate": 0.00017760233918128656, + "loss": 3.747, + "step": 3815 + }, + { + "epoch": 0.2008421052631579, + "grad_norm": 0.6484375, + "learning_rate": 0.0001775906432748538, + "loss": 3.7621, + "step": 3816 + }, + { + "epoch": 0.20089473684210526, + "grad_norm": 0.6328125, + "learning_rate": 0.00017757894736842106, + "loss": 3.8096, + "step": 3817 + }, + { + "epoch": 0.20094736842105262, + "grad_norm": 0.6171875, + "learning_rate": 0.0001775672514619883, + "loss": 3.6032, + "step": 3818 + }, + { + "epoch": 0.201, + "grad_norm": 0.50390625, + "learning_rate": 0.00017755555555555556, + "loss": 4.2838, + "step": 3819 + }, + { + "epoch": 0.20105263157894737, + "grad_norm": 0.765625, + "learning_rate": 0.0001775438596491228, + "loss": 3.7754, + "step": 3820 + }, + { + "epoch": 0.20110526315789473, + "grad_norm": 0.56640625, + "learning_rate": 0.00017753216374269005, + "loss": 4.2805, + "step": 3821 + }, + { + "epoch": 0.20115789473684212, + "grad_norm": 0.57421875, + "learning_rate": 0.0001775204678362573, + "loss": 3.5718, + "step": 3822 + }, + { + "epoch": 0.20121052631578948, + "grad_norm": 0.53515625, + "learning_rate": 0.00017750877192982458, + "loss": 4.2724, + "step": 3823 + }, + { + "epoch": 0.20126315789473684, + "grad_norm": 0.58984375, + "learning_rate": 0.00017749707602339183, + "loss": 4.1648, + "step": 3824 + }, + { + "epoch": 0.2013157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00017748538011695908, + "loss": 4.2831, + "step": 3825 + }, + { + "epoch": 0.20136842105263159, + "grad_norm": 0.5859375, + "learning_rate": 0.00017747368421052633, + "loss": 4.1799, + "step": 3826 + }, + { + "epoch": 0.20142105263157895, + "grad_norm": 0.63671875, + "learning_rate": 0.00017746198830409358, + "loss": 4.5345, + "step": 3827 + }, + { + "epoch": 0.2014736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.00017745029239766083, + "loss": 4.5241, + "step": 3828 + }, + { + "epoch": 0.2015263157894737, + "grad_norm": 0.61328125, + "learning_rate": 0.00017743859649122808, + "loss": 4.2353, + "step": 3829 + }, + { + "epoch": 0.20157894736842105, + "grad_norm": 0.55078125, + "learning_rate": 0.00017742690058479533, + "loss": 4.3277, + "step": 3830 + }, + { + "epoch": 0.2016315789473684, + "grad_norm": 0.7109375, + "learning_rate": 0.00017741520467836258, + "loss": 4.0349, + "step": 3831 + }, + { + "epoch": 0.2016842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.00017740350877192983, + "loss": 3.6097, + "step": 3832 + }, + { + "epoch": 0.20173684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00017739181286549708, + "loss": 4.1299, + "step": 3833 + }, + { + "epoch": 0.20178947368421052, + "grad_norm": 0.7421875, + "learning_rate": 0.00017738011695906433, + "loss": 4.2182, + "step": 3834 + }, + { + "epoch": 0.2018421052631579, + "grad_norm": 0.640625, + "learning_rate": 0.00017736842105263158, + "loss": 4.0145, + "step": 3835 + }, + { + "epoch": 0.20189473684210527, + "grad_norm": 0.703125, + "learning_rate": 0.00017735672514619883, + "loss": 4.1395, + "step": 3836 + }, + { + "epoch": 0.20194736842105263, + "grad_norm": 0.69140625, + "learning_rate": 0.00017734502923976608, + "loss": 4.1016, + "step": 3837 + }, + { + "epoch": 0.202, + "grad_norm": 0.6484375, + "learning_rate": 0.00017733333333333335, + "loss": 3.8633, + "step": 3838 + }, + { + "epoch": 0.20205263157894737, + "grad_norm": 0.5703125, + "learning_rate": 0.0001773216374269006, + "loss": 4.1836, + "step": 3839 + }, + { + "epoch": 0.20210526315789473, + "grad_norm": 0.53125, + "learning_rate": 0.00017730994152046785, + "loss": 3.9621, + "step": 3840 + }, + { + "epoch": 0.2021578947368421, + "grad_norm": 0.66796875, + "learning_rate": 0.0001772982456140351, + "loss": 4.0794, + "step": 3841 + }, + { + "epoch": 0.20221052631578948, + "grad_norm": 0.4609375, + "learning_rate": 0.00017728654970760235, + "loss": 4.4131, + "step": 3842 + }, + { + "epoch": 0.20226315789473684, + "grad_norm": 0.609375, + "learning_rate": 0.0001772748538011696, + "loss": 4.4914, + "step": 3843 + }, + { + "epoch": 0.2023157894736842, + "grad_norm": 0.6015625, + "learning_rate": 0.00017726315789473685, + "loss": 4.1304, + "step": 3844 + }, + { + "epoch": 0.2023684210526316, + "grad_norm": 0.59375, + "learning_rate": 0.0001772514619883041, + "loss": 4.0569, + "step": 3845 + }, + { + "epoch": 0.20242105263157895, + "grad_norm": 0.67578125, + "learning_rate": 0.00017723976608187135, + "loss": 3.754, + "step": 3846 + }, + { + "epoch": 0.2024736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.0001772280701754386, + "loss": 4.0079, + "step": 3847 + }, + { + "epoch": 0.2025263157894737, + "grad_norm": 0.53125, + "learning_rate": 0.00017721637426900585, + "loss": 3.8237, + "step": 3848 + }, + { + "epoch": 0.20257894736842105, + "grad_norm": 0.546875, + "learning_rate": 0.0001772046783625731, + "loss": 4.3675, + "step": 3849 + }, + { + "epoch": 0.2026315789473684, + "grad_norm": 0.5390625, + "learning_rate": 0.00017719298245614035, + "loss": 4.1667, + "step": 3850 + }, + { + "epoch": 0.2026842105263158, + "grad_norm": 0.98828125, + "learning_rate": 0.0001771812865497076, + "loss": 3.8399, + "step": 3851 + }, + { + "epoch": 0.20273684210526316, + "grad_norm": 0.4765625, + "learning_rate": 0.00017716959064327485, + "loss": 3.9896, + "step": 3852 + }, + { + "epoch": 0.20278947368421052, + "grad_norm": 0.515625, + "learning_rate": 0.00017715789473684212, + "loss": 4.1452, + "step": 3853 + }, + { + "epoch": 0.2028421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00017714619883040937, + "loss": 4.183, + "step": 3854 + }, + { + "epoch": 0.20289473684210527, + "grad_norm": 0.53125, + "learning_rate": 0.00017713450292397662, + "loss": 4.1374, + "step": 3855 + }, + { + "epoch": 0.20294736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00017712280701754387, + "loss": 3.9762, + "step": 3856 + }, + { + "epoch": 0.203, + "grad_norm": 0.65234375, + "learning_rate": 0.00017711111111111112, + "loss": 3.4972, + "step": 3857 + }, + { + "epoch": 0.20305263157894737, + "grad_norm": 0.51953125, + "learning_rate": 0.00017709941520467837, + "loss": 4.6289, + "step": 3858 + }, + { + "epoch": 0.20310526315789473, + "grad_norm": 0.75, + "learning_rate": 0.00017708771929824562, + "loss": 4.7072, + "step": 3859 + }, + { + "epoch": 0.2031578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.0001770760233918129, + "loss": 4.3537, + "step": 3860 + }, + { + "epoch": 0.20321052631578948, + "grad_norm": 0.60546875, + "learning_rate": 0.00017706432748538012, + "loss": 4.3408, + "step": 3861 + }, + { + "epoch": 0.20326315789473684, + "grad_norm": 0.578125, + "learning_rate": 0.00017705263157894737, + "loss": 3.994, + "step": 3862 + }, + { + "epoch": 0.2033157894736842, + "grad_norm": 0.63671875, + "learning_rate": 0.00017704093567251462, + "loss": 4.2945, + "step": 3863 + }, + { + "epoch": 0.2033684210526316, + "grad_norm": 0.61328125, + "learning_rate": 0.00017702923976608187, + "loss": 3.7943, + "step": 3864 + }, + { + "epoch": 0.20342105263157895, + "grad_norm": 0.5859375, + "learning_rate": 0.00017701754385964912, + "loss": 3.8164, + "step": 3865 + }, + { + "epoch": 0.2034736842105263, + "grad_norm": 0.4921875, + "learning_rate": 0.00017700584795321637, + "loss": 4.3162, + "step": 3866 + }, + { + "epoch": 0.2035263157894737, + "grad_norm": 0.578125, + "learning_rate": 0.00017699415204678362, + "loss": 4.1558, + "step": 3867 + }, + { + "epoch": 0.20357894736842105, + "grad_norm": 0.66015625, + "learning_rate": 0.0001769824561403509, + "loss": 3.7824, + "step": 3868 + }, + { + "epoch": 0.2036315789473684, + "grad_norm": 0.5859375, + "learning_rate": 0.00017697076023391814, + "loss": 4.073, + "step": 3869 + }, + { + "epoch": 0.2036842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.0001769590643274854, + "loss": 4.2674, + "step": 3870 + }, + { + "epoch": 0.20373684210526316, + "grad_norm": 0.55078125, + "learning_rate": 0.00017694736842105264, + "loss": 4.0264, + "step": 3871 + }, + { + "epoch": 0.20378947368421052, + "grad_norm": 0.52734375, + "learning_rate": 0.0001769356725146199, + "loss": 3.9812, + "step": 3872 + }, + { + "epoch": 0.2038421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00017692397660818714, + "loss": 4.2281, + "step": 3873 + }, + { + "epoch": 0.20389473684210527, + "grad_norm": 0.58203125, + "learning_rate": 0.0001769122807017544, + "loss": 4.1728, + "step": 3874 + }, + { + "epoch": 0.20394736842105263, + "grad_norm": 0.62890625, + "learning_rate": 0.00017690058479532167, + "loss": 4.1982, + "step": 3875 + }, + { + "epoch": 0.204, + "grad_norm": 0.85546875, + "learning_rate": 0.0001768888888888889, + "loss": 3.9098, + "step": 3876 + }, + { + "epoch": 0.20405263157894737, + "grad_norm": 0.498046875, + "learning_rate": 0.00017687719298245614, + "loss": 3.706, + "step": 3877 + }, + { + "epoch": 0.20410526315789473, + "grad_norm": 0.57421875, + "learning_rate": 0.0001768654970760234, + "loss": 4.0157, + "step": 3878 + }, + { + "epoch": 0.2041578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00017685380116959064, + "loss": 4.1541, + "step": 3879 + }, + { + "epoch": 0.20421052631578948, + "grad_norm": 0.55859375, + "learning_rate": 0.0001768421052631579, + "loss": 4.0453, + "step": 3880 + }, + { + "epoch": 0.20426315789473684, + "grad_norm": 0.62109375, + "learning_rate": 0.00017683040935672514, + "loss": 3.9596, + "step": 3881 + }, + { + "epoch": 0.2043157894736842, + "grad_norm": 0.62109375, + "learning_rate": 0.00017681871345029242, + "loss": 3.5991, + "step": 3882 + }, + { + "epoch": 0.2043684210526316, + "grad_norm": 0.60546875, + "learning_rate": 0.00017680701754385967, + "loss": 4.1287, + "step": 3883 + }, + { + "epoch": 0.20442105263157895, + "grad_norm": 0.60546875, + "learning_rate": 0.00017679532163742691, + "loss": 3.786, + "step": 3884 + }, + { + "epoch": 0.2044736842105263, + "grad_norm": 0.6171875, + "learning_rate": 0.00017678362573099416, + "loss": 3.7853, + "step": 3885 + }, + { + "epoch": 0.2045263157894737, + "grad_norm": 0.53515625, + "learning_rate": 0.00017677192982456141, + "loss": 4.3675, + "step": 3886 + }, + { + "epoch": 0.20457894736842105, + "grad_norm": 0.6484375, + "learning_rate": 0.00017676023391812866, + "loss": 4.4681, + "step": 3887 + }, + { + "epoch": 0.20463157894736841, + "grad_norm": 0.53515625, + "learning_rate": 0.0001767485380116959, + "loss": 4.2055, + "step": 3888 + }, + { + "epoch": 0.2046842105263158, + "grad_norm": 0.5703125, + "learning_rate": 0.00017673684210526316, + "loss": 4.6839, + "step": 3889 + }, + { + "epoch": 0.20473684210526316, + "grad_norm": 0.56640625, + "learning_rate": 0.00017672514619883044, + "loss": 4.3946, + "step": 3890 + }, + { + "epoch": 0.20478947368421052, + "grad_norm": 0.59765625, + "learning_rate": 0.00017671345029239766, + "loss": 3.331, + "step": 3891 + }, + { + "epoch": 0.20484210526315788, + "grad_norm": 0.6015625, + "learning_rate": 0.0001767017543859649, + "loss": 4.059, + "step": 3892 + }, + { + "epoch": 0.20489473684210527, + "grad_norm": 0.7109375, + "learning_rate": 0.00017669005847953216, + "loss": 4.2641, + "step": 3893 + }, + { + "epoch": 0.20494736842105263, + "grad_norm": 0.80078125, + "learning_rate": 0.0001766783625730994, + "loss": 4.2415, + "step": 3894 + }, + { + "epoch": 0.205, + "grad_norm": 0.54296875, + "learning_rate": 0.00017666666666666666, + "loss": 4.2404, + "step": 3895 + }, + { + "epoch": 0.20505263157894738, + "grad_norm": 0.5703125, + "learning_rate": 0.0001766549707602339, + "loss": 3.6536, + "step": 3896 + }, + { + "epoch": 0.20510526315789473, + "grad_norm": 0.56640625, + "learning_rate": 0.0001766432748538012, + "loss": 3.995, + "step": 3897 + }, + { + "epoch": 0.2051578947368421, + "grad_norm": 0.6015625, + "learning_rate": 0.00017663157894736844, + "loss": 4.2778, + "step": 3898 + }, + { + "epoch": 0.20521052631578948, + "grad_norm": 0.546875, + "learning_rate": 0.00017661988304093569, + "loss": 4.4362, + "step": 3899 + }, + { + "epoch": 0.20526315789473684, + "grad_norm": 0.5234375, + "learning_rate": 0.00017660818713450294, + "loss": 3.9706, + "step": 3900 + }, + { + "epoch": 0.2053157894736842, + "grad_norm": 0.53515625, + "learning_rate": 0.00017659649122807018, + "loss": 4.3064, + "step": 3901 + }, + { + "epoch": 0.2053684210526316, + "grad_norm": 0.515625, + "learning_rate": 0.00017658479532163743, + "loss": 4.2363, + "step": 3902 + }, + { + "epoch": 0.20542105263157895, + "grad_norm": 0.55078125, + "learning_rate": 0.00017657309941520468, + "loss": 4.2121, + "step": 3903 + }, + { + "epoch": 0.2054736842105263, + "grad_norm": 0.55078125, + "learning_rate": 0.00017656140350877193, + "loss": 4.0262, + "step": 3904 + }, + { + "epoch": 0.2055263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.0001765497076023392, + "loss": 4.0439, + "step": 3905 + }, + { + "epoch": 0.20557894736842106, + "grad_norm": 0.51171875, + "learning_rate": 0.00017653801169590646, + "loss": 4.1132, + "step": 3906 + }, + { + "epoch": 0.20563157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00017652631578947368, + "loss": 4.0593, + "step": 3907 + }, + { + "epoch": 0.2056842105263158, + "grad_norm": 0.52734375, + "learning_rate": 0.00017651461988304093, + "loss": 3.7408, + "step": 3908 + }, + { + "epoch": 0.20573684210526316, + "grad_norm": 0.48828125, + "learning_rate": 0.00017650292397660818, + "loss": 4.5807, + "step": 3909 + }, + { + "epoch": 0.20578947368421052, + "grad_norm": 0.578125, + "learning_rate": 0.00017649122807017543, + "loss": 4.3015, + "step": 3910 + }, + { + "epoch": 0.20584210526315788, + "grad_norm": 0.60546875, + "learning_rate": 0.00017647953216374268, + "loss": 4.1781, + "step": 3911 + }, + { + "epoch": 0.20589473684210527, + "grad_norm": 0.6171875, + "learning_rate": 0.00017646783625730996, + "loss": 4.2268, + "step": 3912 + }, + { + "epoch": 0.20594736842105263, + "grad_norm": 0.7265625, + "learning_rate": 0.0001764561403508772, + "loss": 3.7284, + "step": 3913 + }, + { + "epoch": 0.206, + "grad_norm": 0.5390625, + "learning_rate": 0.00017644444444444446, + "loss": 3.8878, + "step": 3914 + }, + { + "epoch": 0.20605263157894738, + "grad_norm": 0.484375, + "learning_rate": 0.0001764327485380117, + "loss": 4.3446, + "step": 3915 + }, + { + "epoch": 0.20610526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.00017642105263157896, + "loss": 3.8727, + "step": 3916 + }, + { + "epoch": 0.2061578947368421, + "grad_norm": 0.5703125, + "learning_rate": 0.0001764093567251462, + "loss": 3.9115, + "step": 3917 + }, + { + "epoch": 0.20621052631578948, + "grad_norm": 0.61328125, + "learning_rate": 0.00017639766081871346, + "loss": 3.8841, + "step": 3918 + }, + { + "epoch": 0.20626315789473684, + "grad_norm": 0.59375, + "learning_rate": 0.00017638596491228073, + "loss": 4.449, + "step": 3919 + }, + { + "epoch": 0.2063157894736842, + "grad_norm": 0.59375, + "learning_rate": 0.00017637426900584798, + "loss": 4.2529, + "step": 3920 + }, + { + "epoch": 0.2063684210526316, + "grad_norm": 0.5859375, + "learning_rate": 0.00017636257309941523, + "loss": 4.5566, + "step": 3921 + }, + { + "epoch": 0.20642105263157895, + "grad_norm": 0.70703125, + "learning_rate": 0.00017635087719298245, + "loss": 3.7935, + "step": 3922 + }, + { + "epoch": 0.2064736842105263, + "grad_norm": 0.5390625, + "learning_rate": 0.0001763391812865497, + "loss": 4.2801, + "step": 3923 + }, + { + "epoch": 0.2065263157894737, + "grad_norm": 0.58203125, + "learning_rate": 0.00017632748538011695, + "loss": 4.2071, + "step": 3924 + }, + { + "epoch": 0.20657894736842106, + "grad_norm": 0.640625, + "learning_rate": 0.0001763157894736842, + "loss": 4.1546, + "step": 3925 + }, + { + "epoch": 0.20663157894736842, + "grad_norm": 0.59765625, + "learning_rate": 0.00017630409356725145, + "loss": 4.2354, + "step": 3926 + }, + { + "epoch": 0.20668421052631578, + "grad_norm": 0.55859375, + "learning_rate": 0.00017629239766081873, + "loss": 3.5994, + "step": 3927 + }, + { + "epoch": 0.20673684210526316, + "grad_norm": 0.60546875, + "learning_rate": 0.00017628070175438598, + "loss": 3.8941, + "step": 3928 + }, + { + "epoch": 0.20678947368421052, + "grad_norm": 0.6171875, + "learning_rate": 0.00017626900584795323, + "loss": 3.9004, + "step": 3929 + }, + { + "epoch": 0.20684210526315788, + "grad_norm": 0.57421875, + "learning_rate": 0.00017625730994152048, + "loss": 4.011, + "step": 3930 + }, + { + "epoch": 0.20689473684210527, + "grad_norm": 0.609375, + "learning_rate": 0.00017624561403508773, + "loss": 4.1001, + "step": 3931 + }, + { + "epoch": 0.20694736842105263, + "grad_norm": 0.625, + "learning_rate": 0.00017623391812865498, + "loss": 4.2049, + "step": 3932 + }, + { + "epoch": 0.207, + "grad_norm": 0.6015625, + "learning_rate": 0.00017622222222222223, + "loss": 4.2941, + "step": 3933 + }, + { + "epoch": 0.20705263157894738, + "grad_norm": 0.56640625, + "learning_rate": 0.0001762105263157895, + "loss": 4.2975, + "step": 3934 + }, + { + "epoch": 0.20710526315789474, + "grad_norm": 0.64453125, + "learning_rate": 0.00017619883040935675, + "loss": 4.2175, + "step": 3935 + }, + { + "epoch": 0.2071578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.000176187134502924, + "loss": 3.9077, + "step": 3936 + }, + { + "epoch": 0.20721052631578948, + "grad_norm": 0.53125, + "learning_rate": 0.00017617543859649122, + "loss": 4.1573, + "step": 3937 + }, + { + "epoch": 0.20726315789473684, + "grad_norm": 0.5390625, + "learning_rate": 0.00017616374269005847, + "loss": 3.7617, + "step": 3938 + }, + { + "epoch": 0.2073157894736842, + "grad_norm": 0.609375, + "learning_rate": 0.00017615204678362572, + "loss": 4.0053, + "step": 3939 + }, + { + "epoch": 0.2073684210526316, + "grad_norm": 0.51171875, + "learning_rate": 0.00017614035087719297, + "loss": 3.9909, + "step": 3940 + }, + { + "epoch": 0.20742105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00017612865497076022, + "loss": 3.5871, + "step": 3941 + }, + { + "epoch": 0.2074736842105263, + "grad_norm": 0.53515625, + "learning_rate": 0.0001761169590643275, + "loss": 4.426, + "step": 3942 + }, + { + "epoch": 0.2075263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.00017610526315789475, + "loss": 3.6332, + "step": 3943 + }, + { + "epoch": 0.20757894736842106, + "grad_norm": 0.4765625, + "learning_rate": 0.000176093567251462, + "loss": 4.1967, + "step": 3944 + }, + { + "epoch": 0.20763157894736842, + "grad_norm": 0.69921875, + "learning_rate": 0.00017608187134502925, + "loss": 3.8811, + "step": 3945 + }, + { + "epoch": 0.20768421052631578, + "grad_norm": 0.55078125, + "learning_rate": 0.0001760701754385965, + "loss": 3.9389, + "step": 3946 + }, + { + "epoch": 0.20773684210526316, + "grad_norm": 0.5546875, + "learning_rate": 0.00017605847953216375, + "loss": 4.0558, + "step": 3947 + }, + { + "epoch": 0.20778947368421052, + "grad_norm": 0.56640625, + "learning_rate": 0.000176046783625731, + "loss": 4.2078, + "step": 3948 + }, + { + "epoch": 0.20784210526315788, + "grad_norm": 0.5625, + "learning_rate": 0.00017603508771929827, + "loss": 3.8766, + "step": 3949 + }, + { + "epoch": 0.20789473684210527, + "grad_norm": 0.546875, + "learning_rate": 0.00017602339181286552, + "loss": 4.074, + "step": 3950 + }, + { + "epoch": 0.20794736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00017601169590643277, + "loss": 4.0404, + "step": 3951 + }, + { + "epoch": 0.208, + "grad_norm": 0.8359375, + "learning_rate": 0.00017600000000000002, + "loss": 3.7338, + "step": 3952 + }, + { + "epoch": 0.20805263157894738, + "grad_norm": 0.50390625, + "learning_rate": 0.00017598830409356725, + "loss": 4.1752, + "step": 3953 + }, + { + "epoch": 0.20810526315789474, + "grad_norm": 0.62109375, + "learning_rate": 0.0001759766081871345, + "loss": 4.1875, + "step": 3954 + }, + { + "epoch": 0.2081578947368421, + "grad_norm": 0.76171875, + "learning_rate": 0.00017596491228070174, + "loss": 3.8812, + "step": 3955 + }, + { + "epoch": 0.20821052631578948, + "grad_norm": 0.54296875, + "learning_rate": 0.000175953216374269, + "loss": 4.2274, + "step": 3956 + }, + { + "epoch": 0.20826315789473684, + "grad_norm": 0.6015625, + "learning_rate": 0.00017594152046783627, + "loss": 4.1974, + "step": 3957 + }, + { + "epoch": 0.2083157894736842, + "grad_norm": 0.53515625, + "learning_rate": 0.00017592982456140352, + "loss": 4.4842, + "step": 3958 + }, + { + "epoch": 0.2083684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00017591812865497077, + "loss": 4.3143, + "step": 3959 + }, + { + "epoch": 0.20842105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00017590643274853802, + "loss": 4.1087, + "step": 3960 + }, + { + "epoch": 0.2084736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00017589473684210527, + "loss": 4.3197, + "step": 3961 + }, + { + "epoch": 0.20852631578947367, + "grad_norm": 0.64453125, + "learning_rate": 0.00017588304093567252, + "loss": 4.142, + "step": 3962 + }, + { + "epoch": 0.20857894736842106, + "grad_norm": 0.64453125, + "learning_rate": 0.00017587134502923977, + "loss": 4.2838, + "step": 3963 + }, + { + "epoch": 0.20863157894736842, + "grad_norm": 0.67578125, + "learning_rate": 0.00017585964912280704, + "loss": 4.3858, + "step": 3964 + }, + { + "epoch": 0.20868421052631578, + "grad_norm": 0.64453125, + "learning_rate": 0.0001758479532163743, + "loss": 3.933, + "step": 3965 + }, + { + "epoch": 0.20873684210526317, + "grad_norm": 0.8046875, + "learning_rate": 0.00017583625730994154, + "loss": 4.3459, + "step": 3966 + }, + { + "epoch": 0.20878947368421052, + "grad_norm": 0.49609375, + "learning_rate": 0.0001758245614035088, + "loss": 4.0454, + "step": 3967 + }, + { + "epoch": 0.20884210526315788, + "grad_norm": 0.53125, + "learning_rate": 0.00017581286549707602, + "loss": 3.671, + "step": 3968 + }, + { + "epoch": 0.20889473684210527, + "grad_norm": 0.546875, + "learning_rate": 0.00017580116959064327, + "loss": 4.0595, + "step": 3969 + }, + { + "epoch": 0.20894736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.00017578947368421052, + "loss": 3.8829, + "step": 3970 + }, + { + "epoch": 0.209, + "grad_norm": 0.5546875, + "learning_rate": 0.0001757777777777778, + "loss": 4.126, + "step": 3971 + }, + { + "epoch": 0.20905263157894738, + "grad_norm": 0.52734375, + "learning_rate": 0.00017576608187134504, + "loss": 4.2737, + "step": 3972 + }, + { + "epoch": 0.20910526315789474, + "grad_norm": 0.52734375, + "learning_rate": 0.0001757543859649123, + "loss": 4.0914, + "step": 3973 + }, + { + "epoch": 0.2091578947368421, + "grad_norm": 0.71875, + "learning_rate": 0.00017574269005847954, + "loss": 3.6789, + "step": 3974 + }, + { + "epoch": 0.20921052631578949, + "grad_norm": 0.79296875, + "learning_rate": 0.0001757309941520468, + "loss": 3.9909, + "step": 3975 + }, + { + "epoch": 0.20926315789473685, + "grad_norm": 0.5546875, + "learning_rate": 0.00017571929824561404, + "loss": 4.1033, + "step": 3976 + }, + { + "epoch": 0.2093157894736842, + "grad_norm": 0.58203125, + "learning_rate": 0.0001757076023391813, + "loss": 4.1394, + "step": 3977 + }, + { + "epoch": 0.2093684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00017569590643274854, + "loss": 4.2234, + "step": 3978 + }, + { + "epoch": 0.20942105263157895, + "grad_norm": 0.578125, + "learning_rate": 0.00017568421052631582, + "loss": 3.885, + "step": 3979 + }, + { + "epoch": 0.2094736842105263, + "grad_norm": 0.5703125, + "learning_rate": 0.00017567251461988307, + "loss": 4.1313, + "step": 3980 + }, + { + "epoch": 0.20952631578947367, + "grad_norm": 0.5625, + "learning_rate": 0.00017566081871345031, + "loss": 4.2159, + "step": 3981 + }, + { + "epoch": 0.20957894736842106, + "grad_norm": 0.494140625, + "learning_rate": 0.00017564912280701756, + "loss": 4.1394, + "step": 3982 + }, + { + "epoch": 0.20963157894736842, + "grad_norm": 0.59375, + "learning_rate": 0.0001756374269005848, + "loss": 4.3493, + "step": 3983 + }, + { + "epoch": 0.20968421052631578, + "grad_norm": 0.5703125, + "learning_rate": 0.00017562573099415204, + "loss": 4.1893, + "step": 3984 + }, + { + "epoch": 0.20973684210526317, + "grad_norm": 0.52734375, + "learning_rate": 0.00017561403508771929, + "loss": 4.2625, + "step": 3985 + }, + { + "epoch": 0.20978947368421053, + "grad_norm": 0.52734375, + "learning_rate": 0.00017560233918128656, + "loss": 4.1808, + "step": 3986 + }, + { + "epoch": 0.20984210526315789, + "grad_norm": 0.62890625, + "learning_rate": 0.0001755906432748538, + "loss": 4.1176, + "step": 3987 + }, + { + "epoch": 0.20989473684210527, + "grad_norm": 0.578125, + "learning_rate": 0.00017557894736842106, + "loss": 4.1133, + "step": 3988 + }, + { + "epoch": 0.20994736842105263, + "grad_norm": 0.6484375, + "learning_rate": 0.0001755672514619883, + "loss": 4.554, + "step": 3989 + }, + { + "epoch": 0.21, + "grad_norm": 0.59375, + "learning_rate": 0.00017555555555555556, + "loss": 4.3226, + "step": 3990 + }, + { + "epoch": 0.21005263157894738, + "grad_norm": 0.6015625, + "learning_rate": 0.0001755438596491228, + "loss": 4.4555, + "step": 3991 + }, + { + "epoch": 0.21010526315789474, + "grad_norm": 0.63671875, + "learning_rate": 0.00017553216374269006, + "loss": 4.3043, + "step": 3992 + }, + { + "epoch": 0.2101578947368421, + "grad_norm": 0.59375, + "learning_rate": 0.0001755204678362573, + "loss": 4.5643, + "step": 3993 + }, + { + "epoch": 0.2102105263157895, + "grad_norm": 0.55859375, + "learning_rate": 0.0001755087719298246, + "loss": 4.2424, + "step": 3994 + }, + { + "epoch": 0.21026315789473685, + "grad_norm": 0.63671875, + "learning_rate": 0.00017549707602339184, + "loss": 3.7624, + "step": 3995 + }, + { + "epoch": 0.2103157894736842, + "grad_norm": 0.734375, + "learning_rate": 0.00017548538011695909, + "loss": 3.8263, + "step": 3996 + }, + { + "epoch": 0.21036842105263157, + "grad_norm": 0.546875, + "learning_rate": 0.00017547368421052634, + "loss": 3.955, + "step": 3997 + }, + { + "epoch": 0.21042105263157895, + "grad_norm": 0.546875, + "learning_rate": 0.00017546198830409359, + "loss": 4.3731, + "step": 3998 + }, + { + "epoch": 0.2104736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.0001754502923976608, + "loss": 4.4247, + "step": 3999 + }, + { + "epoch": 0.21052631578947367, + "grad_norm": 0.58984375, + "learning_rate": 0.00017543859649122806, + "loss": 4.3106, + "step": 4000 + }, + { + "epoch": 0.21057894736842106, + "grad_norm": 0.58984375, + "learning_rate": 0.00017542690058479533, + "loss": 4.2689, + "step": 4001 + }, + { + "epoch": 0.21063157894736842, + "grad_norm": 0.71875, + "learning_rate": 0.00017541520467836258, + "loss": 3.9069, + "step": 4002 + }, + { + "epoch": 0.21068421052631578, + "grad_norm": 0.67578125, + "learning_rate": 0.00017540350877192983, + "loss": 4.0693, + "step": 4003 + }, + { + "epoch": 0.21073684210526317, + "grad_norm": 0.63671875, + "learning_rate": 0.00017539181286549708, + "loss": 4.0126, + "step": 4004 + }, + { + "epoch": 0.21078947368421053, + "grad_norm": 0.5703125, + "learning_rate": 0.00017538011695906433, + "loss": 4.2646, + "step": 4005 + }, + { + "epoch": 0.2108421052631579, + "grad_norm": 0.578125, + "learning_rate": 0.00017536842105263158, + "loss": 3.7431, + "step": 4006 + }, + { + "epoch": 0.21089473684210527, + "grad_norm": 0.71875, + "learning_rate": 0.00017535672514619883, + "loss": 3.7922, + "step": 4007 + }, + { + "epoch": 0.21094736842105263, + "grad_norm": 0.69921875, + "learning_rate": 0.0001753450292397661, + "loss": 4.0962, + "step": 4008 + }, + { + "epoch": 0.211, + "grad_norm": 0.57421875, + "learning_rate": 0.00017533333333333336, + "loss": 4.1389, + "step": 4009 + }, + { + "epoch": 0.21105263157894738, + "grad_norm": 0.77734375, + "learning_rate": 0.0001753216374269006, + "loss": 3.7719, + "step": 4010 + }, + { + "epoch": 0.21110526315789474, + "grad_norm": 0.5703125, + "learning_rate": 0.00017530994152046786, + "loss": 3.7264, + "step": 4011 + }, + { + "epoch": 0.2111578947368421, + "grad_norm": 0.5546875, + "learning_rate": 0.0001752982456140351, + "loss": 4.1699, + "step": 4012 + }, + { + "epoch": 0.2112105263157895, + "grad_norm": 0.6328125, + "learning_rate": 0.00017528654970760236, + "loss": 3.9462, + "step": 4013 + }, + { + "epoch": 0.21126315789473685, + "grad_norm": 0.5390625, + "learning_rate": 0.00017527485380116958, + "loss": 4.4209, + "step": 4014 + }, + { + "epoch": 0.2113157894736842, + "grad_norm": 0.62890625, + "learning_rate": 0.00017526315789473683, + "loss": 4.0371, + "step": 4015 + }, + { + "epoch": 0.21136842105263157, + "grad_norm": 0.796875, + "learning_rate": 0.0001752514619883041, + "loss": 4.2673, + "step": 4016 + }, + { + "epoch": 0.21142105263157895, + "grad_norm": 0.5859375, + "learning_rate": 0.00017523976608187135, + "loss": 4.1779, + "step": 4017 + }, + { + "epoch": 0.2114736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.0001752280701754386, + "loss": 3.7956, + "step": 4018 + }, + { + "epoch": 0.21152631578947367, + "grad_norm": 0.515625, + "learning_rate": 0.00017521637426900585, + "loss": 4.2318, + "step": 4019 + }, + { + "epoch": 0.21157894736842106, + "grad_norm": 0.5546875, + "learning_rate": 0.0001752046783625731, + "loss": 4.3094, + "step": 4020 + }, + { + "epoch": 0.21163157894736842, + "grad_norm": 0.5703125, + "learning_rate": 0.00017519298245614035, + "loss": 4.2458, + "step": 4021 + }, + { + "epoch": 0.21168421052631578, + "grad_norm": 0.54296875, + "learning_rate": 0.0001751812865497076, + "loss": 4.3067, + "step": 4022 + }, + { + "epoch": 0.21173684210526317, + "grad_norm": 0.54296875, + "learning_rate": 0.00017516959064327488, + "loss": 4.0074, + "step": 4023 + }, + { + "epoch": 0.21178947368421053, + "grad_norm": 0.55859375, + "learning_rate": 0.00017515789473684213, + "loss": 4.2897, + "step": 4024 + }, + { + "epoch": 0.2118421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00017514619883040938, + "loss": 4.3338, + "step": 4025 + }, + { + "epoch": 0.21189473684210527, + "grad_norm": 0.546875, + "learning_rate": 0.00017513450292397663, + "loss": 3.9469, + "step": 4026 + }, + { + "epoch": 0.21194736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00017512280701754388, + "loss": 4.047, + "step": 4027 + }, + { + "epoch": 0.212, + "grad_norm": 0.498046875, + "learning_rate": 0.00017511111111111113, + "loss": 3.8864, + "step": 4028 + }, + { + "epoch": 0.21205263157894738, + "grad_norm": 0.53515625, + "learning_rate": 0.00017509941520467835, + "loss": 3.9238, + "step": 4029 + }, + { + "epoch": 0.21210526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.0001750877192982456, + "loss": 3.948, + "step": 4030 + }, + { + "epoch": 0.2121578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.00017507602339181288, + "loss": 3.7246, + "step": 4031 + }, + { + "epoch": 0.21221052631578946, + "grad_norm": 0.52734375, + "learning_rate": 0.00017506432748538013, + "loss": 4.1882, + "step": 4032 + }, + { + "epoch": 0.21226315789473685, + "grad_norm": 0.486328125, + "learning_rate": 0.00017505263157894738, + "loss": 4.3241, + "step": 4033 + }, + { + "epoch": 0.2123157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00017504093567251462, + "loss": 3.7557, + "step": 4034 + }, + { + "epoch": 0.21236842105263157, + "grad_norm": 0.515625, + "learning_rate": 0.00017502923976608187, + "loss": 4.1666, + "step": 4035 + }, + { + "epoch": 0.21242105263157895, + "grad_norm": 0.50390625, + "learning_rate": 0.00017501754385964912, + "loss": 4.3767, + "step": 4036 + }, + { + "epoch": 0.21247368421052631, + "grad_norm": 0.6953125, + "learning_rate": 0.00017500584795321637, + "loss": 3.8329, + "step": 4037 + }, + { + "epoch": 0.21252631578947367, + "grad_norm": 0.62890625, + "learning_rate": 0.00017499415204678365, + "loss": 3.8589, + "step": 4038 + }, + { + "epoch": 0.21257894736842106, + "grad_norm": 0.5078125, + "learning_rate": 0.0001749824561403509, + "loss": 4.3643, + "step": 4039 + }, + { + "epoch": 0.21263157894736842, + "grad_norm": 0.640625, + "learning_rate": 0.00017497076023391815, + "loss": 3.8113, + "step": 4040 + }, + { + "epoch": 0.21268421052631578, + "grad_norm": 0.5625, + "learning_rate": 0.0001749590643274854, + "loss": 3.5223, + "step": 4041 + }, + { + "epoch": 0.21273684210526317, + "grad_norm": 0.7734375, + "learning_rate": 0.00017494736842105265, + "loss": 3.7338, + "step": 4042 + }, + { + "epoch": 0.21278947368421053, + "grad_norm": 0.56640625, + "learning_rate": 0.0001749356725146199, + "loss": 4.1875, + "step": 4043 + }, + { + "epoch": 0.2128421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00017492397660818715, + "loss": 4.2962, + "step": 4044 + }, + { + "epoch": 0.21289473684210528, + "grad_norm": 0.53515625, + "learning_rate": 0.0001749122807017544, + "loss": 3.8051, + "step": 4045 + }, + { + "epoch": 0.21294736842105264, + "grad_norm": 0.54296875, + "learning_rate": 0.00017490058479532165, + "loss": 3.9056, + "step": 4046 + }, + { + "epoch": 0.213, + "grad_norm": 0.5234375, + "learning_rate": 0.0001748888888888889, + "loss": 4.0788, + "step": 4047 + }, + { + "epoch": 0.21305263157894735, + "grad_norm": 0.640625, + "learning_rate": 0.00017487719298245615, + "loss": 3.9731, + "step": 4048 + }, + { + "epoch": 0.21310526315789474, + "grad_norm": 0.60546875, + "learning_rate": 0.0001748654970760234, + "loss": 3.8913, + "step": 4049 + }, + { + "epoch": 0.2131578947368421, + "grad_norm": 0.62109375, + "learning_rate": 0.00017485380116959065, + "loss": 3.8738, + "step": 4050 + }, + { + "epoch": 0.21321052631578946, + "grad_norm": 0.5390625, + "learning_rate": 0.0001748421052631579, + "loss": 4.4293, + "step": 4051 + }, + { + "epoch": 0.21326315789473685, + "grad_norm": 0.58203125, + "learning_rate": 0.00017483040935672514, + "loss": 4.1855, + "step": 4052 + }, + { + "epoch": 0.2133157894736842, + "grad_norm": 0.60546875, + "learning_rate": 0.00017481871345029242, + "loss": 4.1863, + "step": 4053 + }, + { + "epoch": 0.21336842105263157, + "grad_norm": 0.4609375, + "learning_rate": 0.00017480701754385967, + "loss": 3.9905, + "step": 4054 + }, + { + "epoch": 0.21342105263157896, + "grad_norm": 0.50390625, + "learning_rate": 0.00017479532163742692, + "loss": 4.3594, + "step": 4055 + }, + { + "epoch": 0.21347368421052632, + "grad_norm": 0.62890625, + "learning_rate": 0.00017478362573099417, + "loss": 3.8611, + "step": 4056 + }, + { + "epoch": 0.21352631578947368, + "grad_norm": 0.5, + "learning_rate": 0.00017477192982456142, + "loss": 4.1403, + "step": 4057 + }, + { + "epoch": 0.21357894736842106, + "grad_norm": 0.53125, + "learning_rate": 0.00017476023391812867, + "loss": 3.9333, + "step": 4058 + }, + { + "epoch": 0.21363157894736842, + "grad_norm": 0.58203125, + "learning_rate": 0.00017474853801169592, + "loss": 4.1458, + "step": 4059 + }, + { + "epoch": 0.21368421052631578, + "grad_norm": 0.546875, + "learning_rate": 0.00017473684210526317, + "loss": 4.2279, + "step": 4060 + }, + { + "epoch": 0.21373684210526317, + "grad_norm": 0.63671875, + "learning_rate": 0.00017472514619883042, + "loss": 4.0847, + "step": 4061 + }, + { + "epoch": 0.21378947368421053, + "grad_norm": 0.60546875, + "learning_rate": 0.00017471345029239767, + "loss": 4.3968, + "step": 4062 + }, + { + "epoch": 0.2138421052631579, + "grad_norm": 0.59765625, + "learning_rate": 0.00017470175438596492, + "loss": 4.0573, + "step": 4063 + }, + { + "epoch": 0.21389473684210528, + "grad_norm": 0.61328125, + "learning_rate": 0.00017469005847953217, + "loss": 3.745, + "step": 4064 + }, + { + "epoch": 0.21394736842105264, + "grad_norm": 0.5859375, + "learning_rate": 0.00017467836257309942, + "loss": 4.108, + "step": 4065 + }, + { + "epoch": 0.214, + "grad_norm": 0.5546875, + "learning_rate": 0.00017466666666666667, + "loss": 4.2655, + "step": 4066 + }, + { + "epoch": 0.21405263157894736, + "grad_norm": 0.56640625, + "learning_rate": 0.00017465497076023392, + "loss": 4.2494, + "step": 4067 + }, + { + "epoch": 0.21410526315789474, + "grad_norm": 0.56640625, + "learning_rate": 0.0001746432748538012, + "loss": 3.5678, + "step": 4068 + }, + { + "epoch": 0.2141578947368421, + "grad_norm": 0.59375, + "learning_rate": 0.00017463157894736844, + "loss": 4.0745, + "step": 4069 + }, + { + "epoch": 0.21421052631578946, + "grad_norm": 0.609375, + "learning_rate": 0.0001746198830409357, + "loss": 4.0145, + "step": 4070 + }, + { + "epoch": 0.21426315789473685, + "grad_norm": 0.65234375, + "learning_rate": 0.00017460818713450294, + "loss": 3.9339, + "step": 4071 + }, + { + "epoch": 0.2143157894736842, + "grad_norm": 0.59765625, + "learning_rate": 0.0001745964912280702, + "loss": 4.2189, + "step": 4072 + }, + { + "epoch": 0.21436842105263157, + "grad_norm": 0.63671875, + "learning_rate": 0.00017458479532163744, + "loss": 3.6138, + "step": 4073 + }, + { + "epoch": 0.21442105263157896, + "grad_norm": 0.5546875, + "learning_rate": 0.0001745730994152047, + "loss": 4.0425, + "step": 4074 + }, + { + "epoch": 0.21447368421052632, + "grad_norm": 0.5859375, + "learning_rate": 0.00017456140350877194, + "loss": 3.5869, + "step": 4075 + }, + { + "epoch": 0.21452631578947368, + "grad_norm": 0.53125, + "learning_rate": 0.0001745497076023392, + "loss": 4.3171, + "step": 4076 + }, + { + "epoch": 0.21457894736842106, + "grad_norm": 0.58203125, + "learning_rate": 0.00017453801169590644, + "loss": 4.2028, + "step": 4077 + }, + { + "epoch": 0.21463157894736842, + "grad_norm": 0.65625, + "learning_rate": 0.0001745263157894737, + "loss": 3.705, + "step": 4078 + }, + { + "epoch": 0.21468421052631578, + "grad_norm": 0.5703125, + "learning_rate": 0.00017451461988304094, + "loss": 4.3124, + "step": 4079 + }, + { + "epoch": 0.21473684210526317, + "grad_norm": 0.56640625, + "learning_rate": 0.0001745029239766082, + "loss": 4.0936, + "step": 4080 + }, + { + "epoch": 0.21478947368421053, + "grad_norm": 0.53125, + "learning_rate": 0.00017449122807017544, + "loss": 4.1547, + "step": 4081 + }, + { + "epoch": 0.2148421052631579, + "grad_norm": 0.59375, + "learning_rate": 0.0001744795321637427, + "loss": 4.0574, + "step": 4082 + }, + { + "epoch": 0.21489473684210525, + "grad_norm": 0.5546875, + "learning_rate": 0.00017446783625730996, + "loss": 4.0938, + "step": 4083 + }, + { + "epoch": 0.21494736842105264, + "grad_norm": 0.5625, + "learning_rate": 0.0001744561403508772, + "loss": 4.016, + "step": 4084 + }, + { + "epoch": 0.215, + "grad_norm": 1.2578125, + "learning_rate": 0.00017444444444444446, + "loss": 3.5014, + "step": 4085 + }, + { + "epoch": 0.21505263157894736, + "grad_norm": 0.51171875, + "learning_rate": 0.0001744327485380117, + "loss": 3.9463, + "step": 4086 + }, + { + "epoch": 0.21510526315789474, + "grad_norm": 0.703125, + "learning_rate": 0.00017442105263157896, + "loss": 4.0916, + "step": 4087 + }, + { + "epoch": 0.2151578947368421, + "grad_norm": 0.64453125, + "learning_rate": 0.0001744093567251462, + "loss": 4.053, + "step": 4088 + }, + { + "epoch": 0.21521052631578946, + "grad_norm": 0.53125, + "learning_rate": 0.00017439766081871346, + "loss": 3.9646, + "step": 4089 + }, + { + "epoch": 0.21526315789473685, + "grad_norm": 0.484375, + "learning_rate": 0.0001743859649122807, + "loss": 4.3226, + "step": 4090 + }, + { + "epoch": 0.2153157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00017437426900584796, + "loss": 4.2798, + "step": 4091 + }, + { + "epoch": 0.21536842105263157, + "grad_norm": 0.57421875, + "learning_rate": 0.0001743625730994152, + "loss": 4.1985, + "step": 4092 + }, + { + "epoch": 0.21542105263157896, + "grad_norm": 0.5546875, + "learning_rate": 0.00017435087719298246, + "loss": 3.8334, + "step": 4093 + }, + { + "epoch": 0.21547368421052632, + "grad_norm": 0.5625, + "learning_rate": 0.0001743391812865497, + "loss": 4.3957, + "step": 4094 + }, + { + "epoch": 0.21552631578947368, + "grad_norm": 0.609375, + "learning_rate": 0.00017432748538011696, + "loss": 4.0318, + "step": 4095 + }, + { + "epoch": 0.21557894736842106, + "grad_norm": 0.55859375, + "learning_rate": 0.0001743157894736842, + "loss": 4.4289, + "step": 4096 + }, + { + "epoch": 0.21563157894736842, + "grad_norm": 0.57421875, + "learning_rate": 0.00017430409356725148, + "loss": 3.4753, + "step": 4097 + }, + { + "epoch": 0.21568421052631578, + "grad_norm": 0.57421875, + "learning_rate": 0.00017429239766081873, + "loss": 4.1159, + "step": 4098 + }, + { + "epoch": 0.21573684210526317, + "grad_norm": 0.5546875, + "learning_rate": 0.00017428070175438598, + "loss": 3.9717, + "step": 4099 + }, + { + "epoch": 0.21578947368421053, + "grad_norm": 0.6015625, + "learning_rate": 0.00017426900584795323, + "loss": 3.8037, + "step": 4100 + }, + { + "epoch": 0.2158421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.00017425730994152048, + "loss": 4.4184, + "step": 4101 + }, + { + "epoch": 0.21589473684210525, + "grad_norm": 0.5546875, + "learning_rate": 0.00017424561403508773, + "loss": 4.2675, + "step": 4102 + }, + { + "epoch": 0.21594736842105264, + "grad_norm": 0.59765625, + "learning_rate": 0.00017423391812865498, + "loss": 3.6503, + "step": 4103 + }, + { + "epoch": 0.216, + "grad_norm": 0.62109375, + "learning_rate": 0.00017422222222222223, + "loss": 3.7937, + "step": 4104 + }, + { + "epoch": 0.21605263157894736, + "grad_norm": 0.54296875, + "learning_rate": 0.00017421052631578948, + "loss": 4.0418, + "step": 4105 + }, + { + "epoch": 0.21610526315789474, + "grad_norm": 0.51171875, + "learning_rate": 0.00017419883040935673, + "loss": 4.2687, + "step": 4106 + }, + { + "epoch": 0.2161578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.00017418713450292398, + "loss": 3.8496, + "step": 4107 + }, + { + "epoch": 0.21621052631578946, + "grad_norm": 0.5546875, + "learning_rate": 0.00017417543859649123, + "loss": 4.0395, + "step": 4108 + }, + { + "epoch": 0.21626315789473685, + "grad_norm": 0.71875, + "learning_rate": 0.00017416374269005848, + "loss": 4.037, + "step": 4109 + }, + { + "epoch": 0.2163157894736842, + "grad_norm": 0.52734375, + "learning_rate": 0.00017415204678362573, + "loss": 3.7947, + "step": 4110 + }, + { + "epoch": 0.21636842105263157, + "grad_norm": 0.5234375, + "learning_rate": 0.00017414035087719298, + "loss": 4.0966, + "step": 4111 + }, + { + "epoch": 0.21642105263157896, + "grad_norm": 0.55859375, + "learning_rate": 0.00017412865497076026, + "loss": 3.9905, + "step": 4112 + }, + { + "epoch": 0.21647368421052632, + "grad_norm": 0.57421875, + "learning_rate": 0.0001741169590643275, + "loss": 3.8877, + "step": 4113 + }, + { + "epoch": 0.21652631578947368, + "grad_norm": 0.54296875, + "learning_rate": 0.00017410526315789475, + "loss": 4.2643, + "step": 4114 + }, + { + "epoch": 0.21657894736842107, + "grad_norm": 0.56640625, + "learning_rate": 0.000174093567251462, + "loss": 4.0716, + "step": 4115 + }, + { + "epoch": 0.21663157894736843, + "grad_norm": 0.625, + "learning_rate": 0.00017408187134502925, + "loss": 3.8146, + "step": 4116 + }, + { + "epoch": 0.21668421052631578, + "grad_norm": 0.65625, + "learning_rate": 0.0001740701754385965, + "loss": 4.0432, + "step": 4117 + }, + { + "epoch": 0.21673684210526314, + "grad_norm": 0.68359375, + "learning_rate": 0.00017405847953216375, + "loss": 4.2282, + "step": 4118 + }, + { + "epoch": 0.21678947368421053, + "grad_norm": 0.60546875, + "learning_rate": 0.000174046783625731, + "loss": 3.9634, + "step": 4119 + }, + { + "epoch": 0.2168421052631579, + "grad_norm": 0.65234375, + "learning_rate": 0.00017403508771929825, + "loss": 4.0244, + "step": 4120 + }, + { + "epoch": 0.21689473684210525, + "grad_norm": 0.5234375, + "learning_rate": 0.0001740233918128655, + "loss": 4.4799, + "step": 4121 + }, + { + "epoch": 0.21694736842105264, + "grad_norm": 0.6015625, + "learning_rate": 0.00017401169590643275, + "loss": 4.066, + "step": 4122 + }, + { + "epoch": 0.217, + "grad_norm": 0.6484375, + "learning_rate": 0.000174, + "loss": 3.9773, + "step": 4123 + }, + { + "epoch": 0.21705263157894736, + "grad_norm": 0.6171875, + "learning_rate": 0.00017398830409356725, + "loss": 4.4781, + "step": 4124 + }, + { + "epoch": 0.21710526315789475, + "grad_norm": 0.61328125, + "learning_rate": 0.0001739766081871345, + "loss": 3.8075, + "step": 4125 + }, + { + "epoch": 0.2171578947368421, + "grad_norm": 0.6328125, + "learning_rate": 0.00017396491228070175, + "loss": 4.3865, + "step": 4126 + }, + { + "epoch": 0.21721052631578947, + "grad_norm": 0.640625, + "learning_rate": 0.00017395321637426903, + "loss": 3.7627, + "step": 4127 + }, + { + "epoch": 0.21726315789473685, + "grad_norm": 0.54296875, + "learning_rate": 0.00017394152046783628, + "loss": 3.9834, + "step": 4128 + }, + { + "epoch": 0.2173157894736842, + "grad_norm": 0.54296875, + "learning_rate": 0.00017392982456140353, + "loss": 3.8291, + "step": 4129 + }, + { + "epoch": 0.21736842105263157, + "grad_norm": 0.609375, + "learning_rate": 0.00017391812865497078, + "loss": 3.9769, + "step": 4130 + }, + { + "epoch": 0.21742105263157896, + "grad_norm": 0.8046875, + "learning_rate": 0.00017390643274853802, + "loss": 3.9733, + "step": 4131 + }, + { + "epoch": 0.21747368421052632, + "grad_norm": 0.5234375, + "learning_rate": 0.00017389473684210527, + "loss": 3.7153, + "step": 4132 + }, + { + "epoch": 0.21752631578947368, + "grad_norm": 0.52734375, + "learning_rate": 0.00017388304093567252, + "loss": 3.7075, + "step": 4133 + }, + { + "epoch": 0.21757894736842107, + "grad_norm": 0.5546875, + "learning_rate": 0.00017387134502923977, + "loss": 3.5426, + "step": 4134 + }, + { + "epoch": 0.21763157894736843, + "grad_norm": 0.51171875, + "learning_rate": 0.00017385964912280702, + "loss": 3.8728, + "step": 4135 + }, + { + "epoch": 0.21768421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00017384795321637427, + "loss": 4.0197, + "step": 4136 + }, + { + "epoch": 0.21773684210526315, + "grad_norm": 0.65625, + "learning_rate": 0.00017383625730994152, + "loss": 3.7061, + "step": 4137 + }, + { + "epoch": 0.21778947368421053, + "grad_norm": 0.54296875, + "learning_rate": 0.00017382456140350877, + "loss": 4.0523, + "step": 4138 + }, + { + "epoch": 0.2178421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00017381286549707602, + "loss": 4.0784, + "step": 4139 + }, + { + "epoch": 0.21789473684210525, + "grad_norm": 0.51171875, + "learning_rate": 0.00017380116959064327, + "loss": 3.8507, + "step": 4140 + }, + { + "epoch": 0.21794736842105264, + "grad_norm": 0.59375, + "learning_rate": 0.00017378947368421052, + "loss": 3.8052, + "step": 4141 + }, + { + "epoch": 0.218, + "grad_norm": 0.5390625, + "learning_rate": 0.0001737777777777778, + "loss": 3.8014, + "step": 4142 + }, + { + "epoch": 0.21805263157894736, + "grad_norm": 0.5703125, + "learning_rate": 0.00017376608187134505, + "loss": 3.9204, + "step": 4143 + }, + { + "epoch": 0.21810526315789475, + "grad_norm": 1.4140625, + "learning_rate": 0.0001737543859649123, + "loss": 4.1179, + "step": 4144 + }, + { + "epoch": 0.2181578947368421, + "grad_norm": 0.5546875, + "learning_rate": 0.00017374269005847955, + "loss": 4.0914, + "step": 4145 + }, + { + "epoch": 0.21821052631578947, + "grad_norm": 0.58984375, + "learning_rate": 0.0001737309941520468, + "loss": 4.3197, + "step": 4146 + }, + { + "epoch": 0.21826315789473685, + "grad_norm": 0.6796875, + "learning_rate": 0.00017371929824561405, + "loss": 3.5066, + "step": 4147 + }, + { + "epoch": 0.2183157894736842, + "grad_norm": 0.5, + "learning_rate": 0.0001737076023391813, + "loss": 3.9691, + "step": 4148 + }, + { + "epoch": 0.21836842105263157, + "grad_norm": 0.64453125, + "learning_rate": 0.00017369590643274854, + "loss": 4.051, + "step": 4149 + }, + { + "epoch": 0.21842105263157896, + "grad_norm": 0.69921875, + "learning_rate": 0.0001736842105263158, + "loss": 3.9078, + "step": 4150 + }, + { + "epoch": 0.21847368421052632, + "grad_norm": 0.53515625, + "learning_rate": 0.00017367251461988304, + "loss": 3.9657, + "step": 4151 + }, + { + "epoch": 0.21852631578947368, + "grad_norm": 0.62109375, + "learning_rate": 0.0001736608187134503, + "loss": 4.0366, + "step": 4152 + }, + { + "epoch": 0.21857894736842104, + "grad_norm": 0.5546875, + "learning_rate": 0.00017364912280701754, + "loss": 3.6283, + "step": 4153 + }, + { + "epoch": 0.21863157894736843, + "grad_norm": 0.5390625, + "learning_rate": 0.0001736374269005848, + "loss": 3.9727, + "step": 4154 + }, + { + "epoch": 0.2186842105263158, + "grad_norm": 0.63671875, + "learning_rate": 0.00017362573099415204, + "loss": 3.6974, + "step": 4155 + }, + { + "epoch": 0.21873684210526315, + "grad_norm": 0.62109375, + "learning_rate": 0.0001736140350877193, + "loss": 3.825, + "step": 4156 + }, + { + "epoch": 0.21878947368421053, + "grad_norm": 0.546875, + "learning_rate": 0.00017360233918128657, + "loss": 4.0002, + "step": 4157 + }, + { + "epoch": 0.2188421052631579, + "grad_norm": 0.6171875, + "learning_rate": 0.00017359064327485382, + "loss": 3.9227, + "step": 4158 + }, + { + "epoch": 0.21889473684210525, + "grad_norm": 0.5859375, + "learning_rate": 0.00017357894736842107, + "loss": 3.9319, + "step": 4159 + }, + { + "epoch": 0.21894736842105264, + "grad_norm": 1.0078125, + "learning_rate": 0.00017356725146198832, + "loss": 4.2159, + "step": 4160 + }, + { + "epoch": 0.219, + "grad_norm": 0.9140625, + "learning_rate": 0.00017355555555555557, + "loss": 4.2439, + "step": 4161 + }, + { + "epoch": 0.21905263157894736, + "grad_norm": 0.65234375, + "learning_rate": 0.00017354385964912282, + "loss": 3.7928, + "step": 4162 + }, + { + "epoch": 0.21910526315789475, + "grad_norm": 0.53515625, + "learning_rate": 0.00017353216374269007, + "loss": 4.0095, + "step": 4163 + }, + { + "epoch": 0.2191578947368421, + "grad_norm": 0.5546875, + "learning_rate": 0.00017352046783625732, + "loss": 4.0697, + "step": 4164 + }, + { + "epoch": 0.21921052631578947, + "grad_norm": 0.53515625, + "learning_rate": 0.00017350877192982457, + "loss": 4.1464, + "step": 4165 + }, + { + "epoch": 0.21926315789473685, + "grad_norm": 0.6015625, + "learning_rate": 0.00017349707602339181, + "loss": 4.1433, + "step": 4166 + }, + { + "epoch": 0.21931578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00017348538011695906, + "loss": 4.1977, + "step": 4167 + }, + { + "epoch": 0.21936842105263157, + "grad_norm": 0.5390625, + "learning_rate": 0.00017347368421052631, + "loss": 3.9803, + "step": 4168 + }, + { + "epoch": 0.21942105263157896, + "grad_norm": 0.5703125, + "learning_rate": 0.00017346198830409356, + "loss": 3.9897, + "step": 4169 + }, + { + "epoch": 0.21947368421052632, + "grad_norm": 0.59765625, + "learning_rate": 0.0001734502923976608, + "loss": 4.1551, + "step": 4170 + }, + { + "epoch": 0.21952631578947368, + "grad_norm": 0.59765625, + "learning_rate": 0.0001734385964912281, + "loss": 4.2826, + "step": 4171 + }, + { + "epoch": 0.21957894736842104, + "grad_norm": 0.66796875, + "learning_rate": 0.00017342690058479534, + "loss": 4.1447, + "step": 4172 + }, + { + "epoch": 0.21963157894736843, + "grad_norm": 0.6796875, + "learning_rate": 0.0001734152046783626, + "loss": 3.98, + "step": 4173 + }, + { + "epoch": 0.2196842105263158, + "grad_norm": 0.6953125, + "learning_rate": 0.00017340350877192984, + "loss": 3.8934, + "step": 4174 + }, + { + "epoch": 0.21973684210526315, + "grad_norm": 0.48828125, + "learning_rate": 0.0001733918128654971, + "loss": 4.2527, + "step": 4175 + }, + { + "epoch": 0.21978947368421053, + "grad_norm": 0.6015625, + "learning_rate": 0.00017338011695906434, + "loss": 3.8493, + "step": 4176 + }, + { + "epoch": 0.2198421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.0001733684210526316, + "loss": 3.9772, + "step": 4177 + }, + { + "epoch": 0.21989473684210525, + "grad_norm": 0.59765625, + "learning_rate": 0.00017335672514619884, + "loss": 3.8478, + "step": 4178 + }, + { + "epoch": 0.21994736842105264, + "grad_norm": 0.55078125, + "learning_rate": 0.0001733450292397661, + "loss": 4.066, + "step": 4179 + }, + { + "epoch": 0.22, + "grad_norm": 0.5625, + "learning_rate": 0.00017333333333333334, + "loss": 3.6795, + "step": 4180 + }, + { + "epoch": 0.22005263157894736, + "grad_norm": 0.6796875, + "learning_rate": 0.00017332163742690059, + "loss": 3.7091, + "step": 4181 + }, + { + "epoch": 0.22010526315789475, + "grad_norm": 0.55078125, + "learning_rate": 0.00017330994152046784, + "loss": 3.9886, + "step": 4182 + }, + { + "epoch": 0.2201578947368421, + "grad_norm": 0.57421875, + "learning_rate": 0.00017329824561403508, + "loss": 4.4191, + "step": 4183 + }, + { + "epoch": 0.22021052631578947, + "grad_norm": 0.5546875, + "learning_rate": 0.00017328654970760233, + "loss": 4.3776, + "step": 4184 + }, + { + "epoch": 0.22026315789473686, + "grad_norm": 0.5625, + "learning_rate": 0.00017327485380116958, + "loss": 3.8807, + "step": 4185 + }, + { + "epoch": 0.22031578947368421, + "grad_norm": 0.59375, + "learning_rate": 0.00017326315789473686, + "loss": 3.9082, + "step": 4186 + }, + { + "epoch": 0.22036842105263157, + "grad_norm": 0.60546875, + "learning_rate": 0.0001732514619883041, + "loss": 3.71, + "step": 4187 + }, + { + "epoch": 0.22042105263157893, + "grad_norm": 0.5703125, + "learning_rate": 0.00017323976608187136, + "loss": 3.9596, + "step": 4188 + }, + { + "epoch": 0.22047368421052632, + "grad_norm": 0.5078125, + "learning_rate": 0.0001732280701754386, + "loss": 4.0207, + "step": 4189 + }, + { + "epoch": 0.22052631578947368, + "grad_norm": 0.51171875, + "learning_rate": 0.00017321637426900586, + "loss": 3.8133, + "step": 4190 + }, + { + "epoch": 0.22057894736842104, + "grad_norm": 0.6015625, + "learning_rate": 0.0001732046783625731, + "loss": 4.0014, + "step": 4191 + }, + { + "epoch": 0.22063157894736843, + "grad_norm": 0.6015625, + "learning_rate": 0.00017319298245614036, + "loss": 4.4406, + "step": 4192 + }, + { + "epoch": 0.2206842105263158, + "grad_norm": 0.65625, + "learning_rate": 0.0001731812865497076, + "loss": 4.0468, + "step": 4193 + }, + { + "epoch": 0.22073684210526315, + "grad_norm": 0.6015625, + "learning_rate": 0.00017316959064327486, + "loss": 4.1814, + "step": 4194 + }, + { + "epoch": 0.22078947368421054, + "grad_norm": 0.66796875, + "learning_rate": 0.0001731578947368421, + "loss": 4.0284, + "step": 4195 + }, + { + "epoch": 0.2208421052631579, + "grad_norm": 0.5546875, + "learning_rate": 0.00017314619883040936, + "loss": 3.6332, + "step": 4196 + }, + { + "epoch": 0.22089473684210525, + "grad_norm": 0.53515625, + "learning_rate": 0.0001731345029239766, + "loss": 4.3161, + "step": 4197 + }, + { + "epoch": 0.22094736842105264, + "grad_norm": 0.94921875, + "learning_rate": 0.00017312280701754386, + "loss": 4.0288, + "step": 4198 + }, + { + "epoch": 0.221, + "grad_norm": 0.67578125, + "learning_rate": 0.0001731111111111111, + "loss": 4.3875, + "step": 4199 + }, + { + "epoch": 0.22105263157894736, + "grad_norm": 0.6328125, + "learning_rate": 0.00017309941520467836, + "loss": 3.7196, + "step": 4200 + }, + { + "epoch": 0.22110526315789475, + "grad_norm": 0.53515625, + "learning_rate": 0.00017308771929824563, + "loss": 4.1048, + "step": 4201 + }, + { + "epoch": 0.2211578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.00017307602339181288, + "loss": 4.1663, + "step": 4202 + }, + { + "epoch": 0.22121052631578947, + "grad_norm": 0.6015625, + "learning_rate": 0.00017306432748538013, + "loss": 4.2109, + "step": 4203 + }, + { + "epoch": 0.22126315789473683, + "grad_norm": 0.53515625, + "learning_rate": 0.00017305263157894738, + "loss": 4.0222, + "step": 4204 + }, + { + "epoch": 0.22131578947368422, + "grad_norm": 0.53125, + "learning_rate": 0.00017304093567251463, + "loss": 3.8096, + "step": 4205 + }, + { + "epoch": 0.22136842105263158, + "grad_norm": 0.6171875, + "learning_rate": 0.00017302923976608188, + "loss": 3.9295, + "step": 4206 + }, + { + "epoch": 0.22142105263157894, + "grad_norm": 0.50390625, + "learning_rate": 0.00017301754385964913, + "loss": 4.6557, + "step": 4207 + }, + { + "epoch": 0.22147368421052632, + "grad_norm": 0.62890625, + "learning_rate": 0.0001730058479532164, + "loss": 3.8245, + "step": 4208 + }, + { + "epoch": 0.22152631578947368, + "grad_norm": 0.6484375, + "learning_rate": 0.00017299415204678363, + "loss": 4.2921, + "step": 4209 + }, + { + "epoch": 0.22157894736842104, + "grad_norm": 0.55859375, + "learning_rate": 0.00017298245614035088, + "loss": 4.0578, + "step": 4210 + }, + { + "epoch": 0.22163157894736843, + "grad_norm": 0.5546875, + "learning_rate": 0.00017297076023391813, + "loss": 4.2693, + "step": 4211 + }, + { + "epoch": 0.2216842105263158, + "grad_norm": 0.69140625, + "learning_rate": 0.00017295906432748538, + "loss": 4.4653, + "step": 4212 + }, + { + "epoch": 0.22173684210526315, + "grad_norm": 0.73046875, + "learning_rate": 0.00017294736842105263, + "loss": 4.5197, + "step": 4213 + }, + { + "epoch": 0.22178947368421054, + "grad_norm": 0.640625, + "learning_rate": 0.00017293567251461988, + "loss": 4.4665, + "step": 4214 + }, + { + "epoch": 0.2218421052631579, + "grad_norm": 0.70703125, + "learning_rate": 0.00017292397660818713, + "loss": 3.2359, + "step": 4215 + }, + { + "epoch": 0.22189473684210526, + "grad_norm": 0.5546875, + "learning_rate": 0.0001729122807017544, + "loss": 3.8003, + "step": 4216 + }, + { + "epoch": 0.22194736842105264, + "grad_norm": 0.515625, + "learning_rate": 0.00017290058479532165, + "loss": 3.7631, + "step": 4217 + }, + { + "epoch": 0.222, + "grad_norm": 0.50390625, + "learning_rate": 0.0001728888888888889, + "loss": 4.2523, + "step": 4218 + }, + { + "epoch": 0.22205263157894736, + "grad_norm": 0.5390625, + "learning_rate": 0.00017287719298245615, + "loss": 3.9169, + "step": 4219 + }, + { + "epoch": 0.22210526315789475, + "grad_norm": 0.498046875, + "learning_rate": 0.0001728654970760234, + "loss": 4.5516, + "step": 4220 + }, + { + "epoch": 0.2221578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00017285380116959065, + "loss": 4.1971, + "step": 4221 + }, + { + "epoch": 0.22221052631578947, + "grad_norm": 0.8359375, + "learning_rate": 0.0001728421052631579, + "loss": 4.4014, + "step": 4222 + }, + { + "epoch": 0.22226315789473683, + "grad_norm": 0.51953125, + "learning_rate": 0.00017283040935672518, + "loss": 4.1391, + "step": 4223 + }, + { + "epoch": 0.22231578947368422, + "grad_norm": 0.59765625, + "learning_rate": 0.00017281871345029243, + "loss": 4.1892, + "step": 4224 + }, + { + "epoch": 0.22236842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.00017280701754385965, + "loss": 3.7542, + "step": 4225 + }, + { + "epoch": 0.22242105263157894, + "grad_norm": 0.59765625, + "learning_rate": 0.0001727953216374269, + "loss": 3.8961, + "step": 4226 + }, + { + "epoch": 0.22247368421052632, + "grad_norm": 0.55859375, + "learning_rate": 0.00017278362573099415, + "loss": 4.4504, + "step": 4227 + }, + { + "epoch": 0.22252631578947368, + "grad_norm": 0.54296875, + "learning_rate": 0.0001727719298245614, + "loss": 3.9146, + "step": 4228 + }, + { + "epoch": 0.22257894736842104, + "grad_norm": 0.56640625, + "learning_rate": 0.00017276023391812865, + "loss": 4.3936, + "step": 4229 + }, + { + "epoch": 0.22263157894736843, + "grad_norm": 0.5625, + "learning_rate": 0.0001727485380116959, + "loss": 3.8397, + "step": 4230 + }, + { + "epoch": 0.2226842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00017273684210526317, + "loss": 4.3424, + "step": 4231 + }, + { + "epoch": 0.22273684210526315, + "grad_norm": 0.61328125, + "learning_rate": 0.00017272514619883042, + "loss": 3.8166, + "step": 4232 + }, + { + "epoch": 0.22278947368421054, + "grad_norm": 0.5546875, + "learning_rate": 0.00017271345029239767, + "loss": 4.2086, + "step": 4233 + }, + { + "epoch": 0.2228421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00017270175438596492, + "loss": 4.3944, + "step": 4234 + }, + { + "epoch": 0.22289473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.00017269005847953217, + "loss": 3.7555, + "step": 4235 + }, + { + "epoch": 0.22294736842105264, + "grad_norm": 0.51953125, + "learning_rate": 0.00017267836257309942, + "loss": 3.9293, + "step": 4236 + }, + { + "epoch": 0.223, + "grad_norm": 0.546875, + "learning_rate": 0.00017266666666666667, + "loss": 3.9718, + "step": 4237 + }, + { + "epoch": 0.22305263157894736, + "grad_norm": 0.5546875, + "learning_rate": 0.00017265497076023395, + "loss": 4.3189, + "step": 4238 + }, + { + "epoch": 0.22310526315789472, + "grad_norm": 0.55078125, + "learning_rate": 0.0001726432748538012, + "loss": 4.2746, + "step": 4239 + }, + { + "epoch": 0.2231578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00017263157894736842, + "loss": 4.2914, + "step": 4240 + }, + { + "epoch": 0.22321052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00017261988304093567, + "loss": 4.0845, + "step": 4241 + }, + { + "epoch": 0.22326315789473683, + "grad_norm": 0.66015625, + "learning_rate": 0.00017260818713450292, + "loss": 3.6687, + "step": 4242 + }, + { + "epoch": 0.22331578947368422, + "grad_norm": 0.494140625, + "learning_rate": 0.00017259649122807017, + "loss": 3.8168, + "step": 4243 + }, + { + "epoch": 0.22336842105263158, + "grad_norm": 0.57421875, + "learning_rate": 0.00017258479532163742, + "loss": 3.8279, + "step": 4244 + }, + { + "epoch": 0.22342105263157894, + "grad_norm": 0.52734375, + "learning_rate": 0.0001725730994152047, + "loss": 3.8074, + "step": 4245 + }, + { + "epoch": 0.22347368421052632, + "grad_norm": 0.53515625, + "learning_rate": 0.00017256140350877194, + "loss": 3.8591, + "step": 4246 + }, + { + "epoch": 0.22352631578947368, + "grad_norm": 0.5, + "learning_rate": 0.0001725497076023392, + "loss": 3.9293, + "step": 4247 + }, + { + "epoch": 0.22357894736842104, + "grad_norm": 0.578125, + "learning_rate": 0.00017253801169590644, + "loss": 3.9359, + "step": 4248 + }, + { + "epoch": 0.22363157894736843, + "grad_norm": 0.64453125, + "learning_rate": 0.0001725263157894737, + "loss": 4.3357, + "step": 4249 + }, + { + "epoch": 0.2236842105263158, + "grad_norm": 1.3515625, + "learning_rate": 0.00017251461988304094, + "loss": 4.3796, + "step": 4250 + }, + { + "epoch": 0.22373684210526315, + "grad_norm": 0.478515625, + "learning_rate": 0.0001725029239766082, + "loss": 4.1769, + "step": 4251 + }, + { + "epoch": 0.22378947368421054, + "grad_norm": 0.50390625, + "learning_rate": 0.00017249122807017544, + "loss": 4.0681, + "step": 4252 + }, + { + "epoch": 0.2238421052631579, + "grad_norm": 0.494140625, + "learning_rate": 0.00017247953216374272, + "loss": 3.8625, + "step": 4253 + }, + { + "epoch": 0.22389473684210526, + "grad_norm": 0.6015625, + "learning_rate": 0.00017246783625730997, + "loss": 4.0292, + "step": 4254 + }, + { + "epoch": 0.22394736842105264, + "grad_norm": 0.54296875, + "learning_rate": 0.0001724561403508772, + "loss": 4.0521, + "step": 4255 + }, + { + "epoch": 0.224, + "grad_norm": 0.62109375, + "learning_rate": 0.00017244444444444444, + "loss": 3.8429, + "step": 4256 + }, + { + "epoch": 0.22405263157894736, + "grad_norm": 0.56640625, + "learning_rate": 0.0001724327485380117, + "loss": 3.8592, + "step": 4257 + }, + { + "epoch": 0.22410526315789472, + "grad_norm": 0.6953125, + "learning_rate": 0.00017242105263157894, + "loss": 4.0491, + "step": 4258 + }, + { + "epoch": 0.2241578947368421, + "grad_norm": 0.58203125, + "learning_rate": 0.0001724093567251462, + "loss": 3.9694, + "step": 4259 + }, + { + "epoch": 0.22421052631578947, + "grad_norm": 0.50390625, + "learning_rate": 0.00017239766081871347, + "loss": 4.3353, + "step": 4260 + }, + { + "epoch": 0.22426315789473683, + "grad_norm": 0.5390625, + "learning_rate": 0.00017238596491228072, + "loss": 3.874, + "step": 4261 + }, + { + "epoch": 0.22431578947368422, + "grad_norm": 0.5078125, + "learning_rate": 0.00017237426900584797, + "loss": 4.1218, + "step": 4262 + }, + { + "epoch": 0.22436842105263158, + "grad_norm": 0.58203125, + "learning_rate": 0.00017236257309941522, + "loss": 3.9352, + "step": 4263 + }, + { + "epoch": 0.22442105263157894, + "grad_norm": 0.57421875, + "learning_rate": 0.00017235087719298246, + "loss": 3.5322, + "step": 4264 + }, + { + "epoch": 0.22447368421052633, + "grad_norm": 0.5859375, + "learning_rate": 0.00017233918128654971, + "loss": 3.7177, + "step": 4265 + }, + { + "epoch": 0.22452631578947368, + "grad_norm": 0.5234375, + "learning_rate": 0.00017232748538011696, + "loss": 4.0704, + "step": 4266 + }, + { + "epoch": 0.22457894736842104, + "grad_norm": 0.58984375, + "learning_rate": 0.0001723157894736842, + "loss": 3.6611, + "step": 4267 + }, + { + "epoch": 0.22463157894736843, + "grad_norm": 0.5546875, + "learning_rate": 0.0001723040935672515, + "loss": 4.066, + "step": 4268 + }, + { + "epoch": 0.2246842105263158, + "grad_norm": 0.5859375, + "learning_rate": 0.00017229239766081874, + "loss": 3.6314, + "step": 4269 + }, + { + "epoch": 0.22473684210526315, + "grad_norm": 0.52734375, + "learning_rate": 0.000172280701754386, + "loss": 3.7148, + "step": 4270 + }, + { + "epoch": 0.22478947368421054, + "grad_norm": 0.57421875, + "learning_rate": 0.0001722690058479532, + "loss": 4.2079, + "step": 4271 + }, + { + "epoch": 0.2248421052631579, + "grad_norm": 0.5703125, + "learning_rate": 0.00017225730994152046, + "loss": 4.4161, + "step": 4272 + }, + { + "epoch": 0.22489473684210526, + "grad_norm": 0.81640625, + "learning_rate": 0.0001722456140350877, + "loss": 3.6807, + "step": 4273 + }, + { + "epoch": 0.22494736842105262, + "grad_norm": 0.53515625, + "learning_rate": 0.00017223391812865496, + "loss": 3.7692, + "step": 4274 + }, + { + "epoch": 0.225, + "grad_norm": 0.55078125, + "learning_rate": 0.00017222222222222224, + "loss": 3.8178, + "step": 4275 + }, + { + "epoch": 0.22505263157894737, + "grad_norm": 0.5625, + "learning_rate": 0.0001722105263157895, + "loss": 4.1153, + "step": 4276 + }, + { + "epoch": 0.22510526315789472, + "grad_norm": 0.55859375, + "learning_rate": 0.00017219883040935674, + "loss": 3.581, + "step": 4277 + }, + { + "epoch": 0.2251578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00017218713450292399, + "loss": 4.2163, + "step": 4278 + }, + { + "epoch": 0.22521052631578947, + "grad_norm": 0.52734375, + "learning_rate": 0.00017217543859649124, + "loss": 4.0946, + "step": 4279 + }, + { + "epoch": 0.22526315789473683, + "grad_norm": 0.55859375, + "learning_rate": 0.00017216374269005849, + "loss": 4.2512, + "step": 4280 + }, + { + "epoch": 0.22531578947368422, + "grad_norm": 0.69140625, + "learning_rate": 0.00017215204678362573, + "loss": 3.5445, + "step": 4281 + }, + { + "epoch": 0.22536842105263158, + "grad_norm": 0.5703125, + "learning_rate": 0.000172140350877193, + "loss": 3.6465, + "step": 4282 + }, + { + "epoch": 0.22542105263157894, + "grad_norm": 0.5, + "learning_rate": 0.00017212865497076026, + "loss": 4.4608, + "step": 4283 + }, + { + "epoch": 0.22547368421052633, + "grad_norm": 0.51953125, + "learning_rate": 0.0001721169590643275, + "loss": 4.0547, + "step": 4284 + }, + { + "epoch": 0.22552631578947369, + "grad_norm": 0.6328125, + "learning_rate": 0.00017210526315789476, + "loss": 4.1497, + "step": 4285 + }, + { + "epoch": 0.22557894736842105, + "grad_norm": 0.5625, + "learning_rate": 0.00017209356725146198, + "loss": 4.2349, + "step": 4286 + }, + { + "epoch": 0.22563157894736843, + "grad_norm": 0.578125, + "learning_rate": 0.00017208187134502923, + "loss": 4.0005, + "step": 4287 + }, + { + "epoch": 0.2256842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.00017207017543859648, + "loss": 4.0151, + "step": 4288 + }, + { + "epoch": 0.22573684210526315, + "grad_norm": 0.546875, + "learning_rate": 0.00017205847953216373, + "loss": 3.54, + "step": 4289 + }, + { + "epoch": 0.22578947368421054, + "grad_norm": 0.6171875, + "learning_rate": 0.000172046783625731, + "loss": 3.8652, + "step": 4290 + }, + { + "epoch": 0.2258421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00017203508771929826, + "loss": 3.7046, + "step": 4291 + }, + { + "epoch": 0.22589473684210526, + "grad_norm": 0.482421875, + "learning_rate": 0.0001720233918128655, + "loss": 4.3101, + "step": 4292 + }, + { + "epoch": 0.22594736842105262, + "grad_norm": 0.53515625, + "learning_rate": 0.00017201169590643276, + "loss": 4.0452, + "step": 4293 + }, + { + "epoch": 0.226, + "grad_norm": 0.53515625, + "learning_rate": 0.000172, + "loss": 3.977, + "step": 4294 + }, + { + "epoch": 0.22605263157894737, + "grad_norm": 0.5546875, + "learning_rate": 0.00017198830409356726, + "loss": 4.2572, + "step": 4295 + }, + { + "epoch": 0.22610526315789473, + "grad_norm": 0.68359375, + "learning_rate": 0.0001719766081871345, + "loss": 4.1319, + "step": 4296 + }, + { + "epoch": 0.2261578947368421, + "grad_norm": 0.63671875, + "learning_rate": 0.00017196491228070178, + "loss": 4.15, + "step": 4297 + }, + { + "epoch": 0.22621052631578947, + "grad_norm": 0.58984375, + "learning_rate": 0.00017195321637426903, + "loss": 3.9503, + "step": 4298 + }, + { + "epoch": 0.22626315789473683, + "grad_norm": 0.734375, + "learning_rate": 0.00017194152046783628, + "loss": 3.8657, + "step": 4299 + }, + { + "epoch": 0.22631578947368422, + "grad_norm": 1.5234375, + "learning_rate": 0.00017192982456140353, + "loss": 4.1522, + "step": 4300 + }, + { + "epoch": 0.22636842105263158, + "grad_norm": 0.46875, + "learning_rate": 0.00017191812865497075, + "loss": 3.7811, + "step": 4301 + }, + { + "epoch": 0.22642105263157894, + "grad_norm": 0.53515625, + "learning_rate": 0.000171906432748538, + "loss": 4.0985, + "step": 4302 + }, + { + "epoch": 0.22647368421052633, + "grad_norm": 0.69921875, + "learning_rate": 0.00017189473684210525, + "loss": 4.0454, + "step": 4303 + }, + { + "epoch": 0.2265263157894737, + "grad_norm": 0.60546875, + "learning_rate": 0.0001718830409356725, + "loss": 3.846, + "step": 4304 + }, + { + "epoch": 0.22657894736842105, + "grad_norm": 0.546875, + "learning_rate": 0.00017187134502923978, + "loss": 3.8551, + "step": 4305 + }, + { + "epoch": 0.22663157894736843, + "grad_norm": 0.58984375, + "learning_rate": 0.00017185964912280703, + "loss": 3.8015, + "step": 4306 + }, + { + "epoch": 0.2266842105263158, + "grad_norm": 0.5625, + "learning_rate": 0.00017184795321637428, + "loss": 3.7799, + "step": 4307 + }, + { + "epoch": 0.22673684210526315, + "grad_norm": 0.5703125, + "learning_rate": 0.00017183625730994153, + "loss": 3.8784, + "step": 4308 + }, + { + "epoch": 0.2267894736842105, + "grad_norm": 0.71875, + "learning_rate": 0.00017182456140350878, + "loss": 4.0197, + "step": 4309 + }, + { + "epoch": 0.2268421052631579, + "grad_norm": 0.6328125, + "learning_rate": 0.00017181286549707603, + "loss": 3.6731, + "step": 4310 + }, + { + "epoch": 0.22689473684210526, + "grad_norm": 0.62109375, + "learning_rate": 0.00017180116959064328, + "loss": 3.8657, + "step": 4311 + }, + { + "epoch": 0.22694736842105262, + "grad_norm": 0.5703125, + "learning_rate": 0.00017178947368421055, + "loss": 4.2054, + "step": 4312 + }, + { + "epoch": 0.227, + "grad_norm": 0.58203125, + "learning_rate": 0.0001717777777777778, + "loss": 4.2706, + "step": 4313 + }, + { + "epoch": 0.22705263157894737, + "grad_norm": 0.60546875, + "learning_rate": 0.00017176608187134505, + "loss": 4.4402, + "step": 4314 + }, + { + "epoch": 0.22710526315789473, + "grad_norm": 0.609375, + "learning_rate": 0.0001717543859649123, + "loss": 4.0825, + "step": 4315 + }, + { + "epoch": 0.22715789473684211, + "grad_norm": 0.57421875, + "learning_rate": 0.00017174269005847955, + "loss": 4.1852, + "step": 4316 + }, + { + "epoch": 0.22721052631578947, + "grad_norm": 0.5546875, + "learning_rate": 0.00017173099415204677, + "loss": 3.7117, + "step": 4317 + }, + { + "epoch": 0.22726315789473683, + "grad_norm": 0.76171875, + "learning_rate": 0.00017171929824561402, + "loss": 3.8767, + "step": 4318 + }, + { + "epoch": 0.22731578947368422, + "grad_norm": 0.55078125, + "learning_rate": 0.00017170760233918127, + "loss": 3.7797, + "step": 4319 + }, + { + "epoch": 0.22736842105263158, + "grad_norm": 0.71875, + "learning_rate": 0.00017169590643274855, + "loss": 3.8651, + "step": 4320 + }, + { + "epoch": 0.22742105263157894, + "grad_norm": 0.57421875, + "learning_rate": 0.0001716842105263158, + "loss": 4.1712, + "step": 4321 + }, + { + "epoch": 0.22747368421052633, + "grad_norm": 0.640625, + "learning_rate": 0.00017167251461988305, + "loss": 3.8838, + "step": 4322 + }, + { + "epoch": 0.2275263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.0001716608187134503, + "loss": 4.0153, + "step": 4323 + }, + { + "epoch": 0.22757894736842105, + "grad_norm": 0.52734375, + "learning_rate": 0.00017164912280701755, + "loss": 4.489, + "step": 4324 + }, + { + "epoch": 0.22763157894736843, + "grad_norm": 0.51171875, + "learning_rate": 0.0001716374269005848, + "loss": 3.8477, + "step": 4325 + }, + { + "epoch": 0.2276842105263158, + "grad_norm": 0.546875, + "learning_rate": 0.00017162573099415205, + "loss": 4.352, + "step": 4326 + }, + { + "epoch": 0.22773684210526315, + "grad_norm": 0.54296875, + "learning_rate": 0.00017161403508771932, + "loss": 4.4542, + "step": 4327 + }, + { + "epoch": 0.22778947368421051, + "grad_norm": 0.58984375, + "learning_rate": 0.00017160233918128657, + "loss": 3.9006, + "step": 4328 + }, + { + "epoch": 0.2278421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00017159064327485382, + "loss": 3.7196, + "step": 4329 + }, + { + "epoch": 0.22789473684210526, + "grad_norm": 0.451171875, + "learning_rate": 0.00017157894736842107, + "loss": 4.3503, + "step": 4330 + }, + { + "epoch": 0.22794736842105262, + "grad_norm": 0.53125, + "learning_rate": 0.00017156725146198832, + "loss": 3.9581, + "step": 4331 + }, + { + "epoch": 0.228, + "grad_norm": 0.50390625, + "learning_rate": 0.00017155555555555555, + "loss": 4.0193, + "step": 4332 + }, + { + "epoch": 0.22805263157894737, + "grad_norm": 0.5859375, + "learning_rate": 0.0001715438596491228, + "loss": 3.7332, + "step": 4333 + }, + { + "epoch": 0.22810526315789473, + "grad_norm": 0.58984375, + "learning_rate": 0.00017153216374269007, + "loss": 4.2221, + "step": 4334 + }, + { + "epoch": 0.22815789473684212, + "grad_norm": 0.578125, + "learning_rate": 0.00017152046783625732, + "loss": 3.7167, + "step": 4335 + }, + { + "epoch": 0.22821052631578947, + "grad_norm": 0.515625, + "learning_rate": 0.00017150877192982457, + "loss": 4.0764, + "step": 4336 + }, + { + "epoch": 0.22826315789473683, + "grad_norm": 0.578125, + "learning_rate": 0.00017149707602339182, + "loss": 4.2948, + "step": 4337 + }, + { + "epoch": 0.22831578947368422, + "grad_norm": 0.6484375, + "learning_rate": 0.00017148538011695907, + "loss": 4.5353, + "step": 4338 + }, + { + "epoch": 0.22836842105263158, + "grad_norm": 0.66796875, + "learning_rate": 0.00017147368421052632, + "loss": 4.2847, + "step": 4339 + }, + { + "epoch": 0.22842105263157894, + "grad_norm": 0.6875, + "learning_rate": 0.00017146198830409357, + "loss": 3.5073, + "step": 4340 + }, + { + "epoch": 0.22847368421052633, + "grad_norm": 0.62109375, + "learning_rate": 0.00017145029239766082, + "loss": 4.273, + "step": 4341 + }, + { + "epoch": 0.2285263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.0001714385964912281, + "loss": 4.0859, + "step": 4342 + }, + { + "epoch": 0.22857894736842105, + "grad_norm": 0.5234375, + "learning_rate": 0.00017142690058479535, + "loss": 3.9606, + "step": 4343 + }, + { + "epoch": 0.2286315789473684, + "grad_norm": 0.59765625, + "learning_rate": 0.0001714152046783626, + "loss": 3.8474, + "step": 4344 + }, + { + "epoch": 0.2286842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.00017140350877192984, + "loss": 3.8266, + "step": 4345 + }, + { + "epoch": 0.22873684210526316, + "grad_norm": 0.57421875, + "learning_rate": 0.0001713918128654971, + "loss": 4.1952, + "step": 4346 + }, + { + "epoch": 0.22878947368421051, + "grad_norm": 0.5625, + "learning_rate": 0.00017138011695906432, + "loss": 3.8947, + "step": 4347 + }, + { + "epoch": 0.2288421052631579, + "grad_norm": 0.60546875, + "learning_rate": 0.00017136842105263157, + "loss": 4.232, + "step": 4348 + }, + { + "epoch": 0.22889473684210526, + "grad_norm": 0.6796875, + "learning_rate": 0.00017135672514619884, + "loss": 3.894, + "step": 4349 + }, + { + "epoch": 0.22894736842105262, + "grad_norm": 0.51953125, + "learning_rate": 0.0001713450292397661, + "loss": 3.9115, + "step": 4350 + }, + { + "epoch": 0.229, + "grad_norm": 0.5, + "learning_rate": 0.00017133333333333334, + "loss": 3.867, + "step": 4351 + }, + { + "epoch": 0.22905263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.0001713216374269006, + "loss": 3.6352, + "step": 4352 + }, + { + "epoch": 0.22910526315789473, + "grad_norm": 0.53515625, + "learning_rate": 0.00017130994152046784, + "loss": 3.6416, + "step": 4353 + }, + { + "epoch": 0.22915789473684212, + "grad_norm": 0.578125, + "learning_rate": 0.0001712982456140351, + "loss": 3.6966, + "step": 4354 + }, + { + "epoch": 0.22921052631578948, + "grad_norm": 0.4921875, + "learning_rate": 0.00017128654970760234, + "loss": 4.2386, + "step": 4355 + }, + { + "epoch": 0.22926315789473684, + "grad_norm": 0.74609375, + "learning_rate": 0.0001712748538011696, + "loss": 4.1805, + "step": 4356 + }, + { + "epoch": 0.22931578947368422, + "grad_norm": 0.63671875, + "learning_rate": 0.00017126315789473687, + "loss": 3.7405, + "step": 4357 + }, + { + "epoch": 0.22936842105263158, + "grad_norm": 0.546875, + "learning_rate": 0.00017125146198830412, + "loss": 3.6571, + "step": 4358 + }, + { + "epoch": 0.22942105263157894, + "grad_norm": 0.53515625, + "learning_rate": 0.00017123976608187137, + "loss": 4.4789, + "step": 4359 + }, + { + "epoch": 0.2294736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00017122807017543862, + "loss": 4.1204, + "step": 4360 + }, + { + "epoch": 0.2295263157894737, + "grad_norm": 0.61328125, + "learning_rate": 0.00017121637426900586, + "loss": 4.3317, + "step": 4361 + }, + { + "epoch": 0.22957894736842105, + "grad_norm": 0.66015625, + "learning_rate": 0.00017120467836257311, + "loss": 3.849, + "step": 4362 + }, + { + "epoch": 0.2296315789473684, + "grad_norm": 0.5625, + "learning_rate": 0.00017119298245614034, + "loss": 3.6477, + "step": 4363 + }, + { + "epoch": 0.2296842105263158, + "grad_norm": 0.578125, + "learning_rate": 0.00017118128654970761, + "loss": 4.2894, + "step": 4364 + }, + { + "epoch": 0.22973684210526316, + "grad_norm": 0.56640625, + "learning_rate": 0.00017116959064327486, + "loss": 3.9505, + "step": 4365 + }, + { + "epoch": 0.22978947368421052, + "grad_norm": 0.5546875, + "learning_rate": 0.0001711578947368421, + "loss": 4.0128, + "step": 4366 + }, + { + "epoch": 0.2298421052631579, + "grad_norm": 0.58203125, + "learning_rate": 0.00017114619883040936, + "loss": 4.3032, + "step": 4367 + }, + { + "epoch": 0.22989473684210526, + "grad_norm": 0.4921875, + "learning_rate": 0.0001711345029239766, + "loss": 3.8017, + "step": 4368 + }, + { + "epoch": 0.22994736842105262, + "grad_norm": 0.52734375, + "learning_rate": 0.00017112280701754386, + "loss": 3.9283, + "step": 4369 + }, + { + "epoch": 0.23, + "grad_norm": 0.53125, + "learning_rate": 0.0001711111111111111, + "loss": 4.3796, + "step": 4370 + }, + { + "epoch": 0.23005263157894737, + "grad_norm": 0.60546875, + "learning_rate": 0.0001710994152046784, + "loss": 4.3832, + "step": 4371 + }, + { + "epoch": 0.23010526315789473, + "grad_norm": 0.58203125, + "learning_rate": 0.00017108771929824564, + "loss": 4.498, + "step": 4372 + }, + { + "epoch": 0.23015789473684212, + "grad_norm": 0.5703125, + "learning_rate": 0.0001710760233918129, + "loss": 4.1062, + "step": 4373 + }, + { + "epoch": 0.23021052631578948, + "grad_norm": 0.625, + "learning_rate": 0.00017106432748538014, + "loss": 4.6327, + "step": 4374 + }, + { + "epoch": 0.23026315789473684, + "grad_norm": 0.6640625, + "learning_rate": 0.00017105263157894739, + "loss": 4.0452, + "step": 4375 + }, + { + "epoch": 0.23031578947368422, + "grad_norm": 0.5859375, + "learning_rate": 0.00017104093567251464, + "loss": 3.6989, + "step": 4376 + }, + { + "epoch": 0.23036842105263158, + "grad_norm": 0.59765625, + "learning_rate": 0.00017102923976608189, + "loss": 3.557, + "step": 4377 + }, + { + "epoch": 0.23042105263157894, + "grad_norm": 0.80859375, + "learning_rate": 0.0001710175438596491, + "loss": 3.8229, + "step": 4378 + }, + { + "epoch": 0.2304736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00017100584795321638, + "loss": 3.7305, + "step": 4379 + }, + { + "epoch": 0.2305263157894737, + "grad_norm": 0.59375, + "learning_rate": 0.00017099415204678363, + "loss": 3.5708, + "step": 4380 + }, + { + "epoch": 0.23057894736842105, + "grad_norm": 0.52734375, + "learning_rate": 0.00017098245614035088, + "loss": 3.5642, + "step": 4381 + }, + { + "epoch": 0.2306315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00017097076023391813, + "loss": 4.5448, + "step": 4382 + }, + { + "epoch": 0.2306842105263158, + "grad_norm": 0.578125, + "learning_rate": 0.00017095906432748538, + "loss": 3.6282, + "step": 4383 + }, + { + "epoch": 0.23073684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00017094736842105263, + "loss": 4.1761, + "step": 4384 + }, + { + "epoch": 0.23078947368421052, + "grad_norm": 0.64453125, + "learning_rate": 0.00017093567251461988, + "loss": 4.1253, + "step": 4385 + }, + { + "epoch": 0.2308421052631579, + "grad_norm": 0.58203125, + "learning_rate": 0.00017092397660818716, + "loss": 3.8117, + "step": 4386 + }, + { + "epoch": 0.23089473684210526, + "grad_norm": 0.6796875, + "learning_rate": 0.0001709122807017544, + "loss": 3.6517, + "step": 4387 + }, + { + "epoch": 0.23094736842105262, + "grad_norm": 0.59375, + "learning_rate": 0.00017090058479532166, + "loss": 3.8646, + "step": 4388 + }, + { + "epoch": 0.231, + "grad_norm": 0.671875, + "learning_rate": 0.0001708888888888889, + "loss": 3.569, + "step": 4389 + }, + { + "epoch": 0.23105263157894737, + "grad_norm": 0.578125, + "learning_rate": 0.00017087719298245616, + "loss": 3.9557, + "step": 4390 + }, + { + "epoch": 0.23110526315789473, + "grad_norm": 0.578125, + "learning_rate": 0.0001708654970760234, + "loss": 3.8999, + "step": 4391 + }, + { + "epoch": 0.23115789473684212, + "grad_norm": 0.5390625, + "learning_rate": 0.00017085380116959066, + "loss": 3.9337, + "step": 4392 + }, + { + "epoch": 0.23121052631578948, + "grad_norm": 0.55859375, + "learning_rate": 0.00017084210526315788, + "loss": 4.0407, + "step": 4393 + }, + { + "epoch": 0.23126315789473684, + "grad_norm": 0.65234375, + "learning_rate": 0.00017083040935672516, + "loss": 4.324, + "step": 4394 + }, + { + "epoch": 0.2313157894736842, + "grad_norm": 0.546875, + "learning_rate": 0.0001708187134502924, + "loss": 4.0322, + "step": 4395 + }, + { + "epoch": 0.23136842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00017080701754385965, + "loss": 4.0259, + "step": 4396 + }, + { + "epoch": 0.23142105263157894, + "grad_norm": 0.5078125, + "learning_rate": 0.0001707953216374269, + "loss": 4.0859, + "step": 4397 + }, + { + "epoch": 0.2314736842105263, + "grad_norm": 0.58203125, + "learning_rate": 0.00017078362573099415, + "loss": 4.427, + "step": 4398 + }, + { + "epoch": 0.2315263157894737, + "grad_norm": 0.63671875, + "learning_rate": 0.0001707719298245614, + "loss": 3.9981, + "step": 4399 + }, + { + "epoch": 0.23157894736842105, + "grad_norm": 0.59765625, + "learning_rate": 0.00017076023391812865, + "loss": 3.8945, + "step": 4400 + }, + { + "epoch": 0.2316315789473684, + "grad_norm": 0.5625, + "learning_rate": 0.00017074853801169593, + "loss": 4.1086, + "step": 4401 + }, + { + "epoch": 0.2316842105263158, + "grad_norm": 0.5859375, + "learning_rate": 0.00017073684210526318, + "loss": 3.8484, + "step": 4402 + }, + { + "epoch": 0.23173684210526316, + "grad_norm": 0.52734375, + "learning_rate": 0.00017072514619883043, + "loss": 3.6619, + "step": 4403 + }, + { + "epoch": 0.23178947368421052, + "grad_norm": 0.494140625, + "learning_rate": 0.00017071345029239768, + "loss": 4.1227, + "step": 4404 + }, + { + "epoch": 0.2318421052631579, + "grad_norm": 0.484375, + "learning_rate": 0.00017070175438596493, + "loss": 4.0496, + "step": 4405 + }, + { + "epoch": 0.23189473684210526, + "grad_norm": 0.5390625, + "learning_rate": 0.00017069005847953218, + "loss": 3.382, + "step": 4406 + }, + { + "epoch": 0.23194736842105262, + "grad_norm": 0.5390625, + "learning_rate": 0.00017067836257309943, + "loss": 4.1733, + "step": 4407 + }, + { + "epoch": 0.232, + "grad_norm": 0.5078125, + "learning_rate": 0.00017066666666666668, + "loss": 3.6139, + "step": 4408 + }, + { + "epoch": 0.23205263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.00017065497076023393, + "loss": 3.8788, + "step": 4409 + }, + { + "epoch": 0.23210526315789473, + "grad_norm": 0.5546875, + "learning_rate": 0.00017064327485380118, + "loss": 4.1797, + "step": 4410 + }, + { + "epoch": 0.23215789473684212, + "grad_norm": 0.52734375, + "learning_rate": 0.00017063157894736843, + "loss": 3.7926, + "step": 4411 + }, + { + "epoch": 0.23221052631578948, + "grad_norm": 0.65234375, + "learning_rate": 0.00017061988304093568, + "loss": 4.2376, + "step": 4412 + }, + { + "epoch": 0.23226315789473684, + "grad_norm": 0.5625, + "learning_rate": 0.00017060818713450292, + "loss": 3.8478, + "step": 4413 + }, + { + "epoch": 0.2323157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00017059649122807017, + "loss": 3.8963, + "step": 4414 + }, + { + "epoch": 0.23236842105263159, + "grad_norm": 0.5390625, + "learning_rate": 0.00017058479532163742, + "loss": 3.7305, + "step": 4415 + }, + { + "epoch": 0.23242105263157894, + "grad_norm": 0.5390625, + "learning_rate": 0.0001705730994152047, + "loss": 3.8724, + "step": 4416 + }, + { + "epoch": 0.2324736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00017056140350877195, + "loss": 4.1084, + "step": 4417 + }, + { + "epoch": 0.2325263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.0001705497076023392, + "loss": 3.6979, + "step": 4418 + }, + { + "epoch": 0.23257894736842105, + "grad_norm": 0.53125, + "learning_rate": 0.00017053801169590645, + "loss": 3.9849, + "step": 4419 + }, + { + "epoch": 0.2326315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.0001705263157894737, + "loss": 3.878, + "step": 4420 + }, + { + "epoch": 0.2326842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00017051461988304095, + "loss": 4.1644, + "step": 4421 + }, + { + "epoch": 0.23273684210526316, + "grad_norm": 0.54296875, + "learning_rate": 0.0001705029239766082, + "loss": 3.6709, + "step": 4422 + }, + { + "epoch": 0.23278947368421052, + "grad_norm": 0.474609375, + "learning_rate": 0.00017049122807017545, + "loss": 4.0139, + "step": 4423 + }, + { + "epoch": 0.2328421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.0001704795321637427, + "loss": 3.8312, + "step": 4424 + }, + { + "epoch": 0.23289473684210527, + "grad_norm": 0.58203125, + "learning_rate": 0.00017046783625730995, + "loss": 3.5528, + "step": 4425 + }, + { + "epoch": 0.23294736842105263, + "grad_norm": 0.59765625, + "learning_rate": 0.0001704561403508772, + "loss": 4.3045, + "step": 4426 + }, + { + "epoch": 0.233, + "grad_norm": 0.671875, + "learning_rate": 0.00017044444444444445, + "loss": 3.7974, + "step": 4427 + }, + { + "epoch": 0.23305263157894737, + "grad_norm": 0.56640625, + "learning_rate": 0.0001704327485380117, + "loss": 4.2348, + "step": 4428 + }, + { + "epoch": 0.23310526315789473, + "grad_norm": 0.734375, + "learning_rate": 0.00017042105263157895, + "loss": 4.1004, + "step": 4429 + }, + { + "epoch": 0.2331578947368421, + "grad_norm": 0.46875, + "learning_rate": 0.0001704093567251462, + "loss": 3.9742, + "step": 4430 + }, + { + "epoch": 0.23321052631578948, + "grad_norm": 0.5625, + "learning_rate": 0.00017039766081871347, + "loss": 3.9513, + "step": 4431 + }, + { + "epoch": 0.23326315789473684, + "grad_norm": 0.578125, + "learning_rate": 0.00017038596491228072, + "loss": 3.9051, + "step": 4432 + }, + { + "epoch": 0.2333157894736842, + "grad_norm": 0.55859375, + "learning_rate": 0.00017037426900584797, + "loss": 3.9341, + "step": 4433 + }, + { + "epoch": 0.23336842105263159, + "grad_norm": 0.5859375, + "learning_rate": 0.00017036257309941522, + "loss": 3.6738, + "step": 4434 + }, + { + "epoch": 0.23342105263157895, + "grad_norm": 0.53515625, + "learning_rate": 0.00017035087719298247, + "loss": 3.8725, + "step": 4435 + }, + { + "epoch": 0.2334736842105263, + "grad_norm": 1.0546875, + "learning_rate": 0.00017033918128654972, + "loss": 4.1667, + "step": 4436 + }, + { + "epoch": 0.2335263157894737, + "grad_norm": 0.6953125, + "learning_rate": 0.00017032748538011697, + "loss": 3.962, + "step": 4437 + }, + { + "epoch": 0.23357894736842105, + "grad_norm": 0.6640625, + "learning_rate": 0.00017031578947368422, + "loss": 3.8532, + "step": 4438 + }, + { + "epoch": 0.2336315789473684, + "grad_norm": 0.494140625, + "learning_rate": 0.00017030409356725147, + "loss": 4.2354, + "step": 4439 + }, + { + "epoch": 0.2336842105263158, + "grad_norm": 0.51953125, + "learning_rate": 0.00017029239766081872, + "loss": 3.7193, + "step": 4440 + }, + { + "epoch": 0.23373684210526316, + "grad_norm": 0.5, + "learning_rate": 0.00017028070175438597, + "loss": 4.3746, + "step": 4441 + }, + { + "epoch": 0.23378947368421052, + "grad_norm": 0.5390625, + "learning_rate": 0.00017026900584795322, + "loss": 3.71, + "step": 4442 + }, + { + "epoch": 0.2338421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00017025730994152047, + "loss": 3.8183, + "step": 4443 + }, + { + "epoch": 0.23389473684210527, + "grad_norm": 0.515625, + "learning_rate": 0.00017024561403508772, + "loss": 4.2946, + "step": 4444 + }, + { + "epoch": 0.23394736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.000170233918128655, + "loss": 4.2361, + "step": 4445 + }, + { + "epoch": 0.234, + "grad_norm": 0.53125, + "learning_rate": 0.00017022222222222224, + "loss": 3.9568, + "step": 4446 + }, + { + "epoch": 0.23405263157894737, + "grad_norm": 0.60546875, + "learning_rate": 0.0001702105263157895, + "loss": 3.8721, + "step": 4447 + }, + { + "epoch": 0.23410526315789473, + "grad_norm": 0.5078125, + "learning_rate": 0.00017019883040935674, + "loss": 4.2133, + "step": 4448 + }, + { + "epoch": 0.2341578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.000170187134502924, + "loss": 4.2425, + "step": 4449 + }, + { + "epoch": 0.23421052631578948, + "grad_norm": 0.5234375, + "learning_rate": 0.00017017543859649124, + "loss": 3.9232, + "step": 4450 + }, + { + "epoch": 0.23426315789473684, + "grad_norm": 0.51953125, + "learning_rate": 0.0001701637426900585, + "loss": 3.8543, + "step": 4451 + }, + { + "epoch": 0.2343157894736842, + "grad_norm": 0.62109375, + "learning_rate": 0.00017015204678362574, + "loss": 3.588, + "step": 4452 + }, + { + "epoch": 0.2343684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.000170140350877193, + "loss": 4.2403, + "step": 4453 + }, + { + "epoch": 0.23442105263157895, + "grad_norm": 0.7265625, + "learning_rate": 0.00017012865497076024, + "loss": 3.6784, + "step": 4454 + }, + { + "epoch": 0.2344736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.0001701169590643275, + "loss": 3.6635, + "step": 4455 + }, + { + "epoch": 0.2345263157894737, + "grad_norm": 0.55078125, + "learning_rate": 0.00017010526315789474, + "loss": 3.9592, + "step": 4456 + }, + { + "epoch": 0.23457894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.000170093567251462, + "loss": 3.7885, + "step": 4457 + }, + { + "epoch": 0.2346315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.00017008187134502924, + "loss": 3.929, + "step": 4458 + }, + { + "epoch": 0.2346842105263158, + "grad_norm": 0.68359375, + "learning_rate": 0.0001700701754385965, + "loss": 3.8675, + "step": 4459 + }, + { + "epoch": 0.23473684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00017005847953216376, + "loss": 4.1187, + "step": 4460 + }, + { + "epoch": 0.23478947368421052, + "grad_norm": 0.68359375, + "learning_rate": 0.00017004678362573101, + "loss": 4.1047, + "step": 4461 + }, + { + "epoch": 0.2348421052631579, + "grad_norm": 0.609375, + "learning_rate": 0.00017003508771929826, + "loss": 3.6045, + "step": 4462 + }, + { + "epoch": 0.23489473684210527, + "grad_norm": 0.66796875, + "learning_rate": 0.0001700233918128655, + "loss": 3.9083, + "step": 4463 + }, + { + "epoch": 0.23494736842105263, + "grad_norm": 0.63671875, + "learning_rate": 0.00017001169590643276, + "loss": 3.3172, + "step": 4464 + }, + { + "epoch": 0.235, + "grad_norm": 0.53125, + "learning_rate": 0.00017, + "loss": 3.4757, + "step": 4465 + }, + { + "epoch": 0.23505263157894737, + "grad_norm": 0.640625, + "learning_rate": 0.00016998830409356726, + "loss": 4.0264, + "step": 4466 + }, + { + "epoch": 0.23510526315789473, + "grad_norm": 0.5234375, + "learning_rate": 0.0001699766081871345, + "loss": 4.2238, + "step": 4467 + }, + { + "epoch": 0.2351578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00016996491228070176, + "loss": 3.9906, + "step": 4468 + }, + { + "epoch": 0.23521052631578948, + "grad_norm": 0.5625, + "learning_rate": 0.000169953216374269, + "loss": 3.9978, + "step": 4469 + }, + { + "epoch": 0.23526315789473684, + "grad_norm": 0.5703125, + "learning_rate": 0.00016994152046783626, + "loss": 3.8193, + "step": 4470 + }, + { + "epoch": 0.2353157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.0001699298245614035, + "loss": 4.2184, + "step": 4471 + }, + { + "epoch": 0.2353684210526316, + "grad_norm": 0.6953125, + "learning_rate": 0.00016991812865497076, + "loss": 3.4411, + "step": 4472 + }, + { + "epoch": 0.23542105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.000169906432748538, + "loss": 3.5446, + "step": 4473 + }, + { + "epoch": 0.2354736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.00016989473684210526, + "loss": 3.9465, + "step": 4474 + }, + { + "epoch": 0.2355263157894737, + "grad_norm": 0.58203125, + "learning_rate": 0.00016988304093567254, + "loss": 3.6619, + "step": 4475 + }, + { + "epoch": 0.23557894736842105, + "grad_norm": 0.5546875, + "learning_rate": 0.00016987134502923978, + "loss": 3.7777, + "step": 4476 + }, + { + "epoch": 0.23563157894736841, + "grad_norm": 0.5078125, + "learning_rate": 0.00016985964912280703, + "loss": 4.0701, + "step": 4477 + }, + { + "epoch": 0.2356842105263158, + "grad_norm": 0.5546875, + "learning_rate": 0.00016984795321637428, + "loss": 3.5562, + "step": 4478 + }, + { + "epoch": 0.23573684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00016983625730994153, + "loss": 4.3567, + "step": 4479 + }, + { + "epoch": 0.23578947368421052, + "grad_norm": 0.60546875, + "learning_rate": 0.00016982456140350878, + "loss": 3.7479, + "step": 4480 + }, + { + "epoch": 0.2358421052631579, + "grad_norm": 1.25, + "learning_rate": 0.00016981286549707603, + "loss": 3.6349, + "step": 4481 + }, + { + "epoch": 0.23589473684210527, + "grad_norm": 0.5546875, + "learning_rate": 0.00016980116959064328, + "loss": 3.9561, + "step": 4482 + }, + { + "epoch": 0.23594736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.00016978947368421053, + "loss": 4.1657, + "step": 4483 + }, + { + "epoch": 0.236, + "grad_norm": 0.58984375, + "learning_rate": 0.00016977777777777778, + "loss": 4.2189, + "step": 4484 + }, + { + "epoch": 0.23605263157894738, + "grad_norm": 0.61328125, + "learning_rate": 0.00016976608187134503, + "loss": 4.0427, + "step": 4485 + }, + { + "epoch": 0.23610526315789473, + "grad_norm": 0.57421875, + "learning_rate": 0.00016975438596491228, + "loss": 3.7941, + "step": 4486 + }, + { + "epoch": 0.2361578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00016974269005847953, + "loss": 4.0608, + "step": 4487 + }, + { + "epoch": 0.23621052631578948, + "grad_norm": 0.5859375, + "learning_rate": 0.00016973099415204678, + "loss": 3.9612, + "step": 4488 + }, + { + "epoch": 0.23626315789473684, + "grad_norm": 0.59375, + "learning_rate": 0.00016971929824561403, + "loss": 3.9324, + "step": 4489 + }, + { + "epoch": 0.2363157894736842, + "grad_norm": 0.66015625, + "learning_rate": 0.0001697076023391813, + "loss": 3.7253, + "step": 4490 + }, + { + "epoch": 0.2363684210526316, + "grad_norm": 0.515625, + "learning_rate": 0.00016969590643274856, + "loss": 3.9382, + "step": 4491 + }, + { + "epoch": 0.23642105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.0001696842105263158, + "loss": 4.1052, + "step": 4492 + }, + { + "epoch": 0.2364736842105263, + "grad_norm": 0.5, + "learning_rate": 0.00016967251461988306, + "loss": 4.0705, + "step": 4493 + }, + { + "epoch": 0.2365263157894737, + "grad_norm": 0.515625, + "learning_rate": 0.0001696608187134503, + "loss": 3.7396, + "step": 4494 + }, + { + "epoch": 0.23657894736842106, + "grad_norm": 0.5703125, + "learning_rate": 0.00016964912280701755, + "loss": 3.2357, + "step": 4495 + }, + { + "epoch": 0.23663157894736842, + "grad_norm": 0.68359375, + "learning_rate": 0.0001696374269005848, + "loss": 3.7383, + "step": 4496 + }, + { + "epoch": 0.2366842105263158, + "grad_norm": 0.5546875, + "learning_rate": 0.00016962573099415205, + "loss": 3.906, + "step": 4497 + }, + { + "epoch": 0.23673684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.0001696140350877193, + "loss": 4.0863, + "step": 4498 + }, + { + "epoch": 0.23678947368421052, + "grad_norm": 0.58203125, + "learning_rate": 0.00016960233918128655, + "loss": 3.6169, + "step": 4499 + }, + { + "epoch": 0.23684210526315788, + "grad_norm": 0.515625, + "learning_rate": 0.0001695906432748538, + "loss": 3.9195, + "step": 4500 + }, + { + "epoch": 0.23689473684210527, + "grad_norm": 0.50390625, + "learning_rate": 0.00016957894736842105, + "loss": 3.8801, + "step": 4501 + }, + { + "epoch": 0.23694736842105263, + "grad_norm": 0.5625, + "learning_rate": 0.0001695672514619883, + "loss": 3.9507, + "step": 4502 + }, + { + "epoch": 0.237, + "grad_norm": 0.56640625, + "learning_rate": 0.00016955555555555555, + "loss": 4.3412, + "step": 4503 + }, + { + "epoch": 0.23705263157894738, + "grad_norm": 0.59375, + "learning_rate": 0.0001695438596491228, + "loss": 3.7632, + "step": 4504 + }, + { + "epoch": 0.23710526315789474, + "grad_norm": 0.52734375, + "learning_rate": 0.00016953216374269008, + "loss": 4.3183, + "step": 4505 + }, + { + "epoch": 0.2371578947368421, + "grad_norm": 0.609375, + "learning_rate": 0.00016952046783625733, + "loss": 4.0676, + "step": 4506 + }, + { + "epoch": 0.23721052631578948, + "grad_norm": 0.55859375, + "learning_rate": 0.00016950877192982458, + "loss": 3.6849, + "step": 4507 + }, + { + "epoch": 0.23726315789473684, + "grad_norm": 0.498046875, + "learning_rate": 0.00016949707602339183, + "loss": 3.759, + "step": 4508 + }, + { + "epoch": 0.2373157894736842, + "grad_norm": 0.57421875, + "learning_rate": 0.00016948538011695908, + "loss": 3.8631, + "step": 4509 + }, + { + "epoch": 0.2373684210526316, + "grad_norm": 0.8515625, + "learning_rate": 0.00016947368421052633, + "loss": 3.452, + "step": 4510 + }, + { + "epoch": 0.23742105263157895, + "grad_norm": 0.458984375, + "learning_rate": 0.00016946198830409357, + "loss": 3.9085, + "step": 4511 + }, + { + "epoch": 0.2374736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.00016945029239766082, + "loss": 3.8478, + "step": 4512 + }, + { + "epoch": 0.2375263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00016943859649122807, + "loss": 3.9128, + "step": 4513 + }, + { + "epoch": 0.23757894736842106, + "grad_norm": 0.54296875, + "learning_rate": 0.00016942690058479532, + "loss": 4.2087, + "step": 4514 + }, + { + "epoch": 0.23763157894736842, + "grad_norm": 0.66015625, + "learning_rate": 0.00016941520467836257, + "loss": 3.7767, + "step": 4515 + }, + { + "epoch": 0.23768421052631578, + "grad_norm": 0.60546875, + "learning_rate": 0.00016940350877192982, + "loss": 4.1082, + "step": 4516 + }, + { + "epoch": 0.23773684210526316, + "grad_norm": 0.6328125, + "learning_rate": 0.00016939181286549707, + "loss": 4.1412, + "step": 4517 + }, + { + "epoch": 0.23778947368421052, + "grad_norm": 0.65234375, + "learning_rate": 0.00016938011695906432, + "loss": 3.8107, + "step": 4518 + }, + { + "epoch": 0.23784210526315788, + "grad_norm": 0.625, + "learning_rate": 0.00016936842105263157, + "loss": 4.4707, + "step": 4519 + }, + { + "epoch": 0.23789473684210527, + "grad_norm": 0.60546875, + "learning_rate": 0.00016935672514619885, + "loss": 4.0316, + "step": 4520 + }, + { + "epoch": 0.23794736842105263, + "grad_norm": 0.60546875, + "learning_rate": 0.0001693450292397661, + "loss": 4.0709, + "step": 4521 + }, + { + "epoch": 0.238, + "grad_norm": 0.57421875, + "learning_rate": 0.00016933333333333335, + "loss": 3.9487, + "step": 4522 + }, + { + "epoch": 0.23805263157894738, + "grad_norm": 0.5390625, + "learning_rate": 0.0001693216374269006, + "loss": 4.0143, + "step": 4523 + }, + { + "epoch": 0.23810526315789474, + "grad_norm": 0.59765625, + "learning_rate": 0.00016930994152046785, + "loss": 4.0282, + "step": 4524 + }, + { + "epoch": 0.2381578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.0001692982456140351, + "loss": 3.892, + "step": 4525 + }, + { + "epoch": 0.23821052631578948, + "grad_norm": 0.53515625, + "learning_rate": 0.00016928654970760235, + "loss": 3.6186, + "step": 4526 + }, + { + "epoch": 0.23826315789473684, + "grad_norm": 0.49609375, + "learning_rate": 0.0001692748538011696, + "loss": 4.2917, + "step": 4527 + }, + { + "epoch": 0.2383157894736842, + "grad_norm": 0.546875, + "learning_rate": 0.00016926315789473684, + "loss": 4.203, + "step": 4528 + }, + { + "epoch": 0.2383684210526316, + "grad_norm": 0.56640625, + "learning_rate": 0.0001692514619883041, + "loss": 3.7105, + "step": 4529 + }, + { + "epoch": 0.23842105263157895, + "grad_norm": 0.58984375, + "learning_rate": 0.00016923976608187134, + "loss": 3.3924, + "step": 4530 + }, + { + "epoch": 0.2384736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.0001692280701754386, + "loss": 4.1472, + "step": 4531 + }, + { + "epoch": 0.2385263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.00016921637426900584, + "loss": 4.3196, + "step": 4532 + }, + { + "epoch": 0.23857894736842106, + "grad_norm": 0.58203125, + "learning_rate": 0.0001692046783625731, + "loss": 4.3535, + "step": 4533 + }, + { + "epoch": 0.23863157894736842, + "grad_norm": 0.6015625, + "learning_rate": 0.00016919298245614037, + "loss": 4.2682, + "step": 4534 + }, + { + "epoch": 0.23868421052631578, + "grad_norm": 0.5625, + "learning_rate": 0.00016918128654970762, + "loss": 3.92, + "step": 4535 + }, + { + "epoch": 0.23873684210526316, + "grad_norm": 0.59375, + "learning_rate": 0.00016916959064327487, + "loss": 3.715, + "step": 4536 + }, + { + "epoch": 0.23878947368421052, + "grad_norm": 0.56640625, + "learning_rate": 0.00016915789473684212, + "loss": 3.5957, + "step": 4537 + }, + { + "epoch": 0.23884210526315788, + "grad_norm": 0.54296875, + "learning_rate": 0.00016914619883040937, + "loss": 4.1854, + "step": 4538 + }, + { + "epoch": 0.23889473684210527, + "grad_norm": 0.54296875, + "learning_rate": 0.00016913450292397662, + "loss": 3.5002, + "step": 4539 + }, + { + "epoch": 0.23894736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00016912280701754387, + "loss": 3.9196, + "step": 4540 + }, + { + "epoch": 0.239, + "grad_norm": 0.53515625, + "learning_rate": 0.00016911111111111112, + "loss": 3.9442, + "step": 4541 + }, + { + "epoch": 0.23905263157894738, + "grad_norm": 0.53125, + "learning_rate": 0.00016909941520467837, + "loss": 3.945, + "step": 4542 + }, + { + "epoch": 0.23910526315789474, + "grad_norm": 0.466796875, + "learning_rate": 0.00016908771929824562, + "loss": 4.0298, + "step": 4543 + }, + { + "epoch": 0.2391578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.00016907602339181287, + "loss": 3.9428, + "step": 4544 + }, + { + "epoch": 0.23921052631578948, + "grad_norm": 0.55078125, + "learning_rate": 0.00016906432748538012, + "loss": 3.8421, + "step": 4545 + }, + { + "epoch": 0.23926315789473684, + "grad_norm": 0.58984375, + "learning_rate": 0.00016905263157894736, + "loss": 3.6494, + "step": 4546 + }, + { + "epoch": 0.2393157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.00016904093567251461, + "loss": 4.1882, + "step": 4547 + }, + { + "epoch": 0.2393684210526316, + "grad_norm": 0.68359375, + "learning_rate": 0.00016902923976608186, + "loss": 3.5169, + "step": 4548 + }, + { + "epoch": 0.23942105263157895, + "grad_norm": 0.49609375, + "learning_rate": 0.00016901754385964914, + "loss": 3.7525, + "step": 4549 + }, + { + "epoch": 0.2394736842105263, + "grad_norm": 0.62109375, + "learning_rate": 0.0001690058479532164, + "loss": 3.3898, + "step": 4550 + }, + { + "epoch": 0.23952631578947367, + "grad_norm": 0.578125, + "learning_rate": 0.00016899415204678364, + "loss": 3.8026, + "step": 4551 + }, + { + "epoch": 0.23957894736842106, + "grad_norm": 0.6640625, + "learning_rate": 0.0001689824561403509, + "loss": 4.0576, + "step": 4552 + }, + { + "epoch": 0.23963157894736842, + "grad_norm": 0.54296875, + "learning_rate": 0.00016897076023391814, + "loss": 3.8993, + "step": 4553 + }, + { + "epoch": 0.23968421052631578, + "grad_norm": 0.58203125, + "learning_rate": 0.0001689590643274854, + "loss": 4.1294, + "step": 4554 + }, + { + "epoch": 0.23973684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.00016894736842105264, + "loss": 3.6221, + "step": 4555 + }, + { + "epoch": 0.23978947368421052, + "grad_norm": 0.5234375, + "learning_rate": 0.0001689356725146199, + "loss": 3.8145, + "step": 4556 + }, + { + "epoch": 0.23984210526315788, + "grad_norm": 0.65625, + "learning_rate": 0.00016892397660818716, + "loss": 3.6462, + "step": 4557 + }, + { + "epoch": 0.23989473684210527, + "grad_norm": 0.46875, + "learning_rate": 0.0001689122807017544, + "loss": 3.9836, + "step": 4558 + }, + { + "epoch": 0.23994736842105263, + "grad_norm": 0.5, + "learning_rate": 0.00016890058479532164, + "loss": 3.8728, + "step": 4559 + }, + { + "epoch": 0.24, + "grad_norm": 0.484375, + "learning_rate": 0.00016888888888888889, + "loss": 3.7361, + "step": 4560 + }, + { + "epoch": 0.24005263157894738, + "grad_norm": 0.5546875, + "learning_rate": 0.00016887719298245614, + "loss": 4.0159, + "step": 4561 + }, + { + "epoch": 0.24010526315789474, + "grad_norm": 0.578125, + "learning_rate": 0.00016886549707602339, + "loss": 3.5823, + "step": 4562 + }, + { + "epoch": 0.2401578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.00016885380116959063, + "loss": 3.753, + "step": 4563 + }, + { + "epoch": 0.24021052631578949, + "grad_norm": 0.57421875, + "learning_rate": 0.0001688421052631579, + "loss": 3.6865, + "step": 4564 + }, + { + "epoch": 0.24026315789473685, + "grad_norm": 0.52734375, + "learning_rate": 0.00016883040935672516, + "loss": 3.403, + "step": 4565 + }, + { + "epoch": 0.2403157894736842, + "grad_norm": 0.52734375, + "learning_rate": 0.0001688187134502924, + "loss": 3.9868, + "step": 4566 + }, + { + "epoch": 0.2403684210526316, + "grad_norm": 0.62109375, + "learning_rate": 0.00016880701754385966, + "loss": 4.1333, + "step": 4567 + }, + { + "epoch": 0.24042105263157895, + "grad_norm": 0.58984375, + "learning_rate": 0.0001687953216374269, + "loss": 3.557, + "step": 4568 + }, + { + "epoch": 0.2404736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.00016878362573099416, + "loss": 4.2235, + "step": 4569 + }, + { + "epoch": 0.24052631578947367, + "grad_norm": 0.51953125, + "learning_rate": 0.0001687719298245614, + "loss": 4.1009, + "step": 4570 + }, + { + "epoch": 0.24057894736842106, + "grad_norm": 0.546875, + "learning_rate": 0.00016876023391812869, + "loss": 4.2706, + "step": 4571 + }, + { + "epoch": 0.24063157894736842, + "grad_norm": 0.5703125, + "learning_rate": 0.00016874853801169594, + "loss": 3.8906, + "step": 4572 + }, + { + "epoch": 0.24068421052631578, + "grad_norm": 0.53515625, + "learning_rate": 0.00016873684210526316, + "loss": 4.0145, + "step": 4573 + }, + { + "epoch": 0.24073684210526317, + "grad_norm": 0.5234375, + "learning_rate": 0.0001687251461988304, + "loss": 4.1479, + "step": 4574 + }, + { + "epoch": 0.24078947368421053, + "grad_norm": 0.53515625, + "learning_rate": 0.00016871345029239766, + "loss": 3.9307, + "step": 4575 + }, + { + "epoch": 0.24084210526315789, + "grad_norm": 0.52734375, + "learning_rate": 0.0001687017543859649, + "loss": 3.7782, + "step": 4576 + }, + { + "epoch": 0.24089473684210527, + "grad_norm": 0.578125, + "learning_rate": 0.00016869005847953216, + "loss": 4.0893, + "step": 4577 + }, + { + "epoch": 0.24094736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.0001686783625730994, + "loss": 3.76, + "step": 4578 + }, + { + "epoch": 0.241, + "grad_norm": 0.5625, + "learning_rate": 0.00016866666666666668, + "loss": 3.554, + "step": 4579 + }, + { + "epoch": 0.24105263157894738, + "grad_norm": 0.466796875, + "learning_rate": 0.00016865497076023393, + "loss": 3.8988, + "step": 4580 + }, + { + "epoch": 0.24110526315789474, + "grad_norm": 0.64453125, + "learning_rate": 0.00016864327485380118, + "loss": 3.4587, + "step": 4581 + }, + { + "epoch": 0.2411578947368421, + "grad_norm": 0.625, + "learning_rate": 0.00016863157894736843, + "loss": 3.9311, + "step": 4582 + }, + { + "epoch": 0.2412105263157895, + "grad_norm": 1.1953125, + "learning_rate": 0.00016861988304093568, + "loss": 4.0908, + "step": 4583 + }, + { + "epoch": 0.24126315789473685, + "grad_norm": 0.6015625, + "learning_rate": 0.00016860818713450293, + "loss": 4.1232, + "step": 4584 + }, + { + "epoch": 0.2413157894736842, + "grad_norm": 0.5625, + "learning_rate": 0.00016859649122807018, + "loss": 4.1409, + "step": 4585 + }, + { + "epoch": 0.24136842105263157, + "grad_norm": 0.5078125, + "learning_rate": 0.00016858479532163746, + "loss": 3.8839, + "step": 4586 + }, + { + "epoch": 0.24142105263157895, + "grad_norm": 0.57421875, + "learning_rate": 0.0001685730994152047, + "loss": 3.5769, + "step": 4587 + }, + { + "epoch": 0.2414736842105263, + "grad_norm": 0.5234375, + "learning_rate": 0.00016856140350877193, + "loss": 3.7763, + "step": 4588 + }, + { + "epoch": 0.24152631578947367, + "grad_norm": 0.5078125, + "learning_rate": 0.00016854970760233918, + "loss": 4.3071, + "step": 4589 + }, + { + "epoch": 0.24157894736842106, + "grad_norm": 0.53515625, + "learning_rate": 0.00016853801169590643, + "loss": 3.8976, + "step": 4590 + }, + { + "epoch": 0.24163157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00016852631578947368, + "loss": 4.0685, + "step": 4591 + }, + { + "epoch": 0.24168421052631578, + "grad_norm": 0.50390625, + "learning_rate": 0.00016851461988304093, + "loss": 4.1363, + "step": 4592 + }, + { + "epoch": 0.24173684210526317, + "grad_norm": 0.49609375, + "learning_rate": 0.00016850292397660818, + "loss": 4.2198, + "step": 4593 + }, + { + "epoch": 0.24178947368421053, + "grad_norm": 0.57421875, + "learning_rate": 0.00016849122807017545, + "loss": 3.8621, + "step": 4594 + }, + { + "epoch": 0.24184210526315789, + "grad_norm": 0.59375, + "learning_rate": 0.0001684795321637427, + "loss": 4.0879, + "step": 4595 + }, + { + "epoch": 0.24189473684210527, + "grad_norm": 0.76171875, + "learning_rate": 0.00016846783625730995, + "loss": 3.832, + "step": 4596 + }, + { + "epoch": 0.24194736842105263, + "grad_norm": 0.578125, + "learning_rate": 0.0001684561403508772, + "loss": 3.9549, + "step": 4597 + }, + { + "epoch": 0.242, + "grad_norm": 0.52734375, + "learning_rate": 0.00016844444444444445, + "loss": 4.0242, + "step": 4598 + }, + { + "epoch": 0.24205263157894738, + "grad_norm": 0.6015625, + "learning_rate": 0.0001684327485380117, + "loss": 4.3283, + "step": 4599 + }, + { + "epoch": 0.24210526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.00016842105263157895, + "loss": 3.6203, + "step": 4600 + }, + { + "epoch": 0.2421578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00016840935672514623, + "loss": 3.4898, + "step": 4601 + }, + { + "epoch": 0.2422105263157895, + "grad_norm": 0.59375, + "learning_rate": 0.00016839766081871348, + "loss": 3.7408, + "step": 4602 + }, + { + "epoch": 0.24226315789473685, + "grad_norm": 0.625, + "learning_rate": 0.00016838596491228073, + "loss": 4.2134, + "step": 4603 + }, + { + "epoch": 0.2423157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.00016837426900584795, + "loss": 3.5589, + "step": 4604 + }, + { + "epoch": 0.24236842105263157, + "grad_norm": 0.546875, + "learning_rate": 0.0001683625730994152, + "loss": 4.022, + "step": 4605 + }, + { + "epoch": 0.24242105263157895, + "grad_norm": 0.6171875, + "learning_rate": 0.00016835087719298245, + "loss": 4.0206, + "step": 4606 + }, + { + "epoch": 0.2424736842105263, + "grad_norm": 0.578125, + "learning_rate": 0.0001683391812865497, + "loss": 4.2074, + "step": 4607 + }, + { + "epoch": 0.24252631578947367, + "grad_norm": 0.5546875, + "learning_rate": 0.00016832748538011698, + "loss": 3.8255, + "step": 4608 + }, + { + "epoch": 0.24257894736842106, + "grad_norm": 0.5390625, + "learning_rate": 0.00016831578947368422, + "loss": 4.1548, + "step": 4609 + }, + { + "epoch": 0.24263157894736842, + "grad_norm": 0.5625, + "learning_rate": 0.00016830409356725147, + "loss": 3.8668, + "step": 4610 + }, + { + "epoch": 0.24268421052631578, + "grad_norm": 0.58984375, + "learning_rate": 0.00016829239766081872, + "loss": 4.3792, + "step": 4611 + }, + { + "epoch": 0.24273684210526317, + "grad_norm": 0.6015625, + "learning_rate": 0.00016828070175438597, + "loss": 3.7887, + "step": 4612 + }, + { + "epoch": 0.24278947368421053, + "grad_norm": 0.58203125, + "learning_rate": 0.00016826900584795322, + "loss": 4.0995, + "step": 4613 + }, + { + "epoch": 0.2428421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.00016825730994152047, + "loss": 4.032, + "step": 4614 + }, + { + "epoch": 0.24289473684210527, + "grad_norm": 0.59765625, + "learning_rate": 0.00016824561403508772, + "loss": 4.1824, + "step": 4615 + }, + { + "epoch": 0.24294736842105263, + "grad_norm": 0.60546875, + "learning_rate": 0.000168233918128655, + "loss": 4.2524, + "step": 4616 + }, + { + "epoch": 0.243, + "grad_norm": 0.62109375, + "learning_rate": 0.00016822222222222225, + "loss": 4.3415, + "step": 4617 + }, + { + "epoch": 0.24305263157894738, + "grad_norm": 0.609375, + "learning_rate": 0.0001682105263157895, + "loss": 4.1393, + "step": 4618 + }, + { + "epoch": 0.24310526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.00016819883040935672, + "loss": 3.7671, + "step": 4619 + }, + { + "epoch": 0.2431578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00016818713450292397, + "loss": 4.0088, + "step": 4620 + }, + { + "epoch": 0.24321052631578946, + "grad_norm": 0.57421875, + "learning_rate": 0.00016817543859649122, + "loss": 3.6932, + "step": 4621 + }, + { + "epoch": 0.24326315789473685, + "grad_norm": 0.68359375, + "learning_rate": 0.00016816374269005847, + "loss": 3.8137, + "step": 4622 + }, + { + "epoch": 0.2433157894736842, + "grad_norm": 0.58984375, + "learning_rate": 0.00016815204678362575, + "loss": 3.949, + "step": 4623 + }, + { + "epoch": 0.24336842105263157, + "grad_norm": 0.578125, + "learning_rate": 0.000168140350877193, + "loss": 4.1102, + "step": 4624 + }, + { + "epoch": 0.24342105263157895, + "grad_norm": 0.49609375, + "learning_rate": 0.00016812865497076025, + "loss": 3.9013, + "step": 4625 + }, + { + "epoch": 0.24347368421052631, + "grad_norm": 0.515625, + "learning_rate": 0.0001681169590643275, + "loss": 3.9776, + "step": 4626 + }, + { + "epoch": 0.24352631578947367, + "grad_norm": 0.51953125, + "learning_rate": 0.00016810526315789474, + "loss": 4.0371, + "step": 4627 + }, + { + "epoch": 0.24357894736842106, + "grad_norm": 0.51171875, + "learning_rate": 0.000168093567251462, + "loss": 3.6772, + "step": 4628 + }, + { + "epoch": 0.24363157894736842, + "grad_norm": 0.546875, + "learning_rate": 0.00016808187134502924, + "loss": 4.0889, + "step": 4629 + }, + { + "epoch": 0.24368421052631578, + "grad_norm": 0.64453125, + "learning_rate": 0.0001680701754385965, + "loss": 3.9287, + "step": 4630 + }, + { + "epoch": 0.24373684210526317, + "grad_norm": 0.55078125, + "learning_rate": 0.00016805847953216377, + "loss": 3.7393, + "step": 4631 + }, + { + "epoch": 0.24378947368421053, + "grad_norm": 0.54296875, + "learning_rate": 0.00016804678362573102, + "loss": 3.822, + "step": 4632 + }, + { + "epoch": 0.2438421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.00016803508771929827, + "loss": 3.8706, + "step": 4633 + }, + { + "epoch": 0.24389473684210528, + "grad_norm": 0.5546875, + "learning_rate": 0.0001680233918128655, + "loss": 3.5631, + "step": 4634 + }, + { + "epoch": 0.24394736842105263, + "grad_norm": 0.78125, + "learning_rate": 0.00016801169590643274, + "loss": 3.7958, + "step": 4635 + }, + { + "epoch": 0.244, + "grad_norm": 0.515625, + "learning_rate": 0.000168, + "loss": 3.7305, + "step": 4636 + }, + { + "epoch": 0.24405263157894738, + "grad_norm": 0.5234375, + "learning_rate": 0.00016798830409356724, + "loss": 3.7733, + "step": 4637 + }, + { + "epoch": 0.24410526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.00016797660818713452, + "loss": 3.6288, + "step": 4638 + }, + { + "epoch": 0.2441578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00016796491228070177, + "loss": 3.8313, + "step": 4639 + }, + { + "epoch": 0.24421052631578946, + "grad_norm": 0.5234375, + "learning_rate": 0.00016795321637426902, + "loss": 3.9454, + "step": 4640 + }, + { + "epoch": 0.24426315789473685, + "grad_norm": 0.59375, + "learning_rate": 0.00016794152046783627, + "loss": 3.4679, + "step": 4641 + }, + { + "epoch": 0.2443157894736842, + "grad_norm": 0.578125, + "learning_rate": 0.00016792982456140352, + "loss": 3.6598, + "step": 4642 + }, + { + "epoch": 0.24436842105263157, + "grad_norm": 0.5546875, + "learning_rate": 0.00016791812865497076, + "loss": 4.17, + "step": 4643 + }, + { + "epoch": 0.24442105263157896, + "grad_norm": 0.515625, + "learning_rate": 0.00016790643274853801, + "loss": 3.7994, + "step": 4644 + }, + { + "epoch": 0.24447368421052632, + "grad_norm": 0.484375, + "learning_rate": 0.00016789473684210526, + "loss": 3.4418, + "step": 4645 + }, + { + "epoch": 0.24452631578947367, + "grad_norm": 0.52734375, + "learning_rate": 0.00016788304093567254, + "loss": 3.7702, + "step": 4646 + }, + { + "epoch": 0.24457894736842106, + "grad_norm": 0.482421875, + "learning_rate": 0.0001678713450292398, + "loss": 4.0058, + "step": 4647 + }, + { + "epoch": 0.24463157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00016785964912280704, + "loss": 4.1144, + "step": 4648 + }, + { + "epoch": 0.24468421052631578, + "grad_norm": 0.55078125, + "learning_rate": 0.0001678479532163743, + "loss": 3.8724, + "step": 4649 + }, + { + "epoch": 0.24473684210526317, + "grad_norm": 0.5390625, + "learning_rate": 0.0001678362573099415, + "loss": 3.7661, + "step": 4650 + }, + { + "epoch": 0.24478947368421053, + "grad_norm": 0.51953125, + "learning_rate": 0.00016782456140350876, + "loss": 3.7611, + "step": 4651 + }, + { + "epoch": 0.2448421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.000167812865497076, + "loss": 3.8236, + "step": 4652 + }, + { + "epoch": 0.24489473684210528, + "grad_norm": 0.5546875, + "learning_rate": 0.0001678011695906433, + "loss": 4.0579, + "step": 4653 + }, + { + "epoch": 0.24494736842105264, + "grad_norm": 0.51953125, + "learning_rate": 0.00016778947368421054, + "loss": 4.1403, + "step": 4654 + }, + { + "epoch": 0.245, + "grad_norm": 0.5390625, + "learning_rate": 0.0001677777777777778, + "loss": 3.8475, + "step": 4655 + }, + { + "epoch": 0.24505263157894736, + "grad_norm": 0.53515625, + "learning_rate": 0.00016776608187134504, + "loss": 3.0549, + "step": 4656 + }, + { + "epoch": 0.24510526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00016775438596491229, + "loss": 3.9042, + "step": 4657 + }, + { + "epoch": 0.2451578947368421, + "grad_norm": 0.5625, + "learning_rate": 0.00016774269005847954, + "loss": 4.1428, + "step": 4658 + }, + { + "epoch": 0.24521052631578946, + "grad_norm": 0.57421875, + "learning_rate": 0.00016773099415204679, + "loss": 4.1325, + "step": 4659 + }, + { + "epoch": 0.24526315789473685, + "grad_norm": 0.5546875, + "learning_rate": 0.00016771929824561406, + "loss": 4.3371, + "step": 4660 + }, + { + "epoch": 0.2453157894736842, + "grad_norm": 0.58984375, + "learning_rate": 0.0001677076023391813, + "loss": 3.9929, + "step": 4661 + }, + { + "epoch": 0.24536842105263157, + "grad_norm": 0.62890625, + "learning_rate": 0.00016769590643274856, + "loss": 3.9466, + "step": 4662 + }, + { + "epoch": 0.24542105263157896, + "grad_norm": 0.6015625, + "learning_rate": 0.0001676842105263158, + "loss": 4.4568, + "step": 4663 + }, + { + "epoch": 0.24547368421052632, + "grad_norm": 0.6171875, + "learning_rate": 0.00016767251461988306, + "loss": 4.0588, + "step": 4664 + }, + { + "epoch": 0.24552631578947368, + "grad_norm": 0.5546875, + "learning_rate": 0.00016766081871345028, + "loss": 3.9472, + "step": 4665 + }, + { + "epoch": 0.24557894736842106, + "grad_norm": 0.55859375, + "learning_rate": 0.00016764912280701753, + "loss": 3.6611, + "step": 4666 + }, + { + "epoch": 0.24563157894736842, + "grad_norm": 0.478515625, + "learning_rate": 0.00016763742690058478, + "loss": 4.2011, + "step": 4667 + }, + { + "epoch": 0.24568421052631578, + "grad_norm": 0.53515625, + "learning_rate": 0.00016762573099415206, + "loss": 4.3074, + "step": 4668 + }, + { + "epoch": 0.24573684210526317, + "grad_norm": 0.6171875, + "learning_rate": 0.0001676140350877193, + "loss": 3.9038, + "step": 4669 + }, + { + "epoch": 0.24578947368421053, + "grad_norm": 0.51171875, + "learning_rate": 0.00016760233918128656, + "loss": 3.7808, + "step": 4670 + }, + { + "epoch": 0.2458421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.0001675906432748538, + "loss": 4.032, + "step": 4671 + }, + { + "epoch": 0.24589473684210525, + "grad_norm": 0.5390625, + "learning_rate": 0.00016757894736842106, + "loss": 3.937, + "step": 4672 + }, + { + "epoch": 0.24594736842105264, + "grad_norm": 0.64453125, + "learning_rate": 0.0001675672514619883, + "loss": 3.806, + "step": 4673 + }, + { + "epoch": 0.246, + "grad_norm": 0.5, + "learning_rate": 0.00016755555555555556, + "loss": 3.7853, + "step": 4674 + }, + { + "epoch": 0.24605263157894736, + "grad_norm": 0.5234375, + "learning_rate": 0.00016754385964912283, + "loss": 3.9525, + "step": 4675 + }, + { + "epoch": 0.24610526315789474, + "grad_norm": 0.57421875, + "learning_rate": 0.00016753216374269008, + "loss": 3.6008, + "step": 4676 + }, + { + "epoch": 0.2461578947368421, + "grad_norm": 0.60546875, + "learning_rate": 0.00016752046783625733, + "loss": 3.8718, + "step": 4677 + }, + { + "epoch": 0.24621052631578946, + "grad_norm": 0.546875, + "learning_rate": 0.00016750877192982458, + "loss": 3.7461, + "step": 4678 + }, + { + "epoch": 0.24626315789473685, + "grad_norm": 0.51953125, + "learning_rate": 0.00016749707602339183, + "loss": 3.9121, + "step": 4679 + }, + { + "epoch": 0.2463157894736842, + "grad_norm": 0.51953125, + "learning_rate": 0.00016748538011695905, + "loss": 4.2086, + "step": 4680 + }, + { + "epoch": 0.24636842105263157, + "grad_norm": 0.55859375, + "learning_rate": 0.0001674736842105263, + "loss": 3.8234, + "step": 4681 + }, + { + "epoch": 0.24642105263157896, + "grad_norm": 0.57421875, + "learning_rate": 0.00016746198830409355, + "loss": 3.8673, + "step": 4682 + }, + { + "epoch": 0.24647368421052632, + "grad_norm": 0.5078125, + "learning_rate": 0.00016745029239766083, + "loss": 3.6596, + "step": 4683 + }, + { + "epoch": 0.24652631578947368, + "grad_norm": 0.7734375, + "learning_rate": 0.00016743859649122808, + "loss": 4.2893, + "step": 4684 + }, + { + "epoch": 0.24657894736842106, + "grad_norm": 0.53515625, + "learning_rate": 0.00016742690058479533, + "loss": 3.9676, + "step": 4685 + }, + { + "epoch": 0.24663157894736842, + "grad_norm": 0.5234375, + "learning_rate": 0.00016741520467836258, + "loss": 3.7965, + "step": 4686 + }, + { + "epoch": 0.24668421052631578, + "grad_norm": 0.55859375, + "learning_rate": 0.00016740350877192983, + "loss": 4.1874, + "step": 4687 + }, + { + "epoch": 0.24673684210526317, + "grad_norm": 0.5546875, + "learning_rate": 0.00016739181286549708, + "loss": 3.192, + "step": 4688 + }, + { + "epoch": 0.24678947368421053, + "grad_norm": 0.486328125, + "learning_rate": 0.00016738011695906433, + "loss": 4.2697, + "step": 4689 + }, + { + "epoch": 0.2468421052631579, + "grad_norm": 0.55078125, + "learning_rate": 0.0001673684210526316, + "loss": 3.9707, + "step": 4690 + }, + { + "epoch": 0.24689473684210525, + "grad_norm": 0.54296875, + "learning_rate": 0.00016735672514619885, + "loss": 3.62, + "step": 4691 + }, + { + "epoch": 0.24694736842105264, + "grad_norm": 0.546875, + "learning_rate": 0.0001673450292397661, + "loss": 4.1079, + "step": 4692 + }, + { + "epoch": 0.247, + "grad_norm": 0.5546875, + "learning_rate": 0.00016733333333333335, + "loss": 4.2704, + "step": 4693 + }, + { + "epoch": 0.24705263157894736, + "grad_norm": 0.60546875, + "learning_rate": 0.0001673216374269006, + "loss": 3.9641, + "step": 4694 + }, + { + "epoch": 0.24710526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00016730994152046785, + "loss": 3.9226, + "step": 4695 + }, + { + "epoch": 0.2471578947368421, + "grad_norm": 0.546875, + "learning_rate": 0.00016729824561403507, + "loss": 3.9657, + "step": 4696 + }, + { + "epoch": 0.24721052631578946, + "grad_norm": 0.5703125, + "learning_rate": 0.00016728654970760235, + "loss": 3.7202, + "step": 4697 + }, + { + "epoch": 0.24726315789473685, + "grad_norm": 0.50390625, + "learning_rate": 0.0001672748538011696, + "loss": 3.4889, + "step": 4698 + }, + { + "epoch": 0.2473157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00016726315789473685, + "loss": 4.045, + "step": 4699 + }, + { + "epoch": 0.24736842105263157, + "grad_norm": 0.55078125, + "learning_rate": 0.0001672514619883041, + "loss": 3.8789, + "step": 4700 + }, + { + "epoch": 0.24742105263157896, + "grad_norm": 0.5234375, + "learning_rate": 0.00016723976608187135, + "loss": 3.8822, + "step": 4701 + }, + { + "epoch": 0.24747368421052632, + "grad_norm": 0.59375, + "learning_rate": 0.0001672280701754386, + "loss": 4.0445, + "step": 4702 + }, + { + "epoch": 0.24752631578947368, + "grad_norm": 0.64453125, + "learning_rate": 0.00016721637426900585, + "loss": 3.9253, + "step": 4703 + }, + { + "epoch": 0.24757894736842107, + "grad_norm": 0.57421875, + "learning_rate": 0.0001672046783625731, + "loss": 3.994, + "step": 4704 + }, + { + "epoch": 0.24763157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.00016719298245614038, + "loss": 4.1755, + "step": 4705 + }, + { + "epoch": 0.24768421052631578, + "grad_norm": 0.55859375, + "learning_rate": 0.00016718128654970762, + "loss": 3.7658, + "step": 4706 + }, + { + "epoch": 0.24773684210526314, + "grad_norm": 0.486328125, + "learning_rate": 0.00016716959064327487, + "loss": 4.0553, + "step": 4707 + }, + { + "epoch": 0.24778947368421053, + "grad_norm": 0.5390625, + "learning_rate": 0.00016715789473684212, + "loss": 3.5219, + "step": 4708 + }, + { + "epoch": 0.2478421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00016714619883040937, + "loss": 4.1021, + "step": 4709 + }, + { + "epoch": 0.24789473684210525, + "grad_norm": 0.58203125, + "learning_rate": 0.00016713450292397662, + "loss": 3.9276, + "step": 4710 + }, + { + "epoch": 0.24794736842105264, + "grad_norm": 0.5234375, + "learning_rate": 0.00016712280701754385, + "loss": 4.0506, + "step": 4711 + }, + { + "epoch": 0.248, + "grad_norm": 0.55078125, + "learning_rate": 0.00016711111111111112, + "loss": 3.9642, + "step": 4712 + }, + { + "epoch": 0.24805263157894736, + "grad_norm": 0.54296875, + "learning_rate": 0.00016709941520467837, + "loss": 3.9903, + "step": 4713 + }, + { + "epoch": 0.24810526315789475, + "grad_norm": 0.58984375, + "learning_rate": 0.00016708771929824562, + "loss": 3.7894, + "step": 4714 + }, + { + "epoch": 0.2481578947368421, + "grad_norm": 0.65625, + "learning_rate": 0.00016707602339181287, + "loss": 4.0807, + "step": 4715 + }, + { + "epoch": 0.24821052631578946, + "grad_norm": 0.515625, + "learning_rate": 0.00016706432748538012, + "loss": 3.7604, + "step": 4716 + }, + { + "epoch": 0.24826315789473685, + "grad_norm": 0.69140625, + "learning_rate": 0.00016705263157894737, + "loss": 3.7254, + "step": 4717 + }, + { + "epoch": 0.2483157894736842, + "grad_norm": 1.5546875, + "learning_rate": 0.00016704093567251462, + "loss": 3.7534, + "step": 4718 + }, + { + "epoch": 0.24836842105263157, + "grad_norm": 0.55859375, + "learning_rate": 0.00016702923976608187, + "loss": 4.0715, + "step": 4719 + }, + { + "epoch": 0.24842105263157896, + "grad_norm": 0.4765625, + "learning_rate": 0.00016701754385964915, + "loss": 3.8176, + "step": 4720 + }, + { + "epoch": 0.24847368421052632, + "grad_norm": 0.5625, + "learning_rate": 0.0001670058479532164, + "loss": 3.528, + "step": 4721 + }, + { + "epoch": 0.24852631578947368, + "grad_norm": 0.82421875, + "learning_rate": 0.00016699415204678365, + "loss": 3.9372, + "step": 4722 + }, + { + "epoch": 0.24857894736842107, + "grad_norm": 0.6171875, + "learning_rate": 0.0001669824561403509, + "loss": 3.8872, + "step": 4723 + }, + { + "epoch": 0.24863157894736843, + "grad_norm": 0.49609375, + "learning_rate": 0.00016697076023391814, + "loss": 3.9955, + "step": 4724 + }, + { + "epoch": 0.24868421052631579, + "grad_norm": 0.64453125, + "learning_rate": 0.0001669590643274854, + "loss": 3.5839, + "step": 4725 + }, + { + "epoch": 0.24873684210526315, + "grad_norm": 0.5390625, + "learning_rate": 0.00016694736842105262, + "loss": 3.7465, + "step": 4726 + }, + { + "epoch": 0.24878947368421053, + "grad_norm": 0.45703125, + "learning_rate": 0.0001669356725146199, + "loss": 3.8426, + "step": 4727 + }, + { + "epoch": 0.2488421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00016692397660818714, + "loss": 3.5585, + "step": 4728 + }, + { + "epoch": 0.24889473684210525, + "grad_norm": 0.515625, + "learning_rate": 0.0001669122807017544, + "loss": 4.0648, + "step": 4729 + }, + { + "epoch": 0.24894736842105264, + "grad_norm": 0.578125, + "learning_rate": 0.00016690058479532164, + "loss": 4.151, + "step": 4730 + }, + { + "epoch": 0.249, + "grad_norm": 0.55859375, + "learning_rate": 0.0001668888888888889, + "loss": 3.6079, + "step": 4731 + }, + { + "epoch": 0.24905263157894736, + "grad_norm": 0.58984375, + "learning_rate": 0.00016687719298245614, + "loss": 3.6374, + "step": 4732 + }, + { + "epoch": 0.24910526315789475, + "grad_norm": 0.5859375, + "learning_rate": 0.0001668654970760234, + "loss": 4.107, + "step": 4733 + }, + { + "epoch": 0.2491578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00016685380116959067, + "loss": 4.0046, + "step": 4734 + }, + { + "epoch": 0.24921052631578947, + "grad_norm": 0.56640625, + "learning_rate": 0.00016684210526315792, + "loss": 4.0446, + "step": 4735 + }, + { + "epoch": 0.24926315789473685, + "grad_norm": 0.51953125, + "learning_rate": 0.00016683040935672517, + "loss": 4.3332, + "step": 4736 + }, + { + "epoch": 0.2493157894736842, + "grad_norm": 0.5625, + "learning_rate": 0.00016681871345029242, + "loss": 4.252, + "step": 4737 + }, + { + "epoch": 0.24936842105263157, + "grad_norm": 0.609375, + "learning_rate": 0.00016680701754385967, + "loss": 4.0382, + "step": 4738 + }, + { + "epoch": 0.24942105263157896, + "grad_norm": 0.51953125, + "learning_rate": 0.00016679532163742692, + "loss": 4.1783, + "step": 4739 + }, + { + "epoch": 0.24947368421052632, + "grad_norm": 0.51171875, + "learning_rate": 0.00016678362573099417, + "loss": 3.9889, + "step": 4740 + }, + { + "epoch": 0.24952631578947368, + "grad_norm": 0.59765625, + "learning_rate": 0.00016677192982456141, + "loss": 3.5195, + "step": 4741 + }, + { + "epoch": 0.24957894736842104, + "grad_norm": 0.57421875, + "learning_rate": 0.00016676023391812866, + "loss": 3.5925, + "step": 4742 + }, + { + "epoch": 0.24963157894736843, + "grad_norm": 0.5859375, + "learning_rate": 0.00016674853801169591, + "loss": 3.6783, + "step": 4743 + }, + { + "epoch": 0.2496842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.00016673684210526316, + "loss": 3.8089, + "step": 4744 + }, + { + "epoch": 0.24973684210526315, + "grad_norm": 0.5234375, + "learning_rate": 0.0001667251461988304, + "loss": 3.8955, + "step": 4745 + }, + { + "epoch": 0.24978947368421053, + "grad_norm": 0.5234375, + "learning_rate": 0.00016671345029239766, + "loss": 4.186, + "step": 4746 + }, + { + "epoch": 0.2498421052631579, + "grad_norm": 0.58984375, + "learning_rate": 0.0001667017543859649, + "loss": 3.9829, + "step": 4747 + }, + { + "epoch": 0.24989473684210525, + "grad_norm": 0.58984375, + "learning_rate": 0.00016669005847953216, + "loss": 3.7234, + "step": 4748 + }, + { + "epoch": 0.24994736842105264, + "grad_norm": 0.5703125, + "learning_rate": 0.00016667836257309944, + "loss": 4.2956, + "step": 4749 + }, + { + "epoch": 0.25, + "grad_norm": 0.6171875, + "learning_rate": 0.0001666666666666667, + "loss": 3.4696, + "step": 4750 + }, + { + "epoch": 0.25005263157894736, + "grad_norm": 0.5703125, + "learning_rate": 0.00016665497076023394, + "loss": 3.669, + "step": 4751 + }, + { + "epoch": 0.2501052631578947, + "grad_norm": 0.5390625, + "learning_rate": 0.0001666432748538012, + "loss": 3.9138, + "step": 4752 + }, + { + "epoch": 0.2501578947368421, + "grad_norm": 0.5625, + "learning_rate": 0.00016663157894736844, + "loss": 3.4312, + "step": 4753 + }, + { + "epoch": 0.2502105263157895, + "grad_norm": 0.6875, + "learning_rate": 0.0001666198830409357, + "loss": 3.4885, + "step": 4754 + }, + { + "epoch": 0.25026315789473685, + "grad_norm": 0.53125, + "learning_rate": 0.00016660818713450294, + "loss": 3.9812, + "step": 4755 + }, + { + "epoch": 0.2503157894736842, + "grad_norm": 0.6328125, + "learning_rate": 0.00016659649122807019, + "loss": 3.6462, + "step": 4756 + }, + { + "epoch": 0.2503684210526316, + "grad_norm": 0.55859375, + "learning_rate": 0.00016658479532163744, + "loss": 3.4133, + "step": 4757 + }, + { + "epoch": 0.25042105263157893, + "grad_norm": 0.53125, + "learning_rate": 0.00016657309941520468, + "loss": 3.7367, + "step": 4758 + }, + { + "epoch": 0.2504736842105263, + "grad_norm": 0.4921875, + "learning_rate": 0.00016656140350877193, + "loss": 3.7763, + "step": 4759 + }, + { + "epoch": 0.2505263157894737, + "grad_norm": 0.52734375, + "learning_rate": 0.00016654970760233918, + "loss": 3.6784, + "step": 4760 + }, + { + "epoch": 0.25057894736842107, + "grad_norm": 0.47265625, + "learning_rate": 0.00016653801169590643, + "loss": 3.6685, + "step": 4761 + }, + { + "epoch": 0.2506315789473684, + "grad_norm": 0.490234375, + "learning_rate": 0.00016652631578947368, + "loss": 3.8747, + "step": 4762 + }, + { + "epoch": 0.2506842105263158, + "grad_norm": 0.484375, + "learning_rate": 0.00016651461988304093, + "loss": 4.1149, + "step": 4763 + }, + { + "epoch": 0.25073684210526315, + "grad_norm": 0.5625, + "learning_rate": 0.0001665029239766082, + "loss": 3.7786, + "step": 4764 + }, + { + "epoch": 0.2507894736842105, + "grad_norm": 0.57421875, + "learning_rate": 0.00016649122807017546, + "loss": 3.3236, + "step": 4765 + }, + { + "epoch": 0.2508421052631579, + "grad_norm": 0.5625, + "learning_rate": 0.0001664795321637427, + "loss": 4.2162, + "step": 4766 + }, + { + "epoch": 0.2508947368421053, + "grad_norm": 0.625, + "learning_rate": 0.00016646783625730996, + "loss": 3.9339, + "step": 4767 + }, + { + "epoch": 0.25094736842105264, + "grad_norm": 0.57421875, + "learning_rate": 0.0001664561403508772, + "loss": 3.8986, + "step": 4768 + }, + { + "epoch": 0.251, + "grad_norm": 0.59375, + "learning_rate": 0.00016644444444444446, + "loss": 3.725, + "step": 4769 + }, + { + "epoch": 0.25105263157894736, + "grad_norm": 0.474609375, + "learning_rate": 0.0001664327485380117, + "loss": 4.0663, + "step": 4770 + }, + { + "epoch": 0.2511052631578947, + "grad_norm": 0.482421875, + "learning_rate": 0.00016642105263157896, + "loss": 3.843, + "step": 4771 + }, + { + "epoch": 0.2511578947368421, + "grad_norm": 0.5859375, + "learning_rate": 0.0001664093567251462, + "loss": 3.7149, + "step": 4772 + }, + { + "epoch": 0.2512105263157895, + "grad_norm": 0.52734375, + "learning_rate": 0.00016639766081871346, + "loss": 3.9111, + "step": 4773 + }, + { + "epoch": 0.25126315789473685, + "grad_norm": 0.546875, + "learning_rate": 0.0001663859649122807, + "loss": 3.7978, + "step": 4774 + }, + { + "epoch": 0.2513157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00016637426900584796, + "loss": 3.8585, + "step": 4775 + }, + { + "epoch": 0.2513684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.0001663625730994152, + "loss": 3.6328, + "step": 4776 + }, + { + "epoch": 0.25142105263157893, + "grad_norm": 0.50390625, + "learning_rate": 0.00016635087719298245, + "loss": 4.0745, + "step": 4777 + }, + { + "epoch": 0.2514736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.0001663391812865497, + "loss": 4.1305, + "step": 4778 + }, + { + "epoch": 0.2515263157894737, + "grad_norm": 0.5390625, + "learning_rate": 0.00016632748538011698, + "loss": 4.1511, + "step": 4779 + }, + { + "epoch": 0.25157894736842107, + "grad_norm": 0.53515625, + "learning_rate": 0.00016631578947368423, + "loss": 4.2411, + "step": 4780 + }, + { + "epoch": 0.25163157894736843, + "grad_norm": 0.57421875, + "learning_rate": 0.00016630409356725148, + "loss": 3.7991, + "step": 4781 + }, + { + "epoch": 0.2516842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00016629239766081873, + "loss": 3.7718, + "step": 4782 + }, + { + "epoch": 0.25173684210526315, + "grad_norm": 0.6484375, + "learning_rate": 0.00016628070175438598, + "loss": 3.6707, + "step": 4783 + }, + { + "epoch": 0.2517894736842105, + "grad_norm": 0.498046875, + "learning_rate": 0.00016626900584795323, + "loss": 4.0117, + "step": 4784 + }, + { + "epoch": 0.25184210526315787, + "grad_norm": 0.60546875, + "learning_rate": 0.00016625730994152048, + "loss": 4.0529, + "step": 4785 + }, + { + "epoch": 0.2518947368421053, + "grad_norm": 0.59765625, + "learning_rate": 0.00016624561403508773, + "loss": 4.0168, + "step": 4786 + }, + { + "epoch": 0.25194736842105264, + "grad_norm": 0.53125, + "learning_rate": 0.00016623391812865498, + "loss": 3.6229, + "step": 4787 + }, + { + "epoch": 0.252, + "grad_norm": 0.546875, + "learning_rate": 0.00016622222222222223, + "loss": 3.5809, + "step": 4788 + }, + { + "epoch": 0.25205263157894736, + "grad_norm": 0.578125, + "learning_rate": 0.00016621052631578948, + "loss": 3.6618, + "step": 4789 + }, + { + "epoch": 0.2521052631578947, + "grad_norm": 0.55078125, + "learning_rate": 0.00016619883040935673, + "loss": 4.0257, + "step": 4790 + }, + { + "epoch": 0.2521578947368421, + "grad_norm": 0.58203125, + "learning_rate": 0.00016618713450292398, + "loss": 4.2264, + "step": 4791 + }, + { + "epoch": 0.2522105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.00016617543859649123, + "loss": 4.2416, + "step": 4792 + }, + { + "epoch": 0.25226315789473686, + "grad_norm": 0.546875, + "learning_rate": 0.00016616374269005847, + "loss": 3.9831, + "step": 4793 + }, + { + "epoch": 0.2523157894736842, + "grad_norm": 1.171875, + "learning_rate": 0.00016615204678362575, + "loss": 3.7369, + "step": 4794 + }, + { + "epoch": 0.2523684210526316, + "grad_norm": 0.62109375, + "learning_rate": 0.000166140350877193, + "loss": 3.3694, + "step": 4795 + }, + { + "epoch": 0.25242105263157893, + "grad_norm": 0.48828125, + "learning_rate": 0.00016612865497076025, + "loss": 3.7973, + "step": 4796 + }, + { + "epoch": 0.2524736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.0001661169590643275, + "loss": 3.7952, + "step": 4797 + }, + { + "epoch": 0.2525263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.00016610526315789475, + "loss": 4.182, + "step": 4798 + }, + { + "epoch": 0.25257894736842107, + "grad_norm": 0.50390625, + "learning_rate": 0.000166093567251462, + "loss": 3.9239, + "step": 4799 + }, + { + "epoch": 0.25263157894736843, + "grad_norm": 0.6171875, + "learning_rate": 0.00016608187134502925, + "loss": 3.5118, + "step": 4800 + }, + { + "epoch": 0.2526842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.0001660701754385965, + "loss": 4.2198, + "step": 4801 + }, + { + "epoch": 0.25273684210526315, + "grad_norm": 0.515625, + "learning_rate": 0.00016605847953216375, + "loss": 3.7362, + "step": 4802 + }, + { + "epoch": 0.2527894736842105, + "grad_norm": 0.53515625, + "learning_rate": 0.000166046783625731, + "loss": 3.6997, + "step": 4803 + }, + { + "epoch": 0.25284210526315787, + "grad_norm": 0.51171875, + "learning_rate": 0.00016603508771929825, + "loss": 4.2338, + "step": 4804 + }, + { + "epoch": 0.2528947368421053, + "grad_norm": 0.51953125, + "learning_rate": 0.0001660233918128655, + "loss": 3.3568, + "step": 4805 + }, + { + "epoch": 0.25294736842105264, + "grad_norm": 0.578125, + "learning_rate": 0.00016601169590643275, + "loss": 3.3725, + "step": 4806 + }, + { + "epoch": 0.253, + "grad_norm": 0.47265625, + "learning_rate": 0.000166, + "loss": 3.5635, + "step": 4807 + }, + { + "epoch": 0.25305263157894736, + "grad_norm": 0.50390625, + "learning_rate": 0.00016598830409356725, + "loss": 3.4314, + "step": 4808 + }, + { + "epoch": 0.2531052631578947, + "grad_norm": 0.6484375, + "learning_rate": 0.00016597660818713452, + "loss": 3.4303, + "step": 4809 + }, + { + "epoch": 0.2531578947368421, + "grad_norm": 0.58203125, + "learning_rate": 0.00016596491228070177, + "loss": 3.9195, + "step": 4810 + }, + { + "epoch": 0.2532105263157895, + "grad_norm": 0.6015625, + "learning_rate": 0.00016595321637426902, + "loss": 3.8764, + "step": 4811 + }, + { + "epoch": 0.25326315789473686, + "grad_norm": 0.640625, + "learning_rate": 0.00016594152046783627, + "loss": 4.3239, + "step": 4812 + }, + { + "epoch": 0.2533157894736842, + "grad_norm": 0.6171875, + "learning_rate": 0.00016592982456140352, + "loss": 3.7631, + "step": 4813 + }, + { + "epoch": 0.2533684210526316, + "grad_norm": 0.58203125, + "learning_rate": 0.00016591812865497077, + "loss": 3.698, + "step": 4814 + }, + { + "epoch": 0.25342105263157894, + "grad_norm": 0.515625, + "learning_rate": 0.00016590643274853802, + "loss": 4.1612, + "step": 4815 + }, + { + "epoch": 0.2534736842105263, + "grad_norm": 0.486328125, + "learning_rate": 0.00016589473684210527, + "loss": 3.9182, + "step": 4816 + }, + { + "epoch": 0.2535263157894737, + "grad_norm": 0.55078125, + "learning_rate": 0.00016588304093567252, + "loss": 3.4855, + "step": 4817 + }, + { + "epoch": 0.25357894736842107, + "grad_norm": 0.55859375, + "learning_rate": 0.00016587134502923977, + "loss": 3.9814, + "step": 4818 + }, + { + "epoch": 0.25363157894736843, + "grad_norm": 0.5625, + "learning_rate": 0.00016585964912280702, + "loss": 3.6939, + "step": 4819 + }, + { + "epoch": 0.2536842105263158, + "grad_norm": 0.48046875, + "learning_rate": 0.00016584795321637427, + "loss": 4.0634, + "step": 4820 + }, + { + "epoch": 0.25373684210526315, + "grad_norm": 0.5625, + "learning_rate": 0.00016583625730994152, + "loss": 3.6331, + "step": 4821 + }, + { + "epoch": 0.2537894736842105, + "grad_norm": 0.46875, + "learning_rate": 0.00016582456140350877, + "loss": 4.0926, + "step": 4822 + }, + { + "epoch": 0.25384210526315787, + "grad_norm": 0.53125, + "learning_rate": 0.00016581286549707604, + "loss": 3.9102, + "step": 4823 + }, + { + "epoch": 0.2538947368421053, + "grad_norm": 0.625, + "learning_rate": 0.0001658011695906433, + "loss": 4.3941, + "step": 4824 + }, + { + "epoch": 0.25394736842105264, + "grad_norm": 0.56640625, + "learning_rate": 0.00016578947368421054, + "loss": 4.0285, + "step": 4825 + }, + { + "epoch": 0.254, + "grad_norm": 0.5859375, + "learning_rate": 0.0001657777777777778, + "loss": 3.8158, + "step": 4826 + }, + { + "epoch": 0.25405263157894736, + "grad_norm": 0.6015625, + "learning_rate": 0.00016576608187134504, + "loss": 3.9046, + "step": 4827 + }, + { + "epoch": 0.2541052631578947, + "grad_norm": 0.5078125, + "learning_rate": 0.0001657543859649123, + "loss": 3.549, + "step": 4828 + }, + { + "epoch": 0.2541578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.00016574269005847954, + "loss": 3.6049, + "step": 4829 + }, + { + "epoch": 0.2542105263157895, + "grad_norm": 0.609375, + "learning_rate": 0.0001657309941520468, + "loss": 3.0116, + "step": 4830 + }, + { + "epoch": 0.25426315789473686, + "grad_norm": 0.5234375, + "learning_rate": 0.00016571929824561404, + "loss": 3.818, + "step": 4831 + }, + { + "epoch": 0.2543157894736842, + "grad_norm": 0.4921875, + "learning_rate": 0.0001657076023391813, + "loss": 3.7565, + "step": 4832 + }, + { + "epoch": 0.2543684210526316, + "grad_norm": 0.56640625, + "learning_rate": 0.00016569590643274854, + "loss": 3.8232, + "step": 4833 + }, + { + "epoch": 0.25442105263157894, + "grad_norm": 0.76953125, + "learning_rate": 0.0001656842105263158, + "loss": 4.0631, + "step": 4834 + }, + { + "epoch": 0.2544736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.00016567251461988304, + "loss": 3.4366, + "step": 4835 + }, + { + "epoch": 0.2545263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.0001656608187134503, + "loss": 3.9549, + "step": 4836 + }, + { + "epoch": 0.25457894736842107, + "grad_norm": 0.52734375, + "learning_rate": 0.00016564912280701754, + "loss": 3.8097, + "step": 4837 + }, + { + "epoch": 0.25463157894736843, + "grad_norm": 0.57421875, + "learning_rate": 0.00016563742690058481, + "loss": 3.596, + "step": 4838 + }, + { + "epoch": 0.2546842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00016562573099415206, + "loss": 3.9463, + "step": 4839 + }, + { + "epoch": 0.25473684210526315, + "grad_norm": 0.52734375, + "learning_rate": 0.00016561403508771931, + "loss": 3.8632, + "step": 4840 + }, + { + "epoch": 0.2547894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.00016560233918128656, + "loss": 4.1149, + "step": 4841 + }, + { + "epoch": 0.25484210526315787, + "grad_norm": 0.54296875, + "learning_rate": 0.0001655906432748538, + "loss": 3.9262, + "step": 4842 + }, + { + "epoch": 0.2548947368421053, + "grad_norm": 0.53125, + "learning_rate": 0.00016557894736842106, + "loss": 4.0598, + "step": 4843 + }, + { + "epoch": 0.25494736842105264, + "grad_norm": 0.515625, + "learning_rate": 0.0001655672514619883, + "loss": 3.928, + "step": 4844 + }, + { + "epoch": 0.255, + "grad_norm": 0.51171875, + "learning_rate": 0.00016555555555555556, + "loss": 3.8937, + "step": 4845 + }, + { + "epoch": 0.25505263157894736, + "grad_norm": 0.51953125, + "learning_rate": 0.0001655438596491228, + "loss": 3.8674, + "step": 4846 + }, + { + "epoch": 0.2551052631578947, + "grad_norm": 0.6328125, + "learning_rate": 0.00016553216374269006, + "loss": 4.0406, + "step": 4847 + }, + { + "epoch": 0.2551578947368421, + "grad_norm": 0.494140625, + "learning_rate": 0.0001655204678362573, + "loss": 3.8072, + "step": 4848 + }, + { + "epoch": 0.2552105263157895, + "grad_norm": 0.53515625, + "learning_rate": 0.00016550877192982456, + "loss": 3.4582, + "step": 4849 + }, + { + "epoch": 0.25526315789473686, + "grad_norm": 0.50390625, + "learning_rate": 0.0001654970760233918, + "loss": 3.7213, + "step": 4850 + }, + { + "epoch": 0.2553157894736842, + "grad_norm": 0.578125, + "learning_rate": 0.00016548538011695906, + "loss": 3.5481, + "step": 4851 + }, + { + "epoch": 0.2553684210526316, + "grad_norm": 0.5703125, + "learning_rate": 0.0001654736842105263, + "loss": 3.6685, + "step": 4852 + }, + { + "epoch": 0.25542105263157894, + "grad_norm": 0.5859375, + "learning_rate": 0.00016546198830409359, + "loss": 3.8448, + "step": 4853 + }, + { + "epoch": 0.2554736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00016545029239766084, + "loss": 4.1289, + "step": 4854 + }, + { + "epoch": 0.25552631578947366, + "grad_norm": 0.56640625, + "learning_rate": 0.00016543859649122809, + "loss": 4.3269, + "step": 4855 + }, + { + "epoch": 0.25557894736842107, + "grad_norm": 0.609375, + "learning_rate": 0.00016542690058479533, + "loss": 4.4553, + "step": 4856 + }, + { + "epoch": 0.25563157894736843, + "grad_norm": 0.671875, + "learning_rate": 0.00016541520467836258, + "loss": 4.3993, + "step": 4857 + }, + { + "epoch": 0.2556842105263158, + "grad_norm": 0.6953125, + "learning_rate": 0.00016540350877192983, + "loss": 4.1227, + "step": 4858 + }, + { + "epoch": 0.25573684210526315, + "grad_norm": 0.546875, + "learning_rate": 0.00016539181286549708, + "loss": 4.2636, + "step": 4859 + }, + { + "epoch": 0.2557894736842105, + "grad_norm": 0.56640625, + "learning_rate": 0.00016538011695906433, + "loss": 3.9866, + "step": 4860 + }, + { + "epoch": 0.25584210526315787, + "grad_norm": 0.58984375, + "learning_rate": 0.00016536842105263158, + "loss": 3.8719, + "step": 4861 + }, + { + "epoch": 0.2558947368421053, + "grad_norm": 0.87109375, + "learning_rate": 0.00016535672514619883, + "loss": 3.6095, + "step": 4862 + }, + { + "epoch": 0.25594736842105265, + "grad_norm": 0.4921875, + "learning_rate": 0.00016534502923976608, + "loss": 3.923, + "step": 4863 + }, + { + "epoch": 0.256, + "grad_norm": 0.53515625, + "learning_rate": 0.00016533333333333333, + "loss": 3.5775, + "step": 4864 + }, + { + "epoch": 0.25605263157894737, + "grad_norm": 0.48046875, + "learning_rate": 0.00016532163742690058, + "loss": 4.0116, + "step": 4865 + }, + { + "epoch": 0.2561052631578947, + "grad_norm": 0.474609375, + "learning_rate": 0.00016530994152046783, + "loss": 3.7442, + "step": 4866 + }, + { + "epoch": 0.2561578947368421, + "grad_norm": 0.47265625, + "learning_rate": 0.00016529824561403508, + "loss": 3.8463, + "step": 4867 + }, + { + "epoch": 0.2562105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00016528654970760236, + "loss": 4.3571, + "step": 4868 + }, + { + "epoch": 0.25626315789473686, + "grad_norm": 0.7578125, + "learning_rate": 0.0001652748538011696, + "loss": 3.8496, + "step": 4869 + }, + { + "epoch": 0.2563157894736842, + "grad_norm": 0.52734375, + "learning_rate": 0.00016526315789473686, + "loss": 4.1718, + "step": 4870 + }, + { + "epoch": 0.2563684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.0001652514619883041, + "loss": 3.7269, + "step": 4871 + }, + { + "epoch": 0.25642105263157894, + "grad_norm": 0.5859375, + "learning_rate": 0.00016523976608187136, + "loss": 3.8345, + "step": 4872 + }, + { + "epoch": 0.2564736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.0001652280701754386, + "loss": 4.1496, + "step": 4873 + }, + { + "epoch": 0.25652631578947366, + "grad_norm": 0.62890625, + "learning_rate": 0.00016521637426900585, + "loss": 3.8879, + "step": 4874 + }, + { + "epoch": 0.2565789473684211, + "grad_norm": 0.5234375, + "learning_rate": 0.0001652046783625731, + "loss": 4.0205, + "step": 4875 + }, + { + "epoch": 0.25663157894736843, + "grad_norm": 0.5390625, + "learning_rate": 0.00016519298245614035, + "loss": 4.0242, + "step": 4876 + }, + { + "epoch": 0.2566842105263158, + "grad_norm": 0.5703125, + "learning_rate": 0.0001651812865497076, + "loss": 3.9597, + "step": 4877 + }, + { + "epoch": 0.25673684210526315, + "grad_norm": 0.5703125, + "learning_rate": 0.00016516959064327485, + "loss": 3.8167, + "step": 4878 + }, + { + "epoch": 0.2567894736842105, + "grad_norm": 0.71875, + "learning_rate": 0.0001651578947368421, + "loss": 3.9687, + "step": 4879 + }, + { + "epoch": 0.25684210526315787, + "grad_norm": 0.515625, + "learning_rate": 0.00016514619883040935, + "loss": 3.3968, + "step": 4880 + }, + { + "epoch": 0.2568947368421053, + "grad_norm": 0.515625, + "learning_rate": 0.0001651345029239766, + "loss": 3.654, + "step": 4881 + }, + { + "epoch": 0.25694736842105265, + "grad_norm": 0.52734375, + "learning_rate": 0.00016512280701754385, + "loss": 3.9786, + "step": 4882 + }, + { + "epoch": 0.257, + "grad_norm": 0.55859375, + "learning_rate": 0.00016511111111111113, + "loss": 3.8312, + "step": 4883 + }, + { + "epoch": 0.25705263157894737, + "grad_norm": 0.52734375, + "learning_rate": 0.00016509941520467838, + "loss": 4.0492, + "step": 4884 + }, + { + "epoch": 0.2571052631578947, + "grad_norm": 0.59765625, + "learning_rate": 0.00016508771929824563, + "loss": 3.5587, + "step": 4885 + }, + { + "epoch": 0.2571578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.00016507602339181288, + "loss": 4.0141, + "step": 4886 + }, + { + "epoch": 0.2572105263157895, + "grad_norm": 0.640625, + "learning_rate": 0.00016506432748538013, + "loss": 3.5048, + "step": 4887 + }, + { + "epoch": 0.25726315789473686, + "grad_norm": 0.498046875, + "learning_rate": 0.00016505263157894738, + "loss": 4.0001, + "step": 4888 + }, + { + "epoch": 0.2573157894736842, + "grad_norm": 0.52734375, + "learning_rate": 0.00016504093567251463, + "loss": 3.7611, + "step": 4889 + }, + { + "epoch": 0.2573684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.0001650292397660819, + "loss": 3.8688, + "step": 4890 + }, + { + "epoch": 0.25742105263157894, + "grad_norm": 0.53125, + "learning_rate": 0.00016501754385964912, + "loss": 4.0069, + "step": 4891 + }, + { + "epoch": 0.2574736842105263, + "grad_norm": 0.59375, + "learning_rate": 0.00016500584795321637, + "loss": 4.3017, + "step": 4892 + }, + { + "epoch": 0.25752631578947366, + "grad_norm": 0.68359375, + "learning_rate": 0.00016499415204678362, + "loss": 3.9214, + "step": 4893 + }, + { + "epoch": 0.2575789473684211, + "grad_norm": 0.53125, + "learning_rate": 0.00016498245614035087, + "loss": 3.8468, + "step": 4894 + }, + { + "epoch": 0.25763157894736843, + "grad_norm": 0.52734375, + "learning_rate": 0.00016497076023391812, + "loss": 3.2185, + "step": 4895 + }, + { + "epoch": 0.2576842105263158, + "grad_norm": 0.55078125, + "learning_rate": 0.00016495906432748537, + "loss": 4.1654, + "step": 4896 + }, + { + "epoch": 0.25773684210526315, + "grad_norm": 0.5234375, + "learning_rate": 0.00016494736842105265, + "loss": 4.2754, + "step": 4897 + }, + { + "epoch": 0.2577894736842105, + "grad_norm": 0.58984375, + "learning_rate": 0.0001649356725146199, + "loss": 3.5731, + "step": 4898 + }, + { + "epoch": 0.2578421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.00016492397660818715, + "loss": 4.1102, + "step": 4899 + }, + { + "epoch": 0.2578947368421053, + "grad_norm": 0.62109375, + "learning_rate": 0.0001649122807017544, + "loss": 3.5261, + "step": 4900 + }, + { + "epoch": 0.25794736842105265, + "grad_norm": 0.546875, + "learning_rate": 0.00016490058479532165, + "loss": 4.1663, + "step": 4901 + }, + { + "epoch": 0.258, + "grad_norm": 0.5859375, + "learning_rate": 0.0001648888888888889, + "loss": 4.197, + "step": 4902 + }, + { + "epoch": 0.25805263157894737, + "grad_norm": 0.5546875, + "learning_rate": 0.00016487719298245615, + "loss": 4.127, + "step": 4903 + }, + { + "epoch": 0.2581052631578947, + "grad_norm": 0.5703125, + "learning_rate": 0.0001648654970760234, + "loss": 3.5999, + "step": 4904 + }, + { + "epoch": 0.2581578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00016485380116959067, + "loss": 3.9772, + "step": 4905 + }, + { + "epoch": 0.2582105263157895, + "grad_norm": 0.55859375, + "learning_rate": 0.0001648421052631579, + "loss": 3.4565, + "step": 4906 + }, + { + "epoch": 0.25826315789473686, + "grad_norm": 0.50390625, + "learning_rate": 0.00016483040935672515, + "loss": 3.513, + "step": 4907 + }, + { + "epoch": 0.2583157894736842, + "grad_norm": 0.5, + "learning_rate": 0.0001648187134502924, + "loss": 3.5513, + "step": 4908 + }, + { + "epoch": 0.2583684210526316, + "grad_norm": 0.4921875, + "learning_rate": 0.00016480701754385964, + "loss": 3.2179, + "step": 4909 + }, + { + "epoch": 0.25842105263157894, + "grad_norm": 0.53125, + "learning_rate": 0.0001647953216374269, + "loss": 4.1579, + "step": 4910 + }, + { + "epoch": 0.2584736842105263, + "grad_norm": 0.51953125, + "learning_rate": 0.00016478362573099414, + "loss": 3.9434, + "step": 4911 + }, + { + "epoch": 0.25852631578947366, + "grad_norm": 0.59375, + "learning_rate": 0.00016477192982456142, + "loss": 3.7155, + "step": 4912 + }, + { + "epoch": 0.2585789473684211, + "grad_norm": 0.609375, + "learning_rate": 0.00016476023391812867, + "loss": 3.5448, + "step": 4913 + }, + { + "epoch": 0.25863157894736843, + "grad_norm": 0.515625, + "learning_rate": 0.00016474853801169592, + "loss": 3.4869, + "step": 4914 + }, + { + "epoch": 0.2586842105263158, + "grad_norm": 0.58203125, + "learning_rate": 0.00016473684210526317, + "loss": 3.8885, + "step": 4915 + }, + { + "epoch": 0.25873684210526315, + "grad_norm": 0.59765625, + "learning_rate": 0.00016472514619883042, + "loss": 3.8621, + "step": 4916 + }, + { + "epoch": 0.2587894736842105, + "grad_norm": 0.6640625, + "learning_rate": 0.00016471345029239767, + "loss": 3.5273, + "step": 4917 + }, + { + "epoch": 0.2588421052631579, + "grad_norm": 0.61328125, + "learning_rate": 0.00016470175438596492, + "loss": 4.1006, + "step": 4918 + }, + { + "epoch": 0.2588947368421053, + "grad_norm": 0.59375, + "learning_rate": 0.00016469005847953217, + "loss": 3.7203, + "step": 4919 + }, + { + "epoch": 0.25894736842105265, + "grad_norm": 0.64453125, + "learning_rate": 0.00016467836257309944, + "loss": 3.4035, + "step": 4920 + }, + { + "epoch": 0.259, + "grad_norm": 0.51953125, + "learning_rate": 0.00016466666666666667, + "loss": 3.4963, + "step": 4921 + }, + { + "epoch": 0.25905263157894737, + "grad_norm": 0.478515625, + "learning_rate": 0.00016465497076023392, + "loss": 3.8804, + "step": 4922 + }, + { + "epoch": 0.2591052631578947, + "grad_norm": 0.51953125, + "learning_rate": 0.00016464327485380117, + "loss": 3.7905, + "step": 4923 + }, + { + "epoch": 0.2591578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00016463157894736842, + "loss": 3.6104, + "step": 4924 + }, + { + "epoch": 0.25921052631578945, + "grad_norm": 0.5703125, + "learning_rate": 0.00016461988304093567, + "loss": 3.8292, + "step": 4925 + }, + { + "epoch": 0.25926315789473686, + "grad_norm": 0.55078125, + "learning_rate": 0.00016460818713450291, + "loss": 3.9924, + "step": 4926 + }, + { + "epoch": 0.2593157894736842, + "grad_norm": 0.68359375, + "learning_rate": 0.0001645964912280702, + "loss": 4.0948, + "step": 4927 + }, + { + "epoch": 0.2593684210526316, + "grad_norm": 0.59765625, + "learning_rate": 0.00016458479532163744, + "loss": 3.8222, + "step": 4928 + }, + { + "epoch": 0.25942105263157894, + "grad_norm": 0.578125, + "learning_rate": 0.0001645730994152047, + "loss": 3.7583, + "step": 4929 + }, + { + "epoch": 0.2594736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00016456140350877194, + "loss": 4.0363, + "step": 4930 + }, + { + "epoch": 0.25952631578947366, + "grad_norm": 0.6171875, + "learning_rate": 0.0001645497076023392, + "loss": 4.253, + "step": 4931 + }, + { + "epoch": 0.2595789473684211, + "grad_norm": 0.5546875, + "learning_rate": 0.00016453801169590644, + "loss": 3.5707, + "step": 4932 + }, + { + "epoch": 0.25963157894736844, + "grad_norm": 0.5, + "learning_rate": 0.0001645263157894737, + "loss": 4.1563, + "step": 4933 + }, + { + "epoch": 0.2596842105263158, + "grad_norm": 0.6171875, + "learning_rate": 0.00016451461988304097, + "loss": 3.6045, + "step": 4934 + }, + { + "epoch": 0.25973684210526315, + "grad_norm": 0.51953125, + "learning_rate": 0.00016450292397660822, + "loss": 3.7897, + "step": 4935 + }, + { + "epoch": 0.2597894736842105, + "grad_norm": 0.63671875, + "learning_rate": 0.00016449122807017546, + "loss": 4.0142, + "step": 4936 + }, + { + "epoch": 0.2598421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.0001644795321637427, + "loss": 3.9155, + "step": 4937 + }, + { + "epoch": 0.2598947368421053, + "grad_norm": 0.51953125, + "learning_rate": 0.00016446783625730994, + "loss": 3.9001, + "step": 4938 + }, + { + "epoch": 0.25994736842105265, + "grad_norm": 0.50390625, + "learning_rate": 0.0001644561403508772, + "loss": 3.6151, + "step": 4939 + }, + { + "epoch": 0.26, + "grad_norm": 0.5078125, + "learning_rate": 0.00016444444444444444, + "loss": 3.6034, + "step": 4940 + }, + { + "epoch": 0.26005263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00016443274853801169, + "loss": 4.0001, + "step": 4941 + }, + { + "epoch": 0.26010526315789473, + "grad_norm": 0.5625, + "learning_rate": 0.00016442105263157896, + "loss": 4.1093, + "step": 4942 + }, + { + "epoch": 0.2601578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.0001644093567251462, + "loss": 3.8586, + "step": 4943 + }, + { + "epoch": 0.26021052631578945, + "grad_norm": 0.6796875, + "learning_rate": 0.00016439766081871346, + "loss": 3.7968, + "step": 4944 + }, + { + "epoch": 0.26026315789473686, + "grad_norm": 0.5234375, + "learning_rate": 0.0001643859649122807, + "loss": 3.6965, + "step": 4945 + }, + { + "epoch": 0.2603157894736842, + "grad_norm": 0.59765625, + "learning_rate": 0.00016437426900584796, + "loss": 4.1463, + "step": 4946 + }, + { + "epoch": 0.2603684210526316, + "grad_norm": 0.57421875, + "learning_rate": 0.0001643625730994152, + "loss": 3.6891, + "step": 4947 + }, + { + "epoch": 0.26042105263157894, + "grad_norm": 0.50390625, + "learning_rate": 0.00016435087719298246, + "loss": 3.6278, + "step": 4948 + }, + { + "epoch": 0.2604736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.00016433918128654974, + "loss": 3.7413, + "step": 4949 + }, + { + "epoch": 0.26052631578947366, + "grad_norm": 0.54296875, + "learning_rate": 0.00016432748538011699, + "loss": 3.6244, + "step": 4950 + }, + { + "epoch": 0.2605789473684211, + "grad_norm": 0.49609375, + "learning_rate": 0.00016431578947368424, + "loss": 4.1066, + "step": 4951 + }, + { + "epoch": 0.26063157894736844, + "grad_norm": 0.546875, + "learning_rate": 0.00016430409356725146, + "loss": 3.8362, + "step": 4952 + }, + { + "epoch": 0.2606842105263158, + "grad_norm": 0.55078125, + "learning_rate": 0.0001642923976608187, + "loss": 3.8087, + "step": 4953 + }, + { + "epoch": 0.26073684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00016428070175438596, + "loss": 3.8451, + "step": 4954 + }, + { + "epoch": 0.2607894736842105, + "grad_norm": 0.65234375, + "learning_rate": 0.0001642690058479532, + "loss": 3.5748, + "step": 4955 + }, + { + "epoch": 0.2608421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00016425730994152046, + "loss": 3.6801, + "step": 4956 + }, + { + "epoch": 0.2608947368421053, + "grad_norm": 0.53125, + "learning_rate": 0.00016424561403508773, + "loss": 4.0623, + "step": 4957 + }, + { + "epoch": 0.26094736842105265, + "grad_norm": 0.609375, + "learning_rate": 0.00016423391812865498, + "loss": 4.0336, + "step": 4958 + }, + { + "epoch": 0.261, + "grad_norm": 0.53125, + "learning_rate": 0.00016422222222222223, + "loss": 3.6759, + "step": 4959 + }, + { + "epoch": 0.26105263157894737, + "grad_norm": 0.609375, + "learning_rate": 0.00016421052631578948, + "loss": 4.1324, + "step": 4960 + }, + { + "epoch": 0.26110526315789473, + "grad_norm": 0.58203125, + "learning_rate": 0.00016419883040935673, + "loss": 3.713, + "step": 4961 + }, + { + "epoch": 0.2611578947368421, + "grad_norm": 0.56640625, + "learning_rate": 0.00016418713450292398, + "loss": 3.9126, + "step": 4962 + }, + { + "epoch": 0.26121052631578945, + "grad_norm": 0.5546875, + "learning_rate": 0.00016417543859649123, + "loss": 4.2722, + "step": 4963 + }, + { + "epoch": 0.26126315789473686, + "grad_norm": 0.66015625, + "learning_rate": 0.0001641637426900585, + "loss": 4.101, + "step": 4964 + }, + { + "epoch": 0.2613157894736842, + "grad_norm": 0.64453125, + "learning_rate": 0.00016415204678362576, + "loss": 3.6296, + "step": 4965 + }, + { + "epoch": 0.2613684210526316, + "grad_norm": 0.5546875, + "learning_rate": 0.000164140350877193, + "loss": 3.9935, + "step": 4966 + }, + { + "epoch": 0.26142105263157894, + "grad_norm": 0.6640625, + "learning_rate": 0.00016412865497076023, + "loss": 4.0684, + "step": 4967 + }, + { + "epoch": 0.2614736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00016411695906432748, + "loss": 3.7632, + "step": 4968 + }, + { + "epoch": 0.26152631578947366, + "grad_norm": 0.515625, + "learning_rate": 0.00016410526315789473, + "loss": 3.6718, + "step": 4969 + }, + { + "epoch": 0.2615789473684211, + "grad_norm": 0.52734375, + "learning_rate": 0.00016409356725146198, + "loss": 3.6476, + "step": 4970 + }, + { + "epoch": 0.26163157894736844, + "grad_norm": 0.56640625, + "learning_rate": 0.00016408187134502923, + "loss": 4.1637, + "step": 4971 + }, + { + "epoch": 0.2616842105263158, + "grad_norm": 0.7734375, + "learning_rate": 0.0001640701754385965, + "loss": 3.6642, + "step": 4972 + }, + { + "epoch": 0.26173684210526316, + "grad_norm": 0.578125, + "learning_rate": 0.00016405847953216375, + "loss": 4.0264, + "step": 4973 + }, + { + "epoch": 0.2617894736842105, + "grad_norm": 0.61328125, + "learning_rate": 0.000164046783625731, + "loss": 4.2214, + "step": 4974 + }, + { + "epoch": 0.2618421052631579, + "grad_norm": 0.5703125, + "learning_rate": 0.00016403508771929825, + "loss": 3.4994, + "step": 4975 + }, + { + "epoch": 0.26189473684210524, + "grad_norm": 0.515625, + "learning_rate": 0.0001640233918128655, + "loss": 4.0282, + "step": 4976 + }, + { + "epoch": 0.26194736842105265, + "grad_norm": 0.51171875, + "learning_rate": 0.00016401169590643275, + "loss": 4.0866, + "step": 4977 + }, + { + "epoch": 0.262, + "grad_norm": 0.625, + "learning_rate": 0.000164, + "loss": 4.0631, + "step": 4978 + }, + { + "epoch": 0.26205263157894737, + "grad_norm": 0.546875, + "learning_rate": 0.00016398830409356728, + "loss": 3.4684, + "step": 4979 + }, + { + "epoch": 0.26210526315789473, + "grad_norm": 0.83203125, + "learning_rate": 0.00016397660818713453, + "loss": 3.9525, + "step": 4980 + }, + { + "epoch": 0.2621578947368421, + "grad_norm": 0.498046875, + "learning_rate": 0.00016396491228070178, + "loss": 3.7745, + "step": 4981 + }, + { + "epoch": 0.26221052631578945, + "grad_norm": 0.4765625, + "learning_rate": 0.00016395321637426903, + "loss": 4.0143, + "step": 4982 + }, + { + "epoch": 0.26226315789473686, + "grad_norm": 0.61328125, + "learning_rate": 0.00016394152046783625, + "loss": 3.5447, + "step": 4983 + }, + { + "epoch": 0.2623157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.0001639298245614035, + "loss": 4.079, + "step": 4984 + }, + { + "epoch": 0.2623684210526316, + "grad_norm": 0.5546875, + "learning_rate": 0.00016391812865497075, + "loss": 3.7735, + "step": 4985 + }, + { + "epoch": 0.26242105263157894, + "grad_norm": 0.5390625, + "learning_rate": 0.00016390643274853803, + "loss": 4.0159, + "step": 4986 + }, + { + "epoch": 0.2624736842105263, + "grad_norm": 0.80859375, + "learning_rate": 0.00016389473684210528, + "loss": 3.8543, + "step": 4987 + }, + { + "epoch": 0.26252631578947366, + "grad_norm": 0.5234375, + "learning_rate": 0.00016388304093567252, + "loss": 3.6475, + "step": 4988 + }, + { + "epoch": 0.2625789473684211, + "grad_norm": 0.51171875, + "learning_rate": 0.00016387134502923977, + "loss": 3.8243, + "step": 4989 + }, + { + "epoch": 0.26263157894736844, + "grad_norm": 0.53515625, + "learning_rate": 0.00016385964912280702, + "loss": 3.6746, + "step": 4990 + }, + { + "epoch": 0.2626842105263158, + "grad_norm": 0.61328125, + "learning_rate": 0.00016384795321637427, + "loss": 3.499, + "step": 4991 + }, + { + "epoch": 0.26273684210526316, + "grad_norm": 0.6328125, + "learning_rate": 0.00016383625730994152, + "loss": 3.6046, + "step": 4992 + }, + { + "epoch": 0.2627894736842105, + "grad_norm": 0.490234375, + "learning_rate": 0.00016382456140350877, + "loss": 3.7961, + "step": 4993 + }, + { + "epoch": 0.2628421052631579, + "grad_norm": 0.59375, + "learning_rate": 0.00016381286549707605, + "loss": 3.664, + "step": 4994 + }, + { + "epoch": 0.26289473684210524, + "grad_norm": 0.5234375, + "learning_rate": 0.0001638011695906433, + "loss": 3.853, + "step": 4995 + }, + { + "epoch": 0.26294736842105265, + "grad_norm": 0.6796875, + "learning_rate": 0.00016378947368421055, + "loss": 3.7792, + "step": 4996 + }, + { + "epoch": 0.263, + "grad_norm": 0.54296875, + "learning_rate": 0.0001637777777777778, + "loss": 3.9116, + "step": 4997 + }, + { + "epoch": 0.26305263157894737, + "grad_norm": 0.52734375, + "learning_rate": 0.00016376608187134502, + "loss": 4.1686, + "step": 4998 + }, + { + "epoch": 0.26310526315789473, + "grad_norm": 0.546875, + "learning_rate": 0.00016375438596491227, + "loss": 3.5232, + "step": 4999 + }, + { + "epoch": 0.2631578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00016374269005847952, + "loss": 4.0415, + "step": 5000 + }, + { + "epoch": 0.26321052631578945, + "grad_norm": 0.50390625, + "learning_rate": 0.0001637309941520468, + "loss": 3.6335, + "step": 5001 + }, + { + "epoch": 0.26326315789473687, + "grad_norm": 0.49609375, + "learning_rate": 0.00016371929824561405, + "loss": 4.0741, + "step": 5002 + }, + { + "epoch": 0.2633157894736842, + "grad_norm": 0.546875, + "learning_rate": 0.0001637076023391813, + "loss": 3.848, + "step": 5003 + }, + { + "epoch": 0.2633684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00016369590643274855, + "loss": 3.8405, + "step": 5004 + }, + { + "epoch": 0.26342105263157894, + "grad_norm": 0.515625, + "learning_rate": 0.0001636842105263158, + "loss": 3.8045, + "step": 5005 + }, + { + "epoch": 0.2634736842105263, + "grad_norm": 0.56640625, + "learning_rate": 0.00016367251461988304, + "loss": 4.009, + "step": 5006 + }, + { + "epoch": 0.26352631578947366, + "grad_norm": 0.515625, + "learning_rate": 0.0001636608187134503, + "loss": 3.6751, + "step": 5007 + }, + { + "epoch": 0.2635789473684211, + "grad_norm": 0.51171875, + "learning_rate": 0.00016364912280701754, + "loss": 3.7225, + "step": 5008 + }, + { + "epoch": 0.26363157894736844, + "grad_norm": 0.484375, + "learning_rate": 0.00016363742690058482, + "loss": 4.0261, + "step": 5009 + }, + { + "epoch": 0.2636842105263158, + "grad_norm": 0.54296875, + "learning_rate": 0.00016362573099415207, + "loss": 4.2891, + "step": 5010 + }, + { + "epoch": 0.26373684210526316, + "grad_norm": 0.57421875, + "learning_rate": 0.00016361403508771932, + "loss": 3.7404, + "step": 5011 + }, + { + "epoch": 0.2637894736842105, + "grad_norm": 0.83203125, + "learning_rate": 0.00016360233918128657, + "loss": 3.2998, + "step": 5012 + }, + { + "epoch": 0.2638421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.0001635906432748538, + "loss": 3.6973, + "step": 5013 + }, + { + "epoch": 0.26389473684210524, + "grad_norm": 0.5625, + "learning_rate": 0.00016357894736842104, + "loss": 3.7452, + "step": 5014 + }, + { + "epoch": 0.26394736842105265, + "grad_norm": 0.99609375, + "learning_rate": 0.0001635672514619883, + "loss": 3.6331, + "step": 5015 + }, + { + "epoch": 0.264, + "grad_norm": 0.494140625, + "learning_rate": 0.00016355555555555557, + "loss": 3.7364, + "step": 5016 + }, + { + "epoch": 0.26405263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00016354385964912282, + "loss": 3.7149, + "step": 5017 + }, + { + "epoch": 0.26410526315789473, + "grad_norm": 0.5234375, + "learning_rate": 0.00016353216374269007, + "loss": 3.7679, + "step": 5018 + }, + { + "epoch": 0.2641578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00016352046783625732, + "loss": 4.4798, + "step": 5019 + }, + { + "epoch": 0.26421052631578945, + "grad_norm": 1.0234375, + "learning_rate": 0.00016350877192982457, + "loss": 3.404, + "step": 5020 + }, + { + "epoch": 0.26426315789473687, + "grad_norm": 0.447265625, + "learning_rate": 0.00016349707602339182, + "loss": 3.344, + "step": 5021 + }, + { + "epoch": 0.2643157894736842, + "grad_norm": 0.48828125, + "learning_rate": 0.00016348538011695907, + "loss": 3.787, + "step": 5022 + }, + { + "epoch": 0.2643684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00016347368421052634, + "loss": 3.6089, + "step": 5023 + }, + { + "epoch": 0.26442105263157895, + "grad_norm": 1.15625, + "learning_rate": 0.0001634619883040936, + "loss": 3.6195, + "step": 5024 + }, + { + "epoch": 0.2644736842105263, + "grad_norm": 0.5234375, + "learning_rate": 0.00016345029239766084, + "loss": 4.0361, + "step": 5025 + }, + { + "epoch": 0.26452631578947366, + "grad_norm": 0.4921875, + "learning_rate": 0.0001634385964912281, + "loss": 3.4207, + "step": 5026 + }, + { + "epoch": 0.2645789473684211, + "grad_norm": 0.50390625, + "learning_rate": 0.00016342690058479534, + "loss": 3.9782, + "step": 5027 + }, + { + "epoch": 0.26463157894736844, + "grad_norm": 0.53125, + "learning_rate": 0.0001634152046783626, + "loss": 4.0133, + "step": 5028 + }, + { + "epoch": 0.2646842105263158, + "grad_norm": 0.55078125, + "learning_rate": 0.0001634035087719298, + "loss": 4.0801, + "step": 5029 + }, + { + "epoch": 0.26473684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.00016339181286549706, + "loss": 3.6581, + "step": 5030 + }, + { + "epoch": 0.2647894736842105, + "grad_norm": 0.58984375, + "learning_rate": 0.00016338011695906434, + "loss": 3.1047, + "step": 5031 + }, + { + "epoch": 0.2648421052631579, + "grad_norm": 0.55859375, + "learning_rate": 0.0001633684210526316, + "loss": 3.6357, + "step": 5032 + }, + { + "epoch": 0.26489473684210524, + "grad_norm": 0.494140625, + "learning_rate": 0.00016335672514619884, + "loss": 3.6647, + "step": 5033 + }, + { + "epoch": 0.26494736842105265, + "grad_norm": 0.51171875, + "learning_rate": 0.0001633450292397661, + "loss": 3.4592, + "step": 5034 + }, + { + "epoch": 0.265, + "grad_norm": 0.490234375, + "learning_rate": 0.00016333333333333334, + "loss": 3.7965, + "step": 5035 + }, + { + "epoch": 0.2650526315789474, + "grad_norm": 0.53125, + "learning_rate": 0.0001633216374269006, + "loss": 3.7175, + "step": 5036 + }, + { + "epoch": 0.26510526315789473, + "grad_norm": 0.55078125, + "learning_rate": 0.00016330994152046784, + "loss": 3.9701, + "step": 5037 + }, + { + "epoch": 0.2651578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.0001632982456140351, + "loss": 3.9663, + "step": 5038 + }, + { + "epoch": 0.26521052631578945, + "grad_norm": 0.48828125, + "learning_rate": 0.00016328654970760236, + "loss": 3.7865, + "step": 5039 + }, + { + "epoch": 0.26526315789473687, + "grad_norm": 0.51171875, + "learning_rate": 0.0001632748538011696, + "loss": 4.1897, + "step": 5040 + }, + { + "epoch": 0.2653157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00016326315789473686, + "loss": 4.0059, + "step": 5041 + }, + { + "epoch": 0.2653684210526316, + "grad_norm": 0.546875, + "learning_rate": 0.0001632514619883041, + "loss": 3.7666, + "step": 5042 + }, + { + "epoch": 0.26542105263157895, + "grad_norm": 0.58203125, + "learning_rate": 0.00016323976608187136, + "loss": 3.7323, + "step": 5043 + }, + { + "epoch": 0.2654736842105263, + "grad_norm": 0.6015625, + "learning_rate": 0.00016322807017543858, + "loss": 4.0573, + "step": 5044 + }, + { + "epoch": 0.26552631578947367, + "grad_norm": 0.53125, + "learning_rate": 0.00016321637426900583, + "loss": 3.7089, + "step": 5045 + }, + { + "epoch": 0.265578947368421, + "grad_norm": 0.59375, + "learning_rate": 0.0001632046783625731, + "loss": 3.994, + "step": 5046 + }, + { + "epoch": 0.26563157894736844, + "grad_norm": 0.68359375, + "learning_rate": 0.00016319298245614036, + "loss": 3.7544, + "step": 5047 + }, + { + "epoch": 0.2656842105263158, + "grad_norm": 0.486328125, + "learning_rate": 0.0001631812865497076, + "loss": 3.6318, + "step": 5048 + }, + { + "epoch": 0.26573684210526316, + "grad_norm": 0.60546875, + "learning_rate": 0.00016316959064327486, + "loss": 3.3873, + "step": 5049 + }, + { + "epoch": 0.2657894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.0001631578947368421, + "loss": 3.4476, + "step": 5050 + }, + { + "epoch": 0.2658421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00016314619883040936, + "loss": 4.1999, + "step": 5051 + }, + { + "epoch": 0.26589473684210524, + "grad_norm": 0.51953125, + "learning_rate": 0.0001631345029239766, + "loss": 3.7279, + "step": 5052 + }, + { + "epoch": 0.26594736842105265, + "grad_norm": 0.53125, + "learning_rate": 0.00016312280701754388, + "loss": 3.3851, + "step": 5053 + }, + { + "epoch": 0.266, + "grad_norm": 0.46875, + "learning_rate": 0.00016311111111111113, + "loss": 3.8161, + "step": 5054 + }, + { + "epoch": 0.2660526315789474, + "grad_norm": 0.51171875, + "learning_rate": 0.00016309941520467838, + "loss": 3.9849, + "step": 5055 + }, + { + "epoch": 0.26610526315789473, + "grad_norm": 0.54296875, + "learning_rate": 0.00016308771929824563, + "loss": 3.6812, + "step": 5056 + }, + { + "epoch": 0.2661578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.00016307602339181288, + "loss": 4.1516, + "step": 5057 + }, + { + "epoch": 0.26621052631578945, + "grad_norm": 0.61328125, + "learning_rate": 0.00016306432748538013, + "loss": 3.0885, + "step": 5058 + }, + { + "epoch": 0.26626315789473687, + "grad_norm": 0.62109375, + "learning_rate": 0.00016305263157894735, + "loss": 3.2122, + "step": 5059 + }, + { + "epoch": 0.26631578947368423, + "grad_norm": 0.515625, + "learning_rate": 0.00016304093567251463, + "loss": 4.0213, + "step": 5060 + }, + { + "epoch": 0.2663684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.00016302923976608188, + "loss": 3.7933, + "step": 5061 + }, + { + "epoch": 0.26642105263157895, + "grad_norm": 0.5625, + "learning_rate": 0.00016301754385964913, + "loss": 3.6888, + "step": 5062 + }, + { + "epoch": 0.2664736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.00016300584795321638, + "loss": 3.6566, + "step": 5063 + }, + { + "epoch": 0.26652631578947367, + "grad_norm": 0.470703125, + "learning_rate": 0.00016299415204678363, + "loss": 3.799, + "step": 5064 + }, + { + "epoch": 0.266578947368421, + "grad_norm": 0.46875, + "learning_rate": 0.00016298245614035088, + "loss": 3.7568, + "step": 5065 + }, + { + "epoch": 0.26663157894736844, + "grad_norm": 0.51171875, + "learning_rate": 0.00016297076023391813, + "loss": 3.9807, + "step": 5066 + }, + { + "epoch": 0.2666842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.00016295906432748538, + "loss": 3.7604, + "step": 5067 + }, + { + "epoch": 0.26673684210526316, + "grad_norm": 0.546875, + "learning_rate": 0.00016294736842105265, + "loss": 3.7538, + "step": 5068 + }, + { + "epoch": 0.2667894736842105, + "grad_norm": 0.6171875, + "learning_rate": 0.0001629356725146199, + "loss": 3.6864, + "step": 5069 + }, + { + "epoch": 0.2668421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00016292397660818715, + "loss": 3.7818, + "step": 5070 + }, + { + "epoch": 0.26689473684210524, + "grad_norm": 0.55859375, + "learning_rate": 0.0001629122807017544, + "loss": 3.5653, + "step": 5071 + }, + { + "epoch": 0.26694736842105266, + "grad_norm": 0.6328125, + "learning_rate": 0.00016290058479532165, + "loss": 3.9322, + "step": 5072 + }, + { + "epoch": 0.267, + "grad_norm": 0.53125, + "learning_rate": 0.0001628888888888889, + "loss": 3.8579, + "step": 5073 + }, + { + "epoch": 0.2670526315789474, + "grad_norm": 0.578125, + "learning_rate": 0.00016287719298245615, + "loss": 3.6987, + "step": 5074 + }, + { + "epoch": 0.26710526315789473, + "grad_norm": 0.69140625, + "learning_rate": 0.0001628654970760234, + "loss": 3.7784, + "step": 5075 + }, + { + "epoch": 0.2671578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.00016285380116959065, + "loss": 4.2473, + "step": 5076 + }, + { + "epoch": 0.26721052631578945, + "grad_norm": 0.5625, + "learning_rate": 0.0001628421052631579, + "loss": 4.1557, + "step": 5077 + }, + { + "epoch": 0.26726315789473687, + "grad_norm": 0.53515625, + "learning_rate": 0.00016283040935672515, + "loss": 3.8552, + "step": 5078 + }, + { + "epoch": 0.26731578947368423, + "grad_norm": 0.6171875, + "learning_rate": 0.0001628187134502924, + "loss": 3.8412, + "step": 5079 + }, + { + "epoch": 0.2673684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00016280701754385965, + "loss": 3.8258, + "step": 5080 + }, + { + "epoch": 0.26742105263157895, + "grad_norm": 0.5703125, + "learning_rate": 0.0001627953216374269, + "loss": 3.9417, + "step": 5081 + }, + { + "epoch": 0.2674736842105263, + "grad_norm": 0.59765625, + "learning_rate": 0.00016278362573099415, + "loss": 3.4765, + "step": 5082 + }, + { + "epoch": 0.26752631578947367, + "grad_norm": 0.482421875, + "learning_rate": 0.00016277192982456143, + "loss": 3.8993, + "step": 5083 + }, + { + "epoch": 0.267578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00016276023391812868, + "loss": 4.0789, + "step": 5084 + }, + { + "epoch": 0.26763157894736844, + "grad_norm": 0.7421875, + "learning_rate": 0.00016274853801169593, + "loss": 3.7753, + "step": 5085 + }, + { + "epoch": 0.2676842105263158, + "grad_norm": 0.546875, + "learning_rate": 0.00016273684210526317, + "loss": 4.1837, + "step": 5086 + }, + { + "epoch": 0.26773684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00016272514619883042, + "loss": 3.6478, + "step": 5087 + }, + { + "epoch": 0.2677894736842105, + "grad_norm": 0.5234375, + "learning_rate": 0.00016271345029239767, + "loss": 3.6767, + "step": 5088 + }, + { + "epoch": 0.2678421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00016270175438596492, + "loss": 4.1722, + "step": 5089 + }, + { + "epoch": 0.26789473684210524, + "grad_norm": 0.53515625, + "learning_rate": 0.00016269005847953217, + "loss": 3.5423, + "step": 5090 + }, + { + "epoch": 0.26794736842105266, + "grad_norm": 0.95703125, + "learning_rate": 0.00016267836257309942, + "loss": 3.5099, + "step": 5091 + }, + { + "epoch": 0.268, + "grad_norm": 0.47265625, + "learning_rate": 0.00016266666666666667, + "loss": 3.9942, + "step": 5092 + }, + { + "epoch": 0.2680526315789474, + "grad_norm": 0.494140625, + "learning_rate": 0.00016265497076023392, + "loss": 3.7207, + "step": 5093 + }, + { + "epoch": 0.26810526315789474, + "grad_norm": 0.482421875, + "learning_rate": 0.00016264327485380117, + "loss": 4.1595, + "step": 5094 + }, + { + "epoch": 0.2681578947368421, + "grad_norm": 0.546875, + "learning_rate": 0.00016263157894736842, + "loss": 4.0416, + "step": 5095 + }, + { + "epoch": 0.26821052631578945, + "grad_norm": 0.5078125, + "learning_rate": 0.00016261988304093567, + "loss": 3.811, + "step": 5096 + }, + { + "epoch": 0.2682631578947368, + "grad_norm": 0.5546875, + "learning_rate": 0.00016260818713450295, + "loss": 3.6155, + "step": 5097 + }, + { + "epoch": 0.26831578947368423, + "grad_norm": 0.52734375, + "learning_rate": 0.0001625964912280702, + "loss": 3.9907, + "step": 5098 + }, + { + "epoch": 0.2683684210526316, + "grad_norm": 0.5546875, + "learning_rate": 0.00016258479532163745, + "loss": 3.3231, + "step": 5099 + }, + { + "epoch": 0.26842105263157895, + "grad_norm": 0.484375, + "learning_rate": 0.0001625730994152047, + "loss": 3.8355, + "step": 5100 + }, + { + "epoch": 0.2684736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.00016256140350877195, + "loss": 3.4634, + "step": 5101 + }, + { + "epoch": 0.26852631578947367, + "grad_norm": 0.5625, + "learning_rate": 0.0001625497076023392, + "loss": 3.5325, + "step": 5102 + }, + { + "epoch": 0.26857894736842103, + "grad_norm": 0.515625, + "learning_rate": 0.00016253801169590644, + "loss": 3.7723, + "step": 5103 + }, + { + "epoch": 0.26863157894736844, + "grad_norm": 0.6015625, + "learning_rate": 0.0001625263157894737, + "loss": 3.9337, + "step": 5104 + }, + { + "epoch": 0.2686842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00016251461988304094, + "loss": 3.7753, + "step": 5105 + }, + { + "epoch": 0.26873684210526316, + "grad_norm": 0.54296875, + "learning_rate": 0.0001625029239766082, + "loss": 3.5364, + "step": 5106 + }, + { + "epoch": 0.2687894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.00016249122807017544, + "loss": 3.7156, + "step": 5107 + }, + { + "epoch": 0.2688421052631579, + "grad_norm": 0.6796875, + "learning_rate": 0.0001624795321637427, + "loss": 3.6088, + "step": 5108 + }, + { + "epoch": 0.26889473684210524, + "grad_norm": 0.61328125, + "learning_rate": 0.00016246783625730994, + "loss": 3.3916, + "step": 5109 + }, + { + "epoch": 0.26894736842105266, + "grad_norm": 0.498046875, + "learning_rate": 0.0001624561403508772, + "loss": 3.5194, + "step": 5110 + }, + { + "epoch": 0.269, + "grad_norm": 0.48046875, + "learning_rate": 0.00016244444444444444, + "loss": 3.9343, + "step": 5111 + }, + { + "epoch": 0.2690526315789474, + "grad_norm": 0.515625, + "learning_rate": 0.00016243274853801172, + "loss": 3.6125, + "step": 5112 + }, + { + "epoch": 0.26910526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00016242105263157897, + "loss": 3.6967, + "step": 5113 + }, + { + "epoch": 0.2691578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00016240935672514622, + "loss": 4.0296, + "step": 5114 + }, + { + "epoch": 0.26921052631578946, + "grad_norm": 0.59765625, + "learning_rate": 0.00016239766081871347, + "loss": 3.3486, + "step": 5115 + }, + { + "epoch": 0.2692631578947368, + "grad_norm": 0.58203125, + "learning_rate": 0.00016238596491228072, + "loss": 3.4852, + "step": 5116 + }, + { + "epoch": 0.26931578947368423, + "grad_norm": 0.6015625, + "learning_rate": 0.00016237426900584797, + "loss": 3.6787, + "step": 5117 + }, + { + "epoch": 0.2693684210526316, + "grad_norm": 0.55078125, + "learning_rate": 0.00016236257309941522, + "loss": 3.7856, + "step": 5118 + }, + { + "epoch": 0.26942105263157895, + "grad_norm": 0.6640625, + "learning_rate": 0.00016235087719298247, + "loss": 3.6497, + "step": 5119 + }, + { + "epoch": 0.2694736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00016233918128654972, + "loss": 3.2023, + "step": 5120 + }, + { + "epoch": 0.26952631578947367, + "grad_norm": 0.48828125, + "learning_rate": 0.00016232748538011696, + "loss": 4.1949, + "step": 5121 + }, + { + "epoch": 0.26957894736842103, + "grad_norm": 0.58203125, + "learning_rate": 0.00016231578947368421, + "loss": 3.3025, + "step": 5122 + }, + { + "epoch": 0.26963157894736844, + "grad_norm": 0.6640625, + "learning_rate": 0.00016230409356725146, + "loss": 3.9277, + "step": 5123 + }, + { + "epoch": 0.2696842105263158, + "grad_norm": 0.76953125, + "learning_rate": 0.0001622923976608187, + "loss": 4.0602, + "step": 5124 + }, + { + "epoch": 0.26973684210526316, + "grad_norm": 0.6640625, + "learning_rate": 0.00016228070175438596, + "loss": 3.5693, + "step": 5125 + }, + { + "epoch": 0.2697894736842105, + "grad_norm": 0.56640625, + "learning_rate": 0.0001622690058479532, + "loss": 4.0642, + "step": 5126 + }, + { + "epoch": 0.2698421052631579, + "grad_norm": 0.5703125, + "learning_rate": 0.0001622573099415205, + "loss": 3.3107, + "step": 5127 + }, + { + "epoch": 0.26989473684210524, + "grad_norm": 0.470703125, + "learning_rate": 0.00016224561403508774, + "loss": 4.0788, + "step": 5128 + }, + { + "epoch": 0.26994736842105266, + "grad_norm": 0.482421875, + "learning_rate": 0.000162233918128655, + "loss": 3.9725, + "step": 5129 + }, + { + "epoch": 0.27, + "grad_norm": 0.51953125, + "learning_rate": 0.00016222222222222224, + "loss": 3.9814, + "step": 5130 + }, + { + "epoch": 0.2700526315789474, + "grad_norm": 0.56640625, + "learning_rate": 0.0001622105263157895, + "loss": 3.5065, + "step": 5131 + }, + { + "epoch": 0.27010526315789474, + "grad_norm": 0.55078125, + "learning_rate": 0.00016219883040935674, + "loss": 4.1071, + "step": 5132 + }, + { + "epoch": 0.2701578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.000162187134502924, + "loss": 3.8506, + "step": 5133 + }, + { + "epoch": 0.27021052631578946, + "grad_norm": 0.58984375, + "learning_rate": 0.00016217543859649124, + "loss": 3.8317, + "step": 5134 + }, + { + "epoch": 0.2702631578947368, + "grad_norm": 0.578125, + "learning_rate": 0.00016216374269005849, + "loss": 3.9425, + "step": 5135 + }, + { + "epoch": 0.27031578947368423, + "grad_norm": 0.61328125, + "learning_rate": 0.00016215204678362574, + "loss": 4.0621, + "step": 5136 + }, + { + "epoch": 0.2703684210526316, + "grad_norm": 0.56640625, + "learning_rate": 0.00016214035087719299, + "loss": 4.2104, + "step": 5137 + }, + { + "epoch": 0.27042105263157895, + "grad_norm": 0.6171875, + "learning_rate": 0.00016212865497076023, + "loss": 3.8066, + "step": 5138 + }, + { + "epoch": 0.2704736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.00016211695906432748, + "loss": 3.9455, + "step": 5139 + }, + { + "epoch": 0.27052631578947367, + "grad_norm": 0.515625, + "learning_rate": 0.00016210526315789473, + "loss": 4.1761, + "step": 5140 + }, + { + "epoch": 0.27057894736842103, + "grad_norm": 0.71484375, + "learning_rate": 0.00016209356725146198, + "loss": 4.1206, + "step": 5141 + }, + { + "epoch": 0.27063157894736845, + "grad_norm": 0.53125, + "learning_rate": 0.00016208187134502926, + "loss": 4.1585, + "step": 5142 + }, + { + "epoch": 0.2706842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.0001620701754385965, + "loss": 3.8949, + "step": 5143 + }, + { + "epoch": 0.27073684210526316, + "grad_norm": 0.56640625, + "learning_rate": 0.00016205847953216376, + "loss": 4.0682, + "step": 5144 + }, + { + "epoch": 0.2707894736842105, + "grad_norm": 0.6171875, + "learning_rate": 0.000162046783625731, + "loss": 3.4043, + "step": 5145 + }, + { + "epoch": 0.2708421052631579, + "grad_norm": 0.64453125, + "learning_rate": 0.00016203508771929826, + "loss": 3.639, + "step": 5146 + }, + { + "epoch": 0.27089473684210524, + "grad_norm": 0.5703125, + "learning_rate": 0.0001620233918128655, + "loss": 3.8092, + "step": 5147 + }, + { + "epoch": 0.27094736842105266, + "grad_norm": 0.56640625, + "learning_rate": 0.00016201169590643276, + "loss": 3.8886, + "step": 5148 + }, + { + "epoch": 0.271, + "grad_norm": 0.6015625, + "learning_rate": 0.000162, + "loss": 4.2988, + "step": 5149 + }, + { + "epoch": 0.2710526315789474, + "grad_norm": 0.61328125, + "learning_rate": 0.00016198830409356726, + "loss": 4.1071, + "step": 5150 + }, + { + "epoch": 0.27110526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.0001619766081871345, + "loss": 3.7556, + "step": 5151 + }, + { + "epoch": 0.2711578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00016196491228070176, + "loss": 3.8087, + "step": 5152 + }, + { + "epoch": 0.27121052631578946, + "grad_norm": 0.54296875, + "learning_rate": 0.000161953216374269, + "loss": 3.7016, + "step": 5153 + }, + { + "epoch": 0.2712631578947368, + "grad_norm": 0.55078125, + "learning_rate": 0.00016194152046783626, + "loss": 4.1574, + "step": 5154 + }, + { + "epoch": 0.27131578947368423, + "grad_norm": 0.5859375, + "learning_rate": 0.0001619298245614035, + "loss": 3.6993, + "step": 5155 + }, + { + "epoch": 0.2713684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00016191812865497075, + "loss": 3.795, + "step": 5156 + }, + { + "epoch": 0.27142105263157895, + "grad_norm": 0.52734375, + "learning_rate": 0.00016190643274853803, + "loss": 3.1178, + "step": 5157 + }, + { + "epoch": 0.2714736842105263, + "grad_norm": 0.46875, + "learning_rate": 0.00016189473684210528, + "loss": 3.4253, + "step": 5158 + }, + { + "epoch": 0.27152631578947367, + "grad_norm": 0.466796875, + "learning_rate": 0.00016188304093567253, + "loss": 3.9238, + "step": 5159 + }, + { + "epoch": 0.27157894736842103, + "grad_norm": 0.5078125, + "learning_rate": 0.00016187134502923978, + "loss": 3.8572, + "step": 5160 + }, + { + "epoch": 0.27163157894736845, + "grad_norm": 0.64453125, + "learning_rate": 0.00016185964912280703, + "loss": 4.1581, + "step": 5161 + }, + { + "epoch": 0.2716842105263158, + "grad_norm": 0.58203125, + "learning_rate": 0.00016184795321637428, + "loss": 3.8225, + "step": 5162 + }, + { + "epoch": 0.27173684210526317, + "grad_norm": 0.515625, + "learning_rate": 0.00016183625730994153, + "loss": 3.7232, + "step": 5163 + }, + { + "epoch": 0.2717894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.00016182456140350878, + "loss": 3.6342, + "step": 5164 + }, + { + "epoch": 0.2718421052631579, + "grad_norm": 0.5859375, + "learning_rate": 0.00016181286549707603, + "loss": 3.3561, + "step": 5165 + }, + { + "epoch": 0.27189473684210524, + "grad_norm": 0.447265625, + "learning_rate": 0.00016180116959064328, + "loss": 3.8207, + "step": 5166 + }, + { + "epoch": 0.2719473684210526, + "grad_norm": 0.578125, + "learning_rate": 0.00016178947368421053, + "loss": 3.9759, + "step": 5167 + }, + { + "epoch": 0.272, + "grad_norm": 0.51953125, + "learning_rate": 0.00016177777777777778, + "loss": 3.8048, + "step": 5168 + }, + { + "epoch": 0.2720526315789474, + "grad_norm": 0.5703125, + "learning_rate": 0.00016176608187134503, + "loss": 4.0737, + "step": 5169 + }, + { + "epoch": 0.27210526315789474, + "grad_norm": 0.65234375, + "learning_rate": 0.00016175438596491228, + "loss": 3.6156, + "step": 5170 + }, + { + "epoch": 0.2721578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00016174269005847953, + "loss": 3.5352, + "step": 5171 + }, + { + "epoch": 0.27221052631578946, + "grad_norm": 0.61328125, + "learning_rate": 0.0001617309941520468, + "loss": 3.6914, + "step": 5172 + }, + { + "epoch": 0.2722631578947368, + "grad_norm": 0.515625, + "learning_rate": 0.00016171929824561405, + "loss": 3.4581, + "step": 5173 + }, + { + "epoch": 0.27231578947368423, + "grad_norm": 0.494140625, + "learning_rate": 0.0001617076023391813, + "loss": 3.9372, + "step": 5174 + }, + { + "epoch": 0.2723684210526316, + "grad_norm": 0.498046875, + "learning_rate": 0.00016169590643274855, + "loss": 3.7667, + "step": 5175 + }, + { + "epoch": 0.27242105263157895, + "grad_norm": 0.5, + "learning_rate": 0.0001616842105263158, + "loss": 3.8652, + "step": 5176 + }, + { + "epoch": 0.2724736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00016167251461988305, + "loss": 3.6954, + "step": 5177 + }, + { + "epoch": 0.27252631578947367, + "grad_norm": 0.51953125, + "learning_rate": 0.0001616608187134503, + "loss": 4.002, + "step": 5178 + }, + { + "epoch": 0.27257894736842103, + "grad_norm": 0.515625, + "learning_rate": 0.00016164912280701755, + "loss": 3.923, + "step": 5179 + }, + { + "epoch": 0.27263157894736845, + "grad_norm": 0.5234375, + "learning_rate": 0.0001616374269005848, + "loss": 3.454, + "step": 5180 + }, + { + "epoch": 0.2726842105263158, + "grad_norm": 0.5625, + "learning_rate": 0.00016162573099415205, + "loss": 3.6053, + "step": 5181 + }, + { + "epoch": 0.27273684210526317, + "grad_norm": 0.490234375, + "learning_rate": 0.0001616140350877193, + "loss": 3.6305, + "step": 5182 + }, + { + "epoch": 0.2727894736842105, + "grad_norm": 0.53515625, + "learning_rate": 0.00016160233918128655, + "loss": 4.2329, + "step": 5183 + }, + { + "epoch": 0.2728421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.0001615906432748538, + "loss": 3.6666, + "step": 5184 + }, + { + "epoch": 0.27289473684210525, + "grad_norm": 0.515625, + "learning_rate": 0.00016157894736842105, + "loss": 3.3189, + "step": 5185 + }, + { + "epoch": 0.2729473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00016156725146198832, + "loss": 3.8734, + "step": 5186 + }, + { + "epoch": 0.273, + "grad_norm": 0.49609375, + "learning_rate": 0.00016155555555555557, + "loss": 4.1805, + "step": 5187 + }, + { + "epoch": 0.2730526315789474, + "grad_norm": 0.68359375, + "learning_rate": 0.00016154385964912282, + "loss": 4.4648, + "step": 5188 + }, + { + "epoch": 0.27310526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.00016153216374269007, + "loss": 3.8746, + "step": 5189 + }, + { + "epoch": 0.2731578947368421, + "grad_norm": 0.58203125, + "learning_rate": 0.00016152046783625732, + "loss": 3.6024, + "step": 5190 + }, + { + "epoch": 0.27321052631578946, + "grad_norm": 0.4921875, + "learning_rate": 0.00016150877192982457, + "loss": 3.949, + "step": 5191 + }, + { + "epoch": 0.2732631578947368, + "grad_norm": 0.515625, + "learning_rate": 0.00016149707602339182, + "loss": 3.1931, + "step": 5192 + }, + { + "epoch": 0.27331578947368423, + "grad_norm": 0.486328125, + "learning_rate": 0.00016148538011695907, + "loss": 4.0102, + "step": 5193 + }, + { + "epoch": 0.2733684210526316, + "grad_norm": 0.66796875, + "learning_rate": 0.00016147368421052632, + "loss": 3.3242, + "step": 5194 + }, + { + "epoch": 0.27342105263157895, + "grad_norm": 0.52734375, + "learning_rate": 0.00016146198830409357, + "loss": 4.0422, + "step": 5195 + }, + { + "epoch": 0.2734736842105263, + "grad_norm": 0.61328125, + "learning_rate": 0.00016145029239766082, + "loss": 3.8801, + "step": 5196 + }, + { + "epoch": 0.2735263157894737, + "grad_norm": 0.7890625, + "learning_rate": 0.00016143859649122807, + "loss": 3.2587, + "step": 5197 + }, + { + "epoch": 0.27357894736842103, + "grad_norm": 0.5078125, + "learning_rate": 0.00016142690058479532, + "loss": 3.8195, + "step": 5198 + }, + { + "epoch": 0.27363157894736845, + "grad_norm": 0.5078125, + "learning_rate": 0.00016141520467836257, + "loss": 3.938, + "step": 5199 + }, + { + "epoch": 0.2736842105263158, + "grad_norm": 0.5390625, + "learning_rate": 0.00016140350877192982, + "loss": 3.8406, + "step": 5200 + }, + { + "epoch": 0.27373684210526317, + "grad_norm": 0.671875, + "learning_rate": 0.0001613918128654971, + "loss": 3.6672, + "step": 5201 + }, + { + "epoch": 0.2737894736842105, + "grad_norm": 0.61328125, + "learning_rate": 0.00016138011695906434, + "loss": 4.0395, + "step": 5202 + }, + { + "epoch": 0.2738421052631579, + "grad_norm": 0.64453125, + "learning_rate": 0.0001613684210526316, + "loss": 4.0375, + "step": 5203 + }, + { + "epoch": 0.27389473684210525, + "grad_norm": 0.58984375, + "learning_rate": 0.00016135672514619884, + "loss": 3.7278, + "step": 5204 + }, + { + "epoch": 0.2739473684210526, + "grad_norm": 0.5, + "learning_rate": 0.0001613450292397661, + "loss": 4.313, + "step": 5205 + }, + { + "epoch": 0.274, + "grad_norm": 0.50390625, + "learning_rate": 0.00016133333333333334, + "loss": 3.7669, + "step": 5206 + }, + { + "epoch": 0.2740526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.0001613216374269006, + "loss": 3.8669, + "step": 5207 + }, + { + "epoch": 0.27410526315789474, + "grad_norm": 0.5546875, + "learning_rate": 0.00016130994152046784, + "loss": 3.4412, + "step": 5208 + }, + { + "epoch": 0.2741578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.0001612982456140351, + "loss": 3.5747, + "step": 5209 + }, + { + "epoch": 0.27421052631578946, + "grad_norm": 0.48828125, + "learning_rate": 0.00016128654970760234, + "loss": 4.1237, + "step": 5210 + }, + { + "epoch": 0.2742631578947368, + "grad_norm": 0.52734375, + "learning_rate": 0.0001612748538011696, + "loss": 4.0605, + "step": 5211 + }, + { + "epoch": 0.27431578947368424, + "grad_norm": 0.6171875, + "learning_rate": 0.00016126315789473684, + "loss": 4.1083, + "step": 5212 + }, + { + "epoch": 0.2743684210526316, + "grad_norm": 0.58203125, + "learning_rate": 0.0001612514619883041, + "loss": 4.0079, + "step": 5213 + }, + { + "epoch": 0.27442105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00016123976608187134, + "loss": 3.8625, + "step": 5214 + }, + { + "epoch": 0.2744736842105263, + "grad_norm": 0.486328125, + "learning_rate": 0.0001612280701754386, + "loss": 3.9435, + "step": 5215 + }, + { + "epoch": 0.2745263157894737, + "grad_norm": 0.5546875, + "learning_rate": 0.00016121637426900587, + "loss": 4.0741, + "step": 5216 + }, + { + "epoch": 0.27457894736842103, + "grad_norm": 0.578125, + "learning_rate": 0.00016120467836257312, + "loss": 4.2346, + "step": 5217 + }, + { + "epoch": 0.27463157894736845, + "grad_norm": 0.5078125, + "learning_rate": 0.00016119298245614036, + "loss": 3.6018, + "step": 5218 + }, + { + "epoch": 0.2746842105263158, + "grad_norm": 0.52734375, + "learning_rate": 0.00016118128654970761, + "loss": 3.6769, + "step": 5219 + }, + { + "epoch": 0.27473684210526317, + "grad_norm": 0.5625, + "learning_rate": 0.00016116959064327486, + "loss": 3.7238, + "step": 5220 + }, + { + "epoch": 0.27478947368421053, + "grad_norm": 0.58203125, + "learning_rate": 0.00016115789473684211, + "loss": 4.2101, + "step": 5221 + }, + { + "epoch": 0.2748421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00016114619883040936, + "loss": 3.7237, + "step": 5222 + }, + { + "epoch": 0.27489473684210525, + "grad_norm": 0.5, + "learning_rate": 0.00016113450292397664, + "loss": 3.7284, + "step": 5223 + }, + { + "epoch": 0.2749473684210526, + "grad_norm": 0.56640625, + "learning_rate": 0.00016112280701754386, + "loss": 3.9019, + "step": 5224 + }, + { + "epoch": 0.275, + "grad_norm": 0.5390625, + "learning_rate": 0.0001611111111111111, + "loss": 3.8815, + "step": 5225 + }, + { + "epoch": 0.2750526315789474, + "grad_norm": 0.6015625, + "learning_rate": 0.00016109941520467836, + "loss": 3.9803, + "step": 5226 + }, + { + "epoch": 0.27510526315789474, + "grad_norm": 0.59765625, + "learning_rate": 0.0001610877192982456, + "loss": 3.5913, + "step": 5227 + }, + { + "epoch": 0.2751578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.00016107602339181286, + "loss": 4.0793, + "step": 5228 + }, + { + "epoch": 0.27521052631578946, + "grad_norm": 0.5390625, + "learning_rate": 0.0001610643274853801, + "loss": 3.3145, + "step": 5229 + }, + { + "epoch": 0.2752631578947368, + "grad_norm": 0.52734375, + "learning_rate": 0.00016105263157894736, + "loss": 3.9204, + "step": 5230 + }, + { + "epoch": 0.27531578947368424, + "grad_norm": 0.5703125, + "learning_rate": 0.00016104093567251464, + "loss": 3.9962, + "step": 5231 + }, + { + "epoch": 0.2753684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00016102923976608189, + "loss": 3.446, + "step": 5232 + }, + { + "epoch": 0.27542105263157896, + "grad_norm": 0.5234375, + "learning_rate": 0.00016101754385964914, + "loss": 4.206, + "step": 5233 + }, + { + "epoch": 0.2754736842105263, + "grad_norm": 0.5859375, + "learning_rate": 0.00016100584795321639, + "loss": 3.6825, + "step": 5234 + }, + { + "epoch": 0.2755263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.00016099415204678364, + "loss": 3.7728, + "step": 5235 + }, + { + "epoch": 0.27557894736842103, + "grad_norm": 0.70703125, + "learning_rate": 0.00016098245614035088, + "loss": 3.5161, + "step": 5236 + }, + { + "epoch": 0.2756315789473684, + "grad_norm": 0.55859375, + "learning_rate": 0.00016097076023391813, + "loss": 3.439, + "step": 5237 + }, + { + "epoch": 0.2756842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.0001609590643274854, + "loss": 3.5245, + "step": 5238 + }, + { + "epoch": 0.27573684210526317, + "grad_norm": 0.51953125, + "learning_rate": 0.00016094736842105263, + "loss": 3.9627, + "step": 5239 + }, + { + "epoch": 0.27578947368421053, + "grad_norm": 0.5078125, + "learning_rate": 0.00016093567251461988, + "loss": 3.7726, + "step": 5240 + }, + { + "epoch": 0.2758421052631579, + "grad_norm": 0.546875, + "learning_rate": 0.00016092397660818713, + "loss": 3.2147, + "step": 5241 + }, + { + "epoch": 0.27589473684210525, + "grad_norm": 0.53125, + "learning_rate": 0.00016091228070175438, + "loss": 3.7979, + "step": 5242 + }, + { + "epoch": 0.2759473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00016090058479532163, + "loss": 3.805, + "step": 5243 + }, + { + "epoch": 0.276, + "grad_norm": 0.53515625, + "learning_rate": 0.00016088888888888888, + "loss": 3.7618, + "step": 5244 + }, + { + "epoch": 0.2760526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.00016087719298245613, + "loss": 3.3282, + "step": 5245 + }, + { + "epoch": 0.27610526315789474, + "grad_norm": 0.74609375, + "learning_rate": 0.0001608654970760234, + "loss": 4.0682, + "step": 5246 + }, + { + "epoch": 0.2761578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.00016085380116959066, + "loss": 3.8375, + "step": 5247 + }, + { + "epoch": 0.27621052631578946, + "grad_norm": 0.57421875, + "learning_rate": 0.0001608421052631579, + "loss": 3.7117, + "step": 5248 + }, + { + "epoch": 0.2762631578947368, + "grad_norm": 0.53125, + "learning_rate": 0.00016083040935672516, + "loss": 4.0479, + "step": 5249 + }, + { + "epoch": 0.27631578947368424, + "grad_norm": 0.53515625, + "learning_rate": 0.0001608187134502924, + "loss": 3.5296, + "step": 5250 + }, + { + "epoch": 0.2763684210526316, + "grad_norm": 0.59375, + "learning_rate": 0.00016080701754385966, + "loss": 3.6227, + "step": 5251 + }, + { + "epoch": 0.27642105263157896, + "grad_norm": 0.55078125, + "learning_rate": 0.0001607953216374269, + "loss": 3.675, + "step": 5252 + }, + { + "epoch": 0.2764736842105263, + "grad_norm": 0.5625, + "learning_rate": 0.00016078362573099418, + "loss": 3.8557, + "step": 5253 + }, + { + "epoch": 0.2765263157894737, + "grad_norm": 0.5234375, + "learning_rate": 0.00016077192982456143, + "loss": 3.5912, + "step": 5254 + }, + { + "epoch": 0.27657894736842104, + "grad_norm": 0.51953125, + "learning_rate": 0.00016076023391812865, + "loss": 3.2501, + "step": 5255 + }, + { + "epoch": 0.2766315789473684, + "grad_norm": 0.47265625, + "learning_rate": 0.0001607485380116959, + "loss": 3.4651, + "step": 5256 + }, + { + "epoch": 0.2766842105263158, + "grad_norm": 0.423828125, + "learning_rate": 0.00016073684210526315, + "loss": 3.7871, + "step": 5257 + }, + { + "epoch": 0.27673684210526317, + "grad_norm": 0.53125, + "learning_rate": 0.0001607251461988304, + "loss": 3.4213, + "step": 5258 + }, + { + "epoch": 0.27678947368421053, + "grad_norm": 0.51953125, + "learning_rate": 0.00016071345029239765, + "loss": 3.575, + "step": 5259 + }, + { + "epoch": 0.2768421052631579, + "grad_norm": 0.5703125, + "learning_rate": 0.00016070175438596493, + "loss": 3.8136, + "step": 5260 + }, + { + "epoch": 0.27689473684210525, + "grad_norm": 0.53125, + "learning_rate": 0.00016069005847953218, + "loss": 4.0287, + "step": 5261 + }, + { + "epoch": 0.2769473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00016067836257309943, + "loss": 3.1957, + "step": 5262 + }, + { + "epoch": 0.277, + "grad_norm": 0.5546875, + "learning_rate": 0.00016066666666666668, + "loss": 3.8117, + "step": 5263 + }, + { + "epoch": 0.2770526315789474, + "grad_norm": 0.54296875, + "learning_rate": 0.00016065497076023393, + "loss": 3.7238, + "step": 5264 + }, + { + "epoch": 0.27710526315789474, + "grad_norm": 0.57421875, + "learning_rate": 0.00016064327485380118, + "loss": 4.0433, + "step": 5265 + }, + { + "epoch": 0.2771578947368421, + "grad_norm": 0.60546875, + "learning_rate": 0.00016063157894736843, + "loss": 3.6591, + "step": 5266 + }, + { + "epoch": 0.27721052631578946, + "grad_norm": 0.6328125, + "learning_rate": 0.00016061988304093568, + "loss": 3.8543, + "step": 5267 + }, + { + "epoch": 0.2772631578947368, + "grad_norm": 0.64453125, + "learning_rate": 0.00016060818713450295, + "loss": 4.2623, + "step": 5268 + }, + { + "epoch": 0.27731578947368424, + "grad_norm": 0.5390625, + "learning_rate": 0.0001605964912280702, + "loss": 3.7761, + "step": 5269 + }, + { + "epoch": 0.2773684210526316, + "grad_norm": 0.5859375, + "learning_rate": 0.00016058479532163743, + "loss": 3.9052, + "step": 5270 + }, + { + "epoch": 0.27742105263157896, + "grad_norm": 0.484375, + "learning_rate": 0.00016057309941520467, + "loss": 3.7967, + "step": 5271 + }, + { + "epoch": 0.2774736842105263, + "grad_norm": 0.51953125, + "learning_rate": 0.00016056140350877192, + "loss": 3.7677, + "step": 5272 + }, + { + "epoch": 0.2775263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.00016054970760233917, + "loss": 3.8145, + "step": 5273 + }, + { + "epoch": 0.27757894736842104, + "grad_norm": 0.5390625, + "learning_rate": 0.00016053801169590642, + "loss": 3.9448, + "step": 5274 + }, + { + "epoch": 0.2776315789473684, + "grad_norm": 0.5234375, + "learning_rate": 0.0001605263157894737, + "loss": 3.925, + "step": 5275 + }, + { + "epoch": 0.2776842105263158, + "grad_norm": 0.53515625, + "learning_rate": 0.00016051461988304095, + "loss": 3.7137, + "step": 5276 + }, + { + "epoch": 0.27773684210526317, + "grad_norm": 0.52734375, + "learning_rate": 0.0001605029239766082, + "loss": 3.9911, + "step": 5277 + }, + { + "epoch": 0.27778947368421053, + "grad_norm": 0.5703125, + "learning_rate": 0.00016049122807017545, + "loss": 4.2712, + "step": 5278 + }, + { + "epoch": 0.2778421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.0001604795321637427, + "loss": 3.9491, + "step": 5279 + }, + { + "epoch": 0.27789473684210525, + "grad_norm": 0.5546875, + "learning_rate": 0.00016046783625730995, + "loss": 3.5305, + "step": 5280 + }, + { + "epoch": 0.2779473684210526, + "grad_norm": 0.546875, + "learning_rate": 0.0001604561403508772, + "loss": 3.4317, + "step": 5281 + }, + { + "epoch": 0.278, + "grad_norm": 0.515625, + "learning_rate": 0.00016044444444444445, + "loss": 3.9375, + "step": 5282 + }, + { + "epoch": 0.2780526315789474, + "grad_norm": 0.55078125, + "learning_rate": 0.00016043274853801172, + "loss": 3.6685, + "step": 5283 + }, + { + "epoch": 0.27810526315789474, + "grad_norm": 0.57421875, + "learning_rate": 0.00016042105263157897, + "loss": 3.7462, + "step": 5284 + }, + { + "epoch": 0.2781578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.0001604093567251462, + "loss": 4.1341, + "step": 5285 + }, + { + "epoch": 0.27821052631578946, + "grad_norm": 0.59375, + "learning_rate": 0.00016039766081871345, + "loss": 3.7047, + "step": 5286 + }, + { + "epoch": 0.2782631578947368, + "grad_norm": 0.59375, + "learning_rate": 0.0001603859649122807, + "loss": 3.9148, + "step": 5287 + }, + { + "epoch": 0.2783157894736842, + "grad_norm": 0.6328125, + "learning_rate": 0.00016037426900584794, + "loss": 3.5592, + "step": 5288 + }, + { + "epoch": 0.2783684210526316, + "grad_norm": 0.4609375, + "learning_rate": 0.0001603625730994152, + "loss": 3.585, + "step": 5289 + }, + { + "epoch": 0.27842105263157896, + "grad_norm": 0.51171875, + "learning_rate": 0.00016035087719298247, + "loss": 3.7178, + "step": 5290 + }, + { + "epoch": 0.2784736842105263, + "grad_norm": 0.578125, + "learning_rate": 0.00016033918128654972, + "loss": 3.5498, + "step": 5291 + }, + { + "epoch": 0.2785263157894737, + "grad_norm": 0.5390625, + "learning_rate": 0.00016032748538011697, + "loss": 3.5156, + "step": 5292 + }, + { + "epoch": 0.27857894736842104, + "grad_norm": 0.490234375, + "learning_rate": 0.00016031578947368422, + "loss": 4.0251, + "step": 5293 + }, + { + "epoch": 0.2786315789473684, + "grad_norm": 0.49609375, + "learning_rate": 0.00016030409356725147, + "loss": 4.0008, + "step": 5294 + }, + { + "epoch": 0.2786842105263158, + "grad_norm": 0.59375, + "learning_rate": 0.00016029239766081872, + "loss": 4.2093, + "step": 5295 + }, + { + "epoch": 0.27873684210526317, + "grad_norm": 0.58984375, + "learning_rate": 0.00016028070175438597, + "loss": 3.8589, + "step": 5296 + }, + { + "epoch": 0.27878947368421053, + "grad_norm": 0.58984375, + "learning_rate": 0.00016026900584795322, + "loss": 3.6711, + "step": 5297 + }, + { + "epoch": 0.2788421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.0001602573099415205, + "loss": 3.8336, + "step": 5298 + }, + { + "epoch": 0.27889473684210525, + "grad_norm": 0.60546875, + "learning_rate": 0.00016024561403508774, + "loss": 3.9859, + "step": 5299 + }, + { + "epoch": 0.2789473684210526, + "grad_norm": 0.515625, + "learning_rate": 0.000160233918128655, + "loss": 3.5502, + "step": 5300 + }, + { + "epoch": 0.279, + "grad_norm": 0.494140625, + "learning_rate": 0.00016022222222222222, + "loss": 3.807, + "step": 5301 + }, + { + "epoch": 0.2790526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.00016021052631578947, + "loss": 3.3672, + "step": 5302 + }, + { + "epoch": 0.27910526315789475, + "grad_norm": 0.474609375, + "learning_rate": 0.00016019883040935672, + "loss": 3.6041, + "step": 5303 + }, + { + "epoch": 0.2791578947368421, + "grad_norm": 0.546875, + "learning_rate": 0.00016018713450292397, + "loss": 3.85, + "step": 5304 + }, + { + "epoch": 0.27921052631578946, + "grad_norm": 0.5546875, + "learning_rate": 0.00016017543859649124, + "loss": 3.9142, + "step": 5305 + }, + { + "epoch": 0.2792631578947368, + "grad_norm": 0.546875, + "learning_rate": 0.0001601637426900585, + "loss": 3.6445, + "step": 5306 + }, + { + "epoch": 0.2793157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00016015204678362574, + "loss": 3.9696, + "step": 5307 + }, + { + "epoch": 0.2793684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.000160140350877193, + "loss": 3.979, + "step": 5308 + }, + { + "epoch": 0.27942105263157896, + "grad_norm": 0.4921875, + "learning_rate": 0.00016012865497076024, + "loss": 3.7261, + "step": 5309 + }, + { + "epoch": 0.2794736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.0001601169590643275, + "loss": 3.835, + "step": 5310 + }, + { + "epoch": 0.2795263157894737, + "grad_norm": 0.68359375, + "learning_rate": 0.00016010526315789474, + "loss": 3.1159, + "step": 5311 + }, + { + "epoch": 0.27957894736842104, + "grad_norm": 0.59375, + "learning_rate": 0.00016009356725146202, + "loss": 3.6676, + "step": 5312 + }, + { + "epoch": 0.2796315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.00016008187134502927, + "loss": 3.6006, + "step": 5313 + }, + { + "epoch": 0.2796842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.00016007017543859652, + "loss": 3.7635, + "step": 5314 + }, + { + "epoch": 0.2797368421052632, + "grad_norm": 0.54296875, + "learning_rate": 0.00016005847953216377, + "loss": 3.8946, + "step": 5315 + }, + { + "epoch": 0.27978947368421053, + "grad_norm": 0.52734375, + "learning_rate": 0.000160046783625731, + "loss": 3.7092, + "step": 5316 + }, + { + "epoch": 0.2798421052631579, + "grad_norm": 0.494140625, + "learning_rate": 0.00016003508771929824, + "loss": 3.9041, + "step": 5317 + }, + { + "epoch": 0.27989473684210525, + "grad_norm": 0.453125, + "learning_rate": 0.0001600233918128655, + "loss": 3.6958, + "step": 5318 + }, + { + "epoch": 0.2799473684210526, + "grad_norm": 0.482421875, + "learning_rate": 0.00016001169590643274, + "loss": 4.2814, + "step": 5319 + }, + { + "epoch": 0.28, + "grad_norm": 0.51953125, + "learning_rate": 0.00016, + "loss": 4.0783, + "step": 5320 + }, + { + "epoch": 0.2800526315789474, + "grad_norm": 0.56640625, + "learning_rate": 0.00015998830409356726, + "loss": 4.1573, + "step": 5321 + }, + { + "epoch": 0.28010526315789475, + "grad_norm": 0.5859375, + "learning_rate": 0.0001599766081871345, + "loss": 4.0254, + "step": 5322 + }, + { + "epoch": 0.2801578947368421, + "grad_norm": 0.546875, + "learning_rate": 0.00015996491228070176, + "loss": 3.937, + "step": 5323 + }, + { + "epoch": 0.28021052631578947, + "grad_norm": 0.5859375, + "learning_rate": 0.000159953216374269, + "loss": 3.8881, + "step": 5324 + }, + { + "epoch": 0.2802631578947368, + "grad_norm": 0.51953125, + "learning_rate": 0.00015994152046783626, + "loss": 3.752, + "step": 5325 + }, + { + "epoch": 0.2803157894736842, + "grad_norm": 0.49609375, + "learning_rate": 0.0001599298245614035, + "loss": 3.6328, + "step": 5326 + }, + { + "epoch": 0.2803684210526316, + "grad_norm": 0.470703125, + "learning_rate": 0.0001599181286549708, + "loss": 3.8876, + "step": 5327 + }, + { + "epoch": 0.28042105263157896, + "grad_norm": 0.55859375, + "learning_rate": 0.00015990643274853804, + "loss": 3.905, + "step": 5328 + }, + { + "epoch": 0.2804736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.0001598947368421053, + "loss": 4.1513, + "step": 5329 + }, + { + "epoch": 0.2805263157894737, + "grad_norm": 0.546875, + "learning_rate": 0.00015988304093567254, + "loss": 3.8093, + "step": 5330 + }, + { + "epoch": 0.28057894736842104, + "grad_norm": 0.49609375, + "learning_rate": 0.00015987134502923976, + "loss": 4.1714, + "step": 5331 + }, + { + "epoch": 0.2806315789473684, + "grad_norm": 1.1796875, + "learning_rate": 0.000159859649122807, + "loss": 3.8573, + "step": 5332 + }, + { + "epoch": 0.2806842105263158, + "grad_norm": 0.447265625, + "learning_rate": 0.00015984795321637426, + "loss": 3.8828, + "step": 5333 + }, + { + "epoch": 0.2807368421052632, + "grad_norm": 0.47265625, + "learning_rate": 0.0001598362573099415, + "loss": 3.801, + "step": 5334 + }, + { + "epoch": 0.28078947368421053, + "grad_norm": 0.5625, + "learning_rate": 0.00015982456140350878, + "loss": 3.7336, + "step": 5335 + }, + { + "epoch": 0.2808421052631579, + "grad_norm": 0.5546875, + "learning_rate": 0.00015981286549707603, + "loss": 4.1166, + "step": 5336 + }, + { + "epoch": 0.28089473684210525, + "grad_norm": 0.62109375, + "learning_rate": 0.00015980116959064328, + "loss": 3.4961, + "step": 5337 + }, + { + "epoch": 0.2809473684210526, + "grad_norm": 0.875, + "learning_rate": 0.00015978947368421053, + "loss": 3.5903, + "step": 5338 + }, + { + "epoch": 0.281, + "grad_norm": 0.5234375, + "learning_rate": 0.00015977777777777778, + "loss": 3.6329, + "step": 5339 + }, + { + "epoch": 0.2810526315789474, + "grad_norm": 0.58203125, + "learning_rate": 0.00015976608187134503, + "loss": 3.9902, + "step": 5340 + }, + { + "epoch": 0.28110526315789475, + "grad_norm": 0.6015625, + "learning_rate": 0.00015975438596491228, + "loss": 3.7605, + "step": 5341 + }, + { + "epoch": 0.2811578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.00015974269005847956, + "loss": 3.7782, + "step": 5342 + }, + { + "epoch": 0.28121052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.0001597309941520468, + "loss": 3.6838, + "step": 5343 + }, + { + "epoch": 0.2812631578947368, + "grad_norm": 0.52734375, + "learning_rate": 0.00015971929824561406, + "loss": 4.2612, + "step": 5344 + }, + { + "epoch": 0.2813157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.0001597076023391813, + "loss": 3.2752, + "step": 5345 + }, + { + "epoch": 0.2813684210526316, + "grad_norm": 0.546875, + "learning_rate": 0.00015969590643274856, + "loss": 4.1212, + "step": 5346 + }, + { + "epoch": 0.28142105263157896, + "grad_norm": 0.6171875, + "learning_rate": 0.00015968421052631578, + "loss": 4.1682, + "step": 5347 + }, + { + "epoch": 0.2814736842105263, + "grad_norm": 0.62890625, + "learning_rate": 0.00015967251461988303, + "loss": 3.9274, + "step": 5348 + }, + { + "epoch": 0.2815263157894737, + "grad_norm": 0.59375, + "learning_rate": 0.0001596608187134503, + "loss": 3.729, + "step": 5349 + }, + { + "epoch": 0.28157894736842104, + "grad_norm": 0.57421875, + "learning_rate": 0.00015964912280701756, + "loss": 4.0891, + "step": 5350 + }, + { + "epoch": 0.2816315789473684, + "grad_norm": 0.609375, + "learning_rate": 0.0001596374269005848, + "loss": 3.7785, + "step": 5351 + }, + { + "epoch": 0.2816842105263158, + "grad_norm": 0.57421875, + "learning_rate": 0.00015962573099415205, + "loss": 3.9611, + "step": 5352 + }, + { + "epoch": 0.2817368421052632, + "grad_norm": 0.515625, + "learning_rate": 0.0001596140350877193, + "loss": 3.7765, + "step": 5353 + }, + { + "epoch": 0.28178947368421053, + "grad_norm": 0.6171875, + "learning_rate": 0.00015960233918128655, + "loss": 3.648, + "step": 5354 + }, + { + "epoch": 0.2818421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.0001595906432748538, + "loss": 3.9332, + "step": 5355 + }, + { + "epoch": 0.28189473684210525, + "grad_norm": 0.486328125, + "learning_rate": 0.00015957894736842105, + "loss": 3.6293, + "step": 5356 + }, + { + "epoch": 0.2819473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00015956725146198833, + "loss": 3.9916, + "step": 5357 + }, + { + "epoch": 0.282, + "grad_norm": 0.5078125, + "learning_rate": 0.00015955555555555558, + "loss": 4.0825, + "step": 5358 + }, + { + "epoch": 0.2820526315789474, + "grad_norm": 0.53125, + "learning_rate": 0.00015954385964912283, + "loss": 3.937, + "step": 5359 + }, + { + "epoch": 0.28210526315789475, + "grad_norm": 0.55078125, + "learning_rate": 0.00015953216374269008, + "loss": 3.3229, + "step": 5360 + }, + { + "epoch": 0.2821578947368421, + "grad_norm": 0.5546875, + "learning_rate": 0.00015952046783625733, + "loss": 3.796, + "step": 5361 + }, + { + "epoch": 0.28221052631578947, + "grad_norm": 0.5, + "learning_rate": 0.00015950877192982455, + "loss": 3.7731, + "step": 5362 + }, + { + "epoch": 0.2822631578947368, + "grad_norm": 0.5234375, + "learning_rate": 0.0001594970760233918, + "loss": 3.4611, + "step": 5363 + }, + { + "epoch": 0.2823157894736842, + "grad_norm": 0.4765625, + "learning_rate": 0.00015948538011695908, + "loss": 4.0299, + "step": 5364 + }, + { + "epoch": 0.2823684210526316, + "grad_norm": 0.55859375, + "learning_rate": 0.00015947368421052633, + "loss": 3.7271, + "step": 5365 + }, + { + "epoch": 0.28242105263157896, + "grad_norm": 0.54296875, + "learning_rate": 0.00015946198830409358, + "loss": 3.8124, + "step": 5366 + }, + { + "epoch": 0.2824736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00015945029239766083, + "loss": 3.3478, + "step": 5367 + }, + { + "epoch": 0.2825263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.00015943859649122807, + "loss": 3.8908, + "step": 5368 + }, + { + "epoch": 0.28257894736842104, + "grad_norm": 0.515625, + "learning_rate": 0.00015942690058479532, + "loss": 3.2967, + "step": 5369 + }, + { + "epoch": 0.2826315789473684, + "grad_norm": 0.447265625, + "learning_rate": 0.00015941520467836257, + "loss": 4.3177, + "step": 5370 + }, + { + "epoch": 0.2826842105263158, + "grad_norm": 0.53515625, + "learning_rate": 0.00015940350877192982, + "loss": 3.6918, + "step": 5371 + }, + { + "epoch": 0.2827368421052632, + "grad_norm": 0.52734375, + "learning_rate": 0.0001593918128654971, + "loss": 4.2019, + "step": 5372 + }, + { + "epoch": 0.28278947368421053, + "grad_norm": 0.52734375, + "learning_rate": 0.00015938011695906435, + "loss": 3.8891, + "step": 5373 + }, + { + "epoch": 0.2828421052631579, + "grad_norm": 0.478515625, + "learning_rate": 0.0001593684210526316, + "loss": 3.9834, + "step": 5374 + }, + { + "epoch": 0.28289473684210525, + "grad_norm": 0.48046875, + "learning_rate": 0.00015935672514619885, + "loss": 4.0632, + "step": 5375 + }, + { + "epoch": 0.2829473684210526, + "grad_norm": 0.47265625, + "learning_rate": 0.0001593450292397661, + "loss": 3.9245, + "step": 5376 + }, + { + "epoch": 0.283, + "grad_norm": 0.51953125, + "learning_rate": 0.00015933333333333332, + "loss": 3.9627, + "step": 5377 + }, + { + "epoch": 0.2830526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.00015932163742690057, + "loss": 3.7839, + "step": 5378 + }, + { + "epoch": 0.28310526315789475, + "grad_norm": 0.5234375, + "learning_rate": 0.00015930994152046785, + "loss": 3.606, + "step": 5379 + }, + { + "epoch": 0.2831578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.0001592982456140351, + "loss": 3.7216, + "step": 5380 + }, + { + "epoch": 0.28321052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00015928654970760235, + "loss": 3.3692, + "step": 5381 + }, + { + "epoch": 0.28326315789473683, + "grad_norm": 0.48046875, + "learning_rate": 0.0001592748538011696, + "loss": 3.9261, + "step": 5382 + }, + { + "epoch": 0.2833157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00015926315789473685, + "loss": 4.2375, + "step": 5383 + }, + { + "epoch": 0.2833684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.0001592514619883041, + "loss": 3.843, + "step": 5384 + }, + { + "epoch": 0.28342105263157896, + "grad_norm": 0.51953125, + "learning_rate": 0.00015923976608187135, + "loss": 3.8312, + "step": 5385 + }, + { + "epoch": 0.2834736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.00015922807017543862, + "loss": 3.8354, + "step": 5386 + }, + { + "epoch": 0.2835263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.00015921637426900587, + "loss": 3.6604, + "step": 5387 + }, + { + "epoch": 0.28357894736842104, + "grad_norm": 0.5859375, + "learning_rate": 0.00015920467836257312, + "loss": 3.9329, + "step": 5388 + }, + { + "epoch": 0.2836315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00015919298245614037, + "loss": 4.0088, + "step": 5389 + }, + { + "epoch": 0.2836842105263158, + "grad_norm": 0.6015625, + "learning_rate": 0.00015918128654970762, + "loss": 3.6329, + "step": 5390 + }, + { + "epoch": 0.2837368421052632, + "grad_norm": 0.59375, + "learning_rate": 0.00015916959064327487, + "loss": 3.2197, + "step": 5391 + }, + { + "epoch": 0.28378947368421054, + "grad_norm": 0.515625, + "learning_rate": 0.00015915789473684212, + "loss": 3.9685, + "step": 5392 + }, + { + "epoch": 0.2838421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00015914619883040934, + "loss": 3.4763, + "step": 5393 + }, + { + "epoch": 0.28389473684210526, + "grad_norm": 0.58203125, + "learning_rate": 0.00015913450292397662, + "loss": 3.4838, + "step": 5394 + }, + { + "epoch": 0.2839473684210526, + "grad_norm": 0.6640625, + "learning_rate": 0.00015912280701754387, + "loss": 3.4375, + "step": 5395 + }, + { + "epoch": 0.284, + "grad_norm": 0.486328125, + "learning_rate": 0.00015911111111111112, + "loss": 3.8634, + "step": 5396 + }, + { + "epoch": 0.2840526315789474, + "grad_norm": 0.625, + "learning_rate": 0.00015909941520467837, + "loss": 3.5644, + "step": 5397 + }, + { + "epoch": 0.28410526315789475, + "grad_norm": 0.48046875, + "learning_rate": 0.00015908771929824562, + "loss": 3.7604, + "step": 5398 + }, + { + "epoch": 0.2841578947368421, + "grad_norm": 0.478515625, + "learning_rate": 0.00015907602339181287, + "loss": 4.1135, + "step": 5399 + }, + { + "epoch": 0.28421052631578947, + "grad_norm": 0.63671875, + "learning_rate": 0.00015906432748538012, + "loss": 3.6654, + "step": 5400 + }, + { + "epoch": 0.28426315789473683, + "grad_norm": 0.51953125, + "learning_rate": 0.0001590526315789474, + "loss": 4.1502, + "step": 5401 + }, + { + "epoch": 0.2843157894736842, + "grad_norm": 0.59375, + "learning_rate": 0.00015904093567251464, + "loss": 4.1408, + "step": 5402 + }, + { + "epoch": 0.2843684210526316, + "grad_norm": 0.59765625, + "learning_rate": 0.0001590292397660819, + "loss": 3.7445, + "step": 5403 + }, + { + "epoch": 0.28442105263157896, + "grad_norm": 0.56640625, + "learning_rate": 0.00015901754385964914, + "loss": 3.5536, + "step": 5404 + }, + { + "epoch": 0.2844736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.0001590058479532164, + "loss": 3.7379, + "step": 5405 + }, + { + "epoch": 0.2845263157894737, + "grad_norm": 0.578125, + "learning_rate": 0.00015899415204678364, + "loss": 3.6144, + "step": 5406 + }, + { + "epoch": 0.28457894736842104, + "grad_norm": 0.55078125, + "learning_rate": 0.0001589824561403509, + "loss": 3.4398, + "step": 5407 + }, + { + "epoch": 0.2846315789473684, + "grad_norm": 0.55859375, + "learning_rate": 0.0001589707602339181, + "loss": 3.7182, + "step": 5408 + }, + { + "epoch": 0.28468421052631576, + "grad_norm": 0.53515625, + "learning_rate": 0.0001589590643274854, + "loss": 3.7088, + "step": 5409 + }, + { + "epoch": 0.2847368421052632, + "grad_norm": 0.51953125, + "learning_rate": 0.00015894736842105264, + "loss": 4.0811, + "step": 5410 + }, + { + "epoch": 0.28478947368421054, + "grad_norm": 0.5, + "learning_rate": 0.0001589356725146199, + "loss": 3.7595, + "step": 5411 + }, + { + "epoch": 0.2848421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00015892397660818714, + "loss": 3.9067, + "step": 5412 + }, + { + "epoch": 0.28489473684210526, + "grad_norm": 0.53125, + "learning_rate": 0.0001589122807017544, + "loss": 3.6073, + "step": 5413 + }, + { + "epoch": 0.2849473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.00015890058479532164, + "loss": 3.348, + "step": 5414 + }, + { + "epoch": 0.285, + "grad_norm": 0.474609375, + "learning_rate": 0.0001588888888888889, + "loss": 3.8089, + "step": 5415 + }, + { + "epoch": 0.2850526315789474, + "grad_norm": 0.578125, + "learning_rate": 0.00015887719298245616, + "loss": 3.6771, + "step": 5416 + }, + { + "epoch": 0.28510526315789475, + "grad_norm": 0.4921875, + "learning_rate": 0.0001588654970760234, + "loss": 3.8272, + "step": 5417 + }, + { + "epoch": 0.2851578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.00015885380116959066, + "loss": 3.9541, + "step": 5418 + }, + { + "epoch": 0.28521052631578947, + "grad_norm": 0.55078125, + "learning_rate": 0.0001588421052631579, + "loss": 3.8316, + "step": 5419 + }, + { + "epoch": 0.28526315789473683, + "grad_norm": 0.53125, + "learning_rate": 0.00015883040935672516, + "loss": 4.2009, + "step": 5420 + }, + { + "epoch": 0.2853157894736842, + "grad_norm": 0.57421875, + "learning_rate": 0.0001588187134502924, + "loss": 3.7667, + "step": 5421 + }, + { + "epoch": 0.2853684210526316, + "grad_norm": 0.5625, + "learning_rate": 0.00015880701754385966, + "loss": 3.8774, + "step": 5422 + }, + { + "epoch": 0.28542105263157896, + "grad_norm": 0.61328125, + "learning_rate": 0.0001587953216374269, + "loss": 3.8746, + "step": 5423 + }, + { + "epoch": 0.2854736842105263, + "grad_norm": 0.56640625, + "learning_rate": 0.00015878362573099416, + "loss": 4.0849, + "step": 5424 + }, + { + "epoch": 0.2855263157894737, + "grad_norm": 0.5390625, + "learning_rate": 0.0001587719298245614, + "loss": 3.9918, + "step": 5425 + }, + { + "epoch": 0.28557894736842104, + "grad_norm": 0.51953125, + "learning_rate": 0.00015876023391812866, + "loss": 3.5504, + "step": 5426 + }, + { + "epoch": 0.2856315789473684, + "grad_norm": 0.494140625, + "learning_rate": 0.0001587485380116959, + "loss": 3.9003, + "step": 5427 + }, + { + "epoch": 0.28568421052631576, + "grad_norm": 0.609375, + "learning_rate": 0.00015873684210526316, + "loss": 3.9648, + "step": 5428 + }, + { + "epoch": 0.2857368421052632, + "grad_norm": 0.64453125, + "learning_rate": 0.0001587251461988304, + "loss": 4.106, + "step": 5429 + }, + { + "epoch": 0.28578947368421054, + "grad_norm": 0.5546875, + "learning_rate": 0.00015871345029239766, + "loss": 3.8584, + "step": 5430 + }, + { + "epoch": 0.2858421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00015870175438596493, + "loss": 3.8083, + "step": 5431 + }, + { + "epoch": 0.28589473684210526, + "grad_norm": 0.482421875, + "learning_rate": 0.00015869005847953218, + "loss": 3.9025, + "step": 5432 + }, + { + "epoch": 0.2859473684210526, + "grad_norm": 0.46484375, + "learning_rate": 0.00015867836257309943, + "loss": 3.8085, + "step": 5433 + }, + { + "epoch": 0.286, + "grad_norm": 0.51171875, + "learning_rate": 0.00015866666666666668, + "loss": 3.8456, + "step": 5434 + }, + { + "epoch": 0.2860526315789474, + "grad_norm": 0.609375, + "learning_rate": 0.00015865497076023393, + "loss": 3.9294, + "step": 5435 + }, + { + "epoch": 0.28610526315789475, + "grad_norm": 0.53125, + "learning_rate": 0.00015864327485380118, + "loss": 3.6726, + "step": 5436 + }, + { + "epoch": 0.2861578947368421, + "grad_norm": 0.58984375, + "learning_rate": 0.00015863157894736843, + "loss": 3.7502, + "step": 5437 + }, + { + "epoch": 0.28621052631578947, + "grad_norm": 0.546875, + "learning_rate": 0.00015861988304093568, + "loss": 3.7539, + "step": 5438 + }, + { + "epoch": 0.28626315789473683, + "grad_norm": 0.51171875, + "learning_rate": 0.00015860818713450293, + "loss": 3.9751, + "step": 5439 + }, + { + "epoch": 0.2863157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00015859649122807018, + "loss": 3.5646, + "step": 5440 + }, + { + "epoch": 0.2863684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00015858479532163743, + "loss": 3.3254, + "step": 5441 + }, + { + "epoch": 0.28642105263157897, + "grad_norm": 0.58984375, + "learning_rate": 0.00015857309941520468, + "loss": 3.8784, + "step": 5442 + }, + { + "epoch": 0.2864736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.00015856140350877193, + "loss": 3.6579, + "step": 5443 + }, + { + "epoch": 0.2865263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.00015854970760233918, + "loss": 3.5916, + "step": 5444 + }, + { + "epoch": 0.28657894736842104, + "grad_norm": 0.53125, + "learning_rate": 0.00015853801169590643, + "loss": 3.4891, + "step": 5445 + }, + { + "epoch": 0.2866315789473684, + "grad_norm": 0.625, + "learning_rate": 0.0001585263157894737, + "loss": 3.8334, + "step": 5446 + }, + { + "epoch": 0.28668421052631576, + "grad_norm": 0.51953125, + "learning_rate": 0.00015851461988304096, + "loss": 3.8883, + "step": 5447 + }, + { + "epoch": 0.2867368421052632, + "grad_norm": 0.52734375, + "learning_rate": 0.0001585029239766082, + "loss": 3.481, + "step": 5448 + }, + { + "epoch": 0.28678947368421054, + "grad_norm": 0.51171875, + "learning_rate": 0.00015849122807017545, + "loss": 3.9705, + "step": 5449 + }, + { + "epoch": 0.2868421052631579, + "grad_norm": 0.62109375, + "learning_rate": 0.0001584795321637427, + "loss": 3.5547, + "step": 5450 + }, + { + "epoch": 0.28689473684210526, + "grad_norm": 0.53125, + "learning_rate": 0.00015846783625730995, + "loss": 3.4139, + "step": 5451 + }, + { + "epoch": 0.2869473684210526, + "grad_norm": 0.55078125, + "learning_rate": 0.0001584561403508772, + "loss": 3.6254, + "step": 5452 + }, + { + "epoch": 0.287, + "grad_norm": 0.5859375, + "learning_rate": 0.00015844444444444445, + "loss": 4.0518, + "step": 5453 + }, + { + "epoch": 0.2870526315789474, + "grad_norm": 0.6328125, + "learning_rate": 0.0001584327485380117, + "loss": 3.9871, + "step": 5454 + }, + { + "epoch": 0.28710526315789475, + "grad_norm": 0.58984375, + "learning_rate": 0.00015842105263157895, + "loss": 4.0106, + "step": 5455 + }, + { + "epoch": 0.2871578947368421, + "grad_norm": 1.375, + "learning_rate": 0.0001584093567251462, + "loss": 3.9477, + "step": 5456 + }, + { + "epoch": 0.28721052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.00015839766081871345, + "loss": 3.7735, + "step": 5457 + }, + { + "epoch": 0.28726315789473683, + "grad_norm": 0.515625, + "learning_rate": 0.0001583859649122807, + "loss": 3.8291, + "step": 5458 + }, + { + "epoch": 0.2873157894736842, + "grad_norm": 0.5625, + "learning_rate": 0.00015837426900584795, + "loss": 3.2443, + "step": 5459 + }, + { + "epoch": 0.2873684210526316, + "grad_norm": 0.4921875, + "learning_rate": 0.0001583625730994152, + "loss": 3.4697, + "step": 5460 + }, + { + "epoch": 0.28742105263157897, + "grad_norm": 0.57421875, + "learning_rate": 0.00015835087719298248, + "loss": 3.3228, + "step": 5461 + }, + { + "epoch": 0.2874736842105263, + "grad_norm": 0.48828125, + "learning_rate": 0.00015833918128654973, + "loss": 3.7613, + "step": 5462 + }, + { + "epoch": 0.2875263157894737, + "grad_norm": 0.482421875, + "learning_rate": 0.00015832748538011698, + "loss": 3.8363, + "step": 5463 + }, + { + "epoch": 0.28757894736842105, + "grad_norm": 0.5625, + "learning_rate": 0.00015831578947368423, + "loss": 4.2735, + "step": 5464 + }, + { + "epoch": 0.2876315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.00015830409356725148, + "loss": 3.5429, + "step": 5465 + }, + { + "epoch": 0.28768421052631576, + "grad_norm": 0.5859375, + "learning_rate": 0.00015829239766081872, + "loss": 3.7164, + "step": 5466 + }, + { + "epoch": 0.2877368421052632, + "grad_norm": 0.5078125, + "learning_rate": 0.00015828070175438597, + "loss": 4.2311, + "step": 5467 + }, + { + "epoch": 0.28778947368421054, + "grad_norm": 0.5859375, + "learning_rate": 0.00015826900584795322, + "loss": 3.9108, + "step": 5468 + }, + { + "epoch": 0.2878421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00015825730994152047, + "loss": 3.732, + "step": 5469 + }, + { + "epoch": 0.28789473684210526, + "grad_norm": 0.498046875, + "learning_rate": 0.00015824561403508772, + "loss": 3.5382, + "step": 5470 + }, + { + "epoch": 0.2879473684210526, + "grad_norm": 0.5390625, + "learning_rate": 0.00015823391812865497, + "loss": 3.9765, + "step": 5471 + }, + { + "epoch": 0.288, + "grad_norm": 0.5078125, + "learning_rate": 0.00015822222222222222, + "loss": 3.8043, + "step": 5472 + }, + { + "epoch": 0.2880526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.00015821052631578947, + "loss": 3.358, + "step": 5473 + }, + { + "epoch": 0.28810526315789475, + "grad_norm": 0.5078125, + "learning_rate": 0.00015819883040935672, + "loss": 3.6795, + "step": 5474 + }, + { + "epoch": 0.2881578947368421, + "grad_norm": 0.494140625, + "learning_rate": 0.000158187134502924, + "loss": 3.662, + "step": 5475 + }, + { + "epoch": 0.2882105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.00015817543859649125, + "loss": 3.6815, + "step": 5476 + }, + { + "epoch": 0.28826315789473683, + "grad_norm": 0.57421875, + "learning_rate": 0.0001581637426900585, + "loss": 3.5161, + "step": 5477 + }, + { + "epoch": 0.2883157894736842, + "grad_norm": 0.59375, + "learning_rate": 0.00015815204678362575, + "loss": 4.0031, + "step": 5478 + }, + { + "epoch": 0.28836842105263155, + "grad_norm": 0.59765625, + "learning_rate": 0.000158140350877193, + "loss": 3.9141, + "step": 5479 + }, + { + "epoch": 0.28842105263157897, + "grad_norm": 0.5625, + "learning_rate": 0.00015812865497076025, + "loss": 3.9437, + "step": 5480 + }, + { + "epoch": 0.2884736842105263, + "grad_norm": 0.5234375, + "learning_rate": 0.0001581169590643275, + "loss": 3.7343, + "step": 5481 + }, + { + "epoch": 0.2885263157894737, + "grad_norm": 0.53515625, + "learning_rate": 0.00015810526315789475, + "loss": 3.8497, + "step": 5482 + }, + { + "epoch": 0.28857894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.000158093567251462, + "loss": 3.9098, + "step": 5483 + }, + { + "epoch": 0.2886315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00015808187134502924, + "loss": 4.1713, + "step": 5484 + }, + { + "epoch": 0.28868421052631577, + "grad_norm": 0.49609375, + "learning_rate": 0.0001580701754385965, + "loss": 3.9871, + "step": 5485 + }, + { + "epoch": 0.2887368421052632, + "grad_norm": 0.49609375, + "learning_rate": 0.00015805847953216374, + "loss": 3.9119, + "step": 5486 + }, + { + "epoch": 0.28878947368421054, + "grad_norm": 0.546875, + "learning_rate": 0.000158046783625731, + "loss": 3.5618, + "step": 5487 + }, + { + "epoch": 0.2888421052631579, + "grad_norm": 0.486328125, + "learning_rate": 0.00015803508771929824, + "loss": 4.1245, + "step": 5488 + }, + { + "epoch": 0.28889473684210526, + "grad_norm": 0.53125, + "learning_rate": 0.0001580233918128655, + "loss": 3.5354, + "step": 5489 + }, + { + "epoch": 0.2889473684210526, + "grad_norm": 0.51953125, + "learning_rate": 0.00015801169590643277, + "loss": 3.8325, + "step": 5490 + }, + { + "epoch": 0.289, + "grad_norm": 0.5078125, + "learning_rate": 0.00015800000000000002, + "loss": 3.8257, + "step": 5491 + }, + { + "epoch": 0.2890526315789474, + "grad_norm": 0.58203125, + "learning_rate": 0.00015798830409356727, + "loss": 3.3822, + "step": 5492 + }, + { + "epoch": 0.28910526315789475, + "grad_norm": 0.482421875, + "learning_rate": 0.00015797660818713452, + "loss": 3.8384, + "step": 5493 + }, + { + "epoch": 0.2891578947368421, + "grad_norm": 0.5234375, + "learning_rate": 0.00015796491228070177, + "loss": 3.7891, + "step": 5494 + }, + { + "epoch": 0.2892105263157895, + "grad_norm": 0.55859375, + "learning_rate": 0.00015795321637426902, + "loss": 3.9851, + "step": 5495 + }, + { + "epoch": 0.28926315789473683, + "grad_norm": 0.61328125, + "learning_rate": 0.00015794152046783627, + "loss": 4.1526, + "step": 5496 + }, + { + "epoch": 0.2893157894736842, + "grad_norm": 0.59375, + "learning_rate": 0.00015792982456140352, + "loss": 3.8106, + "step": 5497 + }, + { + "epoch": 0.28936842105263155, + "grad_norm": 0.60546875, + "learning_rate": 0.00015791812865497077, + "loss": 4.0111, + "step": 5498 + }, + { + "epoch": 0.28942105263157897, + "grad_norm": 0.64453125, + "learning_rate": 0.00015790643274853802, + "loss": 3.6863, + "step": 5499 + }, + { + "epoch": 0.2894736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00015789473684210527, + "loss": 4.089, + "step": 5500 + }, + { + "epoch": 0.2895263157894737, + "grad_norm": 0.609375, + "learning_rate": 0.00015788304093567251, + "loss": 3.8523, + "step": 5501 + }, + { + "epoch": 0.28957894736842105, + "grad_norm": 0.498046875, + "learning_rate": 0.00015787134502923976, + "loss": 3.6763, + "step": 5502 + }, + { + "epoch": 0.2896315789473684, + "grad_norm": 0.47265625, + "learning_rate": 0.00015785964912280701, + "loss": 4.2837, + "step": 5503 + }, + { + "epoch": 0.28968421052631577, + "grad_norm": 0.5390625, + "learning_rate": 0.00015784795321637426, + "loss": 3.6165, + "step": 5504 + }, + { + "epoch": 0.2897368421052632, + "grad_norm": 0.474609375, + "learning_rate": 0.00015783625730994154, + "loss": 3.6283, + "step": 5505 + }, + { + "epoch": 0.28978947368421054, + "grad_norm": 0.4453125, + "learning_rate": 0.0001578245614035088, + "loss": 3.6719, + "step": 5506 + }, + { + "epoch": 0.2898421052631579, + "grad_norm": 0.4765625, + "learning_rate": 0.00015781286549707604, + "loss": 3.9273, + "step": 5507 + }, + { + "epoch": 0.28989473684210526, + "grad_norm": 0.51953125, + "learning_rate": 0.0001578011695906433, + "loss": 3.7888, + "step": 5508 + }, + { + "epoch": 0.2899473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.00015778947368421054, + "loss": 3.8579, + "step": 5509 + }, + { + "epoch": 0.29, + "grad_norm": 0.48046875, + "learning_rate": 0.0001577777777777778, + "loss": 3.4059, + "step": 5510 + }, + { + "epoch": 0.2900526315789474, + "grad_norm": 0.5, + "learning_rate": 0.00015776608187134504, + "loss": 3.0301, + "step": 5511 + }, + { + "epoch": 0.29010526315789475, + "grad_norm": 0.4921875, + "learning_rate": 0.0001577543859649123, + "loss": 3.6457, + "step": 5512 + }, + { + "epoch": 0.2901578947368421, + "grad_norm": 0.59375, + "learning_rate": 0.00015774269005847954, + "loss": 3.4904, + "step": 5513 + }, + { + "epoch": 0.2902105263157895, + "grad_norm": 0.578125, + "learning_rate": 0.00015773099415204679, + "loss": 3.274, + "step": 5514 + }, + { + "epoch": 0.29026315789473683, + "grad_norm": 0.5703125, + "learning_rate": 0.00015771929824561404, + "loss": 3.8212, + "step": 5515 + }, + { + "epoch": 0.2903157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00015770760233918129, + "loss": 3.4638, + "step": 5516 + }, + { + "epoch": 0.29036842105263155, + "grad_norm": 0.83203125, + "learning_rate": 0.00015769590643274854, + "loss": 3.9069, + "step": 5517 + }, + { + "epoch": 0.29042105263157897, + "grad_norm": 0.55859375, + "learning_rate": 0.00015768421052631578, + "loss": 3.8437, + "step": 5518 + }, + { + "epoch": 0.29047368421052633, + "grad_norm": 0.5625, + "learning_rate": 0.00015767251461988303, + "loss": 3.5492, + "step": 5519 + }, + { + "epoch": 0.2905263157894737, + "grad_norm": 0.546875, + "learning_rate": 0.0001576608187134503, + "loss": 3.614, + "step": 5520 + }, + { + "epoch": 0.29057894736842105, + "grad_norm": 0.484375, + "learning_rate": 0.00015764912280701756, + "loss": 3.94, + "step": 5521 + }, + { + "epoch": 0.2906315789473684, + "grad_norm": 0.5234375, + "learning_rate": 0.0001576374269005848, + "loss": 3.6591, + "step": 5522 + }, + { + "epoch": 0.29068421052631577, + "grad_norm": 0.5078125, + "learning_rate": 0.00015762573099415206, + "loss": 4.0728, + "step": 5523 + }, + { + "epoch": 0.2907368421052632, + "grad_norm": 0.49609375, + "learning_rate": 0.0001576140350877193, + "loss": 4.0279, + "step": 5524 + }, + { + "epoch": 0.29078947368421054, + "grad_norm": 0.55078125, + "learning_rate": 0.00015760233918128656, + "loss": 3.4844, + "step": 5525 + }, + { + "epoch": 0.2908421052631579, + "grad_norm": 0.48046875, + "learning_rate": 0.0001575906432748538, + "loss": 3.7706, + "step": 5526 + }, + { + "epoch": 0.29089473684210526, + "grad_norm": 0.494140625, + "learning_rate": 0.00015757894736842106, + "loss": 4.022, + "step": 5527 + }, + { + "epoch": 0.2909473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.0001575672514619883, + "loss": 3.3777, + "step": 5528 + }, + { + "epoch": 0.291, + "grad_norm": 0.4921875, + "learning_rate": 0.00015755555555555556, + "loss": 3.4205, + "step": 5529 + }, + { + "epoch": 0.2910526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.0001575438596491228, + "loss": 3.8522, + "step": 5530 + }, + { + "epoch": 0.29110526315789476, + "grad_norm": 0.478515625, + "learning_rate": 0.00015753216374269006, + "loss": 4.0328, + "step": 5531 + }, + { + "epoch": 0.2911578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.0001575204678362573, + "loss": 3.0746, + "step": 5532 + }, + { + "epoch": 0.2912105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00015750877192982456, + "loss": 3.8774, + "step": 5533 + }, + { + "epoch": 0.29126315789473683, + "grad_norm": 0.578125, + "learning_rate": 0.0001574970760233918, + "loss": 3.4873, + "step": 5534 + }, + { + "epoch": 0.2913157894736842, + "grad_norm": 0.546875, + "learning_rate": 0.00015748538011695908, + "loss": 3.5495, + "step": 5535 + }, + { + "epoch": 0.29136842105263155, + "grad_norm": 0.56640625, + "learning_rate": 0.00015747368421052633, + "loss": 3.4939, + "step": 5536 + }, + { + "epoch": 0.29142105263157897, + "grad_norm": 0.5625, + "learning_rate": 0.00015746198830409358, + "loss": 4.0353, + "step": 5537 + }, + { + "epoch": 0.29147368421052633, + "grad_norm": 0.578125, + "learning_rate": 0.00015745029239766083, + "loss": 3.7795, + "step": 5538 + }, + { + "epoch": 0.2915263157894737, + "grad_norm": 0.5703125, + "learning_rate": 0.00015743859649122808, + "loss": 3.7997, + "step": 5539 + }, + { + "epoch": 0.29157894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.00015742690058479533, + "loss": 3.617, + "step": 5540 + }, + { + "epoch": 0.2916315789473684, + "grad_norm": 0.4921875, + "learning_rate": 0.00015741520467836258, + "loss": 3.3101, + "step": 5541 + }, + { + "epoch": 0.29168421052631577, + "grad_norm": 0.478515625, + "learning_rate": 0.00015740350877192983, + "loss": 4.2609, + "step": 5542 + }, + { + "epoch": 0.2917368421052632, + "grad_norm": 0.51953125, + "learning_rate": 0.00015739181286549708, + "loss": 3.7345, + "step": 5543 + }, + { + "epoch": 0.29178947368421054, + "grad_norm": 0.462890625, + "learning_rate": 0.00015738011695906433, + "loss": 3.5651, + "step": 5544 + }, + { + "epoch": 0.2918421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00015736842105263158, + "loss": 3.9433, + "step": 5545 + }, + { + "epoch": 0.29189473684210526, + "grad_norm": 0.53125, + "learning_rate": 0.00015735672514619883, + "loss": 3.6513, + "step": 5546 + }, + { + "epoch": 0.2919473684210526, + "grad_norm": 0.515625, + "learning_rate": 0.00015734502923976608, + "loss": 3.8456, + "step": 5547 + }, + { + "epoch": 0.292, + "grad_norm": 0.515625, + "learning_rate": 0.00015733333333333333, + "loss": 3.6877, + "step": 5548 + }, + { + "epoch": 0.29205263157894734, + "grad_norm": 0.51171875, + "learning_rate": 0.0001573216374269006, + "loss": 4.0408, + "step": 5549 + }, + { + "epoch": 0.29210526315789476, + "grad_norm": 0.515625, + "learning_rate": 0.00015730994152046785, + "loss": 3.7642, + "step": 5550 + }, + { + "epoch": 0.2921578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.0001572982456140351, + "loss": 3.6588, + "step": 5551 + }, + { + "epoch": 0.2922105263157895, + "grad_norm": 0.50390625, + "learning_rate": 0.00015728654970760235, + "loss": 3.7891, + "step": 5552 + }, + { + "epoch": 0.29226315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.0001572748538011696, + "loss": 3.2326, + "step": 5553 + }, + { + "epoch": 0.2923157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00015726315789473685, + "loss": 3.7145, + "step": 5554 + }, + { + "epoch": 0.29236842105263156, + "grad_norm": 0.5625, + "learning_rate": 0.0001572514619883041, + "loss": 3.5917, + "step": 5555 + }, + { + "epoch": 0.29242105263157897, + "grad_norm": 0.5625, + "learning_rate": 0.00015723976608187135, + "loss": 3.9458, + "step": 5556 + }, + { + "epoch": 0.29247368421052633, + "grad_norm": 0.515625, + "learning_rate": 0.0001572280701754386, + "loss": 3.7681, + "step": 5557 + }, + { + "epoch": 0.2925263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.00015721637426900585, + "loss": 4.1251, + "step": 5558 + }, + { + "epoch": 0.29257894736842105, + "grad_norm": 0.5234375, + "learning_rate": 0.0001572046783625731, + "loss": 3.6632, + "step": 5559 + }, + { + "epoch": 0.2926315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00015719298245614035, + "loss": 3.9552, + "step": 5560 + }, + { + "epoch": 0.29268421052631577, + "grad_norm": 0.4609375, + "learning_rate": 0.0001571812865497076, + "loss": 3.8138, + "step": 5561 + }, + { + "epoch": 0.2927368421052632, + "grad_norm": 0.51953125, + "learning_rate": 0.00015716959064327485, + "loss": 4.2577, + "step": 5562 + }, + { + "epoch": 0.29278947368421054, + "grad_norm": 0.67578125, + "learning_rate": 0.0001571578947368421, + "loss": 3.4935, + "step": 5563 + }, + { + "epoch": 0.2928421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00015714619883040937, + "loss": 3.8223, + "step": 5564 + }, + { + "epoch": 0.29289473684210526, + "grad_norm": 0.546875, + "learning_rate": 0.00015713450292397662, + "loss": 3.5861, + "step": 5565 + }, + { + "epoch": 0.2929473684210526, + "grad_norm": 0.5234375, + "learning_rate": 0.00015712280701754387, + "loss": 3.9151, + "step": 5566 + }, + { + "epoch": 0.293, + "grad_norm": 0.5, + "learning_rate": 0.00015711111111111112, + "loss": 3.8069, + "step": 5567 + }, + { + "epoch": 0.29305263157894734, + "grad_norm": 0.55859375, + "learning_rate": 0.00015709941520467837, + "loss": 3.9293, + "step": 5568 + }, + { + "epoch": 0.29310526315789476, + "grad_norm": 0.51953125, + "learning_rate": 0.00015708771929824562, + "loss": 3.7376, + "step": 5569 + }, + { + "epoch": 0.2931578947368421, + "grad_norm": 0.68359375, + "learning_rate": 0.00015707602339181287, + "loss": 4.1626, + "step": 5570 + }, + { + "epoch": 0.2932105263157895, + "grad_norm": 0.5703125, + "learning_rate": 0.00015706432748538012, + "loss": 3.9006, + "step": 5571 + }, + { + "epoch": 0.29326315789473684, + "grad_norm": 0.6171875, + "learning_rate": 0.00015705263157894737, + "loss": 3.6847, + "step": 5572 + }, + { + "epoch": 0.2933157894736842, + "grad_norm": 0.5546875, + "learning_rate": 0.00015704093567251462, + "loss": 3.5489, + "step": 5573 + }, + { + "epoch": 0.29336842105263156, + "grad_norm": 0.55078125, + "learning_rate": 0.00015702923976608187, + "loss": 4.0255, + "step": 5574 + }, + { + "epoch": 0.29342105263157897, + "grad_norm": 0.515625, + "learning_rate": 0.00015701754385964912, + "loss": 3.5516, + "step": 5575 + }, + { + "epoch": 0.29347368421052633, + "grad_norm": 0.546875, + "learning_rate": 0.00015700584795321637, + "loss": 3.6166, + "step": 5576 + }, + { + "epoch": 0.2935263157894737, + "grad_norm": 0.48046875, + "learning_rate": 0.00015699415204678362, + "loss": 3.5147, + "step": 5577 + }, + { + "epoch": 0.29357894736842105, + "grad_norm": 0.46484375, + "learning_rate": 0.00015698245614035087, + "loss": 3.9199, + "step": 5578 + }, + { + "epoch": 0.2936315789473684, + "grad_norm": 0.498046875, + "learning_rate": 0.00015697076023391815, + "loss": 4.0373, + "step": 5579 + }, + { + "epoch": 0.29368421052631577, + "grad_norm": 0.50390625, + "learning_rate": 0.0001569590643274854, + "loss": 3.6636, + "step": 5580 + }, + { + "epoch": 0.2937368421052632, + "grad_norm": 0.515625, + "learning_rate": 0.00015694736842105264, + "loss": 3.4392, + "step": 5581 + }, + { + "epoch": 0.29378947368421054, + "grad_norm": 0.515625, + "learning_rate": 0.0001569356725146199, + "loss": 3.5768, + "step": 5582 + }, + { + "epoch": 0.2938421052631579, + "grad_norm": 0.48046875, + "learning_rate": 0.00015692397660818714, + "loss": 3.9837, + "step": 5583 + }, + { + "epoch": 0.29389473684210526, + "grad_norm": 0.482421875, + "learning_rate": 0.0001569122807017544, + "loss": 3.4264, + "step": 5584 + }, + { + "epoch": 0.2939473684210526, + "grad_norm": 0.55078125, + "learning_rate": 0.00015690058479532164, + "loss": 3.9291, + "step": 5585 + }, + { + "epoch": 0.294, + "grad_norm": 0.478515625, + "learning_rate": 0.00015688888888888892, + "loss": 3.5377, + "step": 5586 + }, + { + "epoch": 0.29405263157894734, + "grad_norm": 0.50390625, + "learning_rate": 0.00015687719298245617, + "loss": 3.4728, + "step": 5587 + }, + { + "epoch": 0.29410526315789476, + "grad_norm": 0.51171875, + "learning_rate": 0.0001568654970760234, + "loss": 3.8948, + "step": 5588 + }, + { + "epoch": 0.2941578947368421, + "grad_norm": 0.46484375, + "learning_rate": 0.00015685380116959064, + "loss": 3.5764, + "step": 5589 + }, + { + "epoch": 0.2942105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.0001568421052631579, + "loss": 3.961, + "step": 5590 + }, + { + "epoch": 0.29426315789473684, + "grad_norm": 0.5625, + "learning_rate": 0.00015683040935672514, + "loss": 3.8095, + "step": 5591 + }, + { + "epoch": 0.2943157894736842, + "grad_norm": 0.59375, + "learning_rate": 0.0001568187134502924, + "loss": 3.9435, + "step": 5592 + }, + { + "epoch": 0.29436842105263156, + "grad_norm": 0.52734375, + "learning_rate": 0.00015680701754385964, + "loss": 3.6358, + "step": 5593 + }, + { + "epoch": 0.29442105263157897, + "grad_norm": 0.498046875, + "learning_rate": 0.00015679532163742692, + "loss": 3.4541, + "step": 5594 + }, + { + "epoch": 0.29447368421052633, + "grad_norm": 0.5234375, + "learning_rate": 0.00015678362573099417, + "loss": 3.789, + "step": 5595 + }, + { + "epoch": 0.2945263157894737, + "grad_norm": 0.5390625, + "learning_rate": 0.00015677192982456142, + "loss": 3.8585, + "step": 5596 + }, + { + "epoch": 0.29457894736842105, + "grad_norm": 0.5625, + "learning_rate": 0.00015676023391812867, + "loss": 3.7005, + "step": 5597 + }, + { + "epoch": 0.2946315789473684, + "grad_norm": 0.5234375, + "learning_rate": 0.00015674853801169591, + "loss": 3.4009, + "step": 5598 + }, + { + "epoch": 0.29468421052631577, + "grad_norm": 0.515625, + "learning_rate": 0.00015673684210526316, + "loss": 3.5975, + "step": 5599 + }, + { + "epoch": 0.29473684210526313, + "grad_norm": 0.5234375, + "learning_rate": 0.00015672514619883041, + "loss": 3.8006, + "step": 5600 + }, + { + "epoch": 0.29478947368421055, + "grad_norm": 0.5390625, + "learning_rate": 0.0001567134502923977, + "loss": 3.6482, + "step": 5601 + }, + { + "epoch": 0.2948421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00015670175438596494, + "loss": 3.5332, + "step": 5602 + }, + { + "epoch": 0.29489473684210527, + "grad_norm": 0.5078125, + "learning_rate": 0.00015669005847953216, + "loss": 4.06, + "step": 5603 + }, + { + "epoch": 0.2949473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.0001566783625730994, + "loss": 3.9615, + "step": 5604 + }, + { + "epoch": 0.295, + "grad_norm": 0.55078125, + "learning_rate": 0.00015666666666666666, + "loss": 3.2113, + "step": 5605 + }, + { + "epoch": 0.29505263157894734, + "grad_norm": 0.478515625, + "learning_rate": 0.0001566549707602339, + "loss": 3.4892, + "step": 5606 + }, + { + "epoch": 0.29510526315789476, + "grad_norm": 0.5234375, + "learning_rate": 0.00015664327485380116, + "loss": 3.7812, + "step": 5607 + }, + { + "epoch": 0.2951578947368421, + "grad_norm": 0.5703125, + "learning_rate": 0.0001566315789473684, + "loss": 3.9927, + "step": 5608 + }, + { + "epoch": 0.2952105263157895, + "grad_norm": 0.58203125, + "learning_rate": 0.0001566198830409357, + "loss": 3.5407, + "step": 5609 + }, + { + "epoch": 0.29526315789473684, + "grad_norm": 0.53125, + "learning_rate": 0.00015660818713450294, + "loss": 3.7558, + "step": 5610 + }, + { + "epoch": 0.2953157894736842, + "grad_norm": 0.52734375, + "learning_rate": 0.0001565964912280702, + "loss": 3.6484, + "step": 5611 + }, + { + "epoch": 0.29536842105263156, + "grad_norm": 0.62890625, + "learning_rate": 0.00015658479532163744, + "loss": 3.3341, + "step": 5612 + }, + { + "epoch": 0.295421052631579, + "grad_norm": 0.58203125, + "learning_rate": 0.00015657309941520469, + "loss": 3.886, + "step": 5613 + }, + { + "epoch": 0.29547368421052633, + "grad_norm": 0.56640625, + "learning_rate": 0.00015656140350877194, + "loss": 3.393, + "step": 5614 + }, + { + "epoch": 0.2955263157894737, + "grad_norm": 0.60546875, + "learning_rate": 0.00015654970760233919, + "loss": 3.8723, + "step": 5615 + }, + { + "epoch": 0.29557894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.00015653801169590646, + "loss": 3.995, + "step": 5616 + }, + { + "epoch": 0.2956315789473684, + "grad_norm": 0.57421875, + "learning_rate": 0.0001565263157894737, + "loss": 4.2567, + "step": 5617 + }, + { + "epoch": 0.29568421052631577, + "grad_norm": 0.53125, + "learning_rate": 0.00015651461988304093, + "loss": 3.9103, + "step": 5618 + }, + { + "epoch": 0.29573684210526313, + "grad_norm": 0.48828125, + "learning_rate": 0.00015650292397660818, + "loss": 3.248, + "step": 5619 + }, + { + "epoch": 0.29578947368421055, + "grad_norm": 0.4609375, + "learning_rate": 0.00015649122807017543, + "loss": 4.1374, + "step": 5620 + }, + { + "epoch": 0.2958421052631579, + "grad_norm": 0.490234375, + "learning_rate": 0.00015647953216374268, + "loss": 3.866, + "step": 5621 + }, + { + "epoch": 0.29589473684210527, + "grad_norm": 0.50390625, + "learning_rate": 0.00015646783625730993, + "loss": 4.1733, + "step": 5622 + }, + { + "epoch": 0.2959473684210526, + "grad_norm": 0.58203125, + "learning_rate": 0.00015645614035087718, + "loss": 4.1307, + "step": 5623 + }, + { + "epoch": 0.296, + "grad_norm": 0.52734375, + "learning_rate": 0.00015644444444444446, + "loss": 3.5329, + "step": 5624 + }, + { + "epoch": 0.29605263157894735, + "grad_norm": 0.5390625, + "learning_rate": 0.0001564327485380117, + "loss": 3.9817, + "step": 5625 + }, + { + "epoch": 0.29610526315789476, + "grad_norm": 0.55078125, + "learning_rate": 0.00015642105263157896, + "loss": 3.4078, + "step": 5626 + }, + { + "epoch": 0.2961578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.0001564093567251462, + "loss": 3.8244, + "step": 5627 + }, + { + "epoch": 0.2962105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00015639766081871346, + "loss": 3.847, + "step": 5628 + }, + { + "epoch": 0.29626315789473684, + "grad_norm": 0.54296875, + "learning_rate": 0.0001563859649122807, + "loss": 4.207, + "step": 5629 + }, + { + "epoch": 0.2963157894736842, + "grad_norm": 0.57421875, + "learning_rate": 0.00015637426900584796, + "loss": 3.6547, + "step": 5630 + }, + { + "epoch": 0.29636842105263156, + "grad_norm": 0.55078125, + "learning_rate": 0.00015636257309941523, + "loss": 3.4066, + "step": 5631 + }, + { + "epoch": 0.296421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00015635087719298248, + "loss": 3.5285, + "step": 5632 + }, + { + "epoch": 0.29647368421052633, + "grad_norm": 0.57421875, + "learning_rate": 0.00015633918128654973, + "loss": 4.0425, + "step": 5633 + }, + { + "epoch": 0.2965263157894737, + "grad_norm": 0.66015625, + "learning_rate": 0.00015632748538011695, + "loss": 3.9784, + "step": 5634 + }, + { + "epoch": 0.29657894736842105, + "grad_norm": 0.7109375, + "learning_rate": 0.0001563157894736842, + "loss": 3.8612, + "step": 5635 + }, + { + "epoch": 0.2966315789473684, + "grad_norm": 0.66015625, + "learning_rate": 0.00015630409356725145, + "loss": 3.6019, + "step": 5636 + }, + { + "epoch": 0.2966842105263158, + "grad_norm": 0.63671875, + "learning_rate": 0.0001562923976608187, + "loss": 3.8361, + "step": 5637 + }, + { + "epoch": 0.29673684210526313, + "grad_norm": 0.6953125, + "learning_rate": 0.00015628070175438598, + "loss": 3.8827, + "step": 5638 + }, + { + "epoch": 0.29678947368421055, + "grad_norm": 0.52734375, + "learning_rate": 0.00015626900584795323, + "loss": 3.8075, + "step": 5639 + }, + { + "epoch": 0.2968421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00015625730994152048, + "loss": 3.7459, + "step": 5640 + }, + { + "epoch": 0.29689473684210527, + "grad_norm": 0.5546875, + "learning_rate": 0.00015624561403508773, + "loss": 3.5549, + "step": 5641 + }, + { + "epoch": 0.2969473684210526, + "grad_norm": 0.53125, + "learning_rate": 0.00015623391812865498, + "loss": 3.7382, + "step": 5642 + }, + { + "epoch": 0.297, + "grad_norm": 0.5546875, + "learning_rate": 0.00015622222222222223, + "loss": 3.4744, + "step": 5643 + }, + { + "epoch": 0.29705263157894735, + "grad_norm": 0.52734375, + "learning_rate": 0.00015621052631578948, + "loss": 3.9513, + "step": 5644 + }, + { + "epoch": 0.29710526315789476, + "grad_norm": 0.54296875, + "learning_rate": 0.00015619883040935673, + "loss": 3.5671, + "step": 5645 + }, + { + "epoch": 0.2971578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.000156187134502924, + "loss": 3.382, + "step": 5646 + }, + { + "epoch": 0.2972105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.00015617543859649125, + "loss": 3.5013, + "step": 5647 + }, + { + "epoch": 0.29726315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.0001561637426900585, + "loss": 3.8236, + "step": 5648 + }, + { + "epoch": 0.2973157894736842, + "grad_norm": 0.640625, + "learning_rate": 0.00015615204678362573, + "loss": 3.5581, + "step": 5649 + }, + { + "epoch": 0.29736842105263156, + "grad_norm": 0.53125, + "learning_rate": 0.00015614035087719297, + "loss": 3.3083, + "step": 5650 + }, + { + "epoch": 0.297421052631579, + "grad_norm": 0.55078125, + "learning_rate": 0.00015612865497076022, + "loss": 4.0265, + "step": 5651 + }, + { + "epoch": 0.29747368421052633, + "grad_norm": 0.59375, + "learning_rate": 0.00015611695906432747, + "loss": 3.485, + "step": 5652 + }, + { + "epoch": 0.2975263157894737, + "grad_norm": 0.48828125, + "learning_rate": 0.00015610526315789475, + "loss": 3.9311, + "step": 5653 + }, + { + "epoch": 0.29757894736842105, + "grad_norm": 0.60546875, + "learning_rate": 0.000156093567251462, + "loss": 3.7908, + "step": 5654 + }, + { + "epoch": 0.2976315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.00015608187134502925, + "loss": 3.6448, + "step": 5655 + }, + { + "epoch": 0.2976842105263158, + "grad_norm": 0.609375, + "learning_rate": 0.0001560701754385965, + "loss": 3.5994, + "step": 5656 + }, + { + "epoch": 0.29773684210526313, + "grad_norm": 0.52734375, + "learning_rate": 0.00015605847953216375, + "loss": 3.7942, + "step": 5657 + }, + { + "epoch": 0.29778947368421055, + "grad_norm": 0.4921875, + "learning_rate": 0.000156046783625731, + "loss": 4.1539, + "step": 5658 + }, + { + "epoch": 0.2978421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00015603508771929825, + "loss": 3.7616, + "step": 5659 + }, + { + "epoch": 0.29789473684210527, + "grad_norm": 0.5234375, + "learning_rate": 0.0001560233918128655, + "loss": 3.6847, + "step": 5660 + }, + { + "epoch": 0.2979473684210526, + "grad_norm": 0.54296875, + "learning_rate": 0.00015601169590643277, + "loss": 3.9497, + "step": 5661 + }, + { + "epoch": 0.298, + "grad_norm": 0.5078125, + "learning_rate": 0.00015600000000000002, + "loss": 4.0904, + "step": 5662 + }, + { + "epoch": 0.29805263157894735, + "grad_norm": 0.515625, + "learning_rate": 0.00015598830409356727, + "loss": 3.8147, + "step": 5663 + }, + { + "epoch": 0.29810526315789476, + "grad_norm": 0.515625, + "learning_rate": 0.0001559766081871345, + "loss": 3.5806, + "step": 5664 + }, + { + "epoch": 0.2981578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00015596491228070175, + "loss": 4.0683, + "step": 5665 + }, + { + "epoch": 0.2982105263157895, + "grad_norm": 0.515625, + "learning_rate": 0.000155953216374269, + "loss": 4.0667, + "step": 5666 + }, + { + "epoch": 0.29826315789473684, + "grad_norm": 0.46484375, + "learning_rate": 0.00015594152046783625, + "loss": 3.6936, + "step": 5667 + }, + { + "epoch": 0.2983157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00015592982456140352, + "loss": 4.0911, + "step": 5668 + }, + { + "epoch": 0.29836842105263156, + "grad_norm": 0.58203125, + "learning_rate": 0.00015591812865497077, + "loss": 4.0929, + "step": 5669 + }, + { + "epoch": 0.2984210526315789, + "grad_norm": 0.59765625, + "learning_rate": 0.00015590643274853802, + "loss": 3.6772, + "step": 5670 + }, + { + "epoch": 0.29847368421052634, + "grad_norm": 0.90234375, + "learning_rate": 0.00015589473684210527, + "loss": 3.9268, + "step": 5671 + }, + { + "epoch": 0.2985263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.00015588304093567252, + "loss": 3.8856, + "step": 5672 + }, + { + "epoch": 0.29857894736842105, + "grad_norm": 0.56640625, + "learning_rate": 0.00015587134502923977, + "loss": 3.6927, + "step": 5673 + }, + { + "epoch": 0.2986315789473684, + "grad_norm": 0.58984375, + "learning_rate": 0.00015585964912280702, + "loss": 3.8843, + "step": 5674 + }, + { + "epoch": 0.2986842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.0001558479532163743, + "loss": 3.9493, + "step": 5675 + }, + { + "epoch": 0.29873684210526313, + "grad_norm": 0.57421875, + "learning_rate": 0.00015583625730994155, + "loss": 3.8967, + "step": 5676 + }, + { + "epoch": 0.29878947368421055, + "grad_norm": 0.58203125, + "learning_rate": 0.0001558245614035088, + "loss": 4.1445, + "step": 5677 + }, + { + "epoch": 0.2988421052631579, + "grad_norm": 0.5859375, + "learning_rate": 0.00015581286549707604, + "loss": 3.7854, + "step": 5678 + }, + { + "epoch": 0.29889473684210527, + "grad_norm": 0.53125, + "learning_rate": 0.0001558011695906433, + "loss": 3.8722, + "step": 5679 + }, + { + "epoch": 0.29894736842105263, + "grad_norm": 0.6171875, + "learning_rate": 0.00015578947368421052, + "loss": 3.3681, + "step": 5680 + }, + { + "epoch": 0.299, + "grad_norm": 0.49609375, + "learning_rate": 0.00015577777777777777, + "loss": 3.9658, + "step": 5681 + }, + { + "epoch": 0.29905263157894735, + "grad_norm": 0.484375, + "learning_rate": 0.00015576608187134502, + "loss": 3.7778, + "step": 5682 + }, + { + "epoch": 0.29910526315789476, + "grad_norm": 0.494140625, + "learning_rate": 0.0001557543859649123, + "loss": 3.292, + "step": 5683 + }, + { + "epoch": 0.2991578947368421, + "grad_norm": 0.46875, + "learning_rate": 0.00015574269005847954, + "loss": 3.8699, + "step": 5684 + }, + { + "epoch": 0.2992105263157895, + "grad_norm": 0.52734375, + "learning_rate": 0.0001557309941520468, + "loss": 3.6636, + "step": 5685 + }, + { + "epoch": 0.29926315789473684, + "grad_norm": 0.52734375, + "learning_rate": 0.00015571929824561404, + "loss": 3.8268, + "step": 5686 + }, + { + "epoch": 0.2993157894736842, + "grad_norm": 0.486328125, + "learning_rate": 0.0001557076023391813, + "loss": 3.847, + "step": 5687 + }, + { + "epoch": 0.29936842105263156, + "grad_norm": 0.5, + "learning_rate": 0.00015569590643274854, + "loss": 3.795, + "step": 5688 + }, + { + "epoch": 0.2994210526315789, + "grad_norm": 0.49609375, + "learning_rate": 0.0001556842105263158, + "loss": 3.573, + "step": 5689 + }, + { + "epoch": 0.29947368421052634, + "grad_norm": 0.466796875, + "learning_rate": 0.00015567251461988307, + "loss": 3.8941, + "step": 5690 + }, + { + "epoch": 0.2995263157894737, + "grad_norm": 0.60546875, + "learning_rate": 0.00015566081871345032, + "loss": 3.8258, + "step": 5691 + }, + { + "epoch": 0.29957894736842106, + "grad_norm": 0.5, + "learning_rate": 0.00015564912280701757, + "loss": 3.6707, + "step": 5692 + }, + { + "epoch": 0.2996315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00015563742690058482, + "loss": 3.4816, + "step": 5693 + }, + { + "epoch": 0.2996842105263158, + "grad_norm": 0.484375, + "learning_rate": 0.00015562573099415207, + "loss": 3.954, + "step": 5694 + }, + { + "epoch": 0.29973684210526313, + "grad_norm": 0.51171875, + "learning_rate": 0.0001556140350877193, + "loss": 3.9993, + "step": 5695 + }, + { + "epoch": 0.29978947368421055, + "grad_norm": 0.53515625, + "learning_rate": 0.00015560233918128654, + "loss": 3.8898, + "step": 5696 + }, + { + "epoch": 0.2998421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.0001555906432748538, + "loss": 3.8095, + "step": 5697 + }, + { + "epoch": 0.29989473684210527, + "grad_norm": 0.51171875, + "learning_rate": 0.00015557894736842106, + "loss": 3.7302, + "step": 5698 + }, + { + "epoch": 0.29994736842105263, + "grad_norm": 0.48046875, + "learning_rate": 0.0001555672514619883, + "loss": 3.8538, + "step": 5699 + }, + { + "epoch": 0.3, + "grad_norm": 0.51171875, + "learning_rate": 0.00015555555555555556, + "loss": 3.505, + "step": 5700 + }, + { + "epoch": 0.30005263157894735, + "grad_norm": 0.5546875, + "learning_rate": 0.0001555438596491228, + "loss": 3.6582, + "step": 5701 + }, + { + "epoch": 0.30010526315789476, + "grad_norm": 0.5, + "learning_rate": 0.00015553216374269006, + "loss": 4.3363, + "step": 5702 + }, + { + "epoch": 0.3001578947368421, + "grad_norm": 0.5703125, + "learning_rate": 0.0001555204678362573, + "loss": 4.0567, + "step": 5703 + }, + { + "epoch": 0.3002105263157895, + "grad_norm": 0.5390625, + "learning_rate": 0.00015550877192982456, + "loss": 3.006, + "step": 5704 + }, + { + "epoch": 0.30026315789473684, + "grad_norm": 0.46875, + "learning_rate": 0.00015549707602339184, + "loss": 3.7929, + "step": 5705 + }, + { + "epoch": 0.3003157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.0001554853801169591, + "loss": 3.8278, + "step": 5706 + }, + { + "epoch": 0.30036842105263156, + "grad_norm": 0.54296875, + "learning_rate": 0.00015547368421052634, + "loss": 3.2787, + "step": 5707 + }, + { + "epoch": 0.3004210526315789, + "grad_norm": 0.494140625, + "learning_rate": 0.0001554619883040936, + "loss": 3.9711, + "step": 5708 + }, + { + "epoch": 0.30047368421052634, + "grad_norm": 0.466796875, + "learning_rate": 0.00015545029239766084, + "loss": 3.4465, + "step": 5709 + }, + { + "epoch": 0.3005263157894737, + "grad_norm": 0.52734375, + "learning_rate": 0.00015543859649122806, + "loss": 3.6146, + "step": 5710 + }, + { + "epoch": 0.30057894736842106, + "grad_norm": 0.515625, + "learning_rate": 0.0001554269005847953, + "loss": 3.7763, + "step": 5711 + }, + { + "epoch": 0.3006315789473684, + "grad_norm": 0.490234375, + "learning_rate": 0.00015541520467836259, + "loss": 3.489, + "step": 5712 + }, + { + "epoch": 0.3006842105263158, + "grad_norm": 0.578125, + "learning_rate": 0.00015540350877192983, + "loss": 3.5511, + "step": 5713 + }, + { + "epoch": 0.30073684210526314, + "grad_norm": 0.5078125, + "learning_rate": 0.00015539181286549708, + "loss": 3.8016, + "step": 5714 + }, + { + "epoch": 0.30078947368421055, + "grad_norm": 0.52734375, + "learning_rate": 0.00015538011695906433, + "loss": 3.67, + "step": 5715 + }, + { + "epoch": 0.3008421052631579, + "grad_norm": 0.578125, + "learning_rate": 0.00015536842105263158, + "loss": 3.7889, + "step": 5716 + }, + { + "epoch": 0.30089473684210527, + "grad_norm": 0.5703125, + "learning_rate": 0.00015535672514619883, + "loss": 3.9437, + "step": 5717 + }, + { + "epoch": 0.30094736842105263, + "grad_norm": 0.58984375, + "learning_rate": 0.00015534502923976608, + "loss": 3.6501, + "step": 5718 + }, + { + "epoch": 0.301, + "grad_norm": 0.5078125, + "learning_rate": 0.00015533333333333333, + "loss": 3.7551, + "step": 5719 + }, + { + "epoch": 0.30105263157894735, + "grad_norm": 0.4765625, + "learning_rate": 0.0001553216374269006, + "loss": 3.8669, + "step": 5720 + }, + { + "epoch": 0.3011052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.00015530994152046786, + "loss": 4.1814, + "step": 5721 + }, + { + "epoch": 0.3011578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.0001552982456140351, + "loss": 3.6027, + "step": 5722 + }, + { + "epoch": 0.3012105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.00015528654970760236, + "loss": 3.6281, + "step": 5723 + }, + { + "epoch": 0.30126315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.0001552748538011696, + "loss": 3.9969, + "step": 5724 + }, + { + "epoch": 0.3013157894736842, + "grad_norm": 0.48828125, + "learning_rate": 0.00015526315789473686, + "loss": 3.2776, + "step": 5725 + }, + { + "epoch": 0.30136842105263156, + "grad_norm": 0.46875, + "learning_rate": 0.00015525146198830408, + "loss": 3.8819, + "step": 5726 + }, + { + "epoch": 0.3014210526315789, + "grad_norm": 0.50390625, + "learning_rate": 0.00015523976608187136, + "loss": 3.8788, + "step": 5727 + }, + { + "epoch": 0.30147368421052634, + "grad_norm": 1.296875, + "learning_rate": 0.0001552280701754386, + "loss": 3.4244, + "step": 5728 + }, + { + "epoch": 0.3015263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.00015521637426900586, + "loss": 3.7697, + "step": 5729 + }, + { + "epoch": 0.30157894736842106, + "grad_norm": 0.5, + "learning_rate": 0.0001552046783625731, + "loss": 3.659, + "step": 5730 + }, + { + "epoch": 0.3016315789473684, + "grad_norm": 0.52734375, + "learning_rate": 0.00015519298245614035, + "loss": 4.2266, + "step": 5731 + }, + { + "epoch": 0.3016842105263158, + "grad_norm": 0.5390625, + "learning_rate": 0.0001551812865497076, + "loss": 3.4899, + "step": 5732 + }, + { + "epoch": 0.30173684210526314, + "grad_norm": 0.50390625, + "learning_rate": 0.00015516959064327485, + "loss": 3.5962, + "step": 5733 + }, + { + "epoch": 0.30178947368421055, + "grad_norm": 0.498046875, + "learning_rate": 0.0001551578947368421, + "loss": 3.7448, + "step": 5734 + }, + { + "epoch": 0.3018421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00015514619883040938, + "loss": 3.1208, + "step": 5735 + }, + { + "epoch": 0.30189473684210527, + "grad_norm": 0.490234375, + "learning_rate": 0.00015513450292397663, + "loss": 3.8238, + "step": 5736 + }, + { + "epoch": 0.30194736842105263, + "grad_norm": 0.51953125, + "learning_rate": 0.00015512280701754388, + "loss": 3.6077, + "step": 5737 + }, + { + "epoch": 0.302, + "grad_norm": 0.50390625, + "learning_rate": 0.00015511111111111113, + "loss": 3.8341, + "step": 5738 + }, + { + "epoch": 0.30205263157894735, + "grad_norm": 0.5078125, + "learning_rate": 0.00015509941520467838, + "loss": 3.755, + "step": 5739 + }, + { + "epoch": 0.3021052631578947, + "grad_norm": 0.48828125, + "learning_rate": 0.00015508771929824563, + "loss": 3.5098, + "step": 5740 + }, + { + "epoch": 0.3021578947368421, + "grad_norm": 0.4765625, + "learning_rate": 0.00015507602339181285, + "loss": 3.846, + "step": 5741 + }, + { + "epoch": 0.3022105263157895, + "grad_norm": 0.486328125, + "learning_rate": 0.00015506432748538013, + "loss": 3.5521, + "step": 5742 + }, + { + "epoch": 0.30226315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.00015505263157894738, + "loss": 3.7609, + "step": 5743 + }, + { + "epoch": 0.3023157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00015504093567251463, + "loss": 3.6603, + "step": 5744 + }, + { + "epoch": 0.30236842105263156, + "grad_norm": 0.55078125, + "learning_rate": 0.00015502923976608188, + "loss": 3.8098, + "step": 5745 + }, + { + "epoch": 0.3024210526315789, + "grad_norm": 0.55859375, + "learning_rate": 0.00015501754385964913, + "loss": 3.602, + "step": 5746 + }, + { + "epoch": 0.30247368421052634, + "grad_norm": 0.51171875, + "learning_rate": 0.00015500584795321638, + "loss": 3.5554, + "step": 5747 + }, + { + "epoch": 0.3025263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00015499415204678362, + "loss": 3.8398, + "step": 5748 + }, + { + "epoch": 0.30257894736842106, + "grad_norm": 0.55859375, + "learning_rate": 0.0001549824561403509, + "loss": 3.6961, + "step": 5749 + }, + { + "epoch": 0.3026315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00015497076023391815, + "loss": 3.6817, + "step": 5750 + }, + { + "epoch": 0.3026842105263158, + "grad_norm": 0.49609375, + "learning_rate": 0.0001549590643274854, + "loss": 3.6899, + "step": 5751 + }, + { + "epoch": 0.30273684210526314, + "grad_norm": 0.54296875, + "learning_rate": 0.00015494736842105265, + "loss": 3.8911, + "step": 5752 + }, + { + "epoch": 0.30278947368421055, + "grad_norm": 0.51953125, + "learning_rate": 0.0001549356725146199, + "loss": 3.5605, + "step": 5753 + }, + { + "epoch": 0.3028421052631579, + "grad_norm": 0.55078125, + "learning_rate": 0.00015492397660818715, + "loss": 3.2653, + "step": 5754 + }, + { + "epoch": 0.30289473684210527, + "grad_norm": 0.53125, + "learning_rate": 0.0001549122807017544, + "loss": 3.922, + "step": 5755 + }, + { + "epoch": 0.30294736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.00015490058479532162, + "loss": 3.6217, + "step": 5756 + }, + { + "epoch": 0.303, + "grad_norm": 0.48046875, + "learning_rate": 0.0001548888888888889, + "loss": 3.6212, + "step": 5757 + }, + { + "epoch": 0.30305263157894735, + "grad_norm": 0.494140625, + "learning_rate": 0.00015487719298245615, + "loss": 4.0302, + "step": 5758 + }, + { + "epoch": 0.3031052631578947, + "grad_norm": 0.48046875, + "learning_rate": 0.0001548654970760234, + "loss": 4.0718, + "step": 5759 + }, + { + "epoch": 0.3031578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00015485380116959065, + "loss": 3.8792, + "step": 5760 + }, + { + "epoch": 0.3032105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.0001548421052631579, + "loss": 3.9396, + "step": 5761 + }, + { + "epoch": 0.30326315789473685, + "grad_norm": 0.5546875, + "learning_rate": 0.00015483040935672515, + "loss": 3.7013, + "step": 5762 + }, + { + "epoch": 0.3033157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.0001548187134502924, + "loss": 3.9839, + "step": 5763 + }, + { + "epoch": 0.30336842105263156, + "grad_norm": 0.59765625, + "learning_rate": 0.00015480701754385967, + "loss": 3.3077, + "step": 5764 + }, + { + "epoch": 0.3034210526315789, + "grad_norm": 0.515625, + "learning_rate": 0.00015479532163742692, + "loss": 3.3876, + "step": 5765 + }, + { + "epoch": 0.30347368421052634, + "grad_norm": 0.50390625, + "learning_rate": 0.00015478362573099417, + "loss": 3.6457, + "step": 5766 + }, + { + "epoch": 0.3035263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00015477192982456142, + "loss": 4.0253, + "step": 5767 + }, + { + "epoch": 0.30357894736842106, + "grad_norm": 0.56640625, + "learning_rate": 0.00015476023391812867, + "loss": 3.8348, + "step": 5768 + }, + { + "epoch": 0.3036315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.00015474853801169592, + "loss": 3.5386, + "step": 5769 + }, + { + "epoch": 0.3036842105263158, + "grad_norm": 0.494140625, + "learning_rate": 0.00015473684210526317, + "loss": 3.6692, + "step": 5770 + }, + { + "epoch": 0.30373684210526314, + "grad_norm": 0.5078125, + "learning_rate": 0.00015472514619883042, + "loss": 3.5945, + "step": 5771 + }, + { + "epoch": 0.30378947368421055, + "grad_norm": 0.4765625, + "learning_rate": 0.00015471345029239767, + "loss": 3.7532, + "step": 5772 + }, + { + "epoch": 0.3038421052631579, + "grad_norm": 0.640625, + "learning_rate": 0.00015470175438596492, + "loss": 3.7754, + "step": 5773 + }, + { + "epoch": 0.3038947368421053, + "grad_norm": 0.458984375, + "learning_rate": 0.00015469005847953217, + "loss": 4.0476, + "step": 5774 + }, + { + "epoch": 0.30394736842105263, + "grad_norm": 0.50390625, + "learning_rate": 0.00015467836257309942, + "loss": 3.7702, + "step": 5775 + }, + { + "epoch": 0.304, + "grad_norm": 0.6875, + "learning_rate": 0.00015466666666666667, + "loss": 3.8099, + "step": 5776 + }, + { + "epoch": 0.30405263157894735, + "grad_norm": 0.5078125, + "learning_rate": 0.00015465497076023392, + "loss": 3.2662, + "step": 5777 + }, + { + "epoch": 0.3041052631578947, + "grad_norm": 0.609375, + "learning_rate": 0.00015464327485380117, + "loss": 3.7823, + "step": 5778 + }, + { + "epoch": 0.3041578947368421, + "grad_norm": 0.490234375, + "learning_rate": 0.00015463157894736844, + "loss": 3.5571, + "step": 5779 + }, + { + "epoch": 0.3042105263157895, + "grad_norm": 0.58984375, + "learning_rate": 0.0001546198830409357, + "loss": 3.9926, + "step": 5780 + }, + { + "epoch": 0.30426315789473685, + "grad_norm": 0.54296875, + "learning_rate": 0.00015460818713450294, + "loss": 3.9185, + "step": 5781 + }, + { + "epoch": 0.3043157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.0001545964912280702, + "loss": 3.2947, + "step": 5782 + }, + { + "epoch": 0.30436842105263157, + "grad_norm": 0.474609375, + "learning_rate": 0.00015458479532163744, + "loss": 4.0254, + "step": 5783 + }, + { + "epoch": 0.3044210526315789, + "grad_norm": 0.474609375, + "learning_rate": 0.0001545730994152047, + "loss": 3.671, + "step": 5784 + }, + { + "epoch": 0.30447368421052634, + "grad_norm": 0.53515625, + "learning_rate": 0.00015456140350877194, + "loss": 3.6811, + "step": 5785 + }, + { + "epoch": 0.3045263157894737, + "grad_norm": 0.58203125, + "learning_rate": 0.0001545497076023392, + "loss": 3.8301, + "step": 5786 + }, + { + "epoch": 0.30457894736842106, + "grad_norm": 0.6015625, + "learning_rate": 0.00015453801169590644, + "loss": 3.8059, + "step": 5787 + }, + { + "epoch": 0.3046315789473684, + "grad_norm": 0.54296875, + "learning_rate": 0.0001545263157894737, + "loss": 3.5797, + "step": 5788 + }, + { + "epoch": 0.3046842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00015451461988304094, + "loss": 3.5182, + "step": 5789 + }, + { + "epoch": 0.30473684210526314, + "grad_norm": 0.5234375, + "learning_rate": 0.0001545029239766082, + "loss": 3.8896, + "step": 5790 + }, + { + "epoch": 0.3047894736842105, + "grad_norm": 0.5234375, + "learning_rate": 0.00015449122807017544, + "loss": 3.7298, + "step": 5791 + }, + { + "epoch": 0.3048421052631579, + "grad_norm": 0.5859375, + "learning_rate": 0.0001544795321637427, + "loss": 4.0858, + "step": 5792 + }, + { + "epoch": 0.3048947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.00015446783625730994, + "loss": 4.0133, + "step": 5793 + }, + { + "epoch": 0.30494736842105263, + "grad_norm": 0.55078125, + "learning_rate": 0.00015445614035087721, + "loss": 3.5831, + "step": 5794 + }, + { + "epoch": 0.305, + "grad_norm": 0.56640625, + "learning_rate": 0.00015444444444444446, + "loss": 4.2708, + "step": 5795 + }, + { + "epoch": 0.30505263157894735, + "grad_norm": 0.58984375, + "learning_rate": 0.00015443274853801171, + "loss": 3.8134, + "step": 5796 + }, + { + "epoch": 0.3051052631578947, + "grad_norm": 0.5078125, + "learning_rate": 0.00015442105263157896, + "loss": 3.966, + "step": 5797 + }, + { + "epoch": 0.30515789473684213, + "grad_norm": 0.58984375, + "learning_rate": 0.0001544093567251462, + "loss": 3.4743, + "step": 5798 + }, + { + "epoch": 0.3052105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00015439766081871346, + "loss": 3.5917, + "step": 5799 + }, + { + "epoch": 0.30526315789473685, + "grad_norm": 0.5, + "learning_rate": 0.0001543859649122807, + "loss": 4.0175, + "step": 5800 + }, + { + "epoch": 0.3053157894736842, + "grad_norm": 0.55859375, + "learning_rate": 0.00015437426900584796, + "loss": 3.8283, + "step": 5801 + }, + { + "epoch": 0.30536842105263157, + "grad_norm": 0.60546875, + "learning_rate": 0.0001543625730994152, + "loss": 3.3792, + "step": 5802 + }, + { + "epoch": 0.3054210526315789, + "grad_norm": 0.52734375, + "learning_rate": 0.00015435087719298246, + "loss": 3.7383, + "step": 5803 + }, + { + "epoch": 0.30547368421052634, + "grad_norm": 0.5, + "learning_rate": 0.0001543391812865497, + "loss": 3.8115, + "step": 5804 + }, + { + "epoch": 0.3055263157894737, + "grad_norm": 0.494140625, + "learning_rate": 0.00015432748538011696, + "loss": 3.3939, + "step": 5805 + }, + { + "epoch": 0.30557894736842106, + "grad_norm": 0.5078125, + "learning_rate": 0.0001543157894736842, + "loss": 3.4528, + "step": 5806 + }, + { + "epoch": 0.3056315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.00015430409356725146, + "loss": 3.7083, + "step": 5807 + }, + { + "epoch": 0.3056842105263158, + "grad_norm": 0.6953125, + "learning_rate": 0.0001542923976608187, + "loss": 4.0485, + "step": 5808 + }, + { + "epoch": 0.30573684210526314, + "grad_norm": 0.5234375, + "learning_rate": 0.00015428070175438599, + "loss": 3.6792, + "step": 5809 + }, + { + "epoch": 0.3057894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.00015426900584795324, + "loss": 3.6314, + "step": 5810 + }, + { + "epoch": 0.3058421052631579, + "grad_norm": 0.58203125, + "learning_rate": 0.00015425730994152048, + "loss": 3.8777, + "step": 5811 + }, + { + "epoch": 0.3058947368421053, + "grad_norm": 0.578125, + "learning_rate": 0.00015424561403508773, + "loss": 4.0157, + "step": 5812 + }, + { + "epoch": 0.30594736842105263, + "grad_norm": 0.5625, + "learning_rate": 0.00015423391812865498, + "loss": 3.9803, + "step": 5813 + }, + { + "epoch": 0.306, + "grad_norm": 0.52734375, + "learning_rate": 0.00015422222222222223, + "loss": 3.6766, + "step": 5814 + }, + { + "epoch": 0.30605263157894735, + "grad_norm": 0.54296875, + "learning_rate": 0.00015421052631578948, + "loss": 3.7365, + "step": 5815 + }, + { + "epoch": 0.3061052631578947, + "grad_norm": 0.56640625, + "learning_rate": 0.00015419883040935673, + "loss": 3.8301, + "step": 5816 + }, + { + "epoch": 0.30615789473684213, + "grad_norm": 0.55078125, + "learning_rate": 0.00015418713450292398, + "loss": 3.3219, + "step": 5817 + }, + { + "epoch": 0.3062105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.00015417543859649123, + "loss": 3.7202, + "step": 5818 + }, + { + "epoch": 0.30626315789473685, + "grad_norm": 0.51171875, + "learning_rate": 0.00015416374269005848, + "loss": 3.8384, + "step": 5819 + }, + { + "epoch": 0.3063157894736842, + "grad_norm": 0.484375, + "learning_rate": 0.00015415204678362573, + "loss": 3.3987, + "step": 5820 + }, + { + "epoch": 0.30636842105263157, + "grad_norm": 0.5390625, + "learning_rate": 0.00015414035087719298, + "loss": 3.319, + "step": 5821 + }, + { + "epoch": 0.3064210526315789, + "grad_norm": 0.498046875, + "learning_rate": 0.00015412865497076023, + "loss": 3.7322, + "step": 5822 + }, + { + "epoch": 0.30647368421052634, + "grad_norm": 0.54296875, + "learning_rate": 0.00015411695906432748, + "loss": 3.8237, + "step": 5823 + }, + { + "epoch": 0.3065263157894737, + "grad_norm": 0.58203125, + "learning_rate": 0.00015410526315789476, + "loss": 3.0451, + "step": 5824 + }, + { + "epoch": 0.30657894736842106, + "grad_norm": 0.5546875, + "learning_rate": 0.000154093567251462, + "loss": 3.5308, + "step": 5825 + }, + { + "epoch": 0.3066315789473684, + "grad_norm": 0.58203125, + "learning_rate": 0.00015408187134502926, + "loss": 3.8539, + "step": 5826 + }, + { + "epoch": 0.3066842105263158, + "grad_norm": 0.5390625, + "learning_rate": 0.0001540701754385965, + "loss": 3.6215, + "step": 5827 + }, + { + "epoch": 0.30673684210526314, + "grad_norm": 0.546875, + "learning_rate": 0.00015405847953216375, + "loss": 3.9787, + "step": 5828 + }, + { + "epoch": 0.3067894736842105, + "grad_norm": 0.62890625, + "learning_rate": 0.000154046783625731, + "loss": 3.9639, + "step": 5829 + }, + { + "epoch": 0.3068421052631579, + "grad_norm": 0.59375, + "learning_rate": 0.00015403508771929825, + "loss": 3.7953, + "step": 5830 + }, + { + "epoch": 0.3068947368421053, + "grad_norm": 0.5625, + "learning_rate": 0.0001540233918128655, + "loss": 3.8024, + "step": 5831 + }, + { + "epoch": 0.30694736842105264, + "grad_norm": 0.5703125, + "learning_rate": 0.00015401169590643275, + "loss": 3.5059, + "step": 5832 + }, + { + "epoch": 0.307, + "grad_norm": 0.54296875, + "learning_rate": 0.000154, + "loss": 3.9885, + "step": 5833 + }, + { + "epoch": 0.30705263157894735, + "grad_norm": 0.52734375, + "learning_rate": 0.00015398830409356725, + "loss": 3.3263, + "step": 5834 + }, + { + "epoch": 0.3071052631578947, + "grad_norm": 0.5234375, + "learning_rate": 0.0001539766081871345, + "loss": 3.7533, + "step": 5835 + }, + { + "epoch": 0.30715789473684213, + "grad_norm": 0.6484375, + "learning_rate": 0.00015396491228070175, + "loss": 3.4141, + "step": 5836 + }, + { + "epoch": 0.3072105263157895, + "grad_norm": 0.55078125, + "learning_rate": 0.000153953216374269, + "loss": 4.1902, + "step": 5837 + }, + { + "epoch": 0.30726315789473685, + "grad_norm": 0.515625, + "learning_rate": 0.00015394152046783628, + "loss": 3.6316, + "step": 5838 + }, + { + "epoch": 0.3073157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.00015392982456140353, + "loss": 3.3123, + "step": 5839 + }, + { + "epoch": 0.30736842105263157, + "grad_norm": 0.4765625, + "learning_rate": 0.00015391812865497078, + "loss": 3.6266, + "step": 5840 + }, + { + "epoch": 0.30742105263157893, + "grad_norm": 0.498046875, + "learning_rate": 0.00015390643274853803, + "loss": 3.9026, + "step": 5841 + }, + { + "epoch": 0.30747368421052634, + "grad_norm": 0.57421875, + "learning_rate": 0.00015389473684210528, + "loss": 3.6976, + "step": 5842 + }, + { + "epoch": 0.3075263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00015388304093567253, + "loss": 3.829, + "step": 5843 + }, + { + "epoch": 0.30757894736842106, + "grad_norm": 0.50390625, + "learning_rate": 0.00015387134502923978, + "loss": 3.9067, + "step": 5844 + }, + { + "epoch": 0.3076315789473684, + "grad_norm": 0.59765625, + "learning_rate": 0.00015385964912280702, + "loss": 3.2424, + "step": 5845 + }, + { + "epoch": 0.3076842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00015384795321637427, + "loss": 3.662, + "step": 5846 + }, + { + "epoch": 0.30773684210526314, + "grad_norm": 0.59765625, + "learning_rate": 0.00015383625730994152, + "loss": 3.2372, + "step": 5847 + }, + { + "epoch": 0.3077894736842105, + "grad_norm": 0.47265625, + "learning_rate": 0.00015382456140350877, + "loss": 3.8111, + "step": 5848 + }, + { + "epoch": 0.3078421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00015381286549707602, + "loss": 3.2541, + "step": 5849 + }, + { + "epoch": 0.3078947368421053, + "grad_norm": 0.57421875, + "learning_rate": 0.00015380116959064327, + "loss": 3.5841, + "step": 5850 + }, + { + "epoch": 0.30794736842105264, + "grad_norm": 0.546875, + "learning_rate": 0.00015378947368421052, + "loss": 3.3821, + "step": 5851 + }, + { + "epoch": 0.308, + "grad_norm": 0.54296875, + "learning_rate": 0.00015377777777777777, + "loss": 3.4727, + "step": 5852 + }, + { + "epoch": 0.30805263157894736, + "grad_norm": 0.455078125, + "learning_rate": 0.00015376608187134505, + "loss": 3.4392, + "step": 5853 + }, + { + "epoch": 0.3081052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.0001537543859649123, + "loss": 3.4238, + "step": 5854 + }, + { + "epoch": 0.30815789473684213, + "grad_norm": 0.451171875, + "learning_rate": 0.00015374269005847955, + "loss": 3.8399, + "step": 5855 + }, + { + "epoch": 0.3082105263157895, + "grad_norm": 0.498046875, + "learning_rate": 0.0001537309941520468, + "loss": 4.0091, + "step": 5856 + }, + { + "epoch": 0.30826315789473685, + "grad_norm": 0.640625, + "learning_rate": 0.00015371929824561405, + "loss": 3.2902, + "step": 5857 + }, + { + "epoch": 0.3083157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.0001537076023391813, + "loss": 3.7755, + "step": 5858 + }, + { + "epoch": 0.30836842105263157, + "grad_norm": 0.62890625, + "learning_rate": 0.00015369590643274855, + "loss": 3.7942, + "step": 5859 + }, + { + "epoch": 0.30842105263157893, + "grad_norm": 0.50390625, + "learning_rate": 0.0001536842105263158, + "loss": 3.785, + "step": 5860 + }, + { + "epoch": 0.3084736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.00015367251461988305, + "loss": 3.8784, + "step": 5861 + }, + { + "epoch": 0.3085263157894737, + "grad_norm": 0.494140625, + "learning_rate": 0.0001536608187134503, + "loss": 3.8751, + "step": 5862 + }, + { + "epoch": 0.30857894736842106, + "grad_norm": 0.5703125, + "learning_rate": 0.00015364912280701754, + "loss": 3.6331, + "step": 5863 + }, + { + "epoch": 0.3086315789473684, + "grad_norm": 0.6328125, + "learning_rate": 0.0001536374269005848, + "loss": 3.4264, + "step": 5864 + }, + { + "epoch": 0.3086842105263158, + "grad_norm": 0.68359375, + "learning_rate": 0.00015362573099415204, + "loss": 3.6462, + "step": 5865 + }, + { + "epoch": 0.30873684210526314, + "grad_norm": 0.57421875, + "learning_rate": 0.0001536140350877193, + "loss": 3.6231, + "step": 5866 + }, + { + "epoch": 0.3087894736842105, + "grad_norm": 0.546875, + "learning_rate": 0.00015360233918128654, + "loss": 3.3534, + "step": 5867 + }, + { + "epoch": 0.3088421052631579, + "grad_norm": 0.494140625, + "learning_rate": 0.00015359064327485382, + "loss": 3.8722, + "step": 5868 + }, + { + "epoch": 0.3088947368421053, + "grad_norm": 0.54296875, + "learning_rate": 0.00015357894736842107, + "loss": 3.8086, + "step": 5869 + }, + { + "epoch": 0.30894736842105264, + "grad_norm": 0.51171875, + "learning_rate": 0.00015356725146198832, + "loss": 3.5352, + "step": 5870 + }, + { + "epoch": 0.309, + "grad_norm": 0.48046875, + "learning_rate": 0.00015355555555555557, + "loss": 4.1695, + "step": 5871 + }, + { + "epoch": 0.30905263157894736, + "grad_norm": 0.49609375, + "learning_rate": 0.00015354385964912282, + "loss": 3.4987, + "step": 5872 + }, + { + "epoch": 0.3091052631578947, + "grad_norm": 0.5078125, + "learning_rate": 0.00015353216374269007, + "loss": 4.0122, + "step": 5873 + }, + { + "epoch": 0.30915789473684213, + "grad_norm": 0.5078125, + "learning_rate": 0.00015352046783625732, + "loss": 3.994, + "step": 5874 + }, + { + "epoch": 0.3092105263157895, + "grad_norm": 0.490234375, + "learning_rate": 0.00015350877192982457, + "loss": 3.6416, + "step": 5875 + }, + { + "epoch": 0.30926315789473685, + "grad_norm": 0.5625, + "learning_rate": 0.00015349707602339182, + "loss": 3.3292, + "step": 5876 + }, + { + "epoch": 0.3093157894736842, + "grad_norm": 0.4765625, + "learning_rate": 0.00015348538011695907, + "loss": 4.1384, + "step": 5877 + }, + { + "epoch": 0.30936842105263157, + "grad_norm": 0.8515625, + "learning_rate": 0.00015347368421052632, + "loss": 3.9195, + "step": 5878 + }, + { + "epoch": 0.30942105263157893, + "grad_norm": 0.546875, + "learning_rate": 0.00015346198830409357, + "loss": 4.0331, + "step": 5879 + }, + { + "epoch": 0.3094736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.00015345029239766081, + "loss": 3.7584, + "step": 5880 + }, + { + "epoch": 0.3095263157894737, + "grad_norm": 0.55078125, + "learning_rate": 0.00015343859649122806, + "loss": 3.5864, + "step": 5881 + }, + { + "epoch": 0.30957894736842106, + "grad_norm": 0.515625, + "learning_rate": 0.00015342690058479531, + "loss": 3.8722, + "step": 5882 + }, + { + "epoch": 0.3096315789473684, + "grad_norm": 0.53125, + "learning_rate": 0.0001534152046783626, + "loss": 3.7367, + "step": 5883 + }, + { + "epoch": 0.3096842105263158, + "grad_norm": 0.53515625, + "learning_rate": 0.00015340350877192984, + "loss": 3.5628, + "step": 5884 + }, + { + "epoch": 0.30973684210526314, + "grad_norm": 0.59765625, + "learning_rate": 0.0001533918128654971, + "loss": 3.712, + "step": 5885 + }, + { + "epoch": 0.3097894736842105, + "grad_norm": 0.58203125, + "learning_rate": 0.00015338011695906434, + "loss": 4.1367, + "step": 5886 + }, + { + "epoch": 0.3098421052631579, + "grad_norm": 0.66015625, + "learning_rate": 0.0001533684210526316, + "loss": 3.4465, + "step": 5887 + }, + { + "epoch": 0.3098947368421053, + "grad_norm": 0.5625, + "learning_rate": 0.00015335672514619884, + "loss": 3.8703, + "step": 5888 + }, + { + "epoch": 0.30994736842105264, + "grad_norm": 0.53515625, + "learning_rate": 0.0001533450292397661, + "loss": 3.4772, + "step": 5889 + }, + { + "epoch": 0.31, + "grad_norm": 0.51953125, + "learning_rate": 0.00015333333333333334, + "loss": 3.7662, + "step": 5890 + }, + { + "epoch": 0.31005263157894736, + "grad_norm": 0.5625, + "learning_rate": 0.0001533216374269006, + "loss": 3.4667, + "step": 5891 + }, + { + "epoch": 0.3101052631578947, + "grad_norm": 0.546875, + "learning_rate": 0.00015330994152046784, + "loss": 3.769, + "step": 5892 + }, + { + "epoch": 0.31015789473684213, + "grad_norm": 0.4921875, + "learning_rate": 0.0001532982456140351, + "loss": 3.5992, + "step": 5893 + }, + { + "epoch": 0.3102105263157895, + "grad_norm": 0.5, + "learning_rate": 0.00015328654970760234, + "loss": 3.7033, + "step": 5894 + }, + { + "epoch": 0.31026315789473685, + "grad_norm": 0.49609375, + "learning_rate": 0.00015327485380116959, + "loss": 3.8535, + "step": 5895 + }, + { + "epoch": 0.3103157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00015326315789473684, + "loss": 3.5648, + "step": 5896 + }, + { + "epoch": 0.31036842105263157, + "grad_norm": 0.53515625, + "learning_rate": 0.00015325146198830409, + "loss": 3.7245, + "step": 5897 + }, + { + "epoch": 0.31042105263157893, + "grad_norm": 0.546875, + "learning_rate": 0.00015323976608187136, + "loss": 4.1161, + "step": 5898 + }, + { + "epoch": 0.3104736842105263, + "grad_norm": 0.5234375, + "learning_rate": 0.0001532280701754386, + "loss": 3.6636, + "step": 5899 + }, + { + "epoch": 0.3105263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00015321637426900586, + "loss": 4.1188, + "step": 5900 + }, + { + "epoch": 0.31057894736842107, + "grad_norm": 0.69140625, + "learning_rate": 0.0001532046783625731, + "loss": 3.885, + "step": 5901 + }, + { + "epoch": 0.3106315789473684, + "grad_norm": 0.57421875, + "learning_rate": 0.00015319298245614036, + "loss": 3.4765, + "step": 5902 + }, + { + "epoch": 0.3106842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.0001531812865497076, + "loss": 3.5147, + "step": 5903 + }, + { + "epoch": 0.31073684210526314, + "grad_norm": 0.51171875, + "learning_rate": 0.00015316959064327486, + "loss": 3.4798, + "step": 5904 + }, + { + "epoch": 0.3107894736842105, + "grad_norm": 0.486328125, + "learning_rate": 0.0001531578947368421, + "loss": 3.7205, + "step": 5905 + }, + { + "epoch": 0.3108421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.00015314619883040936, + "loss": 3.1777, + "step": 5906 + }, + { + "epoch": 0.3108947368421053, + "grad_norm": 0.5234375, + "learning_rate": 0.0001531345029239766, + "loss": 3.8597, + "step": 5907 + }, + { + "epoch": 0.31094736842105264, + "grad_norm": 0.5390625, + "learning_rate": 0.00015312280701754386, + "loss": 3.3183, + "step": 5908 + }, + { + "epoch": 0.311, + "grad_norm": 0.474609375, + "learning_rate": 0.0001531111111111111, + "loss": 4.0329, + "step": 5909 + }, + { + "epoch": 0.31105263157894736, + "grad_norm": 0.4609375, + "learning_rate": 0.00015309941520467836, + "loss": 3.661, + "step": 5910 + }, + { + "epoch": 0.3111052631578947, + "grad_norm": 0.451171875, + "learning_rate": 0.0001530877192982456, + "loss": 3.5359, + "step": 5911 + }, + { + "epoch": 0.3111578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.00015307602339181288, + "loss": 3.1965, + "step": 5912 + }, + { + "epoch": 0.3112105263157895, + "grad_norm": 0.703125, + "learning_rate": 0.00015306432748538013, + "loss": 4.0139, + "step": 5913 + }, + { + "epoch": 0.31126315789473685, + "grad_norm": 0.5390625, + "learning_rate": 0.00015305263157894738, + "loss": 3.9414, + "step": 5914 + }, + { + "epoch": 0.3113157894736842, + "grad_norm": 0.498046875, + "learning_rate": 0.00015304093567251463, + "loss": 3.8075, + "step": 5915 + }, + { + "epoch": 0.31136842105263157, + "grad_norm": 0.5078125, + "learning_rate": 0.00015302923976608188, + "loss": 3.6282, + "step": 5916 + }, + { + "epoch": 0.31142105263157893, + "grad_norm": 0.63671875, + "learning_rate": 0.00015301754385964913, + "loss": 3.586, + "step": 5917 + }, + { + "epoch": 0.3114736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00015300584795321638, + "loss": 3.6683, + "step": 5918 + }, + { + "epoch": 0.3115263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.00015299415204678363, + "loss": 4.0041, + "step": 5919 + }, + { + "epoch": 0.31157894736842107, + "grad_norm": 0.51171875, + "learning_rate": 0.0001529824561403509, + "loss": 3.906, + "step": 5920 + }, + { + "epoch": 0.3116315789473684, + "grad_norm": 0.5859375, + "learning_rate": 0.00015297076023391813, + "loss": 3.7501, + "step": 5921 + }, + { + "epoch": 0.3116842105263158, + "grad_norm": 0.494140625, + "learning_rate": 0.00015295906432748538, + "loss": 3.8896, + "step": 5922 + }, + { + "epoch": 0.31173684210526315, + "grad_norm": 0.51953125, + "learning_rate": 0.00015294736842105263, + "loss": 3.8833, + "step": 5923 + }, + { + "epoch": 0.3117894736842105, + "grad_norm": 0.5234375, + "learning_rate": 0.00015293567251461988, + "loss": 3.7953, + "step": 5924 + }, + { + "epoch": 0.3118421052631579, + "grad_norm": 0.5546875, + "learning_rate": 0.00015292397660818713, + "loss": 4.0197, + "step": 5925 + }, + { + "epoch": 0.3118947368421053, + "grad_norm": 0.58984375, + "learning_rate": 0.00015291228070175438, + "loss": 3.5755, + "step": 5926 + }, + { + "epoch": 0.31194736842105264, + "grad_norm": 0.6171875, + "learning_rate": 0.00015290058479532165, + "loss": 3.4685, + "step": 5927 + }, + { + "epoch": 0.312, + "grad_norm": 0.55859375, + "learning_rate": 0.0001528888888888889, + "loss": 4.0842, + "step": 5928 + }, + { + "epoch": 0.31205263157894736, + "grad_norm": 0.58203125, + "learning_rate": 0.00015287719298245615, + "loss": 3.605, + "step": 5929 + }, + { + "epoch": 0.3121052631578947, + "grad_norm": 0.5546875, + "learning_rate": 0.0001528654970760234, + "loss": 3.2555, + "step": 5930 + }, + { + "epoch": 0.3121578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.00015285380116959065, + "loss": 4.0158, + "step": 5931 + }, + { + "epoch": 0.3122105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.0001528421052631579, + "loss": 3.8861, + "step": 5932 + }, + { + "epoch": 0.31226315789473685, + "grad_norm": 0.51953125, + "learning_rate": 0.00015283040935672515, + "loss": 3.9408, + "step": 5933 + }, + { + "epoch": 0.3123157894736842, + "grad_norm": 0.546875, + "learning_rate": 0.0001528187134502924, + "loss": 3.6589, + "step": 5934 + }, + { + "epoch": 0.3123684210526316, + "grad_norm": 0.52734375, + "learning_rate": 0.00015280701754385968, + "loss": 3.464, + "step": 5935 + }, + { + "epoch": 0.31242105263157893, + "grad_norm": 0.578125, + "learning_rate": 0.0001527953216374269, + "loss": 3.5861, + "step": 5936 + }, + { + "epoch": 0.3124736842105263, + "grad_norm": 0.59375, + "learning_rate": 0.00015278362573099415, + "loss": 4.1138, + "step": 5937 + }, + { + "epoch": 0.3125263157894737, + "grad_norm": 0.59375, + "learning_rate": 0.0001527719298245614, + "loss": 3.943, + "step": 5938 + }, + { + "epoch": 0.31257894736842107, + "grad_norm": 0.60546875, + "learning_rate": 0.00015276023391812865, + "loss": 3.9817, + "step": 5939 + }, + { + "epoch": 0.3126315789473684, + "grad_norm": 0.578125, + "learning_rate": 0.0001527485380116959, + "loss": 3.4683, + "step": 5940 + }, + { + "epoch": 0.3126842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.00015273684210526315, + "loss": 4.0777, + "step": 5941 + }, + { + "epoch": 0.31273684210526315, + "grad_norm": 0.5234375, + "learning_rate": 0.00015272514619883043, + "loss": 3.7945, + "step": 5942 + }, + { + "epoch": 0.3127894736842105, + "grad_norm": 0.58203125, + "learning_rate": 0.00015271345029239767, + "loss": 3.4128, + "step": 5943 + }, + { + "epoch": 0.3128421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00015270175438596492, + "loss": 3.7763, + "step": 5944 + }, + { + "epoch": 0.3128947368421053, + "grad_norm": 0.5, + "learning_rate": 0.00015269005847953217, + "loss": 3.724, + "step": 5945 + }, + { + "epoch": 0.31294736842105264, + "grad_norm": 0.546875, + "learning_rate": 0.00015267836257309942, + "loss": 4.0072, + "step": 5946 + }, + { + "epoch": 0.313, + "grad_norm": 0.6015625, + "learning_rate": 0.00015266666666666667, + "loss": 3.9749, + "step": 5947 + }, + { + "epoch": 0.31305263157894736, + "grad_norm": 0.546875, + "learning_rate": 0.00015265497076023392, + "loss": 3.7876, + "step": 5948 + }, + { + "epoch": 0.3131052631578947, + "grad_norm": 0.55859375, + "learning_rate": 0.00015264327485380117, + "loss": 3.9425, + "step": 5949 + }, + { + "epoch": 0.3131578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00015263157894736845, + "loss": 3.6351, + "step": 5950 + }, + { + "epoch": 0.3132105263157895, + "grad_norm": 0.52734375, + "learning_rate": 0.00015261988304093567, + "loss": 3.6552, + "step": 5951 + }, + { + "epoch": 0.31326315789473685, + "grad_norm": 0.515625, + "learning_rate": 0.00015260818713450292, + "loss": 3.5478, + "step": 5952 + }, + { + "epoch": 0.3133157894736842, + "grad_norm": 0.55078125, + "learning_rate": 0.00015259649122807017, + "loss": 3.661, + "step": 5953 + }, + { + "epoch": 0.3133684210526316, + "grad_norm": 1.1484375, + "learning_rate": 0.00015258479532163742, + "loss": 3.592, + "step": 5954 + }, + { + "epoch": 0.31342105263157893, + "grad_norm": 0.50390625, + "learning_rate": 0.00015257309941520467, + "loss": 3.6668, + "step": 5955 + }, + { + "epoch": 0.3134736842105263, + "grad_norm": 0.474609375, + "learning_rate": 0.00015256140350877192, + "loss": 3.557, + "step": 5956 + }, + { + "epoch": 0.3135263157894737, + "grad_norm": 0.57421875, + "learning_rate": 0.0001525497076023392, + "loss": 3.3907, + "step": 5957 + }, + { + "epoch": 0.31357894736842107, + "grad_norm": 0.51953125, + "learning_rate": 0.00015253801169590645, + "loss": 3.873, + "step": 5958 + }, + { + "epoch": 0.31363157894736843, + "grad_norm": 0.55078125, + "learning_rate": 0.0001525263157894737, + "loss": 3.7613, + "step": 5959 + }, + { + "epoch": 0.3136842105263158, + "grad_norm": 0.4921875, + "learning_rate": 0.00015251461988304094, + "loss": 3.5085, + "step": 5960 + }, + { + "epoch": 0.31373684210526315, + "grad_norm": 0.4765625, + "learning_rate": 0.0001525029239766082, + "loss": 3.8855, + "step": 5961 + }, + { + "epoch": 0.3137894736842105, + "grad_norm": 0.5859375, + "learning_rate": 0.00015249122807017544, + "loss": 3.5369, + "step": 5962 + }, + { + "epoch": 0.3138421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.0001524795321637427, + "loss": 3.3329, + "step": 5963 + }, + { + "epoch": 0.3138947368421053, + "grad_norm": 0.49609375, + "learning_rate": 0.00015246783625730997, + "loss": 3.6118, + "step": 5964 + }, + { + "epoch": 0.31394736842105264, + "grad_norm": 0.5, + "learning_rate": 0.00015245614035087722, + "loss": 3.7343, + "step": 5965 + }, + { + "epoch": 0.314, + "grad_norm": 0.4765625, + "learning_rate": 0.00015244444444444447, + "loss": 3.8578, + "step": 5966 + }, + { + "epoch": 0.31405263157894736, + "grad_norm": 0.50390625, + "learning_rate": 0.0001524327485380117, + "loss": 3.7786, + "step": 5967 + }, + { + "epoch": 0.3141052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.00015242105263157894, + "loss": 3.6631, + "step": 5968 + }, + { + "epoch": 0.3141578947368421, + "grad_norm": 0.5703125, + "learning_rate": 0.0001524093567251462, + "loss": 3.5946, + "step": 5969 + }, + { + "epoch": 0.3142105263157895, + "grad_norm": 0.498046875, + "learning_rate": 0.00015239766081871344, + "loss": 3.6495, + "step": 5970 + }, + { + "epoch": 0.31426315789473686, + "grad_norm": 0.59765625, + "learning_rate": 0.0001523859649122807, + "loss": 3.9018, + "step": 5971 + }, + { + "epoch": 0.3143157894736842, + "grad_norm": 0.486328125, + "learning_rate": 0.00015237426900584797, + "loss": 3.4806, + "step": 5972 + }, + { + "epoch": 0.3143684210526316, + "grad_norm": 0.48046875, + "learning_rate": 0.00015236257309941522, + "loss": 3.6649, + "step": 5973 + }, + { + "epoch": 0.31442105263157893, + "grad_norm": 0.5390625, + "learning_rate": 0.00015235087719298247, + "loss": 4.0086, + "step": 5974 + }, + { + "epoch": 0.3144736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.00015233918128654972, + "loss": 4.1097, + "step": 5975 + }, + { + "epoch": 0.3145263157894737, + "grad_norm": 0.5234375, + "learning_rate": 0.00015232748538011697, + "loss": 3.684, + "step": 5976 + }, + { + "epoch": 0.31457894736842107, + "grad_norm": 0.515625, + "learning_rate": 0.00015231578947368422, + "loss": 3.6696, + "step": 5977 + }, + { + "epoch": 0.31463157894736843, + "grad_norm": 0.515625, + "learning_rate": 0.00015230409356725146, + "loss": 3.9816, + "step": 5978 + }, + { + "epoch": 0.3146842105263158, + "grad_norm": 0.52734375, + "learning_rate": 0.00015229239766081874, + "loss": 3.4711, + "step": 5979 + }, + { + "epoch": 0.31473684210526315, + "grad_norm": 0.546875, + "learning_rate": 0.000152280701754386, + "loss": 3.7137, + "step": 5980 + }, + { + "epoch": 0.3147894736842105, + "grad_norm": 0.59375, + "learning_rate": 0.00015226900584795324, + "loss": 3.7626, + "step": 5981 + }, + { + "epoch": 0.31484210526315787, + "grad_norm": 0.57421875, + "learning_rate": 0.00015225730994152046, + "loss": 3.1894, + "step": 5982 + }, + { + "epoch": 0.3148947368421053, + "grad_norm": 0.44921875, + "learning_rate": 0.0001522456140350877, + "loss": 3.8481, + "step": 5983 + }, + { + "epoch": 0.31494736842105264, + "grad_norm": 0.57421875, + "learning_rate": 0.00015223391812865496, + "loss": 3.4019, + "step": 5984 + }, + { + "epoch": 0.315, + "grad_norm": 0.515625, + "learning_rate": 0.0001522222222222222, + "loss": 3.4968, + "step": 5985 + }, + { + "epoch": 0.31505263157894736, + "grad_norm": 0.5078125, + "learning_rate": 0.00015221052631578946, + "loss": 3.6046, + "step": 5986 + }, + { + "epoch": 0.3151052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.00015219883040935674, + "loss": 3.5364, + "step": 5987 + }, + { + "epoch": 0.3151578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.000152187134502924, + "loss": 3.967, + "step": 5988 + }, + { + "epoch": 0.3152105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00015217543859649124, + "loss": 4.0591, + "step": 5989 + }, + { + "epoch": 0.31526315789473686, + "grad_norm": 0.55078125, + "learning_rate": 0.0001521637426900585, + "loss": 4.0964, + "step": 5990 + }, + { + "epoch": 0.3153157894736842, + "grad_norm": 0.6484375, + "learning_rate": 0.00015215204678362574, + "loss": 3.4878, + "step": 5991 + }, + { + "epoch": 0.3153684210526316, + "grad_norm": 0.478515625, + "learning_rate": 0.00015214035087719299, + "loss": 3.9921, + "step": 5992 + }, + { + "epoch": 0.31542105263157894, + "grad_norm": 0.65625, + "learning_rate": 0.00015212865497076024, + "loss": 3.7158, + "step": 5993 + }, + { + "epoch": 0.3154736842105263, + "grad_norm": 0.484375, + "learning_rate": 0.0001521169590643275, + "loss": 3.6852, + "step": 5994 + }, + { + "epoch": 0.3155263157894737, + "grad_norm": 0.546875, + "learning_rate": 0.00015210526315789476, + "loss": 3.2012, + "step": 5995 + }, + { + "epoch": 0.31557894736842107, + "grad_norm": 0.515625, + "learning_rate": 0.000152093567251462, + "loss": 3.7255, + "step": 5996 + }, + { + "epoch": 0.31563157894736843, + "grad_norm": 0.486328125, + "learning_rate": 0.00015208187134502923, + "loss": 3.4094, + "step": 5997 + }, + { + "epoch": 0.3156842105263158, + "grad_norm": 0.5703125, + "learning_rate": 0.00015207017543859648, + "loss": 3.3846, + "step": 5998 + }, + { + "epoch": 0.31573684210526315, + "grad_norm": 0.51171875, + "learning_rate": 0.00015205847953216373, + "loss": 3.7821, + "step": 5999 + }, + { + "epoch": 0.3157894736842105, + "grad_norm": 0.58984375, + "learning_rate": 0.00015204678362573098, + "loss": 3.6523, + "step": 6000 + }, + { + "epoch": 0.31584210526315787, + "grad_norm": 0.53515625, + "learning_rate": 0.00015203508771929826, + "loss": 3.4299, + "step": 6001 + }, + { + "epoch": 0.3158947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.0001520233918128655, + "loss": 3.9747, + "step": 6002 + }, + { + "epoch": 0.31594736842105264, + "grad_norm": 0.490234375, + "learning_rate": 0.00015201169590643276, + "loss": 3.9299, + "step": 6003 + }, + { + "epoch": 0.316, + "grad_norm": 0.54296875, + "learning_rate": 0.000152, + "loss": 3.5097, + "step": 6004 + }, + { + "epoch": 0.31605263157894736, + "grad_norm": 0.49609375, + "learning_rate": 0.00015198830409356726, + "loss": 4.0023, + "step": 6005 + }, + { + "epoch": 0.3161052631578947, + "grad_norm": 0.59765625, + "learning_rate": 0.0001519766081871345, + "loss": 3.9497, + "step": 6006 + }, + { + "epoch": 0.3161578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00015196491228070176, + "loss": 3.9634, + "step": 6007 + }, + { + "epoch": 0.3162105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.000151953216374269, + "loss": 3.5929, + "step": 6008 + }, + { + "epoch": 0.31626315789473686, + "grad_norm": 0.515625, + "learning_rate": 0.00015194152046783628, + "loss": 3.6713, + "step": 6009 + }, + { + "epoch": 0.3163157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00015192982456140353, + "loss": 3.7246, + "step": 6010 + }, + { + "epoch": 0.3163684210526316, + "grad_norm": 0.494140625, + "learning_rate": 0.00015191812865497078, + "loss": 3.9159, + "step": 6011 + }, + { + "epoch": 0.31642105263157894, + "grad_norm": 0.5390625, + "learning_rate": 0.00015190643274853803, + "loss": 3.5476, + "step": 6012 + }, + { + "epoch": 0.3164736842105263, + "grad_norm": 0.51953125, + "learning_rate": 0.00015189473684210525, + "loss": 3.4928, + "step": 6013 + }, + { + "epoch": 0.3165263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.0001518830409356725, + "loss": 3.5956, + "step": 6014 + }, + { + "epoch": 0.31657894736842107, + "grad_norm": 0.5859375, + "learning_rate": 0.00015187134502923975, + "loss": 3.8693, + "step": 6015 + }, + { + "epoch": 0.31663157894736843, + "grad_norm": 0.609375, + "learning_rate": 0.00015185964912280703, + "loss": 3.8102, + "step": 6016 + }, + { + "epoch": 0.3166842105263158, + "grad_norm": 0.498046875, + "learning_rate": 0.00015184795321637428, + "loss": 4.001, + "step": 6017 + }, + { + "epoch": 0.31673684210526315, + "grad_norm": 0.58203125, + "learning_rate": 0.00015183625730994153, + "loss": 3.7229, + "step": 6018 + }, + { + "epoch": 0.3167894736842105, + "grad_norm": 0.546875, + "learning_rate": 0.00015182456140350878, + "loss": 3.7834, + "step": 6019 + }, + { + "epoch": 0.31684210526315787, + "grad_norm": 0.578125, + "learning_rate": 0.00015181286549707603, + "loss": 4.0976, + "step": 6020 + }, + { + "epoch": 0.3168947368421053, + "grad_norm": 0.59765625, + "learning_rate": 0.00015180116959064328, + "loss": 3.8916, + "step": 6021 + }, + { + "epoch": 0.31694736842105264, + "grad_norm": 0.53515625, + "learning_rate": 0.00015178947368421053, + "loss": 3.8866, + "step": 6022 + }, + { + "epoch": 0.317, + "grad_norm": 0.80859375, + "learning_rate": 0.00015177777777777778, + "loss": 3.8245, + "step": 6023 + }, + { + "epoch": 0.31705263157894736, + "grad_norm": 0.53125, + "learning_rate": 0.00015176608187134505, + "loss": 3.5827, + "step": 6024 + }, + { + "epoch": 0.3171052631578947, + "grad_norm": 0.53515625, + "learning_rate": 0.0001517543859649123, + "loss": 3.7912, + "step": 6025 + }, + { + "epoch": 0.3171578947368421, + "grad_norm": 0.55078125, + "learning_rate": 0.00015174269005847955, + "loss": 3.6218, + "step": 6026 + }, + { + "epoch": 0.3172105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.0001517309941520468, + "loss": 3.3337, + "step": 6027 + }, + { + "epoch": 0.31726315789473686, + "grad_norm": 0.498046875, + "learning_rate": 0.00015171929824561403, + "loss": 3.1939, + "step": 6028 + }, + { + "epoch": 0.3173157894736842, + "grad_norm": 0.84765625, + "learning_rate": 0.00015170760233918128, + "loss": 3.7295, + "step": 6029 + }, + { + "epoch": 0.3173684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00015169590643274852, + "loss": 3.7995, + "step": 6030 + }, + { + "epoch": 0.31742105263157894, + "grad_norm": 0.5625, + "learning_rate": 0.0001516842105263158, + "loss": 3.6689, + "step": 6031 + }, + { + "epoch": 0.3174736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00015167251461988305, + "loss": 3.921, + "step": 6032 + }, + { + "epoch": 0.31752631578947366, + "grad_norm": 0.67578125, + "learning_rate": 0.0001516608187134503, + "loss": 3.3631, + "step": 6033 + }, + { + "epoch": 0.31757894736842107, + "grad_norm": 0.9140625, + "learning_rate": 0.00015164912280701755, + "loss": 3.8293, + "step": 6034 + }, + { + "epoch": 0.31763157894736843, + "grad_norm": 0.51171875, + "learning_rate": 0.0001516374269005848, + "loss": 3.61, + "step": 6035 + }, + { + "epoch": 0.3176842105263158, + "grad_norm": 0.466796875, + "learning_rate": 0.00015162573099415205, + "loss": 3.9374, + "step": 6036 + }, + { + "epoch": 0.31773684210526315, + "grad_norm": 0.51953125, + "learning_rate": 0.0001516140350877193, + "loss": 3.1688, + "step": 6037 + }, + { + "epoch": 0.3177894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.00015160233918128658, + "loss": 3.3784, + "step": 6038 + }, + { + "epoch": 0.31784210526315787, + "grad_norm": 0.48828125, + "learning_rate": 0.00015159064327485383, + "loss": 3.6918, + "step": 6039 + }, + { + "epoch": 0.3178947368421053, + "grad_norm": 0.4921875, + "learning_rate": 0.00015157894736842108, + "loss": 3.454, + "step": 6040 + }, + { + "epoch": 0.31794736842105265, + "grad_norm": 0.5390625, + "learning_rate": 0.00015156725146198832, + "loss": 3.3399, + "step": 6041 + }, + { + "epoch": 0.318, + "grad_norm": 0.51953125, + "learning_rate": 0.00015155555555555557, + "loss": 3.822, + "step": 6042 + }, + { + "epoch": 0.31805263157894736, + "grad_norm": 0.50390625, + "learning_rate": 0.0001515438596491228, + "loss": 3.6407, + "step": 6043 + }, + { + "epoch": 0.3181052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00015153216374269005, + "loss": 3.5898, + "step": 6044 + }, + { + "epoch": 0.3181578947368421, + "grad_norm": 1.3359375, + "learning_rate": 0.0001515204678362573, + "loss": 3.8945, + "step": 6045 + }, + { + "epoch": 0.3182105263157895, + "grad_norm": 0.52734375, + "learning_rate": 0.00015150877192982457, + "loss": 3.4802, + "step": 6046 + }, + { + "epoch": 0.31826315789473686, + "grad_norm": 0.58984375, + "learning_rate": 0.00015149707602339182, + "loss": 3.7733, + "step": 6047 + }, + { + "epoch": 0.3183157894736842, + "grad_norm": 0.5703125, + "learning_rate": 0.00015148538011695907, + "loss": 4.0472, + "step": 6048 + }, + { + "epoch": 0.3183684210526316, + "grad_norm": 0.69921875, + "learning_rate": 0.00015147368421052632, + "loss": 3.6272, + "step": 6049 + }, + { + "epoch": 0.31842105263157894, + "grad_norm": 0.4921875, + "learning_rate": 0.00015146198830409357, + "loss": 3.9619, + "step": 6050 + }, + { + "epoch": 0.3184736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00015145029239766082, + "loss": 3.9842, + "step": 6051 + }, + { + "epoch": 0.31852631578947366, + "grad_norm": 0.5625, + "learning_rate": 0.00015143859649122807, + "loss": 3.6375, + "step": 6052 + }, + { + "epoch": 0.3185789473684211, + "grad_norm": 0.546875, + "learning_rate": 0.00015142690058479535, + "loss": 3.7971, + "step": 6053 + }, + { + "epoch": 0.31863157894736843, + "grad_norm": 0.494140625, + "learning_rate": 0.0001514152046783626, + "loss": 3.2666, + "step": 6054 + }, + { + "epoch": 0.3186842105263158, + "grad_norm": 0.546875, + "learning_rate": 0.00015140350877192985, + "loss": 3.6187, + "step": 6055 + }, + { + "epoch": 0.31873684210526315, + "grad_norm": 0.5703125, + "learning_rate": 0.0001513918128654971, + "loss": 3.4919, + "step": 6056 + }, + { + "epoch": 0.3187894736842105, + "grad_norm": 0.486328125, + "learning_rate": 0.00015138011695906435, + "loss": 3.5226, + "step": 6057 + }, + { + "epoch": 0.31884210526315787, + "grad_norm": 0.498046875, + "learning_rate": 0.0001513684210526316, + "loss": 3.4968, + "step": 6058 + }, + { + "epoch": 0.3188947368421053, + "grad_norm": 0.69921875, + "learning_rate": 0.00015135672514619882, + "loss": 3.6428, + "step": 6059 + }, + { + "epoch": 0.31894736842105265, + "grad_norm": 0.55078125, + "learning_rate": 0.00015134502923976607, + "loss": 3.3025, + "step": 6060 + }, + { + "epoch": 0.319, + "grad_norm": 0.5, + "learning_rate": 0.00015133333333333334, + "loss": 3.4131, + "step": 6061 + }, + { + "epoch": 0.31905263157894737, + "grad_norm": 0.609375, + "learning_rate": 0.0001513216374269006, + "loss": 3.8958, + "step": 6062 + }, + { + "epoch": 0.3191052631578947, + "grad_norm": 0.466796875, + "learning_rate": 0.00015130994152046784, + "loss": 3.3234, + "step": 6063 + }, + { + "epoch": 0.3191578947368421, + "grad_norm": 0.482421875, + "learning_rate": 0.0001512982456140351, + "loss": 3.7486, + "step": 6064 + }, + { + "epoch": 0.3192105263157895, + "grad_norm": 0.470703125, + "learning_rate": 0.00015128654970760234, + "loss": 3.4172, + "step": 6065 + }, + { + "epoch": 0.31926315789473686, + "grad_norm": 0.5078125, + "learning_rate": 0.0001512748538011696, + "loss": 3.4422, + "step": 6066 + }, + { + "epoch": 0.3193157894736842, + "grad_norm": 0.494140625, + "learning_rate": 0.00015126315789473684, + "loss": 3.7725, + "step": 6067 + }, + { + "epoch": 0.3193684210526316, + "grad_norm": 0.6015625, + "learning_rate": 0.00015125146198830412, + "loss": 3.3249, + "step": 6068 + }, + { + "epoch": 0.31942105263157894, + "grad_norm": 0.498046875, + "learning_rate": 0.00015123976608187137, + "loss": 3.5247, + "step": 6069 + }, + { + "epoch": 0.3194736842105263, + "grad_norm": 0.4921875, + "learning_rate": 0.00015122807017543862, + "loss": 3.5833, + "step": 6070 + }, + { + "epoch": 0.31952631578947366, + "grad_norm": 0.498046875, + "learning_rate": 0.00015121637426900587, + "loss": 3.8694, + "step": 6071 + }, + { + "epoch": 0.3195789473684211, + "grad_norm": 0.6015625, + "learning_rate": 0.00015120467836257312, + "loss": 3.5798, + "step": 6072 + }, + { + "epoch": 0.31963157894736843, + "grad_norm": 0.52734375, + "learning_rate": 0.00015119298245614037, + "loss": 3.6082, + "step": 6073 + }, + { + "epoch": 0.3196842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.0001511812865497076, + "loss": 3.9032, + "step": 6074 + }, + { + "epoch": 0.31973684210526315, + "grad_norm": 0.55859375, + "learning_rate": 0.00015116959064327486, + "loss": 3.302, + "step": 6075 + }, + { + "epoch": 0.3197894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.00015115789473684211, + "loss": 3.9682, + "step": 6076 + }, + { + "epoch": 0.31984210526315787, + "grad_norm": 0.578125, + "learning_rate": 0.00015114619883040936, + "loss": 3.9918, + "step": 6077 + }, + { + "epoch": 0.3198947368421053, + "grad_norm": 0.54296875, + "learning_rate": 0.00015113450292397661, + "loss": 3.8124, + "step": 6078 + }, + { + "epoch": 0.31994736842105265, + "grad_norm": 0.484375, + "learning_rate": 0.00015112280701754386, + "loss": 3.4277, + "step": 6079 + }, + { + "epoch": 0.32, + "grad_norm": 0.458984375, + "learning_rate": 0.0001511111111111111, + "loss": 4.2428, + "step": 6080 + }, + { + "epoch": 0.32005263157894737, + "grad_norm": 0.59375, + "learning_rate": 0.00015109941520467836, + "loss": 3.7765, + "step": 6081 + }, + { + "epoch": 0.3201052631578947, + "grad_norm": 0.5703125, + "learning_rate": 0.0001510877192982456, + "loss": 3.4557, + "step": 6082 + }, + { + "epoch": 0.3201578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.0001510760233918129, + "loss": 3.512, + "step": 6083 + }, + { + "epoch": 0.3202105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.00015106432748538014, + "loss": 3.8455, + "step": 6084 + }, + { + "epoch": 0.32026315789473686, + "grad_norm": 0.56640625, + "learning_rate": 0.0001510526315789474, + "loss": 3.4575, + "step": 6085 + }, + { + "epoch": 0.3203157894736842, + "grad_norm": 1.0234375, + "learning_rate": 0.00015104093567251464, + "loss": 3.4753, + "step": 6086 + }, + { + "epoch": 0.3203684210526316, + "grad_norm": 0.48828125, + "learning_rate": 0.0001510292397660819, + "loss": 3.9084, + "step": 6087 + }, + { + "epoch": 0.32042105263157894, + "grad_norm": 0.55078125, + "learning_rate": 0.00015101754385964914, + "loss": 3.6149, + "step": 6088 + }, + { + "epoch": 0.3204736842105263, + "grad_norm": 0.609375, + "learning_rate": 0.00015100584795321636, + "loss": 3.709, + "step": 6089 + }, + { + "epoch": 0.32052631578947366, + "grad_norm": 0.5546875, + "learning_rate": 0.00015099415204678364, + "loss": 3.1066, + "step": 6090 + }, + { + "epoch": 0.3205789473684211, + "grad_norm": 0.50390625, + "learning_rate": 0.00015098245614035089, + "loss": 3.3497, + "step": 6091 + }, + { + "epoch": 0.32063157894736843, + "grad_norm": 0.54296875, + "learning_rate": 0.00015097076023391814, + "loss": 3.2282, + "step": 6092 + }, + { + "epoch": 0.3206842105263158, + "grad_norm": 0.44140625, + "learning_rate": 0.00015095906432748538, + "loss": 3.6426, + "step": 6093 + }, + { + "epoch": 0.32073684210526315, + "grad_norm": 0.5390625, + "learning_rate": 0.00015094736842105263, + "loss": 3.7814, + "step": 6094 + }, + { + "epoch": 0.3207894736842105, + "grad_norm": 0.55859375, + "learning_rate": 0.00015093567251461988, + "loss": 3.1162, + "step": 6095 + }, + { + "epoch": 0.3208421052631579, + "grad_norm": 0.58203125, + "learning_rate": 0.00015092397660818713, + "loss": 3.7301, + "step": 6096 + }, + { + "epoch": 0.3208947368421053, + "grad_norm": 0.67578125, + "learning_rate": 0.00015091228070175438, + "loss": 3.8572, + "step": 6097 + }, + { + "epoch": 0.32094736842105265, + "grad_norm": 0.54296875, + "learning_rate": 0.00015090058479532166, + "loss": 3.2747, + "step": 6098 + }, + { + "epoch": 0.321, + "grad_norm": 0.47265625, + "learning_rate": 0.0001508888888888889, + "loss": 3.861, + "step": 6099 + }, + { + "epoch": 0.32105263157894737, + "grad_norm": 0.482421875, + "learning_rate": 0.00015087719298245616, + "loss": 4.1602, + "step": 6100 + }, + { + "epoch": 0.3211052631578947, + "grad_norm": 0.5625, + "learning_rate": 0.0001508654970760234, + "loss": 3.9747, + "step": 6101 + }, + { + "epoch": 0.3211578947368421, + "grad_norm": 0.57421875, + "learning_rate": 0.00015085380116959066, + "loss": 3.6606, + "step": 6102 + }, + { + "epoch": 0.32121052631578945, + "grad_norm": 0.53125, + "learning_rate": 0.0001508421052631579, + "loss": 4.0715, + "step": 6103 + }, + { + "epoch": 0.32126315789473686, + "grad_norm": 0.5703125, + "learning_rate": 0.00015083040935672516, + "loss": 3.957, + "step": 6104 + }, + { + "epoch": 0.3213157894736842, + "grad_norm": 0.609375, + "learning_rate": 0.0001508187134502924, + "loss": 3.2177, + "step": 6105 + }, + { + "epoch": 0.3213684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00015080701754385966, + "loss": 3.5857, + "step": 6106 + }, + { + "epoch": 0.32142105263157894, + "grad_norm": 0.498046875, + "learning_rate": 0.0001507953216374269, + "loss": 3.6249, + "step": 6107 + }, + { + "epoch": 0.3214736842105263, + "grad_norm": 0.61328125, + "learning_rate": 0.00015078362573099416, + "loss": 3.4168, + "step": 6108 + }, + { + "epoch": 0.32152631578947366, + "grad_norm": 0.490234375, + "learning_rate": 0.0001507719298245614, + "loss": 3.6375, + "step": 6109 + }, + { + "epoch": 0.3215789473684211, + "grad_norm": 0.62109375, + "learning_rate": 0.00015076023391812865, + "loss": 3.1888, + "step": 6110 + }, + { + "epoch": 0.32163157894736844, + "grad_norm": 0.5, + "learning_rate": 0.0001507485380116959, + "loss": 3.9537, + "step": 6111 + }, + { + "epoch": 0.3216842105263158, + "grad_norm": 0.54296875, + "learning_rate": 0.00015073684210526315, + "loss": 3.6033, + "step": 6112 + }, + { + "epoch": 0.32173684210526315, + "grad_norm": 0.5859375, + "learning_rate": 0.00015072514619883043, + "loss": 3.7516, + "step": 6113 + }, + { + "epoch": 0.3217894736842105, + "grad_norm": 0.53125, + "learning_rate": 0.00015071345029239768, + "loss": 3.5725, + "step": 6114 + }, + { + "epoch": 0.3218421052631579, + "grad_norm": 0.55078125, + "learning_rate": 0.00015070175438596493, + "loss": 3.4357, + "step": 6115 + }, + { + "epoch": 0.3218947368421053, + "grad_norm": 0.46484375, + "learning_rate": 0.00015069005847953218, + "loss": 3.901, + "step": 6116 + }, + { + "epoch": 0.32194736842105265, + "grad_norm": 0.484375, + "learning_rate": 0.00015067836257309943, + "loss": 3.9312, + "step": 6117 + }, + { + "epoch": 0.322, + "grad_norm": 0.5, + "learning_rate": 0.00015066666666666668, + "loss": 3.7392, + "step": 6118 + }, + { + "epoch": 0.32205263157894737, + "grad_norm": 0.49609375, + "learning_rate": 0.00015065497076023393, + "loss": 3.7966, + "step": 6119 + }, + { + "epoch": 0.32210526315789473, + "grad_norm": 0.53125, + "learning_rate": 0.00015064327485380118, + "loss": 3.4644, + "step": 6120 + }, + { + "epoch": 0.3221578947368421, + "grad_norm": 0.47265625, + "learning_rate": 0.00015063157894736843, + "loss": 3.3281, + "step": 6121 + }, + { + "epoch": 0.32221052631578945, + "grad_norm": 0.61328125, + "learning_rate": 0.00015061988304093568, + "loss": 3.4138, + "step": 6122 + }, + { + "epoch": 0.32226315789473686, + "grad_norm": 0.4765625, + "learning_rate": 0.00015060818713450293, + "loss": 3.3499, + "step": 6123 + }, + { + "epoch": 0.3223157894736842, + "grad_norm": 0.498046875, + "learning_rate": 0.00015059649122807018, + "loss": 3.5158, + "step": 6124 + }, + { + "epoch": 0.3223684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00015058479532163743, + "loss": 3.6907, + "step": 6125 + }, + { + "epoch": 0.32242105263157894, + "grad_norm": 0.57421875, + "learning_rate": 0.00015057309941520468, + "loss": 3.2629, + "step": 6126 + }, + { + "epoch": 0.3224736842105263, + "grad_norm": 0.58984375, + "learning_rate": 0.00015056140350877195, + "loss": 3.7492, + "step": 6127 + }, + { + "epoch": 0.32252631578947366, + "grad_norm": 0.6171875, + "learning_rate": 0.0001505497076023392, + "loss": 3.5749, + "step": 6128 + }, + { + "epoch": 0.3225789473684211, + "grad_norm": 0.50390625, + "learning_rate": 0.00015053801169590645, + "loss": 3.7854, + "step": 6129 + }, + { + "epoch": 0.32263157894736844, + "grad_norm": 0.4609375, + "learning_rate": 0.0001505263157894737, + "loss": 3.5813, + "step": 6130 + }, + { + "epoch": 0.3226842105263158, + "grad_norm": 0.58203125, + "learning_rate": 0.00015051461988304095, + "loss": 3.4458, + "step": 6131 + }, + { + "epoch": 0.32273684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.0001505029239766082, + "loss": 3.4444, + "step": 6132 + }, + { + "epoch": 0.3227894736842105, + "grad_norm": 0.58203125, + "learning_rate": 0.00015049122807017545, + "loss": 4.121, + "step": 6133 + }, + { + "epoch": 0.3228421052631579, + "grad_norm": 0.68359375, + "learning_rate": 0.0001504795321637427, + "loss": 3.1003, + "step": 6134 + }, + { + "epoch": 0.3228947368421053, + "grad_norm": 0.49609375, + "learning_rate": 0.00015046783625730995, + "loss": 3.9033, + "step": 6135 + }, + { + "epoch": 0.32294736842105265, + "grad_norm": 0.478515625, + "learning_rate": 0.0001504561403508772, + "loss": 3.6514, + "step": 6136 + }, + { + "epoch": 0.323, + "grad_norm": 0.5234375, + "learning_rate": 0.00015044444444444445, + "loss": 3.4685, + "step": 6137 + }, + { + "epoch": 0.32305263157894737, + "grad_norm": 0.5546875, + "learning_rate": 0.0001504327485380117, + "loss": 3.7041, + "step": 6138 + }, + { + "epoch": 0.32310526315789473, + "grad_norm": 0.5546875, + "learning_rate": 0.00015042105263157895, + "loss": 3.4764, + "step": 6139 + }, + { + "epoch": 0.3231578947368421, + "grad_norm": 0.47265625, + "learning_rate": 0.0001504093567251462, + "loss": 3.8526, + "step": 6140 + }, + { + "epoch": 0.32321052631578945, + "grad_norm": 0.474609375, + "learning_rate": 0.00015039766081871345, + "loss": 3.8299, + "step": 6141 + }, + { + "epoch": 0.32326315789473686, + "grad_norm": 0.78515625, + "learning_rate": 0.00015038596491228072, + "loss": 3.2968, + "step": 6142 + }, + { + "epoch": 0.3233157894736842, + "grad_norm": 0.4609375, + "learning_rate": 0.00015037426900584797, + "loss": 3.8869, + "step": 6143 + }, + { + "epoch": 0.3233684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00015036257309941522, + "loss": 3.6942, + "step": 6144 + }, + { + "epoch": 0.32342105263157894, + "grad_norm": 0.5078125, + "learning_rate": 0.00015035087719298247, + "loss": 3.4491, + "step": 6145 + }, + { + "epoch": 0.3234736842105263, + "grad_norm": 0.5, + "learning_rate": 0.00015033918128654972, + "loss": 4.0715, + "step": 6146 + }, + { + "epoch": 0.32352631578947366, + "grad_norm": 0.45703125, + "learning_rate": 0.00015032748538011697, + "loss": 3.5623, + "step": 6147 + }, + { + "epoch": 0.3235789473684211, + "grad_norm": 0.546875, + "learning_rate": 0.00015031578947368422, + "loss": 3.6296, + "step": 6148 + }, + { + "epoch": 0.32363157894736844, + "grad_norm": 0.5234375, + "learning_rate": 0.00015030409356725147, + "loss": 3.8841, + "step": 6149 + }, + { + "epoch": 0.3236842105263158, + "grad_norm": 0.51953125, + "learning_rate": 0.00015029239766081872, + "loss": 3.6844, + "step": 6150 + }, + { + "epoch": 0.32373684210526316, + "grad_norm": 0.4765625, + "learning_rate": 0.00015028070175438597, + "loss": 3.6618, + "step": 6151 + }, + { + "epoch": 0.3237894736842105, + "grad_norm": 0.5, + "learning_rate": 0.00015026900584795322, + "loss": 3.413, + "step": 6152 + }, + { + "epoch": 0.3238421052631579, + "grad_norm": 0.49609375, + "learning_rate": 0.00015025730994152047, + "loss": 3.9375, + "step": 6153 + }, + { + "epoch": 0.3238947368421053, + "grad_norm": 0.58203125, + "learning_rate": 0.00015024561403508772, + "loss": 3.8435, + "step": 6154 + }, + { + "epoch": 0.32394736842105265, + "grad_norm": 0.4921875, + "learning_rate": 0.00015023391812865497, + "loss": 3.7272, + "step": 6155 + }, + { + "epoch": 0.324, + "grad_norm": 0.59375, + "learning_rate": 0.00015022222222222222, + "loss": 3.519, + "step": 6156 + }, + { + "epoch": 0.32405263157894737, + "grad_norm": 0.515625, + "learning_rate": 0.0001502105263157895, + "loss": 3.5612, + "step": 6157 + }, + { + "epoch": 0.32410526315789473, + "grad_norm": 0.494140625, + "learning_rate": 0.00015019883040935674, + "loss": 3.7149, + "step": 6158 + }, + { + "epoch": 0.3241578947368421, + "grad_norm": 0.490234375, + "learning_rate": 0.000150187134502924, + "loss": 3.6118, + "step": 6159 + }, + { + "epoch": 0.32421052631578945, + "grad_norm": 0.5234375, + "learning_rate": 0.00015017543859649124, + "loss": 3.5292, + "step": 6160 + }, + { + "epoch": 0.32426315789473686, + "grad_norm": 0.56640625, + "learning_rate": 0.0001501637426900585, + "loss": 3.656, + "step": 6161 + }, + { + "epoch": 0.3243157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.00015015204678362574, + "loss": 3.9734, + "step": 6162 + }, + { + "epoch": 0.3243684210526316, + "grad_norm": 0.56640625, + "learning_rate": 0.000150140350877193, + "loss": 3.796, + "step": 6163 + }, + { + "epoch": 0.32442105263157894, + "grad_norm": 0.5234375, + "learning_rate": 0.00015012865497076024, + "loss": 3.4692, + "step": 6164 + }, + { + "epoch": 0.3244736842105263, + "grad_norm": 0.578125, + "learning_rate": 0.0001501169590643275, + "loss": 3.7627, + "step": 6165 + }, + { + "epoch": 0.32452631578947366, + "grad_norm": 0.55078125, + "learning_rate": 0.00015010526315789474, + "loss": 3.3226, + "step": 6166 + }, + { + "epoch": 0.3245789473684211, + "grad_norm": 0.5859375, + "learning_rate": 0.000150093567251462, + "loss": 3.7346, + "step": 6167 + }, + { + "epoch": 0.32463157894736844, + "grad_norm": 0.470703125, + "learning_rate": 0.00015008187134502924, + "loss": 3.5563, + "step": 6168 + }, + { + "epoch": 0.3246842105263158, + "grad_norm": 0.498046875, + "learning_rate": 0.0001500701754385965, + "loss": 3.8929, + "step": 6169 + }, + { + "epoch": 0.32473684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00015005847953216374, + "loss": 3.4352, + "step": 6170 + }, + { + "epoch": 0.3247894736842105, + "grad_norm": 0.53125, + "learning_rate": 0.000150046783625731, + "loss": 3.8701, + "step": 6171 + }, + { + "epoch": 0.3248421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00015003508771929827, + "loss": 4.0722, + "step": 6172 + }, + { + "epoch": 0.32489473684210524, + "grad_norm": 0.53515625, + "learning_rate": 0.00015002339181286551, + "loss": 3.3925, + "step": 6173 + }, + { + "epoch": 0.32494736842105265, + "grad_norm": 0.5, + "learning_rate": 0.00015001169590643276, + "loss": 3.6982, + "step": 6174 + }, + { + "epoch": 0.325, + "grad_norm": 0.5546875, + "learning_rate": 0.00015000000000000001, + "loss": 3.9274, + "step": 6175 + }, + { + "epoch": 0.32505263157894737, + "grad_norm": 0.53515625, + "learning_rate": 0.00014998830409356726, + "loss": 3.3698, + "step": 6176 + }, + { + "epoch": 0.32510526315789473, + "grad_norm": 0.65234375, + "learning_rate": 0.0001499766081871345, + "loss": 3.9593, + "step": 6177 + }, + { + "epoch": 0.3251578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00014996491228070176, + "loss": 3.8893, + "step": 6178 + }, + { + "epoch": 0.32521052631578945, + "grad_norm": 0.5, + "learning_rate": 0.000149953216374269, + "loss": 3.793, + "step": 6179 + }, + { + "epoch": 0.32526315789473687, + "grad_norm": 0.5703125, + "learning_rate": 0.00014994152046783626, + "loss": 4.1298, + "step": 6180 + }, + { + "epoch": 0.3253157894736842, + "grad_norm": 0.55859375, + "learning_rate": 0.0001499298245614035, + "loss": 3.9356, + "step": 6181 + }, + { + "epoch": 0.3253684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00014991812865497076, + "loss": 3.727, + "step": 6182 + }, + { + "epoch": 0.32542105263157894, + "grad_norm": 0.53125, + "learning_rate": 0.000149906432748538, + "loss": 3.758, + "step": 6183 + }, + { + "epoch": 0.3254736842105263, + "grad_norm": 0.5234375, + "learning_rate": 0.00014989473684210526, + "loss": 3.7894, + "step": 6184 + }, + { + "epoch": 0.32552631578947366, + "grad_norm": 0.55078125, + "learning_rate": 0.0001498830409356725, + "loss": 3.5265, + "step": 6185 + }, + { + "epoch": 0.3255789473684211, + "grad_norm": 0.482421875, + "learning_rate": 0.00014987134502923976, + "loss": 3.109, + "step": 6186 + }, + { + "epoch": 0.32563157894736844, + "grad_norm": 0.46875, + "learning_rate": 0.00014985964912280704, + "loss": 4.2255, + "step": 6187 + }, + { + "epoch": 0.3256842105263158, + "grad_norm": 0.54296875, + "learning_rate": 0.00014984795321637429, + "loss": 4.0243, + "step": 6188 + }, + { + "epoch": 0.32573684210526316, + "grad_norm": 0.5625, + "learning_rate": 0.00014983625730994154, + "loss": 3.9512, + "step": 6189 + }, + { + "epoch": 0.3257894736842105, + "grad_norm": 0.53515625, + "learning_rate": 0.00014982456140350878, + "loss": 3.3223, + "step": 6190 + }, + { + "epoch": 0.3258421052631579, + "grad_norm": 0.482421875, + "learning_rate": 0.00014981286549707603, + "loss": 3.3539, + "step": 6191 + }, + { + "epoch": 0.32589473684210524, + "grad_norm": 0.46484375, + "learning_rate": 0.00014980116959064328, + "loss": 3.6502, + "step": 6192 + }, + { + "epoch": 0.32594736842105265, + "grad_norm": 0.46484375, + "learning_rate": 0.00014978947368421053, + "loss": 3.8878, + "step": 6193 + }, + { + "epoch": 0.326, + "grad_norm": 0.451171875, + "learning_rate": 0.00014977777777777778, + "loss": 3.7837, + "step": 6194 + }, + { + "epoch": 0.32605263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.00014976608187134503, + "loss": 3.4816, + "step": 6195 + }, + { + "epoch": 0.32610526315789473, + "grad_norm": 0.5234375, + "learning_rate": 0.00014975438596491228, + "loss": 3.5517, + "step": 6196 + }, + { + "epoch": 0.3261578947368421, + "grad_norm": 0.57421875, + "learning_rate": 0.00014974269005847953, + "loss": 3.6458, + "step": 6197 + }, + { + "epoch": 0.32621052631578945, + "grad_norm": 0.515625, + "learning_rate": 0.00014973099415204678, + "loss": 3.2166, + "step": 6198 + }, + { + "epoch": 0.32626315789473687, + "grad_norm": 0.46875, + "learning_rate": 0.00014971929824561403, + "loss": 3.895, + "step": 6199 + }, + { + "epoch": 0.3263157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00014970760233918128, + "loss": 4.1567, + "step": 6200 + }, + { + "epoch": 0.3263684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00014969590643274856, + "loss": 3.6389, + "step": 6201 + }, + { + "epoch": 0.32642105263157895, + "grad_norm": 0.490234375, + "learning_rate": 0.0001496842105263158, + "loss": 3.7635, + "step": 6202 + }, + { + "epoch": 0.3264736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00014967251461988306, + "loss": 3.6478, + "step": 6203 + }, + { + "epoch": 0.32652631578947366, + "grad_norm": 0.54296875, + "learning_rate": 0.0001496608187134503, + "loss": 3.7101, + "step": 6204 + }, + { + "epoch": 0.3265789473684211, + "grad_norm": 0.6015625, + "learning_rate": 0.00014964912280701756, + "loss": 3.9938, + "step": 6205 + }, + { + "epoch": 0.32663157894736844, + "grad_norm": 0.5703125, + "learning_rate": 0.0001496374269005848, + "loss": 3.7889, + "step": 6206 + }, + { + "epoch": 0.3266842105263158, + "grad_norm": 0.546875, + "learning_rate": 0.00014962573099415206, + "loss": 3.7777, + "step": 6207 + }, + { + "epoch": 0.32673684210526316, + "grad_norm": 0.470703125, + "learning_rate": 0.0001496140350877193, + "loss": 4.0204, + "step": 6208 + }, + { + "epoch": 0.3267894736842105, + "grad_norm": 0.53515625, + "learning_rate": 0.00014960233918128655, + "loss": 3.6782, + "step": 6209 + }, + { + "epoch": 0.3268421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.0001495906432748538, + "loss": 3.7258, + "step": 6210 + }, + { + "epoch": 0.32689473684210524, + "grad_norm": 0.494140625, + "learning_rate": 0.00014957894736842105, + "loss": 3.8964, + "step": 6211 + }, + { + "epoch": 0.32694736842105265, + "grad_norm": 0.5234375, + "learning_rate": 0.0001495672514619883, + "loss": 3.5256, + "step": 6212 + }, + { + "epoch": 0.327, + "grad_norm": 0.52734375, + "learning_rate": 0.00014955555555555555, + "loss": 4.0507, + "step": 6213 + }, + { + "epoch": 0.3270526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.0001495438596491228, + "loss": 3.5251, + "step": 6214 + }, + { + "epoch": 0.32710526315789473, + "grad_norm": 0.53125, + "learning_rate": 0.00014953216374269005, + "loss": 3.7099, + "step": 6215 + }, + { + "epoch": 0.3271578947368421, + "grad_norm": 0.56640625, + "learning_rate": 0.00014952046783625733, + "loss": 3.8583, + "step": 6216 + }, + { + "epoch": 0.32721052631578945, + "grad_norm": 0.53515625, + "learning_rate": 0.00014950877192982458, + "loss": 3.6829, + "step": 6217 + }, + { + "epoch": 0.32726315789473687, + "grad_norm": 0.53125, + "learning_rate": 0.00014949707602339183, + "loss": 3.7221, + "step": 6218 + }, + { + "epoch": 0.3273157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00014948538011695908, + "loss": 3.642, + "step": 6219 + }, + { + "epoch": 0.3273684210526316, + "grad_norm": 0.49609375, + "learning_rate": 0.00014947368421052633, + "loss": 3.768, + "step": 6220 + }, + { + "epoch": 0.32742105263157895, + "grad_norm": 0.470703125, + "learning_rate": 0.00014946198830409358, + "loss": 3.3934, + "step": 6221 + }, + { + "epoch": 0.3274736842105263, + "grad_norm": 0.462890625, + "learning_rate": 0.00014945029239766083, + "loss": 3.8342, + "step": 6222 + }, + { + "epoch": 0.32752631578947367, + "grad_norm": 0.515625, + "learning_rate": 0.00014943859649122808, + "loss": 4.0018, + "step": 6223 + }, + { + "epoch": 0.327578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.00014942690058479533, + "loss": 3.9383, + "step": 6224 + }, + { + "epoch": 0.32763157894736844, + "grad_norm": 0.52734375, + "learning_rate": 0.00014941520467836257, + "loss": 3.6369, + "step": 6225 + }, + { + "epoch": 0.3276842105263158, + "grad_norm": 0.474609375, + "learning_rate": 0.00014940350877192982, + "loss": 3.4896, + "step": 6226 + }, + { + "epoch": 0.32773684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00014939181286549707, + "loss": 3.5721, + "step": 6227 + }, + { + "epoch": 0.3277894736842105, + "grad_norm": 0.4609375, + "learning_rate": 0.00014938011695906432, + "loss": 3.4651, + "step": 6228 + }, + { + "epoch": 0.3278421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00014936842105263157, + "loss": 3.8571, + "step": 6229 + }, + { + "epoch": 0.32789473684210524, + "grad_norm": 0.578125, + "learning_rate": 0.00014935672514619882, + "loss": 3.6174, + "step": 6230 + }, + { + "epoch": 0.32794736842105265, + "grad_norm": 0.60546875, + "learning_rate": 0.0001493450292397661, + "loss": 4.0472, + "step": 6231 + }, + { + "epoch": 0.328, + "grad_norm": 0.59375, + "learning_rate": 0.00014933333333333335, + "loss": 3.4808, + "step": 6232 + }, + { + "epoch": 0.3280526315789474, + "grad_norm": 0.5546875, + "learning_rate": 0.0001493216374269006, + "loss": 3.8689, + "step": 6233 + }, + { + "epoch": 0.32810526315789473, + "grad_norm": 0.68359375, + "learning_rate": 0.00014930994152046785, + "loss": 3.8761, + "step": 6234 + }, + { + "epoch": 0.3281578947368421, + "grad_norm": 0.494140625, + "learning_rate": 0.0001492982456140351, + "loss": 3.943, + "step": 6235 + }, + { + "epoch": 0.32821052631578945, + "grad_norm": 0.55859375, + "learning_rate": 0.00014928654970760235, + "loss": 3.6521, + "step": 6236 + }, + { + "epoch": 0.32826315789473687, + "grad_norm": 0.48046875, + "learning_rate": 0.0001492748538011696, + "loss": 3.6818, + "step": 6237 + }, + { + "epoch": 0.3283157894736842, + "grad_norm": 0.486328125, + "learning_rate": 0.00014926315789473687, + "loss": 3.338, + "step": 6238 + }, + { + "epoch": 0.3283684210526316, + "grad_norm": 0.54296875, + "learning_rate": 0.0001492514619883041, + "loss": 3.0526, + "step": 6239 + }, + { + "epoch": 0.32842105263157895, + "grad_norm": 0.55859375, + "learning_rate": 0.00014923976608187135, + "loss": 3.3533, + "step": 6240 + }, + { + "epoch": 0.3284736842105263, + "grad_norm": 0.5390625, + "learning_rate": 0.0001492280701754386, + "loss": 3.78, + "step": 6241 + }, + { + "epoch": 0.32852631578947367, + "grad_norm": 0.490234375, + "learning_rate": 0.00014921637426900585, + "loss": 4.1489, + "step": 6242 + }, + { + "epoch": 0.328578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.0001492046783625731, + "loss": 4.055, + "step": 6243 + }, + { + "epoch": 0.32863157894736844, + "grad_norm": 0.54296875, + "learning_rate": 0.00014919298245614034, + "loss": 3.5548, + "step": 6244 + }, + { + "epoch": 0.3286842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.0001491812865497076, + "loss": 3.9854, + "step": 6245 + }, + { + "epoch": 0.32873684210526316, + "grad_norm": 0.59765625, + "learning_rate": 0.00014916959064327487, + "loss": 3.359, + "step": 6246 + }, + { + "epoch": 0.3287894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.00014915789473684212, + "loss": 3.7014, + "step": 6247 + }, + { + "epoch": 0.3288421052631579, + "grad_norm": 0.59765625, + "learning_rate": 0.00014914619883040937, + "loss": 3.7544, + "step": 6248 + }, + { + "epoch": 0.32889473684210524, + "grad_norm": 0.5546875, + "learning_rate": 0.00014913450292397662, + "loss": 3.4983, + "step": 6249 + }, + { + "epoch": 0.32894736842105265, + "grad_norm": 0.48828125, + "learning_rate": 0.00014912280701754387, + "loss": 3.7958, + "step": 6250 + }, + { + "epoch": 0.329, + "grad_norm": 0.486328125, + "learning_rate": 0.00014911111111111112, + "loss": 3.5989, + "step": 6251 + }, + { + "epoch": 0.3290526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.00014909941520467837, + "loss": 3.4961, + "step": 6252 + }, + { + "epoch": 0.32910526315789473, + "grad_norm": 0.65234375, + "learning_rate": 0.00014908771929824564, + "loss": 3.3178, + "step": 6253 + }, + { + "epoch": 0.3291578947368421, + "grad_norm": 0.490234375, + "learning_rate": 0.00014907602339181287, + "loss": 3.5751, + "step": 6254 + }, + { + "epoch": 0.32921052631578945, + "grad_norm": 0.466796875, + "learning_rate": 0.00014906432748538012, + "loss": 3.571, + "step": 6255 + }, + { + "epoch": 0.32926315789473687, + "grad_norm": 0.49609375, + "learning_rate": 0.00014905263157894737, + "loss": 3.6907, + "step": 6256 + }, + { + "epoch": 0.32931578947368423, + "grad_norm": 0.56640625, + "learning_rate": 0.00014904093567251462, + "loss": 3.6607, + "step": 6257 + }, + { + "epoch": 0.3293684210526316, + "grad_norm": 0.55078125, + "learning_rate": 0.00014902923976608187, + "loss": 3.6786, + "step": 6258 + }, + { + "epoch": 0.32942105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00014901754385964912, + "loss": 3.823, + "step": 6259 + }, + { + "epoch": 0.3294736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.00014900584795321636, + "loss": 4.1662, + "step": 6260 + }, + { + "epoch": 0.32952631578947367, + "grad_norm": 0.57421875, + "learning_rate": 0.00014899415204678364, + "loss": 3.2976, + "step": 6261 + }, + { + "epoch": 0.329578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.0001489824561403509, + "loss": 3.6643, + "step": 6262 + }, + { + "epoch": 0.32963157894736844, + "grad_norm": 0.54296875, + "learning_rate": 0.00014897076023391814, + "loss": 3.5328, + "step": 6263 + }, + { + "epoch": 0.3296842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.0001489590643274854, + "loss": 3.1603, + "step": 6264 + }, + { + "epoch": 0.32973684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00014894736842105264, + "loss": 3.6604, + "step": 6265 + }, + { + "epoch": 0.3297894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.0001489356725146199, + "loss": 4.1627, + "step": 6266 + }, + { + "epoch": 0.3298421052631579, + "grad_norm": 0.484375, + "learning_rate": 0.00014892397660818714, + "loss": 3.7517, + "step": 6267 + }, + { + "epoch": 0.32989473684210524, + "grad_norm": 0.4765625, + "learning_rate": 0.00014891228070175442, + "loss": 3.8129, + "step": 6268 + }, + { + "epoch": 0.32994736842105266, + "grad_norm": 0.5078125, + "learning_rate": 0.00014890058479532164, + "loss": 3.5116, + "step": 6269 + }, + { + "epoch": 0.33, + "grad_norm": 0.50390625, + "learning_rate": 0.0001488888888888889, + "loss": 3.8158, + "step": 6270 + }, + { + "epoch": 0.3300526315789474, + "grad_norm": 0.5, + "learning_rate": 0.00014887719298245614, + "loss": 3.483, + "step": 6271 + }, + { + "epoch": 0.33010526315789473, + "grad_norm": 0.51171875, + "learning_rate": 0.0001488654970760234, + "loss": 3.808, + "step": 6272 + }, + { + "epoch": 0.3301578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.00014885380116959064, + "loss": 3.8012, + "step": 6273 + }, + { + "epoch": 0.33021052631578945, + "grad_norm": 0.515625, + "learning_rate": 0.00014884210526315789, + "loss": 3.4825, + "step": 6274 + }, + { + "epoch": 0.33026315789473687, + "grad_norm": 0.60546875, + "learning_rate": 0.00014883040935672514, + "loss": 3.8282, + "step": 6275 + }, + { + "epoch": 0.33031578947368423, + "grad_norm": 0.54296875, + "learning_rate": 0.0001488187134502924, + "loss": 3.7312, + "step": 6276 + }, + { + "epoch": 0.3303684210526316, + "grad_norm": 0.484375, + "learning_rate": 0.00014880701754385966, + "loss": 3.6309, + "step": 6277 + }, + { + "epoch": 0.33042105263157895, + "grad_norm": 0.52734375, + "learning_rate": 0.0001487953216374269, + "loss": 4.1526, + "step": 6278 + }, + { + "epoch": 0.3304736842105263, + "grad_norm": 0.5703125, + "learning_rate": 0.00014878362573099416, + "loss": 3.7115, + "step": 6279 + }, + { + "epoch": 0.33052631578947367, + "grad_norm": 0.5625, + "learning_rate": 0.0001487719298245614, + "loss": 3.4209, + "step": 6280 + }, + { + "epoch": 0.33057894736842103, + "grad_norm": 0.5, + "learning_rate": 0.00014876023391812866, + "loss": 3.8638, + "step": 6281 + }, + { + "epoch": 0.33063157894736844, + "grad_norm": 0.57421875, + "learning_rate": 0.0001487485380116959, + "loss": 3.9567, + "step": 6282 + }, + { + "epoch": 0.3306842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.0001487368421052632, + "loss": 3.9528, + "step": 6283 + }, + { + "epoch": 0.33073684210526316, + "grad_norm": 0.55078125, + "learning_rate": 0.00014872514619883044, + "loss": 3.9112, + "step": 6284 + }, + { + "epoch": 0.3307894736842105, + "grad_norm": 0.58984375, + "learning_rate": 0.00014871345029239766, + "loss": 3.4838, + "step": 6285 + }, + { + "epoch": 0.3308421052631579, + "grad_norm": 0.46875, + "learning_rate": 0.0001487017543859649, + "loss": 3.619, + "step": 6286 + }, + { + "epoch": 0.33089473684210524, + "grad_norm": 0.490234375, + "learning_rate": 0.00014869005847953216, + "loss": 3.7387, + "step": 6287 + }, + { + "epoch": 0.33094736842105266, + "grad_norm": 0.50390625, + "learning_rate": 0.0001486783625730994, + "loss": 3.635, + "step": 6288 + }, + { + "epoch": 0.331, + "grad_norm": 0.482421875, + "learning_rate": 0.00014866666666666666, + "loss": 3.5346, + "step": 6289 + }, + { + "epoch": 0.3310526315789474, + "grad_norm": 0.47265625, + "learning_rate": 0.00014865497076023393, + "loss": 3.8365, + "step": 6290 + }, + { + "epoch": 0.33110526315789474, + "grad_norm": 0.474609375, + "learning_rate": 0.00014864327485380118, + "loss": 3.8183, + "step": 6291 + }, + { + "epoch": 0.3311578947368421, + "grad_norm": 0.48828125, + "learning_rate": 0.00014863157894736843, + "loss": 3.6569, + "step": 6292 + }, + { + "epoch": 0.33121052631578946, + "grad_norm": 0.490234375, + "learning_rate": 0.00014861988304093568, + "loss": 3.8289, + "step": 6293 + }, + { + "epoch": 0.3312631578947368, + "grad_norm": 0.54296875, + "learning_rate": 0.00014860818713450293, + "loss": 3.5154, + "step": 6294 + }, + { + "epoch": 0.33131578947368423, + "grad_norm": 0.55859375, + "learning_rate": 0.00014859649122807018, + "loss": 3.6762, + "step": 6295 + }, + { + "epoch": 0.3313684210526316, + "grad_norm": 0.57421875, + "learning_rate": 0.00014858479532163743, + "loss": 3.7868, + "step": 6296 + }, + { + "epoch": 0.33142105263157895, + "grad_norm": 0.56640625, + "learning_rate": 0.00014857309941520468, + "loss": 3.8371, + "step": 6297 + }, + { + "epoch": 0.3314736842105263, + "grad_norm": 0.59765625, + "learning_rate": 0.00014856140350877196, + "loss": 4.2352, + "step": 6298 + }, + { + "epoch": 0.33152631578947367, + "grad_norm": 0.5625, + "learning_rate": 0.0001485497076023392, + "loss": 3.8392, + "step": 6299 + }, + { + "epoch": 0.33157894736842103, + "grad_norm": 0.53515625, + "learning_rate": 0.00014853801169590643, + "loss": 3.7052, + "step": 6300 + }, + { + "epoch": 0.33163157894736844, + "grad_norm": 0.55859375, + "learning_rate": 0.00014852631578947368, + "loss": 3.7344, + "step": 6301 + }, + { + "epoch": 0.3316842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.00014851461988304093, + "loss": 3.2412, + "step": 6302 + }, + { + "epoch": 0.33173684210526316, + "grad_norm": 0.49609375, + "learning_rate": 0.00014850292397660818, + "loss": 3.6266, + "step": 6303 + }, + { + "epoch": 0.3317894736842105, + "grad_norm": 0.53515625, + "learning_rate": 0.00014849122807017543, + "loss": 3.5641, + "step": 6304 + }, + { + "epoch": 0.3318421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.0001484795321637427, + "loss": 3.2345, + "step": 6305 + }, + { + "epoch": 0.33189473684210524, + "grad_norm": 0.5, + "learning_rate": 0.00014846783625730995, + "loss": 3.9926, + "step": 6306 + }, + { + "epoch": 0.33194736842105266, + "grad_norm": 0.56640625, + "learning_rate": 0.0001484561403508772, + "loss": 3.98, + "step": 6307 + }, + { + "epoch": 0.332, + "grad_norm": 0.48046875, + "learning_rate": 0.00014844444444444445, + "loss": 3.5113, + "step": 6308 + }, + { + "epoch": 0.3320526315789474, + "grad_norm": 0.498046875, + "learning_rate": 0.0001484327485380117, + "loss": 3.2083, + "step": 6309 + }, + { + "epoch": 0.33210526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.00014842105263157895, + "loss": 3.2766, + "step": 6310 + }, + { + "epoch": 0.3321578947368421, + "grad_norm": 0.466796875, + "learning_rate": 0.0001484093567251462, + "loss": 3.6297, + "step": 6311 + }, + { + "epoch": 0.33221052631578946, + "grad_norm": 0.52734375, + "learning_rate": 0.00014839766081871345, + "loss": 3.609, + "step": 6312 + }, + { + "epoch": 0.3322631578947368, + "grad_norm": 0.546875, + "learning_rate": 0.00014838596491228073, + "loss": 3.6388, + "step": 6313 + }, + { + "epoch": 0.33231578947368423, + "grad_norm": 0.58203125, + "learning_rate": 0.00014837426900584798, + "loss": 3.6447, + "step": 6314 + }, + { + "epoch": 0.3323684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.0001483625730994152, + "loss": 3.7031, + "step": 6315 + }, + { + "epoch": 0.33242105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00014835087719298245, + "loss": 3.6116, + "step": 6316 + }, + { + "epoch": 0.3324736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.0001483391812865497, + "loss": 3.6367, + "step": 6317 + }, + { + "epoch": 0.33252631578947367, + "grad_norm": 0.52734375, + "learning_rate": 0.00014832748538011695, + "loss": 3.5193, + "step": 6318 + }, + { + "epoch": 0.33257894736842103, + "grad_norm": 0.482421875, + "learning_rate": 0.0001483157894736842, + "loss": 3.3894, + "step": 6319 + }, + { + "epoch": 0.33263157894736844, + "grad_norm": 0.515625, + "learning_rate": 0.00014830409356725148, + "loss": 3.5591, + "step": 6320 + }, + { + "epoch": 0.3326842105263158, + "grad_norm": 0.498046875, + "learning_rate": 0.00014829239766081873, + "loss": 3.5119, + "step": 6321 + }, + { + "epoch": 0.33273684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.00014828070175438598, + "loss": 3.5494, + "step": 6322 + }, + { + "epoch": 0.3327894736842105, + "grad_norm": 0.5546875, + "learning_rate": 0.00014826900584795322, + "loss": 3.9875, + "step": 6323 + }, + { + "epoch": 0.3328421052631579, + "grad_norm": 0.55078125, + "learning_rate": 0.00014825730994152047, + "loss": 4.0064, + "step": 6324 + }, + { + "epoch": 0.33289473684210524, + "grad_norm": 0.55078125, + "learning_rate": 0.00014824561403508772, + "loss": 3.4425, + "step": 6325 + }, + { + "epoch": 0.33294736842105266, + "grad_norm": 0.5078125, + "learning_rate": 0.00014823391812865497, + "loss": 3.7582, + "step": 6326 + }, + { + "epoch": 0.333, + "grad_norm": 0.5859375, + "learning_rate": 0.00014822222222222225, + "loss": 3.3833, + "step": 6327 + }, + { + "epoch": 0.3330526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.0001482105263157895, + "loss": 3.5436, + "step": 6328 + }, + { + "epoch": 0.33310526315789474, + "grad_norm": 0.4921875, + "learning_rate": 0.00014819883040935675, + "loss": 3.8148, + "step": 6329 + }, + { + "epoch": 0.3331578947368421, + "grad_norm": 0.55078125, + "learning_rate": 0.000148187134502924, + "loss": 3.4607, + "step": 6330 + }, + { + "epoch": 0.33321052631578946, + "grad_norm": 0.6328125, + "learning_rate": 0.00014817543859649122, + "loss": 3.6224, + "step": 6331 + }, + { + "epoch": 0.3332631578947368, + "grad_norm": 0.65625, + "learning_rate": 0.00014816374269005847, + "loss": 3.0698, + "step": 6332 + }, + { + "epoch": 0.33331578947368423, + "grad_norm": 0.478515625, + "learning_rate": 0.00014815204678362572, + "loss": 3.7362, + "step": 6333 + }, + { + "epoch": 0.3333684210526316, + "grad_norm": 0.5625, + "learning_rate": 0.00014814035087719297, + "loss": 3.347, + "step": 6334 + }, + { + "epoch": 0.33342105263157895, + "grad_norm": 0.5, + "learning_rate": 0.00014812865497076025, + "loss": 3.6931, + "step": 6335 + }, + { + "epoch": 0.3334736842105263, + "grad_norm": 0.478515625, + "learning_rate": 0.0001481169590643275, + "loss": 4.1615, + "step": 6336 + }, + { + "epoch": 0.33352631578947367, + "grad_norm": 0.48828125, + "learning_rate": 0.00014810526315789475, + "loss": 3.5965, + "step": 6337 + }, + { + "epoch": 0.33357894736842103, + "grad_norm": 0.484375, + "learning_rate": 0.000148093567251462, + "loss": 3.7743, + "step": 6338 + }, + { + "epoch": 0.33363157894736845, + "grad_norm": 0.61328125, + "learning_rate": 0.00014808187134502925, + "loss": 3.4183, + "step": 6339 + }, + { + "epoch": 0.3336842105263158, + "grad_norm": 0.455078125, + "learning_rate": 0.0001480701754385965, + "loss": 3.4574, + "step": 6340 + }, + { + "epoch": 0.33373684210526317, + "grad_norm": 0.46484375, + "learning_rate": 0.00014805847953216374, + "loss": 3.6677, + "step": 6341 + }, + { + "epoch": 0.3337894736842105, + "grad_norm": 0.478515625, + "learning_rate": 0.00014804678362573102, + "loss": 3.8541, + "step": 6342 + }, + { + "epoch": 0.3338421052631579, + "grad_norm": 0.48828125, + "learning_rate": 0.00014803508771929827, + "loss": 3.5896, + "step": 6343 + }, + { + "epoch": 0.33389473684210524, + "grad_norm": 0.48046875, + "learning_rate": 0.00014802339181286552, + "loss": 3.4241, + "step": 6344 + }, + { + "epoch": 0.3339473684210526, + "grad_norm": 0.51953125, + "learning_rate": 0.00014801169590643277, + "loss": 3.513, + "step": 6345 + }, + { + "epoch": 0.334, + "grad_norm": 0.5, + "learning_rate": 0.000148, + "loss": 3.7737, + "step": 6346 + }, + { + "epoch": 0.3340526315789474, + "grad_norm": 0.98828125, + "learning_rate": 0.00014798830409356724, + "loss": 3.8793, + "step": 6347 + }, + { + "epoch": 0.33410526315789474, + "grad_norm": 0.52734375, + "learning_rate": 0.0001479766081871345, + "loss": 3.6405, + "step": 6348 + }, + { + "epoch": 0.3341578947368421, + "grad_norm": 0.69921875, + "learning_rate": 0.00014796491228070174, + "loss": 3.6985, + "step": 6349 + }, + { + "epoch": 0.33421052631578946, + "grad_norm": 0.51171875, + "learning_rate": 0.00014795321637426902, + "loss": 3.7345, + "step": 6350 + }, + { + "epoch": 0.3342631578947368, + "grad_norm": 0.53125, + "learning_rate": 0.00014794152046783627, + "loss": 3.7064, + "step": 6351 + }, + { + "epoch": 0.33431578947368423, + "grad_norm": 0.52734375, + "learning_rate": 0.00014792982456140352, + "loss": 4.0243, + "step": 6352 + }, + { + "epoch": 0.3343684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.00014791812865497077, + "loss": 3.788, + "step": 6353 + }, + { + "epoch": 0.33442105263157895, + "grad_norm": 0.5859375, + "learning_rate": 0.00014790643274853802, + "loss": 3.074, + "step": 6354 + }, + { + "epoch": 0.3344736842105263, + "grad_norm": 0.59765625, + "learning_rate": 0.00014789473684210527, + "loss": 3.3576, + "step": 6355 + }, + { + "epoch": 0.33452631578947367, + "grad_norm": 0.58984375, + "learning_rate": 0.00014788304093567252, + "loss": 3.364, + "step": 6356 + }, + { + "epoch": 0.33457894736842103, + "grad_norm": 0.7109375, + "learning_rate": 0.0001478713450292398, + "loss": 3.3632, + "step": 6357 + }, + { + "epoch": 0.33463157894736845, + "grad_norm": 0.5, + "learning_rate": 0.00014785964912280704, + "loss": 3.5448, + "step": 6358 + }, + { + "epoch": 0.3346842105263158, + "grad_norm": 0.482421875, + "learning_rate": 0.0001478479532163743, + "loss": 3.8016, + "step": 6359 + }, + { + "epoch": 0.33473684210526317, + "grad_norm": 0.52734375, + "learning_rate": 0.00014783625730994154, + "loss": 3.2526, + "step": 6360 + }, + { + "epoch": 0.3347894736842105, + "grad_norm": 0.5859375, + "learning_rate": 0.00014782456140350876, + "loss": 3.7055, + "step": 6361 + }, + { + "epoch": 0.3348421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.000147812865497076, + "loss": 3.6805, + "step": 6362 + }, + { + "epoch": 0.33489473684210525, + "grad_norm": 0.5625, + "learning_rate": 0.00014780116959064326, + "loss": 4.2114, + "step": 6363 + }, + { + "epoch": 0.3349473684210526, + "grad_norm": 0.65625, + "learning_rate": 0.00014778947368421054, + "loss": 3.8048, + "step": 6364 + }, + { + "epoch": 0.335, + "grad_norm": 0.54296875, + "learning_rate": 0.0001477777777777778, + "loss": 3.7141, + "step": 6365 + }, + { + "epoch": 0.3350526315789474, + "grad_norm": 0.625, + "learning_rate": 0.00014776608187134504, + "loss": 3.7213, + "step": 6366 + }, + { + "epoch": 0.33510526315789474, + "grad_norm": 0.45703125, + "learning_rate": 0.0001477543859649123, + "loss": 3.5761, + "step": 6367 + }, + { + "epoch": 0.3351578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.00014774269005847954, + "loss": 3.7319, + "step": 6368 + }, + { + "epoch": 0.33521052631578946, + "grad_norm": 0.515625, + "learning_rate": 0.0001477309941520468, + "loss": 3.6444, + "step": 6369 + }, + { + "epoch": 0.3352631578947368, + "grad_norm": 0.53125, + "learning_rate": 0.00014771929824561404, + "loss": 3.4402, + "step": 6370 + }, + { + "epoch": 0.33531578947368423, + "grad_norm": 0.53515625, + "learning_rate": 0.0001477076023391813, + "loss": 3.8197, + "step": 6371 + }, + { + "epoch": 0.3353684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00014769590643274856, + "loss": 3.6193, + "step": 6372 + }, + { + "epoch": 0.33542105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.0001476842105263158, + "loss": 3.727, + "step": 6373 + }, + { + "epoch": 0.3354736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.00014767251461988306, + "loss": 3.1869, + "step": 6374 + }, + { + "epoch": 0.3355263157894737, + "grad_norm": 0.5, + "learning_rate": 0.0001476608187134503, + "loss": 3.9763, + "step": 6375 + }, + { + "epoch": 0.33557894736842103, + "grad_norm": 0.51171875, + "learning_rate": 0.00014764912280701756, + "loss": 3.3455, + "step": 6376 + }, + { + "epoch": 0.33563157894736845, + "grad_norm": 0.50390625, + "learning_rate": 0.00014763742690058478, + "loss": 3.8688, + "step": 6377 + }, + { + "epoch": 0.3356842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00014762573099415203, + "loss": 3.5319, + "step": 6378 + }, + { + "epoch": 0.33573684210526317, + "grad_norm": 0.62890625, + "learning_rate": 0.0001476140350877193, + "loss": 3.7393, + "step": 6379 + }, + { + "epoch": 0.3357894736842105, + "grad_norm": 0.5625, + "learning_rate": 0.00014760233918128656, + "loss": 4.0357, + "step": 6380 + }, + { + "epoch": 0.3358421052631579, + "grad_norm": 0.6640625, + "learning_rate": 0.0001475906432748538, + "loss": 3.0378, + "step": 6381 + }, + { + "epoch": 0.33589473684210525, + "grad_norm": 0.59375, + "learning_rate": 0.00014757894736842106, + "loss": 3.5166, + "step": 6382 + }, + { + "epoch": 0.3359473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.0001475672514619883, + "loss": 3.4072, + "step": 6383 + }, + { + "epoch": 0.336, + "grad_norm": 0.50390625, + "learning_rate": 0.00014755555555555556, + "loss": 3.4852, + "step": 6384 + }, + { + "epoch": 0.3360526315789474, + "grad_norm": 0.515625, + "learning_rate": 0.0001475438596491228, + "loss": 3.4839, + "step": 6385 + }, + { + "epoch": 0.33610526315789474, + "grad_norm": 0.6796875, + "learning_rate": 0.00014753216374269006, + "loss": 3.1921, + "step": 6386 + }, + { + "epoch": 0.3361578947368421, + "grad_norm": 0.48828125, + "learning_rate": 0.00014752046783625733, + "loss": 3.7161, + "step": 6387 + }, + { + "epoch": 0.33621052631578946, + "grad_norm": 0.490234375, + "learning_rate": 0.00014750877192982458, + "loss": 4.1923, + "step": 6388 + }, + { + "epoch": 0.3362631578947368, + "grad_norm": 0.482421875, + "learning_rate": 0.00014749707602339183, + "loss": 3.2548, + "step": 6389 + }, + { + "epoch": 0.33631578947368423, + "grad_norm": 0.52734375, + "learning_rate": 0.00014748538011695908, + "loss": 3.8758, + "step": 6390 + }, + { + "epoch": 0.3363684210526316, + "grad_norm": 0.5, + "learning_rate": 0.00014747368421052633, + "loss": 3.92, + "step": 6391 + }, + { + "epoch": 0.33642105263157895, + "grad_norm": 0.470703125, + "learning_rate": 0.00014746198830409356, + "loss": 3.5598, + "step": 6392 + }, + { + "epoch": 0.3364736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.0001474502923976608, + "loss": 3.8715, + "step": 6393 + }, + { + "epoch": 0.3365263157894737, + "grad_norm": 0.484375, + "learning_rate": 0.00014743859649122808, + "loss": 3.8015, + "step": 6394 + }, + { + "epoch": 0.33657894736842103, + "grad_norm": 0.52734375, + "learning_rate": 0.00014742690058479533, + "loss": 3.8437, + "step": 6395 + }, + { + "epoch": 0.33663157894736845, + "grad_norm": 0.5859375, + "learning_rate": 0.00014741520467836258, + "loss": 3.6626, + "step": 6396 + }, + { + "epoch": 0.3366842105263158, + "grad_norm": 0.482421875, + "learning_rate": 0.00014740350877192983, + "loss": 3.2076, + "step": 6397 + }, + { + "epoch": 0.33673684210526317, + "grad_norm": 0.51953125, + "learning_rate": 0.00014739181286549708, + "loss": 3.4502, + "step": 6398 + }, + { + "epoch": 0.3367894736842105, + "grad_norm": 0.48046875, + "learning_rate": 0.00014738011695906433, + "loss": 3.6632, + "step": 6399 + }, + { + "epoch": 0.3368421052631579, + "grad_norm": 0.478515625, + "learning_rate": 0.00014736842105263158, + "loss": 4.2054, + "step": 6400 + }, + { + "epoch": 0.33689473684210525, + "grad_norm": 0.50390625, + "learning_rate": 0.00014735672514619886, + "loss": 3.5846, + "step": 6401 + }, + { + "epoch": 0.3369473684210526, + "grad_norm": 0.55859375, + "learning_rate": 0.0001473450292397661, + "loss": 3.6702, + "step": 6402 + }, + { + "epoch": 0.337, + "grad_norm": 0.55078125, + "learning_rate": 0.00014733333333333335, + "loss": 3.6254, + "step": 6403 + }, + { + "epoch": 0.3370526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.0001473216374269006, + "loss": 3.8474, + "step": 6404 + }, + { + "epoch": 0.33710526315789474, + "grad_norm": 0.515625, + "learning_rate": 0.00014730994152046785, + "loss": 3.8332, + "step": 6405 + }, + { + "epoch": 0.3371578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.0001472982456140351, + "loss": 3.4266, + "step": 6406 + }, + { + "epoch": 0.33721052631578946, + "grad_norm": 0.55078125, + "learning_rate": 0.00014728654970760233, + "loss": 3.812, + "step": 6407 + }, + { + "epoch": 0.3372631578947368, + "grad_norm": 0.61328125, + "learning_rate": 0.00014727485380116958, + "loss": 3.8649, + "step": 6408 + }, + { + "epoch": 0.33731578947368424, + "grad_norm": 0.578125, + "learning_rate": 0.00014726315789473685, + "loss": 4.0012, + "step": 6409 + }, + { + "epoch": 0.3373684210526316, + "grad_norm": 0.578125, + "learning_rate": 0.0001472514619883041, + "loss": 3.5215, + "step": 6410 + }, + { + "epoch": 0.33742105263157895, + "grad_norm": 0.58203125, + "learning_rate": 0.00014723976608187135, + "loss": 3.974, + "step": 6411 + }, + { + "epoch": 0.3374736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.0001472280701754386, + "loss": 3.7217, + "step": 6412 + }, + { + "epoch": 0.3375263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.00014721637426900585, + "loss": 4.0708, + "step": 6413 + }, + { + "epoch": 0.33757894736842103, + "grad_norm": 0.5, + "learning_rate": 0.0001472046783625731, + "loss": 3.856, + "step": 6414 + }, + { + "epoch": 0.3376315789473684, + "grad_norm": 0.58203125, + "learning_rate": 0.00014719298245614035, + "loss": 3.6523, + "step": 6415 + }, + { + "epoch": 0.3376842105263158, + "grad_norm": 0.53125, + "learning_rate": 0.00014718128654970763, + "loss": 3.8433, + "step": 6416 + }, + { + "epoch": 0.33773684210526317, + "grad_norm": 0.671875, + "learning_rate": 0.00014716959064327488, + "loss": 3.8648, + "step": 6417 + }, + { + "epoch": 0.33778947368421053, + "grad_norm": 0.490234375, + "learning_rate": 0.00014715789473684213, + "loss": 3.651, + "step": 6418 + }, + { + "epoch": 0.3378421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.00014714619883040938, + "loss": 3.4104, + "step": 6419 + }, + { + "epoch": 0.33789473684210525, + "grad_norm": 0.46875, + "learning_rate": 0.00014713450292397662, + "loss": 3.7507, + "step": 6420 + }, + { + "epoch": 0.3379473684210526, + "grad_norm": 0.484375, + "learning_rate": 0.00014712280701754387, + "loss": 3.6502, + "step": 6421 + }, + { + "epoch": 0.338, + "grad_norm": 0.455078125, + "learning_rate": 0.00014711111111111112, + "loss": 3.9278, + "step": 6422 + }, + { + "epoch": 0.3380526315789474, + "grad_norm": 0.48046875, + "learning_rate": 0.00014709941520467835, + "loss": 3.6352, + "step": 6423 + }, + { + "epoch": 0.33810526315789474, + "grad_norm": 0.4453125, + "learning_rate": 0.00014708771929824562, + "loss": 3.7654, + "step": 6424 + }, + { + "epoch": 0.3381578947368421, + "grad_norm": 0.84765625, + "learning_rate": 0.00014707602339181287, + "loss": 3.4305, + "step": 6425 + }, + { + "epoch": 0.33821052631578946, + "grad_norm": 0.4609375, + "learning_rate": 0.00014706432748538012, + "loss": 3.3702, + "step": 6426 + }, + { + "epoch": 0.3382631578947368, + "grad_norm": 0.52734375, + "learning_rate": 0.00014705263157894737, + "loss": 4.1592, + "step": 6427 + }, + { + "epoch": 0.33831578947368424, + "grad_norm": 0.5546875, + "learning_rate": 0.00014704093567251462, + "loss": 3.6354, + "step": 6428 + }, + { + "epoch": 0.3383684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00014702923976608187, + "loss": 3.9156, + "step": 6429 + }, + { + "epoch": 0.33842105263157896, + "grad_norm": 0.56640625, + "learning_rate": 0.00014701754385964912, + "loss": 3.9856, + "step": 6430 + }, + { + "epoch": 0.3384736842105263, + "grad_norm": 0.494140625, + "learning_rate": 0.0001470058479532164, + "loss": 3.5133, + "step": 6431 + }, + { + "epoch": 0.3385263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00014699415204678365, + "loss": 3.7244, + "step": 6432 + }, + { + "epoch": 0.33857894736842103, + "grad_norm": 0.5390625, + "learning_rate": 0.0001469824561403509, + "loss": 4.079, + "step": 6433 + }, + { + "epoch": 0.3386315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00014697076023391815, + "loss": 3.6194, + "step": 6434 + }, + { + "epoch": 0.3386842105263158, + "grad_norm": 0.5546875, + "learning_rate": 0.0001469590643274854, + "loss": 3.244, + "step": 6435 + }, + { + "epoch": 0.33873684210526317, + "grad_norm": 0.58984375, + "learning_rate": 0.00014694736842105265, + "loss": 3.7342, + "step": 6436 + }, + { + "epoch": 0.33878947368421053, + "grad_norm": 0.486328125, + "learning_rate": 0.0001469356725146199, + "loss": 3.6733, + "step": 6437 + }, + { + "epoch": 0.3388421052631579, + "grad_norm": 0.486328125, + "learning_rate": 0.00014692397660818712, + "loss": 3.167, + "step": 6438 + }, + { + "epoch": 0.33889473684210525, + "grad_norm": 0.486328125, + "learning_rate": 0.0001469122807017544, + "loss": 3.2662, + "step": 6439 + }, + { + "epoch": 0.3389473684210526, + "grad_norm": 0.474609375, + "learning_rate": 0.00014690058479532164, + "loss": 3.6689, + "step": 6440 + }, + { + "epoch": 0.339, + "grad_norm": 0.48046875, + "learning_rate": 0.0001468888888888889, + "loss": 3.8325, + "step": 6441 + }, + { + "epoch": 0.3390526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.00014687719298245614, + "loss": 3.8095, + "step": 6442 + }, + { + "epoch": 0.33910526315789474, + "grad_norm": 0.54296875, + "learning_rate": 0.0001468654970760234, + "loss": 3.0169, + "step": 6443 + }, + { + "epoch": 0.3391578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00014685380116959064, + "loss": 3.6935, + "step": 6444 + }, + { + "epoch": 0.33921052631578946, + "grad_norm": 1.03125, + "learning_rate": 0.0001468421052631579, + "loss": 3.4129, + "step": 6445 + }, + { + "epoch": 0.3392631578947368, + "grad_norm": 0.55078125, + "learning_rate": 0.00014683040935672517, + "loss": 3.845, + "step": 6446 + }, + { + "epoch": 0.33931578947368424, + "grad_norm": 0.5234375, + "learning_rate": 0.00014681871345029242, + "loss": 3.4121, + "step": 6447 + }, + { + "epoch": 0.3393684210526316, + "grad_norm": 0.52734375, + "learning_rate": 0.00014680701754385967, + "loss": 3.7939, + "step": 6448 + }, + { + "epoch": 0.33942105263157896, + "grad_norm": 0.51171875, + "learning_rate": 0.00014679532163742692, + "loss": 3.7366, + "step": 6449 + }, + { + "epoch": 0.3394736842105263, + "grad_norm": 0.5625, + "learning_rate": 0.00014678362573099417, + "loss": 3.7062, + "step": 6450 + }, + { + "epoch": 0.3395263157894737, + "grad_norm": 0.5234375, + "learning_rate": 0.00014677192982456142, + "loss": 3.8279, + "step": 6451 + }, + { + "epoch": 0.33957894736842104, + "grad_norm": 0.5703125, + "learning_rate": 0.00014676023391812867, + "loss": 3.8841, + "step": 6452 + }, + { + "epoch": 0.3396315789473684, + "grad_norm": 0.6015625, + "learning_rate": 0.00014674853801169592, + "loss": 4.0508, + "step": 6453 + }, + { + "epoch": 0.3396842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.00014673684210526317, + "loss": 3.6654, + "step": 6454 + }, + { + "epoch": 0.33973684210526317, + "grad_norm": 0.5390625, + "learning_rate": 0.00014672514619883041, + "loss": 3.9273, + "step": 6455 + }, + { + "epoch": 0.33978947368421053, + "grad_norm": 0.609375, + "learning_rate": 0.00014671345029239766, + "loss": 4.1883, + "step": 6456 + }, + { + "epoch": 0.3398421052631579, + "grad_norm": 0.6171875, + "learning_rate": 0.00014670175438596491, + "loss": 3.529, + "step": 6457 + }, + { + "epoch": 0.33989473684210525, + "grad_norm": 0.640625, + "learning_rate": 0.00014669005847953216, + "loss": 4.1368, + "step": 6458 + }, + { + "epoch": 0.3399473684210526, + "grad_norm": 0.478515625, + "learning_rate": 0.0001466783625730994, + "loss": 3.9034, + "step": 6459 + }, + { + "epoch": 0.34, + "grad_norm": 0.56640625, + "learning_rate": 0.00014666666666666666, + "loss": 3.8532, + "step": 6460 + }, + { + "epoch": 0.3400526315789474, + "grad_norm": 0.5859375, + "learning_rate": 0.00014665497076023394, + "loss": 3.7497, + "step": 6461 + }, + { + "epoch": 0.34010526315789474, + "grad_norm": 0.5859375, + "learning_rate": 0.0001466432748538012, + "loss": 3.6857, + "step": 6462 + }, + { + "epoch": 0.3401578947368421, + "grad_norm": 0.625, + "learning_rate": 0.00014663157894736844, + "loss": 3.8114, + "step": 6463 + }, + { + "epoch": 0.34021052631578946, + "grad_norm": 0.50390625, + "learning_rate": 0.0001466198830409357, + "loss": 3.472, + "step": 6464 + }, + { + "epoch": 0.3402631578947368, + "grad_norm": 0.67578125, + "learning_rate": 0.00014660818713450294, + "loss": 3.3668, + "step": 6465 + }, + { + "epoch": 0.34031578947368424, + "grad_norm": 0.4921875, + "learning_rate": 0.0001465964912280702, + "loss": 3.8746, + "step": 6466 + }, + { + "epoch": 0.3403684210526316, + "grad_norm": 0.46875, + "learning_rate": 0.00014658479532163744, + "loss": 3.7443, + "step": 6467 + }, + { + "epoch": 0.34042105263157896, + "grad_norm": 0.6015625, + "learning_rate": 0.0001465730994152047, + "loss": 3.3398, + "step": 6468 + }, + { + "epoch": 0.3404736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.00014656140350877194, + "loss": 3.8619, + "step": 6469 + }, + { + "epoch": 0.3405263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00014654970760233919, + "loss": 3.952, + "step": 6470 + }, + { + "epoch": 0.34057894736842104, + "grad_norm": 0.54296875, + "learning_rate": 0.00014653801169590644, + "loss": 3.3137, + "step": 6471 + }, + { + "epoch": 0.3406315789473684, + "grad_norm": 0.48828125, + "learning_rate": 0.00014652631578947369, + "loss": 3.7964, + "step": 6472 + }, + { + "epoch": 0.3406842105263158, + "grad_norm": 0.48046875, + "learning_rate": 0.00014651461988304093, + "loss": 3.5352, + "step": 6473 + }, + { + "epoch": 0.34073684210526317, + "grad_norm": 0.49609375, + "learning_rate": 0.00014650292397660818, + "loss": 3.3118, + "step": 6474 + }, + { + "epoch": 0.34078947368421053, + "grad_norm": 0.49609375, + "learning_rate": 0.00014649122807017543, + "loss": 3.5584, + "step": 6475 + }, + { + "epoch": 0.3408421052631579, + "grad_norm": 0.5859375, + "learning_rate": 0.0001464795321637427, + "loss": 3.7903, + "step": 6476 + }, + { + "epoch": 0.34089473684210525, + "grad_norm": 0.56640625, + "learning_rate": 0.00014646783625730996, + "loss": 4.0741, + "step": 6477 + }, + { + "epoch": 0.3409473684210526, + "grad_norm": 0.48046875, + "learning_rate": 0.0001464561403508772, + "loss": 3.6916, + "step": 6478 + }, + { + "epoch": 0.341, + "grad_norm": 0.494140625, + "learning_rate": 0.00014644444444444446, + "loss": 3.4206, + "step": 6479 + }, + { + "epoch": 0.3410526315789474, + "grad_norm": 0.46875, + "learning_rate": 0.0001464327485380117, + "loss": 3.9321, + "step": 6480 + }, + { + "epoch": 0.34110526315789474, + "grad_norm": 0.5625, + "learning_rate": 0.00014642105263157896, + "loss": 3.7141, + "step": 6481 + }, + { + "epoch": 0.3411578947368421, + "grad_norm": 0.474609375, + "learning_rate": 0.0001464093567251462, + "loss": 3.5496, + "step": 6482 + }, + { + "epoch": 0.34121052631578946, + "grad_norm": 0.47265625, + "learning_rate": 0.00014639766081871346, + "loss": 3.8536, + "step": 6483 + }, + { + "epoch": 0.3412631578947368, + "grad_norm": 0.5390625, + "learning_rate": 0.0001463859649122807, + "loss": 4.1407, + "step": 6484 + }, + { + "epoch": 0.3413157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.00014637426900584796, + "loss": 4.0302, + "step": 6485 + }, + { + "epoch": 0.3413684210526316, + "grad_norm": 0.494140625, + "learning_rate": 0.0001463625730994152, + "loss": 3.7594, + "step": 6486 + }, + { + "epoch": 0.34142105263157896, + "grad_norm": 0.53125, + "learning_rate": 0.00014635087719298246, + "loss": 3.4902, + "step": 6487 + }, + { + "epoch": 0.3414736842105263, + "grad_norm": 0.474609375, + "learning_rate": 0.0001463391812865497, + "loss": 3.2757, + "step": 6488 + }, + { + "epoch": 0.3415263157894737, + "grad_norm": 0.515625, + "learning_rate": 0.00014632748538011696, + "loss": 3.648, + "step": 6489 + }, + { + "epoch": 0.34157894736842104, + "grad_norm": 0.50390625, + "learning_rate": 0.00014631578947368423, + "loss": 3.778, + "step": 6490 + }, + { + "epoch": 0.3416315789473684, + "grad_norm": 0.52734375, + "learning_rate": 0.00014630409356725148, + "loss": 3.9493, + "step": 6491 + }, + { + "epoch": 0.3416842105263158, + "grad_norm": 0.53515625, + "learning_rate": 0.00014629239766081873, + "loss": 3.5389, + "step": 6492 + }, + { + "epoch": 0.34173684210526317, + "grad_norm": 0.53515625, + "learning_rate": 0.00014628070175438598, + "loss": 3.7871, + "step": 6493 + }, + { + "epoch": 0.34178947368421053, + "grad_norm": 0.53125, + "learning_rate": 0.00014626900584795323, + "loss": 3.6021, + "step": 6494 + }, + { + "epoch": 0.3418421052631579, + "grad_norm": 0.455078125, + "learning_rate": 0.00014625730994152048, + "loss": 2.8236, + "step": 6495 + }, + { + "epoch": 0.34189473684210525, + "grad_norm": 0.48046875, + "learning_rate": 0.00014624561403508773, + "loss": 3.582, + "step": 6496 + }, + { + "epoch": 0.3419473684210526, + "grad_norm": 0.47265625, + "learning_rate": 0.00014623391812865498, + "loss": 3.8116, + "step": 6497 + }, + { + "epoch": 0.342, + "grad_norm": 0.50390625, + "learning_rate": 0.00014622222222222223, + "loss": 3.568, + "step": 6498 + }, + { + "epoch": 0.3420526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.00014621052631578948, + "loss": 3.9533, + "step": 6499 + }, + { + "epoch": 0.34210526315789475, + "grad_norm": 0.55859375, + "learning_rate": 0.00014619883040935673, + "loss": 3.4693, + "step": 6500 + }, + { + "epoch": 0.3421578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.00014618713450292398, + "loss": 3.5425, + "step": 6501 + }, + { + "epoch": 0.34221052631578947, + "grad_norm": 0.482421875, + "learning_rate": 0.00014617543859649123, + "loss": 3.4228, + "step": 6502 + }, + { + "epoch": 0.3422631578947368, + "grad_norm": 0.4609375, + "learning_rate": 0.00014616374269005848, + "loss": 3.5885, + "step": 6503 + }, + { + "epoch": 0.3423157894736842, + "grad_norm": 0.486328125, + "learning_rate": 0.00014615204678362573, + "loss": 3.7292, + "step": 6504 + }, + { + "epoch": 0.3423684210526316, + "grad_norm": 0.494140625, + "learning_rate": 0.000146140350877193, + "loss": 3.5953, + "step": 6505 + }, + { + "epoch": 0.34242105263157896, + "grad_norm": 0.54296875, + "learning_rate": 0.00014612865497076025, + "loss": 3.5284, + "step": 6506 + }, + { + "epoch": 0.3424736842105263, + "grad_norm": 0.5390625, + "learning_rate": 0.0001461169590643275, + "loss": 3.9332, + "step": 6507 + }, + { + "epoch": 0.3425263157894737, + "grad_norm": 0.55078125, + "learning_rate": 0.00014610526315789475, + "loss": 3.5283, + "step": 6508 + }, + { + "epoch": 0.34257894736842104, + "grad_norm": 0.490234375, + "learning_rate": 0.000146093567251462, + "loss": 4.0925, + "step": 6509 + }, + { + "epoch": 0.3426315789473684, + "grad_norm": 0.5234375, + "learning_rate": 0.00014608187134502925, + "loss": 3.8813, + "step": 6510 + }, + { + "epoch": 0.3426842105263158, + "grad_norm": 0.49609375, + "learning_rate": 0.0001460701754385965, + "loss": 3.155, + "step": 6511 + }, + { + "epoch": 0.3427368421052632, + "grad_norm": 0.5078125, + "learning_rate": 0.00014605847953216375, + "loss": 3.4898, + "step": 6512 + }, + { + "epoch": 0.34278947368421053, + "grad_norm": 0.4921875, + "learning_rate": 0.000146046783625731, + "loss": 3.6525, + "step": 6513 + }, + { + "epoch": 0.3428421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00014603508771929825, + "loss": 3.2218, + "step": 6514 + }, + { + "epoch": 0.34289473684210525, + "grad_norm": 0.55078125, + "learning_rate": 0.0001460233918128655, + "loss": 4.0203, + "step": 6515 + }, + { + "epoch": 0.3429473684210526, + "grad_norm": 0.48828125, + "learning_rate": 0.00014601169590643275, + "loss": 3.716, + "step": 6516 + }, + { + "epoch": 0.343, + "grad_norm": 0.578125, + "learning_rate": 0.000146, + "loss": 3.8469, + "step": 6517 + }, + { + "epoch": 0.3430526315789474, + "grad_norm": 0.61328125, + "learning_rate": 0.00014598830409356725, + "loss": 3.8025, + "step": 6518 + }, + { + "epoch": 0.34310526315789475, + "grad_norm": 0.5703125, + "learning_rate": 0.0001459766081871345, + "loss": 3.5393, + "step": 6519 + }, + { + "epoch": 0.3431578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.00014596491228070177, + "loss": 3.7586, + "step": 6520 + }, + { + "epoch": 0.34321052631578947, + "grad_norm": 0.6796875, + "learning_rate": 0.00014595321637426902, + "loss": 3.2485, + "step": 6521 + }, + { + "epoch": 0.3432631578947368, + "grad_norm": 0.66015625, + "learning_rate": 0.00014594152046783627, + "loss": 3.7449, + "step": 6522 + }, + { + "epoch": 0.3433157894736842, + "grad_norm": 0.78125, + "learning_rate": 0.00014592982456140352, + "loss": 3.7275, + "step": 6523 + }, + { + "epoch": 0.3433684210526316, + "grad_norm": 0.6640625, + "learning_rate": 0.00014591812865497077, + "loss": 3.808, + "step": 6524 + }, + { + "epoch": 0.34342105263157896, + "grad_norm": 0.546875, + "learning_rate": 0.00014590643274853802, + "loss": 3.4621, + "step": 6525 + }, + { + "epoch": 0.3434736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.00014589473684210527, + "loss": 3.6952, + "step": 6526 + }, + { + "epoch": 0.3435263157894737, + "grad_norm": 0.78515625, + "learning_rate": 0.00014588304093567252, + "loss": 3.6584, + "step": 6527 + }, + { + "epoch": 0.34357894736842104, + "grad_norm": 0.61328125, + "learning_rate": 0.00014587134502923977, + "loss": 3.28, + "step": 6528 + }, + { + "epoch": 0.3436315789473684, + "grad_norm": 0.57421875, + "learning_rate": 0.00014585964912280702, + "loss": 3.8219, + "step": 6529 + }, + { + "epoch": 0.3436842105263158, + "grad_norm": 0.58984375, + "learning_rate": 0.00014584795321637427, + "loss": 3.98, + "step": 6530 + }, + { + "epoch": 0.3437368421052632, + "grad_norm": 0.55859375, + "learning_rate": 0.00014583625730994152, + "loss": 3.6009, + "step": 6531 + }, + { + "epoch": 0.34378947368421053, + "grad_norm": 0.5625, + "learning_rate": 0.00014582456140350877, + "loss": 3.5331, + "step": 6532 + }, + { + "epoch": 0.3438421052631579, + "grad_norm": 0.474609375, + "learning_rate": 0.00014581286549707602, + "loss": 3.6582, + "step": 6533 + }, + { + "epoch": 0.34389473684210525, + "grad_norm": 0.482421875, + "learning_rate": 0.00014580116959064327, + "loss": 3.3868, + "step": 6534 + }, + { + "epoch": 0.3439473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00014578947368421054, + "loss": 3.8616, + "step": 6535 + }, + { + "epoch": 0.344, + "grad_norm": 0.5234375, + "learning_rate": 0.0001457777777777778, + "loss": 3.7469, + "step": 6536 + }, + { + "epoch": 0.3440526315789474, + "grad_norm": 0.5859375, + "learning_rate": 0.00014576608187134504, + "loss": 3.539, + "step": 6537 + }, + { + "epoch": 0.34410526315789475, + "grad_norm": 0.5625, + "learning_rate": 0.0001457543859649123, + "loss": 3.7048, + "step": 6538 + }, + { + "epoch": 0.3441578947368421, + "grad_norm": 0.58203125, + "learning_rate": 0.00014574269005847954, + "loss": 3.8843, + "step": 6539 + }, + { + "epoch": 0.34421052631578947, + "grad_norm": 0.56640625, + "learning_rate": 0.0001457309941520468, + "loss": 3.48, + "step": 6540 + }, + { + "epoch": 0.3442631578947368, + "grad_norm": 0.54296875, + "learning_rate": 0.00014571929824561404, + "loss": 4.0346, + "step": 6541 + }, + { + "epoch": 0.3443157894736842, + "grad_norm": 0.546875, + "learning_rate": 0.0001457076023391813, + "loss": 3.661, + "step": 6542 + }, + { + "epoch": 0.3443684210526316, + "grad_norm": 0.51953125, + "learning_rate": 0.00014569590643274854, + "loss": 3.5772, + "step": 6543 + }, + { + "epoch": 0.34442105263157896, + "grad_norm": 0.482421875, + "learning_rate": 0.0001456842105263158, + "loss": 3.6531, + "step": 6544 + }, + { + "epoch": 0.3444736842105263, + "grad_norm": 0.671875, + "learning_rate": 0.00014567251461988304, + "loss": 3.7349, + "step": 6545 + }, + { + "epoch": 0.3445263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.0001456608187134503, + "loss": 3.6892, + "step": 6546 + }, + { + "epoch": 0.34457894736842104, + "grad_norm": 0.51953125, + "learning_rate": 0.00014564912280701754, + "loss": 3.3668, + "step": 6547 + }, + { + "epoch": 0.3446315789473684, + "grad_norm": 0.52734375, + "learning_rate": 0.0001456374269005848, + "loss": 3.8874, + "step": 6548 + }, + { + "epoch": 0.3446842105263158, + "grad_norm": 0.51953125, + "learning_rate": 0.00014562573099415204, + "loss": 3.2098, + "step": 6549 + }, + { + "epoch": 0.3447368421052632, + "grad_norm": 0.5390625, + "learning_rate": 0.00014561403508771932, + "loss": 3.8098, + "step": 6550 + }, + { + "epoch": 0.34478947368421053, + "grad_norm": 0.62109375, + "learning_rate": 0.00014560233918128657, + "loss": 3.412, + "step": 6551 + }, + { + "epoch": 0.3448421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.00014559064327485382, + "loss": 3.7139, + "step": 6552 + }, + { + "epoch": 0.34489473684210525, + "grad_norm": 0.56640625, + "learning_rate": 0.00014557894736842106, + "loss": 3.5396, + "step": 6553 + }, + { + "epoch": 0.3449473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00014556725146198831, + "loss": 3.8975, + "step": 6554 + }, + { + "epoch": 0.345, + "grad_norm": 0.53515625, + "learning_rate": 0.00014555555555555556, + "loss": 3.8757, + "step": 6555 + }, + { + "epoch": 0.3450526315789474, + "grad_norm": 0.53515625, + "learning_rate": 0.0001455438596491228, + "loss": 3.7981, + "step": 6556 + }, + { + "epoch": 0.34510526315789475, + "grad_norm": 0.515625, + "learning_rate": 0.00014553216374269006, + "loss": 3.5609, + "step": 6557 + }, + { + "epoch": 0.3451578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.0001455204678362573, + "loss": 3.7887, + "step": 6558 + }, + { + "epoch": 0.34521052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00014550877192982456, + "loss": 3.4794, + "step": 6559 + }, + { + "epoch": 0.3452631578947368, + "grad_norm": 0.52734375, + "learning_rate": 0.0001454970760233918, + "loss": 3.6857, + "step": 6560 + }, + { + "epoch": 0.3453157894736842, + "grad_norm": 0.5625, + "learning_rate": 0.00014548538011695906, + "loss": 3.6851, + "step": 6561 + }, + { + "epoch": 0.3453684210526316, + "grad_norm": 0.546875, + "learning_rate": 0.0001454736842105263, + "loss": 3.2009, + "step": 6562 + }, + { + "epoch": 0.34542105263157896, + "grad_norm": 0.49609375, + "learning_rate": 0.00014546198830409356, + "loss": 3.7131, + "step": 6563 + }, + { + "epoch": 0.3454736842105263, + "grad_norm": 0.45703125, + "learning_rate": 0.00014545029239766084, + "loss": 3.5622, + "step": 6564 + }, + { + "epoch": 0.3455263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.0001454385964912281, + "loss": 3.5869, + "step": 6565 + }, + { + "epoch": 0.34557894736842104, + "grad_norm": 0.57421875, + "learning_rate": 0.00014542690058479534, + "loss": 3.6209, + "step": 6566 + }, + { + "epoch": 0.3456315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.00014541520467836259, + "loss": 3.4126, + "step": 6567 + }, + { + "epoch": 0.3456842105263158, + "grad_norm": 0.70703125, + "learning_rate": 0.00014540350877192984, + "loss": 3.3648, + "step": 6568 + }, + { + "epoch": 0.3457368421052632, + "grad_norm": 0.484375, + "learning_rate": 0.00014539181286549709, + "loss": 4.178, + "step": 6569 + }, + { + "epoch": 0.34578947368421054, + "grad_norm": 0.47265625, + "learning_rate": 0.00014538011695906433, + "loss": 3.8161, + "step": 6570 + }, + { + "epoch": 0.3458421052631579, + "grad_norm": 0.65625, + "learning_rate": 0.00014536842105263158, + "loss": 3.5434, + "step": 6571 + }, + { + "epoch": 0.34589473684210525, + "grad_norm": 0.50390625, + "learning_rate": 0.00014535672514619883, + "loss": 3.8216, + "step": 6572 + }, + { + "epoch": 0.3459473684210526, + "grad_norm": 0.5234375, + "learning_rate": 0.00014534502923976608, + "loss": 3.9291, + "step": 6573 + }, + { + "epoch": 0.346, + "grad_norm": 0.52734375, + "learning_rate": 0.00014533333333333333, + "loss": 3.766, + "step": 6574 + }, + { + "epoch": 0.3460526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.00014532163742690058, + "loss": 3.6054, + "step": 6575 + }, + { + "epoch": 0.34610526315789475, + "grad_norm": 0.5546875, + "learning_rate": 0.00014530994152046783, + "loss": 3.1113, + "step": 6576 + }, + { + "epoch": 0.3461578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00014529824561403508, + "loss": 3.5871, + "step": 6577 + }, + { + "epoch": 0.34621052631578947, + "grad_norm": 0.462890625, + "learning_rate": 0.00014528654970760233, + "loss": 3.9304, + "step": 6578 + }, + { + "epoch": 0.34626315789473683, + "grad_norm": 0.498046875, + "learning_rate": 0.0001452748538011696, + "loss": 3.8727, + "step": 6579 + }, + { + "epoch": 0.3463157894736842, + "grad_norm": 0.58203125, + "learning_rate": 0.00014526315789473686, + "loss": 3.7462, + "step": 6580 + }, + { + "epoch": 0.3463684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.0001452514619883041, + "loss": 3.4097, + "step": 6581 + }, + { + "epoch": 0.34642105263157896, + "grad_norm": 0.7421875, + "learning_rate": 0.00014523976608187136, + "loss": 3.8653, + "step": 6582 + }, + { + "epoch": 0.3464736842105263, + "grad_norm": 0.56640625, + "learning_rate": 0.0001452280701754386, + "loss": 3.369, + "step": 6583 + }, + { + "epoch": 0.3465263157894737, + "grad_norm": 0.458984375, + "learning_rate": 0.00014521637426900586, + "loss": 3.9046, + "step": 6584 + }, + { + "epoch": 0.34657894736842104, + "grad_norm": 0.453125, + "learning_rate": 0.0001452046783625731, + "loss": 3.725, + "step": 6585 + }, + { + "epoch": 0.3466315789473684, + "grad_norm": 0.466796875, + "learning_rate": 0.00014519298245614036, + "loss": 4.1091, + "step": 6586 + }, + { + "epoch": 0.3466842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.0001451812865497076, + "loss": 3.5048, + "step": 6587 + }, + { + "epoch": 0.3467368421052632, + "grad_norm": 0.58203125, + "learning_rate": 0.00014516959064327485, + "loss": 3.4052, + "step": 6588 + }, + { + "epoch": 0.34678947368421054, + "grad_norm": 0.494140625, + "learning_rate": 0.0001451578947368421, + "loss": 3.0964, + "step": 6589 + }, + { + "epoch": 0.3468421052631579, + "grad_norm": 0.47265625, + "learning_rate": 0.00014514619883040935, + "loss": 3.4049, + "step": 6590 + }, + { + "epoch": 0.34689473684210526, + "grad_norm": 0.453125, + "learning_rate": 0.0001451345029239766, + "loss": 4.1653, + "step": 6591 + }, + { + "epoch": 0.3469473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00014512280701754385, + "loss": 3.6677, + "step": 6592 + }, + { + "epoch": 0.347, + "grad_norm": 0.5625, + "learning_rate": 0.0001451111111111111, + "loss": 3.9807, + "step": 6593 + }, + { + "epoch": 0.3470526315789474, + "grad_norm": 0.953125, + "learning_rate": 0.00014509941520467838, + "loss": 3.6824, + "step": 6594 + }, + { + "epoch": 0.34710526315789475, + "grad_norm": 0.46875, + "learning_rate": 0.00014508771929824563, + "loss": 4.1329, + "step": 6595 + }, + { + "epoch": 0.3471578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.00014507602339181288, + "loss": 3.3447, + "step": 6596 + }, + { + "epoch": 0.34721052631578947, + "grad_norm": 0.515625, + "learning_rate": 0.00014506432748538013, + "loss": 3.7162, + "step": 6597 + }, + { + "epoch": 0.34726315789473683, + "grad_norm": 0.625, + "learning_rate": 0.00014505263157894738, + "loss": 3.673, + "step": 6598 + }, + { + "epoch": 0.3473157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00014504093567251463, + "loss": 3.5048, + "step": 6599 + }, + { + "epoch": 0.3473684210526316, + "grad_norm": 0.5546875, + "learning_rate": 0.00014502923976608188, + "loss": 3.8893, + "step": 6600 + }, + { + "epoch": 0.34742105263157896, + "grad_norm": 0.55859375, + "learning_rate": 0.00014501754385964913, + "loss": 4.0435, + "step": 6601 + }, + { + "epoch": 0.3474736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00014500584795321638, + "loss": 3.7916, + "step": 6602 + }, + { + "epoch": 0.3475263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00014499415204678363, + "loss": 3.8095, + "step": 6603 + }, + { + "epoch": 0.34757894736842104, + "grad_norm": 0.51171875, + "learning_rate": 0.00014498245614035088, + "loss": 3.6807, + "step": 6604 + }, + { + "epoch": 0.3476315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.00014497076023391812, + "loss": 3.4756, + "step": 6605 + }, + { + "epoch": 0.34768421052631576, + "grad_norm": 0.55078125, + "learning_rate": 0.00014495906432748537, + "loss": 3.2179, + "step": 6606 + }, + { + "epoch": 0.3477368421052632, + "grad_norm": 0.5, + "learning_rate": 0.00014494736842105262, + "loss": 3.9134, + "step": 6607 + }, + { + "epoch": 0.34778947368421054, + "grad_norm": 0.478515625, + "learning_rate": 0.00014493567251461987, + "loss": 3.7737, + "step": 6608 + }, + { + "epoch": 0.3478421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00014492397660818715, + "loss": 3.8469, + "step": 6609 + }, + { + "epoch": 0.34789473684210526, + "grad_norm": 0.494140625, + "learning_rate": 0.0001449122807017544, + "loss": 3.879, + "step": 6610 + }, + { + "epoch": 0.3479473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00014490058479532165, + "loss": 3.6229, + "step": 6611 + }, + { + "epoch": 0.348, + "grad_norm": 0.49609375, + "learning_rate": 0.0001448888888888889, + "loss": 3.5194, + "step": 6612 + }, + { + "epoch": 0.3480526315789474, + "grad_norm": 0.59765625, + "learning_rate": 0.00014487719298245615, + "loss": 3.7464, + "step": 6613 + }, + { + "epoch": 0.34810526315789475, + "grad_norm": 0.578125, + "learning_rate": 0.0001448654970760234, + "loss": 3.8533, + "step": 6614 + }, + { + "epoch": 0.3481578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.00014485380116959065, + "loss": 3.8968, + "step": 6615 + }, + { + "epoch": 0.34821052631578947, + "grad_norm": 0.515625, + "learning_rate": 0.00014484210526315792, + "loss": 3.7862, + "step": 6616 + }, + { + "epoch": 0.34826315789473683, + "grad_norm": 0.51171875, + "learning_rate": 0.00014483040935672517, + "loss": 3.566, + "step": 6617 + }, + { + "epoch": 0.3483157894736842, + "grad_norm": 0.5234375, + "learning_rate": 0.0001448187134502924, + "loss": 3.4155, + "step": 6618 + }, + { + "epoch": 0.3483684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.00014480701754385965, + "loss": 3.5425, + "step": 6619 + }, + { + "epoch": 0.34842105263157896, + "grad_norm": 0.58984375, + "learning_rate": 0.0001447953216374269, + "loss": 3.5744, + "step": 6620 + }, + { + "epoch": 0.3484736842105263, + "grad_norm": 0.59375, + "learning_rate": 0.00014478362573099415, + "loss": 3.5601, + "step": 6621 + }, + { + "epoch": 0.3485263157894737, + "grad_norm": 0.5625, + "learning_rate": 0.0001447719298245614, + "loss": 3.842, + "step": 6622 + }, + { + "epoch": 0.34857894736842104, + "grad_norm": 0.52734375, + "learning_rate": 0.00014476023391812864, + "loss": 4.1471, + "step": 6623 + }, + { + "epoch": 0.3486315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.00014474853801169592, + "loss": 3.8945, + "step": 6624 + }, + { + "epoch": 0.34868421052631576, + "grad_norm": 0.52734375, + "learning_rate": 0.00014473684210526317, + "loss": 3.3385, + "step": 6625 + }, + { + "epoch": 0.3487368421052632, + "grad_norm": 0.474609375, + "learning_rate": 0.00014472514619883042, + "loss": 3.703, + "step": 6626 + }, + { + "epoch": 0.34878947368421054, + "grad_norm": 0.4921875, + "learning_rate": 0.00014471345029239767, + "loss": 3.614, + "step": 6627 + }, + { + "epoch": 0.3488421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00014470175438596492, + "loss": 3.7957, + "step": 6628 + }, + { + "epoch": 0.34889473684210526, + "grad_norm": 0.48828125, + "learning_rate": 0.00014469005847953217, + "loss": 3.4569, + "step": 6629 + }, + { + "epoch": 0.3489473684210526, + "grad_norm": 0.5390625, + "learning_rate": 0.00014467836257309942, + "loss": 3.3856, + "step": 6630 + }, + { + "epoch": 0.349, + "grad_norm": 0.45703125, + "learning_rate": 0.0001446666666666667, + "loss": 3.6326, + "step": 6631 + }, + { + "epoch": 0.3490526315789474, + "grad_norm": 0.482421875, + "learning_rate": 0.00014465497076023395, + "loss": 3.5385, + "step": 6632 + }, + { + "epoch": 0.34910526315789475, + "grad_norm": 0.51171875, + "learning_rate": 0.00014464327485380117, + "loss": 3.8348, + "step": 6633 + }, + { + "epoch": 0.3491578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00014463157894736842, + "loss": 3.9678, + "step": 6634 + }, + { + "epoch": 0.34921052631578947, + "grad_norm": 0.71484375, + "learning_rate": 0.00014461988304093567, + "loss": 3.9314, + "step": 6635 + }, + { + "epoch": 0.34926315789473683, + "grad_norm": 0.5078125, + "learning_rate": 0.00014460818713450292, + "loss": 3.886, + "step": 6636 + }, + { + "epoch": 0.3493157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00014459649122807017, + "loss": 3.7137, + "step": 6637 + }, + { + "epoch": 0.3493684210526316, + "grad_norm": 0.55078125, + "learning_rate": 0.00014458479532163742, + "loss": 3.5965, + "step": 6638 + }, + { + "epoch": 0.34942105263157897, + "grad_norm": 0.55078125, + "learning_rate": 0.0001445730994152047, + "loss": 3.8912, + "step": 6639 + }, + { + "epoch": 0.3494736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.00014456140350877194, + "loss": 3.9812, + "step": 6640 + }, + { + "epoch": 0.3495263157894737, + "grad_norm": 0.53515625, + "learning_rate": 0.0001445497076023392, + "loss": 3.5433, + "step": 6641 + }, + { + "epoch": 0.34957894736842104, + "grad_norm": 0.55859375, + "learning_rate": 0.00014453801169590644, + "loss": 3.1285, + "step": 6642 + }, + { + "epoch": 0.3496315789473684, + "grad_norm": 0.482421875, + "learning_rate": 0.0001445263157894737, + "loss": 3.221, + "step": 6643 + }, + { + "epoch": 0.34968421052631576, + "grad_norm": 0.546875, + "learning_rate": 0.00014451461988304094, + "loss": 3.7244, + "step": 6644 + }, + { + "epoch": 0.3497368421052632, + "grad_norm": 0.5234375, + "learning_rate": 0.0001445029239766082, + "loss": 4.1703, + "step": 6645 + }, + { + "epoch": 0.34978947368421054, + "grad_norm": 0.5703125, + "learning_rate": 0.00014449122807017547, + "loss": 3.3943, + "step": 6646 + }, + { + "epoch": 0.3498421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00014447953216374272, + "loss": 3.7392, + "step": 6647 + }, + { + "epoch": 0.34989473684210526, + "grad_norm": 0.57421875, + "learning_rate": 0.00014446783625730994, + "loss": 3.6994, + "step": 6648 + }, + { + "epoch": 0.3499473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.0001444561403508772, + "loss": 4.1344, + "step": 6649 + }, + { + "epoch": 0.35, + "grad_norm": 0.515625, + "learning_rate": 0.00014444444444444444, + "loss": 3.6124, + "step": 6650 + }, + { + "epoch": 0.3500526315789474, + "grad_norm": 0.494140625, + "learning_rate": 0.0001444327485380117, + "loss": 3.4642, + "step": 6651 + }, + { + "epoch": 0.35010526315789475, + "grad_norm": 0.462890625, + "learning_rate": 0.00014442105263157894, + "loss": 3.6859, + "step": 6652 + }, + { + "epoch": 0.3501578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00014440935672514621, + "loss": 3.4589, + "step": 6653 + }, + { + "epoch": 0.35021052631578947, + "grad_norm": 0.484375, + "learning_rate": 0.00014439766081871346, + "loss": 3.8594, + "step": 6654 + }, + { + "epoch": 0.35026315789473683, + "grad_norm": 0.49609375, + "learning_rate": 0.0001443859649122807, + "loss": 3.6228, + "step": 6655 + }, + { + "epoch": 0.3503157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00014437426900584796, + "loss": 3.7393, + "step": 6656 + }, + { + "epoch": 0.35036842105263155, + "grad_norm": 0.55859375, + "learning_rate": 0.0001443625730994152, + "loss": 3.8453, + "step": 6657 + }, + { + "epoch": 0.35042105263157897, + "grad_norm": 0.494140625, + "learning_rate": 0.00014435087719298246, + "loss": 3.5481, + "step": 6658 + }, + { + "epoch": 0.3504736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.0001443391812865497, + "loss": 3.4673, + "step": 6659 + }, + { + "epoch": 0.3505263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00014432748538011696, + "loss": 3.962, + "step": 6660 + }, + { + "epoch": 0.35057894736842105, + "grad_norm": 0.57421875, + "learning_rate": 0.00014431578947368424, + "loss": 3.4042, + "step": 6661 + }, + { + "epoch": 0.3506315789473684, + "grad_norm": 0.53125, + "learning_rate": 0.0001443040935672515, + "loss": 3.6064, + "step": 6662 + }, + { + "epoch": 0.35068421052631576, + "grad_norm": 0.53515625, + "learning_rate": 0.00014429239766081874, + "loss": 3.6669, + "step": 6663 + }, + { + "epoch": 0.3507368421052632, + "grad_norm": 0.58984375, + "learning_rate": 0.00014428070175438596, + "loss": 3.5169, + "step": 6664 + }, + { + "epoch": 0.35078947368421054, + "grad_norm": 0.53125, + "learning_rate": 0.0001442690058479532, + "loss": 3.6894, + "step": 6665 + }, + { + "epoch": 0.3508421052631579, + "grad_norm": 0.46484375, + "learning_rate": 0.00014425730994152046, + "loss": 4.0431, + "step": 6666 + }, + { + "epoch": 0.35089473684210526, + "grad_norm": 1.359375, + "learning_rate": 0.0001442456140350877, + "loss": 3.5618, + "step": 6667 + }, + { + "epoch": 0.3509473684210526, + "grad_norm": 0.5078125, + "learning_rate": 0.00014423391812865498, + "loss": 3.6085, + "step": 6668 + }, + { + "epoch": 0.351, + "grad_norm": 0.486328125, + "learning_rate": 0.00014422222222222223, + "loss": 4.0192, + "step": 6669 + }, + { + "epoch": 0.3510526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.00014421052631578948, + "loss": 3.4812, + "step": 6670 + }, + { + "epoch": 0.35110526315789475, + "grad_norm": 0.50390625, + "learning_rate": 0.00014419883040935673, + "loss": 3.2904, + "step": 6671 + }, + { + "epoch": 0.3511578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00014418713450292398, + "loss": 3.5169, + "step": 6672 + }, + { + "epoch": 0.3512105263157895, + "grad_norm": 0.498046875, + "learning_rate": 0.00014417543859649123, + "loss": 3.6194, + "step": 6673 + }, + { + "epoch": 0.35126315789473683, + "grad_norm": 0.494140625, + "learning_rate": 0.00014416374269005848, + "loss": 4.2345, + "step": 6674 + }, + { + "epoch": 0.3513157894736842, + "grad_norm": 0.53515625, + "learning_rate": 0.00014415204678362573, + "loss": 3.2949, + "step": 6675 + }, + { + "epoch": 0.35136842105263155, + "grad_norm": 0.515625, + "learning_rate": 0.000144140350877193, + "loss": 3.3763, + "step": 6676 + }, + { + "epoch": 0.35142105263157897, + "grad_norm": 0.46484375, + "learning_rate": 0.00014412865497076026, + "loss": 3.5992, + "step": 6677 + }, + { + "epoch": 0.3514736842105263, + "grad_norm": 0.51171875, + "learning_rate": 0.0001441169590643275, + "loss": 3.4989, + "step": 6678 + }, + { + "epoch": 0.3515263157894737, + "grad_norm": 0.6484375, + "learning_rate": 0.00014410526315789473, + "loss": 3.3917, + "step": 6679 + }, + { + "epoch": 0.35157894736842105, + "grad_norm": 0.46875, + "learning_rate": 0.00014409356725146198, + "loss": 3.6076, + "step": 6680 + }, + { + "epoch": 0.3516315789473684, + "grad_norm": 0.5234375, + "learning_rate": 0.00014408187134502923, + "loss": 3.271, + "step": 6681 + }, + { + "epoch": 0.35168421052631577, + "grad_norm": 0.46875, + "learning_rate": 0.00014407017543859648, + "loss": 3.5519, + "step": 6682 + }, + { + "epoch": 0.3517368421052632, + "grad_norm": 0.52734375, + "learning_rate": 0.00014405847953216376, + "loss": 3.7356, + "step": 6683 + }, + { + "epoch": 0.35178947368421054, + "grad_norm": 0.58203125, + "learning_rate": 0.000144046783625731, + "loss": 4.1347, + "step": 6684 + }, + { + "epoch": 0.3518421052631579, + "grad_norm": 0.7421875, + "learning_rate": 0.00014403508771929825, + "loss": 3.6515, + "step": 6685 + }, + { + "epoch": 0.35189473684210526, + "grad_norm": 0.470703125, + "learning_rate": 0.0001440233918128655, + "loss": 3.4855, + "step": 6686 + }, + { + "epoch": 0.3519473684210526, + "grad_norm": 0.46875, + "learning_rate": 0.00014401169590643275, + "loss": 3.8298, + "step": 6687 + }, + { + "epoch": 0.352, + "grad_norm": 0.47265625, + "learning_rate": 0.000144, + "loss": 3.9604, + "step": 6688 + }, + { + "epoch": 0.3520526315789474, + "grad_norm": 0.61328125, + "learning_rate": 0.00014398830409356725, + "loss": 3.2442, + "step": 6689 + }, + { + "epoch": 0.35210526315789475, + "grad_norm": 0.4921875, + "learning_rate": 0.00014397660818713453, + "loss": 3.7656, + "step": 6690 + }, + { + "epoch": 0.3521578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00014396491228070178, + "loss": 3.7797, + "step": 6691 + }, + { + "epoch": 0.3522105263157895, + "grad_norm": 0.58203125, + "learning_rate": 0.00014395321637426903, + "loss": 3.5547, + "step": 6692 + }, + { + "epoch": 0.35226315789473683, + "grad_norm": 0.578125, + "learning_rate": 0.00014394152046783628, + "loss": 3.6986, + "step": 6693 + }, + { + "epoch": 0.3523157894736842, + "grad_norm": 0.609375, + "learning_rate": 0.0001439298245614035, + "loss": 3.9236, + "step": 6694 + }, + { + "epoch": 0.35236842105263155, + "grad_norm": 0.6171875, + "learning_rate": 0.00014391812865497075, + "loss": 3.7761, + "step": 6695 + }, + { + "epoch": 0.35242105263157897, + "grad_norm": 0.58984375, + "learning_rate": 0.000143906432748538, + "loss": 3.5608, + "step": 6696 + }, + { + "epoch": 0.35247368421052633, + "grad_norm": 0.5078125, + "learning_rate": 0.00014389473684210525, + "loss": 3.7137, + "step": 6697 + }, + { + "epoch": 0.3525263157894737, + "grad_norm": 0.53515625, + "learning_rate": 0.00014388304093567253, + "loss": 3.8387, + "step": 6698 + }, + { + "epoch": 0.35257894736842105, + "grad_norm": 0.5234375, + "learning_rate": 0.00014387134502923978, + "loss": 3.5792, + "step": 6699 + }, + { + "epoch": 0.3526315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.00014385964912280703, + "loss": 3.6142, + "step": 6700 + }, + { + "epoch": 0.35268421052631577, + "grad_norm": 0.466796875, + "learning_rate": 0.00014384795321637428, + "loss": 4.1393, + "step": 6701 + }, + { + "epoch": 0.3527368421052632, + "grad_norm": 0.55078125, + "learning_rate": 0.00014383625730994153, + "loss": 3.362, + "step": 6702 + }, + { + "epoch": 0.35278947368421054, + "grad_norm": 0.52734375, + "learning_rate": 0.00014382456140350877, + "loss": 3.8338, + "step": 6703 + }, + { + "epoch": 0.3528421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00014381286549707602, + "loss": 3.8688, + "step": 6704 + }, + { + "epoch": 0.35289473684210526, + "grad_norm": 0.51953125, + "learning_rate": 0.0001438011695906433, + "loss": 3.4014, + "step": 6705 + }, + { + "epoch": 0.3529473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00014378947368421055, + "loss": 3.5464, + "step": 6706 + }, + { + "epoch": 0.353, + "grad_norm": 0.498046875, + "learning_rate": 0.0001437777777777778, + "loss": 3.933, + "step": 6707 + }, + { + "epoch": 0.3530526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00014376608187134505, + "loss": 3.5674, + "step": 6708 + }, + { + "epoch": 0.35310526315789476, + "grad_norm": 0.58984375, + "learning_rate": 0.0001437543859649123, + "loss": 3.8446, + "step": 6709 + }, + { + "epoch": 0.3531578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.00014374269005847952, + "loss": 3.6281, + "step": 6710 + }, + { + "epoch": 0.3532105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00014373099415204677, + "loss": 3.6792, + "step": 6711 + }, + { + "epoch": 0.35326315789473683, + "grad_norm": 0.515625, + "learning_rate": 0.00014371929824561402, + "loss": 3.7716, + "step": 6712 + }, + { + "epoch": 0.3533157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.0001437076023391813, + "loss": 3.6049, + "step": 6713 + }, + { + "epoch": 0.35336842105263155, + "grad_norm": 0.546875, + "learning_rate": 0.00014369590643274855, + "loss": 3.6918, + "step": 6714 + }, + { + "epoch": 0.35342105263157897, + "grad_norm": 0.5703125, + "learning_rate": 0.0001436842105263158, + "loss": 3.7652, + "step": 6715 + }, + { + "epoch": 0.35347368421052633, + "grad_norm": 0.5390625, + "learning_rate": 0.00014367251461988305, + "loss": 3.5358, + "step": 6716 + }, + { + "epoch": 0.3535263157894737, + "grad_norm": 0.515625, + "learning_rate": 0.0001436608187134503, + "loss": 3.2883, + "step": 6717 + }, + { + "epoch": 0.35357894736842105, + "grad_norm": 0.6171875, + "learning_rate": 0.00014364912280701755, + "loss": 3.5992, + "step": 6718 + }, + { + "epoch": 0.3536315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.0001436374269005848, + "loss": 3.5707, + "step": 6719 + }, + { + "epoch": 0.35368421052631577, + "grad_norm": 0.61328125, + "learning_rate": 0.00014362573099415207, + "loss": 3.7035, + "step": 6720 + }, + { + "epoch": 0.3537368421052632, + "grad_norm": 0.64453125, + "learning_rate": 0.00014361403508771932, + "loss": 3.9694, + "step": 6721 + }, + { + "epoch": 0.35378947368421054, + "grad_norm": 0.55859375, + "learning_rate": 0.00014360233918128657, + "loss": 4.0389, + "step": 6722 + }, + { + "epoch": 0.3538421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00014359064327485382, + "loss": 3.4924, + "step": 6723 + }, + { + "epoch": 0.35389473684210526, + "grad_norm": 0.46484375, + "learning_rate": 0.00014357894736842107, + "loss": 3.8095, + "step": 6724 + }, + { + "epoch": 0.3539473684210526, + "grad_norm": 0.546875, + "learning_rate": 0.0001435672514619883, + "loss": 3.2401, + "step": 6725 + }, + { + "epoch": 0.354, + "grad_norm": 0.466796875, + "learning_rate": 0.00014355555555555554, + "loss": 3.6011, + "step": 6726 + }, + { + "epoch": 0.35405263157894734, + "grad_norm": 0.443359375, + "learning_rate": 0.00014354385964912282, + "loss": 3.874, + "step": 6727 + }, + { + "epoch": 0.35410526315789476, + "grad_norm": 0.57421875, + "learning_rate": 0.00014353216374269007, + "loss": 3.8534, + "step": 6728 + }, + { + "epoch": 0.3541578947368421, + "grad_norm": 0.5625, + "learning_rate": 0.00014352046783625732, + "loss": 3.3698, + "step": 6729 + }, + { + "epoch": 0.3542105263157895, + "grad_norm": 0.55078125, + "learning_rate": 0.00014350877192982457, + "loss": 3.7998, + "step": 6730 + }, + { + "epoch": 0.35426315789473684, + "grad_norm": 0.494140625, + "learning_rate": 0.00014349707602339182, + "loss": 3.9075, + "step": 6731 + }, + { + "epoch": 0.3543157894736842, + "grad_norm": 0.482421875, + "learning_rate": 0.00014348538011695907, + "loss": 3.4688, + "step": 6732 + }, + { + "epoch": 0.35436842105263155, + "grad_norm": 0.46484375, + "learning_rate": 0.00014347368421052632, + "loss": 3.6688, + "step": 6733 + }, + { + "epoch": 0.35442105263157897, + "grad_norm": 0.51171875, + "learning_rate": 0.00014346198830409357, + "loss": 3.1972, + "step": 6734 + }, + { + "epoch": 0.35447368421052633, + "grad_norm": 0.51953125, + "learning_rate": 0.00014345029239766084, + "loss": 3.8238, + "step": 6735 + }, + { + "epoch": 0.3545263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.0001434385964912281, + "loss": 3.6414, + "step": 6736 + }, + { + "epoch": 0.35457894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.00014342690058479534, + "loss": 3.3896, + "step": 6737 + }, + { + "epoch": 0.3546315789473684, + "grad_norm": 0.48828125, + "learning_rate": 0.0001434152046783626, + "loss": 3.7705, + "step": 6738 + }, + { + "epoch": 0.35468421052631577, + "grad_norm": 0.6171875, + "learning_rate": 0.00014340350877192984, + "loss": 3.5769, + "step": 6739 + }, + { + "epoch": 0.3547368421052632, + "grad_norm": 0.6015625, + "learning_rate": 0.00014339181286549706, + "loss": 3.7951, + "step": 6740 + }, + { + "epoch": 0.35478947368421054, + "grad_norm": 0.494140625, + "learning_rate": 0.0001433801169590643, + "loss": 4.0083, + "step": 6741 + }, + { + "epoch": 0.3548421052631579, + "grad_norm": 0.546875, + "learning_rate": 0.0001433684210526316, + "loss": 3.3278, + "step": 6742 + }, + { + "epoch": 0.35489473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00014335672514619884, + "loss": 3.8184, + "step": 6743 + }, + { + "epoch": 0.3549473684210526, + "grad_norm": 0.490234375, + "learning_rate": 0.0001433450292397661, + "loss": 3.8827, + "step": 6744 + }, + { + "epoch": 0.355, + "grad_norm": 0.55859375, + "learning_rate": 0.00014333333333333334, + "loss": 3.7311, + "step": 6745 + }, + { + "epoch": 0.35505263157894734, + "grad_norm": 0.5234375, + "learning_rate": 0.0001433216374269006, + "loss": 3.6804, + "step": 6746 + }, + { + "epoch": 0.35510526315789476, + "grad_norm": 0.578125, + "learning_rate": 0.00014330994152046784, + "loss": 3.6784, + "step": 6747 + }, + { + "epoch": 0.3551578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.0001432982456140351, + "loss": 3.6237, + "step": 6748 + }, + { + "epoch": 0.3552105263157895, + "grad_norm": 0.50390625, + "learning_rate": 0.00014328654970760234, + "loss": 3.4076, + "step": 6749 + }, + { + "epoch": 0.35526315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.00014327485380116961, + "loss": 3.78, + "step": 6750 + }, + { + "epoch": 0.3553157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00014326315789473686, + "loss": 3.6293, + "step": 6751 + }, + { + "epoch": 0.35536842105263156, + "grad_norm": 0.55078125, + "learning_rate": 0.0001432514619883041, + "loss": 3.7533, + "step": 6752 + }, + { + "epoch": 0.35542105263157897, + "grad_norm": 0.5390625, + "learning_rate": 0.00014323976608187136, + "loss": 3.7949, + "step": 6753 + }, + { + "epoch": 0.35547368421052633, + "grad_norm": 0.578125, + "learning_rate": 0.0001432280701754386, + "loss": 3.3208, + "step": 6754 + }, + { + "epoch": 0.3555263157894737, + "grad_norm": 0.52734375, + "learning_rate": 0.00014321637426900586, + "loss": 4.1808, + "step": 6755 + }, + { + "epoch": 0.35557894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.00014320467836257308, + "loss": 3.8461, + "step": 6756 + }, + { + "epoch": 0.3556315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.00014319298245614036, + "loss": 3.7822, + "step": 6757 + }, + { + "epoch": 0.35568421052631577, + "grad_norm": 0.55859375, + "learning_rate": 0.0001431812865497076, + "loss": 3.6226, + "step": 6758 + }, + { + "epoch": 0.3557368421052632, + "grad_norm": 0.54296875, + "learning_rate": 0.00014316959064327486, + "loss": 3.7734, + "step": 6759 + }, + { + "epoch": 0.35578947368421054, + "grad_norm": 0.56640625, + "learning_rate": 0.0001431578947368421, + "loss": 4.0922, + "step": 6760 + }, + { + "epoch": 0.3558421052631579, + "grad_norm": 0.55078125, + "learning_rate": 0.00014314619883040936, + "loss": 3.8156, + "step": 6761 + }, + { + "epoch": 0.35589473684210526, + "grad_norm": 0.515625, + "learning_rate": 0.0001431345029239766, + "loss": 4.0452, + "step": 6762 + }, + { + "epoch": 0.3559473684210526, + "grad_norm": 0.57421875, + "learning_rate": 0.00014312280701754386, + "loss": 3.8403, + "step": 6763 + }, + { + "epoch": 0.356, + "grad_norm": 0.48046875, + "learning_rate": 0.0001431111111111111, + "loss": 3.5771, + "step": 6764 + }, + { + "epoch": 0.35605263157894734, + "grad_norm": 0.48046875, + "learning_rate": 0.00014309941520467838, + "loss": 3.4526, + "step": 6765 + }, + { + "epoch": 0.35610526315789476, + "grad_norm": 0.52734375, + "learning_rate": 0.00014308771929824563, + "loss": 3.488, + "step": 6766 + }, + { + "epoch": 0.3561578947368421, + "grad_norm": 0.58984375, + "learning_rate": 0.00014307602339181288, + "loss": 3.3905, + "step": 6767 + }, + { + "epoch": 0.3562105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00014306432748538013, + "loss": 3.926, + "step": 6768 + }, + { + "epoch": 0.35626315789473684, + "grad_norm": 0.5703125, + "learning_rate": 0.00014305263157894738, + "loss": 3.3928, + "step": 6769 + }, + { + "epoch": 0.3563157894736842, + "grad_norm": 0.53515625, + "learning_rate": 0.00014304093567251463, + "loss": 3.5052, + "step": 6770 + }, + { + "epoch": 0.35636842105263156, + "grad_norm": 0.47265625, + "learning_rate": 0.00014302923976608186, + "loss": 3.644, + "step": 6771 + }, + { + "epoch": 0.35642105263157897, + "grad_norm": 0.482421875, + "learning_rate": 0.00014301754385964913, + "loss": 3.5389, + "step": 6772 + }, + { + "epoch": 0.35647368421052633, + "grad_norm": 0.482421875, + "learning_rate": 0.00014300584795321638, + "loss": 3.2508, + "step": 6773 + }, + { + "epoch": 0.3565263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.00014299415204678363, + "loss": 4.0143, + "step": 6774 + }, + { + "epoch": 0.35657894736842105, + "grad_norm": 0.64453125, + "learning_rate": 0.00014298245614035088, + "loss": 3.499, + "step": 6775 + }, + { + "epoch": 0.3566315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00014297076023391813, + "loss": 3.6066, + "step": 6776 + }, + { + "epoch": 0.35668421052631577, + "grad_norm": 0.494140625, + "learning_rate": 0.00014295906432748538, + "loss": 3.9042, + "step": 6777 + }, + { + "epoch": 0.3567368421052632, + "grad_norm": 0.515625, + "learning_rate": 0.00014294736842105263, + "loss": 3.9295, + "step": 6778 + }, + { + "epoch": 0.35678947368421055, + "grad_norm": 0.58203125, + "learning_rate": 0.0001429356725146199, + "loss": 3.9541, + "step": 6779 + }, + { + "epoch": 0.3568421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00014292397660818716, + "loss": 3.4083, + "step": 6780 + }, + { + "epoch": 0.35689473684210526, + "grad_norm": 0.5078125, + "learning_rate": 0.0001429122807017544, + "loss": 3.7231, + "step": 6781 + }, + { + "epoch": 0.3569473684210526, + "grad_norm": 0.478515625, + "learning_rate": 0.00014290058479532166, + "loss": 3.6541, + "step": 6782 + }, + { + "epoch": 0.357, + "grad_norm": 0.486328125, + "learning_rate": 0.0001428888888888889, + "loss": 3.661, + "step": 6783 + }, + { + "epoch": 0.35705263157894734, + "grad_norm": 0.46875, + "learning_rate": 0.00014287719298245615, + "loss": 4.0908, + "step": 6784 + }, + { + "epoch": 0.35710526315789476, + "grad_norm": 0.515625, + "learning_rate": 0.0001428654970760234, + "loss": 3.5715, + "step": 6785 + }, + { + "epoch": 0.3571578947368421, + "grad_norm": 0.498046875, + "learning_rate": 0.00014285380116959063, + "loss": 3.5774, + "step": 6786 + }, + { + "epoch": 0.3572105263157895, + "grad_norm": 0.490234375, + "learning_rate": 0.0001428421052631579, + "loss": 3.5947, + "step": 6787 + }, + { + "epoch": 0.35726315789473684, + "grad_norm": 0.6015625, + "learning_rate": 0.00014283040935672515, + "loss": 3.3953, + "step": 6788 + }, + { + "epoch": 0.3573157894736842, + "grad_norm": 0.55859375, + "learning_rate": 0.0001428187134502924, + "loss": 3.595, + "step": 6789 + }, + { + "epoch": 0.35736842105263156, + "grad_norm": 0.53125, + "learning_rate": 0.00014280701754385965, + "loss": 4.1215, + "step": 6790 + }, + { + "epoch": 0.357421052631579, + "grad_norm": 0.46484375, + "learning_rate": 0.0001427953216374269, + "loss": 3.3766, + "step": 6791 + }, + { + "epoch": 0.35747368421052633, + "grad_norm": 0.48046875, + "learning_rate": 0.00014278362573099415, + "loss": 3.6486, + "step": 6792 + }, + { + "epoch": 0.3575263157894737, + "grad_norm": 0.55078125, + "learning_rate": 0.0001427719298245614, + "loss": 3.638, + "step": 6793 + }, + { + "epoch": 0.35757894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.00014276023391812868, + "loss": 3.7527, + "step": 6794 + }, + { + "epoch": 0.3576315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00014274853801169593, + "loss": 4.1428, + "step": 6795 + }, + { + "epoch": 0.35768421052631577, + "grad_norm": 0.55859375, + "learning_rate": 0.00014273684210526318, + "loss": 3.6461, + "step": 6796 + }, + { + "epoch": 0.35773684210526313, + "grad_norm": 0.5234375, + "learning_rate": 0.00014272514619883043, + "loss": 3.4997, + "step": 6797 + }, + { + "epoch": 0.35778947368421055, + "grad_norm": 0.482421875, + "learning_rate": 0.00014271345029239768, + "loss": 3.7079, + "step": 6798 + }, + { + "epoch": 0.3578421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00014270175438596493, + "loss": 3.6421, + "step": 6799 + }, + { + "epoch": 0.35789473684210527, + "grad_norm": 0.5078125, + "learning_rate": 0.00014269005847953217, + "loss": 3.78, + "step": 6800 + }, + { + "epoch": 0.3579473684210526, + "grad_norm": 0.515625, + "learning_rate": 0.00014267836257309942, + "loss": 2.9779, + "step": 6801 + }, + { + "epoch": 0.358, + "grad_norm": 0.50390625, + "learning_rate": 0.00014266666666666667, + "loss": 3.542, + "step": 6802 + }, + { + "epoch": 0.35805263157894734, + "grad_norm": 0.51171875, + "learning_rate": 0.00014265497076023392, + "loss": 3.6116, + "step": 6803 + }, + { + "epoch": 0.35810526315789476, + "grad_norm": 0.53515625, + "learning_rate": 0.00014264327485380117, + "loss": 3.4572, + "step": 6804 + }, + { + "epoch": 0.3581578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.00014263157894736842, + "loss": 3.9401, + "step": 6805 + }, + { + "epoch": 0.3582105263157895, + "grad_norm": 0.484375, + "learning_rate": 0.00014261988304093567, + "loss": 3.4015, + "step": 6806 + }, + { + "epoch": 0.35826315789473684, + "grad_norm": 0.482421875, + "learning_rate": 0.00014260818713450292, + "loss": 3.5621, + "step": 6807 + }, + { + "epoch": 0.3583157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00014259649122807017, + "loss": 3.3194, + "step": 6808 + }, + { + "epoch": 0.35836842105263156, + "grad_norm": 0.59375, + "learning_rate": 0.00014258479532163745, + "loss": 3.3884, + "step": 6809 + }, + { + "epoch": 0.358421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.0001425730994152047, + "loss": 4.0346, + "step": 6810 + }, + { + "epoch": 0.35847368421052633, + "grad_norm": 0.53125, + "learning_rate": 0.00014256140350877195, + "loss": 3.8227, + "step": 6811 + }, + { + "epoch": 0.3585263157894737, + "grad_norm": 0.498046875, + "learning_rate": 0.0001425497076023392, + "loss": 3.5589, + "step": 6812 + }, + { + "epoch": 0.35857894736842105, + "grad_norm": 0.546875, + "learning_rate": 0.00014253801169590645, + "loss": 3.7844, + "step": 6813 + }, + { + "epoch": 0.3586315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.0001425263157894737, + "loss": 3.6782, + "step": 6814 + }, + { + "epoch": 0.35868421052631577, + "grad_norm": 0.50390625, + "learning_rate": 0.00014251461988304095, + "loss": 3.7167, + "step": 6815 + }, + { + "epoch": 0.35873684210526313, + "grad_norm": 0.5078125, + "learning_rate": 0.0001425029239766082, + "loss": 3.3779, + "step": 6816 + }, + { + "epoch": 0.35878947368421055, + "grad_norm": 0.44921875, + "learning_rate": 0.00014249122807017545, + "loss": 3.7466, + "step": 6817 + }, + { + "epoch": 0.3588421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.0001424795321637427, + "loss": 3.5341, + "step": 6818 + }, + { + "epoch": 0.35889473684210527, + "grad_norm": 0.50390625, + "learning_rate": 0.00014246783625730994, + "loss": 3.6356, + "step": 6819 + }, + { + "epoch": 0.3589473684210526, + "grad_norm": 0.46875, + "learning_rate": 0.0001424561403508772, + "loss": 3.804, + "step": 6820 + }, + { + "epoch": 0.359, + "grad_norm": 0.484375, + "learning_rate": 0.00014244444444444444, + "loss": 3.8725, + "step": 6821 + }, + { + "epoch": 0.35905263157894735, + "grad_norm": 0.51953125, + "learning_rate": 0.0001424327485380117, + "loss": 3.4408, + "step": 6822 + }, + { + "epoch": 0.35910526315789476, + "grad_norm": 0.486328125, + "learning_rate": 0.00014242105263157894, + "loss": 3.7789, + "step": 6823 + }, + { + "epoch": 0.3591578947368421, + "grad_norm": 0.5859375, + "learning_rate": 0.00014240935672514622, + "loss": 3.7718, + "step": 6824 + }, + { + "epoch": 0.3592105263157895, + "grad_norm": 0.5546875, + "learning_rate": 0.00014239766081871347, + "loss": 3.6841, + "step": 6825 + }, + { + "epoch": 0.35926315789473684, + "grad_norm": 0.55078125, + "learning_rate": 0.00014238596491228072, + "loss": 3.9944, + "step": 6826 + }, + { + "epoch": 0.3593157894736842, + "grad_norm": 0.6015625, + "learning_rate": 0.00014237426900584797, + "loss": 3.6439, + "step": 6827 + }, + { + "epoch": 0.35936842105263156, + "grad_norm": 0.51953125, + "learning_rate": 0.00014236257309941522, + "loss": 3.3706, + "step": 6828 + }, + { + "epoch": 0.359421052631579, + "grad_norm": 0.46875, + "learning_rate": 0.00014235087719298247, + "loss": 3.6027, + "step": 6829 + }, + { + "epoch": 0.35947368421052633, + "grad_norm": 0.486328125, + "learning_rate": 0.00014233918128654972, + "loss": 3.5637, + "step": 6830 + }, + { + "epoch": 0.3595263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.00014232748538011697, + "loss": 3.6694, + "step": 6831 + }, + { + "epoch": 0.35957894736842105, + "grad_norm": 0.5, + "learning_rate": 0.00014231578947368422, + "loss": 3.3379, + "step": 6832 + }, + { + "epoch": 0.3596315789473684, + "grad_norm": 0.470703125, + "learning_rate": 0.00014230409356725147, + "loss": 3.3377, + "step": 6833 + }, + { + "epoch": 0.3596842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.00014229239766081872, + "loss": 3.4641, + "step": 6834 + }, + { + "epoch": 0.35973684210526313, + "grad_norm": 0.515625, + "learning_rate": 0.00014228070175438596, + "loss": 3.7496, + "step": 6835 + }, + { + "epoch": 0.35978947368421055, + "grad_norm": 0.5078125, + "learning_rate": 0.00014226900584795321, + "loss": 3.4443, + "step": 6836 + }, + { + "epoch": 0.3598421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00014225730994152046, + "loss": 3.4988, + "step": 6837 + }, + { + "epoch": 0.35989473684210527, + "grad_norm": 0.447265625, + "learning_rate": 0.00014224561403508771, + "loss": 3.8803, + "step": 6838 + }, + { + "epoch": 0.3599473684210526, + "grad_norm": 0.478515625, + "learning_rate": 0.000142233918128655, + "loss": 3.6913, + "step": 6839 + }, + { + "epoch": 0.36, + "grad_norm": 0.470703125, + "learning_rate": 0.00014222222222222224, + "loss": 3.9774, + "step": 6840 + }, + { + "epoch": 0.36005263157894735, + "grad_norm": 0.578125, + "learning_rate": 0.0001422105263157895, + "loss": 3.6468, + "step": 6841 + }, + { + "epoch": 0.36010526315789476, + "grad_norm": 0.57421875, + "learning_rate": 0.00014219883040935674, + "loss": 3.5979, + "step": 6842 + }, + { + "epoch": 0.3601578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.000142187134502924, + "loss": 3.8988, + "step": 6843 + }, + { + "epoch": 0.3602105263157895, + "grad_norm": 0.55078125, + "learning_rate": 0.00014217543859649124, + "loss": 3.4625, + "step": 6844 + }, + { + "epoch": 0.36026315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.0001421637426900585, + "loss": 3.6751, + "step": 6845 + }, + { + "epoch": 0.3603157894736842, + "grad_norm": 0.486328125, + "learning_rate": 0.00014215204678362574, + "loss": 3.5674, + "step": 6846 + }, + { + "epoch": 0.36036842105263156, + "grad_norm": 0.484375, + "learning_rate": 0.000142140350877193, + "loss": 3.6497, + "step": 6847 + }, + { + "epoch": 0.3604210526315789, + "grad_norm": 0.53515625, + "learning_rate": 0.00014212865497076024, + "loss": 3.9569, + "step": 6848 + }, + { + "epoch": 0.36047368421052634, + "grad_norm": 0.447265625, + "learning_rate": 0.00014211695906432749, + "loss": 3.6223, + "step": 6849 + }, + { + "epoch": 0.3605263157894737, + "grad_norm": 0.44921875, + "learning_rate": 0.00014210526315789474, + "loss": 3.5698, + "step": 6850 + }, + { + "epoch": 0.36057894736842105, + "grad_norm": 0.51171875, + "learning_rate": 0.00014209356725146199, + "loss": 3.5717, + "step": 6851 + }, + { + "epoch": 0.3606315789473684, + "grad_norm": 0.59375, + "learning_rate": 0.00014208187134502923, + "loss": 3.4498, + "step": 6852 + }, + { + "epoch": 0.3606842105263158, + "grad_norm": 0.474609375, + "learning_rate": 0.0001420701754385965, + "loss": 3.8068, + "step": 6853 + }, + { + "epoch": 0.36073684210526313, + "grad_norm": 0.48828125, + "learning_rate": 0.00014205847953216376, + "loss": 3.5215, + "step": 6854 + }, + { + "epoch": 0.36078947368421055, + "grad_norm": 0.54296875, + "learning_rate": 0.000142046783625731, + "loss": 3.4889, + "step": 6855 + }, + { + "epoch": 0.3608421052631579, + "grad_norm": 0.490234375, + "learning_rate": 0.00014203508771929826, + "loss": 3.3855, + "step": 6856 + }, + { + "epoch": 0.36089473684210527, + "grad_norm": 0.486328125, + "learning_rate": 0.0001420233918128655, + "loss": 3.4083, + "step": 6857 + }, + { + "epoch": 0.36094736842105263, + "grad_norm": 0.494140625, + "learning_rate": 0.00014201169590643276, + "loss": 3.6645, + "step": 6858 + }, + { + "epoch": 0.361, + "grad_norm": 0.546875, + "learning_rate": 0.000142, + "loss": 3.3228, + "step": 6859 + }, + { + "epoch": 0.36105263157894735, + "grad_norm": 0.5390625, + "learning_rate": 0.00014198830409356726, + "loss": 3.2281, + "step": 6860 + }, + { + "epoch": 0.36110526315789476, + "grad_norm": 0.462890625, + "learning_rate": 0.0001419766081871345, + "loss": 3.6752, + "step": 6861 + }, + { + "epoch": 0.3611578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00014196491228070176, + "loss": 3.6865, + "step": 6862 + }, + { + "epoch": 0.3612105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.000141953216374269, + "loss": 3.8781, + "step": 6863 + }, + { + "epoch": 0.36126315789473684, + "grad_norm": 0.62890625, + "learning_rate": 0.00014194152046783626, + "loss": 3.5668, + "step": 6864 + }, + { + "epoch": 0.3613157894736842, + "grad_norm": 0.5625, + "learning_rate": 0.0001419298245614035, + "loss": 3.6006, + "step": 6865 + }, + { + "epoch": 0.36136842105263156, + "grad_norm": 0.51171875, + "learning_rate": 0.00014191812865497076, + "loss": 3.5927, + "step": 6866 + }, + { + "epoch": 0.3614210526315789, + "grad_norm": 0.50390625, + "learning_rate": 0.000141906432748538, + "loss": 3.3804, + "step": 6867 + }, + { + "epoch": 0.36147368421052634, + "grad_norm": 0.53125, + "learning_rate": 0.00014189473684210528, + "loss": 3.7927, + "step": 6868 + }, + { + "epoch": 0.3615263157894737, + "grad_norm": 0.48828125, + "learning_rate": 0.00014188304093567253, + "loss": 3.6485, + "step": 6869 + }, + { + "epoch": 0.36157894736842106, + "grad_norm": 0.9375, + "learning_rate": 0.00014187134502923978, + "loss": 3.2506, + "step": 6870 + }, + { + "epoch": 0.3616315789473684, + "grad_norm": 0.4609375, + "learning_rate": 0.00014185964912280703, + "loss": 3.3485, + "step": 6871 + }, + { + "epoch": 0.3616842105263158, + "grad_norm": 0.49609375, + "learning_rate": 0.00014184795321637428, + "loss": 3.9797, + "step": 6872 + }, + { + "epoch": 0.36173684210526313, + "grad_norm": 0.640625, + "learning_rate": 0.00014183625730994153, + "loss": 3.7895, + "step": 6873 + }, + { + "epoch": 0.36178947368421055, + "grad_norm": 0.490234375, + "learning_rate": 0.00014182456140350878, + "loss": 3.8819, + "step": 6874 + }, + { + "epoch": 0.3618421052631579, + "grad_norm": 0.55078125, + "learning_rate": 0.00014181286549707603, + "loss": 3.3006, + "step": 6875 + }, + { + "epoch": 0.36189473684210527, + "grad_norm": 0.482421875, + "learning_rate": 0.00014180116959064328, + "loss": 3.6807, + "step": 6876 + }, + { + "epoch": 0.36194736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.00014178947368421053, + "loss": 3.6148, + "step": 6877 + }, + { + "epoch": 0.362, + "grad_norm": 0.4921875, + "learning_rate": 0.00014177777777777778, + "loss": 3.4842, + "step": 6878 + }, + { + "epoch": 0.36205263157894735, + "grad_norm": 0.5234375, + "learning_rate": 0.00014176608187134503, + "loss": 3.5113, + "step": 6879 + }, + { + "epoch": 0.36210526315789476, + "grad_norm": 0.490234375, + "learning_rate": 0.00014175438596491228, + "loss": 3.5754, + "step": 6880 + }, + { + "epoch": 0.3621578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.00014174269005847953, + "loss": 3.7496, + "step": 6881 + }, + { + "epoch": 0.3622105263157895, + "grad_norm": 0.515625, + "learning_rate": 0.00014173099415204678, + "loss": 3.7778, + "step": 6882 + }, + { + "epoch": 0.36226315789473684, + "grad_norm": 0.54296875, + "learning_rate": 0.00014171929824561405, + "loss": 3.6601, + "step": 6883 + }, + { + "epoch": 0.3623157894736842, + "grad_norm": 0.53515625, + "learning_rate": 0.0001417076023391813, + "loss": 3.6413, + "step": 6884 + }, + { + "epoch": 0.36236842105263156, + "grad_norm": 0.5078125, + "learning_rate": 0.00014169590643274855, + "loss": 3.7599, + "step": 6885 + }, + { + "epoch": 0.3624210526315789, + "grad_norm": 0.45703125, + "learning_rate": 0.0001416842105263158, + "loss": 2.9381, + "step": 6886 + }, + { + "epoch": 0.36247368421052634, + "grad_norm": 0.478515625, + "learning_rate": 0.00014167251461988305, + "loss": 3.7221, + "step": 6887 + }, + { + "epoch": 0.3625263157894737, + "grad_norm": 0.490234375, + "learning_rate": 0.0001416608187134503, + "loss": 3.2969, + "step": 6888 + }, + { + "epoch": 0.36257894736842106, + "grad_norm": 0.52734375, + "learning_rate": 0.00014164912280701755, + "loss": 3.5667, + "step": 6889 + }, + { + "epoch": 0.3626315789473684, + "grad_norm": 0.478515625, + "learning_rate": 0.0001416374269005848, + "loss": 3.7495, + "step": 6890 + }, + { + "epoch": 0.3626842105263158, + "grad_norm": 0.486328125, + "learning_rate": 0.00014162573099415205, + "loss": 3.4906, + "step": 6891 + }, + { + "epoch": 0.36273684210526314, + "grad_norm": 0.48046875, + "learning_rate": 0.0001416140350877193, + "loss": 3.7738, + "step": 6892 + }, + { + "epoch": 0.36278947368421055, + "grad_norm": 0.578125, + "learning_rate": 0.00014160233918128655, + "loss": 3.4785, + "step": 6893 + }, + { + "epoch": 0.3628421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.0001415906432748538, + "loss": 3.5137, + "step": 6894 + }, + { + "epoch": 0.36289473684210527, + "grad_norm": 0.498046875, + "learning_rate": 0.00014157894736842105, + "loss": 3.5536, + "step": 6895 + }, + { + "epoch": 0.36294736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.0001415672514619883, + "loss": 2.6624, + "step": 6896 + }, + { + "epoch": 0.363, + "grad_norm": 0.470703125, + "learning_rate": 0.00014155555555555555, + "loss": 3.6174, + "step": 6897 + }, + { + "epoch": 0.36305263157894735, + "grad_norm": 0.51953125, + "learning_rate": 0.00014154385964912282, + "loss": 3.7053, + "step": 6898 + }, + { + "epoch": 0.36310526315789476, + "grad_norm": 0.498046875, + "learning_rate": 0.00014153216374269007, + "loss": 3.7579, + "step": 6899 + }, + { + "epoch": 0.3631578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00014152046783625732, + "loss": 3.7342, + "step": 6900 + }, + { + "epoch": 0.3632105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.00014150877192982457, + "loss": 3.6033, + "step": 6901 + }, + { + "epoch": 0.36326315789473684, + "grad_norm": 0.51953125, + "learning_rate": 0.00014149707602339182, + "loss": 3.4917, + "step": 6902 + }, + { + "epoch": 0.3633157894736842, + "grad_norm": 0.474609375, + "learning_rate": 0.00014148538011695907, + "loss": 3.7327, + "step": 6903 + }, + { + "epoch": 0.36336842105263156, + "grad_norm": 0.53515625, + "learning_rate": 0.00014147368421052632, + "loss": 3.495, + "step": 6904 + }, + { + "epoch": 0.3634210526315789, + "grad_norm": 0.625, + "learning_rate": 0.00014146198830409357, + "loss": 3.4863, + "step": 6905 + }, + { + "epoch": 0.36347368421052634, + "grad_norm": 0.5859375, + "learning_rate": 0.00014145029239766082, + "loss": 3.6688, + "step": 6906 + }, + { + "epoch": 0.3635263157894737, + "grad_norm": 0.55078125, + "learning_rate": 0.00014143859649122807, + "loss": 2.9116, + "step": 6907 + }, + { + "epoch": 0.36357894736842106, + "grad_norm": 0.59765625, + "learning_rate": 0.00014142690058479532, + "loss": 3.2481, + "step": 6908 + }, + { + "epoch": 0.3636315789473684, + "grad_norm": 0.53125, + "learning_rate": 0.00014141520467836257, + "loss": 3.6102, + "step": 6909 + }, + { + "epoch": 0.3636842105263158, + "grad_norm": 0.53125, + "learning_rate": 0.00014140350877192982, + "loss": 3.4763, + "step": 6910 + }, + { + "epoch": 0.36373684210526314, + "grad_norm": 0.52734375, + "learning_rate": 0.00014139181286549707, + "loss": 3.3059, + "step": 6911 + }, + { + "epoch": 0.36378947368421055, + "grad_norm": 0.51953125, + "learning_rate": 0.00014138011695906432, + "loss": 3.1514, + "step": 6912 + }, + { + "epoch": 0.3638421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.0001413684210526316, + "loss": 3.5368, + "step": 6913 + }, + { + "epoch": 0.36389473684210527, + "grad_norm": 0.50390625, + "learning_rate": 0.00014135672514619885, + "loss": 3.6404, + "step": 6914 + }, + { + "epoch": 0.36394736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.0001413450292397661, + "loss": 3.6418, + "step": 6915 + }, + { + "epoch": 0.364, + "grad_norm": 0.52734375, + "learning_rate": 0.00014133333333333334, + "loss": 3.7372, + "step": 6916 + }, + { + "epoch": 0.36405263157894735, + "grad_norm": 0.5859375, + "learning_rate": 0.0001413216374269006, + "loss": 3.4387, + "step": 6917 + }, + { + "epoch": 0.3641052631578947, + "grad_norm": 0.55078125, + "learning_rate": 0.00014130994152046784, + "loss": 3.4428, + "step": 6918 + }, + { + "epoch": 0.3641578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.0001412982456140351, + "loss": 3.337, + "step": 6919 + }, + { + "epoch": 0.3642105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.00014128654970760234, + "loss": 3.8644, + "step": 6920 + }, + { + "epoch": 0.36426315789473684, + "grad_norm": 0.5390625, + "learning_rate": 0.0001412748538011696, + "loss": 3.2323, + "step": 6921 + }, + { + "epoch": 0.3643157894736842, + "grad_norm": 0.490234375, + "learning_rate": 0.00014126315789473684, + "loss": 4.0717, + "step": 6922 + }, + { + "epoch": 0.36436842105263156, + "grad_norm": 0.49609375, + "learning_rate": 0.0001412514619883041, + "loss": 3.4993, + "step": 6923 + }, + { + "epoch": 0.3644210526315789, + "grad_norm": 0.478515625, + "learning_rate": 0.00014123976608187134, + "loss": 3.698, + "step": 6924 + }, + { + "epoch": 0.36447368421052634, + "grad_norm": 0.478515625, + "learning_rate": 0.0001412280701754386, + "loss": 3.8925, + "step": 6925 + }, + { + "epoch": 0.3645263157894737, + "grad_norm": 0.494140625, + "learning_rate": 0.00014121637426900584, + "loss": 3.2191, + "step": 6926 + }, + { + "epoch": 0.36457894736842106, + "grad_norm": 0.51171875, + "learning_rate": 0.0001412046783625731, + "loss": 3.6357, + "step": 6927 + }, + { + "epoch": 0.3646315789473684, + "grad_norm": 0.48828125, + "learning_rate": 0.00014119298245614037, + "loss": 3.7248, + "step": 6928 + }, + { + "epoch": 0.3646842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.00014118128654970762, + "loss": 3.4597, + "step": 6929 + }, + { + "epoch": 0.36473684210526314, + "grad_norm": 0.53515625, + "learning_rate": 0.00014116959064327487, + "loss": 3.6174, + "step": 6930 + }, + { + "epoch": 0.36478947368421055, + "grad_norm": 0.59375, + "learning_rate": 0.00014115789473684212, + "loss": 3.2579, + "step": 6931 + }, + { + "epoch": 0.3648421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00014114619883040937, + "loss": 3.8182, + "step": 6932 + }, + { + "epoch": 0.36489473684210527, + "grad_norm": 0.52734375, + "learning_rate": 0.00014113450292397661, + "loss": 3.3168, + "step": 6933 + }, + { + "epoch": 0.36494736842105263, + "grad_norm": 0.58984375, + "learning_rate": 0.00014112280701754386, + "loss": 3.6249, + "step": 6934 + }, + { + "epoch": 0.365, + "grad_norm": 0.484375, + "learning_rate": 0.00014111111111111111, + "loss": 3.6689, + "step": 6935 + }, + { + "epoch": 0.36505263157894735, + "grad_norm": 0.50390625, + "learning_rate": 0.00014109941520467836, + "loss": 3.7793, + "step": 6936 + }, + { + "epoch": 0.3651052631578947, + "grad_norm": 0.470703125, + "learning_rate": 0.0001410877192982456, + "loss": 3.7018, + "step": 6937 + }, + { + "epoch": 0.3651578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.00014107602339181286, + "loss": 3.7571, + "step": 6938 + }, + { + "epoch": 0.3652105263157895, + "grad_norm": 0.59765625, + "learning_rate": 0.0001410643274853801, + "loss": 3.8367, + "step": 6939 + }, + { + "epoch": 0.36526315789473685, + "grad_norm": 0.52734375, + "learning_rate": 0.00014105263157894736, + "loss": 3.688, + "step": 6940 + }, + { + "epoch": 0.3653157894736842, + "grad_norm": 0.54296875, + "learning_rate": 0.0001410409356725146, + "loss": 3.8181, + "step": 6941 + }, + { + "epoch": 0.36536842105263156, + "grad_norm": 0.5546875, + "learning_rate": 0.0001410292397660819, + "loss": 3.7199, + "step": 6942 + }, + { + "epoch": 0.3654210526315789, + "grad_norm": 0.55859375, + "learning_rate": 0.00014101754385964914, + "loss": 3.6994, + "step": 6943 + }, + { + "epoch": 0.36547368421052634, + "grad_norm": 0.515625, + "learning_rate": 0.0001410058479532164, + "loss": 3.7684, + "step": 6944 + }, + { + "epoch": 0.3655263157894737, + "grad_norm": 0.515625, + "learning_rate": 0.00014099415204678364, + "loss": 3.8312, + "step": 6945 + }, + { + "epoch": 0.36557894736842106, + "grad_norm": 0.63671875, + "learning_rate": 0.00014098245614035089, + "loss": 3.6419, + "step": 6946 + }, + { + "epoch": 0.3656315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00014097076023391814, + "loss": 3.2981, + "step": 6947 + }, + { + "epoch": 0.3656842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00014095906432748539, + "loss": 3.7051, + "step": 6948 + }, + { + "epoch": 0.36573684210526314, + "grad_norm": 0.55859375, + "learning_rate": 0.00014094736842105264, + "loss": 3.7471, + "step": 6949 + }, + { + "epoch": 0.36578947368421055, + "grad_norm": 0.53515625, + "learning_rate": 0.0001409356725146199, + "loss": 3.9272, + "step": 6950 + }, + { + "epoch": 0.3658421052631579, + "grad_norm": 0.61328125, + "learning_rate": 0.00014092397660818713, + "loss": 4.029, + "step": 6951 + }, + { + "epoch": 0.3658947368421053, + "grad_norm": 0.53125, + "learning_rate": 0.00014091228070175438, + "loss": 3.7396, + "step": 6952 + }, + { + "epoch": 0.36594736842105263, + "grad_norm": 0.53515625, + "learning_rate": 0.00014090058479532163, + "loss": 3.5955, + "step": 6953 + }, + { + "epoch": 0.366, + "grad_norm": 0.51953125, + "learning_rate": 0.00014088888888888888, + "loss": 3.4537, + "step": 6954 + }, + { + "epoch": 0.36605263157894735, + "grad_norm": 0.53125, + "learning_rate": 0.00014087719298245613, + "loss": 3.7339, + "step": 6955 + }, + { + "epoch": 0.3661052631578947, + "grad_norm": 0.578125, + "learning_rate": 0.00014086549707602338, + "loss": 3.4349, + "step": 6956 + }, + { + "epoch": 0.3661578947368421, + "grad_norm": 0.58203125, + "learning_rate": 0.00014085380116959066, + "loss": 3.6452, + "step": 6957 + }, + { + "epoch": 0.3662105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.0001408421052631579, + "loss": 3.428, + "step": 6958 + }, + { + "epoch": 0.36626315789473685, + "grad_norm": 0.51171875, + "learning_rate": 0.00014083040935672516, + "loss": 3.3881, + "step": 6959 + }, + { + "epoch": 0.3663157894736842, + "grad_norm": 0.5234375, + "learning_rate": 0.0001408187134502924, + "loss": 3.8929, + "step": 6960 + }, + { + "epoch": 0.36636842105263157, + "grad_norm": 0.52734375, + "learning_rate": 0.00014080701754385966, + "loss": 3.4752, + "step": 6961 + }, + { + "epoch": 0.3664210526315789, + "grad_norm": 0.50390625, + "learning_rate": 0.0001407953216374269, + "loss": 3.8327, + "step": 6962 + }, + { + "epoch": 0.36647368421052634, + "grad_norm": 0.6953125, + "learning_rate": 0.00014078362573099416, + "loss": 3.9624, + "step": 6963 + }, + { + "epoch": 0.3665263157894737, + "grad_norm": 0.494140625, + "learning_rate": 0.0001407719298245614, + "loss": 3.4424, + "step": 6964 + }, + { + "epoch": 0.36657894736842106, + "grad_norm": 0.498046875, + "learning_rate": 0.00014076023391812868, + "loss": 3.7581, + "step": 6965 + }, + { + "epoch": 0.3666315789473684, + "grad_norm": 0.55078125, + "learning_rate": 0.0001407485380116959, + "loss": 3.7955, + "step": 6966 + }, + { + "epoch": 0.3666842105263158, + "grad_norm": 0.5390625, + "learning_rate": 0.00014073684210526315, + "loss": 4.0319, + "step": 6967 + }, + { + "epoch": 0.36673684210526314, + "grad_norm": 0.58984375, + "learning_rate": 0.0001407251461988304, + "loss": 3.8219, + "step": 6968 + }, + { + "epoch": 0.3667894736842105, + "grad_norm": 0.5703125, + "learning_rate": 0.00014071345029239765, + "loss": 3.6733, + "step": 6969 + }, + { + "epoch": 0.3668421052631579, + "grad_norm": 0.58984375, + "learning_rate": 0.0001407017543859649, + "loss": 3.524, + "step": 6970 + }, + { + "epoch": 0.3668947368421053, + "grad_norm": 0.53125, + "learning_rate": 0.00014069005847953215, + "loss": 3.873, + "step": 6971 + }, + { + "epoch": 0.36694736842105263, + "grad_norm": 0.466796875, + "learning_rate": 0.00014067836257309943, + "loss": 3.2903, + "step": 6972 + }, + { + "epoch": 0.367, + "grad_norm": 0.486328125, + "learning_rate": 0.00014066666666666668, + "loss": 3.4191, + "step": 6973 + }, + { + "epoch": 0.36705263157894735, + "grad_norm": 0.5390625, + "learning_rate": 0.00014065497076023393, + "loss": 3.7119, + "step": 6974 + }, + { + "epoch": 0.3671052631578947, + "grad_norm": 0.54296875, + "learning_rate": 0.00014064327485380118, + "loss": 3.7319, + "step": 6975 + }, + { + "epoch": 0.3671578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00014063157894736843, + "loss": 3.3616, + "step": 6976 + }, + { + "epoch": 0.3672105263157895, + "grad_norm": 0.490234375, + "learning_rate": 0.00014061988304093568, + "loss": 3.6883, + "step": 6977 + }, + { + "epoch": 0.36726315789473685, + "grad_norm": 0.5, + "learning_rate": 0.00014060818713450293, + "loss": 3.6722, + "step": 6978 + }, + { + "epoch": 0.3673157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.0001405964912280702, + "loss": 3.7523, + "step": 6979 + }, + { + "epoch": 0.36736842105263157, + "grad_norm": 0.48828125, + "learning_rate": 0.00014058479532163745, + "loss": 3.6324, + "step": 6980 + }, + { + "epoch": 0.3674210526315789, + "grad_norm": 0.50390625, + "learning_rate": 0.00014057309941520468, + "loss": 3.5361, + "step": 6981 + }, + { + "epoch": 0.36747368421052634, + "grad_norm": 0.52734375, + "learning_rate": 0.00014056140350877193, + "loss": 3.6471, + "step": 6982 + }, + { + "epoch": 0.3675263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.00014054970760233918, + "loss": 3.314, + "step": 6983 + }, + { + "epoch": 0.36757894736842106, + "grad_norm": 0.453125, + "learning_rate": 0.00014053801169590643, + "loss": 3.5255, + "step": 6984 + }, + { + "epoch": 0.3676315789473684, + "grad_norm": 0.47265625, + "learning_rate": 0.00014052631578947367, + "loss": 3.5671, + "step": 6985 + }, + { + "epoch": 0.3676842105263158, + "grad_norm": 0.482421875, + "learning_rate": 0.00014051461988304092, + "loss": 3.3801, + "step": 6986 + }, + { + "epoch": 0.36773684210526314, + "grad_norm": 0.48828125, + "learning_rate": 0.0001405029239766082, + "loss": 3.7821, + "step": 6987 + }, + { + "epoch": 0.3677894736842105, + "grad_norm": 0.48828125, + "learning_rate": 0.00014049122807017545, + "loss": 3.9414, + "step": 6988 + }, + { + "epoch": 0.3678421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.0001404795321637427, + "loss": 3.9844, + "step": 6989 + }, + { + "epoch": 0.3678947368421053, + "grad_norm": 0.498046875, + "learning_rate": 0.00014046783625730995, + "loss": 3.2574, + "step": 6990 + }, + { + "epoch": 0.36794736842105263, + "grad_norm": 0.5625, + "learning_rate": 0.0001404561403508772, + "loss": 3.6384, + "step": 6991 + }, + { + "epoch": 0.368, + "grad_norm": 0.6953125, + "learning_rate": 0.00014044444444444445, + "loss": 3.5955, + "step": 6992 + }, + { + "epoch": 0.36805263157894735, + "grad_norm": 0.486328125, + "learning_rate": 0.0001404327485380117, + "loss": 3.2997, + "step": 6993 + }, + { + "epoch": 0.3681052631578947, + "grad_norm": 0.5234375, + "learning_rate": 0.00014042105263157898, + "loss": 3.4269, + "step": 6994 + }, + { + "epoch": 0.36815789473684213, + "grad_norm": 0.482421875, + "learning_rate": 0.00014040935672514622, + "loss": 3.8133, + "step": 6995 + }, + { + "epoch": 0.3682105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00014039766081871347, + "loss": 3.8102, + "step": 6996 + }, + { + "epoch": 0.36826315789473685, + "grad_norm": 0.51953125, + "learning_rate": 0.0001403859649122807, + "loss": 4.0291, + "step": 6997 + }, + { + "epoch": 0.3683157894736842, + "grad_norm": 0.5234375, + "learning_rate": 0.00014037426900584795, + "loss": 3.4018, + "step": 6998 + }, + { + "epoch": 0.36836842105263157, + "grad_norm": 0.5078125, + "learning_rate": 0.0001403625730994152, + "loss": 3.6248, + "step": 6999 + }, + { + "epoch": 0.3684210526315789, + "grad_norm": 0.51953125, + "learning_rate": 0.00014035087719298245, + "loss": 3.194, + "step": 7000 + }, + { + "epoch": 0.36847368421052634, + "grad_norm": 0.5390625, + "learning_rate": 0.0001403391812865497, + "loss": 3.2362, + "step": 7001 + }, + { + "epoch": 0.3685263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.00014032748538011697, + "loss": 3.7797, + "step": 7002 + }, + { + "epoch": 0.36857894736842106, + "grad_norm": 0.46484375, + "learning_rate": 0.00014031578947368422, + "loss": 3.2856, + "step": 7003 + }, + { + "epoch": 0.3686315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.00014030409356725147, + "loss": 3.7312, + "step": 7004 + }, + { + "epoch": 0.3686842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00014029239766081872, + "loss": 3.963, + "step": 7005 + }, + { + "epoch": 0.36873684210526314, + "grad_norm": 0.515625, + "learning_rate": 0.00014028070175438597, + "loss": 3.4671, + "step": 7006 + }, + { + "epoch": 0.3687894736842105, + "grad_norm": 0.59375, + "learning_rate": 0.00014026900584795322, + "loss": 3.5815, + "step": 7007 + }, + { + "epoch": 0.3688421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.00014025730994152047, + "loss": 3.761, + "step": 7008 + }, + { + "epoch": 0.3688947368421053, + "grad_norm": 0.671875, + "learning_rate": 0.00014024561403508775, + "loss": 3.6425, + "step": 7009 + }, + { + "epoch": 0.36894736842105263, + "grad_norm": 0.4921875, + "learning_rate": 0.000140233918128655, + "loss": 3.9392, + "step": 7010 + }, + { + "epoch": 0.369, + "grad_norm": 0.53515625, + "learning_rate": 0.00014022222222222225, + "loss": 3.7556, + "step": 7011 + }, + { + "epoch": 0.36905263157894735, + "grad_norm": 0.54296875, + "learning_rate": 0.00014021052631578947, + "loss": 4.0514, + "step": 7012 + }, + { + "epoch": 0.3691052631578947, + "grad_norm": 0.546875, + "learning_rate": 0.00014019883040935672, + "loss": 3.7708, + "step": 7013 + }, + { + "epoch": 0.36915789473684213, + "grad_norm": 0.59375, + "learning_rate": 0.00014018713450292397, + "loss": 3.6248, + "step": 7014 + }, + { + "epoch": 0.3692105263157895, + "grad_norm": 0.515625, + "learning_rate": 0.00014017543859649122, + "loss": 3.1413, + "step": 7015 + }, + { + "epoch": 0.36926315789473685, + "grad_norm": 0.478515625, + "learning_rate": 0.0001401637426900585, + "loss": 3.7721, + "step": 7016 + }, + { + "epoch": 0.3693157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.00014015204678362574, + "loss": 3.3439, + "step": 7017 + }, + { + "epoch": 0.36936842105263157, + "grad_norm": 1.6796875, + "learning_rate": 0.000140140350877193, + "loss": 4.0952, + "step": 7018 + }, + { + "epoch": 0.36942105263157893, + "grad_norm": 0.5625, + "learning_rate": 0.00014012865497076024, + "loss": 3.5979, + "step": 7019 + }, + { + "epoch": 0.36947368421052634, + "grad_norm": 0.50390625, + "learning_rate": 0.0001401169590643275, + "loss": 3.1897, + "step": 7020 + }, + { + "epoch": 0.3695263157894737, + "grad_norm": 0.490234375, + "learning_rate": 0.00014010526315789474, + "loss": 3.5971, + "step": 7021 + }, + { + "epoch": 0.36957894736842106, + "grad_norm": 0.52734375, + "learning_rate": 0.000140093567251462, + "loss": 3.2597, + "step": 7022 + }, + { + "epoch": 0.3696315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.00014008187134502924, + "loss": 3.4189, + "step": 7023 + }, + { + "epoch": 0.3696842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00014007017543859652, + "loss": 3.3535, + "step": 7024 + }, + { + "epoch": 0.36973684210526314, + "grad_norm": 0.53515625, + "learning_rate": 0.00014005847953216377, + "loss": 3.5296, + "step": 7025 + }, + { + "epoch": 0.3697894736842105, + "grad_norm": 0.5859375, + "learning_rate": 0.00014004678362573102, + "loss": 3.1568, + "step": 7026 + }, + { + "epoch": 0.3698421052631579, + "grad_norm": 0.57421875, + "learning_rate": 0.00014003508771929824, + "loss": 3.0357, + "step": 7027 + }, + { + "epoch": 0.3698947368421053, + "grad_norm": 0.5, + "learning_rate": 0.0001400233918128655, + "loss": 3.402, + "step": 7028 + }, + { + "epoch": 0.36994736842105264, + "grad_norm": 0.4765625, + "learning_rate": 0.00014001169590643274, + "loss": 3.4917, + "step": 7029 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.00014, + "loss": 3.1279, + "step": 7030 + }, + { + "epoch": 0.37005263157894736, + "grad_norm": 0.490234375, + "learning_rate": 0.00013998830409356726, + "loss": 3.905, + "step": 7031 + }, + { + "epoch": 0.3701052631578947, + "grad_norm": 0.49609375, + "learning_rate": 0.00013997660818713451, + "loss": 3.5651, + "step": 7032 + }, + { + "epoch": 0.37015789473684213, + "grad_norm": 0.53515625, + "learning_rate": 0.00013996491228070176, + "loss": 3.663, + "step": 7033 + }, + { + "epoch": 0.3702105263157895, + "grad_norm": 0.474609375, + "learning_rate": 0.000139953216374269, + "loss": 3.3607, + "step": 7034 + }, + { + "epoch": 0.37026315789473685, + "grad_norm": 0.48828125, + "learning_rate": 0.00013994152046783626, + "loss": 3.2604, + "step": 7035 + }, + { + "epoch": 0.3703157894736842, + "grad_norm": 0.498046875, + "learning_rate": 0.0001399298245614035, + "loss": 3.7522, + "step": 7036 + }, + { + "epoch": 0.37036842105263157, + "grad_norm": 0.48828125, + "learning_rate": 0.00013991812865497076, + "loss": 3.7703, + "step": 7037 + }, + { + "epoch": 0.37042105263157893, + "grad_norm": 0.49609375, + "learning_rate": 0.000139906432748538, + "loss": 3.2735, + "step": 7038 + }, + { + "epoch": 0.3704736842105263, + "grad_norm": 0.4921875, + "learning_rate": 0.0001398947368421053, + "loss": 3.2588, + "step": 7039 + }, + { + "epoch": 0.3705263157894737, + "grad_norm": 0.451171875, + "learning_rate": 0.00013988304093567254, + "loss": 3.6588, + "step": 7040 + }, + { + "epoch": 0.37057894736842106, + "grad_norm": 0.49609375, + "learning_rate": 0.0001398713450292398, + "loss": 3.8187, + "step": 7041 + }, + { + "epoch": 0.3706315789473684, + "grad_norm": 0.59375, + "learning_rate": 0.00013985964912280704, + "loss": 3.8473, + "step": 7042 + }, + { + "epoch": 0.3706842105263158, + "grad_norm": 0.486328125, + "learning_rate": 0.00013984795321637426, + "loss": 3.2322, + "step": 7043 + }, + { + "epoch": 0.37073684210526314, + "grad_norm": 0.478515625, + "learning_rate": 0.0001398362573099415, + "loss": 3.7862, + "step": 7044 + }, + { + "epoch": 0.3707894736842105, + "grad_norm": 0.494140625, + "learning_rate": 0.00013982456140350876, + "loss": 3.3857, + "step": 7045 + }, + { + "epoch": 0.3708421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00013981286549707604, + "loss": 3.6251, + "step": 7046 + }, + { + "epoch": 0.3708947368421053, + "grad_norm": 0.6484375, + "learning_rate": 0.00013980116959064329, + "loss": 3.2724, + "step": 7047 + }, + { + "epoch": 0.37094736842105264, + "grad_norm": 0.51953125, + "learning_rate": 0.00013978947368421053, + "loss": 3.9603, + "step": 7048 + }, + { + "epoch": 0.371, + "grad_norm": 0.51171875, + "learning_rate": 0.00013977777777777778, + "loss": 3.6779, + "step": 7049 + }, + { + "epoch": 0.37105263157894736, + "grad_norm": 0.51953125, + "learning_rate": 0.00013976608187134503, + "loss": 3.831, + "step": 7050 + }, + { + "epoch": 0.3711052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.00013975438596491228, + "loss": 3.7756, + "step": 7051 + }, + { + "epoch": 0.37115789473684213, + "grad_norm": 0.66015625, + "learning_rate": 0.00013974269005847953, + "loss": 3.6738, + "step": 7052 + }, + { + "epoch": 0.3712105263157895, + "grad_norm": 0.50390625, + "learning_rate": 0.0001397309941520468, + "loss": 3.9744, + "step": 7053 + }, + { + "epoch": 0.37126315789473685, + "grad_norm": 0.6640625, + "learning_rate": 0.00013971929824561406, + "loss": 3.3942, + "step": 7054 + }, + { + "epoch": 0.3713157894736842, + "grad_norm": 0.46484375, + "learning_rate": 0.0001397076023391813, + "loss": 3.4415, + "step": 7055 + }, + { + "epoch": 0.37136842105263157, + "grad_norm": 0.478515625, + "learning_rate": 0.00013969590643274856, + "loss": 3.4134, + "step": 7056 + }, + { + "epoch": 0.37142105263157893, + "grad_norm": 0.57421875, + "learning_rate": 0.0001396842105263158, + "loss": 3.5425, + "step": 7057 + }, + { + "epoch": 0.3714736842105263, + "grad_norm": 0.470703125, + "learning_rate": 0.00013967251461988303, + "loss": 3.8362, + "step": 7058 + }, + { + "epoch": 0.3715263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.00013966081871345028, + "loss": 3.7569, + "step": 7059 + }, + { + "epoch": 0.37157894736842106, + "grad_norm": 0.55859375, + "learning_rate": 0.00013964912280701753, + "loss": 3.3905, + "step": 7060 + }, + { + "epoch": 0.3716315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.0001396374269005848, + "loss": 3.3599, + "step": 7061 + }, + { + "epoch": 0.3716842105263158, + "grad_norm": 0.486328125, + "learning_rate": 0.00013962573099415206, + "loss": 3.5089, + "step": 7062 + }, + { + "epoch": 0.37173684210526314, + "grad_norm": 0.48828125, + "learning_rate": 0.0001396140350877193, + "loss": 3.7203, + "step": 7063 + }, + { + "epoch": 0.3717894736842105, + "grad_norm": 0.4921875, + "learning_rate": 0.00013960233918128656, + "loss": 3.1298, + "step": 7064 + }, + { + "epoch": 0.3718421052631579, + "grad_norm": 0.5, + "learning_rate": 0.0001395906432748538, + "loss": 3.6472, + "step": 7065 + }, + { + "epoch": 0.3718947368421053, + "grad_norm": 0.48828125, + "learning_rate": 0.00013957894736842105, + "loss": 3.5852, + "step": 7066 + }, + { + "epoch": 0.37194736842105264, + "grad_norm": 0.5625, + "learning_rate": 0.0001395672514619883, + "loss": 3.1741, + "step": 7067 + }, + { + "epoch": 0.372, + "grad_norm": 0.47265625, + "learning_rate": 0.00013955555555555558, + "loss": 4.1095, + "step": 7068 + }, + { + "epoch": 0.37205263157894736, + "grad_norm": 0.5078125, + "learning_rate": 0.00013954385964912283, + "loss": 3.5694, + "step": 7069 + }, + { + "epoch": 0.3721052631578947, + "grad_norm": 0.50390625, + "learning_rate": 0.00013953216374269008, + "loss": 3.7618, + "step": 7070 + }, + { + "epoch": 0.37215789473684213, + "grad_norm": 0.51171875, + "learning_rate": 0.00013952046783625733, + "loss": 4.0918, + "step": 7071 + }, + { + "epoch": 0.3722105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00013950877192982458, + "loss": 3.7071, + "step": 7072 + }, + { + "epoch": 0.37226315789473685, + "grad_norm": 0.7109375, + "learning_rate": 0.0001394970760233918, + "loss": 3.7841, + "step": 7073 + }, + { + "epoch": 0.3723157894736842, + "grad_norm": 0.474609375, + "learning_rate": 0.00013948538011695905, + "loss": 3.4196, + "step": 7074 + }, + { + "epoch": 0.37236842105263157, + "grad_norm": 0.53125, + "learning_rate": 0.0001394736842105263, + "loss": 3.6695, + "step": 7075 + }, + { + "epoch": 0.37242105263157893, + "grad_norm": 0.498046875, + "learning_rate": 0.00013946198830409358, + "loss": 3.5993, + "step": 7076 + }, + { + "epoch": 0.3724736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00013945029239766083, + "loss": 3.8534, + "step": 7077 + }, + { + "epoch": 0.3725263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.00013943859649122808, + "loss": 3.4496, + "step": 7078 + }, + { + "epoch": 0.37257894736842107, + "grad_norm": 0.498046875, + "learning_rate": 0.00013942690058479533, + "loss": 3.8523, + "step": 7079 + }, + { + "epoch": 0.3726315789473684, + "grad_norm": 0.55078125, + "learning_rate": 0.00013941520467836258, + "loss": 3.7309, + "step": 7080 + }, + { + "epoch": 0.3726842105263158, + "grad_norm": 0.5703125, + "learning_rate": 0.00013940350877192983, + "loss": 3.3976, + "step": 7081 + }, + { + "epoch": 0.37273684210526314, + "grad_norm": 0.515625, + "learning_rate": 0.00013939181286549707, + "loss": 3.3411, + "step": 7082 + }, + { + "epoch": 0.3727894736842105, + "grad_norm": 0.5703125, + "learning_rate": 0.00013938011695906435, + "loss": 3.6191, + "step": 7083 + }, + { + "epoch": 0.3728421052631579, + "grad_norm": 0.5, + "learning_rate": 0.0001393684210526316, + "loss": 3.4962, + "step": 7084 + }, + { + "epoch": 0.3728947368421053, + "grad_norm": 0.66015625, + "learning_rate": 0.00013935672514619885, + "loss": 3.6697, + "step": 7085 + }, + { + "epoch": 0.37294736842105264, + "grad_norm": 0.451171875, + "learning_rate": 0.0001393450292397661, + "loss": 3.8097, + "step": 7086 + }, + { + "epoch": 0.373, + "grad_norm": 0.48828125, + "learning_rate": 0.00013933333333333335, + "loss": 3.7626, + "step": 7087 + }, + { + "epoch": 0.37305263157894736, + "grad_norm": 0.50390625, + "learning_rate": 0.0001393216374269006, + "loss": 3.7963, + "step": 7088 + }, + { + "epoch": 0.3731052631578947, + "grad_norm": 0.5625, + "learning_rate": 0.00013930994152046782, + "loss": 3.4703, + "step": 7089 + }, + { + "epoch": 0.37315789473684213, + "grad_norm": 0.45703125, + "learning_rate": 0.00013929824561403507, + "loss": 3.5604, + "step": 7090 + }, + { + "epoch": 0.3732105263157895, + "grad_norm": 0.455078125, + "learning_rate": 0.00013928654970760235, + "loss": 3.317, + "step": 7091 + }, + { + "epoch": 0.37326315789473685, + "grad_norm": 0.59375, + "learning_rate": 0.0001392748538011696, + "loss": 3.589, + "step": 7092 + }, + { + "epoch": 0.3733157894736842, + "grad_norm": 0.478515625, + "learning_rate": 0.00013926315789473685, + "loss": 3.9087, + "step": 7093 + }, + { + "epoch": 0.37336842105263157, + "grad_norm": 0.494140625, + "learning_rate": 0.0001392514619883041, + "loss": 3.8487, + "step": 7094 + }, + { + "epoch": 0.37342105263157893, + "grad_norm": 0.50390625, + "learning_rate": 0.00013923976608187135, + "loss": 3.4065, + "step": 7095 + }, + { + "epoch": 0.3734736842105263, + "grad_norm": 0.490234375, + "learning_rate": 0.0001392280701754386, + "loss": 3.2633, + "step": 7096 + }, + { + "epoch": 0.3735263157894737, + "grad_norm": 0.5234375, + "learning_rate": 0.00013921637426900585, + "loss": 3.5487, + "step": 7097 + }, + { + "epoch": 0.37357894736842107, + "grad_norm": 0.49609375, + "learning_rate": 0.00013920467836257312, + "loss": 3.6456, + "step": 7098 + }, + { + "epoch": 0.3736315789473684, + "grad_norm": 0.54296875, + "learning_rate": 0.00013919298245614037, + "loss": 3.6717, + "step": 7099 + }, + { + "epoch": 0.3736842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.00013918128654970762, + "loss": 3.3397, + "step": 7100 + }, + { + "epoch": 0.37373684210526315, + "grad_norm": 0.486328125, + "learning_rate": 0.00013916959064327487, + "loss": 3.4102, + "step": 7101 + }, + { + "epoch": 0.3737894736842105, + "grad_norm": 0.46484375, + "learning_rate": 0.00013915789473684212, + "loss": 3.9915, + "step": 7102 + }, + { + "epoch": 0.3738421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00013914619883040937, + "loss": 3.7734, + "step": 7103 + }, + { + "epoch": 0.3738947368421053, + "grad_norm": 0.498046875, + "learning_rate": 0.0001391345029239766, + "loss": 4.1262, + "step": 7104 + }, + { + "epoch": 0.37394736842105264, + "grad_norm": 0.58984375, + "learning_rate": 0.00013912280701754387, + "loss": 3.4468, + "step": 7105 + }, + { + "epoch": 0.374, + "grad_norm": 0.50390625, + "learning_rate": 0.00013911111111111112, + "loss": 3.3162, + "step": 7106 + }, + { + "epoch": 0.37405263157894736, + "grad_norm": 0.5625, + "learning_rate": 0.00013909941520467837, + "loss": 3.3279, + "step": 7107 + }, + { + "epoch": 0.3741052631578947, + "grad_norm": 0.546875, + "learning_rate": 0.00013908771929824562, + "loss": 4.0245, + "step": 7108 + }, + { + "epoch": 0.3741578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00013907602339181287, + "loss": 3.5874, + "step": 7109 + }, + { + "epoch": 0.3742105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.00013906432748538012, + "loss": 3.714, + "step": 7110 + }, + { + "epoch": 0.37426315789473685, + "grad_norm": 0.48828125, + "learning_rate": 0.00013905263157894737, + "loss": 3.8858, + "step": 7111 + }, + { + "epoch": 0.3743157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00013904093567251462, + "loss": 3.7578, + "step": 7112 + }, + { + "epoch": 0.3743684210526316, + "grad_norm": 0.47265625, + "learning_rate": 0.0001390292397660819, + "loss": 3.5498, + "step": 7113 + }, + { + "epoch": 0.37442105263157893, + "grad_norm": 0.48828125, + "learning_rate": 0.00013901754385964914, + "loss": 3.3846, + "step": 7114 + }, + { + "epoch": 0.3744736842105263, + "grad_norm": 0.5234375, + "learning_rate": 0.0001390058479532164, + "loss": 3.6329, + "step": 7115 + }, + { + "epoch": 0.3745263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00013899415204678364, + "loss": 3.4583, + "step": 7116 + }, + { + "epoch": 0.37457894736842107, + "grad_norm": 0.52734375, + "learning_rate": 0.0001389824561403509, + "loss": 3.4861, + "step": 7117 + }, + { + "epoch": 0.3746315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00013897076023391814, + "loss": 3.5783, + "step": 7118 + }, + { + "epoch": 0.3746842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00013895906432748536, + "loss": 3.3254, + "step": 7119 + }, + { + "epoch": 0.37473684210526315, + "grad_norm": 0.51171875, + "learning_rate": 0.00013894736842105264, + "loss": 3.3937, + "step": 7120 + }, + { + "epoch": 0.3747894736842105, + "grad_norm": 0.486328125, + "learning_rate": 0.0001389356725146199, + "loss": 3.891, + "step": 7121 + }, + { + "epoch": 0.3748421052631579, + "grad_norm": 0.486328125, + "learning_rate": 0.00013892397660818714, + "loss": 3.7826, + "step": 7122 + }, + { + "epoch": 0.3748947368421053, + "grad_norm": 0.5078125, + "learning_rate": 0.0001389122807017544, + "loss": 3.6325, + "step": 7123 + }, + { + "epoch": 0.37494736842105264, + "grad_norm": 0.470703125, + "learning_rate": 0.00013890058479532164, + "loss": 3.1927, + "step": 7124 + }, + { + "epoch": 0.375, + "grad_norm": 0.55078125, + "learning_rate": 0.0001388888888888889, + "loss": 3.4731, + "step": 7125 + }, + { + "epoch": 0.37505263157894736, + "grad_norm": 0.53125, + "learning_rate": 0.00013887719298245614, + "loss": 3.6939, + "step": 7126 + }, + { + "epoch": 0.3751052631578947, + "grad_norm": 0.5625, + "learning_rate": 0.0001388654970760234, + "loss": 3.705, + "step": 7127 + }, + { + "epoch": 0.3751578947368421, + "grad_norm": 0.56640625, + "learning_rate": 0.00013885380116959066, + "loss": 3.4811, + "step": 7128 + }, + { + "epoch": 0.3752105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.00013884210526315791, + "loss": 3.2377, + "step": 7129 + }, + { + "epoch": 0.37526315789473685, + "grad_norm": 0.5078125, + "learning_rate": 0.00013883040935672516, + "loss": 3.6368, + "step": 7130 + }, + { + "epoch": 0.3753157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.0001388187134502924, + "loss": 3.8193, + "step": 7131 + }, + { + "epoch": 0.3753684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.00013880701754385966, + "loss": 3.7757, + "step": 7132 + }, + { + "epoch": 0.37542105263157893, + "grad_norm": 0.51953125, + "learning_rate": 0.0001387953216374269, + "loss": 3.3467, + "step": 7133 + }, + { + "epoch": 0.3754736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00013878362573099416, + "loss": 3.3087, + "step": 7134 + }, + { + "epoch": 0.3755263157894737, + "grad_norm": 0.5, + "learning_rate": 0.0001387719298245614, + "loss": 3.3003, + "step": 7135 + }, + { + "epoch": 0.37557894736842107, + "grad_norm": 0.498046875, + "learning_rate": 0.00013876023391812866, + "loss": 3.5959, + "step": 7136 + }, + { + "epoch": 0.3756315789473684, + "grad_norm": 0.470703125, + "learning_rate": 0.0001387485380116959, + "loss": 3.4628, + "step": 7137 + }, + { + "epoch": 0.3756842105263158, + "grad_norm": 0.51953125, + "learning_rate": 0.00013873684210526316, + "loss": 3.3494, + "step": 7138 + }, + { + "epoch": 0.37573684210526315, + "grad_norm": 0.5078125, + "learning_rate": 0.0001387251461988304, + "loss": 4.1713, + "step": 7139 + }, + { + "epoch": 0.3757894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.00013871345029239766, + "loss": 3.6765, + "step": 7140 + }, + { + "epoch": 0.3758421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.0001387017543859649, + "loss": 3.6918, + "step": 7141 + }, + { + "epoch": 0.3758947368421053, + "grad_norm": 0.57421875, + "learning_rate": 0.00013869005847953219, + "loss": 3.4267, + "step": 7142 + }, + { + "epoch": 0.37594736842105264, + "grad_norm": 0.4765625, + "learning_rate": 0.00013867836257309944, + "loss": 3.247, + "step": 7143 + }, + { + "epoch": 0.376, + "grad_norm": 0.55859375, + "learning_rate": 0.00013866666666666669, + "loss": 3.9422, + "step": 7144 + }, + { + "epoch": 0.37605263157894736, + "grad_norm": 0.53515625, + "learning_rate": 0.00013865497076023393, + "loss": 3.4689, + "step": 7145 + }, + { + "epoch": 0.3761052631578947, + "grad_norm": 0.5234375, + "learning_rate": 0.00013864327485380118, + "loss": 3.6883, + "step": 7146 + }, + { + "epoch": 0.3761578947368421, + "grad_norm": 0.5703125, + "learning_rate": 0.00013863157894736843, + "loss": 3.5678, + "step": 7147 + }, + { + "epoch": 0.3762105263157895, + "grad_norm": 0.4765625, + "learning_rate": 0.00013861988304093568, + "loss": 3.8359, + "step": 7148 + }, + { + "epoch": 0.37626315789473685, + "grad_norm": 0.494140625, + "learning_rate": 0.00013860818713450293, + "loss": 3.9099, + "step": 7149 + }, + { + "epoch": 0.3763157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00013859649122807018, + "loss": 3.5323, + "step": 7150 + }, + { + "epoch": 0.3763684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00013858479532163743, + "loss": 3.6936, + "step": 7151 + }, + { + "epoch": 0.37642105263157893, + "grad_norm": 0.5234375, + "learning_rate": 0.00013857309941520468, + "loss": 3.9693, + "step": 7152 + }, + { + "epoch": 0.3764736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.00013856140350877193, + "loss": 3.965, + "step": 7153 + }, + { + "epoch": 0.3765263157894737, + "grad_norm": 0.8203125, + "learning_rate": 0.00013854970760233918, + "loss": 3.2248, + "step": 7154 + }, + { + "epoch": 0.37657894736842107, + "grad_norm": 0.484375, + "learning_rate": 0.00013853801169590643, + "loss": 3.7407, + "step": 7155 + }, + { + "epoch": 0.37663157894736843, + "grad_norm": 0.55078125, + "learning_rate": 0.00013852631578947368, + "loss": 3.4231, + "step": 7156 + }, + { + "epoch": 0.3766842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00013851461988304096, + "loss": 3.6176, + "step": 7157 + }, + { + "epoch": 0.37673684210526315, + "grad_norm": 0.67578125, + "learning_rate": 0.0001385029239766082, + "loss": 3.2403, + "step": 7158 + }, + { + "epoch": 0.3767894736842105, + "grad_norm": 0.48046875, + "learning_rate": 0.00013849122807017546, + "loss": 3.5226, + "step": 7159 + }, + { + "epoch": 0.37684210526315787, + "grad_norm": 0.546875, + "learning_rate": 0.0001384795321637427, + "loss": 3.5935, + "step": 7160 + }, + { + "epoch": 0.3768947368421053, + "grad_norm": 0.6015625, + "learning_rate": 0.00013846783625730996, + "loss": 3.6858, + "step": 7161 + }, + { + "epoch": 0.37694736842105264, + "grad_norm": 0.47265625, + "learning_rate": 0.0001384561403508772, + "loss": 3.4243, + "step": 7162 + }, + { + "epoch": 0.377, + "grad_norm": 0.53125, + "learning_rate": 0.00013844444444444445, + "loss": 3.3489, + "step": 7163 + }, + { + "epoch": 0.37705263157894736, + "grad_norm": 0.57421875, + "learning_rate": 0.0001384327485380117, + "loss": 3.6448, + "step": 7164 + }, + { + "epoch": 0.3771052631578947, + "grad_norm": 0.451171875, + "learning_rate": 0.00013842105263157895, + "loss": 3.5228, + "step": 7165 + }, + { + "epoch": 0.3771578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.0001384093567251462, + "loss": 3.6833, + "step": 7166 + }, + { + "epoch": 0.3772105263157895, + "grad_norm": 0.4765625, + "learning_rate": 0.00013839766081871345, + "loss": 3.5413, + "step": 7167 + }, + { + "epoch": 0.37726315789473686, + "grad_norm": 0.4921875, + "learning_rate": 0.0001383859649122807, + "loss": 3.9528, + "step": 7168 + }, + { + "epoch": 0.3773157894736842, + "grad_norm": 0.53515625, + "learning_rate": 0.00013837426900584795, + "loss": 3.8773, + "step": 7169 + }, + { + "epoch": 0.3773684210526316, + "grad_norm": 0.52734375, + "learning_rate": 0.0001383625730994152, + "loss": 3.5209, + "step": 7170 + }, + { + "epoch": 0.37742105263157893, + "grad_norm": 0.51171875, + "learning_rate": 0.00013835087719298245, + "loss": 3.4159, + "step": 7171 + }, + { + "epoch": 0.3774736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00013833918128654973, + "loss": 3.655, + "step": 7172 + }, + { + "epoch": 0.3775263157894737, + "grad_norm": 0.48828125, + "learning_rate": 0.00013832748538011698, + "loss": 3.3398, + "step": 7173 + }, + { + "epoch": 0.37757894736842107, + "grad_norm": 0.6328125, + "learning_rate": 0.00013831578947368423, + "loss": 3.3894, + "step": 7174 + }, + { + "epoch": 0.37763157894736843, + "grad_norm": 0.47265625, + "learning_rate": 0.00013830409356725148, + "loss": 3.6835, + "step": 7175 + }, + { + "epoch": 0.3776842105263158, + "grad_norm": 0.490234375, + "learning_rate": 0.00013829239766081873, + "loss": 3.609, + "step": 7176 + }, + { + "epoch": 0.37773684210526315, + "grad_norm": 0.53125, + "learning_rate": 0.00013828070175438598, + "loss": 3.7685, + "step": 7177 + }, + { + "epoch": 0.3777894736842105, + "grad_norm": 0.49609375, + "learning_rate": 0.00013826900584795323, + "loss": 3.6059, + "step": 7178 + }, + { + "epoch": 0.37784210526315787, + "grad_norm": 0.490234375, + "learning_rate": 0.00013825730994152048, + "loss": 3.3412, + "step": 7179 + }, + { + "epoch": 0.3778947368421053, + "grad_norm": 0.4765625, + "learning_rate": 0.00013824561403508772, + "loss": 3.537, + "step": 7180 + }, + { + "epoch": 0.37794736842105264, + "grad_norm": 0.439453125, + "learning_rate": 0.00013823391812865497, + "loss": 3.5378, + "step": 7181 + }, + { + "epoch": 0.378, + "grad_norm": 0.462890625, + "learning_rate": 0.00013822222222222222, + "loss": 3.4155, + "step": 7182 + }, + { + "epoch": 0.37805263157894736, + "grad_norm": 0.50390625, + "learning_rate": 0.00013821052631578947, + "loss": 3.7798, + "step": 7183 + }, + { + "epoch": 0.3781052631578947, + "grad_norm": 0.61328125, + "learning_rate": 0.00013819883040935672, + "loss": 3.2517, + "step": 7184 + }, + { + "epoch": 0.3781578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.00013818713450292397, + "loss": 3.8216, + "step": 7185 + }, + { + "epoch": 0.3782105263157895, + "grad_norm": 0.490234375, + "learning_rate": 0.00013817543859649122, + "loss": 3.79, + "step": 7186 + }, + { + "epoch": 0.37826315789473686, + "grad_norm": 0.486328125, + "learning_rate": 0.0001381637426900585, + "loss": 3.2352, + "step": 7187 + }, + { + "epoch": 0.3783157894736842, + "grad_norm": 0.490234375, + "learning_rate": 0.00013815204678362575, + "loss": 3.326, + "step": 7188 + }, + { + "epoch": 0.3783684210526316, + "grad_norm": 0.453125, + "learning_rate": 0.000138140350877193, + "loss": 3.6577, + "step": 7189 + }, + { + "epoch": 0.37842105263157894, + "grad_norm": 0.498046875, + "learning_rate": 0.00013812865497076025, + "loss": 3.6394, + "step": 7190 + }, + { + "epoch": 0.3784736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.0001381169590643275, + "loss": 3.6626, + "step": 7191 + }, + { + "epoch": 0.3785263157894737, + "grad_norm": 0.474609375, + "learning_rate": 0.00013810526315789475, + "loss": 3.7578, + "step": 7192 + }, + { + "epoch": 0.37857894736842107, + "grad_norm": 0.48046875, + "learning_rate": 0.000138093567251462, + "loss": 3.4885, + "step": 7193 + }, + { + "epoch": 0.37863157894736843, + "grad_norm": 0.47265625, + "learning_rate": 0.00013808187134502925, + "loss": 3.7384, + "step": 7194 + }, + { + "epoch": 0.3786842105263158, + "grad_norm": 0.478515625, + "learning_rate": 0.0001380701754385965, + "loss": 3.6898, + "step": 7195 + }, + { + "epoch": 0.37873684210526315, + "grad_norm": 0.515625, + "learning_rate": 0.00013805847953216375, + "loss": 3.4473, + "step": 7196 + }, + { + "epoch": 0.3787894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.000138046783625731, + "loss": 3.275, + "step": 7197 + }, + { + "epoch": 0.37884210526315787, + "grad_norm": 0.5078125, + "learning_rate": 0.00013803508771929824, + "loss": 3.6186, + "step": 7198 + }, + { + "epoch": 0.3788947368421053, + "grad_norm": 0.470703125, + "learning_rate": 0.0001380233918128655, + "loss": 3.528, + "step": 7199 + }, + { + "epoch": 0.37894736842105264, + "grad_norm": 0.4765625, + "learning_rate": 0.00013801169590643274, + "loss": 3.3121, + "step": 7200 + }, + { + "epoch": 0.379, + "grad_norm": 0.458984375, + "learning_rate": 0.000138, + "loss": 3.7054, + "step": 7201 + }, + { + "epoch": 0.37905263157894736, + "grad_norm": 0.5625, + "learning_rate": 0.00013798830409356727, + "loss": 3.5035, + "step": 7202 + }, + { + "epoch": 0.3791052631578947, + "grad_norm": 0.57421875, + "learning_rate": 0.00013797660818713452, + "loss": 3.4092, + "step": 7203 + }, + { + "epoch": 0.3791578947368421, + "grad_norm": 0.58203125, + "learning_rate": 0.00013796491228070177, + "loss": 3.4578, + "step": 7204 + }, + { + "epoch": 0.3792105263157895, + "grad_norm": 0.5703125, + "learning_rate": 0.00013795321637426902, + "loss": 3.3728, + "step": 7205 + }, + { + "epoch": 0.37926315789473686, + "grad_norm": 0.474609375, + "learning_rate": 0.00013794152046783627, + "loss": 3.44, + "step": 7206 + }, + { + "epoch": 0.3793157894736842, + "grad_norm": 0.4609375, + "learning_rate": 0.00013792982456140352, + "loss": 3.8157, + "step": 7207 + }, + { + "epoch": 0.3793684210526316, + "grad_norm": 0.62890625, + "learning_rate": 0.00013791812865497077, + "loss": 3.6995, + "step": 7208 + }, + { + "epoch": 0.37942105263157894, + "grad_norm": 0.5234375, + "learning_rate": 0.00013790643274853802, + "loss": 3.7489, + "step": 7209 + }, + { + "epoch": 0.3794736842105263, + "grad_norm": 0.5, + "learning_rate": 0.00013789473684210527, + "loss": 3.34, + "step": 7210 + }, + { + "epoch": 0.3795263157894737, + "grad_norm": 0.478515625, + "learning_rate": 0.00013788304093567252, + "loss": 3.7269, + "step": 7211 + }, + { + "epoch": 0.37957894736842107, + "grad_norm": 0.484375, + "learning_rate": 0.00013787134502923977, + "loss": 3.4026, + "step": 7212 + }, + { + "epoch": 0.37963157894736843, + "grad_norm": 0.474609375, + "learning_rate": 0.00013785964912280702, + "loss": 3.6639, + "step": 7213 + }, + { + "epoch": 0.3796842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.00013784795321637427, + "loss": 3.3681, + "step": 7214 + }, + { + "epoch": 0.37973684210526315, + "grad_norm": 0.462890625, + "learning_rate": 0.00013783625730994151, + "loss": 3.714, + "step": 7215 + }, + { + "epoch": 0.3797894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.0001378245614035088, + "loss": 3.3378, + "step": 7216 + }, + { + "epoch": 0.37984210526315787, + "grad_norm": 0.53125, + "learning_rate": 0.00013781286549707604, + "loss": 3.7765, + "step": 7217 + }, + { + "epoch": 0.3798947368421053, + "grad_norm": 0.53125, + "learning_rate": 0.0001378011695906433, + "loss": 3.4714, + "step": 7218 + }, + { + "epoch": 0.37994736842105264, + "grad_norm": 0.53515625, + "learning_rate": 0.00013778947368421054, + "loss": 3.4723, + "step": 7219 + }, + { + "epoch": 0.38, + "grad_norm": 0.51171875, + "learning_rate": 0.0001377777777777778, + "loss": 3.673, + "step": 7220 + }, + { + "epoch": 0.38005263157894736, + "grad_norm": 0.48828125, + "learning_rate": 0.00013776608187134504, + "loss": 3.6938, + "step": 7221 + }, + { + "epoch": 0.3801052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.0001377543859649123, + "loss": 3.3147, + "step": 7222 + }, + { + "epoch": 0.3801578947368421, + "grad_norm": 0.56640625, + "learning_rate": 0.00013774269005847954, + "loss": 3.6046, + "step": 7223 + }, + { + "epoch": 0.3802105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.0001377309941520468, + "loss": 3.7447, + "step": 7224 + }, + { + "epoch": 0.38026315789473686, + "grad_norm": 0.515625, + "learning_rate": 0.00013771929824561404, + "loss": 3.4571, + "step": 7225 + }, + { + "epoch": 0.3803157894736842, + "grad_norm": 0.478515625, + "learning_rate": 0.0001377076023391813, + "loss": 3.6764, + "step": 7226 + }, + { + "epoch": 0.3803684210526316, + "grad_norm": 0.55859375, + "learning_rate": 0.00013769590643274854, + "loss": 3.4353, + "step": 7227 + }, + { + "epoch": 0.38042105263157894, + "grad_norm": 0.482421875, + "learning_rate": 0.0001376842105263158, + "loss": 3.4293, + "step": 7228 + }, + { + "epoch": 0.3804736842105263, + "grad_norm": 0.48828125, + "learning_rate": 0.00013767251461988304, + "loss": 3.5163, + "step": 7229 + }, + { + "epoch": 0.38052631578947366, + "grad_norm": 0.515625, + "learning_rate": 0.00013766081871345029, + "loss": 3.4639, + "step": 7230 + }, + { + "epoch": 0.38057894736842107, + "grad_norm": 0.49609375, + "learning_rate": 0.00013764912280701756, + "loss": 3.8794, + "step": 7231 + }, + { + "epoch": 0.38063157894736843, + "grad_norm": 0.50390625, + "learning_rate": 0.0001376374269005848, + "loss": 3.5608, + "step": 7232 + }, + { + "epoch": 0.3806842105263158, + "grad_norm": 0.5, + "learning_rate": 0.00013762573099415206, + "loss": 4.1401, + "step": 7233 + }, + { + "epoch": 0.38073684210526315, + "grad_norm": 0.51953125, + "learning_rate": 0.0001376140350877193, + "loss": 3.612, + "step": 7234 + }, + { + "epoch": 0.3807894736842105, + "grad_norm": 0.51171875, + "learning_rate": 0.00013760233918128656, + "loss": 3.6353, + "step": 7235 + }, + { + "epoch": 0.38084210526315787, + "grad_norm": 0.5078125, + "learning_rate": 0.0001375906432748538, + "loss": 3.3278, + "step": 7236 + }, + { + "epoch": 0.3808947368421053, + "grad_norm": 0.73046875, + "learning_rate": 0.00013757894736842106, + "loss": 3.5422, + "step": 7237 + }, + { + "epoch": 0.38094736842105265, + "grad_norm": 0.470703125, + "learning_rate": 0.0001375672514619883, + "loss": 3.678, + "step": 7238 + }, + { + "epoch": 0.381, + "grad_norm": 0.61328125, + "learning_rate": 0.00013755555555555556, + "loss": 3.3726, + "step": 7239 + }, + { + "epoch": 0.38105263157894737, + "grad_norm": 0.52734375, + "learning_rate": 0.0001375438596491228, + "loss": 3.5578, + "step": 7240 + }, + { + "epoch": 0.3811052631578947, + "grad_norm": 0.51953125, + "learning_rate": 0.00013753216374269006, + "loss": 3.4637, + "step": 7241 + }, + { + "epoch": 0.3811578947368421, + "grad_norm": 0.57421875, + "learning_rate": 0.0001375204678362573, + "loss": 3.3391, + "step": 7242 + }, + { + "epoch": 0.3812105263157895, + "grad_norm": 0.462890625, + "learning_rate": 0.00013750877192982456, + "loss": 3.8151, + "step": 7243 + }, + { + "epoch": 0.38126315789473686, + "grad_norm": 0.5, + "learning_rate": 0.0001374970760233918, + "loss": 3.4601, + "step": 7244 + }, + { + "epoch": 0.3813157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00013748538011695906, + "loss": 3.4736, + "step": 7245 + }, + { + "epoch": 0.3813684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00013747368421052633, + "loss": 3.7459, + "step": 7246 + }, + { + "epoch": 0.38142105263157894, + "grad_norm": 0.54296875, + "learning_rate": 0.00013746198830409358, + "loss": 3.4592, + "step": 7247 + }, + { + "epoch": 0.3814736842105263, + "grad_norm": 0.50390625, + "learning_rate": 0.00013745029239766083, + "loss": 3.651, + "step": 7248 + }, + { + "epoch": 0.38152631578947366, + "grad_norm": 0.48046875, + "learning_rate": 0.00013743859649122808, + "loss": 3.6149, + "step": 7249 + }, + { + "epoch": 0.3815789473684211, + "grad_norm": 0.51953125, + "learning_rate": 0.00013742690058479533, + "loss": 3.5537, + "step": 7250 + }, + { + "epoch": 0.38163157894736843, + "grad_norm": 0.494140625, + "learning_rate": 0.00013741520467836258, + "loss": 3.3085, + "step": 7251 + }, + { + "epoch": 0.3816842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00013740350877192983, + "loss": 3.5994, + "step": 7252 + }, + { + "epoch": 0.38173684210526315, + "grad_norm": 0.51171875, + "learning_rate": 0.00013739181286549708, + "loss": 3.7777, + "step": 7253 + }, + { + "epoch": 0.3817894736842105, + "grad_norm": 0.474609375, + "learning_rate": 0.00013738011695906433, + "loss": 3.546, + "step": 7254 + }, + { + "epoch": 0.38184210526315787, + "grad_norm": 0.48046875, + "learning_rate": 0.00013736842105263158, + "loss": 3.452, + "step": 7255 + }, + { + "epoch": 0.3818947368421053, + "grad_norm": 0.55859375, + "learning_rate": 0.00013735672514619883, + "loss": 3.2398, + "step": 7256 + }, + { + "epoch": 0.38194736842105265, + "grad_norm": 0.486328125, + "learning_rate": 0.00013734502923976608, + "loss": 3.4165, + "step": 7257 + }, + { + "epoch": 0.382, + "grad_norm": 0.46875, + "learning_rate": 0.00013733333333333333, + "loss": 3.0812, + "step": 7258 + }, + { + "epoch": 0.38205263157894737, + "grad_norm": 0.4765625, + "learning_rate": 0.00013732163742690058, + "loss": 3.7698, + "step": 7259 + }, + { + "epoch": 0.3821052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00013730994152046783, + "loss": 3.7134, + "step": 7260 + }, + { + "epoch": 0.3821578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.0001372982456140351, + "loss": 3.4088, + "step": 7261 + }, + { + "epoch": 0.3822105263157895, + "grad_norm": 0.47265625, + "learning_rate": 0.00013728654970760235, + "loss": 3.7234, + "step": 7262 + }, + { + "epoch": 0.38226315789473686, + "grad_norm": 0.48828125, + "learning_rate": 0.0001372748538011696, + "loss": 3.604, + "step": 7263 + }, + { + "epoch": 0.3823157894736842, + "grad_norm": 0.47265625, + "learning_rate": 0.00013726315789473685, + "loss": 3.3603, + "step": 7264 + }, + { + "epoch": 0.3823684210526316, + "grad_norm": 0.484375, + "learning_rate": 0.0001372514619883041, + "loss": 3.6482, + "step": 7265 + }, + { + "epoch": 0.38242105263157894, + "grad_norm": 0.482421875, + "learning_rate": 0.00013723976608187135, + "loss": 3.2854, + "step": 7266 + }, + { + "epoch": 0.3824736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.0001372280701754386, + "loss": 3.3451, + "step": 7267 + }, + { + "epoch": 0.38252631578947366, + "grad_norm": 0.49609375, + "learning_rate": 0.00013721637426900588, + "loss": 3.7102, + "step": 7268 + }, + { + "epoch": 0.3825789473684211, + "grad_norm": 0.478515625, + "learning_rate": 0.0001372046783625731, + "loss": 3.3502, + "step": 7269 + }, + { + "epoch": 0.38263157894736843, + "grad_norm": 0.5078125, + "learning_rate": 0.00013719298245614035, + "loss": 2.9784, + "step": 7270 + }, + { + "epoch": 0.3826842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.0001371812865497076, + "loss": 3.9672, + "step": 7271 + }, + { + "epoch": 0.38273684210526315, + "grad_norm": 0.49609375, + "learning_rate": 0.00013716959064327485, + "loss": 3.5203, + "step": 7272 + }, + { + "epoch": 0.3827894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.0001371578947368421, + "loss": 3.428, + "step": 7273 + }, + { + "epoch": 0.3828421052631579, + "grad_norm": 0.478515625, + "learning_rate": 0.00013714619883040935, + "loss": 3.5746, + "step": 7274 + }, + { + "epoch": 0.3828947368421053, + "grad_norm": 0.5546875, + "learning_rate": 0.0001371345029239766, + "loss": 4.019, + "step": 7275 + }, + { + "epoch": 0.38294736842105265, + "grad_norm": 0.54296875, + "learning_rate": 0.00013712280701754388, + "loss": 3.5092, + "step": 7276 + }, + { + "epoch": 0.383, + "grad_norm": 0.578125, + "learning_rate": 0.00013711111111111113, + "loss": 3.3364, + "step": 7277 + }, + { + "epoch": 0.38305263157894737, + "grad_norm": 0.5703125, + "learning_rate": 0.00013709941520467837, + "loss": 3.7545, + "step": 7278 + }, + { + "epoch": 0.3831052631578947, + "grad_norm": 0.55859375, + "learning_rate": 0.00013708771929824562, + "loss": 3.6622, + "step": 7279 + }, + { + "epoch": 0.3831578947368421, + "grad_norm": 0.5234375, + "learning_rate": 0.00013707602339181287, + "loss": 3.5862, + "step": 7280 + }, + { + "epoch": 0.3832105263157895, + "grad_norm": 0.515625, + "learning_rate": 0.00013706432748538012, + "loss": 3.6342, + "step": 7281 + }, + { + "epoch": 0.38326315789473686, + "grad_norm": 0.53515625, + "learning_rate": 0.00013705263157894737, + "loss": 3.3883, + "step": 7282 + }, + { + "epoch": 0.3833157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00013704093567251465, + "loss": 3.4945, + "step": 7283 + }, + { + "epoch": 0.3833684210526316, + "grad_norm": 0.498046875, + "learning_rate": 0.00013702923976608187, + "loss": 3.6124, + "step": 7284 + }, + { + "epoch": 0.38342105263157894, + "grad_norm": 0.5859375, + "learning_rate": 0.00013701754385964912, + "loss": 3.7245, + "step": 7285 + }, + { + "epoch": 0.3834736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.00013700584795321637, + "loss": 3.46, + "step": 7286 + }, + { + "epoch": 0.38352631578947366, + "grad_norm": 0.48828125, + "learning_rate": 0.00013699415204678362, + "loss": 3.4883, + "step": 7287 + }, + { + "epoch": 0.3835789473684211, + "grad_norm": 0.46875, + "learning_rate": 0.00013698245614035087, + "loss": 3.4962, + "step": 7288 + }, + { + "epoch": 0.38363157894736843, + "grad_norm": 0.48046875, + "learning_rate": 0.00013697076023391812, + "loss": 3.7618, + "step": 7289 + }, + { + "epoch": 0.3836842105263158, + "grad_norm": 0.48046875, + "learning_rate": 0.00013695906432748537, + "loss": 3.7483, + "step": 7290 + }, + { + "epoch": 0.38373684210526315, + "grad_norm": 0.53515625, + "learning_rate": 0.00013694736842105265, + "loss": 3.2857, + "step": 7291 + }, + { + "epoch": 0.3837894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.0001369356725146199, + "loss": 3.7083, + "step": 7292 + }, + { + "epoch": 0.3838421052631579, + "grad_norm": 0.55859375, + "learning_rate": 0.00013692397660818715, + "loss": 3.3076, + "step": 7293 + }, + { + "epoch": 0.3838947368421053, + "grad_norm": 0.55859375, + "learning_rate": 0.0001369122807017544, + "loss": 3.5926, + "step": 7294 + }, + { + "epoch": 0.38394736842105265, + "grad_norm": 0.54296875, + "learning_rate": 0.00013690058479532164, + "loss": 3.5287, + "step": 7295 + }, + { + "epoch": 0.384, + "grad_norm": 0.498046875, + "learning_rate": 0.0001368888888888889, + "loss": 3.301, + "step": 7296 + }, + { + "epoch": 0.38405263157894737, + "grad_norm": 0.64453125, + "learning_rate": 0.00013687719298245614, + "loss": 3.7213, + "step": 7297 + }, + { + "epoch": 0.3841052631578947, + "grad_norm": 0.4765625, + "learning_rate": 0.00013686549707602342, + "loss": 3.2729, + "step": 7298 + }, + { + "epoch": 0.3841578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.00013685380116959064, + "loss": 3.8799, + "step": 7299 + }, + { + "epoch": 0.38421052631578945, + "grad_norm": 0.51171875, + "learning_rate": 0.0001368421052631579, + "loss": 3.2689, + "step": 7300 + }, + { + "epoch": 0.38426315789473686, + "grad_norm": 0.4765625, + "learning_rate": 0.00013683040935672514, + "loss": 3.6117, + "step": 7301 + }, + { + "epoch": 0.3843157894736842, + "grad_norm": 0.490234375, + "learning_rate": 0.0001368187134502924, + "loss": 3.6172, + "step": 7302 + }, + { + "epoch": 0.3843684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00013680701754385964, + "loss": 3.5512, + "step": 7303 + }, + { + "epoch": 0.38442105263157894, + "grad_norm": 0.52734375, + "learning_rate": 0.0001367953216374269, + "loss": 3.7343, + "step": 7304 + }, + { + "epoch": 0.3844736842105263, + "grad_norm": 0.494140625, + "learning_rate": 0.00013678362573099417, + "loss": 3.9169, + "step": 7305 + }, + { + "epoch": 0.38452631578947366, + "grad_norm": 0.490234375, + "learning_rate": 0.00013677192982456142, + "loss": 3.6627, + "step": 7306 + }, + { + "epoch": 0.3845789473684211, + "grad_norm": 0.48046875, + "learning_rate": 0.00013676023391812867, + "loss": 3.7108, + "step": 7307 + }, + { + "epoch": 0.38463157894736844, + "grad_norm": 0.5, + "learning_rate": 0.00013674853801169592, + "loss": 3.3057, + "step": 7308 + }, + { + "epoch": 0.3846842105263158, + "grad_norm": 0.458984375, + "learning_rate": 0.00013673684210526317, + "loss": 3.0894, + "step": 7309 + }, + { + "epoch": 0.38473684210526315, + "grad_norm": 0.494140625, + "learning_rate": 0.00013672514619883042, + "loss": 3.4151, + "step": 7310 + }, + { + "epoch": 0.3847894736842105, + "grad_norm": 0.4765625, + "learning_rate": 0.00013671345029239767, + "loss": 3.8189, + "step": 7311 + }, + { + "epoch": 0.3848421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00013670175438596491, + "loss": 3.8943, + "step": 7312 + }, + { + "epoch": 0.3848947368421053, + "grad_norm": 0.55859375, + "learning_rate": 0.0001366900584795322, + "loss": 3.1774, + "step": 7313 + }, + { + "epoch": 0.38494736842105265, + "grad_norm": 0.5078125, + "learning_rate": 0.00013667836257309944, + "loss": 3.717, + "step": 7314 + }, + { + "epoch": 0.385, + "grad_norm": 0.515625, + "learning_rate": 0.00013666666666666666, + "loss": 3.9377, + "step": 7315 + }, + { + "epoch": 0.38505263157894737, + "grad_norm": 0.53515625, + "learning_rate": 0.0001366549707602339, + "loss": 3.7914, + "step": 7316 + }, + { + "epoch": 0.38510526315789473, + "grad_norm": 0.546875, + "learning_rate": 0.00013664327485380116, + "loss": 3.8594, + "step": 7317 + }, + { + "epoch": 0.3851578947368421, + "grad_norm": 0.58984375, + "learning_rate": 0.0001366315789473684, + "loss": 3.5135, + "step": 7318 + }, + { + "epoch": 0.38521052631578945, + "grad_norm": 0.5, + "learning_rate": 0.00013661988304093566, + "loss": 3.312, + "step": 7319 + }, + { + "epoch": 0.38526315789473686, + "grad_norm": 0.48828125, + "learning_rate": 0.00013660818713450294, + "loss": 3.2921, + "step": 7320 + }, + { + "epoch": 0.3853157894736842, + "grad_norm": 0.51953125, + "learning_rate": 0.0001365964912280702, + "loss": 3.54, + "step": 7321 + }, + { + "epoch": 0.3853684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00013658479532163744, + "loss": 3.7286, + "step": 7322 + }, + { + "epoch": 0.38542105263157894, + "grad_norm": 0.470703125, + "learning_rate": 0.0001365730994152047, + "loss": 3.7761, + "step": 7323 + }, + { + "epoch": 0.3854736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00013656140350877194, + "loss": 3.86, + "step": 7324 + }, + { + "epoch": 0.38552631578947366, + "grad_norm": 0.5, + "learning_rate": 0.0001365497076023392, + "loss": 3.5025, + "step": 7325 + }, + { + "epoch": 0.3855789473684211, + "grad_norm": 0.5, + "learning_rate": 0.00013653801169590644, + "loss": 3.4501, + "step": 7326 + }, + { + "epoch": 0.38563157894736844, + "grad_norm": 0.52734375, + "learning_rate": 0.00013652631578947369, + "loss": 3.8078, + "step": 7327 + }, + { + "epoch": 0.3856842105263158, + "grad_norm": 0.54296875, + "learning_rate": 0.00013651461988304096, + "loss": 3.7844, + "step": 7328 + }, + { + "epoch": 0.38573684210526316, + "grad_norm": 0.494140625, + "learning_rate": 0.0001365029239766082, + "loss": 3.5867, + "step": 7329 + }, + { + "epoch": 0.3857894736842105, + "grad_norm": 0.48828125, + "learning_rate": 0.00013649122807017543, + "loss": 3.4112, + "step": 7330 + }, + { + "epoch": 0.3858421052631579, + "grad_norm": 0.470703125, + "learning_rate": 0.00013647953216374268, + "loss": 3.6896, + "step": 7331 + }, + { + "epoch": 0.3858947368421053, + "grad_norm": 0.48046875, + "learning_rate": 0.00013646783625730993, + "loss": 3.7315, + "step": 7332 + }, + { + "epoch": 0.38594736842105265, + "grad_norm": 0.46484375, + "learning_rate": 0.00013645614035087718, + "loss": 3.5648, + "step": 7333 + }, + { + "epoch": 0.386, + "grad_norm": 0.5546875, + "learning_rate": 0.00013644444444444443, + "loss": 3.9156, + "step": 7334 + }, + { + "epoch": 0.38605263157894737, + "grad_norm": 0.48046875, + "learning_rate": 0.0001364327485380117, + "loss": 3.7953, + "step": 7335 + }, + { + "epoch": 0.38610526315789473, + "grad_norm": 0.57421875, + "learning_rate": 0.00013642105263157896, + "loss": 3.4117, + "step": 7336 + }, + { + "epoch": 0.3861578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.0001364093567251462, + "loss": 3.6712, + "step": 7337 + }, + { + "epoch": 0.38621052631578945, + "grad_norm": 0.609375, + "learning_rate": 0.00013639766081871346, + "loss": 3.4999, + "step": 7338 + }, + { + "epoch": 0.38626315789473686, + "grad_norm": 0.48046875, + "learning_rate": 0.0001363859649122807, + "loss": 3.312, + "step": 7339 + }, + { + "epoch": 0.3863157894736842, + "grad_norm": 0.498046875, + "learning_rate": 0.00013637426900584796, + "loss": 3.3589, + "step": 7340 + }, + { + "epoch": 0.3863684210526316, + "grad_norm": 0.54296875, + "learning_rate": 0.0001363625730994152, + "loss": 3.9134, + "step": 7341 + }, + { + "epoch": 0.38642105263157894, + "grad_norm": 0.53515625, + "learning_rate": 0.00013635087719298248, + "loss": 3.2909, + "step": 7342 + }, + { + "epoch": 0.3864736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00013633918128654973, + "loss": 3.5267, + "step": 7343 + }, + { + "epoch": 0.38652631578947366, + "grad_norm": 0.57421875, + "learning_rate": 0.00013632748538011698, + "loss": 3.7673, + "step": 7344 + }, + { + "epoch": 0.3865789473684211, + "grad_norm": 0.49609375, + "learning_rate": 0.0001363157894736842, + "loss": 3.8439, + "step": 7345 + }, + { + "epoch": 0.38663157894736844, + "grad_norm": 0.515625, + "learning_rate": 0.00013630409356725146, + "loss": 3.5629, + "step": 7346 + }, + { + "epoch": 0.3866842105263158, + "grad_norm": 0.484375, + "learning_rate": 0.0001362923976608187, + "loss": 3.6769, + "step": 7347 + }, + { + "epoch": 0.38673684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00013628070175438595, + "loss": 3.5917, + "step": 7348 + }, + { + "epoch": 0.3867894736842105, + "grad_norm": 0.5859375, + "learning_rate": 0.0001362690058479532, + "loss": 3.7382, + "step": 7349 + }, + { + "epoch": 0.3868421052631579, + "grad_norm": 0.4609375, + "learning_rate": 0.00013625730994152048, + "loss": 3.4525, + "step": 7350 + }, + { + "epoch": 0.38689473684210524, + "grad_norm": 0.46875, + "learning_rate": 0.00013624561403508773, + "loss": 3.5296, + "step": 7351 + }, + { + "epoch": 0.38694736842105265, + "grad_norm": 0.4765625, + "learning_rate": 0.00013623391812865498, + "loss": 3.4437, + "step": 7352 + }, + { + "epoch": 0.387, + "grad_norm": 0.50390625, + "learning_rate": 0.00013622222222222223, + "loss": 3.5436, + "step": 7353 + }, + { + "epoch": 0.38705263157894737, + "grad_norm": 0.486328125, + "learning_rate": 0.00013621052631578948, + "loss": 3.7335, + "step": 7354 + }, + { + "epoch": 0.38710526315789473, + "grad_norm": 0.515625, + "learning_rate": 0.00013619883040935673, + "loss": 3.3778, + "step": 7355 + }, + { + "epoch": 0.3871578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.00013618713450292398, + "loss": 3.3844, + "step": 7356 + }, + { + "epoch": 0.38721052631578945, + "grad_norm": 0.482421875, + "learning_rate": 0.00013617543859649126, + "loss": 3.6268, + "step": 7357 + }, + { + "epoch": 0.38726315789473686, + "grad_norm": 0.46875, + "learning_rate": 0.0001361637426900585, + "loss": 3.7693, + "step": 7358 + }, + { + "epoch": 0.3873157894736842, + "grad_norm": 0.494140625, + "learning_rate": 0.00013615204678362575, + "loss": 3.6289, + "step": 7359 + }, + { + "epoch": 0.3873684210526316, + "grad_norm": 0.490234375, + "learning_rate": 0.000136140350877193, + "loss": 3.3061, + "step": 7360 + }, + { + "epoch": 0.38742105263157894, + "grad_norm": 0.48046875, + "learning_rate": 0.00013612865497076023, + "loss": 3.4335, + "step": 7361 + }, + { + "epoch": 0.3874736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00013611695906432748, + "loss": 3.3208, + "step": 7362 + }, + { + "epoch": 0.38752631578947366, + "grad_norm": 0.48828125, + "learning_rate": 0.00013610526315789473, + "loss": 3.5587, + "step": 7363 + }, + { + "epoch": 0.3875789473684211, + "grad_norm": 0.482421875, + "learning_rate": 0.00013609356725146198, + "loss": 3.2156, + "step": 7364 + }, + { + "epoch": 0.38763157894736844, + "grad_norm": 0.46875, + "learning_rate": 0.00013608187134502925, + "loss": 3.5054, + "step": 7365 + }, + { + "epoch": 0.3876842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.0001360701754385965, + "loss": 3.8711, + "step": 7366 + }, + { + "epoch": 0.38773684210526316, + "grad_norm": 0.48046875, + "learning_rate": 0.00013605847953216375, + "loss": 3.3135, + "step": 7367 + }, + { + "epoch": 0.3877894736842105, + "grad_norm": 0.54296875, + "learning_rate": 0.000136046783625731, + "loss": 3.4639, + "step": 7368 + }, + { + "epoch": 0.3878421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00013603508771929825, + "loss": 3.6574, + "step": 7369 + }, + { + "epoch": 0.38789473684210524, + "grad_norm": 0.54296875, + "learning_rate": 0.0001360233918128655, + "loss": 3.2524, + "step": 7370 + }, + { + "epoch": 0.38794736842105265, + "grad_norm": 0.48828125, + "learning_rate": 0.00013601169590643275, + "loss": 3.1312, + "step": 7371 + }, + { + "epoch": 0.388, + "grad_norm": 0.478515625, + "learning_rate": 0.00013600000000000003, + "loss": 3.3351, + "step": 7372 + }, + { + "epoch": 0.38805263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00013598830409356728, + "loss": 3.5535, + "step": 7373 + }, + { + "epoch": 0.38810526315789473, + "grad_norm": 0.486328125, + "learning_rate": 0.00013597660818713453, + "loss": 3.3142, + "step": 7374 + }, + { + "epoch": 0.3881578947368421, + "grad_norm": 0.48828125, + "learning_rate": 0.00013596491228070177, + "loss": 3.8077, + "step": 7375 + }, + { + "epoch": 0.38821052631578945, + "grad_norm": 0.5234375, + "learning_rate": 0.000135953216374269, + "loss": 3.3115, + "step": 7376 + }, + { + "epoch": 0.38826315789473687, + "grad_norm": 0.5, + "learning_rate": 0.00013594152046783625, + "loss": 3.6004, + "step": 7377 + }, + { + "epoch": 0.3883157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.0001359298245614035, + "loss": 3.4075, + "step": 7378 + }, + { + "epoch": 0.3883684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00013591812865497077, + "loss": 3.6253, + "step": 7379 + }, + { + "epoch": 0.38842105263157894, + "grad_norm": 0.478515625, + "learning_rate": 0.00013590643274853802, + "loss": 3.4257, + "step": 7380 + }, + { + "epoch": 0.3884736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00013589473684210527, + "loss": 3.8423, + "step": 7381 + }, + { + "epoch": 0.38852631578947366, + "grad_norm": 0.498046875, + "learning_rate": 0.00013588304093567252, + "loss": 3.2843, + "step": 7382 + }, + { + "epoch": 0.3885789473684211, + "grad_norm": 0.484375, + "learning_rate": 0.00013587134502923977, + "loss": 3.4355, + "step": 7383 + }, + { + "epoch": 0.38863157894736844, + "grad_norm": 0.470703125, + "learning_rate": 0.00013585964912280702, + "loss": 3.703, + "step": 7384 + }, + { + "epoch": 0.3886842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00013584795321637427, + "loss": 3.6738, + "step": 7385 + }, + { + "epoch": 0.38873684210526316, + "grad_norm": 0.474609375, + "learning_rate": 0.00013583625730994152, + "loss": 3.4433, + "step": 7386 + }, + { + "epoch": 0.3887894736842105, + "grad_norm": 0.49609375, + "learning_rate": 0.0001358245614035088, + "loss": 3.8173, + "step": 7387 + }, + { + "epoch": 0.3888421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00013581286549707605, + "loss": 3.9771, + "step": 7388 + }, + { + "epoch": 0.38889473684210524, + "grad_norm": 0.58203125, + "learning_rate": 0.0001358011695906433, + "loss": 3.7318, + "step": 7389 + }, + { + "epoch": 0.38894736842105265, + "grad_norm": 0.58984375, + "learning_rate": 0.00013578947368421055, + "loss": 3.7332, + "step": 7390 + }, + { + "epoch": 0.389, + "grad_norm": 0.625, + "learning_rate": 0.00013577777777777777, + "loss": 3.9598, + "step": 7391 + }, + { + "epoch": 0.38905263157894737, + "grad_norm": 0.55859375, + "learning_rate": 0.00013576608187134502, + "loss": 3.5417, + "step": 7392 + }, + { + "epoch": 0.38910526315789473, + "grad_norm": 0.6484375, + "learning_rate": 0.00013575438596491227, + "loss": 3.1946, + "step": 7393 + }, + { + "epoch": 0.3891578947368421, + "grad_norm": 0.5703125, + "learning_rate": 0.00013574269005847954, + "loss": 4.0259, + "step": 7394 + }, + { + "epoch": 0.38921052631578945, + "grad_norm": 0.51953125, + "learning_rate": 0.0001357309941520468, + "loss": 3.4856, + "step": 7395 + }, + { + "epoch": 0.38926315789473687, + "grad_norm": 0.53125, + "learning_rate": 0.00013571929824561404, + "loss": 3.5412, + "step": 7396 + }, + { + "epoch": 0.3893157894736842, + "grad_norm": 0.498046875, + "learning_rate": 0.0001357076023391813, + "loss": 3.9496, + "step": 7397 + }, + { + "epoch": 0.3893684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00013569590643274854, + "loss": 3.9038, + "step": 7398 + }, + { + "epoch": 0.38942105263157895, + "grad_norm": 0.484375, + "learning_rate": 0.0001356842105263158, + "loss": 3.1896, + "step": 7399 + }, + { + "epoch": 0.3894736842105263, + "grad_norm": 0.478515625, + "learning_rate": 0.00013567251461988304, + "loss": 3.7315, + "step": 7400 + }, + { + "epoch": 0.38952631578947366, + "grad_norm": 0.51171875, + "learning_rate": 0.0001356608187134503, + "loss": 3.5654, + "step": 7401 + }, + { + "epoch": 0.3895789473684211, + "grad_norm": 0.515625, + "learning_rate": 0.00013564912280701757, + "loss": 3.6444, + "step": 7402 + }, + { + "epoch": 0.38963157894736844, + "grad_norm": 0.51171875, + "learning_rate": 0.00013563742690058482, + "loss": 3.4084, + "step": 7403 + }, + { + "epoch": 0.3896842105263158, + "grad_norm": 0.60546875, + "learning_rate": 0.00013562573099415207, + "loss": 4.0316, + "step": 7404 + }, + { + "epoch": 0.38973684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00013561403508771932, + "loss": 4.0077, + "step": 7405 + }, + { + "epoch": 0.3897894736842105, + "grad_norm": 0.494140625, + "learning_rate": 0.00013560233918128657, + "loss": 3.7735, + "step": 7406 + }, + { + "epoch": 0.3898421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.0001355906432748538, + "loss": 4.2427, + "step": 7407 + }, + { + "epoch": 0.38989473684210524, + "grad_norm": 0.55078125, + "learning_rate": 0.00013557894736842104, + "loss": 3.7102, + "step": 7408 + }, + { + "epoch": 0.38994736842105265, + "grad_norm": 0.65625, + "learning_rate": 0.00013556725146198832, + "loss": 3.3196, + "step": 7409 + }, + { + "epoch": 0.39, + "grad_norm": 0.4609375, + "learning_rate": 0.00013555555555555556, + "loss": 3.1502, + "step": 7410 + }, + { + "epoch": 0.3900526315789474, + "grad_norm": 0.4921875, + "learning_rate": 0.00013554385964912281, + "loss": 3.7595, + "step": 7411 + }, + { + "epoch": 0.39010526315789473, + "grad_norm": 0.5390625, + "learning_rate": 0.00013553216374269006, + "loss": 3.9203, + "step": 7412 + }, + { + "epoch": 0.3901578947368421, + "grad_norm": 0.640625, + "learning_rate": 0.0001355204678362573, + "loss": 3.3111, + "step": 7413 + }, + { + "epoch": 0.39021052631578945, + "grad_norm": 0.5625, + "learning_rate": 0.00013550877192982456, + "loss": 3.4691, + "step": 7414 + }, + { + "epoch": 0.39026315789473687, + "grad_norm": 0.55078125, + "learning_rate": 0.0001354970760233918, + "loss": 3.8396, + "step": 7415 + }, + { + "epoch": 0.3903157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00013548538011695906, + "loss": 3.5245, + "step": 7416 + }, + { + "epoch": 0.3903684210526316, + "grad_norm": 0.482421875, + "learning_rate": 0.00013547368421052634, + "loss": 3.8464, + "step": 7417 + }, + { + "epoch": 0.39042105263157895, + "grad_norm": 0.474609375, + "learning_rate": 0.0001354619883040936, + "loss": 3.2696, + "step": 7418 + }, + { + "epoch": 0.3904736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00013545029239766084, + "loss": 3.6545, + "step": 7419 + }, + { + "epoch": 0.39052631578947367, + "grad_norm": 0.45703125, + "learning_rate": 0.0001354385964912281, + "loss": 3.6612, + "step": 7420 + }, + { + "epoch": 0.390578947368421, + "grad_norm": 0.46484375, + "learning_rate": 0.00013542690058479534, + "loss": 3.3697, + "step": 7421 + }, + { + "epoch": 0.39063157894736844, + "grad_norm": 0.51953125, + "learning_rate": 0.00013541520467836256, + "loss": 3.2615, + "step": 7422 + }, + { + "epoch": 0.3906842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.0001354035087719298, + "loss": 3.793, + "step": 7423 + }, + { + "epoch": 0.39073684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00013539181286549709, + "loss": 3.5886, + "step": 7424 + }, + { + "epoch": 0.3907894736842105, + "grad_norm": 0.52734375, + "learning_rate": 0.00013538011695906434, + "loss": 3.5107, + "step": 7425 + }, + { + "epoch": 0.3908421052631579, + "grad_norm": 0.478515625, + "learning_rate": 0.00013536842105263159, + "loss": 3.8103, + "step": 7426 + }, + { + "epoch": 0.39089473684210524, + "grad_norm": 0.48828125, + "learning_rate": 0.00013535672514619883, + "loss": 3.5268, + "step": 7427 + }, + { + "epoch": 0.39094736842105265, + "grad_norm": 0.486328125, + "learning_rate": 0.00013534502923976608, + "loss": 3.3761, + "step": 7428 + }, + { + "epoch": 0.391, + "grad_norm": 0.49609375, + "learning_rate": 0.00013533333333333333, + "loss": 3.6724, + "step": 7429 + }, + { + "epoch": 0.3910526315789474, + "grad_norm": 0.47265625, + "learning_rate": 0.00013532163742690058, + "loss": 3.5709, + "step": 7430 + }, + { + "epoch": 0.39110526315789473, + "grad_norm": 0.458984375, + "learning_rate": 0.00013530994152046786, + "loss": 3.5127, + "step": 7431 + }, + { + "epoch": 0.3911578947368421, + "grad_norm": 0.482421875, + "learning_rate": 0.0001352982456140351, + "loss": 3.8475, + "step": 7432 + }, + { + "epoch": 0.39121052631578945, + "grad_norm": 0.46875, + "learning_rate": 0.00013528654970760236, + "loss": 3.8472, + "step": 7433 + }, + { + "epoch": 0.39126315789473687, + "grad_norm": 0.474609375, + "learning_rate": 0.0001352748538011696, + "loss": 3.734, + "step": 7434 + }, + { + "epoch": 0.39131578947368423, + "grad_norm": 0.490234375, + "learning_rate": 0.00013526315789473686, + "loss": 3.8081, + "step": 7435 + }, + { + "epoch": 0.3913684210526316, + "grad_norm": 0.515625, + "learning_rate": 0.0001352514619883041, + "loss": 3.509, + "step": 7436 + }, + { + "epoch": 0.39142105263157895, + "grad_norm": 0.46875, + "learning_rate": 0.00013523976608187133, + "loss": 3.3763, + "step": 7437 + }, + { + "epoch": 0.3914736842105263, + "grad_norm": 0.578125, + "learning_rate": 0.00013522807017543858, + "loss": 3.5003, + "step": 7438 + }, + { + "epoch": 0.39152631578947367, + "grad_norm": 0.55859375, + "learning_rate": 0.00013521637426900586, + "loss": 3.3802, + "step": 7439 + }, + { + "epoch": 0.391578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.0001352046783625731, + "loss": 3.2961, + "step": 7440 + }, + { + "epoch": 0.39163157894736844, + "grad_norm": 0.94921875, + "learning_rate": 0.00013519298245614036, + "loss": 3.8494, + "step": 7441 + }, + { + "epoch": 0.3916842105263158, + "grad_norm": 0.48046875, + "learning_rate": 0.0001351812865497076, + "loss": 4.0912, + "step": 7442 + }, + { + "epoch": 0.39173684210526316, + "grad_norm": 0.515625, + "learning_rate": 0.00013516959064327486, + "loss": 3.6464, + "step": 7443 + }, + { + "epoch": 0.3917894736842105, + "grad_norm": 0.52734375, + "learning_rate": 0.0001351578947368421, + "loss": 3.776, + "step": 7444 + }, + { + "epoch": 0.3918421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00013514619883040935, + "loss": 3.7785, + "step": 7445 + }, + { + "epoch": 0.39189473684210524, + "grad_norm": 0.546875, + "learning_rate": 0.00013513450292397663, + "loss": 3.1642, + "step": 7446 + }, + { + "epoch": 0.39194736842105266, + "grad_norm": 0.486328125, + "learning_rate": 0.00013512280701754388, + "loss": 3.4858, + "step": 7447 + }, + { + "epoch": 0.392, + "grad_norm": 0.46875, + "learning_rate": 0.00013511111111111113, + "loss": 3.4888, + "step": 7448 + }, + { + "epoch": 0.3920526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.00013509941520467838, + "loss": 3.7864, + "step": 7449 + }, + { + "epoch": 0.39210526315789473, + "grad_norm": 0.5546875, + "learning_rate": 0.00013508771929824563, + "loss": 3.6009, + "step": 7450 + }, + { + "epoch": 0.3921578947368421, + "grad_norm": 0.48046875, + "learning_rate": 0.00013507602339181288, + "loss": 3.8981, + "step": 7451 + }, + { + "epoch": 0.39221052631578945, + "grad_norm": 0.5390625, + "learning_rate": 0.00013506432748538013, + "loss": 4.1359, + "step": 7452 + }, + { + "epoch": 0.39226315789473687, + "grad_norm": 0.494140625, + "learning_rate": 0.00013505263157894735, + "loss": 2.9329, + "step": 7453 + }, + { + "epoch": 0.39231578947368423, + "grad_norm": 0.5390625, + "learning_rate": 0.00013504093567251463, + "loss": 3.546, + "step": 7454 + }, + { + "epoch": 0.3923684210526316, + "grad_norm": 0.54296875, + "learning_rate": 0.00013502923976608188, + "loss": 3.5735, + "step": 7455 + }, + { + "epoch": 0.39242105263157895, + "grad_norm": 0.5390625, + "learning_rate": 0.00013501754385964913, + "loss": 3.7483, + "step": 7456 + }, + { + "epoch": 0.3924736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.00013500584795321638, + "loss": 3.532, + "step": 7457 + }, + { + "epoch": 0.39252631578947367, + "grad_norm": 0.5390625, + "learning_rate": 0.00013499415204678363, + "loss": 3.5453, + "step": 7458 + }, + { + "epoch": 0.392578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00013498245614035088, + "loss": 3.654, + "step": 7459 + }, + { + "epoch": 0.39263157894736844, + "grad_norm": 0.50390625, + "learning_rate": 0.00013497076023391813, + "loss": 3.9906, + "step": 7460 + }, + { + "epoch": 0.3926842105263158, + "grad_norm": 0.48828125, + "learning_rate": 0.0001349590643274854, + "loss": 3.338, + "step": 7461 + }, + { + "epoch": 0.39273684210526316, + "grad_norm": 0.52734375, + "learning_rate": 0.00013494736842105265, + "loss": 3.7407, + "step": 7462 + }, + { + "epoch": 0.3927894736842105, + "grad_norm": 0.474609375, + "learning_rate": 0.0001349356725146199, + "loss": 3.3276, + "step": 7463 + }, + { + "epoch": 0.3928421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00013492397660818715, + "loss": 3.2969, + "step": 7464 + }, + { + "epoch": 0.39289473684210524, + "grad_norm": 0.52734375, + "learning_rate": 0.0001349122807017544, + "loss": 3.4608, + "step": 7465 + }, + { + "epoch": 0.39294736842105266, + "grad_norm": 0.5, + "learning_rate": 0.00013490058479532165, + "loss": 3.7164, + "step": 7466 + }, + { + "epoch": 0.393, + "grad_norm": 0.52734375, + "learning_rate": 0.0001348888888888889, + "loss": 3.2386, + "step": 7467 + }, + { + "epoch": 0.3930526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.00013487719298245615, + "loss": 3.7757, + "step": 7468 + }, + { + "epoch": 0.39310526315789474, + "grad_norm": 0.484375, + "learning_rate": 0.0001348654970760234, + "loss": 3.2474, + "step": 7469 + }, + { + "epoch": 0.3931578947368421, + "grad_norm": 0.466796875, + "learning_rate": 0.00013485380116959065, + "loss": 3.9681, + "step": 7470 + }, + { + "epoch": 0.39321052631578945, + "grad_norm": 0.50390625, + "learning_rate": 0.0001348421052631579, + "loss": 3.4676, + "step": 7471 + }, + { + "epoch": 0.3932631578947368, + "grad_norm": 0.466796875, + "learning_rate": 0.00013483040935672515, + "loss": 3.7501, + "step": 7472 + }, + { + "epoch": 0.39331578947368423, + "grad_norm": 0.44140625, + "learning_rate": 0.0001348187134502924, + "loss": 3.2572, + "step": 7473 + }, + { + "epoch": 0.3933684210526316, + "grad_norm": 0.4609375, + "learning_rate": 0.00013480701754385965, + "loss": 3.3119, + "step": 7474 + }, + { + "epoch": 0.39342105263157895, + "grad_norm": 2.703125, + "learning_rate": 0.0001347953216374269, + "loss": 3.8276, + "step": 7475 + }, + { + "epoch": 0.3934736842105263, + "grad_norm": 0.47265625, + "learning_rate": 0.00013478362573099417, + "loss": 3.8133, + "step": 7476 + }, + { + "epoch": 0.39352631578947367, + "grad_norm": 0.51953125, + "learning_rate": 0.00013477192982456142, + "loss": 3.8953, + "step": 7477 + }, + { + "epoch": 0.39357894736842103, + "grad_norm": 0.484375, + "learning_rate": 0.00013476023391812867, + "loss": 3.4786, + "step": 7478 + }, + { + "epoch": 0.39363157894736844, + "grad_norm": 0.51953125, + "learning_rate": 0.00013474853801169592, + "loss": 3.4503, + "step": 7479 + }, + { + "epoch": 0.3936842105263158, + "grad_norm": 0.494140625, + "learning_rate": 0.00013473684210526317, + "loss": 3.3607, + "step": 7480 + }, + { + "epoch": 0.39373684210526316, + "grad_norm": 0.498046875, + "learning_rate": 0.00013472514619883042, + "loss": 3.7708, + "step": 7481 + }, + { + "epoch": 0.3937894736842105, + "grad_norm": 0.58984375, + "learning_rate": 0.00013471345029239767, + "loss": 3.3065, + "step": 7482 + }, + { + "epoch": 0.3938421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00013470175438596492, + "loss": 3.521, + "step": 7483 + }, + { + "epoch": 0.39389473684210524, + "grad_norm": 0.5546875, + "learning_rate": 0.00013469005847953217, + "loss": 3.8584, + "step": 7484 + }, + { + "epoch": 0.39394736842105266, + "grad_norm": 0.55859375, + "learning_rate": 0.00013467836257309942, + "loss": 3.8568, + "step": 7485 + }, + { + "epoch": 0.394, + "grad_norm": 0.52734375, + "learning_rate": 0.00013466666666666667, + "loss": 3.7919, + "step": 7486 + }, + { + "epoch": 0.3940526315789474, + "grad_norm": 0.515625, + "learning_rate": 0.00013465497076023392, + "loss": 3.547, + "step": 7487 + }, + { + "epoch": 0.39410526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.00013464327485380117, + "loss": 3.5455, + "step": 7488 + }, + { + "epoch": 0.3941578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00013463157894736842, + "loss": 3.4579, + "step": 7489 + }, + { + "epoch": 0.39421052631578946, + "grad_norm": 0.6171875, + "learning_rate": 0.00013461988304093567, + "loss": 3.4618, + "step": 7490 + }, + { + "epoch": 0.3942631578947368, + "grad_norm": 0.51171875, + "learning_rate": 0.00013460818713450294, + "loss": 3.6663, + "step": 7491 + }, + { + "epoch": 0.39431578947368423, + "grad_norm": 0.5078125, + "learning_rate": 0.0001345964912280702, + "loss": 3.4127, + "step": 7492 + }, + { + "epoch": 0.3943684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.00013458479532163744, + "loss": 3.3895, + "step": 7493 + }, + { + "epoch": 0.39442105263157895, + "grad_norm": 0.46875, + "learning_rate": 0.0001345730994152047, + "loss": 3.8119, + "step": 7494 + }, + { + "epoch": 0.3944736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00013456140350877194, + "loss": 3.4241, + "step": 7495 + }, + { + "epoch": 0.39452631578947367, + "grad_norm": 0.56640625, + "learning_rate": 0.0001345497076023392, + "loss": 3.9028, + "step": 7496 + }, + { + "epoch": 0.39457894736842103, + "grad_norm": 0.625, + "learning_rate": 0.00013453801169590644, + "loss": 3.666, + "step": 7497 + }, + { + "epoch": 0.39463157894736844, + "grad_norm": 0.47265625, + "learning_rate": 0.0001345263157894737, + "loss": 3.6711, + "step": 7498 + }, + { + "epoch": 0.3946842105263158, + "grad_norm": 0.443359375, + "learning_rate": 0.00013451461988304094, + "loss": 3.5873, + "step": 7499 + }, + { + "epoch": 0.39473684210526316, + "grad_norm": 0.455078125, + "learning_rate": 0.0001345029239766082, + "loss": 3.2074, + "step": 7500 + }, + { + "epoch": 0.3947894736842105, + "grad_norm": 0.44921875, + "learning_rate": 0.00013449122807017544, + "loss": 3.772, + "step": 7501 + }, + { + "epoch": 0.3948421052631579, + "grad_norm": 0.46484375, + "learning_rate": 0.0001344795321637427, + "loss": 3.4707, + "step": 7502 + }, + { + "epoch": 0.39489473684210524, + "grad_norm": 0.4609375, + "learning_rate": 0.00013446783625730994, + "loss": 3.3237, + "step": 7503 + }, + { + "epoch": 0.39494736842105266, + "grad_norm": 0.470703125, + "learning_rate": 0.0001344561403508772, + "loss": 3.7315, + "step": 7504 + }, + { + "epoch": 0.395, + "grad_norm": 0.45703125, + "learning_rate": 0.00013444444444444447, + "loss": 3.7291, + "step": 7505 + }, + { + "epoch": 0.3950526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.00013443274853801172, + "loss": 3.8937, + "step": 7506 + }, + { + "epoch": 0.39510526315789474, + "grad_norm": 0.5625, + "learning_rate": 0.00013442105263157896, + "loss": 3.6963, + "step": 7507 + }, + { + "epoch": 0.3951578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00013440935672514621, + "loss": 3.5063, + "step": 7508 + }, + { + "epoch": 0.39521052631578946, + "grad_norm": 0.53515625, + "learning_rate": 0.00013439766081871346, + "loss": 3.5853, + "step": 7509 + }, + { + "epoch": 0.3952631578947368, + "grad_norm": 0.470703125, + "learning_rate": 0.00013438596491228071, + "loss": 4.006, + "step": 7510 + }, + { + "epoch": 0.39531578947368423, + "grad_norm": 0.5625, + "learning_rate": 0.00013437426900584796, + "loss": 3.4026, + "step": 7511 + }, + { + "epoch": 0.3953684210526316, + "grad_norm": 0.4765625, + "learning_rate": 0.0001343625730994152, + "loss": 3.816, + "step": 7512 + }, + { + "epoch": 0.39542105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00013435087719298246, + "loss": 3.6851, + "step": 7513 + }, + { + "epoch": 0.3954736842105263, + "grad_norm": 0.490234375, + "learning_rate": 0.0001343391812865497, + "loss": 3.7755, + "step": 7514 + }, + { + "epoch": 0.39552631578947367, + "grad_norm": 0.515625, + "learning_rate": 0.00013432748538011696, + "loss": 3.5305, + "step": 7515 + }, + { + "epoch": 0.39557894736842103, + "grad_norm": 0.48828125, + "learning_rate": 0.0001343157894736842, + "loss": 3.5226, + "step": 7516 + }, + { + "epoch": 0.39563157894736845, + "grad_norm": 0.50390625, + "learning_rate": 0.00013430409356725146, + "loss": 3.5402, + "step": 7517 + }, + { + "epoch": 0.3956842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.0001342923976608187, + "loss": 3.7659, + "step": 7518 + }, + { + "epoch": 0.39573684210526316, + "grad_norm": 0.470703125, + "learning_rate": 0.00013428070175438596, + "loss": 3.6578, + "step": 7519 + }, + { + "epoch": 0.3957894736842105, + "grad_norm": 0.67578125, + "learning_rate": 0.00013426900584795324, + "loss": 3.2795, + "step": 7520 + }, + { + "epoch": 0.3958421052631579, + "grad_norm": 0.5703125, + "learning_rate": 0.00013425730994152049, + "loss": 3.5061, + "step": 7521 + }, + { + "epoch": 0.39589473684210524, + "grad_norm": 0.470703125, + "learning_rate": 0.00013424561403508774, + "loss": 3.6243, + "step": 7522 + }, + { + "epoch": 0.39594736842105266, + "grad_norm": 0.515625, + "learning_rate": 0.00013423391812865499, + "loss": 3.1447, + "step": 7523 + }, + { + "epoch": 0.396, + "grad_norm": 0.486328125, + "learning_rate": 0.00013422222222222224, + "loss": 4.0009, + "step": 7524 + }, + { + "epoch": 0.3960526315789474, + "grad_norm": 0.478515625, + "learning_rate": 0.00013421052631578948, + "loss": 3.2708, + "step": 7525 + }, + { + "epoch": 0.39610526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.00013419883040935673, + "loss": 3.942, + "step": 7526 + }, + { + "epoch": 0.3961578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.00013418713450292398, + "loss": 3.48, + "step": 7527 + }, + { + "epoch": 0.39621052631578946, + "grad_norm": 0.5625, + "learning_rate": 0.00013417543859649123, + "loss": 3.3038, + "step": 7528 + }, + { + "epoch": 0.3962631578947368, + "grad_norm": 0.484375, + "learning_rate": 0.00013416374269005848, + "loss": 3.718, + "step": 7529 + }, + { + "epoch": 0.39631578947368423, + "grad_norm": 0.48828125, + "learning_rate": 0.00013415204678362573, + "loss": 3.6129, + "step": 7530 + }, + { + "epoch": 0.3963684210526316, + "grad_norm": 0.48828125, + "learning_rate": 0.00013414035087719298, + "loss": 3.366, + "step": 7531 + }, + { + "epoch": 0.39642105263157895, + "grad_norm": 0.50390625, + "learning_rate": 0.00013412865497076023, + "loss": 3.6207, + "step": 7532 + }, + { + "epoch": 0.3964736842105263, + "grad_norm": 0.8828125, + "learning_rate": 0.00013411695906432748, + "loss": 3.4724, + "step": 7533 + }, + { + "epoch": 0.39652631578947367, + "grad_norm": 0.4921875, + "learning_rate": 0.00013410526315789473, + "loss": 3.5681, + "step": 7534 + }, + { + "epoch": 0.39657894736842103, + "grad_norm": 0.498046875, + "learning_rate": 0.000134093567251462, + "loss": 3.5565, + "step": 7535 + }, + { + "epoch": 0.39663157894736845, + "grad_norm": 0.515625, + "learning_rate": 0.00013408187134502926, + "loss": 3.4436, + "step": 7536 + }, + { + "epoch": 0.3966842105263158, + "grad_norm": 0.5390625, + "learning_rate": 0.0001340701754385965, + "loss": 3.9436, + "step": 7537 + }, + { + "epoch": 0.39673684210526317, + "grad_norm": 0.498046875, + "learning_rate": 0.00013405847953216376, + "loss": 3.5587, + "step": 7538 + }, + { + "epoch": 0.3967894736842105, + "grad_norm": 0.486328125, + "learning_rate": 0.000134046783625731, + "loss": 3.7646, + "step": 7539 + }, + { + "epoch": 0.3968421052631579, + "grad_norm": 0.4765625, + "learning_rate": 0.00013403508771929826, + "loss": 3.8403, + "step": 7540 + }, + { + "epoch": 0.39689473684210524, + "grad_norm": 0.51171875, + "learning_rate": 0.0001340233918128655, + "loss": 3.7575, + "step": 7541 + }, + { + "epoch": 0.3969473684210526, + "grad_norm": 0.515625, + "learning_rate": 0.00013401169590643275, + "loss": 3.3388, + "step": 7542 + }, + { + "epoch": 0.397, + "grad_norm": 0.51953125, + "learning_rate": 0.000134, + "loss": 3.8296, + "step": 7543 + }, + { + "epoch": 0.3970526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.00013398830409356725, + "loss": 3.955, + "step": 7544 + }, + { + "epoch": 0.39710526315789474, + "grad_norm": 0.54296875, + "learning_rate": 0.0001339766081871345, + "loss": 3.4783, + "step": 7545 + }, + { + "epoch": 0.3971578947368421, + "grad_norm": 0.65625, + "learning_rate": 0.00013396491228070175, + "loss": 3.9896, + "step": 7546 + }, + { + "epoch": 0.39721052631578946, + "grad_norm": 0.49609375, + "learning_rate": 0.000133953216374269, + "loss": 3.4218, + "step": 7547 + }, + { + "epoch": 0.3972631578947368, + "grad_norm": 0.4765625, + "learning_rate": 0.00013394152046783625, + "loss": 3.487, + "step": 7548 + }, + { + "epoch": 0.39731578947368423, + "grad_norm": 0.4609375, + "learning_rate": 0.0001339298245614035, + "loss": 3.7013, + "step": 7549 + }, + { + "epoch": 0.3973684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.00013391812865497078, + "loss": 3.5976, + "step": 7550 + }, + { + "epoch": 0.39742105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00013390643274853803, + "loss": 3.6029, + "step": 7551 + }, + { + "epoch": 0.3974736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.00013389473684210528, + "loss": 3.4635, + "step": 7552 + }, + { + "epoch": 0.39752631578947367, + "grad_norm": 0.55078125, + "learning_rate": 0.00013388304093567253, + "loss": 3.3801, + "step": 7553 + }, + { + "epoch": 0.39757894736842103, + "grad_norm": 0.53125, + "learning_rate": 0.00013387134502923978, + "loss": 3.3162, + "step": 7554 + }, + { + "epoch": 0.39763157894736845, + "grad_norm": 0.498046875, + "learning_rate": 0.00013385964912280703, + "loss": 3.6734, + "step": 7555 + }, + { + "epoch": 0.3976842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.00013384795321637428, + "loss": 3.6612, + "step": 7556 + }, + { + "epoch": 0.39773684210526317, + "grad_norm": 0.51953125, + "learning_rate": 0.00013383625730994153, + "loss": 3.5126, + "step": 7557 + }, + { + "epoch": 0.3977894736842105, + "grad_norm": 0.58984375, + "learning_rate": 0.00013382456140350878, + "loss": 3.6697, + "step": 7558 + }, + { + "epoch": 0.3978421052631579, + "grad_norm": 0.5859375, + "learning_rate": 0.00013381286549707603, + "loss": 3.9206, + "step": 7559 + }, + { + "epoch": 0.39789473684210525, + "grad_norm": 0.51171875, + "learning_rate": 0.00013380116959064327, + "loss": 3.9515, + "step": 7560 + }, + { + "epoch": 0.3979473684210526, + "grad_norm": 0.498046875, + "learning_rate": 0.00013378947368421052, + "loss": 3.4268, + "step": 7561 + }, + { + "epoch": 0.398, + "grad_norm": 0.5234375, + "learning_rate": 0.00013377777777777777, + "loss": 3.6014, + "step": 7562 + }, + { + "epoch": 0.3980526315789474, + "grad_norm": 0.51171875, + "learning_rate": 0.00013376608187134502, + "loss": 3.5507, + "step": 7563 + }, + { + "epoch": 0.39810526315789474, + "grad_norm": 0.64453125, + "learning_rate": 0.00013375438596491227, + "loss": 3.5096, + "step": 7564 + }, + { + "epoch": 0.3981578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00013374269005847955, + "loss": 3.6704, + "step": 7565 + }, + { + "epoch": 0.39821052631578946, + "grad_norm": 0.50390625, + "learning_rate": 0.0001337309941520468, + "loss": 3.3117, + "step": 7566 + }, + { + "epoch": 0.3982631578947368, + "grad_norm": 0.4765625, + "learning_rate": 0.00013371929824561405, + "loss": 3.6571, + "step": 7567 + }, + { + "epoch": 0.39831578947368423, + "grad_norm": 0.470703125, + "learning_rate": 0.0001337076023391813, + "loss": 3.9701, + "step": 7568 + }, + { + "epoch": 0.3983684210526316, + "grad_norm": 0.490234375, + "learning_rate": 0.00013369590643274855, + "loss": 3.8796, + "step": 7569 + }, + { + "epoch": 0.39842105263157895, + "grad_norm": 0.5390625, + "learning_rate": 0.0001336842105263158, + "loss": 4.0369, + "step": 7570 + }, + { + "epoch": 0.3984736842105263, + "grad_norm": 0.57421875, + "learning_rate": 0.00013367251461988305, + "loss": 4.0223, + "step": 7571 + }, + { + "epoch": 0.3985263157894737, + "grad_norm": 0.5625, + "learning_rate": 0.0001336608187134503, + "loss": 3.5693, + "step": 7572 + }, + { + "epoch": 0.39857894736842103, + "grad_norm": 0.53125, + "learning_rate": 0.00013364912280701755, + "loss": 3.5104, + "step": 7573 + }, + { + "epoch": 0.39863157894736845, + "grad_norm": 0.54296875, + "learning_rate": 0.0001336374269005848, + "loss": 3.6528, + "step": 7574 + }, + { + "epoch": 0.3986842105263158, + "grad_norm": 0.5390625, + "learning_rate": 0.00013362573099415205, + "loss": 3.4892, + "step": 7575 + }, + { + "epoch": 0.39873684210526317, + "grad_norm": 0.578125, + "learning_rate": 0.0001336140350877193, + "loss": 3.5738, + "step": 7576 + }, + { + "epoch": 0.3987894736842105, + "grad_norm": 0.55078125, + "learning_rate": 0.00013360233918128654, + "loss": 3.4186, + "step": 7577 + }, + { + "epoch": 0.3988421052631579, + "grad_norm": 0.46875, + "learning_rate": 0.0001335906432748538, + "loss": 3.7068, + "step": 7578 + }, + { + "epoch": 0.39889473684210525, + "grad_norm": 0.5078125, + "learning_rate": 0.00013357894736842104, + "loss": 3.7613, + "step": 7579 + }, + { + "epoch": 0.3989473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00013356725146198832, + "loss": 3.2527, + "step": 7580 + }, + { + "epoch": 0.399, + "grad_norm": 0.46875, + "learning_rate": 0.00013355555555555557, + "loss": 3.3894, + "step": 7581 + }, + { + "epoch": 0.3990526315789474, + "grad_norm": 0.490234375, + "learning_rate": 0.00013354385964912282, + "loss": 3.5221, + "step": 7582 + }, + { + "epoch": 0.39910526315789474, + "grad_norm": 0.48828125, + "learning_rate": 0.00013353216374269007, + "loss": 4.0698, + "step": 7583 + }, + { + "epoch": 0.3991578947368421, + "grad_norm": 0.4765625, + "learning_rate": 0.00013352046783625732, + "loss": 3.8639, + "step": 7584 + }, + { + "epoch": 0.39921052631578946, + "grad_norm": 0.48828125, + "learning_rate": 0.00013350877192982457, + "loss": 3.5222, + "step": 7585 + }, + { + "epoch": 0.3992631578947368, + "grad_norm": 0.5546875, + "learning_rate": 0.00013349707602339182, + "loss": 3.3388, + "step": 7586 + }, + { + "epoch": 0.39931578947368424, + "grad_norm": 0.55859375, + "learning_rate": 0.00013348538011695907, + "loss": 3.5475, + "step": 7587 + }, + { + "epoch": 0.3993684210526316, + "grad_norm": 0.48046875, + "learning_rate": 0.00013347368421052632, + "loss": 3.4079, + "step": 7588 + }, + { + "epoch": 0.39942105263157895, + "grad_norm": 0.484375, + "learning_rate": 0.00013346198830409357, + "loss": 3.7369, + "step": 7589 + }, + { + "epoch": 0.3994736842105263, + "grad_norm": 0.4765625, + "learning_rate": 0.00013345029239766082, + "loss": 3.5326, + "step": 7590 + }, + { + "epoch": 0.3995263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.00013343859649122807, + "loss": 3.6871, + "step": 7591 + }, + { + "epoch": 0.39957894736842103, + "grad_norm": 0.484375, + "learning_rate": 0.00013342690058479532, + "loss": 3.4629, + "step": 7592 + }, + { + "epoch": 0.39963157894736845, + "grad_norm": 0.51953125, + "learning_rate": 0.00013341520467836257, + "loss": 3.6919, + "step": 7593 + }, + { + "epoch": 0.3996842105263158, + "grad_norm": 0.484375, + "learning_rate": 0.00013340350877192984, + "loss": 3.4939, + "step": 7594 + }, + { + "epoch": 0.39973684210526317, + "grad_norm": 0.5078125, + "learning_rate": 0.0001333918128654971, + "loss": 2.8022, + "step": 7595 + }, + { + "epoch": 0.39978947368421053, + "grad_norm": 0.484375, + "learning_rate": 0.00013338011695906434, + "loss": 3.2436, + "step": 7596 + }, + { + "epoch": 0.3998421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.0001333684210526316, + "loss": 3.1698, + "step": 7597 + }, + { + "epoch": 0.39989473684210525, + "grad_norm": 0.490234375, + "learning_rate": 0.00013335672514619884, + "loss": 3.3589, + "step": 7598 + }, + { + "epoch": 0.3999473684210526, + "grad_norm": 0.498046875, + "learning_rate": 0.0001333450292397661, + "loss": 3.7561, + "step": 7599 + }, + { + "epoch": 0.4, + "grad_norm": 0.52734375, + "learning_rate": 0.00013333333333333334, + "loss": 3.8478, + "step": 7600 + }, + { + "epoch": 0.4000526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.0001333216374269006, + "loss": 3.7545, + "step": 7601 + }, + { + "epoch": 0.40010526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.00013330994152046784, + "loss": 3.8264, + "step": 7602 + }, + { + "epoch": 0.4001578947368421, + "grad_norm": 0.4765625, + "learning_rate": 0.0001332982456140351, + "loss": 3.531, + "step": 7603 + }, + { + "epoch": 0.40021052631578946, + "grad_norm": 0.51171875, + "learning_rate": 0.00013328654970760234, + "loss": 3.7169, + "step": 7604 + }, + { + "epoch": 0.4002631578947368, + "grad_norm": 0.6484375, + "learning_rate": 0.0001332748538011696, + "loss": 3.7893, + "step": 7605 + }, + { + "epoch": 0.40031578947368424, + "grad_norm": 0.60546875, + "learning_rate": 0.00013326315789473684, + "loss": 3.8584, + "step": 7606 + }, + { + "epoch": 0.4003684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.0001332514619883041, + "loss": 3.5561, + "step": 7607 + }, + { + "epoch": 0.40042105263157896, + "grad_norm": 0.5, + "learning_rate": 0.00013323976608187134, + "loss": 3.4791, + "step": 7608 + }, + { + "epoch": 0.4004736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.0001332280701754386, + "loss": 3.4986, + "step": 7609 + }, + { + "epoch": 0.4005263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00013321637426900586, + "loss": 4.0525, + "step": 7610 + }, + { + "epoch": 0.40057894736842103, + "grad_norm": 0.59375, + "learning_rate": 0.0001332046783625731, + "loss": 3.5096, + "step": 7611 + }, + { + "epoch": 0.4006315789473684, + "grad_norm": 0.55078125, + "learning_rate": 0.00013319298245614036, + "loss": 3.4836, + "step": 7612 + }, + { + "epoch": 0.4006842105263158, + "grad_norm": 0.4765625, + "learning_rate": 0.0001331812865497076, + "loss": 3.4078, + "step": 7613 + }, + { + "epoch": 0.40073684210526317, + "grad_norm": 0.4609375, + "learning_rate": 0.00013316959064327486, + "loss": 3.769, + "step": 7614 + }, + { + "epoch": 0.40078947368421053, + "grad_norm": 0.4609375, + "learning_rate": 0.0001331578947368421, + "loss": 3.4082, + "step": 7615 + }, + { + "epoch": 0.4008421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00013314619883040936, + "loss": 3.6124, + "step": 7616 + }, + { + "epoch": 0.40089473684210525, + "grad_norm": 0.5859375, + "learning_rate": 0.0001331345029239766, + "loss": 3.7105, + "step": 7617 + }, + { + "epoch": 0.4009473684210526, + "grad_norm": 0.490234375, + "learning_rate": 0.00013312280701754386, + "loss": 3.7113, + "step": 7618 + }, + { + "epoch": 0.401, + "grad_norm": 1.6953125, + "learning_rate": 0.0001331111111111111, + "loss": 3.4941, + "step": 7619 + }, + { + "epoch": 0.4010526315789474, + "grad_norm": 0.498046875, + "learning_rate": 0.00013309941520467836, + "loss": 3.8138, + "step": 7620 + }, + { + "epoch": 0.40110526315789474, + "grad_norm": 0.8046875, + "learning_rate": 0.0001330877192982456, + "loss": 3.1527, + "step": 7621 + }, + { + "epoch": 0.4011578947368421, + "grad_norm": 0.49609375, + "learning_rate": 0.00013307602339181286, + "loss": 3.8677, + "step": 7622 + }, + { + "epoch": 0.40121052631578946, + "grad_norm": 0.67578125, + "learning_rate": 0.0001330643274853801, + "loss": 3.1783, + "step": 7623 + }, + { + "epoch": 0.4012631578947368, + "grad_norm": 0.51171875, + "learning_rate": 0.00013305263157894738, + "loss": 3.7057, + "step": 7624 + }, + { + "epoch": 0.40131578947368424, + "grad_norm": 0.5390625, + "learning_rate": 0.00013304093567251463, + "loss": 3.6732, + "step": 7625 + }, + { + "epoch": 0.4013684210526316, + "grad_norm": 0.5, + "learning_rate": 0.00013302923976608188, + "loss": 3.5085, + "step": 7626 + }, + { + "epoch": 0.40142105263157896, + "grad_norm": 0.65234375, + "learning_rate": 0.00013301754385964913, + "loss": 3.4131, + "step": 7627 + }, + { + "epoch": 0.4014736842105263, + "grad_norm": 0.5, + "learning_rate": 0.00013300584795321638, + "loss": 3.396, + "step": 7628 + }, + { + "epoch": 0.4015263157894737, + "grad_norm": 0.478515625, + "learning_rate": 0.00013299415204678363, + "loss": 3.3842, + "step": 7629 + }, + { + "epoch": 0.40157894736842104, + "grad_norm": 0.478515625, + "learning_rate": 0.00013298245614035088, + "loss": 3.1884, + "step": 7630 + }, + { + "epoch": 0.4016315789473684, + "grad_norm": 0.46875, + "learning_rate": 0.00013297076023391816, + "loss": 3.4998, + "step": 7631 + }, + { + "epoch": 0.4016842105263158, + "grad_norm": 0.51953125, + "learning_rate": 0.00013295906432748538, + "loss": 3.8815, + "step": 7632 + }, + { + "epoch": 0.40173684210526317, + "grad_norm": 0.515625, + "learning_rate": 0.00013294736842105263, + "loss": 3.1177, + "step": 7633 + }, + { + "epoch": 0.40178947368421053, + "grad_norm": 0.58984375, + "learning_rate": 0.00013293567251461988, + "loss": 3.4147, + "step": 7634 + }, + { + "epoch": 0.4018421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00013292397660818713, + "loss": 3.6436, + "step": 7635 + }, + { + "epoch": 0.40189473684210525, + "grad_norm": 0.51953125, + "learning_rate": 0.00013291228070175438, + "loss": 3.5595, + "step": 7636 + }, + { + "epoch": 0.4019473684210526, + "grad_norm": 0.498046875, + "learning_rate": 0.00013290058479532163, + "loss": 3.5209, + "step": 7637 + }, + { + "epoch": 0.402, + "grad_norm": 0.55078125, + "learning_rate": 0.00013288888888888888, + "loss": 3.3409, + "step": 7638 + }, + { + "epoch": 0.4020526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.00013287719298245616, + "loss": 3.5869, + "step": 7639 + }, + { + "epoch": 0.40210526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.0001328654970760234, + "loss": 3.5364, + "step": 7640 + }, + { + "epoch": 0.4021578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.00013285380116959065, + "loss": 3.6001, + "step": 7641 + }, + { + "epoch": 0.40221052631578946, + "grad_norm": 0.48046875, + "learning_rate": 0.0001328421052631579, + "loss": 3.6091, + "step": 7642 + }, + { + "epoch": 0.4022631578947368, + "grad_norm": 0.490234375, + "learning_rate": 0.00013283040935672515, + "loss": 3.8212, + "step": 7643 + }, + { + "epoch": 0.40231578947368424, + "grad_norm": 0.478515625, + "learning_rate": 0.0001328187134502924, + "loss": 3.708, + "step": 7644 + }, + { + "epoch": 0.4023684210526316, + "grad_norm": 0.47265625, + "learning_rate": 0.00013280701754385965, + "loss": 3.999, + "step": 7645 + }, + { + "epoch": 0.40242105263157896, + "grad_norm": 0.55078125, + "learning_rate": 0.00013279532163742693, + "loss": 3.6315, + "step": 7646 + }, + { + "epoch": 0.4024736842105263, + "grad_norm": 0.51953125, + "learning_rate": 0.00013278362573099418, + "loss": 3.6195, + "step": 7647 + }, + { + "epoch": 0.4025263157894737, + "grad_norm": 0.5, + "learning_rate": 0.0001327719298245614, + "loss": 3.6646, + "step": 7648 + }, + { + "epoch": 0.40257894736842104, + "grad_norm": 0.49609375, + "learning_rate": 0.00013276023391812865, + "loss": 3.1367, + "step": 7649 + }, + { + "epoch": 0.4026315789473684, + "grad_norm": 0.46875, + "learning_rate": 0.0001327485380116959, + "loss": 3.6187, + "step": 7650 + }, + { + "epoch": 0.4026842105263158, + "grad_norm": 0.47265625, + "learning_rate": 0.00013273684210526315, + "loss": 3.9305, + "step": 7651 + }, + { + "epoch": 0.40273684210526317, + "grad_norm": 0.4765625, + "learning_rate": 0.0001327251461988304, + "loss": 3.4344, + "step": 7652 + }, + { + "epoch": 0.40278947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.00013271345029239765, + "loss": 3.9853, + "step": 7653 + }, + { + "epoch": 0.4028421052631579, + "grad_norm": 0.470703125, + "learning_rate": 0.00013270175438596493, + "loss": 3.4525, + "step": 7654 + }, + { + "epoch": 0.40289473684210525, + "grad_norm": 0.482421875, + "learning_rate": 0.00013269005847953218, + "loss": 3.8304, + "step": 7655 + }, + { + "epoch": 0.4029473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00013267836257309943, + "loss": 3.8963, + "step": 7656 + }, + { + "epoch": 0.403, + "grad_norm": 0.46484375, + "learning_rate": 0.00013266666666666667, + "loss": 3.861, + "step": 7657 + }, + { + "epoch": 0.4030526315789474, + "grad_norm": 0.4921875, + "learning_rate": 0.00013265497076023392, + "loss": 3.052, + "step": 7658 + }, + { + "epoch": 0.40310526315789474, + "grad_norm": 0.474609375, + "learning_rate": 0.00013264327485380117, + "loss": 3.7878, + "step": 7659 + }, + { + "epoch": 0.4031578947368421, + "grad_norm": 0.453125, + "learning_rate": 0.00013263157894736842, + "loss": 3.5806, + "step": 7660 + }, + { + "epoch": 0.40321052631578946, + "grad_norm": 0.470703125, + "learning_rate": 0.0001326198830409357, + "loss": 3.5716, + "step": 7661 + }, + { + "epoch": 0.4032631578947368, + "grad_norm": 0.474609375, + "learning_rate": 0.00013260818713450295, + "loss": 3.7455, + "step": 7662 + }, + { + "epoch": 0.4033157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00013259649122807017, + "loss": 3.9315, + "step": 7663 + }, + { + "epoch": 0.4033684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.00013258479532163742, + "loss": 3.263, + "step": 7664 + }, + { + "epoch": 0.40342105263157896, + "grad_norm": 0.5625, + "learning_rate": 0.00013257309941520467, + "loss": 3.1979, + "step": 7665 + }, + { + "epoch": 0.4034736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.00013256140350877192, + "loss": 3.558, + "step": 7666 + }, + { + "epoch": 0.4035263157894737, + "grad_norm": 0.47265625, + "learning_rate": 0.00013254970760233917, + "loss": 3.3447, + "step": 7667 + }, + { + "epoch": 0.40357894736842104, + "grad_norm": 0.50390625, + "learning_rate": 0.00013253801169590645, + "loss": 3.4371, + "step": 7668 + }, + { + "epoch": 0.4036315789473684, + "grad_norm": 0.4765625, + "learning_rate": 0.0001325263157894737, + "loss": 3.7243, + "step": 7669 + }, + { + "epoch": 0.4036842105263158, + "grad_norm": 0.48828125, + "learning_rate": 0.00013251461988304095, + "loss": 3.4971, + "step": 7670 + }, + { + "epoch": 0.40373684210526317, + "grad_norm": 0.484375, + "learning_rate": 0.0001325029239766082, + "loss": 3.6076, + "step": 7671 + }, + { + "epoch": 0.40378947368421053, + "grad_norm": 0.484375, + "learning_rate": 0.00013249122807017545, + "loss": 3.8641, + "step": 7672 + }, + { + "epoch": 0.4038421052631579, + "grad_norm": 0.4765625, + "learning_rate": 0.0001324795321637427, + "loss": 3.505, + "step": 7673 + }, + { + "epoch": 0.40389473684210525, + "grad_norm": 0.470703125, + "learning_rate": 0.00013246783625730995, + "loss": 3.586, + "step": 7674 + }, + { + "epoch": 0.4039473684210526, + "grad_norm": 0.5546875, + "learning_rate": 0.0001324561403508772, + "loss": 3.3443, + "step": 7675 + }, + { + "epoch": 0.404, + "grad_norm": 0.455078125, + "learning_rate": 0.00013244444444444447, + "loss": 3.4761, + "step": 7676 + }, + { + "epoch": 0.4040526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.00013243274853801172, + "loss": 3.9235, + "step": 7677 + }, + { + "epoch": 0.40410526315789475, + "grad_norm": 0.51953125, + "learning_rate": 0.00013242105263157894, + "loss": 3.4202, + "step": 7678 + }, + { + "epoch": 0.4041578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.0001324093567251462, + "loss": 3.4249, + "step": 7679 + }, + { + "epoch": 0.40421052631578946, + "grad_norm": 0.498046875, + "learning_rate": 0.00013239766081871344, + "loss": 3.5073, + "step": 7680 + }, + { + "epoch": 0.4042631578947368, + "grad_norm": 0.5234375, + "learning_rate": 0.0001323859649122807, + "loss": 3.0534, + "step": 7681 + }, + { + "epoch": 0.4043157894736842, + "grad_norm": 0.48828125, + "learning_rate": 0.00013237426900584794, + "loss": 3.435, + "step": 7682 + }, + { + "epoch": 0.4043684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00013236257309941522, + "loss": 3.5899, + "step": 7683 + }, + { + "epoch": 0.40442105263157896, + "grad_norm": 0.5, + "learning_rate": 0.00013235087719298247, + "loss": 3.6253, + "step": 7684 + }, + { + "epoch": 0.4044736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00013233918128654972, + "loss": 3.9226, + "step": 7685 + }, + { + "epoch": 0.4045263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.00013232748538011697, + "loss": 3.3838, + "step": 7686 + }, + { + "epoch": 0.40457894736842104, + "grad_norm": 0.48828125, + "learning_rate": 0.00013231578947368422, + "loss": 3.8192, + "step": 7687 + }, + { + "epoch": 0.4046315789473684, + "grad_norm": 0.4765625, + "learning_rate": 0.00013230409356725147, + "loss": 3.7136, + "step": 7688 + }, + { + "epoch": 0.4046842105263158, + "grad_norm": 0.5390625, + "learning_rate": 0.00013229239766081872, + "loss": 4.067, + "step": 7689 + }, + { + "epoch": 0.4047368421052632, + "grad_norm": 0.5234375, + "learning_rate": 0.00013228070175438597, + "loss": 3.5065, + "step": 7690 + }, + { + "epoch": 0.40478947368421053, + "grad_norm": 0.53515625, + "learning_rate": 0.00013226900584795324, + "loss": 3.5394, + "step": 7691 + }, + { + "epoch": 0.4048421052631579, + "grad_norm": 0.5, + "learning_rate": 0.0001322573099415205, + "loss": 3.5305, + "step": 7692 + }, + { + "epoch": 0.40489473684210525, + "grad_norm": 0.51953125, + "learning_rate": 0.00013224561403508774, + "loss": 3.9508, + "step": 7693 + }, + { + "epoch": 0.4049473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00013223391812865496, + "loss": 3.0705, + "step": 7694 + }, + { + "epoch": 0.405, + "grad_norm": 0.4921875, + "learning_rate": 0.00013222222222222221, + "loss": 3.6293, + "step": 7695 + }, + { + "epoch": 0.4050526315789474, + "grad_norm": 0.447265625, + "learning_rate": 0.00013221052631578946, + "loss": 3.6793, + "step": 7696 + }, + { + "epoch": 0.40510526315789475, + "grad_norm": 0.498046875, + "learning_rate": 0.0001321988304093567, + "loss": 3.4994, + "step": 7697 + }, + { + "epoch": 0.4051578947368421, + "grad_norm": 0.462890625, + "learning_rate": 0.000132187134502924, + "loss": 3.7183, + "step": 7698 + }, + { + "epoch": 0.40521052631578947, + "grad_norm": 0.5, + "learning_rate": 0.00013217543859649124, + "loss": 3.5305, + "step": 7699 + }, + { + "epoch": 0.4052631578947368, + "grad_norm": 0.515625, + "learning_rate": 0.0001321637426900585, + "loss": 3.8582, + "step": 7700 + }, + { + "epoch": 0.4053157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00013215204678362574, + "loss": 3.8156, + "step": 7701 + }, + { + "epoch": 0.4053684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.000132140350877193, + "loss": 3.1632, + "step": 7702 + }, + { + "epoch": 0.40542105263157896, + "grad_norm": 0.5234375, + "learning_rate": 0.00013212865497076024, + "loss": 3.5255, + "step": 7703 + }, + { + "epoch": 0.4054736842105263, + "grad_norm": 0.51953125, + "learning_rate": 0.0001321169590643275, + "loss": 3.4144, + "step": 7704 + }, + { + "epoch": 0.4055263157894737, + "grad_norm": 0.47265625, + "learning_rate": 0.00013210526315789476, + "loss": 4.1176, + "step": 7705 + }, + { + "epoch": 0.40557894736842104, + "grad_norm": 0.5390625, + "learning_rate": 0.000132093567251462, + "loss": 3.5381, + "step": 7706 + }, + { + "epoch": 0.4056315789473684, + "grad_norm": 0.54296875, + "learning_rate": 0.00013208187134502926, + "loss": 3.9229, + "step": 7707 + }, + { + "epoch": 0.4056842105263158, + "grad_norm": 0.48046875, + "learning_rate": 0.0001320701754385965, + "loss": 3.3336, + "step": 7708 + }, + { + "epoch": 0.4057368421052632, + "grad_norm": 0.482421875, + "learning_rate": 0.00013205847953216374, + "loss": 3.5673, + "step": 7709 + }, + { + "epoch": 0.40578947368421053, + "grad_norm": 0.462890625, + "learning_rate": 0.00013204678362573098, + "loss": 4.0079, + "step": 7710 + }, + { + "epoch": 0.4058421052631579, + "grad_norm": 0.53515625, + "learning_rate": 0.00013203508771929823, + "loss": 3.8493, + "step": 7711 + }, + { + "epoch": 0.40589473684210525, + "grad_norm": 0.51171875, + "learning_rate": 0.00013202339181286548, + "loss": 3.5319, + "step": 7712 + }, + { + "epoch": 0.4059473684210526, + "grad_norm": 0.482421875, + "learning_rate": 0.00013201169590643276, + "loss": 3.3906, + "step": 7713 + }, + { + "epoch": 0.406, + "grad_norm": 0.462890625, + "learning_rate": 0.000132, + "loss": 3.5689, + "step": 7714 + }, + { + "epoch": 0.4060526315789474, + "grad_norm": 0.95703125, + "learning_rate": 0.00013198830409356726, + "loss": 3.353, + "step": 7715 + }, + { + "epoch": 0.40610526315789475, + "grad_norm": 0.458984375, + "learning_rate": 0.0001319766081871345, + "loss": 4.0566, + "step": 7716 + }, + { + "epoch": 0.4061578947368421, + "grad_norm": 0.6328125, + "learning_rate": 0.00013196491228070176, + "loss": 3.4322, + "step": 7717 + }, + { + "epoch": 0.40621052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.000131953216374269, + "loss": 3.3124, + "step": 7718 + }, + { + "epoch": 0.4062631578947368, + "grad_norm": 0.52734375, + "learning_rate": 0.00013194152046783626, + "loss": 3.4941, + "step": 7719 + }, + { + "epoch": 0.4063157894736842, + "grad_norm": 0.498046875, + "learning_rate": 0.00013192982456140353, + "loss": 3.5138, + "step": 7720 + }, + { + "epoch": 0.4063684210526316, + "grad_norm": 0.52734375, + "learning_rate": 0.00013191812865497078, + "loss": 3.5088, + "step": 7721 + }, + { + "epoch": 0.40642105263157896, + "grad_norm": 0.5, + "learning_rate": 0.00013190643274853803, + "loss": 3.7781, + "step": 7722 + }, + { + "epoch": 0.4064736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00013189473684210528, + "loss": 3.5731, + "step": 7723 + }, + { + "epoch": 0.4065263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.0001318830409356725, + "loss": 3.6766, + "step": 7724 + }, + { + "epoch": 0.40657894736842104, + "grad_norm": 0.515625, + "learning_rate": 0.00013187134502923976, + "loss": 3.5396, + "step": 7725 + }, + { + "epoch": 0.4066315789473684, + "grad_norm": 0.478515625, + "learning_rate": 0.000131859649122807, + "loss": 3.6561, + "step": 7726 + }, + { + "epoch": 0.4066842105263158, + "grad_norm": 0.482421875, + "learning_rate": 0.00013184795321637425, + "loss": 3.0758, + "step": 7727 + }, + { + "epoch": 0.4067368421052632, + "grad_norm": 0.5078125, + "learning_rate": 0.00013183625730994153, + "loss": 3.6872, + "step": 7728 + }, + { + "epoch": 0.40678947368421053, + "grad_norm": 0.5625, + "learning_rate": 0.00013182456140350878, + "loss": 3.8071, + "step": 7729 + }, + { + "epoch": 0.4068421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00013181286549707603, + "loss": 3.506, + "step": 7730 + }, + { + "epoch": 0.40689473684210525, + "grad_norm": 0.54296875, + "learning_rate": 0.00013180116959064328, + "loss": 3.1519, + "step": 7731 + }, + { + "epoch": 0.4069473684210526, + "grad_norm": 0.5859375, + "learning_rate": 0.00013178947368421053, + "loss": 3.5305, + "step": 7732 + }, + { + "epoch": 0.407, + "grad_norm": 0.56640625, + "learning_rate": 0.00013177777777777778, + "loss": 3.4856, + "step": 7733 + }, + { + "epoch": 0.4070526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00013176608187134503, + "loss": 3.6127, + "step": 7734 + }, + { + "epoch": 0.40710526315789475, + "grad_norm": 0.49609375, + "learning_rate": 0.0001317543859649123, + "loss": 3.4815, + "step": 7735 + }, + { + "epoch": 0.4071578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.00013174269005847956, + "loss": 3.306, + "step": 7736 + }, + { + "epoch": 0.40721052631578947, + "grad_norm": 0.462890625, + "learning_rate": 0.0001317309941520468, + "loss": 2.9397, + "step": 7737 + }, + { + "epoch": 0.4072631578947368, + "grad_norm": 0.48828125, + "learning_rate": 0.00013171929824561405, + "loss": 3.593, + "step": 7738 + }, + { + "epoch": 0.4073157894736842, + "grad_norm": 0.458984375, + "learning_rate": 0.0001317076023391813, + "loss": 3.6694, + "step": 7739 + }, + { + "epoch": 0.4073684210526316, + "grad_norm": 0.447265625, + "learning_rate": 0.00013169590643274853, + "loss": 3.987, + "step": 7740 + }, + { + "epoch": 0.40742105263157896, + "grad_norm": 0.486328125, + "learning_rate": 0.00013168421052631578, + "loss": 3.4504, + "step": 7741 + }, + { + "epoch": 0.4074736842105263, + "grad_norm": 0.46484375, + "learning_rate": 0.00013167251461988303, + "loss": 3.8291, + "step": 7742 + }, + { + "epoch": 0.4075263157894737, + "grad_norm": 0.49609375, + "learning_rate": 0.0001316608187134503, + "loss": 3.5261, + "step": 7743 + }, + { + "epoch": 0.40757894736842104, + "grad_norm": 0.466796875, + "learning_rate": 0.00013164912280701755, + "loss": 3.5288, + "step": 7744 + }, + { + "epoch": 0.4076315789473684, + "grad_norm": 0.484375, + "learning_rate": 0.0001316374269005848, + "loss": 3.7708, + "step": 7745 + }, + { + "epoch": 0.4076842105263158, + "grad_norm": 0.46875, + "learning_rate": 0.00013162573099415205, + "loss": 3.7439, + "step": 7746 + }, + { + "epoch": 0.4077368421052632, + "grad_norm": 0.51171875, + "learning_rate": 0.0001316140350877193, + "loss": 3.3603, + "step": 7747 + }, + { + "epoch": 0.40778947368421053, + "grad_norm": 0.498046875, + "learning_rate": 0.00013160233918128655, + "loss": 3.5471, + "step": 7748 + }, + { + "epoch": 0.4078421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.0001315906432748538, + "loss": 3.7908, + "step": 7749 + }, + { + "epoch": 0.40789473684210525, + "grad_norm": 0.51171875, + "learning_rate": 0.00013157894736842108, + "loss": 3.7216, + "step": 7750 + }, + { + "epoch": 0.4079473684210526, + "grad_norm": 0.546875, + "learning_rate": 0.00013156725146198833, + "loss": 3.646, + "step": 7751 + }, + { + "epoch": 0.408, + "grad_norm": 0.478515625, + "learning_rate": 0.00013155555555555558, + "loss": 3.8853, + "step": 7752 + }, + { + "epoch": 0.4080526315789474, + "grad_norm": 0.578125, + "learning_rate": 0.00013154385964912283, + "loss": 3.5297, + "step": 7753 + }, + { + "epoch": 0.40810526315789475, + "grad_norm": 0.484375, + "learning_rate": 0.00013153216374269008, + "loss": 3.3546, + "step": 7754 + }, + { + "epoch": 0.4081578947368421, + "grad_norm": 0.498046875, + "learning_rate": 0.0001315204678362573, + "loss": 3.38, + "step": 7755 + }, + { + "epoch": 0.40821052631578947, + "grad_norm": 0.47265625, + "learning_rate": 0.00013150877192982455, + "loss": 3.378, + "step": 7756 + }, + { + "epoch": 0.40826315789473683, + "grad_norm": 0.484375, + "learning_rate": 0.00013149707602339182, + "loss": 3.6675, + "step": 7757 + }, + { + "epoch": 0.4083157894736842, + "grad_norm": 0.59765625, + "learning_rate": 0.00013148538011695907, + "loss": 3.4599, + "step": 7758 + }, + { + "epoch": 0.4083684210526316, + "grad_norm": 0.44921875, + "learning_rate": 0.00013147368421052632, + "loss": 3.7111, + "step": 7759 + }, + { + "epoch": 0.40842105263157896, + "grad_norm": 0.53125, + "learning_rate": 0.00013146198830409357, + "loss": 3.4484, + "step": 7760 + }, + { + "epoch": 0.4084736842105263, + "grad_norm": 0.484375, + "learning_rate": 0.00013145029239766082, + "loss": 3.6304, + "step": 7761 + }, + { + "epoch": 0.4085263157894737, + "grad_norm": 0.625, + "learning_rate": 0.00013143859649122807, + "loss": 3.519, + "step": 7762 + }, + { + "epoch": 0.40857894736842104, + "grad_norm": 0.51953125, + "learning_rate": 0.00013142690058479532, + "loss": 3.3144, + "step": 7763 + }, + { + "epoch": 0.4086315789473684, + "grad_norm": 0.66796875, + "learning_rate": 0.00013141520467836257, + "loss": 3.43, + "step": 7764 + }, + { + "epoch": 0.4086842105263158, + "grad_norm": 0.482421875, + "learning_rate": 0.00013140350877192985, + "loss": 3.6242, + "step": 7765 + }, + { + "epoch": 0.4087368421052632, + "grad_norm": 0.50390625, + "learning_rate": 0.0001313918128654971, + "loss": 3.4772, + "step": 7766 + }, + { + "epoch": 0.40878947368421054, + "grad_norm": 0.5078125, + "learning_rate": 0.00013138011695906435, + "loss": 3.8228, + "step": 7767 + }, + { + "epoch": 0.4088421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.0001313684210526316, + "loss": 3.04, + "step": 7768 + }, + { + "epoch": 0.40889473684210526, + "grad_norm": 0.462890625, + "learning_rate": 0.00013135672514619885, + "loss": 3.6952, + "step": 7769 + }, + { + "epoch": 0.4089473684210526, + "grad_norm": 0.482421875, + "learning_rate": 0.00013134502923976607, + "loss": 3.6554, + "step": 7770 + }, + { + "epoch": 0.409, + "grad_norm": 0.5234375, + "learning_rate": 0.00013133333333333332, + "loss": 3.4717, + "step": 7771 + }, + { + "epoch": 0.4090526315789474, + "grad_norm": 0.4609375, + "learning_rate": 0.0001313216374269006, + "loss": 3.6951, + "step": 7772 + }, + { + "epoch": 0.40910526315789475, + "grad_norm": 0.47265625, + "learning_rate": 0.00013130994152046784, + "loss": 3.7831, + "step": 7773 + }, + { + "epoch": 0.4091578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.0001312982456140351, + "loss": 3.5001, + "step": 7774 + }, + { + "epoch": 0.40921052631578947, + "grad_norm": 0.5, + "learning_rate": 0.00013128654970760234, + "loss": 3.5911, + "step": 7775 + }, + { + "epoch": 0.40926315789473683, + "grad_norm": 0.52734375, + "learning_rate": 0.0001312748538011696, + "loss": 2.9438, + "step": 7776 + }, + { + "epoch": 0.4093157894736842, + "grad_norm": 0.54296875, + "learning_rate": 0.00013126315789473684, + "loss": 3.3065, + "step": 7777 + }, + { + "epoch": 0.4093684210526316, + "grad_norm": 0.515625, + "learning_rate": 0.0001312514619883041, + "loss": 3.8651, + "step": 7778 + }, + { + "epoch": 0.40942105263157896, + "grad_norm": 0.56640625, + "learning_rate": 0.00013123976608187134, + "loss": 3.4032, + "step": 7779 + }, + { + "epoch": 0.4094736842105263, + "grad_norm": 0.50390625, + "learning_rate": 0.00013122807017543862, + "loss": 3.6858, + "step": 7780 + }, + { + "epoch": 0.4095263157894737, + "grad_norm": 0.53125, + "learning_rate": 0.00013121637426900587, + "loss": 3.8845, + "step": 7781 + }, + { + "epoch": 0.40957894736842104, + "grad_norm": 0.52734375, + "learning_rate": 0.00013120467836257312, + "loss": 3.1399, + "step": 7782 + }, + { + "epoch": 0.4096315789473684, + "grad_norm": 0.52734375, + "learning_rate": 0.00013119298245614037, + "loss": 2.9523, + "step": 7783 + }, + { + "epoch": 0.40968421052631576, + "grad_norm": 0.478515625, + "learning_rate": 0.00013118128654970762, + "loss": 3.5569, + "step": 7784 + }, + { + "epoch": 0.4097368421052632, + "grad_norm": 0.54296875, + "learning_rate": 0.00013116959064327487, + "loss": 3.7368, + "step": 7785 + }, + { + "epoch": 0.40978947368421054, + "grad_norm": 0.52734375, + "learning_rate": 0.0001311578947368421, + "loss": 3.5021, + "step": 7786 + }, + { + "epoch": 0.4098421052631579, + "grad_norm": 0.486328125, + "learning_rate": 0.00013114619883040937, + "loss": 3.8083, + "step": 7787 + }, + { + "epoch": 0.40989473684210526, + "grad_norm": 0.4765625, + "learning_rate": 0.00013113450292397662, + "loss": 3.2093, + "step": 7788 + }, + { + "epoch": 0.4099473684210526, + "grad_norm": 0.48046875, + "learning_rate": 0.00013112280701754387, + "loss": 3.3405, + "step": 7789 + }, + { + "epoch": 0.41, + "grad_norm": 0.5, + "learning_rate": 0.00013111111111111111, + "loss": 3.6834, + "step": 7790 + }, + { + "epoch": 0.4100526315789474, + "grad_norm": 0.498046875, + "learning_rate": 0.00013109941520467836, + "loss": 3.2142, + "step": 7791 + }, + { + "epoch": 0.41010526315789475, + "grad_norm": 0.50390625, + "learning_rate": 0.00013108771929824561, + "loss": 3.4979, + "step": 7792 + }, + { + "epoch": 0.4101578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00013107602339181286, + "loss": 3.7664, + "step": 7793 + }, + { + "epoch": 0.41021052631578947, + "grad_norm": 0.5234375, + "learning_rate": 0.00013106432748538014, + "loss": 3.7468, + "step": 7794 + }, + { + "epoch": 0.41026315789473683, + "grad_norm": 0.462890625, + "learning_rate": 0.0001310526315789474, + "loss": 3.4668, + "step": 7795 + }, + { + "epoch": 0.4103157894736842, + "grad_norm": 0.498046875, + "learning_rate": 0.00013104093567251464, + "loss": 3.6522, + "step": 7796 + }, + { + "epoch": 0.4103684210526316, + "grad_norm": 0.482421875, + "learning_rate": 0.0001310292397660819, + "loss": 3.6159, + "step": 7797 + }, + { + "epoch": 0.41042105263157896, + "grad_norm": 0.5703125, + "learning_rate": 0.00013101754385964914, + "loss": 3.1356, + "step": 7798 + }, + { + "epoch": 0.4104736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.0001310058479532164, + "loss": 3.6485, + "step": 7799 + }, + { + "epoch": 0.4105263157894737, + "grad_norm": 0.455078125, + "learning_rate": 0.00013099415204678364, + "loss": 3.587, + "step": 7800 + }, + { + "epoch": 0.41057894736842104, + "grad_norm": 0.55859375, + "learning_rate": 0.00013098245614035086, + "loss": 3.3034, + "step": 7801 + }, + { + "epoch": 0.4106315789473684, + "grad_norm": 0.478515625, + "learning_rate": 0.00013097076023391814, + "loss": 3.4288, + "step": 7802 + }, + { + "epoch": 0.41068421052631576, + "grad_norm": 0.431640625, + "learning_rate": 0.0001309590643274854, + "loss": 3.5855, + "step": 7803 + }, + { + "epoch": 0.4107368421052632, + "grad_norm": 0.46875, + "learning_rate": 0.00013094736842105264, + "loss": 3.4313, + "step": 7804 + }, + { + "epoch": 0.41078947368421054, + "grad_norm": 0.5078125, + "learning_rate": 0.00013093567251461989, + "loss": 3.6496, + "step": 7805 + }, + { + "epoch": 0.4108421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00013092397660818714, + "loss": 3.7009, + "step": 7806 + }, + { + "epoch": 0.41089473684210526, + "grad_norm": 0.53515625, + "learning_rate": 0.00013091228070175438, + "loss": 3.4904, + "step": 7807 + }, + { + "epoch": 0.4109473684210526, + "grad_norm": 0.4765625, + "learning_rate": 0.00013090058479532163, + "loss": 3.7481, + "step": 7808 + }, + { + "epoch": 0.411, + "grad_norm": 0.4921875, + "learning_rate": 0.0001308888888888889, + "loss": 4.0532, + "step": 7809 + }, + { + "epoch": 0.4110526315789474, + "grad_norm": 0.47265625, + "learning_rate": 0.00013087719298245616, + "loss": 3.377, + "step": 7810 + }, + { + "epoch": 0.41110526315789475, + "grad_norm": 0.482421875, + "learning_rate": 0.0001308654970760234, + "loss": 3.8553, + "step": 7811 + }, + { + "epoch": 0.4111578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.00013085380116959066, + "loss": 3.304, + "step": 7812 + }, + { + "epoch": 0.41121052631578947, + "grad_norm": 0.458984375, + "learning_rate": 0.0001308421052631579, + "loss": 3.5027, + "step": 7813 + }, + { + "epoch": 0.41126315789473683, + "grad_norm": 0.5078125, + "learning_rate": 0.00013083040935672516, + "loss": 3.9057, + "step": 7814 + }, + { + "epoch": 0.4113157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.0001308187134502924, + "loss": 3.4174, + "step": 7815 + }, + { + "epoch": 0.4113684210526316, + "grad_norm": 0.5703125, + "learning_rate": 0.00013080701754385963, + "loss": 3.4196, + "step": 7816 + }, + { + "epoch": 0.41142105263157897, + "grad_norm": 0.49609375, + "learning_rate": 0.0001307953216374269, + "loss": 3.3745, + "step": 7817 + }, + { + "epoch": 0.4114736842105263, + "grad_norm": 0.56640625, + "learning_rate": 0.00013078362573099416, + "loss": 3.3821, + "step": 7818 + }, + { + "epoch": 0.4115263157894737, + "grad_norm": 0.47265625, + "learning_rate": 0.0001307719298245614, + "loss": 3.4647, + "step": 7819 + }, + { + "epoch": 0.41157894736842104, + "grad_norm": 0.5234375, + "learning_rate": 0.00013076023391812866, + "loss": 3.4711, + "step": 7820 + }, + { + "epoch": 0.4116315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.0001307485380116959, + "loss": 3.8704, + "step": 7821 + }, + { + "epoch": 0.41168421052631576, + "grad_norm": 0.48046875, + "learning_rate": 0.00013073684210526316, + "loss": 3.227, + "step": 7822 + }, + { + "epoch": 0.4117368421052632, + "grad_norm": 0.51171875, + "learning_rate": 0.0001307251461988304, + "loss": 3.6867, + "step": 7823 + }, + { + "epoch": 0.41178947368421054, + "grad_norm": 0.482421875, + "learning_rate": 0.00013071345029239768, + "loss": 3.3877, + "step": 7824 + }, + { + "epoch": 0.4118421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.00013070175438596493, + "loss": 3.6335, + "step": 7825 + }, + { + "epoch": 0.41189473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.00013069005847953218, + "loss": 3.4789, + "step": 7826 + }, + { + "epoch": 0.4119473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00013067836257309943, + "loss": 3.6592, + "step": 7827 + }, + { + "epoch": 0.412, + "grad_norm": 0.5078125, + "learning_rate": 0.00013066666666666668, + "loss": 3.389, + "step": 7828 + }, + { + "epoch": 0.4120526315789474, + "grad_norm": 0.515625, + "learning_rate": 0.00013065497076023393, + "loss": 3.8248, + "step": 7829 + }, + { + "epoch": 0.41210526315789475, + "grad_norm": 0.498046875, + "learning_rate": 0.00013064327485380118, + "loss": 3.3436, + "step": 7830 + }, + { + "epoch": 0.4121578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.00013063157894736843, + "loss": 3.9385, + "step": 7831 + }, + { + "epoch": 0.41221052631578947, + "grad_norm": 0.48046875, + "learning_rate": 0.00013061988304093568, + "loss": 3.3089, + "step": 7832 + }, + { + "epoch": 0.41226315789473683, + "grad_norm": 0.46484375, + "learning_rate": 0.00013060818713450293, + "loss": 3.5984, + "step": 7833 + }, + { + "epoch": 0.4123157894736842, + "grad_norm": 0.470703125, + "learning_rate": 0.00013059649122807018, + "loss": 3.5567, + "step": 7834 + }, + { + "epoch": 0.4123684210526316, + "grad_norm": 0.462890625, + "learning_rate": 0.00013058479532163743, + "loss": 3.7638, + "step": 7835 + }, + { + "epoch": 0.41242105263157897, + "grad_norm": 0.470703125, + "learning_rate": 0.00013057309941520468, + "loss": 3.4145, + "step": 7836 + }, + { + "epoch": 0.4124736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00013056140350877193, + "loss": 3.6025, + "step": 7837 + }, + { + "epoch": 0.4125263157894737, + "grad_norm": 0.470703125, + "learning_rate": 0.00013054970760233918, + "loss": 3.6034, + "step": 7838 + }, + { + "epoch": 0.41257894736842105, + "grad_norm": 0.58984375, + "learning_rate": 0.00013053801169590645, + "loss": 3.1217, + "step": 7839 + }, + { + "epoch": 0.4126315789473684, + "grad_norm": 0.484375, + "learning_rate": 0.0001305263157894737, + "loss": 3.7061, + "step": 7840 + }, + { + "epoch": 0.41268421052631576, + "grad_norm": 0.47265625, + "learning_rate": 0.00013051461988304095, + "loss": 3.5206, + "step": 7841 + }, + { + "epoch": 0.4127368421052632, + "grad_norm": 0.56640625, + "learning_rate": 0.0001305029239766082, + "loss": 3.7045, + "step": 7842 + }, + { + "epoch": 0.41278947368421054, + "grad_norm": 0.5, + "learning_rate": 0.00013049122807017545, + "loss": 4.1886, + "step": 7843 + }, + { + "epoch": 0.4128421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.0001304795321637427, + "loss": 3.9397, + "step": 7844 + }, + { + "epoch": 0.41289473684210526, + "grad_norm": 0.53515625, + "learning_rate": 0.00013046783625730995, + "loss": 3.7167, + "step": 7845 + }, + { + "epoch": 0.4129473684210526, + "grad_norm": 0.51953125, + "learning_rate": 0.0001304561403508772, + "loss": 3.7693, + "step": 7846 + }, + { + "epoch": 0.413, + "grad_norm": 0.515625, + "learning_rate": 0.00013044444444444445, + "loss": 3.6835, + "step": 7847 + }, + { + "epoch": 0.4130526315789474, + "grad_norm": 0.51953125, + "learning_rate": 0.0001304327485380117, + "loss": 3.322, + "step": 7848 + }, + { + "epoch": 0.41310526315789475, + "grad_norm": 0.46875, + "learning_rate": 0.00013042105263157895, + "loss": 4.007, + "step": 7849 + }, + { + "epoch": 0.4131578947368421, + "grad_norm": 0.5234375, + "learning_rate": 0.0001304093567251462, + "loss": 3.6556, + "step": 7850 + }, + { + "epoch": 0.4132105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00013039766081871345, + "loss": 3.5373, + "step": 7851 + }, + { + "epoch": 0.41326315789473683, + "grad_norm": 0.46484375, + "learning_rate": 0.0001303859649122807, + "loss": 3.3896, + "step": 7852 + }, + { + "epoch": 0.4133157894736842, + "grad_norm": 0.46875, + "learning_rate": 0.00013037426900584795, + "loss": 3.3203, + "step": 7853 + }, + { + "epoch": 0.41336842105263155, + "grad_norm": 0.46484375, + "learning_rate": 0.00013036257309941522, + "loss": 3.8092, + "step": 7854 + }, + { + "epoch": 0.41342105263157897, + "grad_norm": 0.474609375, + "learning_rate": 0.00013035087719298247, + "loss": 3.4701, + "step": 7855 + }, + { + "epoch": 0.4134736842105263, + "grad_norm": 0.53515625, + "learning_rate": 0.00013033918128654972, + "loss": 3.3808, + "step": 7856 + }, + { + "epoch": 0.4135263157894737, + "grad_norm": 0.490234375, + "learning_rate": 0.00013032748538011697, + "loss": 3.542, + "step": 7857 + }, + { + "epoch": 0.41357894736842105, + "grad_norm": 0.470703125, + "learning_rate": 0.00013031578947368422, + "loss": 3.7463, + "step": 7858 + }, + { + "epoch": 0.4136315789473684, + "grad_norm": 0.462890625, + "learning_rate": 0.00013030409356725147, + "loss": 3.616, + "step": 7859 + }, + { + "epoch": 0.41368421052631577, + "grad_norm": 0.53515625, + "learning_rate": 0.00013029239766081872, + "loss": 3.4473, + "step": 7860 + }, + { + "epoch": 0.4137368421052632, + "grad_norm": 0.486328125, + "learning_rate": 0.00013028070175438597, + "loss": 3.3635, + "step": 7861 + }, + { + "epoch": 0.41378947368421054, + "grad_norm": 0.4765625, + "learning_rate": 0.00013026900584795322, + "loss": 3.4695, + "step": 7862 + }, + { + "epoch": 0.4138421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00013025730994152047, + "loss": 3.8305, + "step": 7863 + }, + { + "epoch": 0.41389473684210526, + "grad_norm": 0.5078125, + "learning_rate": 0.00013024561403508772, + "loss": 3.5193, + "step": 7864 + }, + { + "epoch": 0.4139473684210526, + "grad_norm": 0.5546875, + "learning_rate": 0.00013023391812865497, + "loss": 3.2991, + "step": 7865 + }, + { + "epoch": 0.414, + "grad_norm": 0.51171875, + "learning_rate": 0.00013022222222222222, + "loss": 3.5461, + "step": 7866 + }, + { + "epoch": 0.4140526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.00013021052631578947, + "loss": 3.4592, + "step": 7867 + }, + { + "epoch": 0.41410526315789475, + "grad_norm": 0.58203125, + "learning_rate": 0.00013019883040935675, + "loss": 3.385, + "step": 7868 + }, + { + "epoch": 0.4141578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.000130187134502924, + "loss": 3.3007, + "step": 7869 + }, + { + "epoch": 0.4142105263157895, + "grad_norm": 0.5625, + "learning_rate": 0.00013017543859649124, + "loss": 3.4829, + "step": 7870 + }, + { + "epoch": 0.41426315789473683, + "grad_norm": 0.515625, + "learning_rate": 0.0001301637426900585, + "loss": 3.1301, + "step": 7871 + }, + { + "epoch": 0.4143157894736842, + "grad_norm": 0.453125, + "learning_rate": 0.00013015204678362574, + "loss": 3.6715, + "step": 7872 + }, + { + "epoch": 0.41436842105263155, + "grad_norm": 0.5078125, + "learning_rate": 0.000130140350877193, + "loss": 3.2966, + "step": 7873 + }, + { + "epoch": 0.41442105263157897, + "grad_norm": 0.435546875, + "learning_rate": 0.00013012865497076024, + "loss": 3.3905, + "step": 7874 + }, + { + "epoch": 0.4144736842105263, + "grad_norm": 0.55859375, + "learning_rate": 0.0001301169590643275, + "loss": 3.5426, + "step": 7875 + }, + { + "epoch": 0.4145263157894737, + "grad_norm": 0.4765625, + "learning_rate": 0.00013010526315789474, + "loss": 3.6892, + "step": 7876 + }, + { + "epoch": 0.41457894736842105, + "grad_norm": 0.52734375, + "learning_rate": 0.000130093567251462, + "loss": 3.4154, + "step": 7877 + }, + { + "epoch": 0.4146315789473684, + "grad_norm": 0.55859375, + "learning_rate": 0.00013008187134502924, + "loss": 3.0547, + "step": 7878 + }, + { + "epoch": 0.41468421052631577, + "grad_norm": 0.4765625, + "learning_rate": 0.0001300701754385965, + "loss": 3.5208, + "step": 7879 + }, + { + "epoch": 0.4147368421052632, + "grad_norm": 0.51953125, + "learning_rate": 0.00013005847953216374, + "loss": 3.3587, + "step": 7880 + }, + { + "epoch": 0.41478947368421054, + "grad_norm": 0.53515625, + "learning_rate": 0.000130046783625731, + "loss": 3.6593, + "step": 7881 + }, + { + "epoch": 0.4148421052631579, + "grad_norm": 0.6015625, + "learning_rate": 0.00013003508771929824, + "loss": 3.1343, + "step": 7882 + }, + { + "epoch": 0.41489473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00013002339181286552, + "loss": 4.0073, + "step": 7883 + }, + { + "epoch": 0.4149473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00013001169590643277, + "loss": 3.8051, + "step": 7884 + }, + { + "epoch": 0.415, + "grad_norm": 0.5, + "learning_rate": 0.00013000000000000002, + "loss": 3.5174, + "step": 7885 + }, + { + "epoch": 0.4150526315789474, + "grad_norm": 0.47265625, + "learning_rate": 0.00012998830409356727, + "loss": 3.3585, + "step": 7886 + }, + { + "epoch": 0.41510526315789475, + "grad_norm": 0.51953125, + "learning_rate": 0.00012997660818713451, + "loss": 3.4752, + "step": 7887 + }, + { + "epoch": 0.4151578947368421, + "grad_norm": 0.490234375, + "learning_rate": 0.00012996491228070176, + "loss": 3.7595, + "step": 7888 + }, + { + "epoch": 0.4152105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00012995321637426901, + "loss": 3.5725, + "step": 7889 + }, + { + "epoch": 0.41526315789473683, + "grad_norm": 0.59765625, + "learning_rate": 0.00012994152046783626, + "loss": 3.7296, + "step": 7890 + }, + { + "epoch": 0.4153157894736842, + "grad_norm": 0.5234375, + "learning_rate": 0.0001299298245614035, + "loss": 3.7951, + "step": 7891 + }, + { + "epoch": 0.41536842105263155, + "grad_norm": 0.62890625, + "learning_rate": 0.00012991812865497076, + "loss": 3.4002, + "step": 7892 + }, + { + "epoch": 0.41542105263157897, + "grad_norm": 0.6015625, + "learning_rate": 0.000129906432748538, + "loss": 3.7836, + "step": 7893 + }, + { + "epoch": 0.41547368421052633, + "grad_norm": 0.46875, + "learning_rate": 0.00012989473684210526, + "loss": 3.6954, + "step": 7894 + }, + { + "epoch": 0.4155263157894737, + "grad_norm": 0.498046875, + "learning_rate": 0.0001298830409356725, + "loss": 3.3372, + "step": 7895 + }, + { + "epoch": 0.41557894736842105, + "grad_norm": 0.48046875, + "learning_rate": 0.00012987134502923976, + "loss": 2.8461, + "step": 7896 + }, + { + "epoch": 0.4156315789473684, + "grad_norm": 0.6875, + "learning_rate": 0.000129859649122807, + "loss": 3.8489, + "step": 7897 + }, + { + "epoch": 0.41568421052631577, + "grad_norm": 0.6328125, + "learning_rate": 0.0001298479532163743, + "loss": 3.7235, + "step": 7898 + }, + { + "epoch": 0.4157368421052632, + "grad_norm": 0.55859375, + "learning_rate": 0.00012983625730994154, + "loss": 3.2589, + "step": 7899 + }, + { + "epoch": 0.41578947368421054, + "grad_norm": 0.4765625, + "learning_rate": 0.0001298245614035088, + "loss": 3.0812, + "step": 7900 + }, + { + "epoch": 0.4158421052631579, + "grad_norm": 0.451171875, + "learning_rate": 0.00012981286549707604, + "loss": 3.5516, + "step": 7901 + }, + { + "epoch": 0.41589473684210526, + "grad_norm": 0.48828125, + "learning_rate": 0.00012980116959064329, + "loss": 3.7741, + "step": 7902 + }, + { + "epoch": 0.4159473684210526, + "grad_norm": 0.5078125, + "learning_rate": 0.00012978947368421054, + "loss": 3.885, + "step": 7903 + }, + { + "epoch": 0.416, + "grad_norm": 0.515625, + "learning_rate": 0.00012977777777777779, + "loss": 3.3669, + "step": 7904 + }, + { + "epoch": 0.4160526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00012976608187134503, + "loss": 3.5368, + "step": 7905 + }, + { + "epoch": 0.41610526315789476, + "grad_norm": 0.53515625, + "learning_rate": 0.00012975438596491228, + "loss": 3.5807, + "step": 7906 + }, + { + "epoch": 0.4161578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00012974269005847953, + "loss": 3.5197, + "step": 7907 + }, + { + "epoch": 0.4162105263157895, + "grad_norm": 0.48828125, + "learning_rate": 0.00012973099415204678, + "loss": 3.7675, + "step": 7908 + }, + { + "epoch": 0.41626315789473683, + "grad_norm": 0.53515625, + "learning_rate": 0.00012971929824561403, + "loss": 3.33, + "step": 7909 + }, + { + "epoch": 0.4163157894736842, + "grad_norm": 0.6484375, + "learning_rate": 0.00012970760233918128, + "loss": 3.2068, + "step": 7910 + }, + { + "epoch": 0.41636842105263155, + "grad_norm": 0.474609375, + "learning_rate": 0.00012969590643274853, + "loss": 3.5788, + "step": 7911 + }, + { + "epoch": 0.41642105263157897, + "grad_norm": 0.484375, + "learning_rate": 0.00012968421052631578, + "loss": 3.7637, + "step": 7912 + }, + { + "epoch": 0.41647368421052633, + "grad_norm": 0.51953125, + "learning_rate": 0.00012967251461988306, + "loss": 3.8371, + "step": 7913 + }, + { + "epoch": 0.4165263157894737, + "grad_norm": 0.58984375, + "learning_rate": 0.0001296608187134503, + "loss": 2.9577, + "step": 7914 + }, + { + "epoch": 0.41657894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.00012964912280701756, + "loss": 3.6066, + "step": 7915 + }, + { + "epoch": 0.4166315789473684, + "grad_norm": 0.48046875, + "learning_rate": 0.0001296374269005848, + "loss": 3.6539, + "step": 7916 + }, + { + "epoch": 0.41668421052631577, + "grad_norm": 0.5546875, + "learning_rate": 0.00012962573099415206, + "loss": 3.9004, + "step": 7917 + }, + { + "epoch": 0.4167368421052632, + "grad_norm": 0.625, + "learning_rate": 0.0001296140350877193, + "loss": 3.7925, + "step": 7918 + }, + { + "epoch": 0.41678947368421054, + "grad_norm": 0.51953125, + "learning_rate": 0.00012960233918128656, + "loss": 3.5854, + "step": 7919 + }, + { + "epoch": 0.4168421052631579, + "grad_norm": 0.46484375, + "learning_rate": 0.0001295906432748538, + "loss": 3.6241, + "step": 7920 + }, + { + "epoch": 0.41689473684210526, + "grad_norm": 0.5390625, + "learning_rate": 0.00012957894736842106, + "loss": 3.6275, + "step": 7921 + }, + { + "epoch": 0.4169473684210526, + "grad_norm": 0.5078125, + "learning_rate": 0.0001295672514619883, + "loss": 3.4403, + "step": 7922 + }, + { + "epoch": 0.417, + "grad_norm": 0.482421875, + "learning_rate": 0.00012955555555555555, + "loss": 3.6717, + "step": 7923 + }, + { + "epoch": 0.41705263157894734, + "grad_norm": 0.474609375, + "learning_rate": 0.0001295438596491228, + "loss": 3.6799, + "step": 7924 + }, + { + "epoch": 0.41710526315789476, + "grad_norm": 0.53515625, + "learning_rate": 0.00012953216374269005, + "loss": 3.0744, + "step": 7925 + }, + { + "epoch": 0.4171578947368421, + "grad_norm": 0.76953125, + "learning_rate": 0.0001295204678362573, + "loss": 3.0599, + "step": 7926 + }, + { + "epoch": 0.4172105263157895, + "grad_norm": 0.458984375, + "learning_rate": 0.00012950877192982455, + "loss": 3.7915, + "step": 7927 + }, + { + "epoch": 0.41726315789473684, + "grad_norm": 0.478515625, + "learning_rate": 0.00012949707602339183, + "loss": 3.5923, + "step": 7928 + }, + { + "epoch": 0.4173157894736842, + "grad_norm": 0.47265625, + "learning_rate": 0.00012948538011695908, + "loss": 3.9153, + "step": 7929 + }, + { + "epoch": 0.41736842105263156, + "grad_norm": 0.50390625, + "learning_rate": 0.00012947368421052633, + "loss": 3.2854, + "step": 7930 + }, + { + "epoch": 0.41742105263157897, + "grad_norm": 0.482421875, + "learning_rate": 0.00012946198830409358, + "loss": 3.4077, + "step": 7931 + }, + { + "epoch": 0.41747368421052633, + "grad_norm": 0.515625, + "learning_rate": 0.00012945029239766083, + "loss": 3.5201, + "step": 7932 + }, + { + "epoch": 0.4175263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00012943859649122808, + "loss": 3.4513, + "step": 7933 + }, + { + "epoch": 0.41757894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.00012942690058479533, + "loss": 3.5242, + "step": 7934 + }, + { + "epoch": 0.4176315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00012941520467836258, + "loss": 3.5547, + "step": 7935 + }, + { + "epoch": 0.41768421052631577, + "grad_norm": 0.48828125, + "learning_rate": 0.00012940350877192983, + "loss": 4.0108, + "step": 7936 + }, + { + "epoch": 0.4177368421052632, + "grad_norm": 0.53515625, + "learning_rate": 0.00012939181286549708, + "loss": 3.3339, + "step": 7937 + }, + { + "epoch": 0.41778947368421054, + "grad_norm": 0.7578125, + "learning_rate": 0.00012938011695906433, + "loss": 2.9706, + "step": 7938 + }, + { + "epoch": 0.4178421052631579, + "grad_norm": 0.462890625, + "learning_rate": 0.00012936842105263158, + "loss": 3.8741, + "step": 7939 + }, + { + "epoch": 0.41789473684210526, + "grad_norm": 0.49609375, + "learning_rate": 0.00012935672514619882, + "loss": 3.6864, + "step": 7940 + }, + { + "epoch": 0.4179473684210526, + "grad_norm": 0.5625, + "learning_rate": 0.00012934502923976607, + "loss": 3.6246, + "step": 7941 + }, + { + "epoch": 0.418, + "grad_norm": 0.451171875, + "learning_rate": 0.00012933333333333332, + "loss": 4.033, + "step": 7942 + }, + { + "epoch": 0.41805263157894734, + "grad_norm": 0.4921875, + "learning_rate": 0.0001293216374269006, + "loss": 3.6107, + "step": 7943 + }, + { + "epoch": 0.41810526315789476, + "grad_norm": 0.447265625, + "learning_rate": 0.00012930994152046785, + "loss": 3.4736, + "step": 7944 + }, + { + "epoch": 0.4181578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.0001292982456140351, + "loss": 3.6945, + "step": 7945 + }, + { + "epoch": 0.4182105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.00012928654970760235, + "loss": 3.8183, + "step": 7946 + }, + { + "epoch": 0.41826315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.0001292748538011696, + "loss": 3.2393, + "step": 7947 + }, + { + "epoch": 0.4183157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00012926315789473685, + "loss": 3.478, + "step": 7948 + }, + { + "epoch": 0.41836842105263156, + "grad_norm": 0.4921875, + "learning_rate": 0.0001292514619883041, + "loss": 3.3637, + "step": 7949 + }, + { + "epoch": 0.41842105263157897, + "grad_norm": 0.43359375, + "learning_rate": 0.00012923976608187135, + "loss": 3.7547, + "step": 7950 + }, + { + "epoch": 0.41847368421052633, + "grad_norm": 0.5625, + "learning_rate": 0.0001292280701754386, + "loss": 3.7623, + "step": 7951 + }, + { + "epoch": 0.4185263157894737, + "grad_norm": 0.4453125, + "learning_rate": 0.00012921637426900585, + "loss": 3.637, + "step": 7952 + }, + { + "epoch": 0.41857894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.0001292046783625731, + "loss": 3.408, + "step": 7953 + }, + { + "epoch": 0.4186315789473684, + "grad_norm": 0.453125, + "learning_rate": 0.00012919298245614035, + "loss": 3.4834, + "step": 7954 + }, + { + "epoch": 0.41868421052631577, + "grad_norm": 0.5078125, + "learning_rate": 0.0001291812865497076, + "loss": 3.7546, + "step": 7955 + }, + { + "epoch": 0.4187368421052632, + "grad_norm": 0.6328125, + "learning_rate": 0.00012916959064327485, + "loss": 3.3056, + "step": 7956 + }, + { + "epoch": 0.41878947368421054, + "grad_norm": 0.53515625, + "learning_rate": 0.00012915789473684212, + "loss": 3.7115, + "step": 7957 + }, + { + "epoch": 0.4188421052631579, + "grad_norm": 0.484375, + "learning_rate": 0.00012914619883040937, + "loss": 3.4537, + "step": 7958 + }, + { + "epoch": 0.41889473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00012913450292397662, + "loss": 3.8545, + "step": 7959 + }, + { + "epoch": 0.4189473684210526, + "grad_norm": 0.4921875, + "learning_rate": 0.00012912280701754387, + "loss": 3.5357, + "step": 7960 + }, + { + "epoch": 0.419, + "grad_norm": 0.56640625, + "learning_rate": 0.00012911111111111112, + "loss": 3.2371, + "step": 7961 + }, + { + "epoch": 0.41905263157894734, + "grad_norm": 0.5078125, + "learning_rate": 0.00012909941520467837, + "loss": 3.4031, + "step": 7962 + }, + { + "epoch": 0.41910526315789476, + "grad_norm": 0.47265625, + "learning_rate": 0.00012908771929824562, + "loss": 3.3025, + "step": 7963 + }, + { + "epoch": 0.4191578947368421, + "grad_norm": 0.57421875, + "learning_rate": 0.00012907602339181287, + "loss": 3.27, + "step": 7964 + }, + { + "epoch": 0.4192105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00012906432748538012, + "loss": 2.823, + "step": 7965 + }, + { + "epoch": 0.41926315789473684, + "grad_norm": 0.458984375, + "learning_rate": 0.00012905263157894737, + "loss": 3.4223, + "step": 7966 + }, + { + "epoch": 0.4193157894736842, + "grad_norm": 0.458984375, + "learning_rate": 0.00012904093567251462, + "loss": 3.4691, + "step": 7967 + }, + { + "epoch": 0.41936842105263156, + "grad_norm": 0.458984375, + "learning_rate": 0.00012902923976608187, + "loss": 3.3172, + "step": 7968 + }, + { + "epoch": 0.41942105263157897, + "grad_norm": 0.490234375, + "learning_rate": 0.00012901754385964912, + "loss": 3.3577, + "step": 7969 + }, + { + "epoch": 0.41947368421052633, + "grad_norm": 0.51171875, + "learning_rate": 0.00012900584795321637, + "loss": 3.5739, + "step": 7970 + }, + { + "epoch": 0.4195263157894737, + "grad_norm": 0.5390625, + "learning_rate": 0.00012899415204678362, + "loss": 3.3345, + "step": 7971 + }, + { + "epoch": 0.41957894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.0001289824561403509, + "loss": 3.4618, + "step": 7972 + }, + { + "epoch": 0.4196315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.00012897076023391814, + "loss": 3.8828, + "step": 7973 + }, + { + "epoch": 0.41968421052631577, + "grad_norm": 0.6328125, + "learning_rate": 0.0001289590643274854, + "loss": 4.0023, + "step": 7974 + }, + { + "epoch": 0.41973684210526313, + "grad_norm": 0.50390625, + "learning_rate": 0.00012894736842105264, + "loss": 3.501, + "step": 7975 + }, + { + "epoch": 0.41978947368421055, + "grad_norm": 0.50390625, + "learning_rate": 0.0001289356725146199, + "loss": 3.5448, + "step": 7976 + }, + { + "epoch": 0.4198421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.00012892397660818714, + "loss": 3.6368, + "step": 7977 + }, + { + "epoch": 0.41989473684210527, + "grad_norm": 0.51953125, + "learning_rate": 0.0001289122807017544, + "loss": 3.5067, + "step": 7978 + }, + { + "epoch": 0.4199473684210526, + "grad_norm": 0.52734375, + "learning_rate": 0.00012890058479532164, + "loss": 4.0337, + "step": 7979 + }, + { + "epoch": 0.42, + "grad_norm": 0.53515625, + "learning_rate": 0.00012888888888888892, + "loss": 3.5947, + "step": 7980 + }, + { + "epoch": 0.42005263157894734, + "grad_norm": 0.69921875, + "learning_rate": 0.00012887719298245614, + "loss": 3.7101, + "step": 7981 + }, + { + "epoch": 0.42010526315789476, + "grad_norm": 0.5703125, + "learning_rate": 0.0001288654970760234, + "loss": 3.1218, + "step": 7982 + }, + { + "epoch": 0.4201578947368421, + "grad_norm": 0.76171875, + "learning_rate": 0.00012885380116959064, + "loss": 3.5859, + "step": 7983 + }, + { + "epoch": 0.4202105263157895, + "grad_norm": 0.578125, + "learning_rate": 0.0001288421052631579, + "loss": 3.7733, + "step": 7984 + }, + { + "epoch": 0.42026315789473684, + "grad_norm": 0.484375, + "learning_rate": 0.00012883040935672514, + "loss": 3.6856, + "step": 7985 + }, + { + "epoch": 0.4203157894736842, + "grad_norm": 0.478515625, + "learning_rate": 0.0001288187134502924, + "loss": 3.5908, + "step": 7986 + }, + { + "epoch": 0.42036842105263156, + "grad_norm": 0.46484375, + "learning_rate": 0.00012880701754385966, + "loss": 3.7366, + "step": 7987 + }, + { + "epoch": 0.420421052631579, + "grad_norm": 0.47265625, + "learning_rate": 0.0001287953216374269, + "loss": 3.3422, + "step": 7988 + }, + { + "epoch": 0.42047368421052633, + "grad_norm": 0.5078125, + "learning_rate": 0.00012878362573099416, + "loss": 3.6318, + "step": 7989 + }, + { + "epoch": 0.4205263157894737, + "grad_norm": 0.5, + "learning_rate": 0.0001287719298245614, + "loss": 3.7826, + "step": 7990 + }, + { + "epoch": 0.42057894736842105, + "grad_norm": 0.5703125, + "learning_rate": 0.00012876023391812866, + "loss": 3.5067, + "step": 7991 + }, + { + "epoch": 0.4206315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.0001287485380116959, + "loss": 3.4163, + "step": 7992 + }, + { + "epoch": 0.42068421052631577, + "grad_norm": 0.412109375, + "learning_rate": 0.00012873684210526316, + "loss": 3.0316, + "step": 7993 + }, + { + "epoch": 0.42073684210526313, + "grad_norm": 0.494140625, + "learning_rate": 0.00012872514619883044, + "loss": 3.7401, + "step": 7994 + }, + { + "epoch": 0.42078947368421055, + "grad_norm": 0.48828125, + "learning_rate": 0.0001287134502923977, + "loss": 3.554, + "step": 7995 + }, + { + "epoch": 0.4208421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.0001287017543859649, + "loss": 3.6852, + "step": 7996 + }, + { + "epoch": 0.42089473684210527, + "grad_norm": 0.859375, + "learning_rate": 0.00012869005847953216, + "loss": 3.5124, + "step": 7997 + }, + { + "epoch": 0.4209473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.0001286783625730994, + "loss": 3.3729, + "step": 7998 + }, + { + "epoch": 0.421, + "grad_norm": 0.490234375, + "learning_rate": 0.00012866666666666666, + "loss": 3.2883, + "step": 7999 + }, + { + "epoch": 0.42105263157894735, + "grad_norm": 0.50390625, + "learning_rate": 0.0001286549707602339, + "loss": 3.7136, + "step": 8000 + }, + { + "epoch": 0.42110526315789476, + "grad_norm": 0.55078125, + "learning_rate": 0.00012864327485380116, + "loss": 3.8302, + "step": 8001 + }, + { + "epoch": 0.4211578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.00012863157894736843, + "loss": 3.6854, + "step": 8002 + }, + { + "epoch": 0.4212105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.00012861988304093568, + "loss": 3.9164, + "step": 8003 + }, + { + "epoch": 0.42126315789473684, + "grad_norm": 0.62109375, + "learning_rate": 0.00012860818713450293, + "loss": 3.1547, + "step": 8004 + }, + { + "epoch": 0.4213157894736842, + "grad_norm": 0.482421875, + "learning_rate": 0.00012859649122807018, + "loss": 3.3741, + "step": 8005 + }, + { + "epoch": 0.42136842105263156, + "grad_norm": 0.48046875, + "learning_rate": 0.00012858479532163743, + "loss": 3.5481, + "step": 8006 + }, + { + "epoch": 0.421421052631579, + "grad_norm": 0.46875, + "learning_rate": 0.00012857309941520468, + "loss": 3.6707, + "step": 8007 + }, + { + "epoch": 0.42147368421052633, + "grad_norm": 0.69921875, + "learning_rate": 0.00012856140350877193, + "loss": 3.5718, + "step": 8008 + }, + { + "epoch": 0.4215263157894737, + "grad_norm": 0.59765625, + "learning_rate": 0.0001285497076023392, + "loss": 3.2837, + "step": 8009 + }, + { + "epoch": 0.42157894736842105, + "grad_norm": 0.53125, + "learning_rate": 0.00012853801169590646, + "loss": 3.7426, + "step": 8010 + }, + { + "epoch": 0.4216315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.00012852631578947368, + "loss": 3.0256, + "step": 8011 + }, + { + "epoch": 0.4216842105263158, + "grad_norm": 0.494140625, + "learning_rate": 0.00012851461988304093, + "loss": 3.8046, + "step": 8012 + }, + { + "epoch": 0.42173684210526313, + "grad_norm": 0.5, + "learning_rate": 0.00012850292397660818, + "loss": 3.773, + "step": 8013 + }, + { + "epoch": 0.42178947368421055, + "grad_norm": 0.5, + "learning_rate": 0.00012849122807017543, + "loss": 3.9965, + "step": 8014 + }, + { + "epoch": 0.4218421052631579, + "grad_norm": 0.65625, + "learning_rate": 0.00012847953216374268, + "loss": 4.0177, + "step": 8015 + }, + { + "epoch": 0.42189473684210527, + "grad_norm": 0.6015625, + "learning_rate": 0.00012846783625730993, + "loss": 3.6685, + "step": 8016 + }, + { + "epoch": 0.4219473684210526, + "grad_norm": 0.5625, + "learning_rate": 0.0001284561403508772, + "loss": 3.6206, + "step": 8017 + }, + { + "epoch": 0.422, + "grad_norm": 0.5234375, + "learning_rate": 0.00012844444444444446, + "loss": 3.875, + "step": 8018 + }, + { + "epoch": 0.42205263157894735, + "grad_norm": 0.68359375, + "learning_rate": 0.0001284327485380117, + "loss": 3.8894, + "step": 8019 + }, + { + "epoch": 0.42210526315789476, + "grad_norm": 0.498046875, + "learning_rate": 0.00012842105263157895, + "loss": 3.4569, + "step": 8020 + }, + { + "epoch": 0.4221578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.0001284093567251462, + "loss": 3.4687, + "step": 8021 + }, + { + "epoch": 0.4222105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.00012839766081871345, + "loss": 3.6501, + "step": 8022 + }, + { + "epoch": 0.42226315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.0001283859649122807, + "loss": 3.68, + "step": 8023 + }, + { + "epoch": 0.4223157894736842, + "grad_norm": 0.494140625, + "learning_rate": 0.00012837426900584798, + "loss": 3.4031, + "step": 8024 + }, + { + "epoch": 0.42236842105263156, + "grad_norm": 0.5703125, + "learning_rate": 0.00012836257309941523, + "loss": 3.4253, + "step": 8025 + }, + { + "epoch": 0.422421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00012835087719298248, + "loss": 3.2768, + "step": 8026 + }, + { + "epoch": 0.42247368421052633, + "grad_norm": 0.48828125, + "learning_rate": 0.0001283391812865497, + "loss": 3.3207, + "step": 8027 + }, + { + "epoch": 0.4225263157894737, + "grad_norm": 0.466796875, + "learning_rate": 0.00012832748538011695, + "loss": 3.8076, + "step": 8028 + }, + { + "epoch": 0.42257894736842105, + "grad_norm": 0.48046875, + "learning_rate": 0.0001283157894736842, + "loss": 3.4275, + "step": 8029 + }, + { + "epoch": 0.4226315789473684, + "grad_norm": 0.51953125, + "learning_rate": 0.00012830409356725145, + "loss": 3.256, + "step": 8030 + }, + { + "epoch": 0.4226842105263158, + "grad_norm": 0.46875, + "learning_rate": 0.00012829239766081873, + "loss": 3.4353, + "step": 8031 + }, + { + "epoch": 0.42273684210526313, + "grad_norm": 0.46875, + "learning_rate": 0.00012828070175438598, + "loss": 3.437, + "step": 8032 + }, + { + "epoch": 0.42278947368421055, + "grad_norm": 0.451171875, + "learning_rate": 0.00012826900584795323, + "loss": 3.0577, + "step": 8033 + }, + { + "epoch": 0.4228421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.00012825730994152048, + "loss": 3.2776, + "step": 8034 + }, + { + "epoch": 0.42289473684210527, + "grad_norm": 0.51953125, + "learning_rate": 0.00012824561403508773, + "loss": 3.5017, + "step": 8035 + }, + { + "epoch": 0.4229473684210526, + "grad_norm": 0.4765625, + "learning_rate": 0.00012823391812865498, + "loss": 3.4389, + "step": 8036 + }, + { + "epoch": 0.423, + "grad_norm": 0.5, + "learning_rate": 0.00012822222222222222, + "loss": 3.0833, + "step": 8037 + }, + { + "epoch": 0.42305263157894735, + "grad_norm": 0.478515625, + "learning_rate": 0.00012821052631578947, + "loss": 3.228, + "step": 8038 + }, + { + "epoch": 0.42310526315789476, + "grad_norm": 0.478515625, + "learning_rate": 0.00012819883040935675, + "loss": 3.3896, + "step": 8039 + }, + { + "epoch": 0.4231578947368421, + "grad_norm": 0.5234375, + "learning_rate": 0.000128187134502924, + "loss": 3.7764, + "step": 8040 + }, + { + "epoch": 0.4232105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00012817543859649125, + "loss": 3.4694, + "step": 8041 + }, + { + "epoch": 0.42326315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.00012816374269005847, + "loss": 3.6267, + "step": 8042 + }, + { + "epoch": 0.4233157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00012815204678362572, + "loss": 3.5464, + "step": 8043 + }, + { + "epoch": 0.42336842105263156, + "grad_norm": 0.53515625, + "learning_rate": 0.00012814035087719297, + "loss": 3.3982, + "step": 8044 + }, + { + "epoch": 0.4234210526315789, + "grad_norm": 0.484375, + "learning_rate": 0.00012812865497076022, + "loss": 3.4087, + "step": 8045 + }, + { + "epoch": 0.42347368421052634, + "grad_norm": 0.46484375, + "learning_rate": 0.0001281169590643275, + "loss": 3.7054, + "step": 8046 + }, + { + "epoch": 0.4235263157894737, + "grad_norm": 0.546875, + "learning_rate": 0.00012810526315789475, + "loss": 3.1526, + "step": 8047 + }, + { + "epoch": 0.42357894736842105, + "grad_norm": 0.4921875, + "learning_rate": 0.000128093567251462, + "loss": 3.3749, + "step": 8048 + }, + { + "epoch": 0.4236315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00012808187134502925, + "loss": 3.6327, + "step": 8049 + }, + { + "epoch": 0.4236842105263158, + "grad_norm": 0.55078125, + "learning_rate": 0.0001280701754385965, + "loss": 3.4914, + "step": 8050 + }, + { + "epoch": 0.42373684210526313, + "grad_norm": 0.53125, + "learning_rate": 0.00012805847953216375, + "loss": 3.3373, + "step": 8051 + }, + { + "epoch": 0.42378947368421055, + "grad_norm": 0.515625, + "learning_rate": 0.000128046783625731, + "loss": 3.1188, + "step": 8052 + }, + { + "epoch": 0.4238421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00012803508771929825, + "loss": 3.205, + "step": 8053 + }, + { + "epoch": 0.42389473684210527, + "grad_norm": 0.48046875, + "learning_rate": 0.00012802339181286552, + "loss": 3.6956, + "step": 8054 + }, + { + "epoch": 0.42394736842105263, + "grad_norm": 0.494140625, + "learning_rate": 0.00012801169590643277, + "loss": 3.256, + "step": 8055 + }, + { + "epoch": 0.424, + "grad_norm": 0.462890625, + "learning_rate": 0.00012800000000000002, + "loss": 3.2, + "step": 8056 + }, + { + "epoch": 0.42405263157894735, + "grad_norm": 0.5546875, + "learning_rate": 0.00012798830409356724, + "loss": 3.3999, + "step": 8057 + }, + { + "epoch": 0.42410526315789476, + "grad_norm": 0.6171875, + "learning_rate": 0.0001279766081871345, + "loss": 2.941, + "step": 8058 + }, + { + "epoch": 0.4241578947368421, + "grad_norm": 0.427734375, + "learning_rate": 0.00012796491228070174, + "loss": 3.4039, + "step": 8059 + }, + { + "epoch": 0.4242105263157895, + "grad_norm": 0.77734375, + "learning_rate": 0.000127953216374269, + "loss": 3.5364, + "step": 8060 + }, + { + "epoch": 0.42426315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00012794152046783627, + "loss": 3.3749, + "step": 8061 + }, + { + "epoch": 0.4243157894736842, + "grad_norm": 0.49609375, + "learning_rate": 0.00012792982456140352, + "loss": 3.4273, + "step": 8062 + }, + { + "epoch": 0.42436842105263156, + "grad_norm": 0.5078125, + "learning_rate": 0.00012791812865497077, + "loss": 3.6217, + "step": 8063 + }, + { + "epoch": 0.4244210526315789, + "grad_norm": 0.50390625, + "learning_rate": 0.00012790643274853802, + "loss": 4.0393, + "step": 8064 + }, + { + "epoch": 0.42447368421052634, + "grad_norm": 0.498046875, + "learning_rate": 0.00012789473684210527, + "loss": 3.6192, + "step": 8065 + }, + { + "epoch": 0.4245263157894737, + "grad_norm": 0.5234375, + "learning_rate": 0.00012788304093567252, + "loss": 3.5832, + "step": 8066 + }, + { + "epoch": 0.42457894736842106, + "grad_norm": 0.50390625, + "learning_rate": 0.00012787134502923977, + "loss": 3.8057, + "step": 8067 + }, + { + "epoch": 0.4246315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.00012785964912280702, + "loss": 3.5076, + "step": 8068 + }, + { + "epoch": 0.4246842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.0001278479532163743, + "loss": 3.6181, + "step": 8069 + }, + { + "epoch": 0.42473684210526313, + "grad_norm": 0.72265625, + "learning_rate": 0.00012783625730994154, + "loss": 4.0029, + "step": 8070 + }, + { + "epoch": 0.42478947368421055, + "grad_norm": 0.49609375, + "learning_rate": 0.0001278245614035088, + "loss": 3.5593, + "step": 8071 + }, + { + "epoch": 0.4248421052631579, + "grad_norm": 0.466796875, + "learning_rate": 0.00012781286549707604, + "loss": 3.3344, + "step": 8072 + }, + { + "epoch": 0.42489473684210527, + "grad_norm": 0.470703125, + "learning_rate": 0.00012780116959064326, + "loss": 3.2412, + "step": 8073 + }, + { + "epoch": 0.42494736842105263, + "grad_norm": 0.48828125, + "learning_rate": 0.00012778947368421051, + "loss": 3.7571, + "step": 8074 + }, + { + "epoch": 0.425, + "grad_norm": 0.5, + "learning_rate": 0.00012777777777777776, + "loss": 4.0023, + "step": 8075 + }, + { + "epoch": 0.42505263157894735, + "grad_norm": 0.5234375, + "learning_rate": 0.00012776608187134504, + "loss": 3.5172, + "step": 8076 + }, + { + "epoch": 0.42510526315789476, + "grad_norm": 0.5078125, + "learning_rate": 0.0001277543859649123, + "loss": 3.8262, + "step": 8077 + }, + { + "epoch": 0.4251578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.00012774269005847954, + "loss": 3.2831, + "step": 8078 + }, + { + "epoch": 0.4252105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.0001277309941520468, + "loss": 3.7009, + "step": 8079 + }, + { + "epoch": 0.42526315789473684, + "grad_norm": 0.51953125, + "learning_rate": 0.00012771929824561404, + "loss": 3.093, + "step": 8080 + }, + { + "epoch": 0.4253157894736842, + "grad_norm": 0.6171875, + "learning_rate": 0.0001277076023391813, + "loss": 3.2706, + "step": 8081 + }, + { + "epoch": 0.42536842105263156, + "grad_norm": 0.48828125, + "learning_rate": 0.00012769590643274854, + "loss": 3.6531, + "step": 8082 + }, + { + "epoch": 0.4254210526315789, + "grad_norm": 0.47265625, + "learning_rate": 0.00012768421052631581, + "loss": 3.7085, + "step": 8083 + }, + { + "epoch": 0.42547368421052634, + "grad_norm": 0.486328125, + "learning_rate": 0.00012767251461988306, + "loss": 3.6277, + "step": 8084 + }, + { + "epoch": 0.4255263157894737, + "grad_norm": 0.53125, + "learning_rate": 0.00012766081871345031, + "loss": 3.4956, + "step": 8085 + }, + { + "epoch": 0.42557894736842106, + "grad_norm": 0.486328125, + "learning_rate": 0.00012764912280701756, + "loss": 3.6683, + "step": 8086 + }, + { + "epoch": 0.4256315789473684, + "grad_norm": 0.49609375, + "learning_rate": 0.0001276374269005848, + "loss": 3.9231, + "step": 8087 + }, + { + "epoch": 0.4256842105263158, + "grad_norm": 0.53125, + "learning_rate": 0.00012762573099415204, + "loss": 3.327, + "step": 8088 + }, + { + "epoch": 0.42573684210526314, + "grad_norm": 0.515625, + "learning_rate": 0.00012761403508771928, + "loss": 3.628, + "step": 8089 + }, + { + "epoch": 0.42578947368421055, + "grad_norm": 0.92578125, + "learning_rate": 0.00012760233918128653, + "loss": 3.3422, + "step": 8090 + }, + { + "epoch": 0.4258421052631579, + "grad_norm": 0.494140625, + "learning_rate": 0.0001275906432748538, + "loss": 3.2154, + "step": 8091 + }, + { + "epoch": 0.42589473684210527, + "grad_norm": 0.46875, + "learning_rate": 0.00012757894736842106, + "loss": 3.4103, + "step": 8092 + }, + { + "epoch": 0.42594736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.0001275672514619883, + "loss": 3.9616, + "step": 8093 + }, + { + "epoch": 0.426, + "grad_norm": 0.5, + "learning_rate": 0.00012755555555555556, + "loss": 3.119, + "step": 8094 + }, + { + "epoch": 0.42605263157894735, + "grad_norm": 0.53125, + "learning_rate": 0.0001275438596491228, + "loss": 3.1999, + "step": 8095 + }, + { + "epoch": 0.4261052631578947, + "grad_norm": 0.484375, + "learning_rate": 0.00012753216374269006, + "loss": 3.3526, + "step": 8096 + }, + { + "epoch": 0.4261578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.0001275204678362573, + "loss": 3.0779, + "step": 8097 + }, + { + "epoch": 0.4262105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00012750877192982459, + "loss": 3.4134, + "step": 8098 + }, + { + "epoch": 0.42626315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.00012749707602339184, + "loss": 3.616, + "step": 8099 + }, + { + "epoch": 0.4263157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00012748538011695908, + "loss": 3.8257, + "step": 8100 + }, + { + "epoch": 0.42636842105263156, + "grad_norm": 0.52734375, + "learning_rate": 0.00012747368421052633, + "loss": 3.3112, + "step": 8101 + }, + { + "epoch": 0.4264210526315789, + "grad_norm": 0.478515625, + "learning_rate": 0.00012746198830409358, + "loss": 3.4904, + "step": 8102 + }, + { + "epoch": 0.42647368421052634, + "grad_norm": 0.4921875, + "learning_rate": 0.0001274502923976608, + "loss": 3.4118, + "step": 8103 + }, + { + "epoch": 0.4265263157894737, + "grad_norm": 0.69140625, + "learning_rate": 0.00012743859649122806, + "loss": 3.5731, + "step": 8104 + }, + { + "epoch": 0.42657894736842106, + "grad_norm": 0.52734375, + "learning_rate": 0.0001274269005847953, + "loss": 3.7815, + "step": 8105 + }, + { + "epoch": 0.4266315789473684, + "grad_norm": 0.5390625, + "learning_rate": 0.00012741520467836258, + "loss": 3.1864, + "step": 8106 + }, + { + "epoch": 0.4266842105263158, + "grad_norm": 0.48828125, + "learning_rate": 0.00012740350877192983, + "loss": 3.3985, + "step": 8107 + }, + { + "epoch": 0.42673684210526314, + "grad_norm": 0.5, + "learning_rate": 0.00012739181286549708, + "loss": 3.7255, + "step": 8108 + }, + { + "epoch": 0.42678947368421055, + "grad_norm": 1.3203125, + "learning_rate": 0.00012738011695906433, + "loss": 3.6406, + "step": 8109 + }, + { + "epoch": 0.4268421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00012736842105263158, + "loss": 3.4852, + "step": 8110 + }, + { + "epoch": 0.42689473684210527, + "grad_norm": 0.5, + "learning_rate": 0.00012735672514619883, + "loss": 3.5981, + "step": 8111 + }, + { + "epoch": 0.42694736842105263, + "grad_norm": 0.58203125, + "learning_rate": 0.00012734502923976608, + "loss": 3.6815, + "step": 8112 + }, + { + "epoch": 0.427, + "grad_norm": 0.5234375, + "learning_rate": 0.00012733333333333336, + "loss": 2.9718, + "step": 8113 + }, + { + "epoch": 0.42705263157894735, + "grad_norm": 0.5, + "learning_rate": 0.0001273216374269006, + "loss": 3.6501, + "step": 8114 + }, + { + "epoch": 0.4271052631578947, + "grad_norm": 0.48046875, + "learning_rate": 0.00012730994152046786, + "loss": 3.5536, + "step": 8115 + }, + { + "epoch": 0.4271578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.0001272982456140351, + "loss": 3.2759, + "step": 8116 + }, + { + "epoch": 0.4272105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.00012728654970760235, + "loss": 3.6912, + "step": 8117 + }, + { + "epoch": 0.42726315789473684, + "grad_norm": 0.5859375, + "learning_rate": 0.0001272748538011696, + "loss": 3.646, + "step": 8118 + }, + { + "epoch": 0.4273157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00012726315789473683, + "loss": 3.7382, + "step": 8119 + }, + { + "epoch": 0.42736842105263156, + "grad_norm": 0.66796875, + "learning_rate": 0.0001272514619883041, + "loss": 3.794, + "step": 8120 + }, + { + "epoch": 0.4274210526315789, + "grad_norm": 0.4921875, + "learning_rate": 0.00012723976608187135, + "loss": 3.2435, + "step": 8121 + }, + { + "epoch": 0.42747368421052634, + "grad_norm": 0.470703125, + "learning_rate": 0.0001272280701754386, + "loss": 2.9415, + "step": 8122 + }, + { + "epoch": 0.4275263157894737, + "grad_norm": 0.482421875, + "learning_rate": 0.00012721637426900585, + "loss": 3.8172, + "step": 8123 + }, + { + "epoch": 0.42757894736842106, + "grad_norm": 0.46484375, + "learning_rate": 0.0001272046783625731, + "loss": 3.5094, + "step": 8124 + }, + { + "epoch": 0.4276315789473684, + "grad_norm": 0.482421875, + "learning_rate": 0.00012719298245614035, + "loss": 3.2283, + "step": 8125 + }, + { + "epoch": 0.4276842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.0001271812865497076, + "loss": 3.5559, + "step": 8126 + }, + { + "epoch": 0.42773684210526314, + "grad_norm": 0.5234375, + "learning_rate": 0.00012716959064327485, + "loss": 3.317, + "step": 8127 + }, + { + "epoch": 0.42778947368421055, + "grad_norm": 0.515625, + "learning_rate": 0.00012715789473684213, + "loss": 3.3225, + "step": 8128 + }, + { + "epoch": 0.4278421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00012714619883040938, + "loss": 3.809, + "step": 8129 + }, + { + "epoch": 0.42789473684210527, + "grad_norm": 0.515625, + "learning_rate": 0.00012713450292397663, + "loss": 3.3281, + "step": 8130 + }, + { + "epoch": 0.42794736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00012712280701754388, + "loss": 3.4426, + "step": 8131 + }, + { + "epoch": 0.428, + "grad_norm": 0.470703125, + "learning_rate": 0.00012711111111111113, + "loss": 3.6738, + "step": 8132 + }, + { + "epoch": 0.42805263157894735, + "grad_norm": 0.474609375, + "learning_rate": 0.00012709941520467838, + "loss": 3.535, + "step": 8133 + }, + { + "epoch": 0.4281052631578947, + "grad_norm": 0.466796875, + "learning_rate": 0.0001270877192982456, + "loss": 3.4749, + "step": 8134 + }, + { + "epoch": 0.4281578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.00012707602339181287, + "loss": 3.5213, + "step": 8135 + }, + { + "epoch": 0.4282105263157895, + "grad_norm": 0.76953125, + "learning_rate": 0.00012706432748538012, + "loss": 3.6138, + "step": 8136 + }, + { + "epoch": 0.42826315789473685, + "grad_norm": 0.52734375, + "learning_rate": 0.00012705263157894737, + "loss": 3.7426, + "step": 8137 + }, + { + "epoch": 0.4283157894736842, + "grad_norm": 0.5625, + "learning_rate": 0.00012704093567251462, + "loss": 3.0492, + "step": 8138 + }, + { + "epoch": 0.42836842105263156, + "grad_norm": 0.49609375, + "learning_rate": 0.00012702923976608187, + "loss": 3.5566, + "step": 8139 + }, + { + "epoch": 0.4284210526315789, + "grad_norm": 0.484375, + "learning_rate": 0.00012701754385964912, + "loss": 3.8344, + "step": 8140 + }, + { + "epoch": 0.42847368421052634, + "grad_norm": 0.49609375, + "learning_rate": 0.00012700584795321637, + "loss": 3.9321, + "step": 8141 + }, + { + "epoch": 0.4285263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.00012699415204678362, + "loss": 3.4907, + "step": 8142 + }, + { + "epoch": 0.42857894736842106, + "grad_norm": 0.53515625, + "learning_rate": 0.0001269824561403509, + "loss": 3.7873, + "step": 8143 + }, + { + "epoch": 0.4286315789473684, + "grad_norm": 0.7421875, + "learning_rate": 0.00012697076023391815, + "loss": 3.364, + "step": 8144 + }, + { + "epoch": 0.4286842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.0001269590643274854, + "loss": 3.425, + "step": 8145 + }, + { + "epoch": 0.42873684210526314, + "grad_norm": 0.4765625, + "learning_rate": 0.00012694736842105265, + "loss": 3.2758, + "step": 8146 + }, + { + "epoch": 0.42878947368421055, + "grad_norm": 0.478515625, + "learning_rate": 0.0001269356725146199, + "loss": 3.6199, + "step": 8147 + }, + { + "epoch": 0.4288421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00012692397660818715, + "loss": 3.4662, + "step": 8148 + }, + { + "epoch": 0.4288947368421053, + "grad_norm": 0.478515625, + "learning_rate": 0.00012691228070175437, + "loss": 3.5033, + "step": 8149 + }, + { + "epoch": 0.42894736842105263, + "grad_norm": 0.578125, + "learning_rate": 0.00012690058479532165, + "loss": 3.3951, + "step": 8150 + }, + { + "epoch": 0.429, + "grad_norm": 0.52734375, + "learning_rate": 0.0001268888888888889, + "loss": 3.6567, + "step": 8151 + }, + { + "epoch": 0.42905263157894735, + "grad_norm": 0.55078125, + "learning_rate": 0.00012687719298245614, + "loss": 3.5642, + "step": 8152 + }, + { + "epoch": 0.4291052631578947, + "grad_norm": 0.50390625, + "learning_rate": 0.0001268654970760234, + "loss": 3.8021, + "step": 8153 + }, + { + "epoch": 0.4291578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00012685380116959064, + "loss": 3.2984, + "step": 8154 + }, + { + "epoch": 0.4292105263157895, + "grad_norm": 0.490234375, + "learning_rate": 0.0001268421052631579, + "loss": 3.5269, + "step": 8155 + }, + { + "epoch": 0.42926315789473685, + "grad_norm": 0.609375, + "learning_rate": 0.00012683040935672514, + "loss": 2.9119, + "step": 8156 + }, + { + "epoch": 0.4293157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00012681871345029242, + "loss": 3.7704, + "step": 8157 + }, + { + "epoch": 0.42936842105263157, + "grad_norm": 0.466796875, + "learning_rate": 0.00012680701754385967, + "loss": 3.6285, + "step": 8158 + }, + { + "epoch": 0.4294210526315789, + "grad_norm": 0.498046875, + "learning_rate": 0.00012679532163742692, + "loss": 3.5577, + "step": 8159 + }, + { + "epoch": 0.42947368421052634, + "grad_norm": 0.53125, + "learning_rate": 0.00012678362573099417, + "loss": 3.0371, + "step": 8160 + }, + { + "epoch": 0.4295263157894737, + "grad_norm": 0.486328125, + "learning_rate": 0.00012677192982456142, + "loss": 3.3563, + "step": 8161 + }, + { + "epoch": 0.42957894736842106, + "grad_norm": 0.5078125, + "learning_rate": 0.00012676023391812867, + "loss": 3.5906, + "step": 8162 + }, + { + "epoch": 0.4296315789473684, + "grad_norm": 0.484375, + "learning_rate": 0.00012674853801169592, + "loss": 3.5918, + "step": 8163 + }, + { + "epoch": 0.4296842105263158, + "grad_norm": 0.4921875, + "learning_rate": 0.00012673684210526317, + "loss": 3.6257, + "step": 8164 + }, + { + "epoch": 0.42973684210526314, + "grad_norm": 0.490234375, + "learning_rate": 0.00012672514619883042, + "loss": 3.6694, + "step": 8165 + }, + { + "epoch": 0.4297894736842105, + "grad_norm": 0.48828125, + "learning_rate": 0.00012671345029239767, + "loss": 3.1759, + "step": 8166 + }, + { + "epoch": 0.4298421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00012670175438596492, + "loss": 3.2595, + "step": 8167 + }, + { + "epoch": 0.4298947368421053, + "grad_norm": 0.484375, + "learning_rate": 0.00012669005847953217, + "loss": 3.5363, + "step": 8168 + }, + { + "epoch": 0.42994736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00012667836257309942, + "loss": 3.5683, + "step": 8169 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.00012666666666666666, + "loss": 3.6026, + "step": 8170 + }, + { + "epoch": 0.43005263157894735, + "grad_norm": 0.78125, + "learning_rate": 0.00012665497076023391, + "loss": 2.9284, + "step": 8171 + }, + { + "epoch": 0.4301052631578947, + "grad_norm": 0.58203125, + "learning_rate": 0.0001266432748538012, + "loss": 3.2784, + "step": 8172 + }, + { + "epoch": 0.43015789473684213, + "grad_norm": 0.7109375, + "learning_rate": 0.00012663157894736844, + "loss": 3.6795, + "step": 8173 + }, + { + "epoch": 0.4302105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.0001266198830409357, + "loss": 3.946, + "step": 8174 + }, + { + "epoch": 0.43026315789473685, + "grad_norm": 0.478515625, + "learning_rate": 0.00012660818713450294, + "loss": 3.6308, + "step": 8175 + }, + { + "epoch": 0.4303157894736842, + "grad_norm": 0.46484375, + "learning_rate": 0.0001265964912280702, + "loss": 3.5331, + "step": 8176 + }, + { + "epoch": 0.43036842105263157, + "grad_norm": 0.5859375, + "learning_rate": 0.00012658479532163744, + "loss": 3.6775, + "step": 8177 + }, + { + "epoch": 0.4304210526315789, + "grad_norm": 0.494140625, + "learning_rate": 0.0001265730994152047, + "loss": 3.1045, + "step": 8178 + }, + { + "epoch": 0.43047368421052634, + "grad_norm": 0.53125, + "learning_rate": 0.00012656140350877194, + "loss": 3.4305, + "step": 8179 + }, + { + "epoch": 0.4305263157894737, + "grad_norm": 0.5234375, + "learning_rate": 0.0001265497076023392, + "loss": 3.6208, + "step": 8180 + }, + { + "epoch": 0.43057894736842106, + "grad_norm": 0.51953125, + "learning_rate": 0.00012653801169590644, + "loss": 3.6552, + "step": 8181 + }, + { + "epoch": 0.4306315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.0001265263157894737, + "loss": 3.3604, + "step": 8182 + }, + { + "epoch": 0.4306842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.00012651461988304094, + "loss": 3.5038, + "step": 8183 + }, + { + "epoch": 0.43073684210526314, + "grad_norm": 0.55078125, + "learning_rate": 0.00012650292397660819, + "loss": 3.6944, + "step": 8184 + }, + { + "epoch": 0.4307894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.00012649122807017544, + "loss": 3.7159, + "step": 8185 + }, + { + "epoch": 0.4308421052631579, + "grad_norm": 0.828125, + "learning_rate": 0.00012647953216374269, + "loss": 3.411, + "step": 8186 + }, + { + "epoch": 0.4308947368421053, + "grad_norm": 0.484375, + "learning_rate": 0.00012646783625730996, + "loss": 3.5166, + "step": 8187 + }, + { + "epoch": 0.43094736842105263, + "grad_norm": 0.484375, + "learning_rate": 0.0001264561403508772, + "loss": 3.5694, + "step": 8188 + }, + { + "epoch": 0.431, + "grad_norm": 0.4765625, + "learning_rate": 0.00012644444444444446, + "loss": 3.5313, + "step": 8189 + }, + { + "epoch": 0.43105263157894735, + "grad_norm": 0.478515625, + "learning_rate": 0.0001264327485380117, + "loss": 3.5847, + "step": 8190 + }, + { + "epoch": 0.4311052631578947, + "grad_norm": 0.4921875, + "learning_rate": 0.00012642105263157896, + "loss": 3.0845, + "step": 8191 + }, + { + "epoch": 0.43115789473684213, + "grad_norm": 0.55078125, + "learning_rate": 0.0001264093567251462, + "loss": 3.3519, + "step": 8192 + }, + { + "epoch": 0.4312105263157895, + "grad_norm": 0.53515625, + "learning_rate": 0.00012639766081871346, + "loss": 3.3841, + "step": 8193 + }, + { + "epoch": 0.43126315789473685, + "grad_norm": 0.5078125, + "learning_rate": 0.0001263859649122807, + "loss": 3.5765, + "step": 8194 + }, + { + "epoch": 0.4313157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.00012637426900584796, + "loss": 3.6955, + "step": 8195 + }, + { + "epoch": 0.43136842105263157, + "grad_norm": 0.5859375, + "learning_rate": 0.0001263625730994152, + "loss": 3.6937, + "step": 8196 + }, + { + "epoch": 0.4314210526315789, + "grad_norm": 0.58984375, + "learning_rate": 0.00012635087719298246, + "loss": 3.6931, + "step": 8197 + }, + { + "epoch": 0.43147368421052634, + "grad_norm": 0.54296875, + "learning_rate": 0.0001263391812865497, + "loss": 3.2862, + "step": 8198 + }, + { + "epoch": 0.4315263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.00012632748538011696, + "loss": 3.5614, + "step": 8199 + }, + { + "epoch": 0.43157894736842106, + "grad_norm": 0.546875, + "learning_rate": 0.0001263157894736842, + "loss": 3.7808, + "step": 8200 + }, + { + "epoch": 0.4316315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00012630409356725146, + "loss": 3.3052, + "step": 8201 + }, + { + "epoch": 0.4316842105263158, + "grad_norm": 0.609375, + "learning_rate": 0.00012629239766081873, + "loss": 3.7882, + "step": 8202 + }, + { + "epoch": 0.43173684210526314, + "grad_norm": 0.47265625, + "learning_rate": 0.00012628070175438598, + "loss": 3.2183, + "step": 8203 + }, + { + "epoch": 0.4317894736842105, + "grad_norm": 0.46875, + "learning_rate": 0.00012626900584795323, + "loss": 3.2102, + "step": 8204 + }, + { + "epoch": 0.4318421052631579, + "grad_norm": 0.47265625, + "learning_rate": 0.00012625730994152048, + "loss": 3.6594, + "step": 8205 + }, + { + "epoch": 0.4318947368421053, + "grad_norm": 0.486328125, + "learning_rate": 0.00012624561403508773, + "loss": 3.2155, + "step": 8206 + }, + { + "epoch": 0.43194736842105264, + "grad_norm": 0.498046875, + "learning_rate": 0.00012623391812865498, + "loss": 3.9003, + "step": 8207 + }, + { + "epoch": 0.432, + "grad_norm": 0.51953125, + "learning_rate": 0.00012622222222222223, + "loss": 3.616, + "step": 8208 + }, + { + "epoch": 0.43205263157894735, + "grad_norm": 0.51953125, + "learning_rate": 0.00012621052631578948, + "loss": 3.5599, + "step": 8209 + }, + { + "epoch": 0.4321052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00012619883040935673, + "loss": 3.6919, + "step": 8210 + }, + { + "epoch": 0.43215789473684213, + "grad_norm": 0.5234375, + "learning_rate": 0.00012618713450292398, + "loss": 3.3257, + "step": 8211 + }, + { + "epoch": 0.4322105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.00012617543859649123, + "loss": 3.349, + "step": 8212 + }, + { + "epoch": 0.43226315789473685, + "grad_norm": 0.490234375, + "learning_rate": 0.00012616374269005848, + "loss": 3.9499, + "step": 8213 + }, + { + "epoch": 0.4323157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00012615204678362573, + "loss": 3.3772, + "step": 8214 + }, + { + "epoch": 0.43236842105263157, + "grad_norm": 0.494140625, + "learning_rate": 0.00012614035087719298, + "loss": 3.7152, + "step": 8215 + }, + { + "epoch": 0.43242105263157893, + "grad_norm": 0.53125, + "learning_rate": 0.00012612865497076023, + "loss": 3.58, + "step": 8216 + }, + { + "epoch": 0.43247368421052634, + "grad_norm": 0.50390625, + "learning_rate": 0.0001261169590643275, + "loss": 3.6931, + "step": 8217 + }, + { + "epoch": 0.4325263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00012610526315789475, + "loss": 3.406, + "step": 8218 + }, + { + "epoch": 0.43257894736842106, + "grad_norm": 0.50390625, + "learning_rate": 0.000126093567251462, + "loss": 3.3677, + "step": 8219 + }, + { + "epoch": 0.4326315789473684, + "grad_norm": 0.474609375, + "learning_rate": 0.00012608187134502925, + "loss": 3.2473, + "step": 8220 + }, + { + "epoch": 0.4326842105263158, + "grad_norm": 0.5625, + "learning_rate": 0.0001260701754385965, + "loss": 3.4306, + "step": 8221 + }, + { + "epoch": 0.43273684210526314, + "grad_norm": 0.5078125, + "learning_rate": 0.00012605847953216375, + "loss": 2.9255, + "step": 8222 + }, + { + "epoch": 0.4327894736842105, + "grad_norm": 0.4921875, + "learning_rate": 0.000126046783625731, + "loss": 3.411, + "step": 8223 + }, + { + "epoch": 0.4328421052631579, + "grad_norm": 0.48046875, + "learning_rate": 0.00012603508771929825, + "loss": 3.5741, + "step": 8224 + }, + { + "epoch": 0.4328947368421053, + "grad_norm": 0.486328125, + "learning_rate": 0.0001260233918128655, + "loss": 3.271, + "step": 8225 + }, + { + "epoch": 0.43294736842105264, + "grad_norm": 0.68359375, + "learning_rate": 0.00012601169590643275, + "loss": 3.4821, + "step": 8226 + }, + { + "epoch": 0.433, + "grad_norm": 1.4375, + "learning_rate": 0.000126, + "loss": 3.2714, + "step": 8227 + }, + { + "epoch": 0.43305263157894736, + "grad_norm": 0.4921875, + "learning_rate": 0.00012598830409356725, + "loss": 3.6289, + "step": 8228 + }, + { + "epoch": 0.4331052631578947, + "grad_norm": 0.490234375, + "learning_rate": 0.0001259766081871345, + "loss": 3.8969, + "step": 8229 + }, + { + "epoch": 0.43315789473684213, + "grad_norm": 0.49609375, + "learning_rate": 0.00012596491228070175, + "loss": 3.8847, + "step": 8230 + }, + { + "epoch": 0.4332105263157895, + "grad_norm": 0.515625, + "learning_rate": 0.000125953216374269, + "loss": 3.3946, + "step": 8231 + }, + { + "epoch": 0.43326315789473685, + "grad_norm": 0.53125, + "learning_rate": 0.00012594152046783627, + "loss": 3.8168, + "step": 8232 + }, + { + "epoch": 0.4333157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.00012592982456140352, + "loss": 3.5143, + "step": 8233 + }, + { + "epoch": 0.43336842105263157, + "grad_norm": 0.734375, + "learning_rate": 0.00012591812865497077, + "loss": 3.5101, + "step": 8234 + }, + { + "epoch": 0.43342105263157893, + "grad_norm": 0.51171875, + "learning_rate": 0.00012590643274853802, + "loss": 3.7775, + "step": 8235 + }, + { + "epoch": 0.4334736842105263, + "grad_norm": 0.51953125, + "learning_rate": 0.00012589473684210527, + "loss": 3.4769, + "step": 8236 + }, + { + "epoch": 0.4335263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00012588304093567252, + "loss": 4.0015, + "step": 8237 + }, + { + "epoch": 0.43357894736842106, + "grad_norm": 0.5078125, + "learning_rate": 0.00012587134502923977, + "loss": 3.4923, + "step": 8238 + }, + { + "epoch": 0.4336315789473684, + "grad_norm": 0.5546875, + "learning_rate": 0.00012585964912280702, + "loss": 3.6748, + "step": 8239 + }, + { + "epoch": 0.4336842105263158, + "grad_norm": 0.609375, + "learning_rate": 0.00012584795321637427, + "loss": 3.4753, + "step": 8240 + }, + { + "epoch": 0.43373684210526314, + "grad_norm": 0.5625, + "learning_rate": 0.00012583625730994152, + "loss": 3.6441, + "step": 8241 + }, + { + "epoch": 0.4337894736842105, + "grad_norm": 0.546875, + "learning_rate": 0.00012582456140350877, + "loss": 3.7017, + "step": 8242 + }, + { + "epoch": 0.4338421052631579, + "grad_norm": 0.5546875, + "learning_rate": 0.00012581286549707602, + "loss": 3.4599, + "step": 8243 + }, + { + "epoch": 0.4338947368421053, + "grad_norm": 0.51171875, + "learning_rate": 0.00012580116959064327, + "loss": 3.6152, + "step": 8244 + }, + { + "epoch": 0.43394736842105264, + "grad_norm": 0.5078125, + "learning_rate": 0.00012578947368421052, + "loss": 3.7644, + "step": 8245 + }, + { + "epoch": 0.434, + "grad_norm": 0.54296875, + "learning_rate": 0.0001257777777777778, + "loss": 3.6133, + "step": 8246 + }, + { + "epoch": 0.43405263157894736, + "grad_norm": 0.4765625, + "learning_rate": 0.00012576608187134505, + "loss": 3.499, + "step": 8247 + }, + { + "epoch": 0.4341052631578947, + "grad_norm": 0.546875, + "learning_rate": 0.0001257543859649123, + "loss": 3.6842, + "step": 8248 + }, + { + "epoch": 0.43415789473684213, + "grad_norm": 0.57421875, + "learning_rate": 0.00012574269005847955, + "loss": 3.1408, + "step": 8249 + }, + { + "epoch": 0.4342105263157895, + "grad_norm": 0.48046875, + "learning_rate": 0.0001257309941520468, + "loss": 3.6558, + "step": 8250 + }, + { + "epoch": 0.43426315789473685, + "grad_norm": 0.47265625, + "learning_rate": 0.00012571929824561404, + "loss": 3.6492, + "step": 8251 + }, + { + "epoch": 0.4343157894736842, + "grad_norm": 0.578125, + "learning_rate": 0.0001257076023391813, + "loss": 3.7605, + "step": 8252 + }, + { + "epoch": 0.43436842105263157, + "grad_norm": 0.51953125, + "learning_rate": 0.00012569590643274854, + "loss": 3.5072, + "step": 8253 + }, + { + "epoch": 0.43442105263157893, + "grad_norm": 0.515625, + "learning_rate": 0.0001256842105263158, + "loss": 3.5316, + "step": 8254 + }, + { + "epoch": 0.4344736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00012567251461988304, + "loss": 3.7198, + "step": 8255 + }, + { + "epoch": 0.4345263157894737, + "grad_norm": 0.48046875, + "learning_rate": 0.0001256608187134503, + "loss": 3.5644, + "step": 8256 + }, + { + "epoch": 0.43457894736842106, + "grad_norm": 0.44921875, + "learning_rate": 0.00012564912280701754, + "loss": 3.5152, + "step": 8257 + }, + { + "epoch": 0.4346315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.0001256374269005848, + "loss": 3.2085, + "step": 8258 + }, + { + "epoch": 0.4346842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.00012562573099415204, + "loss": 3.396, + "step": 8259 + }, + { + "epoch": 0.43473684210526314, + "grad_norm": 0.474609375, + "learning_rate": 0.0001256140350877193, + "loss": 3.4108, + "step": 8260 + }, + { + "epoch": 0.4347894736842105, + "grad_norm": 0.6640625, + "learning_rate": 0.00012560233918128657, + "loss": 3.2818, + "step": 8261 + }, + { + "epoch": 0.4348421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00012559064327485382, + "loss": 3.7018, + "step": 8262 + }, + { + "epoch": 0.4348947368421053, + "grad_norm": 0.5703125, + "learning_rate": 0.00012557894736842107, + "loss": 3.678, + "step": 8263 + }, + { + "epoch": 0.43494736842105264, + "grad_norm": 0.482421875, + "learning_rate": 0.00012556725146198832, + "loss": 3.5924, + "step": 8264 + }, + { + "epoch": 0.435, + "grad_norm": 0.515625, + "learning_rate": 0.00012555555555555557, + "loss": 3.722, + "step": 8265 + }, + { + "epoch": 0.43505263157894736, + "grad_norm": 0.490234375, + "learning_rate": 0.00012554385964912282, + "loss": 3.4935, + "step": 8266 + }, + { + "epoch": 0.4351052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.00012553216374269006, + "loss": 3.306, + "step": 8267 + }, + { + "epoch": 0.43515789473684213, + "grad_norm": 0.5234375, + "learning_rate": 0.00012552046783625731, + "loss": 3.4845, + "step": 8268 + }, + { + "epoch": 0.4352105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00012550877192982456, + "loss": 3.5823, + "step": 8269 + }, + { + "epoch": 0.43526315789473685, + "grad_norm": 0.55078125, + "learning_rate": 0.00012549707602339181, + "loss": 3.2131, + "step": 8270 + }, + { + "epoch": 0.4353157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00012548538011695906, + "loss": 3.5282, + "step": 8271 + }, + { + "epoch": 0.43536842105263157, + "grad_norm": 0.43359375, + "learning_rate": 0.0001254736842105263, + "loss": 3.249, + "step": 8272 + }, + { + "epoch": 0.43542105263157893, + "grad_norm": 0.470703125, + "learning_rate": 0.00012546198830409356, + "loss": 3.5978, + "step": 8273 + }, + { + "epoch": 0.4354736842105263, + "grad_norm": 0.47265625, + "learning_rate": 0.0001254502923976608, + "loss": 3.3354, + "step": 8274 + }, + { + "epoch": 0.4355263157894737, + "grad_norm": 0.462890625, + "learning_rate": 0.00012543859649122806, + "loss": 3.7653, + "step": 8275 + }, + { + "epoch": 0.43557894736842107, + "grad_norm": 0.66015625, + "learning_rate": 0.00012542690058479534, + "loss": 3.4987, + "step": 8276 + }, + { + "epoch": 0.4356315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.0001254152046783626, + "loss": 3.2982, + "step": 8277 + }, + { + "epoch": 0.4356842105263158, + "grad_norm": 0.4765625, + "learning_rate": 0.00012540350877192984, + "loss": 4.1079, + "step": 8278 + }, + { + "epoch": 0.43573684210526314, + "grad_norm": 0.50390625, + "learning_rate": 0.0001253918128654971, + "loss": 3.706, + "step": 8279 + }, + { + "epoch": 0.4357894736842105, + "grad_norm": 0.4765625, + "learning_rate": 0.00012538011695906434, + "loss": 3.8354, + "step": 8280 + }, + { + "epoch": 0.4358421052631579, + "grad_norm": 0.5703125, + "learning_rate": 0.00012536842105263159, + "loss": 3.4435, + "step": 8281 + }, + { + "epoch": 0.4358947368421053, + "grad_norm": 0.6015625, + "learning_rate": 0.00012535672514619884, + "loss": 3.1848, + "step": 8282 + }, + { + "epoch": 0.43594736842105264, + "grad_norm": 0.53125, + "learning_rate": 0.00012534502923976609, + "loss": 3.4233, + "step": 8283 + }, + { + "epoch": 0.436, + "grad_norm": 0.494140625, + "learning_rate": 0.00012533333333333334, + "loss": 3.8125, + "step": 8284 + }, + { + "epoch": 0.43605263157894736, + "grad_norm": 0.5078125, + "learning_rate": 0.00012532163742690058, + "loss": 3.2793, + "step": 8285 + }, + { + "epoch": 0.4361052631578947, + "grad_norm": 0.50390625, + "learning_rate": 0.00012530994152046783, + "loss": 3.5064, + "step": 8286 + }, + { + "epoch": 0.4361578947368421, + "grad_norm": 0.5234375, + "learning_rate": 0.00012529824561403508, + "loss": 3.5841, + "step": 8287 + }, + { + "epoch": 0.4362105263157895, + "grad_norm": 0.451171875, + "learning_rate": 0.00012528654970760233, + "loss": 3.3007, + "step": 8288 + }, + { + "epoch": 0.43626315789473685, + "grad_norm": 0.46484375, + "learning_rate": 0.00012527485380116958, + "loss": 3.5365, + "step": 8289 + }, + { + "epoch": 0.4363157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00012526315789473683, + "loss": 3.656, + "step": 8290 + }, + { + "epoch": 0.43636842105263157, + "grad_norm": 0.5234375, + "learning_rate": 0.0001252514619883041, + "loss": 3.2275, + "step": 8291 + }, + { + "epoch": 0.43642105263157893, + "grad_norm": 0.490234375, + "learning_rate": 0.00012523976608187136, + "loss": 3.5235, + "step": 8292 + }, + { + "epoch": 0.4364736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.0001252280701754386, + "loss": 3.4796, + "step": 8293 + }, + { + "epoch": 0.4365263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00012521637426900586, + "loss": 3.4068, + "step": 8294 + }, + { + "epoch": 0.43657894736842107, + "grad_norm": 0.4921875, + "learning_rate": 0.0001252046783625731, + "loss": 3.4611, + "step": 8295 + }, + { + "epoch": 0.4366315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00012519298245614036, + "loss": 3.7895, + "step": 8296 + }, + { + "epoch": 0.4366842105263158, + "grad_norm": 0.4765625, + "learning_rate": 0.0001251812865497076, + "loss": 3.4087, + "step": 8297 + }, + { + "epoch": 0.43673684210526315, + "grad_norm": 0.478515625, + "learning_rate": 0.00012516959064327486, + "loss": 3.206, + "step": 8298 + }, + { + "epoch": 0.4367894736842105, + "grad_norm": 0.48046875, + "learning_rate": 0.0001251578947368421, + "loss": 3.7711, + "step": 8299 + }, + { + "epoch": 0.4368421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00012514619883040936, + "loss": 3.8654, + "step": 8300 + }, + { + "epoch": 0.4368947368421053, + "grad_norm": 0.5546875, + "learning_rate": 0.0001251345029239766, + "loss": 3.333, + "step": 8301 + }, + { + "epoch": 0.43694736842105264, + "grad_norm": 0.55859375, + "learning_rate": 0.00012512280701754385, + "loss": 3.5095, + "step": 8302 + }, + { + "epoch": 0.437, + "grad_norm": 0.70703125, + "learning_rate": 0.0001251111111111111, + "loss": 3.6719, + "step": 8303 + }, + { + "epoch": 0.43705263157894736, + "grad_norm": 0.5078125, + "learning_rate": 0.00012509941520467835, + "loss": 3.0222, + "step": 8304 + }, + { + "epoch": 0.4371052631578947, + "grad_norm": 0.451171875, + "learning_rate": 0.0001250877192982456, + "loss": 3.9594, + "step": 8305 + }, + { + "epoch": 0.4371578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00012507602339181288, + "loss": 2.8902, + "step": 8306 + }, + { + "epoch": 0.4372105263157895, + "grad_norm": 0.482421875, + "learning_rate": 0.00012506432748538013, + "loss": 3.2975, + "step": 8307 + }, + { + "epoch": 0.43726315789473685, + "grad_norm": 0.451171875, + "learning_rate": 0.00012505263157894738, + "loss": 3.6049, + "step": 8308 + }, + { + "epoch": 0.4373157894736842, + "grad_norm": 0.458984375, + "learning_rate": 0.00012504093567251463, + "loss": 3.3727, + "step": 8309 + }, + { + "epoch": 0.4373684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00012502923976608188, + "loss": 3.5047, + "step": 8310 + }, + { + "epoch": 0.43742105263157893, + "grad_norm": 0.4765625, + "learning_rate": 0.00012501754385964913, + "loss": 3.2922, + "step": 8311 + }, + { + "epoch": 0.4374736842105263, + "grad_norm": 0.486328125, + "learning_rate": 0.00012500584795321638, + "loss": 3.6181, + "step": 8312 + }, + { + "epoch": 0.4375263157894737, + "grad_norm": 0.490234375, + "learning_rate": 0.00012499415204678365, + "loss": 3.1359, + "step": 8313 + }, + { + "epoch": 0.43757894736842107, + "grad_norm": 0.4375, + "learning_rate": 0.00012498245614035088, + "loss": 3.6638, + "step": 8314 + }, + { + "epoch": 0.4376315789473684, + "grad_norm": 0.484375, + "learning_rate": 0.00012497076023391813, + "loss": 3.6584, + "step": 8315 + }, + { + "epoch": 0.4376842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00012495906432748538, + "loss": 3.7702, + "step": 8316 + }, + { + "epoch": 0.43773684210526315, + "grad_norm": 0.50390625, + "learning_rate": 0.00012494736842105263, + "loss": 3.5623, + "step": 8317 + }, + { + "epoch": 0.4377894736842105, + "grad_norm": 0.486328125, + "learning_rate": 0.00012493567251461988, + "loss": 3.6987, + "step": 8318 + }, + { + "epoch": 0.4378421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00012492397660818712, + "loss": 3.4843, + "step": 8319 + }, + { + "epoch": 0.4378947368421053, + "grad_norm": 0.7265625, + "learning_rate": 0.0001249122807017544, + "loss": 3.6171, + "step": 8320 + }, + { + "epoch": 0.43794736842105264, + "grad_norm": 0.49609375, + "learning_rate": 0.00012490058479532165, + "loss": 3.2773, + "step": 8321 + }, + { + "epoch": 0.438, + "grad_norm": 0.4453125, + "learning_rate": 0.0001248888888888889, + "loss": 3.9213, + "step": 8322 + }, + { + "epoch": 0.43805263157894736, + "grad_norm": 0.55078125, + "learning_rate": 0.00012487719298245615, + "loss": 3.3571, + "step": 8323 + }, + { + "epoch": 0.4381052631578947, + "grad_norm": 0.494140625, + "learning_rate": 0.0001248654970760234, + "loss": 3.5173, + "step": 8324 + }, + { + "epoch": 0.4381578947368421, + "grad_norm": 0.44921875, + "learning_rate": 0.00012485380116959065, + "loss": 3.4256, + "step": 8325 + }, + { + "epoch": 0.4382105263157895, + "grad_norm": 0.46875, + "learning_rate": 0.0001248421052631579, + "loss": 3.6819, + "step": 8326 + }, + { + "epoch": 0.43826315789473685, + "grad_norm": 0.490234375, + "learning_rate": 0.00012483040935672515, + "loss": 3.5316, + "step": 8327 + }, + { + "epoch": 0.4383157894736842, + "grad_norm": 0.46875, + "learning_rate": 0.00012481871345029243, + "loss": 3.3657, + "step": 8328 + }, + { + "epoch": 0.4383684210526316, + "grad_norm": 0.478515625, + "learning_rate": 0.00012480701754385965, + "loss": 3.159, + "step": 8329 + }, + { + "epoch": 0.43842105263157893, + "grad_norm": 0.4609375, + "learning_rate": 0.0001247953216374269, + "loss": 3.5361, + "step": 8330 + }, + { + "epoch": 0.4384736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00012478362573099415, + "loss": 3.1025, + "step": 8331 + }, + { + "epoch": 0.4385263157894737, + "grad_norm": 0.4765625, + "learning_rate": 0.0001247719298245614, + "loss": 3.9455, + "step": 8332 + }, + { + "epoch": 0.43857894736842107, + "grad_norm": 0.443359375, + "learning_rate": 0.00012476023391812865, + "loss": 3.9312, + "step": 8333 + }, + { + "epoch": 0.43863157894736843, + "grad_norm": 0.56640625, + "learning_rate": 0.0001247485380116959, + "loss": 3.5683, + "step": 8334 + }, + { + "epoch": 0.4386842105263158, + "grad_norm": 0.498046875, + "learning_rate": 0.00012473684210526317, + "loss": 3.5551, + "step": 8335 + }, + { + "epoch": 0.43873684210526315, + "grad_norm": 0.49609375, + "learning_rate": 0.00012472514619883042, + "loss": 3.5492, + "step": 8336 + }, + { + "epoch": 0.4387894736842105, + "grad_norm": 0.51171875, + "learning_rate": 0.00012471345029239767, + "loss": 3.3061, + "step": 8337 + }, + { + "epoch": 0.4388421052631579, + "grad_norm": 0.578125, + "learning_rate": 0.00012470175438596492, + "loss": 3.3108, + "step": 8338 + }, + { + "epoch": 0.4388947368421053, + "grad_norm": 0.57421875, + "learning_rate": 0.00012469005847953217, + "loss": 3.5174, + "step": 8339 + }, + { + "epoch": 0.43894736842105264, + "grad_norm": 0.470703125, + "learning_rate": 0.00012467836257309942, + "loss": 3.6317, + "step": 8340 + }, + { + "epoch": 0.439, + "grad_norm": 0.5, + "learning_rate": 0.00012466666666666667, + "loss": 3.9889, + "step": 8341 + }, + { + "epoch": 0.43905263157894736, + "grad_norm": 0.55078125, + "learning_rate": 0.00012465497076023392, + "loss": 3.4411, + "step": 8342 + }, + { + "epoch": 0.4391052631578947, + "grad_norm": 0.474609375, + "learning_rate": 0.0001246432748538012, + "loss": 3.3037, + "step": 8343 + }, + { + "epoch": 0.4391578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.00012463157894736845, + "loss": 3.5559, + "step": 8344 + }, + { + "epoch": 0.4392105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00012461988304093567, + "loss": 3.7349, + "step": 8345 + }, + { + "epoch": 0.43926315789473686, + "grad_norm": 0.546875, + "learning_rate": 0.00012460818713450292, + "loss": 3.4353, + "step": 8346 + }, + { + "epoch": 0.4393157894736842, + "grad_norm": 0.55859375, + "learning_rate": 0.00012459649122807017, + "loss": 3.5861, + "step": 8347 + }, + { + "epoch": 0.4393684210526316, + "grad_norm": 0.490234375, + "learning_rate": 0.00012458479532163742, + "loss": 3.592, + "step": 8348 + }, + { + "epoch": 0.43942105263157893, + "grad_norm": 0.51953125, + "learning_rate": 0.00012457309941520467, + "loss": 3.1329, + "step": 8349 + }, + { + "epoch": 0.4394736842105263, + "grad_norm": 0.59375, + "learning_rate": 0.00012456140350877194, + "loss": 3.4015, + "step": 8350 + }, + { + "epoch": 0.4395263157894737, + "grad_norm": 0.46484375, + "learning_rate": 0.0001245497076023392, + "loss": 3.375, + "step": 8351 + }, + { + "epoch": 0.43957894736842107, + "grad_norm": 0.490234375, + "learning_rate": 0.00012453801169590644, + "loss": 3.1936, + "step": 8352 + }, + { + "epoch": 0.43963157894736843, + "grad_norm": 0.490234375, + "learning_rate": 0.0001245263157894737, + "loss": 3.3516, + "step": 8353 + }, + { + "epoch": 0.4396842105263158, + "grad_norm": 0.484375, + "learning_rate": 0.00012451461988304094, + "loss": 3.7068, + "step": 8354 + }, + { + "epoch": 0.43973684210526315, + "grad_norm": 0.48828125, + "learning_rate": 0.0001245029239766082, + "loss": 3.5409, + "step": 8355 + }, + { + "epoch": 0.4397894736842105, + "grad_norm": 0.54296875, + "learning_rate": 0.00012449122807017544, + "loss": 3.262, + "step": 8356 + }, + { + "epoch": 0.43984210526315787, + "grad_norm": 0.43359375, + "learning_rate": 0.00012447953216374272, + "loss": 3.765, + "step": 8357 + }, + { + "epoch": 0.4398947368421053, + "grad_norm": 0.6796875, + "learning_rate": 0.00012446783625730997, + "loss": 3.6916, + "step": 8358 + }, + { + "epoch": 0.43994736842105264, + "grad_norm": 0.546875, + "learning_rate": 0.00012445614035087722, + "loss": 3.7359, + "step": 8359 + }, + { + "epoch": 0.44, + "grad_norm": 0.5, + "learning_rate": 0.00012444444444444444, + "loss": 3.2491, + "step": 8360 + }, + { + "epoch": 0.44005263157894736, + "grad_norm": 0.484375, + "learning_rate": 0.0001244327485380117, + "loss": 3.4055, + "step": 8361 + }, + { + "epoch": 0.4401052631578947, + "grad_norm": 0.443359375, + "learning_rate": 0.00012442105263157894, + "loss": 3.6218, + "step": 8362 + }, + { + "epoch": 0.4401578947368421, + "grad_norm": 0.4609375, + "learning_rate": 0.0001244093567251462, + "loss": 3.687, + "step": 8363 + }, + { + "epoch": 0.4402105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.00012439766081871344, + "loss": 3.8164, + "step": 8364 + }, + { + "epoch": 0.44026315789473686, + "grad_norm": 0.486328125, + "learning_rate": 0.00012438596491228071, + "loss": 3.6456, + "step": 8365 + }, + { + "epoch": 0.4403157894736842, + "grad_norm": 0.52734375, + "learning_rate": 0.00012437426900584796, + "loss": 3.4676, + "step": 8366 + }, + { + "epoch": 0.4403684210526316, + "grad_norm": 0.51171875, + "learning_rate": 0.00012436257309941521, + "loss": 3.1518, + "step": 8367 + }, + { + "epoch": 0.44042105263157894, + "grad_norm": 0.51953125, + "learning_rate": 0.00012435087719298246, + "loss": 3.3696, + "step": 8368 + }, + { + "epoch": 0.4404736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.0001243391812865497, + "loss": 3.4915, + "step": 8369 + }, + { + "epoch": 0.4405263157894737, + "grad_norm": 0.466796875, + "learning_rate": 0.00012432748538011696, + "loss": 3.3656, + "step": 8370 + }, + { + "epoch": 0.44057894736842107, + "grad_norm": 0.51953125, + "learning_rate": 0.0001243157894736842, + "loss": 3.3615, + "step": 8371 + }, + { + "epoch": 0.44063157894736843, + "grad_norm": 0.462890625, + "learning_rate": 0.0001243040935672515, + "loss": 3.6945, + "step": 8372 + }, + { + "epoch": 0.4406842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00012429239766081874, + "loss": 3.7161, + "step": 8373 + }, + { + "epoch": 0.44073684210526315, + "grad_norm": 0.51171875, + "learning_rate": 0.000124280701754386, + "loss": 3.5736, + "step": 8374 + }, + { + "epoch": 0.4407894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.0001242690058479532, + "loss": 3.4216, + "step": 8375 + }, + { + "epoch": 0.44084210526315787, + "grad_norm": 0.50390625, + "learning_rate": 0.00012425730994152046, + "loss": 3.8699, + "step": 8376 + }, + { + "epoch": 0.4408947368421053, + "grad_norm": 0.66015625, + "learning_rate": 0.0001242456140350877, + "loss": 3.3212, + "step": 8377 + }, + { + "epoch": 0.44094736842105264, + "grad_norm": 0.451171875, + "learning_rate": 0.00012423391812865496, + "loss": 3.6341, + "step": 8378 + }, + { + "epoch": 0.441, + "grad_norm": 0.470703125, + "learning_rate": 0.0001242222222222222, + "loss": 3.7417, + "step": 8379 + }, + { + "epoch": 0.44105263157894736, + "grad_norm": 0.50390625, + "learning_rate": 0.00012421052631578949, + "loss": 3.7959, + "step": 8380 + }, + { + "epoch": 0.4411052631578947, + "grad_norm": 0.484375, + "learning_rate": 0.00012419883040935674, + "loss": 3.4161, + "step": 8381 + }, + { + "epoch": 0.4411578947368421, + "grad_norm": 0.482421875, + "learning_rate": 0.00012418713450292398, + "loss": 3.8811, + "step": 8382 + }, + { + "epoch": 0.4412105263157895, + "grad_norm": 0.57421875, + "learning_rate": 0.00012417543859649123, + "loss": 3.4995, + "step": 8383 + }, + { + "epoch": 0.44126315789473686, + "grad_norm": 0.4609375, + "learning_rate": 0.00012416374269005848, + "loss": 3.847, + "step": 8384 + }, + { + "epoch": 0.4413157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00012415204678362573, + "loss": 3.7211, + "step": 8385 + }, + { + "epoch": 0.4413684210526316, + "grad_norm": 0.47265625, + "learning_rate": 0.00012414035087719298, + "loss": 3.2476, + "step": 8386 + }, + { + "epoch": 0.44142105263157894, + "grad_norm": 0.5703125, + "learning_rate": 0.00012412865497076026, + "loss": 3.5326, + "step": 8387 + }, + { + "epoch": 0.4414736842105263, + "grad_norm": 0.60546875, + "learning_rate": 0.0001241169590643275, + "loss": 3.2395, + "step": 8388 + }, + { + "epoch": 0.4415263157894737, + "grad_norm": 0.5625, + "learning_rate": 0.00012410526315789476, + "loss": 3.5864, + "step": 8389 + }, + { + "epoch": 0.44157894736842107, + "grad_norm": 0.4921875, + "learning_rate": 0.000124093567251462, + "loss": 3.4574, + "step": 8390 + }, + { + "epoch": 0.44163157894736843, + "grad_norm": 0.546875, + "learning_rate": 0.00012408187134502923, + "loss": 3.5015, + "step": 8391 + }, + { + "epoch": 0.4416842105263158, + "grad_norm": 0.46875, + "learning_rate": 0.00012407017543859648, + "loss": 3.4623, + "step": 8392 + }, + { + "epoch": 0.44173684210526315, + "grad_norm": 0.48828125, + "learning_rate": 0.00012405847953216373, + "loss": 3.4238, + "step": 8393 + }, + { + "epoch": 0.4417894736842105, + "grad_norm": 0.484375, + "learning_rate": 0.000124046783625731, + "loss": 3.4592, + "step": 8394 + }, + { + "epoch": 0.44184210526315787, + "grad_norm": 0.48828125, + "learning_rate": 0.00012403508771929826, + "loss": 3.4786, + "step": 8395 + }, + { + "epoch": 0.4418947368421053, + "grad_norm": 0.5, + "learning_rate": 0.0001240233918128655, + "loss": 3.6963, + "step": 8396 + }, + { + "epoch": 0.44194736842105264, + "grad_norm": 0.4921875, + "learning_rate": 0.00012401169590643276, + "loss": 3.1244, + "step": 8397 + }, + { + "epoch": 0.442, + "grad_norm": 0.484375, + "learning_rate": 0.000124, + "loss": 3.5164, + "step": 8398 + }, + { + "epoch": 0.44205263157894736, + "grad_norm": 0.54296875, + "learning_rate": 0.00012398830409356725, + "loss": 3.4246, + "step": 8399 + }, + { + "epoch": 0.4421052631578947, + "grad_norm": 0.4609375, + "learning_rate": 0.0001239766081871345, + "loss": 3.5809, + "step": 8400 + }, + { + "epoch": 0.4421578947368421, + "grad_norm": 0.498046875, + "learning_rate": 0.00012396491228070175, + "loss": 3.7988, + "step": 8401 + }, + { + "epoch": 0.4422105263157895, + "grad_norm": 0.494140625, + "learning_rate": 0.00012395321637426903, + "loss": 3.6876, + "step": 8402 + }, + { + "epoch": 0.44226315789473686, + "grad_norm": 0.486328125, + "learning_rate": 0.00012394152046783628, + "loss": 3.2739, + "step": 8403 + }, + { + "epoch": 0.4423157894736842, + "grad_norm": 0.466796875, + "learning_rate": 0.00012392982456140353, + "loss": 3.347, + "step": 8404 + }, + { + "epoch": 0.4423684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00012391812865497078, + "loss": 3.53, + "step": 8405 + }, + { + "epoch": 0.44242105263157894, + "grad_norm": 0.47265625, + "learning_rate": 0.000123906432748538, + "loss": 3.3416, + "step": 8406 + }, + { + "epoch": 0.4424736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00012389473684210525, + "loss": 3.2109, + "step": 8407 + }, + { + "epoch": 0.44252631578947366, + "grad_norm": 0.56640625, + "learning_rate": 0.0001238830409356725, + "loss": 3.9452, + "step": 8408 + }, + { + "epoch": 0.44257894736842107, + "grad_norm": 0.466796875, + "learning_rate": 0.00012387134502923978, + "loss": 3.669, + "step": 8409 + }, + { + "epoch": 0.44263157894736843, + "grad_norm": 0.4921875, + "learning_rate": 0.00012385964912280703, + "loss": 3.6009, + "step": 8410 + }, + { + "epoch": 0.4426842105263158, + "grad_norm": 0.4921875, + "learning_rate": 0.00012384795321637428, + "loss": 2.9775, + "step": 8411 + }, + { + "epoch": 0.44273684210526315, + "grad_norm": 0.51953125, + "learning_rate": 0.00012383625730994153, + "loss": 3.4738, + "step": 8412 + }, + { + "epoch": 0.4427894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.00012382456140350878, + "loss": 3.4753, + "step": 8413 + }, + { + "epoch": 0.44284210526315787, + "grad_norm": 0.51953125, + "learning_rate": 0.00012381286549707603, + "loss": 3.6006, + "step": 8414 + }, + { + "epoch": 0.4428947368421053, + "grad_norm": 0.5390625, + "learning_rate": 0.00012380116959064328, + "loss": 3.231, + "step": 8415 + }, + { + "epoch": 0.44294736842105265, + "grad_norm": 0.490234375, + "learning_rate": 0.00012378947368421053, + "loss": 3.483, + "step": 8416 + }, + { + "epoch": 0.443, + "grad_norm": 0.4453125, + "learning_rate": 0.0001237777777777778, + "loss": 3.5617, + "step": 8417 + }, + { + "epoch": 0.44305263157894736, + "grad_norm": 0.4921875, + "learning_rate": 0.00012376608187134505, + "loss": 3.7201, + "step": 8418 + }, + { + "epoch": 0.4431052631578947, + "grad_norm": 0.466796875, + "learning_rate": 0.0001237543859649123, + "loss": 3.2989, + "step": 8419 + }, + { + "epoch": 0.4431578947368421, + "grad_norm": 0.474609375, + "learning_rate": 0.00012374269005847955, + "loss": 3.7425, + "step": 8420 + }, + { + "epoch": 0.4432105263157895, + "grad_norm": 0.5703125, + "learning_rate": 0.00012373099415204677, + "loss": 3.1744, + "step": 8421 + }, + { + "epoch": 0.44326315789473686, + "grad_norm": 0.52734375, + "learning_rate": 0.00012371929824561402, + "loss": 3.4327, + "step": 8422 + }, + { + "epoch": 0.4433157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00012370760233918127, + "loss": 3.6865, + "step": 8423 + }, + { + "epoch": 0.4433684210526316, + "grad_norm": 0.5, + "learning_rate": 0.00012369590643274855, + "loss": 3.6275, + "step": 8424 + }, + { + "epoch": 0.44342105263157894, + "grad_norm": 0.53125, + "learning_rate": 0.0001236842105263158, + "loss": 3.5805, + "step": 8425 + }, + { + "epoch": 0.4434736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00012367251461988305, + "loss": 3.155, + "step": 8426 + }, + { + "epoch": 0.44352631578947366, + "grad_norm": 0.482421875, + "learning_rate": 0.0001236608187134503, + "loss": 3.3803, + "step": 8427 + }, + { + "epoch": 0.4435789473684211, + "grad_norm": 0.4765625, + "learning_rate": 0.00012364912280701755, + "loss": 4.0347, + "step": 8428 + }, + { + "epoch": 0.44363157894736843, + "grad_norm": 0.62109375, + "learning_rate": 0.0001236374269005848, + "loss": 3.3616, + "step": 8429 + }, + { + "epoch": 0.4436842105263158, + "grad_norm": 0.494140625, + "learning_rate": 0.00012362573099415205, + "loss": 3.7629, + "step": 8430 + }, + { + "epoch": 0.44373684210526315, + "grad_norm": 0.5078125, + "learning_rate": 0.0001236140350877193, + "loss": 3.6287, + "step": 8431 + }, + { + "epoch": 0.4437894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.00012360233918128657, + "loss": 3.4932, + "step": 8432 + }, + { + "epoch": 0.44384210526315787, + "grad_norm": 0.482421875, + "learning_rate": 0.00012359064327485382, + "loss": 3.7537, + "step": 8433 + }, + { + "epoch": 0.4438947368421053, + "grad_norm": 0.5859375, + "learning_rate": 0.00012357894736842107, + "loss": 3.838, + "step": 8434 + }, + { + "epoch": 0.44394736842105265, + "grad_norm": 0.56640625, + "learning_rate": 0.00012356725146198832, + "loss": 3.2366, + "step": 8435 + }, + { + "epoch": 0.444, + "grad_norm": 0.609375, + "learning_rate": 0.00012355555555555557, + "loss": 3.738, + "step": 8436 + }, + { + "epoch": 0.44405263157894737, + "grad_norm": 0.515625, + "learning_rate": 0.0001235438596491228, + "loss": 3.9886, + "step": 8437 + }, + { + "epoch": 0.4441052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00012353216374269004, + "loss": 3.216, + "step": 8438 + }, + { + "epoch": 0.4441578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00012352046783625732, + "loss": 3.7889, + "step": 8439 + }, + { + "epoch": 0.4442105263157895, + "grad_norm": 0.490234375, + "learning_rate": 0.00012350877192982457, + "loss": 3.578, + "step": 8440 + }, + { + "epoch": 0.44426315789473686, + "grad_norm": 0.478515625, + "learning_rate": 0.00012349707602339182, + "loss": 3.3954, + "step": 8441 + }, + { + "epoch": 0.4443157894736842, + "grad_norm": 0.4921875, + "learning_rate": 0.00012348538011695907, + "loss": 3.4831, + "step": 8442 + }, + { + "epoch": 0.4443684210526316, + "grad_norm": 0.494140625, + "learning_rate": 0.00012347368421052632, + "loss": 3.1758, + "step": 8443 + }, + { + "epoch": 0.44442105263157894, + "grad_norm": 0.490234375, + "learning_rate": 0.00012346198830409357, + "loss": 3.4947, + "step": 8444 + }, + { + "epoch": 0.4444736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.00012345029239766082, + "loss": 3.7778, + "step": 8445 + }, + { + "epoch": 0.44452631578947366, + "grad_norm": 0.5625, + "learning_rate": 0.0001234385964912281, + "loss": 3.2012, + "step": 8446 + }, + { + "epoch": 0.4445789473684211, + "grad_norm": 0.52734375, + "learning_rate": 0.00012342690058479534, + "loss": 3.6001, + "step": 8447 + }, + { + "epoch": 0.44463157894736843, + "grad_norm": 0.494140625, + "learning_rate": 0.0001234152046783626, + "loss": 3.1288, + "step": 8448 + }, + { + "epoch": 0.4446842105263158, + "grad_norm": 0.478515625, + "learning_rate": 0.00012340350877192984, + "loss": 2.9573, + "step": 8449 + }, + { + "epoch": 0.44473684210526315, + "grad_norm": 0.466796875, + "learning_rate": 0.0001233918128654971, + "loss": 3.8229, + "step": 8450 + }, + { + "epoch": 0.4447894736842105, + "grad_norm": 0.482421875, + "learning_rate": 0.00012338011695906434, + "loss": 3.267, + "step": 8451 + }, + { + "epoch": 0.44484210526315787, + "grad_norm": 0.46484375, + "learning_rate": 0.00012336842105263156, + "loss": 3.6014, + "step": 8452 + }, + { + "epoch": 0.4448947368421053, + "grad_norm": 0.48828125, + "learning_rate": 0.00012335672514619881, + "loss": 3.2873, + "step": 8453 + }, + { + "epoch": 0.44494736842105265, + "grad_norm": 0.490234375, + "learning_rate": 0.0001233450292397661, + "loss": 3.5697, + "step": 8454 + }, + { + "epoch": 0.445, + "grad_norm": 0.57421875, + "learning_rate": 0.00012333333333333334, + "loss": 3.1784, + "step": 8455 + }, + { + "epoch": 0.44505263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.0001233216374269006, + "loss": 3.6551, + "step": 8456 + }, + { + "epoch": 0.4451052631578947, + "grad_norm": 0.490234375, + "learning_rate": 0.00012330994152046784, + "loss": 3.8948, + "step": 8457 + }, + { + "epoch": 0.4451578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.0001232982456140351, + "loss": 3.5294, + "step": 8458 + }, + { + "epoch": 0.4452105263157895, + "grad_norm": 0.46484375, + "learning_rate": 0.00012328654970760234, + "loss": 3.072, + "step": 8459 + }, + { + "epoch": 0.44526315789473686, + "grad_norm": 0.52734375, + "learning_rate": 0.0001232748538011696, + "loss": 3.2868, + "step": 8460 + }, + { + "epoch": 0.4453157894736842, + "grad_norm": 0.51953125, + "learning_rate": 0.00012326315789473687, + "loss": 3.1973, + "step": 8461 + }, + { + "epoch": 0.4453684210526316, + "grad_norm": 0.51171875, + "learning_rate": 0.00012325146198830411, + "loss": 3.5508, + "step": 8462 + }, + { + "epoch": 0.44542105263157894, + "grad_norm": 0.5234375, + "learning_rate": 0.00012323976608187136, + "loss": 3.7188, + "step": 8463 + }, + { + "epoch": 0.4454736842105263, + "grad_norm": 0.50390625, + "learning_rate": 0.00012322807017543861, + "loss": 3.3059, + "step": 8464 + }, + { + "epoch": 0.44552631578947366, + "grad_norm": 0.498046875, + "learning_rate": 0.00012321637426900586, + "loss": 3.3433, + "step": 8465 + }, + { + "epoch": 0.4455789473684211, + "grad_norm": 0.474609375, + "learning_rate": 0.0001232046783625731, + "loss": 4.0301, + "step": 8466 + }, + { + "epoch": 0.44563157894736843, + "grad_norm": 0.5234375, + "learning_rate": 0.00012319298245614034, + "loss": 3.0649, + "step": 8467 + }, + { + "epoch": 0.4456842105263158, + "grad_norm": 0.45703125, + "learning_rate": 0.00012318128654970759, + "loss": 3.4848, + "step": 8468 + }, + { + "epoch": 0.44573684210526315, + "grad_norm": 0.486328125, + "learning_rate": 0.00012316959064327486, + "loss": 3.5968, + "step": 8469 + }, + { + "epoch": 0.4457894736842105, + "grad_norm": 0.462890625, + "learning_rate": 0.0001231578947368421, + "loss": 3.3988, + "step": 8470 + }, + { + "epoch": 0.4458421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.00012314619883040936, + "loss": 3.3592, + "step": 8471 + }, + { + "epoch": 0.4458947368421053, + "grad_norm": 0.470703125, + "learning_rate": 0.0001231345029239766, + "loss": 3.7709, + "step": 8472 + }, + { + "epoch": 0.44594736842105265, + "grad_norm": 0.53125, + "learning_rate": 0.00012312280701754386, + "loss": 3.1984, + "step": 8473 + }, + { + "epoch": 0.446, + "grad_norm": 0.484375, + "learning_rate": 0.0001231111111111111, + "loss": 3.3384, + "step": 8474 + }, + { + "epoch": 0.44605263157894737, + "grad_norm": 0.45703125, + "learning_rate": 0.00012309941520467836, + "loss": 3.5261, + "step": 8475 + }, + { + "epoch": 0.4461052631578947, + "grad_norm": 0.69921875, + "learning_rate": 0.00012308771929824564, + "loss": 3.2348, + "step": 8476 + }, + { + "epoch": 0.4461578947368421, + "grad_norm": 0.498046875, + "learning_rate": 0.00012307602339181289, + "loss": 3.5171, + "step": 8477 + }, + { + "epoch": 0.44621052631578945, + "grad_norm": 0.53515625, + "learning_rate": 0.00012306432748538014, + "loss": 3.3493, + "step": 8478 + }, + { + "epoch": 0.44626315789473686, + "grad_norm": 0.50390625, + "learning_rate": 0.00012305263157894739, + "loss": 3.28, + "step": 8479 + }, + { + "epoch": 0.4463157894736842, + "grad_norm": 0.62890625, + "learning_rate": 0.00012304093567251463, + "loss": 3.5916, + "step": 8480 + }, + { + "epoch": 0.4463684210526316, + "grad_norm": 0.490234375, + "learning_rate": 0.00012302923976608188, + "loss": 2.9345, + "step": 8481 + }, + { + "epoch": 0.44642105263157894, + "grad_norm": 0.486328125, + "learning_rate": 0.00012301754385964913, + "loss": 3.2957, + "step": 8482 + }, + { + "epoch": 0.4464736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00012300584795321638, + "loss": 3.3744, + "step": 8483 + }, + { + "epoch": 0.44652631578947366, + "grad_norm": 0.50390625, + "learning_rate": 0.00012299415204678363, + "loss": 3.7157, + "step": 8484 + }, + { + "epoch": 0.4465789473684211, + "grad_norm": 0.48828125, + "learning_rate": 0.00012298245614035088, + "loss": 3.3044, + "step": 8485 + }, + { + "epoch": 0.44663157894736844, + "grad_norm": 0.55078125, + "learning_rate": 0.00012297076023391813, + "loss": 3.8432, + "step": 8486 + }, + { + "epoch": 0.4466842105263158, + "grad_norm": 0.484375, + "learning_rate": 0.00012295906432748538, + "loss": 3.4907, + "step": 8487 + }, + { + "epoch": 0.44673684210526315, + "grad_norm": 0.458984375, + "learning_rate": 0.00012294736842105263, + "loss": 3.2662, + "step": 8488 + }, + { + "epoch": 0.4467894736842105, + "grad_norm": 0.494140625, + "learning_rate": 0.00012293567251461988, + "loss": 3.3058, + "step": 8489 + }, + { + "epoch": 0.4468421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00012292397660818713, + "loss": 3.2541, + "step": 8490 + }, + { + "epoch": 0.4468947368421053, + "grad_norm": 0.5859375, + "learning_rate": 0.0001229122807017544, + "loss": 3.2581, + "step": 8491 + }, + { + "epoch": 0.44694736842105265, + "grad_norm": 0.66015625, + "learning_rate": 0.00012290058479532166, + "loss": 3.3925, + "step": 8492 + }, + { + "epoch": 0.447, + "grad_norm": 0.486328125, + "learning_rate": 0.0001228888888888889, + "loss": 3.3548, + "step": 8493 + }, + { + "epoch": 0.44705263157894737, + "grad_norm": 0.61328125, + "learning_rate": 0.00012287719298245616, + "loss": 3.7358, + "step": 8494 + }, + { + "epoch": 0.44710526315789473, + "grad_norm": 0.458984375, + "learning_rate": 0.0001228654970760234, + "loss": 3.9805, + "step": 8495 + }, + { + "epoch": 0.4471578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00012285380116959066, + "loss": 3.1351, + "step": 8496 + }, + { + "epoch": 0.44721052631578945, + "grad_norm": 0.515625, + "learning_rate": 0.0001228421052631579, + "loss": 4.0365, + "step": 8497 + }, + { + "epoch": 0.44726315789473686, + "grad_norm": 0.498046875, + "learning_rate": 0.00012283040935672515, + "loss": 3.5274, + "step": 8498 + }, + { + "epoch": 0.4473157894736842, + "grad_norm": 0.486328125, + "learning_rate": 0.0001228187134502924, + "loss": 3.5002, + "step": 8499 + }, + { + "epoch": 0.4473684210526316, + "grad_norm": 0.48828125, + "learning_rate": 0.00012280701754385965, + "loss": 3.5748, + "step": 8500 + }, + { + "epoch": 0.44742105263157894, + "grad_norm": 0.50390625, + "learning_rate": 0.0001227953216374269, + "loss": 3.1759, + "step": 8501 + }, + { + "epoch": 0.4474736842105263, + "grad_norm": 0.490234375, + "learning_rate": 0.00012278362573099415, + "loss": 3.6967, + "step": 8502 + }, + { + "epoch": 0.44752631578947366, + "grad_norm": 0.4921875, + "learning_rate": 0.0001227719298245614, + "loss": 3.2035, + "step": 8503 + }, + { + "epoch": 0.4475789473684211, + "grad_norm": 0.53125, + "learning_rate": 0.00012276023391812865, + "loss": 3.1374, + "step": 8504 + }, + { + "epoch": 0.44763157894736844, + "grad_norm": 0.4609375, + "learning_rate": 0.0001227485380116959, + "loss": 3.7429, + "step": 8505 + }, + { + "epoch": 0.4476842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00012273684210526318, + "loss": 3.8086, + "step": 8506 + }, + { + "epoch": 0.44773684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00012272514619883043, + "loss": 3.9748, + "step": 8507 + }, + { + "epoch": 0.4477894736842105, + "grad_norm": 0.55078125, + "learning_rate": 0.00012271345029239768, + "loss": 3.3508, + "step": 8508 + }, + { + "epoch": 0.4478421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00012270175438596493, + "loss": 3.7583, + "step": 8509 + }, + { + "epoch": 0.4478947368421053, + "grad_norm": 0.83203125, + "learning_rate": 0.00012269005847953218, + "loss": 4.0279, + "step": 8510 + }, + { + "epoch": 0.44794736842105265, + "grad_norm": 0.578125, + "learning_rate": 0.00012267836257309943, + "loss": 3.5021, + "step": 8511 + }, + { + "epoch": 0.448, + "grad_norm": 0.6796875, + "learning_rate": 0.00012266666666666668, + "loss": 3.6979, + "step": 8512 + }, + { + "epoch": 0.44805263157894737, + "grad_norm": 0.49609375, + "learning_rate": 0.00012265497076023393, + "loss": 3.7453, + "step": 8513 + }, + { + "epoch": 0.44810526315789473, + "grad_norm": 0.5, + "learning_rate": 0.00012264327485380117, + "loss": 3.7363, + "step": 8514 + }, + { + "epoch": 0.4481578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.00012263157894736842, + "loss": 3.904, + "step": 8515 + }, + { + "epoch": 0.44821052631578945, + "grad_norm": 0.48828125, + "learning_rate": 0.00012261988304093567, + "loss": 3.1875, + "step": 8516 + }, + { + "epoch": 0.44826315789473686, + "grad_norm": 0.498046875, + "learning_rate": 0.00012260818713450292, + "loss": 3.3884, + "step": 8517 + }, + { + "epoch": 0.4483157894736842, + "grad_norm": 0.455078125, + "learning_rate": 0.00012259649122807017, + "loss": 3.0508, + "step": 8518 + }, + { + "epoch": 0.4483684210526316, + "grad_norm": 0.466796875, + "learning_rate": 0.00012258479532163742, + "loss": 3.3892, + "step": 8519 + }, + { + "epoch": 0.44842105263157894, + "grad_norm": 0.490234375, + "learning_rate": 0.0001225730994152047, + "loss": 3.1881, + "step": 8520 + }, + { + "epoch": 0.4484736842105263, + "grad_norm": 0.458984375, + "learning_rate": 0.00012256140350877195, + "loss": 3.4614, + "step": 8521 + }, + { + "epoch": 0.44852631578947366, + "grad_norm": 0.5078125, + "learning_rate": 0.0001225497076023392, + "loss": 3.164, + "step": 8522 + }, + { + "epoch": 0.4485789473684211, + "grad_norm": 0.51171875, + "learning_rate": 0.00012253801169590645, + "loss": 3.5084, + "step": 8523 + }, + { + "epoch": 0.44863157894736844, + "grad_norm": 0.462890625, + "learning_rate": 0.0001225263157894737, + "loss": 3.7309, + "step": 8524 + }, + { + "epoch": 0.4486842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.00012251461988304095, + "loss": 3.3461, + "step": 8525 + }, + { + "epoch": 0.44873684210526316, + "grad_norm": 0.474609375, + "learning_rate": 0.0001225029239766082, + "loss": 3.62, + "step": 8526 + }, + { + "epoch": 0.4487894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.00012249122807017545, + "loss": 3.7132, + "step": 8527 + }, + { + "epoch": 0.4488421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.0001224795321637427, + "loss": 3.8261, + "step": 8528 + }, + { + "epoch": 0.4488947368421053, + "grad_norm": 0.53515625, + "learning_rate": 0.00012246783625730995, + "loss": 3.3866, + "step": 8529 + }, + { + "epoch": 0.44894736842105265, + "grad_norm": 0.498046875, + "learning_rate": 0.0001224561403508772, + "loss": 3.2012, + "step": 8530 + }, + { + "epoch": 0.449, + "grad_norm": 0.515625, + "learning_rate": 0.00012244444444444445, + "loss": 3.5831, + "step": 8531 + }, + { + "epoch": 0.44905263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.0001224327485380117, + "loss": 3.4322, + "step": 8532 + }, + { + "epoch": 0.44910526315789473, + "grad_norm": 0.470703125, + "learning_rate": 0.00012242105263157894, + "loss": 2.9532, + "step": 8533 + }, + { + "epoch": 0.4491578947368421, + "grad_norm": 0.455078125, + "learning_rate": 0.0001224093567251462, + "loss": 3.6361, + "step": 8534 + }, + { + "epoch": 0.44921052631578945, + "grad_norm": 0.47265625, + "learning_rate": 0.00012239766081871347, + "loss": 3.4889, + "step": 8535 + }, + { + "epoch": 0.44926315789473686, + "grad_norm": 0.494140625, + "learning_rate": 0.00012238596491228072, + "loss": 2.9737, + "step": 8536 + }, + { + "epoch": 0.4493157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00012237426900584797, + "loss": 3.4721, + "step": 8537 + }, + { + "epoch": 0.4493684210526316, + "grad_norm": 0.52734375, + "learning_rate": 0.00012236257309941522, + "loss": 3.803, + "step": 8538 + }, + { + "epoch": 0.44942105263157894, + "grad_norm": 0.55859375, + "learning_rate": 0.00012235087719298247, + "loss": 2.988, + "step": 8539 + }, + { + "epoch": 0.4494736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00012233918128654972, + "loss": 4.1801, + "step": 8540 + }, + { + "epoch": 0.44952631578947366, + "grad_norm": 0.53125, + "learning_rate": 0.00012232748538011697, + "loss": 3.6139, + "step": 8541 + }, + { + "epoch": 0.4495789473684211, + "grad_norm": 0.5234375, + "learning_rate": 0.00012231578947368422, + "loss": 3.4385, + "step": 8542 + }, + { + "epoch": 0.44963157894736844, + "grad_norm": 0.486328125, + "learning_rate": 0.00012230409356725147, + "loss": 3.523, + "step": 8543 + }, + { + "epoch": 0.4496842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00012229239766081872, + "loss": 3.5063, + "step": 8544 + }, + { + "epoch": 0.44973684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00012228070175438597, + "loss": 3.2443, + "step": 8545 + }, + { + "epoch": 0.4497894736842105, + "grad_norm": 0.482421875, + "learning_rate": 0.00012226900584795322, + "loss": 3.4116, + "step": 8546 + }, + { + "epoch": 0.4498421052631579, + "grad_norm": 0.609375, + "learning_rate": 0.00012225730994152047, + "loss": 3.5038, + "step": 8547 + }, + { + "epoch": 0.44989473684210524, + "grad_norm": 0.48828125, + "learning_rate": 0.00012224561403508772, + "loss": 3.3997, + "step": 8548 + }, + { + "epoch": 0.44994736842105265, + "grad_norm": 0.53125, + "learning_rate": 0.00012223391812865496, + "loss": 3.0576, + "step": 8549 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.00012222222222222224, + "loss": 4.0478, + "step": 8550 + }, + { + "epoch": 0.45005263157894737, + "grad_norm": 0.462890625, + "learning_rate": 0.0001222105263157895, + "loss": 3.1803, + "step": 8551 + }, + { + "epoch": 0.45010526315789473, + "grad_norm": 0.61328125, + "learning_rate": 0.00012219883040935674, + "loss": 3.298, + "step": 8552 + }, + { + "epoch": 0.4501578947368421, + "grad_norm": 0.5, + "learning_rate": 0.000122187134502924, + "loss": 3.9415, + "step": 8553 + }, + { + "epoch": 0.45021052631578945, + "grad_norm": 0.50390625, + "learning_rate": 0.00012217543859649124, + "loss": 3.5339, + "step": 8554 + }, + { + "epoch": 0.45026315789473687, + "grad_norm": 0.55859375, + "learning_rate": 0.0001221637426900585, + "loss": 3.3355, + "step": 8555 + }, + { + "epoch": 0.4503157894736842, + "grad_norm": 0.4765625, + "learning_rate": 0.00012215204678362574, + "loss": 3.5049, + "step": 8556 + }, + { + "epoch": 0.4503684210526316, + "grad_norm": 0.51171875, + "learning_rate": 0.000122140350877193, + "loss": 3.4563, + "step": 8557 + }, + { + "epoch": 0.45042105263157894, + "grad_norm": 0.4921875, + "learning_rate": 0.00012212865497076024, + "loss": 3.8057, + "step": 8558 + }, + { + "epoch": 0.4504736842105263, + "grad_norm": 0.55078125, + "learning_rate": 0.0001221169590643275, + "loss": 3.8213, + "step": 8559 + }, + { + "epoch": 0.45052631578947366, + "grad_norm": 0.55859375, + "learning_rate": 0.00012210526315789474, + "loss": 3.6871, + "step": 8560 + }, + { + "epoch": 0.4505789473684211, + "grad_norm": 0.60546875, + "learning_rate": 0.000122093567251462, + "loss": 3.3548, + "step": 8561 + }, + { + "epoch": 0.45063157894736844, + "grad_norm": 0.490234375, + "learning_rate": 0.00012208187134502924, + "loss": 3.1489, + "step": 8562 + }, + { + "epoch": 0.4506842105263158, + "grad_norm": 0.484375, + "learning_rate": 0.00012207017543859649, + "loss": 3.5015, + "step": 8563 + }, + { + "epoch": 0.45073684210526316, + "grad_norm": 0.498046875, + "learning_rate": 0.00012205847953216374, + "loss": 3.6078, + "step": 8564 + }, + { + "epoch": 0.4507894736842105, + "grad_norm": 0.484375, + "learning_rate": 0.00012204678362573101, + "loss": 3.0369, + "step": 8565 + }, + { + "epoch": 0.4508421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00012203508771929826, + "loss": 3.3968, + "step": 8566 + }, + { + "epoch": 0.45089473684210524, + "grad_norm": 0.482421875, + "learning_rate": 0.00012202339181286551, + "loss": 3.561, + "step": 8567 + }, + { + "epoch": 0.45094736842105265, + "grad_norm": 0.54296875, + "learning_rate": 0.00012201169590643276, + "loss": 3.8178, + "step": 8568 + }, + { + "epoch": 0.451, + "grad_norm": 0.515625, + "learning_rate": 0.000122, + "loss": 3.3035, + "step": 8569 + }, + { + "epoch": 0.45105263157894737, + "grad_norm": 0.546875, + "learning_rate": 0.00012198830409356725, + "loss": 3.8271, + "step": 8570 + }, + { + "epoch": 0.45110526315789473, + "grad_norm": 0.51171875, + "learning_rate": 0.0001219766081871345, + "loss": 3.9032, + "step": 8571 + }, + { + "epoch": 0.4511578947368421, + "grad_norm": 0.58984375, + "learning_rate": 0.00012196491228070177, + "loss": 3.6589, + "step": 8572 + }, + { + "epoch": 0.45121052631578945, + "grad_norm": 0.6484375, + "learning_rate": 0.00012195321637426902, + "loss": 3.669, + "step": 8573 + }, + { + "epoch": 0.45126315789473687, + "grad_norm": 0.47265625, + "learning_rate": 0.00012194152046783627, + "loss": 3.6077, + "step": 8574 + }, + { + "epoch": 0.4513157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00012192982456140352, + "loss": 3.321, + "step": 8575 + }, + { + "epoch": 0.4513684210526316, + "grad_norm": 0.490234375, + "learning_rate": 0.00012191812865497076, + "loss": 3.3509, + "step": 8576 + }, + { + "epoch": 0.45142105263157895, + "grad_norm": 0.443359375, + "learning_rate": 0.00012190643274853801, + "loss": 3.2329, + "step": 8577 + }, + { + "epoch": 0.4514736842105263, + "grad_norm": 0.466796875, + "learning_rate": 0.00012189473684210526, + "loss": 3.4975, + "step": 8578 + }, + { + "epoch": 0.45152631578947366, + "grad_norm": 0.4375, + "learning_rate": 0.00012188304093567251, + "loss": 3.52, + "step": 8579 + }, + { + "epoch": 0.4515789473684211, + "grad_norm": 0.498046875, + "learning_rate": 0.00012187134502923978, + "loss": 3.3701, + "step": 8580 + }, + { + "epoch": 0.45163157894736844, + "grad_norm": 0.48046875, + "learning_rate": 0.00012185964912280703, + "loss": 3.215, + "step": 8581 + }, + { + "epoch": 0.4516842105263158, + "grad_norm": 0.46484375, + "learning_rate": 0.00012184795321637428, + "loss": 3.3133, + "step": 8582 + }, + { + "epoch": 0.45173684210526316, + "grad_norm": 0.51953125, + "learning_rate": 0.00012183625730994153, + "loss": 3.619, + "step": 8583 + }, + { + "epoch": 0.4517894736842105, + "grad_norm": 0.486328125, + "learning_rate": 0.00012182456140350877, + "loss": 3.3472, + "step": 8584 + }, + { + "epoch": 0.4518421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00012181286549707602, + "loss": 3.3323, + "step": 8585 + }, + { + "epoch": 0.45189473684210524, + "grad_norm": 0.482421875, + "learning_rate": 0.00012180116959064327, + "loss": 3.7442, + "step": 8586 + }, + { + "epoch": 0.45194736842105265, + "grad_norm": 0.51171875, + "learning_rate": 0.00012178947368421054, + "loss": 3.3387, + "step": 8587 + }, + { + "epoch": 0.452, + "grad_norm": 0.4375, + "learning_rate": 0.0001217777777777778, + "loss": 3.4321, + "step": 8588 + }, + { + "epoch": 0.4520526315789474, + "grad_norm": 0.46875, + "learning_rate": 0.00012176608187134504, + "loss": 3.5754, + "step": 8589 + }, + { + "epoch": 0.45210526315789473, + "grad_norm": 0.48828125, + "learning_rate": 0.00012175438596491229, + "loss": 3.8327, + "step": 8590 + }, + { + "epoch": 0.4521578947368421, + "grad_norm": 0.4765625, + "learning_rate": 0.00012174269005847954, + "loss": 2.9483, + "step": 8591 + }, + { + "epoch": 0.45221052631578945, + "grad_norm": 0.5, + "learning_rate": 0.00012173099415204678, + "loss": 3.5438, + "step": 8592 + }, + { + "epoch": 0.45226315789473687, + "grad_norm": 0.52734375, + "learning_rate": 0.00012171929824561403, + "loss": 3.421, + "step": 8593 + }, + { + "epoch": 0.4523157894736842, + "grad_norm": 0.490234375, + "learning_rate": 0.00012170760233918128, + "loss": 3.8706, + "step": 8594 + }, + { + "epoch": 0.4523684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.00012169590643274855, + "loss": 3.2647, + "step": 8595 + }, + { + "epoch": 0.45242105263157895, + "grad_norm": 0.5859375, + "learning_rate": 0.0001216842105263158, + "loss": 3.3549, + "step": 8596 + }, + { + "epoch": 0.4524736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00012167251461988305, + "loss": 3.3672, + "step": 8597 + }, + { + "epoch": 0.45252631578947367, + "grad_norm": 0.49609375, + "learning_rate": 0.0001216608187134503, + "loss": 3.3718, + "step": 8598 + }, + { + "epoch": 0.452578947368421, + "grad_norm": 0.5234375, + "learning_rate": 0.00012164912280701754, + "loss": 3.4061, + "step": 8599 + }, + { + "epoch": 0.45263157894736844, + "grad_norm": 0.50390625, + "learning_rate": 0.00012163742690058479, + "loss": 3.2345, + "step": 8600 + }, + { + "epoch": 0.4526842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.00012162573099415204, + "loss": 3.6949, + "step": 8601 + }, + { + "epoch": 0.45273684210526316, + "grad_norm": 0.48046875, + "learning_rate": 0.00012161403508771932, + "loss": 3.523, + "step": 8602 + }, + { + "epoch": 0.4527894736842105, + "grad_norm": 0.5, + "learning_rate": 0.00012160233918128656, + "loss": 3.5746, + "step": 8603 + }, + { + "epoch": 0.4528421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00012159064327485381, + "loss": 3.495, + "step": 8604 + }, + { + "epoch": 0.45289473684210524, + "grad_norm": 0.498046875, + "learning_rate": 0.00012157894736842106, + "loss": 3.1461, + "step": 8605 + }, + { + "epoch": 0.45294736842105265, + "grad_norm": 0.498046875, + "learning_rate": 0.00012156725146198831, + "loss": 3.8201, + "step": 8606 + }, + { + "epoch": 0.453, + "grad_norm": 0.53125, + "learning_rate": 0.00012155555555555555, + "loss": 3.6568, + "step": 8607 + }, + { + "epoch": 0.4530526315789474, + "grad_norm": 0.5546875, + "learning_rate": 0.0001215438596491228, + "loss": 3.5761, + "step": 8608 + }, + { + "epoch": 0.45310526315789473, + "grad_norm": 0.7109375, + "learning_rate": 0.00012153216374269008, + "loss": 3.3383, + "step": 8609 + }, + { + "epoch": 0.4531578947368421, + "grad_norm": 0.474609375, + "learning_rate": 0.00012152046783625733, + "loss": 3.3378, + "step": 8610 + }, + { + "epoch": 0.45321052631578945, + "grad_norm": 0.484375, + "learning_rate": 0.00012150877192982458, + "loss": 3.5631, + "step": 8611 + }, + { + "epoch": 0.45326315789473687, + "grad_norm": 0.5234375, + "learning_rate": 0.00012149707602339182, + "loss": 3.5588, + "step": 8612 + }, + { + "epoch": 0.4533157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00012148538011695907, + "loss": 3.457, + "step": 8613 + }, + { + "epoch": 0.4533684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00012147368421052632, + "loss": 3.3171, + "step": 8614 + }, + { + "epoch": 0.45342105263157895, + "grad_norm": 0.47265625, + "learning_rate": 0.00012146198830409356, + "loss": 3.6058, + "step": 8615 + }, + { + "epoch": 0.4534736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.00012145029239766081, + "loss": 3.5909, + "step": 8616 + }, + { + "epoch": 0.45352631578947367, + "grad_norm": 0.52734375, + "learning_rate": 0.00012143859649122809, + "loss": 3.7993, + "step": 8617 + }, + { + "epoch": 0.453578947368421, + "grad_norm": 0.498046875, + "learning_rate": 0.00012142690058479534, + "loss": 3.2158, + "step": 8618 + }, + { + "epoch": 0.45363157894736844, + "grad_norm": 0.4921875, + "learning_rate": 0.00012141520467836259, + "loss": 3.5238, + "step": 8619 + }, + { + "epoch": 0.4536842105263158, + "grad_norm": 0.5625, + "learning_rate": 0.00012140350877192984, + "loss": 3.0177, + "step": 8620 + }, + { + "epoch": 0.45373684210526316, + "grad_norm": 0.4921875, + "learning_rate": 0.00012139181286549708, + "loss": 3.931, + "step": 8621 + }, + { + "epoch": 0.4537894736842105, + "grad_norm": 0.53515625, + "learning_rate": 0.00012138011695906432, + "loss": 3.4429, + "step": 8622 + }, + { + "epoch": 0.4538421052631579, + "grad_norm": 0.47265625, + "learning_rate": 0.00012136842105263157, + "loss": 3.4707, + "step": 8623 + }, + { + "epoch": 0.45389473684210524, + "grad_norm": 0.49609375, + "learning_rate": 0.00012135672514619885, + "loss": 3.2468, + "step": 8624 + }, + { + "epoch": 0.45394736842105265, + "grad_norm": 0.50390625, + "learning_rate": 0.0001213450292397661, + "loss": 3.8801, + "step": 8625 + }, + { + "epoch": 0.454, + "grad_norm": 0.498046875, + "learning_rate": 0.00012133333333333335, + "loss": 3.5546, + "step": 8626 + }, + { + "epoch": 0.4540526315789474, + "grad_norm": 0.73046875, + "learning_rate": 0.0001213216374269006, + "loss": 3.1608, + "step": 8627 + }, + { + "epoch": 0.45410526315789473, + "grad_norm": 0.482421875, + "learning_rate": 0.00012130994152046785, + "loss": 3.3369, + "step": 8628 + }, + { + "epoch": 0.4541578947368421, + "grad_norm": 0.466796875, + "learning_rate": 0.0001212982456140351, + "loss": 3.7207, + "step": 8629 + }, + { + "epoch": 0.45421052631578945, + "grad_norm": 0.46875, + "learning_rate": 0.00012128654970760233, + "loss": 3.7659, + "step": 8630 + }, + { + "epoch": 0.45426315789473687, + "grad_norm": 0.49609375, + "learning_rate": 0.00012127485380116958, + "loss": 3.5499, + "step": 8631 + }, + { + "epoch": 0.45431578947368423, + "grad_norm": 0.5078125, + "learning_rate": 0.00012126315789473686, + "loss": 3.4798, + "step": 8632 + }, + { + "epoch": 0.4543684210526316, + "grad_norm": 0.5390625, + "learning_rate": 0.00012125146198830411, + "loss": 3.4047, + "step": 8633 + }, + { + "epoch": 0.45442105263157895, + "grad_norm": 0.48046875, + "learning_rate": 0.00012123976608187136, + "loss": 3.0359, + "step": 8634 + }, + { + "epoch": 0.4544736842105263, + "grad_norm": 0.5234375, + "learning_rate": 0.0001212280701754386, + "loss": 3.2505, + "step": 8635 + }, + { + "epoch": 0.45452631578947367, + "grad_norm": 0.474609375, + "learning_rate": 0.00012121637426900586, + "loss": 3.6084, + "step": 8636 + }, + { + "epoch": 0.454578947368421, + "grad_norm": 0.462890625, + "learning_rate": 0.0001212046783625731, + "loss": 3.7054, + "step": 8637 + }, + { + "epoch": 0.45463157894736844, + "grad_norm": 0.53125, + "learning_rate": 0.00012119298245614034, + "loss": 3.3982, + "step": 8638 + }, + { + "epoch": 0.4546842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.00012118128654970762, + "loss": 3.6078, + "step": 8639 + }, + { + "epoch": 0.45473684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00012116959064327487, + "loss": 3.4202, + "step": 8640 + }, + { + "epoch": 0.4547894736842105, + "grad_norm": 0.47265625, + "learning_rate": 0.00012115789473684212, + "loss": 4.0547, + "step": 8641 + }, + { + "epoch": 0.4548421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00012114619883040937, + "loss": 3.5141, + "step": 8642 + }, + { + "epoch": 0.45489473684210524, + "grad_norm": 0.51953125, + "learning_rate": 0.00012113450292397662, + "loss": 3.1454, + "step": 8643 + }, + { + "epoch": 0.45494736842105266, + "grad_norm": 0.486328125, + "learning_rate": 0.00012112280701754387, + "loss": 2.9658, + "step": 8644 + }, + { + "epoch": 0.455, + "grad_norm": 0.5, + "learning_rate": 0.0001211111111111111, + "loss": 3.8438, + "step": 8645 + }, + { + "epoch": 0.4550526315789474, + "grad_norm": 0.484375, + "learning_rate": 0.00012109941520467838, + "loss": 3.4607, + "step": 8646 + }, + { + "epoch": 0.45510526315789473, + "grad_norm": 0.4609375, + "learning_rate": 0.00012108771929824563, + "loss": 3.13, + "step": 8647 + }, + { + "epoch": 0.4551578947368421, + "grad_norm": 0.490234375, + "learning_rate": 0.00012107602339181288, + "loss": 3.9521, + "step": 8648 + }, + { + "epoch": 0.45521052631578945, + "grad_norm": 0.57421875, + "learning_rate": 0.00012106432748538013, + "loss": 3.7738, + "step": 8649 + }, + { + "epoch": 0.45526315789473687, + "grad_norm": 0.5078125, + "learning_rate": 0.00012105263157894738, + "loss": 2.9891, + "step": 8650 + }, + { + "epoch": 0.45531578947368423, + "grad_norm": 0.484375, + "learning_rate": 0.00012104093567251463, + "loss": 3.5566, + "step": 8651 + }, + { + "epoch": 0.4553684210526316, + "grad_norm": 0.470703125, + "learning_rate": 0.00012102923976608188, + "loss": 3.962, + "step": 8652 + }, + { + "epoch": 0.45542105263157895, + "grad_norm": 0.494140625, + "learning_rate": 0.00012101754385964911, + "loss": 3.2744, + "step": 8653 + }, + { + "epoch": 0.4554736842105263, + "grad_norm": 0.484375, + "learning_rate": 0.00012100584795321639, + "loss": 3.7489, + "step": 8654 + }, + { + "epoch": 0.45552631578947367, + "grad_norm": 0.494140625, + "learning_rate": 0.00012099415204678364, + "loss": 3.2718, + "step": 8655 + }, + { + "epoch": 0.45557894736842103, + "grad_norm": 0.49609375, + "learning_rate": 0.00012098245614035089, + "loss": 3.47, + "step": 8656 + }, + { + "epoch": 0.45563157894736844, + "grad_norm": 0.52734375, + "learning_rate": 0.00012097076023391814, + "loss": 3.4762, + "step": 8657 + }, + { + "epoch": 0.4556842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00012095906432748539, + "loss": 3.454, + "step": 8658 + }, + { + "epoch": 0.45573684210526316, + "grad_norm": 0.478515625, + "learning_rate": 0.00012094736842105264, + "loss": 3.467, + "step": 8659 + }, + { + "epoch": 0.4557894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.00012093567251461989, + "loss": 3.1877, + "step": 8660 + }, + { + "epoch": 0.4558421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00012092397660818715, + "loss": 3.4938, + "step": 8661 + }, + { + "epoch": 0.45589473684210524, + "grad_norm": 0.486328125, + "learning_rate": 0.0001209122807017544, + "loss": 3.3134, + "step": 8662 + }, + { + "epoch": 0.45594736842105266, + "grad_norm": 0.498046875, + "learning_rate": 0.00012090058479532165, + "loss": 3.1692, + "step": 8663 + }, + { + "epoch": 0.456, + "grad_norm": 0.478515625, + "learning_rate": 0.0001208888888888889, + "loss": 3.4709, + "step": 8664 + }, + { + "epoch": 0.4560526315789474, + "grad_norm": 0.498046875, + "learning_rate": 0.00012087719298245615, + "loss": 3.5238, + "step": 8665 + }, + { + "epoch": 0.45610526315789474, + "grad_norm": 0.55859375, + "learning_rate": 0.0001208654970760234, + "loss": 3.077, + "step": 8666 + }, + { + "epoch": 0.4561578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00012085380116959065, + "loss": 3.296, + "step": 8667 + }, + { + "epoch": 0.45621052631578946, + "grad_norm": 0.5703125, + "learning_rate": 0.00012084210526315788, + "loss": 3.6823, + "step": 8668 + }, + { + "epoch": 0.4562631578947368, + "grad_norm": 0.5390625, + "learning_rate": 0.00012083040935672516, + "loss": 3.7395, + "step": 8669 + }, + { + "epoch": 0.45631578947368423, + "grad_norm": 0.5390625, + "learning_rate": 0.00012081871345029241, + "loss": 3.6099, + "step": 8670 + }, + { + "epoch": 0.4563684210526316, + "grad_norm": 0.51171875, + "learning_rate": 0.00012080701754385966, + "loss": 3.6253, + "step": 8671 + }, + { + "epoch": 0.45642105263157895, + "grad_norm": 0.625, + "learning_rate": 0.00012079532163742691, + "loss": 3.6073, + "step": 8672 + }, + { + "epoch": 0.4564736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.00012078362573099416, + "loss": 3.4173, + "step": 8673 + }, + { + "epoch": 0.45652631578947367, + "grad_norm": 0.5234375, + "learning_rate": 0.00012077192982456141, + "loss": 3.6317, + "step": 8674 + }, + { + "epoch": 0.45657894736842103, + "grad_norm": 0.498046875, + "learning_rate": 0.00012076023391812866, + "loss": 3.6639, + "step": 8675 + }, + { + "epoch": 0.45663157894736844, + "grad_norm": 0.546875, + "learning_rate": 0.00012074853801169592, + "loss": 3.5808, + "step": 8676 + }, + { + "epoch": 0.4566842105263158, + "grad_norm": 0.47265625, + "learning_rate": 0.00012073684210526317, + "loss": 3.5669, + "step": 8677 + }, + { + "epoch": 0.45673684210526316, + "grad_norm": 0.4765625, + "learning_rate": 0.00012072514619883042, + "loss": 3.6038, + "step": 8678 + }, + { + "epoch": 0.4567894736842105, + "grad_norm": 0.4609375, + "learning_rate": 0.00012071345029239767, + "loss": 3.5775, + "step": 8679 + }, + { + "epoch": 0.4568421052631579, + "grad_norm": 0.75, + "learning_rate": 0.00012070175438596492, + "loss": 3.7623, + "step": 8680 + }, + { + "epoch": 0.45689473684210524, + "grad_norm": 0.498046875, + "learning_rate": 0.00012069005847953217, + "loss": 3.1845, + "step": 8681 + }, + { + "epoch": 0.45694736842105266, + "grad_norm": 0.546875, + "learning_rate": 0.00012067836257309942, + "loss": 3.2933, + "step": 8682 + }, + { + "epoch": 0.457, + "grad_norm": 0.494140625, + "learning_rate": 0.00012066666666666668, + "loss": 3.7034, + "step": 8683 + }, + { + "epoch": 0.4570526315789474, + "grad_norm": 0.486328125, + "learning_rate": 0.00012065497076023393, + "loss": 3.6919, + "step": 8684 + }, + { + "epoch": 0.45710526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.00012064327485380118, + "loss": 3.8393, + "step": 8685 + }, + { + "epoch": 0.4571578947368421, + "grad_norm": 0.54296875, + "learning_rate": 0.00012063157894736843, + "loss": 3.2883, + "step": 8686 + }, + { + "epoch": 0.45721052631578946, + "grad_norm": 0.50390625, + "learning_rate": 0.00012061988304093568, + "loss": 4.0042, + "step": 8687 + }, + { + "epoch": 0.4572631578947368, + "grad_norm": 0.53125, + "learning_rate": 0.00012060818713450293, + "loss": 3.3146, + "step": 8688 + }, + { + "epoch": 0.45731578947368423, + "grad_norm": 0.50390625, + "learning_rate": 0.00012059649122807018, + "loss": 3.8464, + "step": 8689 + }, + { + "epoch": 0.4573684210526316, + "grad_norm": 0.5, + "learning_rate": 0.00012058479532163743, + "loss": 3.2089, + "step": 8690 + }, + { + "epoch": 0.45742105263157895, + "grad_norm": 0.482421875, + "learning_rate": 0.00012057309941520469, + "loss": 3.4336, + "step": 8691 + }, + { + "epoch": 0.4574736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00012056140350877194, + "loss": 3.596, + "step": 8692 + }, + { + "epoch": 0.45752631578947367, + "grad_norm": 0.486328125, + "learning_rate": 0.00012054970760233919, + "loss": 3.1902, + "step": 8693 + }, + { + "epoch": 0.45757894736842103, + "grad_norm": 1.09375, + "learning_rate": 0.00012053801169590644, + "loss": 3.341, + "step": 8694 + }, + { + "epoch": 0.45763157894736844, + "grad_norm": 0.45703125, + "learning_rate": 0.00012052631578947369, + "loss": 3.3951, + "step": 8695 + }, + { + "epoch": 0.4576842105263158, + "grad_norm": 0.52734375, + "learning_rate": 0.00012051461988304094, + "loss": 3.697, + "step": 8696 + }, + { + "epoch": 0.45773684210526316, + "grad_norm": 0.5, + "learning_rate": 0.00012050292397660819, + "loss": 3.5715, + "step": 8697 + }, + { + "epoch": 0.4577894736842105, + "grad_norm": 0.5, + "learning_rate": 0.00012049122807017545, + "loss": 3.8179, + "step": 8698 + }, + { + "epoch": 0.4578421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.0001204795321637427, + "loss": 3.5434, + "step": 8699 + }, + { + "epoch": 0.45789473684210524, + "grad_norm": 0.5234375, + "learning_rate": 0.00012046783625730995, + "loss": 3.9922, + "step": 8700 + }, + { + "epoch": 0.45794736842105266, + "grad_norm": 0.5078125, + "learning_rate": 0.0001204561403508772, + "loss": 3.5508, + "step": 8701 + }, + { + "epoch": 0.458, + "grad_norm": 0.470703125, + "learning_rate": 0.00012044444444444445, + "loss": 3.4036, + "step": 8702 + }, + { + "epoch": 0.4580526315789474, + "grad_norm": 0.51171875, + "learning_rate": 0.0001204327485380117, + "loss": 3.2193, + "step": 8703 + }, + { + "epoch": 0.45810526315789474, + "grad_norm": 0.48046875, + "learning_rate": 0.00012042105263157895, + "loss": 3.5659, + "step": 8704 + }, + { + "epoch": 0.4581578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.0001204093567251462, + "loss": 3.5229, + "step": 8705 + }, + { + "epoch": 0.45821052631578946, + "grad_norm": 0.5, + "learning_rate": 0.00012039766081871346, + "loss": 2.9916, + "step": 8706 + }, + { + "epoch": 0.4582631578947368, + "grad_norm": 0.54296875, + "learning_rate": 0.00012038596491228071, + "loss": 3.6924, + "step": 8707 + }, + { + "epoch": 0.45831578947368423, + "grad_norm": 0.55859375, + "learning_rate": 0.00012037426900584796, + "loss": 3.6931, + "step": 8708 + }, + { + "epoch": 0.4583684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.00012036257309941521, + "loss": 3.1638, + "step": 8709 + }, + { + "epoch": 0.45842105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00012035087719298246, + "loss": 3.4306, + "step": 8710 + }, + { + "epoch": 0.4584736842105263, + "grad_norm": 0.625, + "learning_rate": 0.00012033918128654971, + "loss": 3.5966, + "step": 8711 + }, + { + "epoch": 0.45852631578947367, + "grad_norm": 0.55078125, + "learning_rate": 0.00012032748538011696, + "loss": 3.5602, + "step": 8712 + }, + { + "epoch": 0.45857894736842103, + "grad_norm": 0.52734375, + "learning_rate": 0.00012031578947368422, + "loss": 3.6181, + "step": 8713 + }, + { + "epoch": 0.45863157894736845, + "grad_norm": 0.50390625, + "learning_rate": 0.00012030409356725147, + "loss": 3.5081, + "step": 8714 + }, + { + "epoch": 0.4586842105263158, + "grad_norm": 0.49609375, + "learning_rate": 0.00012029239766081872, + "loss": 3.6717, + "step": 8715 + }, + { + "epoch": 0.45873684210526317, + "grad_norm": 0.515625, + "learning_rate": 0.00012028070175438597, + "loss": 3.37, + "step": 8716 + }, + { + "epoch": 0.4587894736842105, + "grad_norm": 0.57421875, + "learning_rate": 0.00012026900584795322, + "loss": 3.3241, + "step": 8717 + }, + { + "epoch": 0.4588421052631579, + "grad_norm": 0.48828125, + "learning_rate": 0.00012025730994152047, + "loss": 3.8571, + "step": 8718 + }, + { + "epoch": 0.45889473684210524, + "grad_norm": 0.5859375, + "learning_rate": 0.00012024561403508772, + "loss": 3.3255, + "step": 8719 + }, + { + "epoch": 0.4589473684210526, + "grad_norm": 0.55859375, + "learning_rate": 0.00012023391812865498, + "loss": 3.7871, + "step": 8720 + }, + { + "epoch": 0.459, + "grad_norm": 0.58203125, + "learning_rate": 0.00012022222222222223, + "loss": 3.4294, + "step": 8721 + }, + { + "epoch": 0.4590526315789474, + "grad_norm": 0.55078125, + "learning_rate": 0.00012021052631578948, + "loss": 3.463, + "step": 8722 + }, + { + "epoch": 0.45910526315789474, + "grad_norm": 0.48046875, + "learning_rate": 0.00012019883040935673, + "loss": 3.5121, + "step": 8723 + }, + { + "epoch": 0.4591578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00012018713450292398, + "loss": 3.627, + "step": 8724 + }, + { + "epoch": 0.45921052631578946, + "grad_norm": 0.474609375, + "learning_rate": 0.00012017543859649123, + "loss": 3.5939, + "step": 8725 + }, + { + "epoch": 0.4592631578947368, + "grad_norm": 0.498046875, + "learning_rate": 0.00012016374269005848, + "loss": 2.9902, + "step": 8726 + }, + { + "epoch": 0.45931578947368423, + "grad_norm": 0.474609375, + "learning_rate": 0.00012015204678362573, + "loss": 3.7293, + "step": 8727 + }, + { + "epoch": 0.4593684210526316, + "grad_norm": 0.474609375, + "learning_rate": 0.000120140350877193, + "loss": 3.4288, + "step": 8728 + }, + { + "epoch": 0.45942105263157895, + "grad_norm": 0.478515625, + "learning_rate": 0.00012012865497076024, + "loss": 3.6289, + "step": 8729 + }, + { + "epoch": 0.4594736842105263, + "grad_norm": 0.474609375, + "learning_rate": 0.0001201169590643275, + "loss": 3.576, + "step": 8730 + }, + { + "epoch": 0.45952631578947367, + "grad_norm": 0.4609375, + "learning_rate": 0.00012010526315789474, + "loss": 3.6085, + "step": 8731 + }, + { + "epoch": 0.45957894736842103, + "grad_norm": 0.5390625, + "learning_rate": 0.00012009356725146199, + "loss": 3.6542, + "step": 8732 + }, + { + "epoch": 0.45963157894736845, + "grad_norm": 0.478515625, + "learning_rate": 0.00012008187134502924, + "loss": 3.1145, + "step": 8733 + }, + { + "epoch": 0.4596842105263158, + "grad_norm": 0.484375, + "learning_rate": 0.00012007017543859649, + "loss": 3.3304, + "step": 8734 + }, + { + "epoch": 0.45973684210526317, + "grad_norm": 0.5, + "learning_rate": 0.00012005847953216376, + "loss": 3.2783, + "step": 8735 + }, + { + "epoch": 0.4597894736842105, + "grad_norm": 0.494140625, + "learning_rate": 0.000120046783625731, + "loss": 3.3965, + "step": 8736 + }, + { + "epoch": 0.4598421052631579, + "grad_norm": 0.458984375, + "learning_rate": 0.00012003508771929825, + "loss": 3.4124, + "step": 8737 + }, + { + "epoch": 0.45989473684210525, + "grad_norm": 0.4609375, + "learning_rate": 0.0001200233918128655, + "loss": 3.4159, + "step": 8738 + }, + { + "epoch": 0.4599473684210526, + "grad_norm": 0.474609375, + "learning_rate": 0.00012001169590643275, + "loss": 3.4361, + "step": 8739 + }, + { + "epoch": 0.46, + "grad_norm": 0.48046875, + "learning_rate": 0.00012, + "loss": 3.5796, + "step": 8740 + }, + { + "epoch": 0.4600526315789474, + "grad_norm": 0.478515625, + "learning_rate": 0.00011998830409356725, + "loss": 3.4988, + "step": 8741 + }, + { + "epoch": 0.46010526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.0001199766081871345, + "loss": 3.5114, + "step": 8742 + }, + { + "epoch": 0.4601578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00011996491228070177, + "loss": 3.4668, + "step": 8743 + }, + { + "epoch": 0.46021052631578946, + "grad_norm": 0.51171875, + "learning_rate": 0.00011995321637426901, + "loss": 3.6374, + "step": 8744 + }, + { + "epoch": 0.4602631578947368, + "grad_norm": 0.5, + "learning_rate": 0.00011994152046783626, + "loss": 3.9259, + "step": 8745 + }, + { + "epoch": 0.46031578947368423, + "grad_norm": 0.48828125, + "learning_rate": 0.00011992982456140351, + "loss": 3.5989, + "step": 8746 + }, + { + "epoch": 0.4603684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00011991812865497076, + "loss": 3.5172, + "step": 8747 + }, + { + "epoch": 0.46042105263157895, + "grad_norm": 0.4609375, + "learning_rate": 0.00011990643274853801, + "loss": 3.6311, + "step": 8748 + }, + { + "epoch": 0.4604736842105263, + "grad_norm": 0.51171875, + "learning_rate": 0.00011989473684210526, + "loss": 3.806, + "step": 8749 + }, + { + "epoch": 0.4605263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.00011988304093567253, + "loss": 3.7395, + "step": 8750 + }, + { + "epoch": 0.46057894736842103, + "grad_norm": 0.515625, + "learning_rate": 0.00011987134502923978, + "loss": 2.9301, + "step": 8751 + }, + { + "epoch": 0.46063157894736845, + "grad_norm": 0.4921875, + "learning_rate": 0.00011985964912280703, + "loss": 3.7496, + "step": 8752 + }, + { + "epoch": 0.4606842105263158, + "grad_norm": 0.4921875, + "learning_rate": 0.00011984795321637427, + "loss": 3.3353, + "step": 8753 + }, + { + "epoch": 0.46073684210526317, + "grad_norm": 0.4765625, + "learning_rate": 0.00011983625730994152, + "loss": 3.8859, + "step": 8754 + }, + { + "epoch": 0.4607894736842105, + "grad_norm": 0.498046875, + "learning_rate": 0.00011982456140350877, + "loss": 3.4213, + "step": 8755 + }, + { + "epoch": 0.4608421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00011981286549707602, + "loss": 3.1851, + "step": 8756 + }, + { + "epoch": 0.46089473684210525, + "grad_norm": 0.478515625, + "learning_rate": 0.00011980116959064327, + "loss": 3.4781, + "step": 8757 + }, + { + "epoch": 0.4609473684210526, + "grad_norm": 0.48046875, + "learning_rate": 0.00011978947368421054, + "loss": 3.8341, + "step": 8758 + }, + { + "epoch": 0.461, + "grad_norm": 0.484375, + "learning_rate": 0.00011977777777777779, + "loss": 3.4028, + "step": 8759 + }, + { + "epoch": 0.4610526315789474, + "grad_norm": 0.48828125, + "learning_rate": 0.00011976608187134504, + "loss": 3.7725, + "step": 8760 + }, + { + "epoch": 0.46110526315789474, + "grad_norm": 0.5, + "learning_rate": 0.00011975438596491229, + "loss": 3.4663, + "step": 8761 + }, + { + "epoch": 0.4611578947368421, + "grad_norm": 0.482421875, + "learning_rate": 0.00011974269005847953, + "loss": 3.6319, + "step": 8762 + }, + { + "epoch": 0.46121052631578946, + "grad_norm": 0.458984375, + "learning_rate": 0.00011973099415204678, + "loss": 3.9111, + "step": 8763 + }, + { + "epoch": 0.4612631578947368, + "grad_norm": 0.5078125, + "learning_rate": 0.00011971929824561403, + "loss": 2.7921, + "step": 8764 + }, + { + "epoch": 0.46131578947368423, + "grad_norm": 0.5234375, + "learning_rate": 0.0001197076023391813, + "loss": 3.7367, + "step": 8765 + }, + { + "epoch": 0.4613684210526316, + "grad_norm": 0.51953125, + "learning_rate": 0.00011969590643274855, + "loss": 3.2899, + "step": 8766 + }, + { + "epoch": 0.46142105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.0001196842105263158, + "loss": 3.8739, + "step": 8767 + }, + { + "epoch": 0.4614736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00011967251461988305, + "loss": 3.2847, + "step": 8768 + }, + { + "epoch": 0.4615263157894737, + "grad_norm": 0.486328125, + "learning_rate": 0.0001196608187134503, + "loss": 3.6519, + "step": 8769 + }, + { + "epoch": 0.46157894736842103, + "grad_norm": 0.515625, + "learning_rate": 0.00011964912280701755, + "loss": 3.6637, + "step": 8770 + }, + { + "epoch": 0.46163157894736845, + "grad_norm": 0.478515625, + "learning_rate": 0.0001196374269005848, + "loss": 3.5806, + "step": 8771 + }, + { + "epoch": 0.4616842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00011962573099415206, + "loss": 3.6044, + "step": 8772 + }, + { + "epoch": 0.46173684210526317, + "grad_norm": 0.54296875, + "learning_rate": 0.00011961403508771931, + "loss": 3.1224, + "step": 8773 + }, + { + "epoch": 0.4617894736842105, + "grad_norm": 0.474609375, + "learning_rate": 0.00011960233918128656, + "loss": 3.3442, + "step": 8774 + }, + { + "epoch": 0.4618421052631579, + "grad_norm": 0.5703125, + "learning_rate": 0.0001195906432748538, + "loss": 3.5593, + "step": 8775 + }, + { + "epoch": 0.46189473684210525, + "grad_norm": 0.47265625, + "learning_rate": 0.00011957894736842106, + "loss": 3.3303, + "step": 8776 + }, + { + "epoch": 0.4619473684210526, + "grad_norm": 0.52734375, + "learning_rate": 0.0001195672514619883, + "loss": 3.4756, + "step": 8777 + }, + { + "epoch": 0.462, + "grad_norm": 0.484375, + "learning_rate": 0.00011955555555555556, + "loss": 3.8589, + "step": 8778 + }, + { + "epoch": 0.4620526315789474, + "grad_norm": 0.458984375, + "learning_rate": 0.0001195438596491228, + "loss": 3.5967, + "step": 8779 + }, + { + "epoch": 0.46210526315789474, + "grad_norm": 0.60546875, + "learning_rate": 0.00011953216374269007, + "loss": 3.2843, + "step": 8780 + }, + { + "epoch": 0.4621578947368421, + "grad_norm": 0.55859375, + "learning_rate": 0.00011952046783625732, + "loss": 3.2907, + "step": 8781 + }, + { + "epoch": 0.46221052631578946, + "grad_norm": 0.47265625, + "learning_rate": 0.00011950877192982457, + "loss": 3.6604, + "step": 8782 + }, + { + "epoch": 0.4622631578947368, + "grad_norm": 0.6953125, + "learning_rate": 0.00011949707602339182, + "loss": 3.5894, + "step": 8783 + }, + { + "epoch": 0.46231578947368424, + "grad_norm": 0.443359375, + "learning_rate": 0.00011948538011695907, + "loss": 3.1996, + "step": 8784 + }, + { + "epoch": 0.4623684210526316, + "grad_norm": 0.8046875, + "learning_rate": 0.00011947368421052632, + "loss": 3.527, + "step": 8785 + }, + { + "epoch": 0.46242105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.00011946198830409357, + "loss": 3.1847, + "step": 8786 + }, + { + "epoch": 0.4624736842105263, + "grad_norm": 0.490234375, + "learning_rate": 0.00011945029239766083, + "loss": 3.4148, + "step": 8787 + }, + { + "epoch": 0.4625263157894737, + "grad_norm": 0.486328125, + "learning_rate": 0.00011943859649122808, + "loss": 3.5367, + "step": 8788 + }, + { + "epoch": 0.46257894736842103, + "grad_norm": 0.6171875, + "learning_rate": 0.00011942690058479533, + "loss": 3.6719, + "step": 8789 + }, + { + "epoch": 0.4626315789473684, + "grad_norm": 0.466796875, + "learning_rate": 0.00011941520467836258, + "loss": 3.4515, + "step": 8790 + }, + { + "epoch": 0.4626842105263158, + "grad_norm": 0.48828125, + "learning_rate": 0.00011940350877192983, + "loss": 3.5442, + "step": 8791 + }, + { + "epoch": 0.46273684210526317, + "grad_norm": 0.52734375, + "learning_rate": 0.00011939181286549708, + "loss": 3.5295, + "step": 8792 + }, + { + "epoch": 0.46278947368421053, + "grad_norm": 0.5, + "learning_rate": 0.00011938011695906433, + "loss": 3.4626, + "step": 8793 + }, + { + "epoch": 0.4628421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00011936842105263158, + "loss": 3.6362, + "step": 8794 + }, + { + "epoch": 0.46289473684210525, + "grad_norm": 0.53515625, + "learning_rate": 0.00011935672514619884, + "loss": 3.6404, + "step": 8795 + }, + { + "epoch": 0.4629473684210526, + "grad_norm": 0.5546875, + "learning_rate": 0.00011934502923976609, + "loss": 4.0311, + "step": 8796 + }, + { + "epoch": 0.463, + "grad_norm": 0.51171875, + "learning_rate": 0.00011933333333333334, + "loss": 3.3873, + "step": 8797 + }, + { + "epoch": 0.4630526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00011932163742690059, + "loss": 3.543, + "step": 8798 + }, + { + "epoch": 0.46310526315789474, + "grad_norm": 0.51171875, + "learning_rate": 0.00011930994152046784, + "loss": 3.2847, + "step": 8799 + }, + { + "epoch": 0.4631578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.00011929824561403509, + "loss": 3.4494, + "step": 8800 + }, + { + "epoch": 0.46321052631578946, + "grad_norm": 0.4765625, + "learning_rate": 0.00011928654970760234, + "loss": 3.5824, + "step": 8801 + }, + { + "epoch": 0.4632631578947368, + "grad_norm": 0.4921875, + "learning_rate": 0.0001192748538011696, + "loss": 3.2503, + "step": 8802 + }, + { + "epoch": 0.46331578947368424, + "grad_norm": 0.494140625, + "learning_rate": 0.00011926315789473685, + "loss": 3.5417, + "step": 8803 + }, + { + "epoch": 0.4633684210526316, + "grad_norm": 0.5390625, + "learning_rate": 0.0001192514619883041, + "loss": 3.1377, + "step": 8804 + }, + { + "epoch": 0.46342105263157896, + "grad_norm": 0.5, + "learning_rate": 0.00011923976608187135, + "loss": 3.4882, + "step": 8805 + }, + { + "epoch": 0.4634736842105263, + "grad_norm": 0.490234375, + "learning_rate": 0.0001192280701754386, + "loss": 3.5434, + "step": 8806 + }, + { + "epoch": 0.4635263157894737, + "grad_norm": 0.455078125, + "learning_rate": 0.00011921637426900585, + "loss": 3.5768, + "step": 8807 + }, + { + "epoch": 0.46357894736842103, + "grad_norm": 0.52734375, + "learning_rate": 0.0001192046783625731, + "loss": 3.6268, + "step": 8808 + }, + { + "epoch": 0.4636315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.00011919298245614037, + "loss": 3.4243, + "step": 8809 + }, + { + "epoch": 0.4636842105263158, + "grad_norm": 0.5703125, + "learning_rate": 0.00011918128654970761, + "loss": 3.5639, + "step": 8810 + }, + { + "epoch": 0.46373684210526317, + "grad_norm": 0.478515625, + "learning_rate": 0.00011916959064327486, + "loss": 3.5019, + "step": 8811 + }, + { + "epoch": 0.46378947368421053, + "grad_norm": 0.486328125, + "learning_rate": 0.00011915789473684211, + "loss": 3.0876, + "step": 8812 + }, + { + "epoch": 0.4638421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.00011914619883040936, + "loss": 3.6022, + "step": 8813 + }, + { + "epoch": 0.46389473684210525, + "grad_norm": 0.494140625, + "learning_rate": 0.00011913450292397661, + "loss": 3.589, + "step": 8814 + }, + { + "epoch": 0.4639473684210526, + "grad_norm": 0.53515625, + "learning_rate": 0.00011912280701754386, + "loss": 3.8829, + "step": 8815 + }, + { + "epoch": 0.464, + "grad_norm": 0.546875, + "learning_rate": 0.00011911111111111111, + "loss": 3.6331, + "step": 8816 + }, + { + "epoch": 0.4640526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00011909941520467837, + "loss": 3.5112, + "step": 8817 + }, + { + "epoch": 0.46410526315789474, + "grad_norm": 0.49609375, + "learning_rate": 0.00011908771929824562, + "loss": 3.5152, + "step": 8818 + }, + { + "epoch": 0.4641578947368421, + "grad_norm": 0.478515625, + "learning_rate": 0.00011907602339181287, + "loss": 3.6096, + "step": 8819 + }, + { + "epoch": 0.46421052631578946, + "grad_norm": 0.55078125, + "learning_rate": 0.00011906432748538012, + "loss": 3.6537, + "step": 8820 + }, + { + "epoch": 0.4642631578947368, + "grad_norm": 0.4921875, + "learning_rate": 0.00011905263157894737, + "loss": 3.5262, + "step": 8821 + }, + { + "epoch": 0.46431578947368424, + "grad_norm": 0.498046875, + "learning_rate": 0.00011904093567251462, + "loss": 2.9452, + "step": 8822 + }, + { + "epoch": 0.4643684210526316, + "grad_norm": 0.4765625, + "learning_rate": 0.00011902923976608187, + "loss": 3.7308, + "step": 8823 + }, + { + "epoch": 0.46442105263157896, + "grad_norm": 0.5234375, + "learning_rate": 0.00011901754385964915, + "loss": 3.6265, + "step": 8824 + }, + { + "epoch": 0.4644736842105263, + "grad_norm": 0.48046875, + "learning_rate": 0.00011900584795321638, + "loss": 3.4878, + "step": 8825 + }, + { + "epoch": 0.4645263157894737, + "grad_norm": 0.486328125, + "learning_rate": 0.00011899415204678363, + "loss": 3.5515, + "step": 8826 + }, + { + "epoch": 0.46457894736842104, + "grad_norm": 0.5, + "learning_rate": 0.00011898245614035088, + "loss": 4.1066, + "step": 8827 + }, + { + "epoch": 0.4646315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00011897076023391813, + "loss": 3.788, + "step": 8828 + }, + { + "epoch": 0.4646842105263158, + "grad_norm": 0.4765625, + "learning_rate": 0.00011895906432748538, + "loss": 3.4489, + "step": 8829 + }, + { + "epoch": 0.46473684210526317, + "grad_norm": 0.49609375, + "learning_rate": 0.00011894736842105263, + "loss": 3.811, + "step": 8830 + }, + { + "epoch": 0.46478947368421053, + "grad_norm": 0.48828125, + "learning_rate": 0.00011893567251461988, + "loss": 3.8106, + "step": 8831 + }, + { + "epoch": 0.4648421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00011892397660818716, + "loss": 3.4193, + "step": 8832 + }, + { + "epoch": 0.46489473684210525, + "grad_norm": 0.498046875, + "learning_rate": 0.00011891228070175439, + "loss": 3.5112, + "step": 8833 + }, + { + "epoch": 0.4649473684210526, + "grad_norm": 0.4765625, + "learning_rate": 0.00011890058479532164, + "loss": 3.4748, + "step": 8834 + }, + { + "epoch": 0.465, + "grad_norm": 0.48046875, + "learning_rate": 0.00011888888888888889, + "loss": 3.76, + "step": 8835 + }, + { + "epoch": 0.4650526315789474, + "grad_norm": 0.5, + "learning_rate": 0.00011887719298245614, + "loss": 3.3869, + "step": 8836 + }, + { + "epoch": 0.46510526315789474, + "grad_norm": 0.49609375, + "learning_rate": 0.00011886549707602339, + "loss": 3.3158, + "step": 8837 + }, + { + "epoch": 0.4651578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00011885380116959064, + "loss": 3.2517, + "step": 8838 + }, + { + "epoch": 0.46521052631578946, + "grad_norm": 0.59765625, + "learning_rate": 0.00011884210526315792, + "loss": 3.1474, + "step": 8839 + }, + { + "epoch": 0.4652631578947368, + "grad_norm": 0.5625, + "learning_rate": 0.00011883040935672515, + "loss": 3.7578, + "step": 8840 + }, + { + "epoch": 0.46531578947368424, + "grad_norm": 0.478515625, + "learning_rate": 0.0001188187134502924, + "loss": 4.0058, + "step": 8841 + }, + { + "epoch": 0.4653684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00011880701754385965, + "loss": 3.4939, + "step": 8842 + }, + { + "epoch": 0.46542105263157896, + "grad_norm": 0.51953125, + "learning_rate": 0.0001187953216374269, + "loss": 3.4151, + "step": 8843 + }, + { + "epoch": 0.4654736842105263, + "grad_norm": 0.51171875, + "learning_rate": 0.00011878362573099415, + "loss": 3.6647, + "step": 8844 + }, + { + "epoch": 0.4655263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.0001187719298245614, + "loss": 3.6684, + "step": 8845 + }, + { + "epoch": 0.46557894736842104, + "grad_norm": 0.458984375, + "learning_rate": 0.00011876023391812868, + "loss": 3.5797, + "step": 8846 + }, + { + "epoch": 0.4656315789473684, + "grad_norm": 0.482421875, + "learning_rate": 0.00011874853801169593, + "loss": 3.5461, + "step": 8847 + }, + { + "epoch": 0.4656842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.00011873684210526316, + "loss": 3.4835, + "step": 8848 + }, + { + "epoch": 0.46573684210526317, + "grad_norm": 0.50390625, + "learning_rate": 0.00011872514619883041, + "loss": 3.3415, + "step": 8849 + }, + { + "epoch": 0.46578947368421053, + "grad_norm": 0.462890625, + "learning_rate": 0.00011871345029239766, + "loss": 3.5926, + "step": 8850 + }, + { + "epoch": 0.4658421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00011870175438596491, + "loss": 3.6227, + "step": 8851 + }, + { + "epoch": 0.46589473684210525, + "grad_norm": 0.58984375, + "learning_rate": 0.00011869005847953216, + "loss": 3.3431, + "step": 8852 + }, + { + "epoch": 0.4659473684210526, + "grad_norm": 0.546875, + "learning_rate": 0.00011867836257309941, + "loss": 3.7075, + "step": 8853 + }, + { + "epoch": 0.466, + "grad_norm": 0.76953125, + "learning_rate": 0.00011866666666666669, + "loss": 3.6934, + "step": 8854 + }, + { + "epoch": 0.4660526315789474, + "grad_norm": 0.462890625, + "learning_rate": 0.00011865497076023394, + "loss": 3.5103, + "step": 8855 + }, + { + "epoch": 0.46610526315789474, + "grad_norm": 0.55078125, + "learning_rate": 0.00011864327485380117, + "loss": 3.4815, + "step": 8856 + }, + { + "epoch": 0.4661578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.00011863157894736842, + "loss": 3.4338, + "step": 8857 + }, + { + "epoch": 0.46621052631578946, + "grad_norm": 0.474609375, + "learning_rate": 0.00011861988304093567, + "loss": 3.2388, + "step": 8858 + }, + { + "epoch": 0.4662631578947368, + "grad_norm": 0.49609375, + "learning_rate": 0.00011860818713450292, + "loss": 2.775, + "step": 8859 + }, + { + "epoch": 0.4663157894736842, + "grad_norm": 0.486328125, + "learning_rate": 0.00011859649122807017, + "loss": 3.5424, + "step": 8860 + }, + { + "epoch": 0.4663684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00011858479532163745, + "loss": 3.6973, + "step": 8861 + }, + { + "epoch": 0.46642105263157896, + "grad_norm": 0.54296875, + "learning_rate": 0.0001185730994152047, + "loss": 3.6973, + "step": 8862 + }, + { + "epoch": 0.4664736842105263, + "grad_norm": 0.466796875, + "learning_rate": 0.00011856140350877193, + "loss": 4.0112, + "step": 8863 + }, + { + "epoch": 0.4665263157894737, + "grad_norm": 0.498046875, + "learning_rate": 0.00011854970760233918, + "loss": 3.6803, + "step": 8864 + }, + { + "epoch": 0.46657894736842104, + "grad_norm": 0.50390625, + "learning_rate": 0.00011853801169590643, + "loss": 3.3237, + "step": 8865 + }, + { + "epoch": 0.4666315789473684, + "grad_norm": 0.466796875, + "learning_rate": 0.00011852631578947368, + "loss": 3.7338, + "step": 8866 + }, + { + "epoch": 0.4666842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00011851461988304093, + "loss": 3.4789, + "step": 8867 + }, + { + "epoch": 0.46673684210526317, + "grad_norm": 0.486328125, + "learning_rate": 0.00011850292397660818, + "loss": 3.4165, + "step": 8868 + }, + { + "epoch": 0.46678947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.00011849122807017546, + "loss": 3.3462, + "step": 8869 + }, + { + "epoch": 0.4668421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.00011847953216374271, + "loss": 3.707, + "step": 8870 + }, + { + "epoch": 0.46689473684210525, + "grad_norm": 0.484375, + "learning_rate": 0.00011846783625730994, + "loss": 3.989, + "step": 8871 + }, + { + "epoch": 0.4669473684210526, + "grad_norm": 0.51953125, + "learning_rate": 0.00011845614035087719, + "loss": 3.9259, + "step": 8872 + }, + { + "epoch": 0.467, + "grad_norm": 0.48046875, + "learning_rate": 0.00011844444444444444, + "loss": 3.5436, + "step": 8873 + }, + { + "epoch": 0.4670526315789474, + "grad_norm": 0.5, + "learning_rate": 0.00011843274853801169, + "loss": 3.5577, + "step": 8874 + }, + { + "epoch": 0.46710526315789475, + "grad_norm": 0.474609375, + "learning_rate": 0.00011842105263157894, + "loss": 3.7429, + "step": 8875 + }, + { + "epoch": 0.4671578947368421, + "grad_norm": 0.482421875, + "learning_rate": 0.00011840935672514622, + "loss": 3.5775, + "step": 8876 + }, + { + "epoch": 0.46721052631578947, + "grad_norm": 0.46875, + "learning_rate": 0.00011839766081871347, + "loss": 3.3155, + "step": 8877 + }, + { + "epoch": 0.4672631578947368, + "grad_norm": 0.5078125, + "learning_rate": 0.00011838596491228072, + "loss": 3.5037, + "step": 8878 + }, + { + "epoch": 0.4673157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00011837426900584795, + "loss": 3.3029, + "step": 8879 + }, + { + "epoch": 0.4673684210526316, + "grad_norm": 0.494140625, + "learning_rate": 0.0001183625730994152, + "loss": 3.2096, + "step": 8880 + }, + { + "epoch": 0.46742105263157896, + "grad_norm": 0.45703125, + "learning_rate": 0.00011835087719298245, + "loss": 3.6711, + "step": 8881 + }, + { + "epoch": 0.4674736842105263, + "grad_norm": 0.46484375, + "learning_rate": 0.0001183391812865497, + "loss": 3.6589, + "step": 8882 + }, + { + "epoch": 0.4675263157894737, + "grad_norm": 0.482421875, + "learning_rate": 0.00011832748538011698, + "loss": 3.37, + "step": 8883 + }, + { + "epoch": 0.46757894736842104, + "grad_norm": 0.470703125, + "learning_rate": 0.00011831578947368423, + "loss": 3.6577, + "step": 8884 + }, + { + "epoch": 0.4676315789473684, + "grad_norm": 0.4765625, + "learning_rate": 0.00011830409356725148, + "loss": 3.5578, + "step": 8885 + }, + { + "epoch": 0.4676842105263158, + "grad_norm": 0.498046875, + "learning_rate": 0.00011829239766081871, + "loss": 3.9213, + "step": 8886 + }, + { + "epoch": 0.4677368421052632, + "grad_norm": 0.482421875, + "learning_rate": 0.00011828070175438596, + "loss": 3.6096, + "step": 8887 + }, + { + "epoch": 0.46778947368421053, + "grad_norm": 0.49609375, + "learning_rate": 0.00011826900584795321, + "loss": 3.5203, + "step": 8888 + }, + { + "epoch": 0.4678421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00011825730994152046, + "loss": 3.7021, + "step": 8889 + }, + { + "epoch": 0.46789473684210525, + "grad_norm": 0.494140625, + "learning_rate": 0.00011824561403508771, + "loss": 3.3908, + "step": 8890 + }, + { + "epoch": 0.4679473684210526, + "grad_norm": 0.46484375, + "learning_rate": 0.00011823391812865499, + "loss": 3.3635, + "step": 8891 + }, + { + "epoch": 0.468, + "grad_norm": 0.490234375, + "learning_rate": 0.00011822222222222224, + "loss": 3.6377, + "step": 8892 + }, + { + "epoch": 0.4680526315789474, + "grad_norm": 0.515625, + "learning_rate": 0.00011821052631578949, + "loss": 3.6628, + "step": 8893 + }, + { + "epoch": 0.46810526315789475, + "grad_norm": 0.51171875, + "learning_rate": 0.00011819883040935672, + "loss": 3.8922, + "step": 8894 + }, + { + "epoch": 0.4681578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00011818713450292397, + "loss": 3.9164, + "step": 8895 + }, + { + "epoch": 0.46821052631578947, + "grad_norm": 0.490234375, + "learning_rate": 0.00011817543859649122, + "loss": 3.4447, + "step": 8896 + }, + { + "epoch": 0.4682631578947368, + "grad_norm": 0.498046875, + "learning_rate": 0.00011816374269005847, + "loss": 3.469, + "step": 8897 + }, + { + "epoch": 0.4683157894736842, + "grad_norm": 0.65625, + "learning_rate": 0.00011815204678362575, + "loss": 3.8116, + "step": 8898 + }, + { + "epoch": 0.4683684210526316, + "grad_norm": 0.59375, + "learning_rate": 0.000118140350877193, + "loss": 3.5785, + "step": 8899 + }, + { + "epoch": 0.46842105263157896, + "grad_norm": 0.55859375, + "learning_rate": 0.00011812865497076025, + "loss": 3.5442, + "step": 8900 + }, + { + "epoch": 0.4684736842105263, + "grad_norm": 0.5390625, + "learning_rate": 0.0001181169590643275, + "loss": 3.5305, + "step": 8901 + }, + { + "epoch": 0.4685263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00011810526315789474, + "loss": 3.882, + "step": 8902 + }, + { + "epoch": 0.46857894736842104, + "grad_norm": 0.5390625, + "learning_rate": 0.00011809356725146198, + "loss": 3.706, + "step": 8903 + }, + { + "epoch": 0.4686315789473684, + "grad_norm": 0.45703125, + "learning_rate": 0.00011808187134502923, + "loss": 3.1258, + "step": 8904 + }, + { + "epoch": 0.4686842105263158, + "grad_norm": 0.4765625, + "learning_rate": 0.00011807017543859648, + "loss": 3.444, + "step": 8905 + }, + { + "epoch": 0.4687368421052632, + "grad_norm": 0.474609375, + "learning_rate": 0.00011805847953216376, + "loss": 3.495, + "step": 8906 + }, + { + "epoch": 0.46878947368421053, + "grad_norm": 0.458984375, + "learning_rate": 0.00011804678362573101, + "loss": 3.6768, + "step": 8907 + }, + { + "epoch": 0.4688421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00011803508771929826, + "loss": 3.6011, + "step": 8908 + }, + { + "epoch": 0.46889473684210525, + "grad_norm": 0.52734375, + "learning_rate": 0.0001180233918128655, + "loss": 3.7105, + "step": 8909 + }, + { + "epoch": 0.4689473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.00011801169590643275, + "loss": 3.4976, + "step": 8910 + }, + { + "epoch": 0.469, + "grad_norm": 0.458984375, + "learning_rate": 0.000118, + "loss": 3.3628, + "step": 8911 + }, + { + "epoch": 0.4690526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00011798830409356724, + "loss": 3.548, + "step": 8912 + }, + { + "epoch": 0.46910526315789475, + "grad_norm": 0.4453125, + "learning_rate": 0.00011797660818713452, + "loss": 3.4063, + "step": 8913 + }, + { + "epoch": 0.4691578947368421, + "grad_norm": 0.578125, + "learning_rate": 0.00011796491228070177, + "loss": 3.4202, + "step": 8914 + }, + { + "epoch": 0.46921052631578947, + "grad_norm": 0.49609375, + "learning_rate": 0.00011795321637426902, + "loss": 3.8715, + "step": 8915 + }, + { + "epoch": 0.4692631578947368, + "grad_norm": 0.50390625, + "learning_rate": 0.00011794152046783627, + "loss": 3.4575, + "step": 8916 + }, + { + "epoch": 0.4693157894736842, + "grad_norm": 0.49609375, + "learning_rate": 0.0001179298245614035, + "loss": 3.8804, + "step": 8917 + }, + { + "epoch": 0.4693684210526316, + "grad_norm": 0.5703125, + "learning_rate": 0.00011791812865497076, + "loss": 3.608, + "step": 8918 + }, + { + "epoch": 0.46942105263157896, + "grad_norm": 0.69140625, + "learning_rate": 0.000117906432748538, + "loss": 3.733, + "step": 8919 + }, + { + "epoch": 0.4694736842105263, + "grad_norm": 0.451171875, + "learning_rate": 0.00011789473684210525, + "loss": 3.4811, + "step": 8920 + }, + { + "epoch": 0.4695263157894737, + "grad_norm": 0.46484375, + "learning_rate": 0.00011788304093567253, + "loss": 3.5517, + "step": 8921 + }, + { + "epoch": 0.46957894736842104, + "grad_norm": 0.5, + "learning_rate": 0.00011787134502923978, + "loss": 3.6751, + "step": 8922 + }, + { + "epoch": 0.4696315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.00011785964912280703, + "loss": 3.9049, + "step": 8923 + }, + { + "epoch": 0.4696842105263158, + "grad_norm": 0.48828125, + "learning_rate": 0.00011784795321637428, + "loss": 3.7243, + "step": 8924 + }, + { + "epoch": 0.4697368421052632, + "grad_norm": 0.48046875, + "learning_rate": 0.00011783625730994152, + "loss": 3.4078, + "step": 8925 + }, + { + "epoch": 0.46978947368421053, + "grad_norm": 0.484375, + "learning_rate": 0.00011782456140350877, + "loss": 3.7861, + "step": 8926 + }, + { + "epoch": 0.4698421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00011781286549707602, + "loss": 3.5684, + "step": 8927 + }, + { + "epoch": 0.46989473684210525, + "grad_norm": 0.54296875, + "learning_rate": 0.00011780116959064329, + "loss": 3.4838, + "step": 8928 + }, + { + "epoch": 0.4699473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00011778947368421054, + "loss": 3.6083, + "step": 8929 + }, + { + "epoch": 0.47, + "grad_norm": 0.5390625, + "learning_rate": 0.00011777777777777779, + "loss": 3.7434, + "step": 8930 + }, + { + "epoch": 0.4700526315789474, + "grad_norm": 0.48828125, + "learning_rate": 0.00011776608187134504, + "loss": 3.6863, + "step": 8931 + }, + { + "epoch": 0.47010526315789475, + "grad_norm": 0.52734375, + "learning_rate": 0.00011775438596491228, + "loss": 3.7617, + "step": 8932 + }, + { + "epoch": 0.4701578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00011774269005847953, + "loss": 3.7015, + "step": 8933 + }, + { + "epoch": 0.47021052631578947, + "grad_norm": 0.50390625, + "learning_rate": 0.00011773099415204678, + "loss": 3.396, + "step": 8934 + }, + { + "epoch": 0.4702631578947368, + "grad_norm": 0.5, + "learning_rate": 0.00011771929824561405, + "loss": 3.6041, + "step": 8935 + }, + { + "epoch": 0.4703157894736842, + "grad_norm": 0.55859375, + "learning_rate": 0.0001177076023391813, + "loss": 3.685, + "step": 8936 + }, + { + "epoch": 0.4703684210526316, + "grad_norm": 0.55859375, + "learning_rate": 0.00011769590643274855, + "loss": 3.5474, + "step": 8937 + }, + { + "epoch": 0.47042105263157896, + "grad_norm": 0.49609375, + "learning_rate": 0.0001176842105263158, + "loss": 3.6913, + "step": 8938 + }, + { + "epoch": 0.4704736842105263, + "grad_norm": 0.451171875, + "learning_rate": 0.00011767251461988305, + "loss": 3.4436, + "step": 8939 + }, + { + "epoch": 0.4705263157894737, + "grad_norm": 0.494140625, + "learning_rate": 0.00011766081871345029, + "loss": 3.5635, + "step": 8940 + }, + { + "epoch": 0.47057894736842104, + "grad_norm": 0.474609375, + "learning_rate": 0.00011764912280701754, + "loss": 4.0885, + "step": 8941 + }, + { + "epoch": 0.4706315789473684, + "grad_norm": 0.5703125, + "learning_rate": 0.00011763742690058479, + "loss": 3.3748, + "step": 8942 + }, + { + "epoch": 0.4706842105263158, + "grad_norm": 0.462890625, + "learning_rate": 0.00011762573099415206, + "loss": 3.1984, + "step": 8943 + }, + { + "epoch": 0.4707368421052632, + "grad_norm": 0.48046875, + "learning_rate": 0.00011761403508771931, + "loss": 3.8764, + "step": 8944 + }, + { + "epoch": 0.47078947368421054, + "grad_norm": 0.53125, + "learning_rate": 0.00011760233918128656, + "loss": 3.6386, + "step": 8945 + }, + { + "epoch": 0.4708421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00011759064327485381, + "loss": 3.6068, + "step": 8946 + }, + { + "epoch": 0.47089473684210525, + "grad_norm": 0.59375, + "learning_rate": 0.00011757894736842106, + "loss": 3.6656, + "step": 8947 + }, + { + "epoch": 0.4709473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.0001175672514619883, + "loss": 3.6705, + "step": 8948 + }, + { + "epoch": 0.471, + "grad_norm": 0.494140625, + "learning_rate": 0.00011755555555555555, + "loss": 3.7445, + "step": 8949 + }, + { + "epoch": 0.4710526315789474, + "grad_norm": 0.486328125, + "learning_rate": 0.00011754385964912282, + "loss": 3.7396, + "step": 8950 + }, + { + "epoch": 0.47110526315789475, + "grad_norm": 0.578125, + "learning_rate": 0.00011753216374269007, + "loss": 3.5125, + "step": 8951 + }, + { + "epoch": 0.4711578947368421, + "grad_norm": 0.46875, + "learning_rate": 0.00011752046783625732, + "loss": 3.7289, + "step": 8952 + }, + { + "epoch": 0.47121052631578947, + "grad_norm": 0.5234375, + "learning_rate": 0.00011750877192982457, + "loss": 3.347, + "step": 8953 + }, + { + "epoch": 0.47126315789473683, + "grad_norm": 0.4609375, + "learning_rate": 0.00011749707602339182, + "loss": 3.531, + "step": 8954 + }, + { + "epoch": 0.4713157894736842, + "grad_norm": 0.498046875, + "learning_rate": 0.00011748538011695906, + "loss": 3.771, + "step": 8955 + }, + { + "epoch": 0.4713684210526316, + "grad_norm": 0.455078125, + "learning_rate": 0.00011747368421052631, + "loss": 3.4863, + "step": 8956 + }, + { + "epoch": 0.47142105263157896, + "grad_norm": 0.462890625, + "learning_rate": 0.00011746198830409356, + "loss": 3.0887, + "step": 8957 + }, + { + "epoch": 0.4714736842105263, + "grad_norm": 0.48828125, + "learning_rate": 0.00011745029239766083, + "loss": 3.4527, + "step": 8958 + }, + { + "epoch": 0.4715263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00011743859649122808, + "loss": 3.6495, + "step": 8959 + }, + { + "epoch": 0.47157894736842104, + "grad_norm": 0.48828125, + "learning_rate": 0.00011742690058479533, + "loss": 3.6974, + "step": 8960 + }, + { + "epoch": 0.4716315789473684, + "grad_norm": 0.486328125, + "learning_rate": 0.00011741520467836258, + "loss": 3.1794, + "step": 8961 + }, + { + "epoch": 0.4716842105263158, + "grad_norm": 0.453125, + "learning_rate": 0.00011740350877192983, + "loss": 3.652, + "step": 8962 + }, + { + "epoch": 0.4717368421052632, + "grad_norm": 0.453125, + "learning_rate": 0.00011739181286549707, + "loss": 3.366, + "step": 8963 + }, + { + "epoch": 0.47178947368421054, + "grad_norm": 0.6953125, + "learning_rate": 0.00011738011695906432, + "loss": 3.3718, + "step": 8964 + }, + { + "epoch": 0.4718421052631579, + "grad_norm": 0.61328125, + "learning_rate": 0.0001173684210526316, + "loss": 3.7048, + "step": 8965 + }, + { + "epoch": 0.47189473684210526, + "grad_norm": 0.5703125, + "learning_rate": 0.00011735672514619884, + "loss": 3.1018, + "step": 8966 + }, + { + "epoch": 0.4719473684210526, + "grad_norm": 0.470703125, + "learning_rate": 0.0001173450292397661, + "loss": 3.3002, + "step": 8967 + }, + { + "epoch": 0.472, + "grad_norm": 0.6640625, + "learning_rate": 0.00011733333333333334, + "loss": 3.3691, + "step": 8968 + }, + { + "epoch": 0.4720526315789474, + "grad_norm": 0.5, + "learning_rate": 0.0001173216374269006, + "loss": 3.4801, + "step": 8969 + }, + { + "epoch": 0.47210526315789475, + "grad_norm": 0.51171875, + "learning_rate": 0.00011730994152046784, + "loss": 3.7994, + "step": 8970 + }, + { + "epoch": 0.4721578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.00011729824561403508, + "loss": 3.2285, + "step": 8971 + }, + { + "epoch": 0.47221052631578947, + "grad_norm": 0.478515625, + "learning_rate": 0.00011728654970760236, + "loss": 3.4427, + "step": 8972 + }, + { + "epoch": 0.47226315789473683, + "grad_norm": 0.48046875, + "learning_rate": 0.0001172748538011696, + "loss": 3.6176, + "step": 8973 + }, + { + "epoch": 0.4723157894736842, + "grad_norm": 0.466796875, + "learning_rate": 0.00011726315789473685, + "loss": 3.5613, + "step": 8974 + }, + { + "epoch": 0.4723684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.0001172514619883041, + "loss": 3.6831, + "step": 8975 + }, + { + "epoch": 0.47242105263157896, + "grad_norm": 0.484375, + "learning_rate": 0.00011723976608187135, + "loss": 3.2991, + "step": 8976 + }, + { + "epoch": 0.4724736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.0001172280701754386, + "loss": 3.4063, + "step": 8977 + }, + { + "epoch": 0.4725263157894737, + "grad_norm": 0.515625, + "learning_rate": 0.00011721637426900584, + "loss": 3.6964, + "step": 8978 + }, + { + "epoch": 0.47257894736842104, + "grad_norm": 0.5, + "learning_rate": 0.00011720467836257309, + "loss": 3.587, + "step": 8979 + }, + { + "epoch": 0.4726315789473684, + "grad_norm": 0.6171875, + "learning_rate": 0.00011719298245614037, + "loss": 3.3213, + "step": 8980 + }, + { + "epoch": 0.47268421052631576, + "grad_norm": 0.52734375, + "learning_rate": 0.00011718128654970762, + "loss": 3.2347, + "step": 8981 + }, + { + "epoch": 0.4727368421052632, + "grad_norm": 0.49609375, + "learning_rate": 0.00011716959064327487, + "loss": 3.4198, + "step": 8982 + }, + { + "epoch": 0.47278947368421054, + "grad_norm": 0.5, + "learning_rate": 0.00011715789473684211, + "loss": 3.1632, + "step": 8983 + }, + { + "epoch": 0.4728421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00011714619883040936, + "loss": 3.2087, + "step": 8984 + }, + { + "epoch": 0.47289473684210526, + "grad_norm": 0.515625, + "learning_rate": 0.00011713450292397661, + "loss": 3.3442, + "step": 8985 + }, + { + "epoch": 0.4729473684210526, + "grad_norm": 0.53125, + "learning_rate": 0.00011712280701754385, + "loss": 3.5148, + "step": 8986 + }, + { + "epoch": 0.473, + "grad_norm": 0.5390625, + "learning_rate": 0.00011711111111111113, + "loss": 3.1706, + "step": 8987 + }, + { + "epoch": 0.4730526315789474, + "grad_norm": 0.515625, + "learning_rate": 0.00011709941520467838, + "loss": 3.8381, + "step": 8988 + }, + { + "epoch": 0.47310526315789475, + "grad_norm": 0.5234375, + "learning_rate": 0.00011708771929824563, + "loss": 3.7612, + "step": 8989 + }, + { + "epoch": 0.4731578947368421, + "grad_norm": 0.49609375, + "learning_rate": 0.00011707602339181288, + "loss": 3.4833, + "step": 8990 + }, + { + "epoch": 0.47321052631578947, + "grad_norm": 0.5078125, + "learning_rate": 0.00011706432748538013, + "loss": 3.9664, + "step": 8991 + }, + { + "epoch": 0.47326315789473683, + "grad_norm": 0.58984375, + "learning_rate": 0.00011705263157894737, + "loss": 3.462, + "step": 8992 + }, + { + "epoch": 0.4733157894736842, + "grad_norm": 0.54296875, + "learning_rate": 0.00011704093567251462, + "loss": 3.8797, + "step": 8993 + }, + { + "epoch": 0.4733684210526316, + "grad_norm": 0.5625, + "learning_rate": 0.00011702923976608186, + "loss": 3.5495, + "step": 8994 + }, + { + "epoch": 0.47342105263157896, + "grad_norm": 0.515625, + "learning_rate": 0.00011701754385964914, + "loss": 3.2536, + "step": 8995 + }, + { + "epoch": 0.4734736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00011700584795321639, + "loss": 3.0558, + "step": 8996 + }, + { + "epoch": 0.4735263157894737, + "grad_norm": 0.4609375, + "learning_rate": 0.00011699415204678364, + "loss": 3.4225, + "step": 8997 + }, + { + "epoch": 0.47357894736842104, + "grad_norm": 0.462890625, + "learning_rate": 0.00011698245614035089, + "loss": 3.8381, + "step": 8998 + }, + { + "epoch": 0.4736315789473684, + "grad_norm": 0.47265625, + "learning_rate": 0.00011697076023391814, + "loss": 3.867, + "step": 8999 + }, + { + "epoch": 0.47368421052631576, + "grad_norm": 0.47265625, + "learning_rate": 0.00011695906432748539, + "loss": 3.3632, + "step": 9000 + }, + { + "epoch": 0.4737368421052632, + "grad_norm": 0.486328125, + "learning_rate": 0.00011694736842105262, + "loss": 2.7531, + "step": 9001 + }, + { + "epoch": 0.47378947368421054, + "grad_norm": 0.515625, + "learning_rate": 0.0001169356725146199, + "loss": 3.5921, + "step": 9002 + }, + { + "epoch": 0.4738421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00011692397660818715, + "loss": 3.6149, + "step": 9003 + }, + { + "epoch": 0.47389473684210526, + "grad_norm": 0.55859375, + "learning_rate": 0.0001169122807017544, + "loss": 3.498, + "step": 9004 + }, + { + "epoch": 0.4739473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00011690058479532165, + "loss": 2.8455, + "step": 9005 + }, + { + "epoch": 0.474, + "grad_norm": 0.482421875, + "learning_rate": 0.0001168888888888889, + "loss": 3.0908, + "step": 9006 + }, + { + "epoch": 0.4740526315789474, + "grad_norm": 0.53125, + "learning_rate": 0.00011687719298245615, + "loss": 3.2807, + "step": 9007 + }, + { + "epoch": 0.47410526315789475, + "grad_norm": 0.494140625, + "learning_rate": 0.0001168654970760234, + "loss": 3.7176, + "step": 9008 + }, + { + "epoch": 0.4741578947368421, + "grad_norm": 0.48828125, + "learning_rate": 0.00011685380116959066, + "loss": 2.9911, + "step": 9009 + }, + { + "epoch": 0.47421052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.00011684210526315791, + "loss": 2.8777, + "step": 9010 + }, + { + "epoch": 0.47426315789473683, + "grad_norm": 0.474609375, + "learning_rate": 0.00011683040935672516, + "loss": 3.5384, + "step": 9011 + }, + { + "epoch": 0.4743157894736842, + "grad_norm": 0.51953125, + "learning_rate": 0.00011681871345029241, + "loss": 3.4782, + "step": 9012 + }, + { + "epoch": 0.4743684210526316, + "grad_norm": 0.49609375, + "learning_rate": 0.00011680701754385966, + "loss": 3.8081, + "step": 9013 + }, + { + "epoch": 0.47442105263157897, + "grad_norm": 0.50390625, + "learning_rate": 0.0001167953216374269, + "loss": 3.602, + "step": 9014 + }, + { + "epoch": 0.4744736842105263, + "grad_norm": 0.51953125, + "learning_rate": 0.00011678362573099416, + "loss": 3.5881, + "step": 9015 + }, + { + "epoch": 0.4745263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.0001167719298245614, + "loss": 3.3711, + "step": 9016 + }, + { + "epoch": 0.47457894736842104, + "grad_norm": 0.578125, + "learning_rate": 0.00011676023391812867, + "loss": 3.5161, + "step": 9017 + }, + { + "epoch": 0.4746315789473684, + "grad_norm": 0.5625, + "learning_rate": 0.00011674853801169592, + "loss": 3.4984, + "step": 9018 + }, + { + "epoch": 0.47468421052631576, + "grad_norm": 1.4609375, + "learning_rate": 0.00011673684210526317, + "loss": 3.4745, + "step": 9019 + }, + { + "epoch": 0.4747368421052632, + "grad_norm": 0.474609375, + "learning_rate": 0.00011672514619883042, + "loss": 3.194, + "step": 9020 + }, + { + "epoch": 0.47478947368421054, + "grad_norm": 0.48828125, + "learning_rate": 0.00011671345029239767, + "loss": 3.8027, + "step": 9021 + }, + { + "epoch": 0.4748421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00011670175438596492, + "loss": 3.3384, + "step": 9022 + }, + { + "epoch": 0.47489473684210526, + "grad_norm": 0.5390625, + "learning_rate": 0.00011669005847953217, + "loss": 3.5946, + "step": 9023 + }, + { + "epoch": 0.4749473684210526, + "grad_norm": 0.53515625, + "learning_rate": 0.00011667836257309943, + "loss": 3.8235, + "step": 9024 + }, + { + "epoch": 0.475, + "grad_norm": 0.5078125, + "learning_rate": 0.00011666666666666668, + "loss": 3.1924, + "step": 9025 + }, + { + "epoch": 0.4750526315789474, + "grad_norm": 0.462890625, + "learning_rate": 0.00011665497076023393, + "loss": 3.5278, + "step": 9026 + }, + { + "epoch": 0.47510526315789475, + "grad_norm": 0.458984375, + "learning_rate": 0.00011664327485380118, + "loss": 3.4465, + "step": 9027 + }, + { + "epoch": 0.4751578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00011663157894736843, + "loss": 3.6307, + "step": 9028 + }, + { + "epoch": 0.47521052631578947, + "grad_norm": 0.486328125, + "learning_rate": 0.00011661988304093568, + "loss": 3.3349, + "step": 9029 + }, + { + "epoch": 0.47526315789473683, + "grad_norm": 0.486328125, + "learning_rate": 0.00011660818713450293, + "loss": 3.5074, + "step": 9030 + }, + { + "epoch": 0.4753157894736842, + "grad_norm": 0.57421875, + "learning_rate": 0.00011659649122807018, + "loss": 3.8508, + "step": 9031 + }, + { + "epoch": 0.47536842105263155, + "grad_norm": 0.462890625, + "learning_rate": 0.00011658479532163744, + "loss": 3.7669, + "step": 9032 + }, + { + "epoch": 0.47542105263157897, + "grad_norm": 0.5078125, + "learning_rate": 0.00011657309941520469, + "loss": 3.8962, + "step": 9033 + }, + { + "epoch": 0.4754736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00011656140350877194, + "loss": 3.5657, + "step": 9034 + }, + { + "epoch": 0.4755263157894737, + "grad_norm": 0.4921875, + "learning_rate": 0.00011654970760233919, + "loss": 3.4012, + "step": 9035 + }, + { + "epoch": 0.47557894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.00011653801169590644, + "loss": 3.3113, + "step": 9036 + }, + { + "epoch": 0.4756315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00011652631578947369, + "loss": 3.8129, + "step": 9037 + }, + { + "epoch": 0.47568421052631576, + "grad_norm": 0.4921875, + "learning_rate": 0.00011651461988304094, + "loss": 3.831, + "step": 9038 + }, + { + "epoch": 0.4757368421052632, + "grad_norm": 0.474609375, + "learning_rate": 0.0001165029239766082, + "loss": 3.3692, + "step": 9039 + }, + { + "epoch": 0.47578947368421054, + "grad_norm": 0.46484375, + "learning_rate": 0.00011649122807017545, + "loss": 3.6063, + "step": 9040 + }, + { + "epoch": 0.4758421052631579, + "grad_norm": 0.484375, + "learning_rate": 0.0001164795321637427, + "loss": 3.5602, + "step": 9041 + }, + { + "epoch": 0.47589473684210526, + "grad_norm": 0.462890625, + "learning_rate": 0.00011646783625730995, + "loss": 3.2271, + "step": 9042 + }, + { + "epoch": 0.4759473684210526, + "grad_norm": 0.5, + "learning_rate": 0.0001164561403508772, + "loss": 3.2948, + "step": 9043 + }, + { + "epoch": 0.476, + "grad_norm": 0.52734375, + "learning_rate": 0.00011644444444444445, + "loss": 3.8966, + "step": 9044 + }, + { + "epoch": 0.4760526315789474, + "grad_norm": 0.55859375, + "learning_rate": 0.0001164327485380117, + "loss": 3.3907, + "step": 9045 + }, + { + "epoch": 0.47610526315789475, + "grad_norm": 0.48046875, + "learning_rate": 0.00011642105263157896, + "loss": 3.119, + "step": 9046 + }, + { + "epoch": 0.4761578947368421, + "grad_norm": 0.46484375, + "learning_rate": 0.00011640935672514621, + "loss": 3.2992, + "step": 9047 + }, + { + "epoch": 0.4762105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00011639766081871346, + "loss": 3.5941, + "step": 9048 + }, + { + "epoch": 0.47626315789473683, + "grad_norm": 0.5, + "learning_rate": 0.00011638596491228071, + "loss": 3.6948, + "step": 9049 + }, + { + "epoch": 0.4763157894736842, + "grad_norm": 0.5546875, + "learning_rate": 0.00011637426900584796, + "loss": 3.71, + "step": 9050 + }, + { + "epoch": 0.47636842105263155, + "grad_norm": 0.625, + "learning_rate": 0.00011636257309941521, + "loss": 3.4626, + "step": 9051 + }, + { + "epoch": 0.47642105263157897, + "grad_norm": 0.47265625, + "learning_rate": 0.00011635087719298246, + "loss": 3.4631, + "step": 9052 + }, + { + "epoch": 0.4764736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00011633918128654971, + "loss": 3.4807, + "step": 9053 + }, + { + "epoch": 0.4765263157894737, + "grad_norm": 0.451171875, + "learning_rate": 0.00011632748538011697, + "loss": 3.2843, + "step": 9054 + }, + { + "epoch": 0.47657894736842105, + "grad_norm": 0.458984375, + "learning_rate": 0.00011631578947368422, + "loss": 3.4107, + "step": 9055 + }, + { + "epoch": 0.4766315789473684, + "grad_norm": 0.490234375, + "learning_rate": 0.00011630409356725147, + "loss": 3.5073, + "step": 9056 + }, + { + "epoch": 0.47668421052631577, + "grad_norm": 0.498046875, + "learning_rate": 0.00011629239766081872, + "loss": 3.2399, + "step": 9057 + }, + { + "epoch": 0.4767368421052632, + "grad_norm": 0.453125, + "learning_rate": 0.00011628070175438597, + "loss": 3.6659, + "step": 9058 + }, + { + "epoch": 0.47678947368421054, + "grad_norm": 0.46875, + "learning_rate": 0.00011626900584795322, + "loss": 3.58, + "step": 9059 + }, + { + "epoch": 0.4768421052631579, + "grad_norm": 0.48828125, + "learning_rate": 0.00011625730994152047, + "loss": 3.3735, + "step": 9060 + }, + { + "epoch": 0.47689473684210526, + "grad_norm": 0.48828125, + "learning_rate": 0.00011624561403508773, + "loss": 3.309, + "step": 9061 + }, + { + "epoch": 0.4769473684210526, + "grad_norm": 0.5625, + "learning_rate": 0.00011623391812865498, + "loss": 3.2818, + "step": 9062 + }, + { + "epoch": 0.477, + "grad_norm": 0.46484375, + "learning_rate": 0.00011622222222222223, + "loss": 3.4147, + "step": 9063 + }, + { + "epoch": 0.4770526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.00011621052631578948, + "loss": 3.8134, + "step": 9064 + }, + { + "epoch": 0.47710526315789475, + "grad_norm": 0.51171875, + "learning_rate": 0.00011619883040935673, + "loss": 3.6788, + "step": 9065 + }, + { + "epoch": 0.4771578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.00011618713450292398, + "loss": 3.5509, + "step": 9066 + }, + { + "epoch": 0.4772105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00011617543859649123, + "loss": 3.2985, + "step": 9067 + }, + { + "epoch": 0.47726315789473683, + "grad_norm": 0.5078125, + "learning_rate": 0.00011616374269005848, + "loss": 3.5453, + "step": 9068 + }, + { + "epoch": 0.4773157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00011615204678362574, + "loss": 3.4479, + "step": 9069 + }, + { + "epoch": 0.47736842105263155, + "grad_norm": 0.55078125, + "learning_rate": 0.00011614035087719299, + "loss": 3.2626, + "step": 9070 + }, + { + "epoch": 0.47742105263157897, + "grad_norm": 0.5, + "learning_rate": 0.00011612865497076024, + "loss": 3.376, + "step": 9071 + }, + { + "epoch": 0.47747368421052633, + "grad_norm": 0.52734375, + "learning_rate": 0.00011611695906432749, + "loss": 2.803, + "step": 9072 + }, + { + "epoch": 0.4775263157894737, + "grad_norm": 0.478515625, + "learning_rate": 0.00011610526315789474, + "loss": 3.4868, + "step": 9073 + }, + { + "epoch": 0.47757894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.00011609356725146199, + "loss": 3.4691, + "step": 9074 + }, + { + "epoch": 0.4776315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.00011608187134502924, + "loss": 3.4177, + "step": 9075 + }, + { + "epoch": 0.47768421052631577, + "grad_norm": 0.515625, + "learning_rate": 0.0001160701754385965, + "loss": 3.3002, + "step": 9076 + }, + { + "epoch": 0.4777368421052632, + "grad_norm": 0.57421875, + "learning_rate": 0.00011605847953216375, + "loss": 3.0756, + "step": 9077 + }, + { + "epoch": 0.47778947368421054, + "grad_norm": 0.494140625, + "learning_rate": 0.000116046783625731, + "loss": 3.6847, + "step": 9078 + }, + { + "epoch": 0.4778421052631579, + "grad_norm": 0.44921875, + "learning_rate": 0.00011603508771929825, + "loss": 3.4497, + "step": 9079 + }, + { + "epoch": 0.47789473684210526, + "grad_norm": 0.474609375, + "learning_rate": 0.0001160233918128655, + "loss": 3.7187, + "step": 9080 + }, + { + "epoch": 0.4779473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00011601169590643275, + "loss": 3.4765, + "step": 9081 + }, + { + "epoch": 0.478, + "grad_norm": 0.486328125, + "learning_rate": 0.000116, + "loss": 3.4912, + "step": 9082 + }, + { + "epoch": 0.4780526315789474, + "grad_norm": 0.5078125, + "learning_rate": 0.00011598830409356725, + "loss": 3.5518, + "step": 9083 + }, + { + "epoch": 0.47810526315789476, + "grad_norm": 0.52734375, + "learning_rate": 0.00011597660818713451, + "loss": 3.1606, + "step": 9084 + }, + { + "epoch": 0.4781578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00011596491228070176, + "loss": 3.4402, + "step": 9085 + }, + { + "epoch": 0.4782105263157895, + "grad_norm": 0.470703125, + "learning_rate": 0.00011595321637426901, + "loss": 3.0691, + "step": 9086 + }, + { + "epoch": 0.47826315789473683, + "grad_norm": 0.486328125, + "learning_rate": 0.00011594152046783626, + "loss": 3.4866, + "step": 9087 + }, + { + "epoch": 0.4783157894736842, + "grad_norm": 0.486328125, + "learning_rate": 0.00011592982456140351, + "loss": 3.6523, + "step": 9088 + }, + { + "epoch": 0.47836842105263155, + "grad_norm": 0.48828125, + "learning_rate": 0.00011591812865497076, + "loss": 3.5434, + "step": 9089 + }, + { + "epoch": 0.47842105263157897, + "grad_norm": 0.515625, + "learning_rate": 0.00011590643274853801, + "loss": 3.4116, + "step": 9090 + }, + { + "epoch": 0.47847368421052633, + "grad_norm": 0.5625, + "learning_rate": 0.00011589473684210527, + "loss": 3.0433, + "step": 9091 + }, + { + "epoch": 0.4785263157894737, + "grad_norm": 0.46875, + "learning_rate": 0.00011588304093567252, + "loss": 3.6456, + "step": 9092 + }, + { + "epoch": 0.47857894736842105, + "grad_norm": 0.5234375, + "learning_rate": 0.00011587134502923977, + "loss": 3.5581, + "step": 9093 + }, + { + "epoch": 0.4786315789473684, + "grad_norm": 0.455078125, + "learning_rate": 0.00011585964912280702, + "loss": 3.7936, + "step": 9094 + }, + { + "epoch": 0.47868421052631577, + "grad_norm": 0.49609375, + "learning_rate": 0.00011584795321637427, + "loss": 3.4054, + "step": 9095 + }, + { + "epoch": 0.4787368421052632, + "grad_norm": 0.494140625, + "learning_rate": 0.00011583625730994152, + "loss": 3.4717, + "step": 9096 + }, + { + "epoch": 0.47878947368421054, + "grad_norm": 0.55859375, + "learning_rate": 0.00011582456140350877, + "loss": 3.2854, + "step": 9097 + }, + { + "epoch": 0.4788421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00011581286549707603, + "loss": 2.805, + "step": 9098 + }, + { + "epoch": 0.47889473684210526, + "grad_norm": 0.490234375, + "learning_rate": 0.00011580116959064328, + "loss": 3.6916, + "step": 9099 + }, + { + "epoch": 0.4789473684210526, + "grad_norm": 0.494140625, + "learning_rate": 0.00011578947368421053, + "loss": 3.7782, + "step": 9100 + }, + { + "epoch": 0.479, + "grad_norm": 0.48828125, + "learning_rate": 0.00011577777777777778, + "loss": 3.7478, + "step": 9101 + }, + { + "epoch": 0.47905263157894734, + "grad_norm": 0.478515625, + "learning_rate": 0.00011576608187134503, + "loss": 3.8469, + "step": 9102 + }, + { + "epoch": 0.47910526315789476, + "grad_norm": 0.5234375, + "learning_rate": 0.00011575438596491228, + "loss": 3.4331, + "step": 9103 + }, + { + "epoch": 0.4791578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00011574269005847953, + "loss": 2.9784, + "step": 9104 + }, + { + "epoch": 0.4792105263157895, + "grad_norm": 0.515625, + "learning_rate": 0.00011573099415204678, + "loss": 3.3478, + "step": 9105 + }, + { + "epoch": 0.47926315789473684, + "grad_norm": 0.466796875, + "learning_rate": 0.00011571929824561405, + "loss": 3.7218, + "step": 9106 + }, + { + "epoch": 0.4793157894736842, + "grad_norm": 0.71875, + "learning_rate": 0.0001157076023391813, + "loss": 3.2069, + "step": 9107 + }, + { + "epoch": 0.47936842105263155, + "grad_norm": 0.470703125, + "learning_rate": 0.00011569590643274854, + "loss": 3.5252, + "step": 9108 + }, + { + "epoch": 0.47942105263157897, + "grad_norm": 0.56640625, + "learning_rate": 0.0001156842105263158, + "loss": 3.717, + "step": 9109 + }, + { + "epoch": 0.47947368421052633, + "grad_norm": 0.51953125, + "learning_rate": 0.00011567251461988304, + "loss": 3.6423, + "step": 9110 + }, + { + "epoch": 0.4795263157894737, + "grad_norm": 0.48828125, + "learning_rate": 0.00011566081871345029, + "loss": 3.38, + "step": 9111 + }, + { + "epoch": 0.47957894736842105, + "grad_norm": 0.47265625, + "learning_rate": 0.00011564912280701754, + "loss": 3.2921, + "step": 9112 + }, + { + "epoch": 0.4796315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.0001156374269005848, + "loss": 3.3233, + "step": 9113 + }, + { + "epoch": 0.47968421052631577, + "grad_norm": 0.53125, + "learning_rate": 0.00011562573099415206, + "loss": 3.5277, + "step": 9114 + }, + { + "epoch": 0.4797368421052632, + "grad_norm": 0.4765625, + "learning_rate": 0.0001156140350877193, + "loss": 3.6254, + "step": 9115 + }, + { + "epoch": 0.47978947368421054, + "grad_norm": 0.47265625, + "learning_rate": 0.00011560233918128655, + "loss": 3.4331, + "step": 9116 + }, + { + "epoch": 0.4798421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.0001155906432748538, + "loss": 3.4842, + "step": 9117 + }, + { + "epoch": 0.47989473684210526, + "grad_norm": 0.8046875, + "learning_rate": 0.00011557894736842105, + "loss": 3.2091, + "step": 9118 + }, + { + "epoch": 0.4799473684210526, + "grad_norm": 0.4453125, + "learning_rate": 0.0001155672514619883, + "loss": 3.4134, + "step": 9119 + }, + { + "epoch": 0.48, + "grad_norm": 0.47265625, + "learning_rate": 0.00011555555555555555, + "loss": 3.6012, + "step": 9120 + }, + { + "epoch": 0.48005263157894734, + "grad_norm": 0.453125, + "learning_rate": 0.00011554385964912282, + "loss": 3.4788, + "step": 9121 + }, + { + "epoch": 0.48010526315789476, + "grad_norm": 0.498046875, + "learning_rate": 0.00011553216374269007, + "loss": 3.6372, + "step": 9122 + }, + { + "epoch": 0.4801578947368421, + "grad_norm": 0.458984375, + "learning_rate": 0.00011552046783625732, + "loss": 3.5978, + "step": 9123 + }, + { + "epoch": 0.4802105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00011550877192982456, + "loss": 3.0891, + "step": 9124 + }, + { + "epoch": 0.48026315789473684, + "grad_norm": 0.4609375, + "learning_rate": 0.00011549707602339181, + "loss": 3.6508, + "step": 9125 + }, + { + "epoch": 0.4803157894736842, + "grad_norm": 0.53515625, + "learning_rate": 0.00011548538011695906, + "loss": 3.3912, + "step": 9126 + }, + { + "epoch": 0.48036842105263156, + "grad_norm": 0.5078125, + "learning_rate": 0.00011547368421052631, + "loss": 3.1011, + "step": 9127 + }, + { + "epoch": 0.48042105263157897, + "grad_norm": 0.478515625, + "learning_rate": 0.00011546198830409358, + "loss": 3.2599, + "step": 9128 + }, + { + "epoch": 0.48047368421052633, + "grad_norm": 0.474609375, + "learning_rate": 0.00011545029239766083, + "loss": 3.2123, + "step": 9129 + }, + { + "epoch": 0.4805263157894737, + "grad_norm": 0.484375, + "learning_rate": 0.00011543859649122808, + "loss": 2.6961, + "step": 9130 + }, + { + "epoch": 0.48057894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.00011542690058479533, + "loss": 3.5945, + "step": 9131 + }, + { + "epoch": 0.4806315789473684, + "grad_norm": 0.494140625, + "learning_rate": 0.00011541520467836258, + "loss": 3.5018, + "step": 9132 + }, + { + "epoch": 0.48068421052631577, + "grad_norm": 0.462890625, + "learning_rate": 0.00011540350877192982, + "loss": 3.2306, + "step": 9133 + }, + { + "epoch": 0.4807368421052632, + "grad_norm": 0.4765625, + "learning_rate": 0.00011539181286549707, + "loss": 3.8046, + "step": 9134 + }, + { + "epoch": 0.48078947368421054, + "grad_norm": 0.48046875, + "learning_rate": 0.00011538011695906434, + "loss": 3.6861, + "step": 9135 + }, + { + "epoch": 0.4808421052631579, + "grad_norm": 0.52734375, + "learning_rate": 0.00011536842105263159, + "loss": 3.8585, + "step": 9136 + }, + { + "epoch": 0.48089473684210526, + "grad_norm": 0.6171875, + "learning_rate": 0.00011535672514619884, + "loss": 3.3716, + "step": 9137 + }, + { + "epoch": 0.4809473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00011534502923976609, + "loss": 3.1637, + "step": 9138 + }, + { + "epoch": 0.481, + "grad_norm": 0.5078125, + "learning_rate": 0.00011533333333333334, + "loss": 3.2693, + "step": 9139 + }, + { + "epoch": 0.48105263157894734, + "grad_norm": 0.484375, + "learning_rate": 0.00011532163742690059, + "loss": 3.2127, + "step": 9140 + }, + { + "epoch": 0.48110526315789476, + "grad_norm": 0.546875, + "learning_rate": 0.00011530994152046784, + "loss": 3.1169, + "step": 9141 + }, + { + "epoch": 0.4811578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.00011529824561403508, + "loss": 3.426, + "step": 9142 + }, + { + "epoch": 0.4812105263157895, + "grad_norm": 0.53125, + "learning_rate": 0.00011528654970760235, + "loss": 3.7182, + "step": 9143 + }, + { + "epoch": 0.48126315789473684, + "grad_norm": 0.51953125, + "learning_rate": 0.0001152748538011696, + "loss": 3.4678, + "step": 9144 + }, + { + "epoch": 0.4813157894736842, + "grad_norm": 0.455078125, + "learning_rate": 0.00011526315789473685, + "loss": 3.1471, + "step": 9145 + }, + { + "epoch": 0.48136842105263156, + "grad_norm": 0.45703125, + "learning_rate": 0.0001152514619883041, + "loss": 3.3185, + "step": 9146 + }, + { + "epoch": 0.48142105263157897, + "grad_norm": 0.474609375, + "learning_rate": 0.00011523976608187135, + "loss": 3.5014, + "step": 9147 + }, + { + "epoch": 0.48147368421052633, + "grad_norm": 0.515625, + "learning_rate": 0.0001152280701754386, + "loss": 3.646, + "step": 9148 + }, + { + "epoch": 0.4815263157894737, + "grad_norm": 0.515625, + "learning_rate": 0.00011521637426900585, + "loss": 3.1561, + "step": 9149 + }, + { + "epoch": 0.48157894736842105, + "grad_norm": 0.494140625, + "learning_rate": 0.00011520467836257311, + "loss": 3.6787, + "step": 9150 + }, + { + "epoch": 0.4816315789473684, + "grad_norm": 0.5234375, + "learning_rate": 0.00011519298245614036, + "loss": 3.3147, + "step": 9151 + }, + { + "epoch": 0.48168421052631577, + "grad_norm": 0.48828125, + "learning_rate": 0.00011518128654970761, + "loss": 3.635, + "step": 9152 + }, + { + "epoch": 0.4817368421052632, + "grad_norm": 0.49609375, + "learning_rate": 0.00011516959064327486, + "loss": 3.1502, + "step": 9153 + }, + { + "epoch": 0.48178947368421055, + "grad_norm": 0.4921875, + "learning_rate": 0.00011515789473684211, + "loss": 3.6007, + "step": 9154 + }, + { + "epoch": 0.4818421052631579, + "grad_norm": 0.478515625, + "learning_rate": 0.00011514619883040936, + "loss": 3.1889, + "step": 9155 + }, + { + "epoch": 0.48189473684210526, + "grad_norm": 0.49609375, + "learning_rate": 0.0001151345029239766, + "loss": 3.8013, + "step": 9156 + }, + { + "epoch": 0.4819473684210526, + "grad_norm": 0.439453125, + "learning_rate": 0.00011512280701754386, + "loss": 3.1969, + "step": 9157 + }, + { + "epoch": 0.482, + "grad_norm": 0.44140625, + "learning_rate": 0.00011511111111111112, + "loss": 3.5822, + "step": 9158 + }, + { + "epoch": 0.48205263157894734, + "grad_norm": 0.482421875, + "learning_rate": 0.00011509941520467837, + "loss": 3.7649, + "step": 9159 + }, + { + "epoch": 0.48210526315789476, + "grad_norm": 0.47265625, + "learning_rate": 0.00011508771929824562, + "loss": 3.3364, + "step": 9160 + }, + { + "epoch": 0.4821578947368421, + "grad_norm": 0.474609375, + "learning_rate": 0.00011507602339181287, + "loss": 3.4539, + "step": 9161 + }, + { + "epoch": 0.4822105263157895, + "grad_norm": 0.484375, + "learning_rate": 0.00011506432748538012, + "loss": 3.6879, + "step": 9162 + }, + { + "epoch": 0.48226315789473684, + "grad_norm": 0.484375, + "learning_rate": 0.00011505263157894737, + "loss": 3.6611, + "step": 9163 + }, + { + "epoch": 0.4823157894736842, + "grad_norm": 0.76171875, + "learning_rate": 0.00011504093567251462, + "loss": 3.5642, + "step": 9164 + }, + { + "epoch": 0.48236842105263156, + "grad_norm": 0.466796875, + "learning_rate": 0.00011502923976608189, + "loss": 3.3856, + "step": 9165 + }, + { + "epoch": 0.482421052631579, + "grad_norm": 0.640625, + "learning_rate": 0.00011501754385964913, + "loss": 3.4382, + "step": 9166 + }, + { + "epoch": 0.48247368421052633, + "grad_norm": 0.546875, + "learning_rate": 0.00011500584795321638, + "loss": 3.4777, + "step": 9167 + }, + { + "epoch": 0.4825263157894737, + "grad_norm": 0.482421875, + "learning_rate": 0.00011499415204678363, + "loss": 3.791, + "step": 9168 + }, + { + "epoch": 0.48257894736842105, + "grad_norm": 0.51171875, + "learning_rate": 0.00011498245614035088, + "loss": 2.9428, + "step": 9169 + }, + { + "epoch": 0.4826315789473684, + "grad_norm": 0.484375, + "learning_rate": 0.00011497076023391813, + "loss": 3.67, + "step": 9170 + }, + { + "epoch": 0.48268421052631577, + "grad_norm": 0.490234375, + "learning_rate": 0.00011495906432748538, + "loss": 3.9987, + "step": 9171 + }, + { + "epoch": 0.48273684210526313, + "grad_norm": 0.474609375, + "learning_rate": 0.00011494736842105265, + "loss": 3.7126, + "step": 9172 + }, + { + "epoch": 0.48278947368421055, + "grad_norm": 0.474609375, + "learning_rate": 0.00011493567251461989, + "loss": 3.4345, + "step": 9173 + }, + { + "epoch": 0.4828421052631579, + "grad_norm": 0.494140625, + "learning_rate": 0.00011492397660818714, + "loss": 3.7166, + "step": 9174 + }, + { + "epoch": 0.48289473684210527, + "grad_norm": 0.4921875, + "learning_rate": 0.00011491228070175439, + "loss": 3.5606, + "step": 9175 + }, + { + "epoch": 0.4829473684210526, + "grad_norm": 0.490234375, + "learning_rate": 0.00011490058479532164, + "loss": 3.6806, + "step": 9176 + }, + { + "epoch": 0.483, + "grad_norm": 0.466796875, + "learning_rate": 0.00011488888888888889, + "loss": 3.2419, + "step": 9177 + }, + { + "epoch": 0.48305263157894734, + "grad_norm": 0.478515625, + "learning_rate": 0.00011487719298245614, + "loss": 3.7298, + "step": 9178 + }, + { + "epoch": 0.48310526315789476, + "grad_norm": 0.546875, + "learning_rate": 0.00011486549707602339, + "loss": 2.9487, + "step": 9179 + }, + { + "epoch": 0.4831578947368421, + "grad_norm": 0.640625, + "learning_rate": 0.00011485380116959066, + "loss": 3.5197, + "step": 9180 + }, + { + "epoch": 0.4832105263157895, + "grad_norm": 0.50390625, + "learning_rate": 0.0001148421052631579, + "loss": 3.7334, + "step": 9181 + }, + { + "epoch": 0.48326315789473684, + "grad_norm": 0.4453125, + "learning_rate": 0.00011483040935672515, + "loss": 3.4256, + "step": 9182 + }, + { + "epoch": 0.4833157894736842, + "grad_norm": 0.51953125, + "learning_rate": 0.0001148187134502924, + "loss": 3.3827, + "step": 9183 + }, + { + "epoch": 0.48336842105263156, + "grad_norm": 0.4453125, + "learning_rate": 0.00011480701754385965, + "loss": 3.2461, + "step": 9184 + }, + { + "epoch": 0.483421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.0001147953216374269, + "loss": 3.2738, + "step": 9185 + }, + { + "epoch": 0.48347368421052633, + "grad_norm": 0.55859375, + "learning_rate": 0.00011478362573099415, + "loss": 2.9585, + "step": 9186 + }, + { + "epoch": 0.4835263157894737, + "grad_norm": 0.48046875, + "learning_rate": 0.00011477192982456142, + "loss": 3.3982, + "step": 9187 + }, + { + "epoch": 0.48357894736842105, + "grad_norm": 0.59375, + "learning_rate": 0.00011476023391812867, + "loss": 3.5125, + "step": 9188 + }, + { + "epoch": 0.4836315789473684, + "grad_norm": 0.52734375, + "learning_rate": 0.00011474853801169591, + "loss": 3.7704, + "step": 9189 + }, + { + "epoch": 0.48368421052631577, + "grad_norm": 0.5625, + "learning_rate": 0.00011473684210526316, + "loss": 3.2893, + "step": 9190 + }, + { + "epoch": 0.48373684210526313, + "grad_norm": 0.48828125, + "learning_rate": 0.00011472514619883041, + "loss": 3.6156, + "step": 9191 + }, + { + "epoch": 0.48378947368421055, + "grad_norm": 0.5, + "learning_rate": 0.00011471345029239766, + "loss": 3.3175, + "step": 9192 + }, + { + "epoch": 0.4838421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00011470175438596491, + "loss": 3.3933, + "step": 9193 + }, + { + "epoch": 0.48389473684210527, + "grad_norm": 0.51171875, + "learning_rate": 0.00011469005847953216, + "loss": 3.2563, + "step": 9194 + }, + { + "epoch": 0.4839473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00011467836257309944, + "loss": 3.5347, + "step": 9195 + }, + { + "epoch": 0.484, + "grad_norm": 0.478515625, + "learning_rate": 0.00011466666666666667, + "loss": 3.2733, + "step": 9196 + }, + { + "epoch": 0.48405263157894735, + "grad_norm": 0.474609375, + "learning_rate": 0.00011465497076023392, + "loss": 3.5456, + "step": 9197 + }, + { + "epoch": 0.48410526315789476, + "grad_norm": 0.51171875, + "learning_rate": 0.00011464327485380117, + "loss": 3.5126, + "step": 9198 + }, + { + "epoch": 0.4841578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.00011463157894736842, + "loss": 3.625, + "step": 9199 + }, + { + "epoch": 0.4842105263157895, + "grad_norm": 0.46875, + "learning_rate": 0.00011461988304093567, + "loss": 3.2862, + "step": 9200 + }, + { + "epoch": 0.48426315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00011460818713450292, + "loss": 3.4815, + "step": 9201 + }, + { + "epoch": 0.4843157894736842, + "grad_norm": 0.47265625, + "learning_rate": 0.0001145964912280702, + "loss": 3.6415, + "step": 9202 + }, + { + "epoch": 0.48436842105263156, + "grad_norm": 0.54296875, + "learning_rate": 0.00011458479532163745, + "loss": 3.0409, + "step": 9203 + }, + { + "epoch": 0.484421052631579, + "grad_norm": 0.470703125, + "learning_rate": 0.00011457309941520468, + "loss": 3.5219, + "step": 9204 + }, + { + "epoch": 0.48447368421052633, + "grad_norm": 0.703125, + "learning_rate": 0.00011456140350877193, + "loss": 3.3695, + "step": 9205 + }, + { + "epoch": 0.4845263157894737, + "grad_norm": 0.494140625, + "learning_rate": 0.00011454970760233918, + "loss": 3.7229, + "step": 9206 + }, + { + "epoch": 0.48457894736842105, + "grad_norm": 0.50390625, + "learning_rate": 0.00011453801169590643, + "loss": 3.4062, + "step": 9207 + }, + { + "epoch": 0.4846315789473684, + "grad_norm": 0.48828125, + "learning_rate": 0.00011452631578947368, + "loss": 3.3171, + "step": 9208 + }, + { + "epoch": 0.4846842105263158, + "grad_norm": 0.4765625, + "learning_rate": 0.00011451461988304096, + "loss": 3.3418, + "step": 9209 + }, + { + "epoch": 0.48473684210526313, + "grad_norm": 0.5078125, + "learning_rate": 0.0001145029239766082, + "loss": 3.3726, + "step": 9210 + }, + { + "epoch": 0.48478947368421055, + "grad_norm": 0.50390625, + "learning_rate": 0.00011449122807017546, + "loss": 3.665, + "step": 9211 + }, + { + "epoch": 0.4848421052631579, + "grad_norm": 0.484375, + "learning_rate": 0.00011447953216374269, + "loss": 3.4819, + "step": 9212 + }, + { + "epoch": 0.48489473684210527, + "grad_norm": 0.494140625, + "learning_rate": 0.00011446783625730994, + "loss": 3.5509, + "step": 9213 + }, + { + "epoch": 0.4849473684210526, + "grad_norm": 0.482421875, + "learning_rate": 0.00011445614035087719, + "loss": 3.2844, + "step": 9214 + }, + { + "epoch": 0.485, + "grad_norm": 0.56640625, + "learning_rate": 0.00011444444444444444, + "loss": 3.1851, + "step": 9215 + }, + { + "epoch": 0.48505263157894735, + "grad_norm": 0.494140625, + "learning_rate": 0.00011443274853801169, + "loss": 3.3194, + "step": 9216 + }, + { + "epoch": 0.48510526315789476, + "grad_norm": 0.474609375, + "learning_rate": 0.00011442105263157897, + "loss": 3.5472, + "step": 9217 + }, + { + "epoch": 0.4851578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.00011440935672514622, + "loss": 3.8349, + "step": 9218 + }, + { + "epoch": 0.4852105263157895, + "grad_norm": 0.48046875, + "learning_rate": 0.00011439766081871345, + "loss": 3.0289, + "step": 9219 + }, + { + "epoch": 0.48526315789473684, + "grad_norm": 0.49609375, + "learning_rate": 0.0001143859649122807, + "loss": 3.0517, + "step": 9220 + }, + { + "epoch": 0.4853157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00011437426900584795, + "loss": 3.3144, + "step": 9221 + }, + { + "epoch": 0.48536842105263156, + "grad_norm": 0.462890625, + "learning_rate": 0.0001143625730994152, + "loss": 3.5171, + "step": 9222 + }, + { + "epoch": 0.4854210526315789, + "grad_norm": 0.48046875, + "learning_rate": 0.00011435087719298245, + "loss": 3.243, + "step": 9223 + }, + { + "epoch": 0.48547368421052634, + "grad_norm": 0.48046875, + "learning_rate": 0.00011433918128654973, + "loss": 3.2675, + "step": 9224 + }, + { + "epoch": 0.4855263157894737, + "grad_norm": 0.48828125, + "learning_rate": 0.00011432748538011698, + "loss": 3.7515, + "step": 9225 + }, + { + "epoch": 0.48557894736842105, + "grad_norm": 0.4921875, + "learning_rate": 0.00011431578947368423, + "loss": 2.897, + "step": 9226 + }, + { + "epoch": 0.4856315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.00011430409356725146, + "loss": 3.2972, + "step": 9227 + }, + { + "epoch": 0.4856842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.00011429239766081871, + "loss": 3.5704, + "step": 9228 + }, + { + "epoch": 0.48573684210526313, + "grad_norm": 0.53515625, + "learning_rate": 0.00011428070175438596, + "loss": 3.4064, + "step": 9229 + }, + { + "epoch": 0.48578947368421055, + "grad_norm": 0.451171875, + "learning_rate": 0.00011426900584795321, + "loss": 3.2525, + "step": 9230 + }, + { + "epoch": 0.4858421052631579, + "grad_norm": 0.6796875, + "learning_rate": 0.00011425730994152046, + "loss": 3.2707, + "step": 9231 + }, + { + "epoch": 0.48589473684210527, + "grad_norm": 0.474609375, + "learning_rate": 0.00011424561403508774, + "loss": 3.469, + "step": 9232 + }, + { + "epoch": 0.48594736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00011423391812865499, + "loss": 3.0657, + "step": 9233 + }, + { + "epoch": 0.486, + "grad_norm": 0.50390625, + "learning_rate": 0.00011422222222222224, + "loss": 3.1292, + "step": 9234 + }, + { + "epoch": 0.48605263157894735, + "grad_norm": 0.59765625, + "learning_rate": 0.00011421052631578947, + "loss": 3.2767, + "step": 9235 + }, + { + "epoch": 0.48610526315789476, + "grad_norm": 0.45703125, + "learning_rate": 0.00011419883040935672, + "loss": 3.7209, + "step": 9236 + }, + { + "epoch": 0.4861578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.00011418713450292397, + "loss": 3.3537, + "step": 9237 + }, + { + "epoch": 0.4862105263157895, + "grad_norm": 0.482421875, + "learning_rate": 0.00011417543859649122, + "loss": 3.548, + "step": 9238 + }, + { + "epoch": 0.48626315789473684, + "grad_norm": 0.490234375, + "learning_rate": 0.0001141637426900585, + "loss": 3.5064, + "step": 9239 + }, + { + "epoch": 0.4863157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.00011415204678362575, + "loss": 3.5724, + "step": 9240 + }, + { + "epoch": 0.48636842105263156, + "grad_norm": 0.470703125, + "learning_rate": 0.000114140350877193, + "loss": 3.4652, + "step": 9241 + }, + { + "epoch": 0.4864210526315789, + "grad_norm": 0.5390625, + "learning_rate": 0.00011412865497076023, + "loss": 4.169, + "step": 9242 + }, + { + "epoch": 0.48647368421052634, + "grad_norm": 0.5625, + "learning_rate": 0.00011411695906432748, + "loss": 3.6411, + "step": 9243 + }, + { + "epoch": 0.4865263157894737, + "grad_norm": 0.5390625, + "learning_rate": 0.00011410526315789473, + "loss": 3.6056, + "step": 9244 + }, + { + "epoch": 0.48657894736842106, + "grad_norm": 0.498046875, + "learning_rate": 0.00011409356725146198, + "loss": 3.6905, + "step": 9245 + }, + { + "epoch": 0.4866315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.00011408187134502923, + "loss": 3.262, + "step": 9246 + }, + { + "epoch": 0.4866842105263158, + "grad_norm": 0.52734375, + "learning_rate": 0.00011407017543859651, + "loss": 3.0168, + "step": 9247 + }, + { + "epoch": 0.48673684210526313, + "grad_norm": 0.5078125, + "learning_rate": 0.00011405847953216376, + "loss": 3.2465, + "step": 9248 + }, + { + "epoch": 0.48678947368421055, + "grad_norm": 0.55859375, + "learning_rate": 0.00011404678362573101, + "loss": 3.6763, + "step": 9249 + }, + { + "epoch": 0.4868421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00011403508771929824, + "loss": 3.4069, + "step": 9250 + }, + { + "epoch": 0.48689473684210527, + "grad_norm": 0.474609375, + "learning_rate": 0.0001140233918128655, + "loss": 3.5952, + "step": 9251 + }, + { + "epoch": 0.48694736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00011401169590643274, + "loss": 3.5378, + "step": 9252 + }, + { + "epoch": 0.487, + "grad_norm": 0.490234375, + "learning_rate": 0.00011399999999999999, + "loss": 3.5324, + "step": 9253 + }, + { + "epoch": 0.48705263157894735, + "grad_norm": 0.5546875, + "learning_rate": 0.00011398830409356727, + "loss": 3.2419, + "step": 9254 + }, + { + "epoch": 0.48710526315789476, + "grad_norm": 0.5078125, + "learning_rate": 0.00011397660818713452, + "loss": 3.3271, + "step": 9255 + }, + { + "epoch": 0.4871578947368421, + "grad_norm": 0.478515625, + "learning_rate": 0.00011396491228070177, + "loss": 3.5472, + "step": 9256 + }, + { + "epoch": 0.4872105263157895, + "grad_norm": 0.46875, + "learning_rate": 0.00011395321637426902, + "loss": 3.772, + "step": 9257 + }, + { + "epoch": 0.48726315789473684, + "grad_norm": 0.47265625, + "learning_rate": 0.00011394152046783625, + "loss": 3.7087, + "step": 9258 + }, + { + "epoch": 0.4873157894736842, + "grad_norm": 0.51953125, + "learning_rate": 0.0001139298245614035, + "loss": 3.1841, + "step": 9259 + }, + { + "epoch": 0.48736842105263156, + "grad_norm": 0.484375, + "learning_rate": 0.00011391812865497075, + "loss": 3.5479, + "step": 9260 + }, + { + "epoch": 0.4874210526315789, + "grad_norm": 0.453125, + "learning_rate": 0.00011390643274853803, + "loss": 3.7015, + "step": 9261 + }, + { + "epoch": 0.48747368421052634, + "grad_norm": 0.5, + "learning_rate": 0.00011389473684210528, + "loss": 3.493, + "step": 9262 + }, + { + "epoch": 0.4875263157894737, + "grad_norm": 0.5546875, + "learning_rate": 0.00011388304093567253, + "loss": 3.4732, + "step": 9263 + }, + { + "epoch": 0.48757894736842106, + "grad_norm": 0.498046875, + "learning_rate": 0.00011387134502923978, + "loss": 3.5144, + "step": 9264 + }, + { + "epoch": 0.4876315789473684, + "grad_norm": 0.49609375, + "learning_rate": 0.00011385964912280701, + "loss": 3.1421, + "step": 9265 + }, + { + "epoch": 0.4876842105263158, + "grad_norm": 0.45703125, + "learning_rate": 0.00011384795321637426, + "loss": 3.645, + "step": 9266 + }, + { + "epoch": 0.48773684210526314, + "grad_norm": 0.490234375, + "learning_rate": 0.00011383625730994151, + "loss": 3.4394, + "step": 9267 + }, + { + "epoch": 0.48778947368421055, + "grad_norm": 0.51171875, + "learning_rate": 0.00011382456140350876, + "loss": 3.3068, + "step": 9268 + }, + { + "epoch": 0.4878421052631579, + "grad_norm": 0.482421875, + "learning_rate": 0.00011381286549707604, + "loss": 3.4755, + "step": 9269 + }, + { + "epoch": 0.48789473684210527, + "grad_norm": 0.5078125, + "learning_rate": 0.00011380116959064329, + "loss": 3.6749, + "step": 9270 + }, + { + "epoch": 0.48794736842105263, + "grad_norm": 0.62109375, + "learning_rate": 0.00011378947368421054, + "loss": 3.3428, + "step": 9271 + }, + { + "epoch": 0.488, + "grad_norm": 0.54296875, + "learning_rate": 0.00011377777777777779, + "loss": 3.2708, + "step": 9272 + }, + { + "epoch": 0.48805263157894735, + "grad_norm": 0.486328125, + "learning_rate": 0.00011376608187134503, + "loss": 3.4544, + "step": 9273 + }, + { + "epoch": 0.48810526315789476, + "grad_norm": 0.4609375, + "learning_rate": 0.00011375438596491227, + "loss": 3.2635, + "step": 9274 + }, + { + "epoch": 0.4881578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00011374269005847952, + "loss": 3.3101, + "step": 9275 + }, + { + "epoch": 0.4882105263157895, + "grad_norm": 0.490234375, + "learning_rate": 0.0001137309941520468, + "loss": 3.3012, + "step": 9276 + }, + { + "epoch": 0.48826315789473684, + "grad_norm": 0.63671875, + "learning_rate": 0.00011371929824561405, + "loss": 3.5111, + "step": 9277 + }, + { + "epoch": 0.4883157894736842, + "grad_norm": 0.455078125, + "learning_rate": 0.0001137076023391813, + "loss": 3.5442, + "step": 9278 + }, + { + "epoch": 0.48836842105263156, + "grad_norm": 0.5625, + "learning_rate": 0.00011369590643274855, + "loss": 3.4546, + "step": 9279 + }, + { + "epoch": 0.4884210526315789, + "grad_norm": 0.48828125, + "learning_rate": 0.0001136842105263158, + "loss": 3.4263, + "step": 9280 + }, + { + "epoch": 0.48847368421052634, + "grad_norm": 0.51953125, + "learning_rate": 0.00011367251461988304, + "loss": 3.7467, + "step": 9281 + }, + { + "epoch": 0.4885263157894737, + "grad_norm": 0.58984375, + "learning_rate": 0.00011366081871345029, + "loss": 3.5655, + "step": 9282 + }, + { + "epoch": 0.48857894736842106, + "grad_norm": 0.5234375, + "learning_rate": 0.00011364912280701753, + "loss": 3.9713, + "step": 9283 + }, + { + "epoch": 0.4886315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.00011363742690058481, + "loss": 3.5561, + "step": 9284 + }, + { + "epoch": 0.4886842105263158, + "grad_norm": 0.53515625, + "learning_rate": 0.00011362573099415206, + "loss": 3.564, + "step": 9285 + }, + { + "epoch": 0.48873684210526314, + "grad_norm": 0.462890625, + "learning_rate": 0.00011361403508771931, + "loss": 3.2213, + "step": 9286 + }, + { + "epoch": 0.48878947368421055, + "grad_norm": 0.470703125, + "learning_rate": 0.00011360233918128656, + "loss": 3.2887, + "step": 9287 + }, + { + "epoch": 0.4888421052631579, + "grad_norm": 0.4765625, + "learning_rate": 0.0001135906432748538, + "loss": 3.3462, + "step": 9288 + }, + { + "epoch": 0.48889473684210527, + "grad_norm": 0.51953125, + "learning_rate": 0.00011357894736842105, + "loss": 3.5618, + "step": 9289 + }, + { + "epoch": 0.48894736842105263, + "grad_norm": 0.4921875, + "learning_rate": 0.0001135672514619883, + "loss": 4.0424, + "step": 9290 + }, + { + "epoch": 0.489, + "grad_norm": 0.5390625, + "learning_rate": 0.00011355555555555557, + "loss": 3.4522, + "step": 9291 + }, + { + "epoch": 0.48905263157894735, + "grad_norm": 0.91015625, + "learning_rate": 0.00011354385964912282, + "loss": 3.3593, + "step": 9292 + }, + { + "epoch": 0.4891052631578947, + "grad_norm": 0.486328125, + "learning_rate": 0.00011353216374269007, + "loss": 3.0732, + "step": 9293 + }, + { + "epoch": 0.4891578947368421, + "grad_norm": 0.470703125, + "learning_rate": 0.00011352046783625732, + "loss": 3.1879, + "step": 9294 + }, + { + "epoch": 0.4892105263157895, + "grad_norm": 0.478515625, + "learning_rate": 0.00011350877192982457, + "loss": 3.5698, + "step": 9295 + }, + { + "epoch": 0.48926315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.0001134970760233918, + "loss": 3.5278, + "step": 9296 + }, + { + "epoch": 0.4893157894736842, + "grad_norm": 0.984375, + "learning_rate": 0.00011348538011695906, + "loss": 2.8191, + "step": 9297 + }, + { + "epoch": 0.48936842105263156, + "grad_norm": 0.486328125, + "learning_rate": 0.00011347368421052633, + "loss": 3.3348, + "step": 9298 + }, + { + "epoch": 0.4894210526315789, + "grad_norm": 0.462890625, + "learning_rate": 0.00011346198830409358, + "loss": 3.7859, + "step": 9299 + }, + { + "epoch": 0.48947368421052634, + "grad_norm": 0.490234375, + "learning_rate": 0.00011345029239766083, + "loss": 3.009, + "step": 9300 + }, + { + "epoch": 0.4895263157894737, + "grad_norm": 0.4609375, + "learning_rate": 0.00011343859649122808, + "loss": 3.2697, + "step": 9301 + }, + { + "epoch": 0.48957894736842106, + "grad_norm": 0.51171875, + "learning_rate": 0.00011342690058479533, + "loss": 3.6648, + "step": 9302 + }, + { + "epoch": 0.4896315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.00011341520467836258, + "loss": 2.8873, + "step": 9303 + }, + { + "epoch": 0.4896842105263158, + "grad_norm": 0.55078125, + "learning_rate": 0.00011340350877192982, + "loss": 3.6181, + "step": 9304 + }, + { + "epoch": 0.48973684210526314, + "grad_norm": 0.5078125, + "learning_rate": 0.00011339181286549707, + "loss": 4.0031, + "step": 9305 + }, + { + "epoch": 0.48978947368421055, + "grad_norm": 0.49609375, + "learning_rate": 0.00011338011695906434, + "loss": 3.3101, + "step": 9306 + }, + { + "epoch": 0.4898421052631579, + "grad_norm": 0.490234375, + "learning_rate": 0.00011336842105263159, + "loss": 3.4935, + "step": 9307 + }, + { + "epoch": 0.48989473684210527, + "grad_norm": 0.54296875, + "learning_rate": 0.00011335672514619884, + "loss": 3.3826, + "step": 9308 + }, + { + "epoch": 0.48994736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00011334502923976609, + "loss": 3.7421, + "step": 9309 + }, + { + "epoch": 0.49, + "grad_norm": 0.486328125, + "learning_rate": 0.00011333333333333334, + "loss": 3.3759, + "step": 9310 + }, + { + "epoch": 0.49005263157894735, + "grad_norm": 0.474609375, + "learning_rate": 0.00011332163742690058, + "loss": 3.011, + "step": 9311 + }, + { + "epoch": 0.4901052631578947, + "grad_norm": 0.66796875, + "learning_rate": 0.00011330994152046783, + "loss": 3.4871, + "step": 9312 + }, + { + "epoch": 0.4901578947368421, + "grad_norm": 0.48046875, + "learning_rate": 0.0001132982456140351, + "loss": 3.2767, + "step": 9313 + }, + { + "epoch": 0.4902105263157895, + "grad_norm": 0.46875, + "learning_rate": 0.00011328654970760235, + "loss": 3.4719, + "step": 9314 + }, + { + "epoch": 0.49026315789473685, + "grad_norm": 0.5, + "learning_rate": 0.0001132748538011696, + "loss": 3.7055, + "step": 9315 + }, + { + "epoch": 0.4903157894736842, + "grad_norm": 0.49609375, + "learning_rate": 0.00011326315789473685, + "loss": 3.7898, + "step": 9316 + }, + { + "epoch": 0.49036842105263156, + "grad_norm": 0.4921875, + "learning_rate": 0.0001132514619883041, + "loss": 3.46, + "step": 9317 + }, + { + "epoch": 0.4904210526315789, + "grad_norm": 0.455078125, + "learning_rate": 0.00011323976608187135, + "loss": 3.3505, + "step": 9318 + }, + { + "epoch": 0.49047368421052634, + "grad_norm": 0.51953125, + "learning_rate": 0.00011322807017543859, + "loss": 3.0257, + "step": 9319 + }, + { + "epoch": 0.4905263157894737, + "grad_norm": 0.482421875, + "learning_rate": 0.00011321637426900584, + "loss": 3.2051, + "step": 9320 + }, + { + "epoch": 0.49057894736842106, + "grad_norm": 0.5, + "learning_rate": 0.00011320467836257311, + "loss": 3.1096, + "step": 9321 + }, + { + "epoch": 0.4906315789473684, + "grad_norm": 0.4765625, + "learning_rate": 0.00011319298245614036, + "loss": 3.573, + "step": 9322 + }, + { + "epoch": 0.4906842105263158, + "grad_norm": 0.4765625, + "learning_rate": 0.00011318128654970761, + "loss": 3.2289, + "step": 9323 + }, + { + "epoch": 0.49073684210526314, + "grad_norm": 0.478515625, + "learning_rate": 0.00011316959064327486, + "loss": 3.4206, + "step": 9324 + }, + { + "epoch": 0.49078947368421055, + "grad_norm": 0.51171875, + "learning_rate": 0.00011315789473684211, + "loss": 3.5143, + "step": 9325 + }, + { + "epoch": 0.4908421052631579, + "grad_norm": 0.478515625, + "learning_rate": 0.00011314619883040936, + "loss": 3.1865, + "step": 9326 + }, + { + "epoch": 0.4908947368421053, + "grad_norm": 0.53515625, + "learning_rate": 0.0001131345029239766, + "loss": 3.336, + "step": 9327 + }, + { + "epoch": 0.49094736842105263, + "grad_norm": 0.462890625, + "learning_rate": 0.00011312280701754387, + "loss": 3.4238, + "step": 9328 + }, + { + "epoch": 0.491, + "grad_norm": 0.490234375, + "learning_rate": 0.00011311111111111112, + "loss": 3.7115, + "step": 9329 + }, + { + "epoch": 0.49105263157894735, + "grad_norm": 0.51171875, + "learning_rate": 0.00011309941520467837, + "loss": 2.9332, + "step": 9330 + }, + { + "epoch": 0.4911052631578947, + "grad_norm": 0.46484375, + "learning_rate": 0.00011308771929824562, + "loss": 3.9394, + "step": 9331 + }, + { + "epoch": 0.4911578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.00011307602339181287, + "loss": 3.3815, + "step": 9332 + }, + { + "epoch": 0.4912105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00011306432748538012, + "loss": 3.4794, + "step": 9333 + }, + { + "epoch": 0.49126315789473685, + "grad_norm": 0.44921875, + "learning_rate": 0.00011305263157894736, + "loss": 3.4318, + "step": 9334 + }, + { + "epoch": 0.4913157894736842, + "grad_norm": 0.4765625, + "learning_rate": 0.00011304093567251464, + "loss": 3.9106, + "step": 9335 + }, + { + "epoch": 0.49136842105263157, + "grad_norm": 0.52734375, + "learning_rate": 0.00011302923976608189, + "loss": 3.6757, + "step": 9336 + }, + { + "epoch": 0.4914210526315789, + "grad_norm": 0.5390625, + "learning_rate": 0.00011301754385964913, + "loss": 3.5983, + "step": 9337 + }, + { + "epoch": 0.49147368421052634, + "grad_norm": 0.515625, + "learning_rate": 0.00011300584795321638, + "loss": 3.2803, + "step": 9338 + }, + { + "epoch": 0.4915263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00011299415204678363, + "loss": 3.8402, + "step": 9339 + }, + { + "epoch": 0.49157894736842106, + "grad_norm": 0.58984375, + "learning_rate": 0.00011298245614035088, + "loss": 3.6812, + "step": 9340 + }, + { + "epoch": 0.4916315789473684, + "grad_norm": 0.51953125, + "learning_rate": 0.00011297076023391813, + "loss": 3.4265, + "step": 9341 + }, + { + "epoch": 0.4916842105263158, + "grad_norm": 0.458984375, + "learning_rate": 0.00011295906432748537, + "loss": 3.5489, + "step": 9342 + }, + { + "epoch": 0.49173684210526314, + "grad_norm": 0.462890625, + "learning_rate": 0.00011294736842105265, + "loss": 3.2069, + "step": 9343 + }, + { + "epoch": 0.4917894736842105, + "grad_norm": 0.51171875, + "learning_rate": 0.0001129356725146199, + "loss": 3.315, + "step": 9344 + }, + { + "epoch": 0.4918421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00011292397660818714, + "loss": 3.7157, + "step": 9345 + }, + { + "epoch": 0.4918947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.0001129122807017544, + "loss": 2.9914, + "step": 9346 + }, + { + "epoch": 0.49194736842105263, + "grad_norm": 0.474609375, + "learning_rate": 0.00011290058479532164, + "loss": 3.217, + "step": 9347 + }, + { + "epoch": 0.492, + "grad_norm": 0.4765625, + "learning_rate": 0.0001128888888888889, + "loss": 3.4367, + "step": 9348 + }, + { + "epoch": 0.49205263157894735, + "grad_norm": 0.53515625, + "learning_rate": 0.00011287719298245614, + "loss": 3.6261, + "step": 9349 + }, + { + "epoch": 0.4921052631578947, + "grad_norm": 0.49609375, + "learning_rate": 0.0001128654970760234, + "loss": 3.0504, + "step": 9350 + }, + { + "epoch": 0.4921578947368421, + "grad_norm": 0.50390625, + "learning_rate": 0.00011285380116959066, + "loss": 3.4492, + "step": 9351 + }, + { + "epoch": 0.4922105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.0001128421052631579, + "loss": 3.5418, + "step": 9352 + }, + { + "epoch": 0.49226315789473685, + "grad_norm": 0.51953125, + "learning_rate": 0.00011283040935672516, + "loss": 3.4033, + "step": 9353 + }, + { + "epoch": 0.4923157894736842, + "grad_norm": 0.5546875, + "learning_rate": 0.0001128187134502924, + "loss": 3.2984, + "step": 9354 + }, + { + "epoch": 0.49236842105263157, + "grad_norm": 0.5, + "learning_rate": 0.00011280701754385965, + "loss": 3.2594, + "step": 9355 + }, + { + "epoch": 0.4924210526315789, + "grad_norm": 0.47265625, + "learning_rate": 0.0001127953216374269, + "loss": 3.9201, + "step": 9356 + }, + { + "epoch": 0.49247368421052634, + "grad_norm": 0.5234375, + "learning_rate": 0.00011278362573099414, + "loss": 3.5673, + "step": 9357 + }, + { + "epoch": 0.4925263157894737, + "grad_norm": 0.4609375, + "learning_rate": 0.00011277192982456142, + "loss": 3.1619, + "step": 9358 + }, + { + "epoch": 0.49257894736842106, + "grad_norm": 0.466796875, + "learning_rate": 0.00011276023391812867, + "loss": 3.5902, + "step": 9359 + }, + { + "epoch": 0.4926315789473684, + "grad_norm": 0.52734375, + "learning_rate": 0.00011274853801169592, + "loss": 3.4192, + "step": 9360 + }, + { + "epoch": 0.4926842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00011273684210526317, + "loss": 3.1668, + "step": 9361 + }, + { + "epoch": 0.49273684210526314, + "grad_norm": 0.470703125, + "learning_rate": 0.00011272514619883042, + "loss": 3.1105, + "step": 9362 + }, + { + "epoch": 0.4927894736842105, + "grad_norm": 0.4765625, + "learning_rate": 0.00011271345029239766, + "loss": 3.2016, + "step": 9363 + }, + { + "epoch": 0.4928421052631579, + "grad_norm": 0.458984375, + "learning_rate": 0.00011270175438596491, + "loss": 3.9192, + "step": 9364 + }, + { + "epoch": 0.4928947368421053, + "grad_norm": 0.451171875, + "learning_rate": 0.00011269005847953218, + "loss": 3.2979, + "step": 9365 + }, + { + "epoch": 0.49294736842105263, + "grad_norm": 0.5, + "learning_rate": 0.00011267836257309943, + "loss": 3.4994, + "step": 9366 + }, + { + "epoch": 0.493, + "grad_norm": 0.52734375, + "learning_rate": 0.00011266666666666668, + "loss": 3.653, + "step": 9367 + }, + { + "epoch": 0.49305263157894735, + "grad_norm": 0.490234375, + "learning_rate": 0.00011265497076023393, + "loss": 3.2225, + "step": 9368 + }, + { + "epoch": 0.4931052631578947, + "grad_norm": 0.470703125, + "learning_rate": 0.00011264327485380118, + "loss": 3.2827, + "step": 9369 + }, + { + "epoch": 0.49315789473684213, + "grad_norm": 0.48046875, + "learning_rate": 0.00011263157894736843, + "loss": 3.7831, + "step": 9370 + }, + { + "epoch": 0.4932105263157895, + "grad_norm": 0.52734375, + "learning_rate": 0.00011261988304093568, + "loss": 3.6315, + "step": 9371 + }, + { + "epoch": 0.49326315789473685, + "grad_norm": 0.5, + "learning_rate": 0.00011260818713450294, + "loss": 3.5768, + "step": 9372 + }, + { + "epoch": 0.4933157894736842, + "grad_norm": 0.494140625, + "learning_rate": 0.00011259649122807019, + "loss": 3.6856, + "step": 9373 + }, + { + "epoch": 0.49336842105263157, + "grad_norm": 0.482421875, + "learning_rate": 0.00011258479532163744, + "loss": 3.9119, + "step": 9374 + }, + { + "epoch": 0.4934210526315789, + "grad_norm": 0.5, + "learning_rate": 0.00011257309941520469, + "loss": 3.0217, + "step": 9375 + }, + { + "epoch": 0.49347368421052634, + "grad_norm": 0.47265625, + "learning_rate": 0.00011256140350877194, + "loss": 3.5271, + "step": 9376 + }, + { + "epoch": 0.4935263157894737, + "grad_norm": 0.486328125, + "learning_rate": 0.00011254970760233919, + "loss": 3.1175, + "step": 9377 + }, + { + "epoch": 0.49357894736842106, + "grad_norm": 0.48046875, + "learning_rate": 0.00011253801169590644, + "loss": 3.7247, + "step": 9378 + }, + { + "epoch": 0.4936315789473684, + "grad_norm": 0.494140625, + "learning_rate": 0.00011252631578947369, + "loss": 3.5183, + "step": 9379 + }, + { + "epoch": 0.4936842105263158, + "grad_norm": 0.51171875, + "learning_rate": 0.00011251461988304095, + "loss": 3.0784, + "step": 9380 + }, + { + "epoch": 0.49373684210526314, + "grad_norm": 0.455078125, + "learning_rate": 0.0001125029239766082, + "loss": 3.3044, + "step": 9381 + }, + { + "epoch": 0.4937894736842105, + "grad_norm": 0.4453125, + "learning_rate": 0.00011249122807017545, + "loss": 3.5768, + "step": 9382 + }, + { + "epoch": 0.4938421052631579, + "grad_norm": 0.47265625, + "learning_rate": 0.0001124795321637427, + "loss": 3.2808, + "step": 9383 + }, + { + "epoch": 0.4938947368421053, + "grad_norm": 0.482421875, + "learning_rate": 0.00011246783625730995, + "loss": 3.4141, + "step": 9384 + }, + { + "epoch": 0.49394736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.0001124561403508772, + "loss": 3.7597, + "step": 9385 + }, + { + "epoch": 0.494, + "grad_norm": 0.515625, + "learning_rate": 0.00011244444444444445, + "loss": 3.1712, + "step": 9386 + }, + { + "epoch": 0.49405263157894735, + "grad_norm": 0.5078125, + "learning_rate": 0.00011243274853801171, + "loss": 2.7324, + "step": 9387 + }, + { + "epoch": 0.4941052631578947, + "grad_norm": 0.486328125, + "learning_rate": 0.00011242105263157896, + "loss": 3.4415, + "step": 9388 + }, + { + "epoch": 0.49415789473684213, + "grad_norm": 0.7421875, + "learning_rate": 0.00011240935672514621, + "loss": 3.5289, + "step": 9389 + }, + { + "epoch": 0.4942105263157895, + "grad_norm": 0.49609375, + "learning_rate": 0.00011239766081871346, + "loss": 3.2951, + "step": 9390 + }, + { + "epoch": 0.49426315789473685, + "grad_norm": 0.462890625, + "learning_rate": 0.00011238596491228071, + "loss": 3.7891, + "step": 9391 + }, + { + "epoch": 0.4943157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.00011237426900584796, + "loss": 3.0268, + "step": 9392 + }, + { + "epoch": 0.49436842105263157, + "grad_norm": 0.478515625, + "learning_rate": 0.0001123625730994152, + "loss": 3.6843, + "step": 9393 + }, + { + "epoch": 0.49442105263157893, + "grad_norm": 0.5703125, + "learning_rate": 0.00011235087719298246, + "loss": 3.2616, + "step": 9394 + }, + { + "epoch": 0.49447368421052634, + "grad_norm": 0.466796875, + "learning_rate": 0.00011233918128654972, + "loss": 3.4037, + "step": 9395 + }, + { + "epoch": 0.4945263157894737, + "grad_norm": 0.48046875, + "learning_rate": 0.00011232748538011697, + "loss": 3.6935, + "step": 9396 + }, + { + "epoch": 0.49457894736842106, + "grad_norm": 0.51171875, + "learning_rate": 0.00011231578947368422, + "loss": 3.8745, + "step": 9397 + }, + { + "epoch": 0.4946315789473684, + "grad_norm": 0.546875, + "learning_rate": 0.00011230409356725147, + "loss": 3.6255, + "step": 9398 + }, + { + "epoch": 0.4946842105263158, + "grad_norm": 0.55859375, + "learning_rate": 0.00011229239766081872, + "loss": 3.8887, + "step": 9399 + }, + { + "epoch": 0.49473684210526314, + "grad_norm": 0.49609375, + "learning_rate": 0.00011228070175438597, + "loss": 3.8255, + "step": 9400 + }, + { + "epoch": 0.4947894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.00011226900584795322, + "loss": 3.3574, + "step": 9401 + }, + { + "epoch": 0.4948421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00011225730994152048, + "loss": 3.4195, + "step": 9402 + }, + { + "epoch": 0.4948947368421053, + "grad_norm": 0.47265625, + "learning_rate": 0.00011224561403508773, + "loss": 3.7846, + "step": 9403 + }, + { + "epoch": 0.49494736842105264, + "grad_norm": 0.48828125, + "learning_rate": 0.00011223391812865498, + "loss": 3.859, + "step": 9404 + }, + { + "epoch": 0.495, + "grad_norm": 0.51171875, + "learning_rate": 0.00011222222222222223, + "loss": 3.2391, + "step": 9405 + }, + { + "epoch": 0.49505263157894736, + "grad_norm": 0.5625, + "learning_rate": 0.00011221052631578948, + "loss": 3.3758, + "step": 9406 + }, + { + "epoch": 0.4951052631578947, + "grad_norm": 0.65234375, + "learning_rate": 0.00011219883040935673, + "loss": 3.4031, + "step": 9407 + }, + { + "epoch": 0.49515789473684213, + "grad_norm": 0.45703125, + "learning_rate": 0.00011218713450292398, + "loss": 3.4072, + "step": 9408 + }, + { + "epoch": 0.4952105263157895, + "grad_norm": 0.482421875, + "learning_rate": 0.00011217543859649123, + "loss": 3.7037, + "step": 9409 + }, + { + "epoch": 0.49526315789473685, + "grad_norm": 0.466796875, + "learning_rate": 0.00011216374269005849, + "loss": 3.1308, + "step": 9410 + }, + { + "epoch": 0.4953157894736842, + "grad_norm": 0.494140625, + "learning_rate": 0.00011215204678362574, + "loss": 3.2931, + "step": 9411 + }, + { + "epoch": 0.49536842105263157, + "grad_norm": 0.5, + "learning_rate": 0.00011214035087719299, + "loss": 3.5242, + "step": 9412 + }, + { + "epoch": 0.49542105263157893, + "grad_norm": 0.447265625, + "learning_rate": 0.00011212865497076024, + "loss": 3.7611, + "step": 9413 + }, + { + "epoch": 0.4954736842105263, + "grad_norm": 0.486328125, + "learning_rate": 0.00011211695906432749, + "loss": 3.3379, + "step": 9414 + }, + { + "epoch": 0.4955263157894737, + "grad_norm": 0.51953125, + "learning_rate": 0.00011210526315789474, + "loss": 3.0751, + "step": 9415 + }, + { + "epoch": 0.49557894736842106, + "grad_norm": 0.62890625, + "learning_rate": 0.00011209356725146199, + "loss": 3.1391, + "step": 9416 + }, + { + "epoch": 0.4956315789473684, + "grad_norm": 0.5, + "learning_rate": 0.00011208187134502925, + "loss": 3.1827, + "step": 9417 + }, + { + "epoch": 0.4956842105263158, + "grad_norm": 0.48828125, + "learning_rate": 0.0001120701754385965, + "loss": 3.6201, + "step": 9418 + }, + { + "epoch": 0.49573684210526314, + "grad_norm": 0.5, + "learning_rate": 0.00011205847953216375, + "loss": 3.3597, + "step": 9419 + }, + { + "epoch": 0.4957894736842105, + "grad_norm": 0.458984375, + "learning_rate": 0.000112046783625731, + "loss": 3.6081, + "step": 9420 + }, + { + "epoch": 0.4958421052631579, + "grad_norm": 0.4609375, + "learning_rate": 0.00011203508771929825, + "loss": 3.4357, + "step": 9421 + }, + { + "epoch": 0.4958947368421053, + "grad_norm": 0.48828125, + "learning_rate": 0.0001120233918128655, + "loss": 3.3002, + "step": 9422 + }, + { + "epoch": 0.49594736842105264, + "grad_norm": 0.5390625, + "learning_rate": 0.00011201169590643275, + "loss": 3.3262, + "step": 9423 + }, + { + "epoch": 0.496, + "grad_norm": 0.515625, + "learning_rate": 0.00011200000000000001, + "loss": 3.6141, + "step": 9424 + }, + { + "epoch": 0.49605263157894736, + "grad_norm": 0.4765625, + "learning_rate": 0.00011198830409356726, + "loss": 3.4759, + "step": 9425 + }, + { + "epoch": 0.4961052631578947, + "grad_norm": 0.6796875, + "learning_rate": 0.00011197660818713451, + "loss": 3.2101, + "step": 9426 + }, + { + "epoch": 0.49615789473684213, + "grad_norm": 0.490234375, + "learning_rate": 0.00011196491228070176, + "loss": 3.0007, + "step": 9427 + }, + { + "epoch": 0.4962105263157895, + "grad_norm": 0.44921875, + "learning_rate": 0.00011195321637426901, + "loss": 3.4142, + "step": 9428 + }, + { + "epoch": 0.49626315789473685, + "grad_norm": 0.470703125, + "learning_rate": 0.00011194152046783626, + "loss": 3.5899, + "step": 9429 + }, + { + "epoch": 0.4963157894736842, + "grad_norm": 0.482421875, + "learning_rate": 0.00011192982456140351, + "loss": 3.1903, + "step": 9430 + }, + { + "epoch": 0.49636842105263157, + "grad_norm": 0.47265625, + "learning_rate": 0.00011191812865497076, + "loss": 3.8775, + "step": 9431 + }, + { + "epoch": 0.49642105263157893, + "grad_norm": 0.51171875, + "learning_rate": 0.00011190643274853802, + "loss": 3.4247, + "step": 9432 + }, + { + "epoch": 0.4964736842105263, + "grad_norm": 0.52734375, + "learning_rate": 0.00011189473684210527, + "loss": 3.4698, + "step": 9433 + }, + { + "epoch": 0.4965263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00011188304093567252, + "loss": 3.7506, + "step": 9434 + }, + { + "epoch": 0.49657894736842106, + "grad_norm": 0.490234375, + "learning_rate": 0.00011187134502923977, + "loss": 3.4159, + "step": 9435 + }, + { + "epoch": 0.4966315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.00011185964912280702, + "loss": 3.8315, + "step": 9436 + }, + { + "epoch": 0.4966842105263158, + "grad_norm": 0.51953125, + "learning_rate": 0.00011184795321637427, + "loss": 3.9731, + "step": 9437 + }, + { + "epoch": 0.49673684210526314, + "grad_norm": 0.55859375, + "learning_rate": 0.00011183625730994152, + "loss": 3.3293, + "step": 9438 + }, + { + "epoch": 0.4967894736842105, + "grad_norm": 0.45703125, + "learning_rate": 0.00011182456140350878, + "loss": 3.5601, + "step": 9439 + }, + { + "epoch": 0.4968421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00011181286549707603, + "loss": 3.3139, + "step": 9440 + }, + { + "epoch": 0.4968947368421053, + "grad_norm": 0.498046875, + "learning_rate": 0.00011180116959064328, + "loss": 3.7978, + "step": 9441 + }, + { + "epoch": 0.49694736842105264, + "grad_norm": 0.54296875, + "learning_rate": 0.00011178947368421053, + "loss": 3.2264, + "step": 9442 + }, + { + "epoch": 0.497, + "grad_norm": 0.474609375, + "learning_rate": 0.00011177777777777778, + "loss": 3.5233, + "step": 9443 + }, + { + "epoch": 0.49705263157894736, + "grad_norm": 0.466796875, + "learning_rate": 0.00011176608187134503, + "loss": 3.4949, + "step": 9444 + }, + { + "epoch": 0.4971052631578947, + "grad_norm": 0.4921875, + "learning_rate": 0.00011175438596491228, + "loss": 3.5665, + "step": 9445 + }, + { + "epoch": 0.49715789473684213, + "grad_norm": 0.494140625, + "learning_rate": 0.00011174269005847953, + "loss": 3.2273, + "step": 9446 + }, + { + "epoch": 0.4972105263157895, + "grad_norm": 0.486328125, + "learning_rate": 0.00011173099415204679, + "loss": 3.8, + "step": 9447 + }, + { + "epoch": 0.49726315789473685, + "grad_norm": 0.57421875, + "learning_rate": 0.00011171929824561404, + "loss": 3.7815, + "step": 9448 + }, + { + "epoch": 0.4973157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.00011170760233918129, + "loss": 3.3009, + "step": 9449 + }, + { + "epoch": 0.49736842105263157, + "grad_norm": 0.515625, + "learning_rate": 0.00011169590643274854, + "loss": 3.6795, + "step": 9450 + }, + { + "epoch": 0.49742105263157893, + "grad_norm": 0.53515625, + "learning_rate": 0.00011168421052631579, + "loss": 3.3883, + "step": 9451 + }, + { + "epoch": 0.4974736842105263, + "grad_norm": 0.53515625, + "learning_rate": 0.00011167251461988304, + "loss": 3.9587, + "step": 9452 + }, + { + "epoch": 0.4975263157894737, + "grad_norm": 0.478515625, + "learning_rate": 0.00011166081871345029, + "loss": 3.7534, + "step": 9453 + }, + { + "epoch": 0.49757894736842107, + "grad_norm": 0.609375, + "learning_rate": 0.00011164912280701755, + "loss": 3.5353, + "step": 9454 + }, + { + "epoch": 0.4976315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.0001116374269005848, + "loss": 3.7511, + "step": 9455 + }, + { + "epoch": 0.4976842105263158, + "grad_norm": 0.48046875, + "learning_rate": 0.00011162573099415205, + "loss": 3.2683, + "step": 9456 + }, + { + "epoch": 0.49773684210526314, + "grad_norm": 0.494140625, + "learning_rate": 0.0001116140350877193, + "loss": 3.697, + "step": 9457 + }, + { + "epoch": 0.4977894736842105, + "grad_norm": 0.46484375, + "learning_rate": 0.00011160233918128655, + "loss": 3.4751, + "step": 9458 + }, + { + "epoch": 0.4978421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.0001115906432748538, + "loss": 3.5565, + "step": 9459 + }, + { + "epoch": 0.4978947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.00011157894736842105, + "loss": 3.2627, + "step": 9460 + }, + { + "epoch": 0.49794736842105264, + "grad_norm": 0.515625, + "learning_rate": 0.00011156725146198831, + "loss": 3.6235, + "step": 9461 + }, + { + "epoch": 0.498, + "grad_norm": 0.5234375, + "learning_rate": 0.00011155555555555556, + "loss": 3.4498, + "step": 9462 + }, + { + "epoch": 0.49805263157894736, + "grad_norm": 0.67578125, + "learning_rate": 0.00011154385964912281, + "loss": 3.5227, + "step": 9463 + }, + { + "epoch": 0.4981052631578947, + "grad_norm": 0.46484375, + "learning_rate": 0.00011153216374269006, + "loss": 3.5663, + "step": 9464 + }, + { + "epoch": 0.49815789473684213, + "grad_norm": 0.6015625, + "learning_rate": 0.00011152046783625731, + "loss": 3.3702, + "step": 9465 + }, + { + "epoch": 0.4982105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.00011150877192982456, + "loss": 3.394, + "step": 9466 + }, + { + "epoch": 0.49826315789473685, + "grad_norm": 0.46875, + "learning_rate": 0.00011149707602339181, + "loss": 3.5411, + "step": 9467 + }, + { + "epoch": 0.4983157894736842, + "grad_norm": 0.46875, + "learning_rate": 0.00011148538011695906, + "loss": 3.4377, + "step": 9468 + }, + { + "epoch": 0.49836842105263157, + "grad_norm": 0.5703125, + "learning_rate": 0.00011147368421052632, + "loss": 3.6344, + "step": 9469 + }, + { + "epoch": 0.49842105263157893, + "grad_norm": 0.482421875, + "learning_rate": 0.00011146198830409357, + "loss": 3.015, + "step": 9470 + }, + { + "epoch": 0.4984736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.00011145029239766082, + "loss": 3.9115, + "step": 9471 + }, + { + "epoch": 0.4985263157894737, + "grad_norm": 0.48046875, + "learning_rate": 0.00011143859649122807, + "loss": 3.684, + "step": 9472 + }, + { + "epoch": 0.49857894736842107, + "grad_norm": 0.51953125, + "learning_rate": 0.00011142690058479532, + "loss": 3.6577, + "step": 9473 + }, + { + "epoch": 0.4986315789473684, + "grad_norm": 0.49609375, + "learning_rate": 0.00011141520467836257, + "loss": 3.8283, + "step": 9474 + }, + { + "epoch": 0.4986842105263158, + "grad_norm": 0.578125, + "learning_rate": 0.00011140350877192982, + "loss": 3.2457, + "step": 9475 + }, + { + "epoch": 0.49873684210526315, + "grad_norm": 0.5546875, + "learning_rate": 0.00011139181286549709, + "loss": 3.0376, + "step": 9476 + }, + { + "epoch": 0.4987894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.00011138011695906434, + "loss": 3.6153, + "step": 9477 + }, + { + "epoch": 0.4988421052631579, + "grad_norm": 0.49609375, + "learning_rate": 0.00011136842105263158, + "loss": 3.5786, + "step": 9478 + }, + { + "epoch": 0.4988947368421053, + "grad_norm": 0.470703125, + "learning_rate": 0.00011135672514619883, + "loss": 3.5932, + "step": 9479 + }, + { + "epoch": 0.49894736842105264, + "grad_norm": 0.490234375, + "learning_rate": 0.00011134502923976608, + "loss": 3.2948, + "step": 9480 + }, + { + "epoch": 0.499, + "grad_norm": 0.474609375, + "learning_rate": 0.00011133333333333333, + "loss": 3.2081, + "step": 9481 + }, + { + "epoch": 0.49905263157894736, + "grad_norm": 0.486328125, + "learning_rate": 0.00011132163742690058, + "loss": 3.6877, + "step": 9482 + }, + { + "epoch": 0.4991052631578947, + "grad_norm": 0.498046875, + "learning_rate": 0.00011130994152046783, + "loss": 3.6276, + "step": 9483 + }, + { + "epoch": 0.4991578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.0001112982456140351, + "loss": 3.6728, + "step": 9484 + }, + { + "epoch": 0.4992105263157895, + "grad_norm": 0.45703125, + "learning_rate": 0.00011128654970760235, + "loss": 3.3607, + "step": 9485 + }, + { + "epoch": 0.49926315789473685, + "grad_norm": 0.478515625, + "learning_rate": 0.0001112748538011696, + "loss": 3.4814, + "step": 9486 + }, + { + "epoch": 0.4993157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00011126315789473684, + "loss": 3.0839, + "step": 9487 + }, + { + "epoch": 0.4993684210526316, + "grad_norm": 0.4609375, + "learning_rate": 0.0001112514619883041, + "loss": 3.4327, + "step": 9488 + }, + { + "epoch": 0.49942105263157893, + "grad_norm": 0.4765625, + "learning_rate": 0.00011123976608187134, + "loss": 3.4939, + "step": 9489 + }, + { + "epoch": 0.4994736842105263, + "grad_norm": 0.484375, + "learning_rate": 0.0001112280701754386, + "loss": 3.3082, + "step": 9490 + }, + { + "epoch": 0.4995263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.00011121637426900586, + "loss": 3.1481, + "step": 9491 + }, + { + "epoch": 0.49957894736842107, + "grad_norm": 0.46875, + "learning_rate": 0.0001112046783625731, + "loss": 3.4645, + "step": 9492 + }, + { + "epoch": 0.4996315789473684, + "grad_norm": 0.46875, + "learning_rate": 0.00011119298245614036, + "loss": 3.383, + "step": 9493 + }, + { + "epoch": 0.4996842105263158, + "grad_norm": 0.470703125, + "learning_rate": 0.0001111812865497076, + "loss": 3.686, + "step": 9494 + }, + { + "epoch": 0.49973684210526315, + "grad_norm": 0.5, + "learning_rate": 0.00011116959064327485, + "loss": 2.7936, + "step": 9495 + }, + { + "epoch": 0.4997894736842105, + "grad_norm": 0.484375, + "learning_rate": 0.0001111578947368421, + "loss": 3.4778, + "step": 9496 + }, + { + "epoch": 0.4998421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00011114619883040935, + "loss": 3.6952, + "step": 9497 + }, + { + "epoch": 0.4998947368421053, + "grad_norm": 0.474609375, + "learning_rate": 0.00011113450292397663, + "loss": 3.3624, + "step": 9498 + }, + { + "epoch": 0.49994736842105264, + "grad_norm": 0.474609375, + "learning_rate": 0.00011112280701754387, + "loss": 3.5561, + "step": 9499 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.00011111111111111112, + "loss": 3.1231, + "step": 9500 + }, + { + "epoch": 0.5000526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.00011109941520467837, + "loss": 3.6552, + "step": 9501 + }, + { + "epoch": 0.5001052631578947, + "grad_norm": 0.50390625, + "learning_rate": 0.00011108771929824562, + "loss": 3.7421, + "step": 9502 + }, + { + "epoch": 0.5001578947368421, + "grad_norm": 0.546875, + "learning_rate": 0.00011107602339181287, + "loss": 3.2504, + "step": 9503 + }, + { + "epoch": 0.5002105263157894, + "grad_norm": 0.5078125, + "learning_rate": 0.00011106432748538011, + "loss": 3.9988, + "step": 9504 + }, + { + "epoch": 0.5002631578947369, + "grad_norm": 0.5, + "learning_rate": 0.00011105263157894736, + "loss": 3.746, + "step": 9505 + }, + { + "epoch": 0.5003157894736842, + "grad_norm": 0.46484375, + "learning_rate": 0.00011104093567251464, + "loss": 3.4142, + "step": 9506 + }, + { + "epoch": 0.5003684210526316, + "grad_norm": 0.466796875, + "learning_rate": 0.00011102923976608188, + "loss": 3.3671, + "step": 9507 + }, + { + "epoch": 0.500421052631579, + "grad_norm": 0.45703125, + "learning_rate": 0.00011101754385964913, + "loss": 3.009, + "step": 9508 + }, + { + "epoch": 0.5004736842105263, + "grad_norm": 0.46875, + "learning_rate": 0.00011100584795321638, + "loss": 3.6728, + "step": 9509 + }, + { + "epoch": 0.5005263157894737, + "grad_norm": 0.51953125, + "learning_rate": 0.00011099415204678363, + "loss": 3.2175, + "step": 9510 + }, + { + "epoch": 0.500578947368421, + "grad_norm": 0.49609375, + "learning_rate": 0.00011098245614035088, + "loss": 3.5356, + "step": 9511 + }, + { + "epoch": 0.5006315789473684, + "grad_norm": 0.51171875, + "learning_rate": 0.00011097076023391813, + "loss": 3.4246, + "step": 9512 + }, + { + "epoch": 0.5006842105263158, + "grad_norm": 0.498046875, + "learning_rate": 0.0001109590643274854, + "loss": 3.3162, + "step": 9513 + }, + { + "epoch": 0.5007368421052631, + "grad_norm": 0.486328125, + "learning_rate": 0.00011094736842105264, + "loss": 3.2218, + "step": 9514 + }, + { + "epoch": 0.5007894736842106, + "grad_norm": 0.50390625, + "learning_rate": 0.00011093567251461989, + "loss": 3.1389, + "step": 9515 + }, + { + "epoch": 0.5008421052631579, + "grad_norm": 0.474609375, + "learning_rate": 0.00011092397660818714, + "loss": 3.4129, + "step": 9516 + }, + { + "epoch": 0.5008947368421053, + "grad_norm": 0.48046875, + "learning_rate": 0.00011091228070175439, + "loss": 3.0759, + "step": 9517 + }, + { + "epoch": 0.5009473684210526, + "grad_norm": 0.455078125, + "learning_rate": 0.00011090058479532164, + "loss": 3.1655, + "step": 9518 + }, + { + "epoch": 0.501, + "grad_norm": 0.46484375, + "learning_rate": 0.00011088888888888889, + "loss": 3.1793, + "step": 9519 + }, + { + "epoch": 0.5010526315789474, + "grad_norm": 0.48046875, + "learning_rate": 0.00011087719298245614, + "loss": 3.3866, + "step": 9520 + }, + { + "epoch": 0.5011052631578947, + "grad_norm": 0.484375, + "learning_rate": 0.00011086549707602341, + "loss": 3.8998, + "step": 9521 + }, + { + "epoch": 0.5011578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00011085380116959065, + "loss": 3.4052, + "step": 9522 + }, + { + "epoch": 0.5012105263157894, + "grad_norm": 0.51171875, + "learning_rate": 0.0001108421052631579, + "loss": 3.5483, + "step": 9523 + }, + { + "epoch": 0.5012631578947369, + "grad_norm": 0.4453125, + "learning_rate": 0.00011083040935672515, + "loss": 3.1453, + "step": 9524 + }, + { + "epoch": 0.5013157894736842, + "grad_norm": 0.478515625, + "learning_rate": 0.0001108187134502924, + "loss": 3.6584, + "step": 9525 + }, + { + "epoch": 0.5013684210526316, + "grad_norm": 0.474609375, + "learning_rate": 0.00011080701754385965, + "loss": 3.207, + "step": 9526 + }, + { + "epoch": 0.501421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.0001107953216374269, + "loss": 2.8761, + "step": 9527 + }, + { + "epoch": 0.5014736842105263, + "grad_norm": 0.4765625, + "learning_rate": 0.00011078362573099417, + "loss": 3.3641, + "step": 9528 + }, + { + "epoch": 0.5015263157894737, + "grad_norm": 0.4453125, + "learning_rate": 0.00011077192982456142, + "loss": 3.4956, + "step": 9529 + }, + { + "epoch": 0.501578947368421, + "grad_norm": 0.45703125, + "learning_rate": 0.00011076023391812866, + "loss": 3.8422, + "step": 9530 + }, + { + "epoch": 0.5016315789473684, + "grad_norm": 0.478515625, + "learning_rate": 0.00011074853801169591, + "loss": 3.4489, + "step": 9531 + }, + { + "epoch": 0.5016842105263158, + "grad_norm": 0.478515625, + "learning_rate": 0.00011073684210526316, + "loss": 3.356, + "step": 9532 + }, + { + "epoch": 0.5017368421052631, + "grad_norm": 0.44140625, + "learning_rate": 0.00011072514619883041, + "loss": 3.5681, + "step": 9533 + }, + { + "epoch": 0.5017894736842106, + "grad_norm": 0.462890625, + "learning_rate": 0.00011071345029239766, + "loss": 3.6783, + "step": 9534 + }, + { + "epoch": 0.5018421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00011070175438596493, + "loss": 3.4976, + "step": 9535 + }, + { + "epoch": 0.5018947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.00011069005847953218, + "loss": 3.2861, + "step": 9536 + }, + { + "epoch": 0.5019473684210526, + "grad_norm": 0.5625, + "learning_rate": 0.00011067836257309942, + "loss": 3.4955, + "step": 9537 + }, + { + "epoch": 0.502, + "grad_norm": 1.125, + "learning_rate": 0.00011066666666666667, + "loss": 3.5272, + "step": 9538 + }, + { + "epoch": 0.5020526315789474, + "grad_norm": 0.5390625, + "learning_rate": 0.00011065497076023392, + "loss": 3.7126, + "step": 9539 + }, + { + "epoch": 0.5021052631578947, + "grad_norm": 0.46484375, + "learning_rate": 0.00011064327485380117, + "loss": 3.5243, + "step": 9540 + }, + { + "epoch": 0.5021578947368421, + "grad_norm": 0.46484375, + "learning_rate": 0.00011063157894736842, + "loss": 3.2633, + "step": 9541 + }, + { + "epoch": 0.5022105263157894, + "grad_norm": 0.451171875, + "learning_rate": 0.00011061988304093567, + "loss": 3.8637, + "step": 9542 + }, + { + "epoch": 0.5022631578947369, + "grad_norm": 0.51171875, + "learning_rate": 0.00011060818713450294, + "loss": 3.5762, + "step": 9543 + }, + { + "epoch": 0.5023157894736842, + "grad_norm": 0.51953125, + "learning_rate": 0.0001105964912280702, + "loss": 3.4228, + "step": 9544 + }, + { + "epoch": 0.5023684210526316, + "grad_norm": 0.51171875, + "learning_rate": 0.00011058479532163743, + "loss": 3.2954, + "step": 9545 + }, + { + "epoch": 0.502421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00011057309941520468, + "loss": 3.2973, + "step": 9546 + }, + { + "epoch": 0.5024736842105263, + "grad_norm": 0.46875, + "learning_rate": 0.00011056140350877193, + "loss": 3.8024, + "step": 9547 + }, + { + "epoch": 0.5025263157894737, + "grad_norm": 0.486328125, + "learning_rate": 0.00011054970760233918, + "loss": 3.3545, + "step": 9548 + }, + { + "epoch": 0.502578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.00011053801169590643, + "loss": 3.8943, + "step": 9549 + }, + { + "epoch": 0.5026315789473684, + "grad_norm": 0.515625, + "learning_rate": 0.0001105263157894737, + "loss": 3.84, + "step": 9550 + }, + { + "epoch": 0.5026842105263158, + "grad_norm": 0.546875, + "learning_rate": 0.00011051461988304095, + "loss": 3.6238, + "step": 9551 + }, + { + "epoch": 0.5027368421052631, + "grad_norm": 0.515625, + "learning_rate": 0.0001105029239766082, + "loss": 3.3691, + "step": 9552 + }, + { + "epoch": 0.5027894736842106, + "grad_norm": 0.51171875, + "learning_rate": 0.00011049122807017544, + "loss": 3.0893, + "step": 9553 + }, + { + "epoch": 0.5028421052631579, + "grad_norm": 0.48046875, + "learning_rate": 0.00011047953216374269, + "loss": 3.523, + "step": 9554 + }, + { + "epoch": 0.5028947368421053, + "grad_norm": 0.470703125, + "learning_rate": 0.00011046783625730994, + "loss": 3.3709, + "step": 9555 + }, + { + "epoch": 0.5029473684210526, + "grad_norm": 0.48046875, + "learning_rate": 0.00011045614035087719, + "loss": 3.6398, + "step": 9556 + }, + { + "epoch": 0.503, + "grad_norm": 0.46875, + "learning_rate": 0.00011044444444444444, + "loss": 3.9641, + "step": 9557 + }, + { + "epoch": 0.5030526315789474, + "grad_norm": 0.53515625, + "learning_rate": 0.00011043274853801171, + "loss": 3.4186, + "step": 9558 + }, + { + "epoch": 0.5031052631578947, + "grad_norm": 0.48828125, + "learning_rate": 0.00011042105263157896, + "loss": 3.6068, + "step": 9559 + }, + { + "epoch": 0.5031578947368421, + "grad_norm": 0.5234375, + "learning_rate": 0.0001104093567251462, + "loss": 2.9614, + "step": 9560 + }, + { + "epoch": 0.5032105263157894, + "grad_norm": 0.47265625, + "learning_rate": 0.00011039766081871345, + "loss": 3.2295, + "step": 9561 + }, + { + "epoch": 0.5032631578947369, + "grad_norm": 0.546875, + "learning_rate": 0.0001103859649122807, + "loss": 3.0199, + "step": 9562 + }, + { + "epoch": 0.5033157894736842, + "grad_norm": 0.478515625, + "learning_rate": 0.00011037426900584795, + "loss": 3.6993, + "step": 9563 + }, + { + "epoch": 0.5033684210526316, + "grad_norm": 0.4609375, + "learning_rate": 0.0001103625730994152, + "loss": 3.7815, + "step": 9564 + }, + { + "epoch": 0.503421052631579, + "grad_norm": 0.490234375, + "learning_rate": 0.00011035087719298248, + "loss": 3.381, + "step": 9565 + }, + { + "epoch": 0.5034736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00011033918128654973, + "loss": 3.2797, + "step": 9566 + }, + { + "epoch": 0.5035263157894737, + "grad_norm": 0.4765625, + "learning_rate": 0.00011032748538011697, + "loss": 3.6733, + "step": 9567 + }, + { + "epoch": 0.503578947368421, + "grad_norm": 0.46875, + "learning_rate": 0.00011031578947368421, + "loss": 3.2868, + "step": 9568 + }, + { + "epoch": 0.5036315789473684, + "grad_norm": 0.53515625, + "learning_rate": 0.00011030409356725146, + "loss": 3.7313, + "step": 9569 + }, + { + "epoch": 0.5036842105263157, + "grad_norm": 0.51171875, + "learning_rate": 0.00011029239766081871, + "loss": 3.6611, + "step": 9570 + }, + { + "epoch": 0.5037368421052631, + "grad_norm": 0.546875, + "learning_rate": 0.00011028070175438596, + "loss": 3.0675, + "step": 9571 + }, + { + "epoch": 0.5037894736842106, + "grad_norm": 0.47265625, + "learning_rate": 0.00011026900584795321, + "loss": 3.6289, + "step": 9572 + }, + { + "epoch": 0.5038421052631579, + "grad_norm": 0.4609375, + "learning_rate": 0.00011025730994152049, + "loss": 4.0793, + "step": 9573 + }, + { + "epoch": 0.5038947368421053, + "grad_norm": 0.474609375, + "learning_rate": 0.00011024561403508774, + "loss": 3.1074, + "step": 9574 + }, + { + "epoch": 0.5039473684210526, + "grad_norm": 0.470703125, + "learning_rate": 0.00011023391812865498, + "loss": 3.5497, + "step": 9575 + }, + { + "epoch": 0.504, + "grad_norm": 0.49609375, + "learning_rate": 0.00011022222222222222, + "loss": 3.665, + "step": 9576 + }, + { + "epoch": 0.5040526315789474, + "grad_norm": 0.5234375, + "learning_rate": 0.00011021052631578947, + "loss": 3.8645, + "step": 9577 + }, + { + "epoch": 0.5041052631578947, + "grad_norm": 0.478515625, + "learning_rate": 0.00011019883040935672, + "loss": 3.8522, + "step": 9578 + }, + { + "epoch": 0.5041578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00011018713450292397, + "loss": 3.8162, + "step": 9579 + }, + { + "epoch": 0.5042105263157894, + "grad_norm": 0.47265625, + "learning_rate": 0.00011017543859649125, + "loss": 3.1248, + "step": 9580 + }, + { + "epoch": 0.5042631578947369, + "grad_norm": 0.478515625, + "learning_rate": 0.0001101637426900585, + "loss": 3.5271, + "step": 9581 + }, + { + "epoch": 0.5043157894736842, + "grad_norm": 0.462890625, + "learning_rate": 0.00011015204678362575, + "loss": 3.2026, + "step": 9582 + }, + { + "epoch": 0.5043684210526316, + "grad_norm": 0.470703125, + "learning_rate": 0.00011014035087719298, + "loss": 3.8883, + "step": 9583 + }, + { + "epoch": 0.504421052631579, + "grad_norm": 0.50390625, + "learning_rate": 0.00011012865497076023, + "loss": 3.5339, + "step": 9584 + }, + { + "epoch": 0.5044736842105263, + "grad_norm": 0.4765625, + "learning_rate": 0.00011011695906432748, + "loss": 3.2641, + "step": 9585 + }, + { + "epoch": 0.5045263157894737, + "grad_norm": 0.439453125, + "learning_rate": 0.00011010526315789473, + "loss": 3.18, + "step": 9586 + }, + { + "epoch": 0.504578947368421, + "grad_norm": 0.470703125, + "learning_rate": 0.00011009356725146201, + "loss": 3.2156, + "step": 9587 + }, + { + "epoch": 0.5046315789473684, + "grad_norm": 0.478515625, + "learning_rate": 0.00011008187134502926, + "loss": 3.3123, + "step": 9588 + }, + { + "epoch": 0.5046842105263157, + "grad_norm": 0.486328125, + "learning_rate": 0.0001100701754385965, + "loss": 3.9002, + "step": 9589 + }, + { + "epoch": 0.5047368421052632, + "grad_norm": 0.462890625, + "learning_rate": 0.00011005847953216376, + "loss": 3.2475, + "step": 9590 + }, + { + "epoch": 0.5047894736842106, + "grad_norm": 0.47265625, + "learning_rate": 0.00011004678362573099, + "loss": 3.1916, + "step": 9591 + }, + { + "epoch": 0.5048421052631579, + "grad_norm": 0.47265625, + "learning_rate": 0.00011003508771929824, + "loss": 3.6708, + "step": 9592 + }, + { + "epoch": 0.5048947368421053, + "grad_norm": 0.462890625, + "learning_rate": 0.00011002339181286549, + "loss": 3.5268, + "step": 9593 + }, + { + "epoch": 0.5049473684210526, + "grad_norm": 0.474609375, + "learning_rate": 0.00011001169590643274, + "loss": 3.8044, + "step": 9594 + }, + { + "epoch": 0.505, + "grad_norm": 0.515625, + "learning_rate": 0.00011000000000000002, + "loss": 3.8075, + "step": 9595 + }, + { + "epoch": 0.5050526315789474, + "grad_norm": 0.53125, + "learning_rate": 0.00010998830409356727, + "loss": 3.4097, + "step": 9596 + }, + { + "epoch": 0.5051052631578947, + "grad_norm": 0.470703125, + "learning_rate": 0.00010997660818713452, + "loss": 3.6385, + "step": 9597 + }, + { + "epoch": 0.5051578947368421, + "grad_norm": 0.48828125, + "learning_rate": 0.00010996491228070177, + "loss": 3.5968, + "step": 9598 + }, + { + "epoch": 0.5052105263157894, + "grad_norm": 0.49609375, + "learning_rate": 0.000109953216374269, + "loss": 3.48, + "step": 9599 + }, + { + "epoch": 0.5052631578947369, + "grad_norm": 0.51171875, + "learning_rate": 0.00010994152046783625, + "loss": 3.8261, + "step": 9600 + }, + { + "epoch": 0.5053157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.0001099298245614035, + "loss": 3.482, + "step": 9601 + }, + { + "epoch": 0.5053684210526316, + "grad_norm": 0.48828125, + "learning_rate": 0.00010991812865497078, + "loss": 3.8155, + "step": 9602 + }, + { + "epoch": 0.505421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00010990643274853803, + "loss": 3.5761, + "step": 9603 + }, + { + "epoch": 0.5054736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.00010989473684210528, + "loss": 3.4391, + "step": 9604 + }, + { + "epoch": 0.5055263157894737, + "grad_norm": 0.5078125, + "learning_rate": 0.00010988304093567253, + "loss": 3.2942, + "step": 9605 + }, + { + "epoch": 0.505578947368421, + "grad_norm": 0.5234375, + "learning_rate": 0.00010987134502923976, + "loss": 3.3991, + "step": 9606 + }, + { + "epoch": 0.5056315789473684, + "grad_norm": 0.48046875, + "learning_rate": 0.00010985964912280701, + "loss": 3.2904, + "step": 9607 + }, + { + "epoch": 0.5056842105263157, + "grad_norm": 0.46484375, + "learning_rate": 0.00010984795321637426, + "loss": 3.1187, + "step": 9608 + }, + { + "epoch": 0.5057368421052632, + "grad_norm": 0.50390625, + "learning_rate": 0.00010983625730994151, + "loss": 3.0711, + "step": 9609 + }, + { + "epoch": 0.5057894736842106, + "grad_norm": 0.5078125, + "learning_rate": 0.00010982456140350879, + "loss": 3.9199, + "step": 9610 + }, + { + "epoch": 0.5058421052631579, + "grad_norm": 0.46875, + "learning_rate": 0.00010981286549707604, + "loss": 3.4212, + "step": 9611 + }, + { + "epoch": 0.5058947368421053, + "grad_norm": 0.498046875, + "learning_rate": 0.00010980116959064329, + "loss": 3.5833, + "step": 9612 + }, + { + "epoch": 0.5059473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00010978947368421054, + "loss": 3.5385, + "step": 9613 + }, + { + "epoch": 0.506, + "grad_norm": 0.494140625, + "learning_rate": 0.00010977777777777777, + "loss": 3.7848, + "step": 9614 + }, + { + "epoch": 0.5060526315789474, + "grad_norm": 0.50390625, + "learning_rate": 0.00010976608187134502, + "loss": 3.5717, + "step": 9615 + }, + { + "epoch": 0.5061052631578947, + "grad_norm": 0.5390625, + "learning_rate": 0.00010975438596491227, + "loss": 3.4283, + "step": 9616 + }, + { + "epoch": 0.5061578947368421, + "grad_norm": 0.5546875, + "learning_rate": 0.00010974269005847955, + "loss": 3.6075, + "step": 9617 + }, + { + "epoch": 0.5062105263157894, + "grad_norm": 0.50390625, + "learning_rate": 0.0001097309941520468, + "loss": 3.1198, + "step": 9618 + }, + { + "epoch": 0.5062631578947369, + "grad_norm": 0.49609375, + "learning_rate": 0.00010971929824561405, + "loss": 3.5182, + "step": 9619 + }, + { + "epoch": 0.5063157894736842, + "grad_norm": 0.470703125, + "learning_rate": 0.0001097076023391813, + "loss": 3.3026, + "step": 9620 + }, + { + "epoch": 0.5063684210526316, + "grad_norm": 0.494140625, + "learning_rate": 0.00010969590643274855, + "loss": 3.3344, + "step": 9621 + }, + { + "epoch": 0.506421052631579, + "grad_norm": 0.470703125, + "learning_rate": 0.00010968421052631578, + "loss": 3.4822, + "step": 9622 + }, + { + "epoch": 0.5064736842105263, + "grad_norm": 0.47265625, + "learning_rate": 0.00010967251461988303, + "loss": 3.1704, + "step": 9623 + }, + { + "epoch": 0.5065263157894737, + "grad_norm": 0.46484375, + "learning_rate": 0.00010966081871345031, + "loss": 2.9895, + "step": 9624 + }, + { + "epoch": 0.506578947368421, + "grad_norm": 0.451171875, + "learning_rate": 0.00010964912280701756, + "loss": 3.7531, + "step": 9625 + }, + { + "epoch": 0.5066315789473684, + "grad_norm": 0.4609375, + "learning_rate": 0.00010963742690058481, + "loss": 2.7846, + "step": 9626 + }, + { + "epoch": 0.5066842105263157, + "grad_norm": 0.486328125, + "learning_rate": 0.00010962573099415206, + "loss": 3.3948, + "step": 9627 + }, + { + "epoch": 0.5067368421052632, + "grad_norm": 0.4609375, + "learning_rate": 0.00010961403508771931, + "loss": 3.5558, + "step": 9628 + }, + { + "epoch": 0.5067894736842106, + "grad_norm": 0.5390625, + "learning_rate": 0.00010960233918128654, + "loss": 3.4178, + "step": 9629 + }, + { + "epoch": 0.5068421052631579, + "grad_norm": 0.462890625, + "learning_rate": 0.0001095906432748538, + "loss": 3.2572, + "step": 9630 + }, + { + "epoch": 0.5068947368421053, + "grad_norm": 0.46484375, + "learning_rate": 0.00010957894736842104, + "loss": 3.8444, + "step": 9631 + }, + { + "epoch": 0.5069473684210526, + "grad_norm": 0.48046875, + "learning_rate": 0.00010956725146198832, + "loss": 3.4998, + "step": 9632 + }, + { + "epoch": 0.507, + "grad_norm": 0.47265625, + "learning_rate": 0.00010955555555555557, + "loss": 3.6872, + "step": 9633 + }, + { + "epoch": 0.5070526315789474, + "grad_norm": 0.48828125, + "learning_rate": 0.00010954385964912282, + "loss": 3.7175, + "step": 9634 + }, + { + "epoch": 0.5071052631578947, + "grad_norm": 0.49609375, + "learning_rate": 0.00010953216374269007, + "loss": 3.5928, + "step": 9635 + }, + { + "epoch": 0.5071578947368421, + "grad_norm": 0.5859375, + "learning_rate": 0.00010952046783625732, + "loss": 3.7861, + "step": 9636 + }, + { + "epoch": 0.5072105263157894, + "grad_norm": 0.4765625, + "learning_rate": 0.00010950877192982455, + "loss": 3.4583, + "step": 9637 + }, + { + "epoch": 0.5072631578947369, + "grad_norm": 1.84375, + "learning_rate": 0.0001094970760233918, + "loss": 3.3631, + "step": 9638 + }, + { + "epoch": 0.5073157894736842, + "grad_norm": 0.4765625, + "learning_rate": 0.00010948538011695908, + "loss": 3.6753, + "step": 9639 + }, + { + "epoch": 0.5073684210526316, + "grad_norm": 0.4921875, + "learning_rate": 0.00010947368421052633, + "loss": 3.3081, + "step": 9640 + }, + { + "epoch": 0.507421052631579, + "grad_norm": 0.466796875, + "learning_rate": 0.00010946198830409358, + "loss": 3.9308, + "step": 9641 + }, + { + "epoch": 0.5074736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00010945029239766083, + "loss": 3.7131, + "step": 9642 + }, + { + "epoch": 0.5075263157894737, + "grad_norm": 0.484375, + "learning_rate": 0.00010943859649122808, + "loss": 3.5753, + "step": 9643 + }, + { + "epoch": 0.507578947368421, + "grad_norm": 0.49609375, + "learning_rate": 0.00010942690058479533, + "loss": 3.3312, + "step": 9644 + }, + { + "epoch": 0.5076315789473684, + "grad_norm": 0.51953125, + "learning_rate": 0.00010941520467836256, + "loss": 3.2879, + "step": 9645 + }, + { + "epoch": 0.5076842105263157, + "grad_norm": 0.671875, + "learning_rate": 0.00010940350877192981, + "loss": 3.4933, + "step": 9646 + }, + { + "epoch": 0.5077368421052632, + "grad_norm": 0.47265625, + "learning_rate": 0.00010939181286549709, + "loss": 3.959, + "step": 9647 + }, + { + "epoch": 0.5077894736842106, + "grad_norm": 0.52734375, + "learning_rate": 0.00010938011695906434, + "loss": 3.0171, + "step": 9648 + }, + { + "epoch": 0.5078421052631579, + "grad_norm": 0.69921875, + "learning_rate": 0.00010936842105263159, + "loss": 3.3723, + "step": 9649 + }, + { + "epoch": 0.5078947368421053, + "grad_norm": 0.498046875, + "learning_rate": 0.00010935672514619884, + "loss": 3.5721, + "step": 9650 + }, + { + "epoch": 0.5079473684210526, + "grad_norm": 0.498046875, + "learning_rate": 0.00010934502923976609, + "loss": 3.3225, + "step": 9651 + }, + { + "epoch": 0.508, + "grad_norm": 0.490234375, + "learning_rate": 0.00010933333333333333, + "loss": 3.2325, + "step": 9652 + }, + { + "epoch": 0.5080526315789474, + "grad_norm": 0.54296875, + "learning_rate": 0.00010932163742690058, + "loss": 3.3141, + "step": 9653 + }, + { + "epoch": 0.5081052631578947, + "grad_norm": 0.474609375, + "learning_rate": 0.00010930994152046785, + "loss": 3.3445, + "step": 9654 + }, + { + "epoch": 0.5081578947368421, + "grad_norm": 0.48828125, + "learning_rate": 0.0001092982456140351, + "loss": 3.4877, + "step": 9655 + }, + { + "epoch": 0.5082105263157894, + "grad_norm": 0.478515625, + "learning_rate": 0.00010928654970760235, + "loss": 3.5905, + "step": 9656 + }, + { + "epoch": 0.5082631578947369, + "grad_norm": 0.466796875, + "learning_rate": 0.0001092748538011696, + "loss": 3.4083, + "step": 9657 + }, + { + "epoch": 0.5083157894736842, + "grad_norm": 0.4765625, + "learning_rate": 0.00010926315789473685, + "loss": 3.6179, + "step": 9658 + }, + { + "epoch": 0.5083684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.0001092514619883041, + "loss": 3.5083, + "step": 9659 + }, + { + "epoch": 0.508421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.00010923976608187134, + "loss": 3.4092, + "step": 9660 + }, + { + "epoch": 0.5084736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00010922807017543861, + "loss": 2.5012, + "step": 9661 + }, + { + "epoch": 0.5085263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00010921637426900586, + "loss": 3.5861, + "step": 9662 + }, + { + "epoch": 0.508578947368421, + "grad_norm": 0.435546875, + "learning_rate": 0.00010920467836257311, + "loss": 3.5841, + "step": 9663 + }, + { + "epoch": 0.5086315789473684, + "grad_norm": 0.462890625, + "learning_rate": 0.00010919298245614036, + "loss": 3.5592, + "step": 9664 + }, + { + "epoch": 0.5086842105263157, + "grad_norm": 0.66015625, + "learning_rate": 0.00010918128654970761, + "loss": 3.4944, + "step": 9665 + }, + { + "epoch": 0.5087368421052632, + "grad_norm": 0.474609375, + "learning_rate": 0.00010916959064327486, + "loss": 3.9203, + "step": 9666 + }, + { + "epoch": 0.5087894736842106, + "grad_norm": 0.51953125, + "learning_rate": 0.00010915789473684211, + "loss": 3.5175, + "step": 9667 + }, + { + "epoch": 0.5088421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.00010914619883040935, + "loss": 3.5668, + "step": 9668 + }, + { + "epoch": 0.5088947368421053, + "grad_norm": 0.546875, + "learning_rate": 0.00010913450292397662, + "loss": 3.3301, + "step": 9669 + }, + { + "epoch": 0.5089473684210526, + "grad_norm": 0.60546875, + "learning_rate": 0.00010912280701754387, + "loss": 3.4037, + "step": 9670 + }, + { + "epoch": 0.509, + "grad_norm": 0.55859375, + "learning_rate": 0.00010911111111111112, + "loss": 3.5265, + "step": 9671 + }, + { + "epoch": 0.5090526315789474, + "grad_norm": 0.498046875, + "learning_rate": 0.00010909941520467837, + "loss": 3.5731, + "step": 9672 + }, + { + "epoch": 0.5091052631578947, + "grad_norm": 0.58203125, + "learning_rate": 0.00010908771929824562, + "loss": 3.2054, + "step": 9673 + }, + { + "epoch": 0.5091578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00010907602339181287, + "loss": 3.1825, + "step": 9674 + }, + { + "epoch": 0.5092105263157894, + "grad_norm": 0.50390625, + "learning_rate": 0.00010906432748538011, + "loss": 3.4363, + "step": 9675 + }, + { + "epoch": 0.5092631578947369, + "grad_norm": 0.48828125, + "learning_rate": 0.00010905263157894738, + "loss": 3.6866, + "step": 9676 + }, + { + "epoch": 0.5093157894736842, + "grad_norm": 0.60546875, + "learning_rate": 0.00010904093567251463, + "loss": 3.5143, + "step": 9677 + }, + { + "epoch": 0.5093684210526316, + "grad_norm": 0.48046875, + "learning_rate": 0.00010902923976608188, + "loss": 3.4313, + "step": 9678 + }, + { + "epoch": 0.509421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.00010901754385964913, + "loss": 3.3863, + "step": 9679 + }, + { + "epoch": 0.5094736842105263, + "grad_norm": 0.5, + "learning_rate": 0.00010900584795321638, + "loss": 3.684, + "step": 9680 + }, + { + "epoch": 0.5095263157894737, + "grad_norm": 0.53125, + "learning_rate": 0.00010899415204678363, + "loss": 3.0287, + "step": 9681 + }, + { + "epoch": 0.509578947368421, + "grad_norm": 0.4765625, + "learning_rate": 0.00010898245614035088, + "loss": 3.4181, + "step": 9682 + }, + { + "epoch": 0.5096315789473684, + "grad_norm": 0.470703125, + "learning_rate": 0.00010897076023391812, + "loss": 3.3403, + "step": 9683 + }, + { + "epoch": 0.5096842105263157, + "grad_norm": 0.470703125, + "learning_rate": 0.0001089590643274854, + "loss": 3.4539, + "step": 9684 + }, + { + "epoch": 0.5097368421052632, + "grad_norm": 0.4765625, + "learning_rate": 0.00010894736842105264, + "loss": 3.6152, + "step": 9685 + }, + { + "epoch": 0.5097894736842106, + "grad_norm": 0.48046875, + "learning_rate": 0.00010893567251461989, + "loss": 3.3303, + "step": 9686 + }, + { + "epoch": 0.5098421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00010892397660818714, + "loss": 3.6253, + "step": 9687 + }, + { + "epoch": 0.5098947368421053, + "grad_norm": 0.49609375, + "learning_rate": 0.00010891228070175439, + "loss": 2.8837, + "step": 9688 + }, + { + "epoch": 0.5099473684210526, + "grad_norm": 0.51171875, + "learning_rate": 0.00010890058479532164, + "loss": 3.652, + "step": 9689 + }, + { + "epoch": 0.51, + "grad_norm": 0.51953125, + "learning_rate": 0.00010888888888888889, + "loss": 3.6416, + "step": 9690 + }, + { + "epoch": 0.5100526315789473, + "grad_norm": 0.50390625, + "learning_rate": 0.00010887719298245615, + "loss": 3.6686, + "step": 9691 + }, + { + "epoch": 0.5101052631578947, + "grad_norm": 0.51171875, + "learning_rate": 0.0001088654970760234, + "loss": 3.5017, + "step": 9692 + }, + { + "epoch": 0.5101578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.00010885380116959065, + "loss": 3.4914, + "step": 9693 + }, + { + "epoch": 0.5102105263157894, + "grad_norm": 0.478515625, + "learning_rate": 0.0001088421052631579, + "loss": 3.489, + "step": 9694 + }, + { + "epoch": 0.5102631578947369, + "grad_norm": 0.46484375, + "learning_rate": 0.00010883040935672515, + "loss": 3.6313, + "step": 9695 + }, + { + "epoch": 0.5103157894736842, + "grad_norm": 0.451171875, + "learning_rate": 0.0001088187134502924, + "loss": 3.5872, + "step": 9696 + }, + { + "epoch": 0.5103684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00010880701754385965, + "loss": 3.4984, + "step": 9697 + }, + { + "epoch": 0.510421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00010879532163742692, + "loss": 3.7243, + "step": 9698 + }, + { + "epoch": 0.5104736842105263, + "grad_norm": 0.51171875, + "learning_rate": 0.00010878362573099416, + "loss": 3.3456, + "step": 9699 + }, + { + "epoch": 0.5105263157894737, + "grad_norm": 0.498046875, + "learning_rate": 0.00010877192982456141, + "loss": 3.7074, + "step": 9700 + }, + { + "epoch": 0.510578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.00010876023391812866, + "loss": 3.1764, + "step": 9701 + }, + { + "epoch": 0.5106315789473684, + "grad_norm": 0.5234375, + "learning_rate": 0.00010874853801169591, + "loss": 3.7748, + "step": 9702 + }, + { + "epoch": 0.5106842105263157, + "grad_norm": 0.494140625, + "learning_rate": 0.00010873684210526316, + "loss": 3.9634, + "step": 9703 + }, + { + "epoch": 0.5107368421052632, + "grad_norm": 0.49609375, + "learning_rate": 0.00010872514619883041, + "loss": 3.2787, + "step": 9704 + }, + { + "epoch": 0.5107894736842106, + "grad_norm": 0.53125, + "learning_rate": 0.00010871345029239766, + "loss": 3.7524, + "step": 9705 + }, + { + "epoch": 0.5108421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.00010870175438596493, + "loss": 3.1657, + "step": 9706 + }, + { + "epoch": 0.5108947368421053, + "grad_norm": 0.455078125, + "learning_rate": 0.00010869005847953218, + "loss": 3.1579, + "step": 9707 + }, + { + "epoch": 0.5109473684210526, + "grad_norm": 0.45703125, + "learning_rate": 0.00010867836257309942, + "loss": 3.6856, + "step": 9708 + }, + { + "epoch": 0.511, + "grad_norm": 0.455078125, + "learning_rate": 0.00010866666666666667, + "loss": 3.9158, + "step": 9709 + }, + { + "epoch": 0.5110526315789473, + "grad_norm": 0.52734375, + "learning_rate": 0.00010865497076023392, + "loss": 3.6154, + "step": 9710 + }, + { + "epoch": 0.5111052631578947, + "grad_norm": 0.494140625, + "learning_rate": 0.00010864327485380117, + "loss": 3.2241, + "step": 9711 + }, + { + "epoch": 0.5111578947368421, + "grad_norm": 0.470703125, + "learning_rate": 0.00010863157894736842, + "loss": 3.5923, + "step": 9712 + }, + { + "epoch": 0.5112105263157894, + "grad_norm": 0.5234375, + "learning_rate": 0.00010861988304093569, + "loss": 3.9204, + "step": 9713 + }, + { + "epoch": 0.5112631578947369, + "grad_norm": 0.515625, + "learning_rate": 0.00010860818713450294, + "loss": 3.2478, + "step": 9714 + }, + { + "epoch": 0.5113157894736842, + "grad_norm": 0.490234375, + "learning_rate": 0.00010859649122807019, + "loss": 3.6108, + "step": 9715 + }, + { + "epoch": 0.5113684210526316, + "grad_norm": 0.53515625, + "learning_rate": 0.00010858479532163744, + "loss": 3.0672, + "step": 9716 + }, + { + "epoch": 0.511421052631579, + "grad_norm": 0.4609375, + "learning_rate": 0.00010857309941520468, + "loss": 3.6023, + "step": 9717 + }, + { + "epoch": 0.5114736842105263, + "grad_norm": 0.478515625, + "learning_rate": 0.00010856140350877193, + "loss": 3.3517, + "step": 9718 + }, + { + "epoch": 0.5115263157894737, + "grad_norm": 0.47265625, + "learning_rate": 0.00010854970760233918, + "loss": 3.1452, + "step": 9719 + }, + { + "epoch": 0.511578947368421, + "grad_norm": 0.470703125, + "learning_rate": 0.00010853801169590643, + "loss": 3.4354, + "step": 9720 + }, + { + "epoch": 0.5116315789473684, + "grad_norm": 0.53125, + "learning_rate": 0.0001085263157894737, + "loss": 4.0017, + "step": 9721 + }, + { + "epoch": 0.5116842105263157, + "grad_norm": 0.48828125, + "learning_rate": 0.00010851461988304095, + "loss": 3.7048, + "step": 9722 + }, + { + "epoch": 0.5117368421052632, + "grad_norm": 0.6171875, + "learning_rate": 0.0001085029239766082, + "loss": 3.3722, + "step": 9723 + }, + { + "epoch": 0.5117894736842106, + "grad_norm": 0.51171875, + "learning_rate": 0.00010849122807017545, + "loss": 3.5062, + "step": 9724 + }, + { + "epoch": 0.5118421052631579, + "grad_norm": 0.5234375, + "learning_rate": 0.0001084795321637427, + "loss": 3.2531, + "step": 9725 + }, + { + "epoch": 0.5118947368421053, + "grad_norm": 0.51953125, + "learning_rate": 0.00010846783625730994, + "loss": 3.1357, + "step": 9726 + }, + { + "epoch": 0.5119473684210526, + "grad_norm": 0.5546875, + "learning_rate": 0.0001084561403508772, + "loss": 3.693, + "step": 9727 + }, + { + "epoch": 0.512, + "grad_norm": 0.53125, + "learning_rate": 0.00010844444444444446, + "loss": 3.6222, + "step": 9728 + }, + { + "epoch": 0.5120526315789473, + "grad_norm": 0.546875, + "learning_rate": 0.00010843274853801171, + "loss": 3.5317, + "step": 9729 + }, + { + "epoch": 0.5121052631578947, + "grad_norm": 0.6875, + "learning_rate": 0.00010842105263157896, + "loss": 3.303, + "step": 9730 + }, + { + "epoch": 0.5121578947368421, + "grad_norm": 0.490234375, + "learning_rate": 0.0001084093567251462, + "loss": 3.7639, + "step": 9731 + }, + { + "epoch": 0.5122105263157894, + "grad_norm": 0.52734375, + "learning_rate": 0.00010839766081871346, + "loss": 3.2264, + "step": 9732 + }, + { + "epoch": 0.5122631578947369, + "grad_norm": 0.54296875, + "learning_rate": 0.0001083859649122807, + "loss": 3.9005, + "step": 9733 + }, + { + "epoch": 0.5123157894736842, + "grad_norm": 0.5078125, + "learning_rate": 0.00010837426900584795, + "loss": 3.9354, + "step": 9734 + }, + { + "epoch": 0.5123684210526316, + "grad_norm": 0.609375, + "learning_rate": 0.0001083625730994152, + "loss": 3.4746, + "step": 9735 + }, + { + "epoch": 0.512421052631579, + "grad_norm": 0.470703125, + "learning_rate": 0.00010835087719298247, + "loss": 2.986, + "step": 9736 + }, + { + "epoch": 0.5124736842105263, + "grad_norm": 0.470703125, + "learning_rate": 0.00010833918128654972, + "loss": 3.268, + "step": 9737 + }, + { + "epoch": 0.5125263157894737, + "grad_norm": 0.4765625, + "learning_rate": 0.00010832748538011697, + "loss": 3.5471, + "step": 9738 + }, + { + "epoch": 0.512578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00010831578947368422, + "loss": 3.3164, + "step": 9739 + }, + { + "epoch": 0.5126315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.00010830409356725147, + "loss": 3.3637, + "step": 9740 + }, + { + "epoch": 0.5126842105263157, + "grad_norm": 0.53125, + "learning_rate": 0.00010829239766081872, + "loss": 3.2836, + "step": 9741 + }, + { + "epoch": 0.5127368421052632, + "grad_norm": 0.48046875, + "learning_rate": 0.00010828070175438597, + "loss": 3.9147, + "step": 9742 + }, + { + "epoch": 0.5127894736842106, + "grad_norm": 0.48046875, + "learning_rate": 0.00010826900584795323, + "loss": 3.5981, + "step": 9743 + }, + { + "epoch": 0.5128421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00010825730994152048, + "loss": 3.627, + "step": 9744 + }, + { + "epoch": 0.5128947368421053, + "grad_norm": 0.458984375, + "learning_rate": 0.00010824561403508773, + "loss": 3.5979, + "step": 9745 + }, + { + "epoch": 0.5129473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00010823391812865498, + "loss": 3.6386, + "step": 9746 + }, + { + "epoch": 0.513, + "grad_norm": 0.4765625, + "learning_rate": 0.00010822222222222223, + "loss": 3.2798, + "step": 9747 + }, + { + "epoch": 0.5130526315789473, + "grad_norm": 0.52734375, + "learning_rate": 0.00010821052631578948, + "loss": 3.7625, + "step": 9748 + }, + { + "epoch": 0.5131052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00010819883040935673, + "loss": 3.2528, + "step": 9749 + }, + { + "epoch": 0.5131578947368421, + "grad_norm": 0.474609375, + "learning_rate": 0.00010818713450292399, + "loss": 3.2015, + "step": 9750 + }, + { + "epoch": 0.5132105263157895, + "grad_norm": 0.474609375, + "learning_rate": 0.00010817543859649124, + "loss": 2.6582, + "step": 9751 + }, + { + "epoch": 0.5132631578947369, + "grad_norm": 0.64453125, + "learning_rate": 0.00010816374269005849, + "loss": 3.6907, + "step": 9752 + }, + { + "epoch": 0.5133157894736842, + "grad_norm": 0.447265625, + "learning_rate": 0.00010815204678362574, + "loss": 3.3337, + "step": 9753 + }, + { + "epoch": 0.5133684210526316, + "grad_norm": 0.48046875, + "learning_rate": 0.00010814035087719299, + "loss": 3.9464, + "step": 9754 + }, + { + "epoch": 0.513421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00010812865497076024, + "loss": 3.2283, + "step": 9755 + }, + { + "epoch": 0.5134736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.00010811695906432749, + "loss": 3.4375, + "step": 9756 + }, + { + "epoch": 0.5135263157894737, + "grad_norm": 0.484375, + "learning_rate": 0.00010810526315789474, + "loss": 3.9124, + "step": 9757 + }, + { + "epoch": 0.513578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.000108093567251462, + "loss": 3.461, + "step": 9758 + }, + { + "epoch": 0.5136315789473684, + "grad_norm": 0.474609375, + "learning_rate": 0.00010808187134502925, + "loss": 3.6464, + "step": 9759 + }, + { + "epoch": 0.5136842105263157, + "grad_norm": 0.48828125, + "learning_rate": 0.0001080701754385965, + "loss": 3.235, + "step": 9760 + }, + { + "epoch": 0.5137368421052632, + "grad_norm": 0.494140625, + "learning_rate": 0.00010805847953216375, + "loss": 3.4492, + "step": 9761 + }, + { + "epoch": 0.5137894736842106, + "grad_norm": 0.482421875, + "learning_rate": 0.000108046783625731, + "loss": 3.0745, + "step": 9762 + }, + { + "epoch": 0.5138421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00010803508771929825, + "loss": 3.3715, + "step": 9763 + }, + { + "epoch": 0.5138947368421053, + "grad_norm": 0.5078125, + "learning_rate": 0.0001080233918128655, + "loss": 3.4474, + "step": 9764 + }, + { + "epoch": 0.5139473684210526, + "grad_norm": 0.49609375, + "learning_rate": 0.00010801169590643276, + "loss": 3.4602, + "step": 9765 + }, + { + "epoch": 0.514, + "grad_norm": 0.51953125, + "learning_rate": 0.00010800000000000001, + "loss": 3.4284, + "step": 9766 + }, + { + "epoch": 0.5140526315789473, + "grad_norm": 0.49609375, + "learning_rate": 0.00010798830409356726, + "loss": 2.8582, + "step": 9767 + }, + { + "epoch": 0.5141052631578947, + "grad_norm": 0.5, + "learning_rate": 0.00010797660818713451, + "loss": 3.4255, + "step": 9768 + }, + { + "epoch": 0.5141578947368421, + "grad_norm": 0.62890625, + "learning_rate": 0.00010796491228070176, + "loss": 3.3977, + "step": 9769 + }, + { + "epoch": 0.5142105263157895, + "grad_norm": 0.478515625, + "learning_rate": 0.00010795321637426901, + "loss": 3.7562, + "step": 9770 + }, + { + "epoch": 0.5142631578947369, + "grad_norm": 0.49609375, + "learning_rate": 0.00010794152046783626, + "loss": 3.58, + "step": 9771 + }, + { + "epoch": 0.5143157894736842, + "grad_norm": 0.482421875, + "learning_rate": 0.00010792982456140351, + "loss": 3.4952, + "step": 9772 + }, + { + "epoch": 0.5143684210526316, + "grad_norm": 0.48828125, + "learning_rate": 0.00010791812865497077, + "loss": 2.9646, + "step": 9773 + }, + { + "epoch": 0.514421052631579, + "grad_norm": 0.474609375, + "learning_rate": 0.00010790643274853802, + "loss": 3.7258, + "step": 9774 + }, + { + "epoch": 0.5144736842105263, + "grad_norm": 0.48828125, + "learning_rate": 0.00010789473684210527, + "loss": 3.6328, + "step": 9775 + }, + { + "epoch": 0.5145263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00010788304093567252, + "loss": 3.4736, + "step": 9776 + }, + { + "epoch": 0.514578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00010787134502923977, + "loss": 3.338, + "step": 9777 + }, + { + "epoch": 0.5146315789473684, + "grad_norm": 0.50390625, + "learning_rate": 0.00010785964912280702, + "loss": 2.9494, + "step": 9778 + }, + { + "epoch": 0.5146842105263157, + "grad_norm": 0.53125, + "learning_rate": 0.00010784795321637427, + "loss": 3.8227, + "step": 9779 + }, + { + "epoch": 0.5147368421052632, + "grad_norm": 0.48046875, + "learning_rate": 0.00010783625730994153, + "loss": 3.3376, + "step": 9780 + }, + { + "epoch": 0.5147894736842106, + "grad_norm": 0.466796875, + "learning_rate": 0.00010782456140350878, + "loss": 3.3518, + "step": 9781 + }, + { + "epoch": 0.5148421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00010781286549707603, + "loss": 3.0129, + "step": 9782 + }, + { + "epoch": 0.5148947368421053, + "grad_norm": 0.478515625, + "learning_rate": 0.00010780116959064328, + "loss": 3.4369, + "step": 9783 + }, + { + "epoch": 0.5149473684210526, + "grad_norm": 0.546875, + "learning_rate": 0.00010778947368421053, + "loss": 3.4167, + "step": 9784 + }, + { + "epoch": 0.515, + "grad_norm": 0.5, + "learning_rate": 0.00010777777777777778, + "loss": 3.4575, + "step": 9785 + }, + { + "epoch": 0.5150526315789473, + "grad_norm": 0.451171875, + "learning_rate": 0.00010776608187134503, + "loss": 3.2882, + "step": 9786 + }, + { + "epoch": 0.5151052631578947, + "grad_norm": 0.5625, + "learning_rate": 0.00010775438596491229, + "loss": 3.2497, + "step": 9787 + }, + { + "epoch": 0.5151578947368421, + "grad_norm": 0.48828125, + "learning_rate": 0.00010774269005847954, + "loss": 3.3941, + "step": 9788 + }, + { + "epoch": 0.5152105263157895, + "grad_norm": 0.47265625, + "learning_rate": 0.00010773099415204679, + "loss": 3.8082, + "step": 9789 + }, + { + "epoch": 0.5152631578947369, + "grad_norm": 0.6875, + "learning_rate": 0.00010771929824561404, + "loss": 3.3104, + "step": 9790 + }, + { + "epoch": 0.5153157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00010770760233918129, + "loss": 3.283, + "step": 9791 + }, + { + "epoch": 0.5153684210526316, + "grad_norm": 0.46875, + "learning_rate": 0.00010769590643274854, + "loss": 3.5206, + "step": 9792 + }, + { + "epoch": 0.515421052631579, + "grad_norm": 0.4921875, + "learning_rate": 0.00010768421052631579, + "loss": 3.5897, + "step": 9793 + }, + { + "epoch": 0.5154736842105263, + "grad_norm": 0.46484375, + "learning_rate": 0.00010767251461988304, + "loss": 3.2503, + "step": 9794 + }, + { + "epoch": 0.5155263157894737, + "grad_norm": 0.484375, + "learning_rate": 0.0001076608187134503, + "loss": 3.7275, + "step": 9795 + }, + { + "epoch": 0.515578947368421, + "grad_norm": 0.45703125, + "learning_rate": 0.00010764912280701755, + "loss": 3.4548, + "step": 9796 + }, + { + "epoch": 0.5156315789473684, + "grad_norm": 0.490234375, + "learning_rate": 0.0001076374269005848, + "loss": 3.5115, + "step": 9797 + }, + { + "epoch": 0.5156842105263157, + "grad_norm": 0.478515625, + "learning_rate": 0.00010762573099415205, + "loss": 3.657, + "step": 9798 + }, + { + "epoch": 0.5157368421052632, + "grad_norm": 0.494140625, + "learning_rate": 0.0001076140350877193, + "loss": 3.7128, + "step": 9799 + }, + { + "epoch": 0.5157894736842106, + "grad_norm": 0.458984375, + "learning_rate": 0.00010760233918128655, + "loss": 3.4749, + "step": 9800 + }, + { + "epoch": 0.5158421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.0001075906432748538, + "loss": 3.2591, + "step": 9801 + }, + { + "epoch": 0.5158947368421053, + "grad_norm": 0.4921875, + "learning_rate": 0.00010757894736842106, + "loss": 3.2677, + "step": 9802 + }, + { + "epoch": 0.5159473684210526, + "grad_norm": 0.48046875, + "learning_rate": 0.00010756725146198831, + "loss": 3.4102, + "step": 9803 + }, + { + "epoch": 0.516, + "grad_norm": 0.486328125, + "learning_rate": 0.00010755555555555556, + "loss": 3.6495, + "step": 9804 + }, + { + "epoch": 0.5160526315789473, + "grad_norm": 0.46875, + "learning_rate": 0.00010754385964912281, + "loss": 3.5511, + "step": 9805 + }, + { + "epoch": 0.5161052631578947, + "grad_norm": 0.5078125, + "learning_rate": 0.00010753216374269006, + "loss": 3.5267, + "step": 9806 + }, + { + "epoch": 0.5161578947368421, + "grad_norm": 0.46875, + "learning_rate": 0.00010752046783625731, + "loss": 3.3051, + "step": 9807 + }, + { + "epoch": 0.5162105263157895, + "grad_norm": 0.486328125, + "learning_rate": 0.00010750877192982456, + "loss": 4.012, + "step": 9808 + }, + { + "epoch": 0.5162631578947369, + "grad_norm": 0.51953125, + "learning_rate": 0.00010749707602339181, + "loss": 3.2923, + "step": 9809 + }, + { + "epoch": 0.5163157894736842, + "grad_norm": 0.57421875, + "learning_rate": 0.00010748538011695907, + "loss": 3.666, + "step": 9810 + }, + { + "epoch": 0.5163684210526316, + "grad_norm": 0.5078125, + "learning_rate": 0.00010747368421052632, + "loss": 3.5689, + "step": 9811 + }, + { + "epoch": 0.516421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00010746198830409357, + "loss": 3.3552, + "step": 9812 + }, + { + "epoch": 0.5164736842105263, + "grad_norm": 0.494140625, + "learning_rate": 0.00010745029239766082, + "loss": 3.8096, + "step": 9813 + }, + { + "epoch": 0.5165263157894737, + "grad_norm": 0.490234375, + "learning_rate": 0.00010743859649122807, + "loss": 3.4322, + "step": 9814 + }, + { + "epoch": 0.516578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.00010742690058479532, + "loss": 2.9867, + "step": 9815 + }, + { + "epoch": 0.5166315789473684, + "grad_norm": 0.498046875, + "learning_rate": 0.00010741520467836257, + "loss": 3.236, + "step": 9816 + }, + { + "epoch": 0.5166842105263157, + "grad_norm": 0.48046875, + "learning_rate": 0.00010740350877192983, + "loss": 3.6225, + "step": 9817 + }, + { + "epoch": 0.5167368421052632, + "grad_norm": 0.458984375, + "learning_rate": 0.00010739181286549708, + "loss": 3.5949, + "step": 9818 + }, + { + "epoch": 0.5167894736842106, + "grad_norm": 0.50390625, + "learning_rate": 0.00010738011695906433, + "loss": 3.3017, + "step": 9819 + }, + { + "epoch": 0.5168421052631579, + "grad_norm": 0.498046875, + "learning_rate": 0.00010736842105263158, + "loss": 3.3628, + "step": 9820 + }, + { + "epoch": 0.5168947368421053, + "grad_norm": 0.5234375, + "learning_rate": 0.00010735672514619883, + "loss": 3.7157, + "step": 9821 + }, + { + "epoch": 0.5169473684210526, + "grad_norm": 0.57421875, + "learning_rate": 0.00010734502923976608, + "loss": 3.3531, + "step": 9822 + }, + { + "epoch": 0.517, + "grad_norm": 0.53515625, + "learning_rate": 0.00010733333333333333, + "loss": 3.2245, + "step": 9823 + }, + { + "epoch": 0.5170526315789473, + "grad_norm": 0.49609375, + "learning_rate": 0.0001073216374269006, + "loss": 3.604, + "step": 9824 + }, + { + "epoch": 0.5171052631578947, + "grad_norm": 0.53125, + "learning_rate": 0.00010730994152046784, + "loss": 3.3234, + "step": 9825 + }, + { + "epoch": 0.5171578947368421, + "grad_norm": 0.498046875, + "learning_rate": 0.0001072982456140351, + "loss": 3.7144, + "step": 9826 + }, + { + "epoch": 0.5172105263157895, + "grad_norm": 0.546875, + "learning_rate": 0.00010728654970760234, + "loss": 3.3864, + "step": 9827 + }, + { + "epoch": 0.5172631578947369, + "grad_norm": 0.51171875, + "learning_rate": 0.00010727485380116959, + "loss": 3.6885, + "step": 9828 + }, + { + "epoch": 0.5173157894736842, + "grad_norm": 0.51953125, + "learning_rate": 0.00010726315789473684, + "loss": 3.6386, + "step": 9829 + }, + { + "epoch": 0.5173684210526316, + "grad_norm": 0.51171875, + "learning_rate": 0.00010725146198830409, + "loss": 3.2684, + "step": 9830 + }, + { + "epoch": 0.5174210526315789, + "grad_norm": 0.482421875, + "learning_rate": 0.00010723976608187134, + "loss": 3.4706, + "step": 9831 + }, + { + "epoch": 0.5174736842105263, + "grad_norm": 0.47265625, + "learning_rate": 0.0001072280701754386, + "loss": 3.2981, + "step": 9832 + }, + { + "epoch": 0.5175263157894737, + "grad_norm": 0.48828125, + "learning_rate": 0.00010721637426900585, + "loss": 3.5584, + "step": 9833 + }, + { + "epoch": 0.517578947368421, + "grad_norm": 0.494140625, + "learning_rate": 0.0001072046783625731, + "loss": 3.6531, + "step": 9834 + }, + { + "epoch": 0.5176315789473684, + "grad_norm": 0.46875, + "learning_rate": 0.00010719298245614035, + "loss": 3.4429, + "step": 9835 + }, + { + "epoch": 0.5176842105263157, + "grad_norm": 0.4765625, + "learning_rate": 0.0001071812865497076, + "loss": 3.4255, + "step": 9836 + }, + { + "epoch": 0.5177368421052632, + "grad_norm": 0.4921875, + "learning_rate": 0.00010716959064327485, + "loss": 3.3714, + "step": 9837 + }, + { + "epoch": 0.5177894736842106, + "grad_norm": 0.53125, + "learning_rate": 0.0001071578947368421, + "loss": 3.2239, + "step": 9838 + }, + { + "epoch": 0.5178421052631579, + "grad_norm": 0.51953125, + "learning_rate": 0.00010714619883040938, + "loss": 3.5445, + "step": 9839 + }, + { + "epoch": 0.5178947368421053, + "grad_norm": 0.48828125, + "learning_rate": 0.00010713450292397661, + "loss": 3.133, + "step": 9840 + }, + { + "epoch": 0.5179473684210526, + "grad_norm": 0.515625, + "learning_rate": 0.00010712280701754386, + "loss": 3.0792, + "step": 9841 + }, + { + "epoch": 0.518, + "grad_norm": 0.53515625, + "learning_rate": 0.00010711111111111111, + "loss": 3.6089, + "step": 9842 + }, + { + "epoch": 0.5180526315789473, + "grad_norm": 0.458984375, + "learning_rate": 0.00010709941520467836, + "loss": 3.1519, + "step": 9843 + }, + { + "epoch": 0.5181052631578947, + "grad_norm": 0.50390625, + "learning_rate": 0.00010708771929824561, + "loss": 3.4491, + "step": 9844 + }, + { + "epoch": 0.5181578947368422, + "grad_norm": 0.6328125, + "learning_rate": 0.00010707602339181286, + "loss": 3.0281, + "step": 9845 + }, + { + "epoch": 0.5182105263157895, + "grad_norm": 0.494140625, + "learning_rate": 0.00010706432748538011, + "loss": 3.2753, + "step": 9846 + }, + { + "epoch": 0.5182631578947369, + "grad_norm": 0.5078125, + "learning_rate": 0.00010705263157894738, + "loss": 3.3001, + "step": 9847 + }, + { + "epoch": 0.5183157894736842, + "grad_norm": 0.494140625, + "learning_rate": 0.00010704093567251463, + "loss": 3.5156, + "step": 9848 + }, + { + "epoch": 0.5183684210526316, + "grad_norm": 0.46875, + "learning_rate": 0.00010702923976608187, + "loss": 3.8511, + "step": 9849 + }, + { + "epoch": 0.5184210526315789, + "grad_norm": 0.54296875, + "learning_rate": 0.00010701754385964912, + "loss": 3.6885, + "step": 9850 + }, + { + "epoch": 0.5184736842105263, + "grad_norm": 0.53515625, + "learning_rate": 0.00010700584795321637, + "loss": 3.7715, + "step": 9851 + }, + { + "epoch": 0.5185263157894737, + "grad_norm": 0.5, + "learning_rate": 0.00010699415204678362, + "loss": 3.4504, + "step": 9852 + }, + { + "epoch": 0.518578947368421, + "grad_norm": 0.470703125, + "learning_rate": 0.00010698245614035087, + "loss": 3.7004, + "step": 9853 + }, + { + "epoch": 0.5186315789473684, + "grad_norm": 0.44140625, + "learning_rate": 0.00010697076023391815, + "loss": 3.2305, + "step": 9854 + }, + { + "epoch": 0.5186842105263157, + "grad_norm": 0.484375, + "learning_rate": 0.00010695906432748539, + "loss": 3.452, + "step": 9855 + }, + { + "epoch": 0.5187368421052632, + "grad_norm": 0.498046875, + "learning_rate": 0.00010694736842105264, + "loss": 3.4372, + "step": 9856 + }, + { + "epoch": 0.5187894736842106, + "grad_norm": 0.515625, + "learning_rate": 0.00010693567251461989, + "loss": 3.5167, + "step": 9857 + }, + { + "epoch": 0.5188421052631579, + "grad_norm": 0.494140625, + "learning_rate": 0.00010692397660818713, + "loss": 3.4785, + "step": 9858 + }, + { + "epoch": 0.5188947368421053, + "grad_norm": 0.49609375, + "learning_rate": 0.00010691228070175438, + "loss": 3.3966, + "step": 9859 + }, + { + "epoch": 0.5189473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00010690058479532163, + "loss": 3.598, + "step": 9860 + }, + { + "epoch": 0.519, + "grad_norm": 0.46875, + "learning_rate": 0.00010688888888888891, + "loss": 3.4247, + "step": 9861 + }, + { + "epoch": 0.5190526315789473, + "grad_norm": 0.46484375, + "learning_rate": 0.00010687719298245616, + "loss": 3.0822, + "step": 9862 + }, + { + "epoch": 0.5191052631578947, + "grad_norm": 0.48828125, + "learning_rate": 0.0001068654970760234, + "loss": 2.9391, + "step": 9863 + }, + { + "epoch": 0.5191578947368422, + "grad_norm": 0.7109375, + "learning_rate": 0.00010685380116959065, + "loss": 3.3333, + "step": 9864 + }, + { + "epoch": 0.5192105263157895, + "grad_norm": 0.462890625, + "learning_rate": 0.0001068421052631579, + "loss": 3.4462, + "step": 9865 + }, + { + "epoch": 0.5192631578947369, + "grad_norm": 0.4765625, + "learning_rate": 0.00010683040935672514, + "loss": 3.222, + "step": 9866 + }, + { + "epoch": 0.5193157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.0001068187134502924, + "loss": 3.9157, + "step": 9867 + }, + { + "epoch": 0.5193684210526316, + "grad_norm": 0.490234375, + "learning_rate": 0.00010680701754385964, + "loss": 3.5523, + "step": 9868 + }, + { + "epoch": 0.5194210526315789, + "grad_norm": 0.5859375, + "learning_rate": 0.00010679532163742692, + "loss": 3.5682, + "step": 9869 + }, + { + "epoch": 0.5194736842105263, + "grad_norm": 0.5078125, + "learning_rate": 0.00010678362573099416, + "loss": 3.6203, + "step": 9870 + }, + { + "epoch": 0.5195263157894737, + "grad_norm": 0.484375, + "learning_rate": 0.0001067719298245614, + "loss": 3.0142, + "step": 9871 + }, + { + "epoch": 0.519578947368421, + "grad_norm": 0.48046875, + "learning_rate": 0.00010676023391812866, + "loss": 3.5746, + "step": 9872 + }, + { + "epoch": 0.5196315789473684, + "grad_norm": 0.498046875, + "learning_rate": 0.0001067485380116959, + "loss": 3.3073, + "step": 9873 + }, + { + "epoch": 0.5196842105263157, + "grad_norm": 0.48046875, + "learning_rate": 0.00010673684210526316, + "loss": 3.8364, + "step": 9874 + }, + { + "epoch": 0.5197368421052632, + "grad_norm": 0.51171875, + "learning_rate": 0.0001067251461988304, + "loss": 3.4512, + "step": 9875 + }, + { + "epoch": 0.5197894736842106, + "grad_norm": 0.4921875, + "learning_rate": 0.00010671345029239768, + "loss": 3.3702, + "step": 9876 + }, + { + "epoch": 0.5198421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00010670175438596493, + "loss": 3.2974, + "step": 9877 + }, + { + "epoch": 0.5198947368421053, + "grad_norm": 1.5546875, + "learning_rate": 0.00010669005847953217, + "loss": 2.9078, + "step": 9878 + }, + { + "epoch": 0.5199473684210526, + "grad_norm": 0.44140625, + "learning_rate": 0.00010667836257309942, + "loss": 3.5886, + "step": 9879 + }, + { + "epoch": 0.52, + "grad_norm": 0.48046875, + "learning_rate": 0.00010666666666666667, + "loss": 3.7199, + "step": 9880 + }, + { + "epoch": 0.5200526315789473, + "grad_norm": 0.498046875, + "learning_rate": 0.00010665497076023392, + "loss": 3.0156, + "step": 9881 + }, + { + "epoch": 0.5201052631578947, + "grad_norm": 0.494140625, + "learning_rate": 0.00010664327485380117, + "loss": 3.3628, + "step": 9882 + }, + { + "epoch": 0.5201578947368422, + "grad_norm": 0.4921875, + "learning_rate": 0.00010663157894736842, + "loss": 3.6104, + "step": 9883 + }, + { + "epoch": 0.5202105263157895, + "grad_norm": 0.4765625, + "learning_rate": 0.00010661988304093569, + "loss": 3.304, + "step": 9884 + }, + { + "epoch": 0.5202631578947369, + "grad_norm": 0.5078125, + "learning_rate": 0.00010660818713450294, + "loss": 3.2337, + "step": 9885 + }, + { + "epoch": 0.5203157894736842, + "grad_norm": 0.49609375, + "learning_rate": 0.00010659649122807018, + "loss": 3.2978, + "step": 9886 + }, + { + "epoch": 0.5203684210526316, + "grad_norm": 0.46875, + "learning_rate": 0.00010658479532163743, + "loss": 3.2402, + "step": 9887 + }, + { + "epoch": 0.5204210526315789, + "grad_norm": 0.44921875, + "learning_rate": 0.00010657309941520468, + "loss": 3.3312, + "step": 9888 + }, + { + "epoch": 0.5204736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00010656140350877193, + "loss": 3.229, + "step": 9889 + }, + { + "epoch": 0.5205263157894737, + "grad_norm": 3.4375, + "learning_rate": 0.00010654970760233918, + "loss": 3.423, + "step": 9890 + }, + { + "epoch": 0.520578947368421, + "grad_norm": 0.6328125, + "learning_rate": 0.00010653801169590645, + "loss": 3.4574, + "step": 9891 + }, + { + "epoch": 0.5206315789473684, + "grad_norm": 0.5078125, + "learning_rate": 0.0001065263157894737, + "loss": 3.664, + "step": 9892 + }, + { + "epoch": 0.5206842105263157, + "grad_norm": 0.53515625, + "learning_rate": 0.00010651461988304094, + "loss": 3.3692, + "step": 9893 + }, + { + "epoch": 0.5207368421052632, + "grad_norm": 0.515625, + "learning_rate": 0.00010650292397660819, + "loss": 3.8204, + "step": 9894 + }, + { + "epoch": 0.5207894736842106, + "grad_norm": 0.453125, + "learning_rate": 0.00010649122807017544, + "loss": 3.3615, + "step": 9895 + }, + { + "epoch": 0.5208421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00010647953216374269, + "loss": 3.1433, + "step": 9896 + }, + { + "epoch": 0.5208947368421053, + "grad_norm": 0.5, + "learning_rate": 0.00010646783625730994, + "loss": 3.6194, + "step": 9897 + }, + { + "epoch": 0.5209473684210526, + "grad_norm": 0.46875, + "learning_rate": 0.00010645614035087719, + "loss": 3.5292, + "step": 9898 + }, + { + "epoch": 0.521, + "grad_norm": 0.51953125, + "learning_rate": 0.00010644444444444446, + "loss": 3.4467, + "step": 9899 + }, + { + "epoch": 0.5210526315789473, + "grad_norm": 0.494140625, + "learning_rate": 0.00010643274853801171, + "loss": 3.1448, + "step": 9900 + }, + { + "epoch": 0.5211052631578947, + "grad_norm": 0.490234375, + "learning_rate": 0.00010642105263157895, + "loss": 3.3876, + "step": 9901 + }, + { + "epoch": 0.5211578947368422, + "grad_norm": 0.48828125, + "learning_rate": 0.0001064093567251462, + "loss": 3.611, + "step": 9902 + }, + { + "epoch": 0.5212105263157895, + "grad_norm": 0.482421875, + "learning_rate": 0.00010639766081871345, + "loss": 3.6629, + "step": 9903 + }, + { + "epoch": 0.5212631578947369, + "grad_norm": 0.494140625, + "learning_rate": 0.0001063859649122807, + "loss": 3.3414, + "step": 9904 + }, + { + "epoch": 0.5213157894736842, + "grad_norm": 0.474609375, + "learning_rate": 0.00010637426900584795, + "loss": 3.5916, + "step": 9905 + }, + { + "epoch": 0.5213684210526316, + "grad_norm": 0.4921875, + "learning_rate": 0.00010636257309941522, + "loss": 3.146, + "step": 9906 + }, + { + "epoch": 0.5214210526315789, + "grad_norm": 0.48828125, + "learning_rate": 0.00010635087719298247, + "loss": 3.3735, + "step": 9907 + }, + { + "epoch": 0.5214736842105263, + "grad_norm": 0.466796875, + "learning_rate": 0.00010633918128654972, + "loss": 3.1552, + "step": 9908 + }, + { + "epoch": 0.5215263157894737, + "grad_norm": 0.478515625, + "learning_rate": 0.00010632748538011696, + "loss": 3.1802, + "step": 9909 + }, + { + "epoch": 0.521578947368421, + "grad_norm": 0.48828125, + "learning_rate": 0.00010631578947368421, + "loss": 3.505, + "step": 9910 + }, + { + "epoch": 0.5216315789473684, + "grad_norm": 0.474609375, + "learning_rate": 0.00010630409356725146, + "loss": 3.6745, + "step": 9911 + }, + { + "epoch": 0.5216842105263158, + "grad_norm": 0.490234375, + "learning_rate": 0.00010629239766081871, + "loss": 3.4064, + "step": 9912 + }, + { + "epoch": 0.5217368421052632, + "grad_norm": 0.5, + "learning_rate": 0.00010628070175438598, + "loss": 3.0269, + "step": 9913 + }, + { + "epoch": 0.5217894736842106, + "grad_norm": 0.51171875, + "learning_rate": 0.00010626900584795323, + "loss": 3.5868, + "step": 9914 + }, + { + "epoch": 0.5218421052631579, + "grad_norm": 0.5390625, + "learning_rate": 0.00010625730994152048, + "loss": 3.812, + "step": 9915 + }, + { + "epoch": 0.5218947368421053, + "grad_norm": 0.49609375, + "learning_rate": 0.00010624561403508772, + "loss": 3.3301, + "step": 9916 + }, + { + "epoch": 0.5219473684210526, + "grad_norm": 0.453125, + "learning_rate": 0.00010623391812865497, + "loss": 2.9973, + "step": 9917 + }, + { + "epoch": 0.522, + "grad_norm": 0.486328125, + "learning_rate": 0.00010622222222222222, + "loss": 3.4131, + "step": 9918 + }, + { + "epoch": 0.5220526315789473, + "grad_norm": 0.466796875, + "learning_rate": 0.00010621052631578947, + "loss": 3.2655, + "step": 9919 + }, + { + "epoch": 0.5221052631578947, + "grad_norm": 0.484375, + "learning_rate": 0.00010619883040935672, + "loss": 3.4448, + "step": 9920 + }, + { + "epoch": 0.5221578947368422, + "grad_norm": 0.462890625, + "learning_rate": 0.000106187134502924, + "loss": 3.4345, + "step": 9921 + }, + { + "epoch": 0.5222105263157895, + "grad_norm": 0.5078125, + "learning_rate": 0.00010617543859649124, + "loss": 3.4318, + "step": 9922 + }, + { + "epoch": 0.5222631578947369, + "grad_norm": 0.5546875, + "learning_rate": 0.0001061637426900585, + "loss": 3.6976, + "step": 9923 + }, + { + "epoch": 0.5223157894736842, + "grad_norm": 0.52734375, + "learning_rate": 0.00010615204678362573, + "loss": 3.2801, + "step": 9924 + }, + { + "epoch": 0.5223684210526316, + "grad_norm": 0.58984375, + "learning_rate": 0.00010614035087719298, + "loss": 3.104, + "step": 9925 + }, + { + "epoch": 0.5224210526315789, + "grad_norm": 0.515625, + "learning_rate": 0.00010612865497076023, + "loss": 3.6544, + "step": 9926 + }, + { + "epoch": 0.5224736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.00010611695906432748, + "loss": 3.64, + "step": 9927 + }, + { + "epoch": 0.5225263157894737, + "grad_norm": 0.490234375, + "learning_rate": 0.00010610526315789476, + "loss": 3.5744, + "step": 9928 + }, + { + "epoch": 0.522578947368421, + "grad_norm": 0.474609375, + "learning_rate": 0.000106093567251462, + "loss": 3.358, + "step": 9929 + }, + { + "epoch": 0.5226315789473684, + "grad_norm": 0.46484375, + "learning_rate": 0.00010608187134502925, + "loss": 2.8739, + "step": 9930 + }, + { + "epoch": 0.5226842105263158, + "grad_norm": 0.462890625, + "learning_rate": 0.0001060701754385965, + "loss": 3.2916, + "step": 9931 + }, + { + "epoch": 0.5227368421052632, + "grad_norm": 0.48046875, + "learning_rate": 0.00010605847953216374, + "loss": 3.505, + "step": 9932 + }, + { + "epoch": 0.5227894736842106, + "grad_norm": 0.5390625, + "learning_rate": 0.00010604678362573099, + "loss": 3.322, + "step": 9933 + }, + { + "epoch": 0.5228421052631579, + "grad_norm": 0.486328125, + "learning_rate": 0.00010603508771929824, + "loss": 3.6015, + "step": 9934 + }, + { + "epoch": 0.5228947368421053, + "grad_norm": 0.4765625, + "learning_rate": 0.00010602339181286549, + "loss": 3.1957, + "step": 9935 + }, + { + "epoch": 0.5229473684210526, + "grad_norm": 0.55859375, + "learning_rate": 0.00010601169590643277, + "loss": 3.1845, + "step": 9936 + }, + { + "epoch": 0.523, + "grad_norm": 0.5078125, + "learning_rate": 0.00010600000000000002, + "loss": 3.5114, + "step": 9937 + }, + { + "epoch": 0.5230526315789473, + "grad_norm": 0.478515625, + "learning_rate": 0.00010598830409356726, + "loss": 3.4301, + "step": 9938 + }, + { + "epoch": 0.5231052631578947, + "grad_norm": 0.46484375, + "learning_rate": 0.0001059766081871345, + "loss": 3.7288, + "step": 9939 + }, + { + "epoch": 0.5231578947368422, + "grad_norm": 0.53515625, + "learning_rate": 0.00010596491228070175, + "loss": 3.3199, + "step": 9940 + }, + { + "epoch": 0.5232105263157895, + "grad_norm": 0.482421875, + "learning_rate": 0.000105953216374269, + "loss": 3.516, + "step": 9941 + }, + { + "epoch": 0.5232631578947369, + "grad_norm": 0.58984375, + "learning_rate": 0.00010594152046783625, + "loss": 3.1005, + "step": 9942 + }, + { + "epoch": 0.5233157894736842, + "grad_norm": 0.48828125, + "learning_rate": 0.00010592982456140353, + "loss": 3.0247, + "step": 9943 + }, + { + "epoch": 0.5233684210526316, + "grad_norm": 0.453125, + "learning_rate": 0.00010591812865497078, + "loss": 3.5275, + "step": 9944 + }, + { + "epoch": 0.5234210526315789, + "grad_norm": 0.51953125, + "learning_rate": 0.00010590643274853803, + "loss": 3.4855, + "step": 9945 + }, + { + "epoch": 0.5234736842105263, + "grad_norm": 0.48828125, + "learning_rate": 0.00010589473684210527, + "loss": 4.0005, + "step": 9946 + }, + { + "epoch": 0.5235263157894737, + "grad_norm": 0.46484375, + "learning_rate": 0.00010588304093567251, + "loss": 3.6187, + "step": 9947 + }, + { + "epoch": 0.523578947368421, + "grad_norm": 0.482421875, + "learning_rate": 0.00010587134502923976, + "loss": 3.0362, + "step": 9948 + }, + { + "epoch": 0.5236315789473684, + "grad_norm": 0.470703125, + "learning_rate": 0.00010585964912280701, + "loss": 3.1123, + "step": 9949 + }, + { + "epoch": 0.5236842105263158, + "grad_norm": 0.50390625, + "learning_rate": 0.00010584795321637429, + "loss": 3.1194, + "step": 9950 + }, + { + "epoch": 0.5237368421052632, + "grad_norm": 0.48046875, + "learning_rate": 0.00010583625730994154, + "loss": 3.3552, + "step": 9951 + }, + { + "epoch": 0.5237894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.00010582456140350879, + "loss": 3.8732, + "step": 9952 + }, + { + "epoch": 0.5238421052631579, + "grad_norm": 0.546875, + "learning_rate": 0.00010581286549707604, + "loss": 2.8752, + "step": 9953 + }, + { + "epoch": 0.5238947368421053, + "grad_norm": 0.48046875, + "learning_rate": 0.00010580116959064329, + "loss": 3.6826, + "step": 9954 + }, + { + "epoch": 0.5239473684210526, + "grad_norm": 0.49609375, + "learning_rate": 0.00010578947368421052, + "loss": 3.7434, + "step": 9955 + }, + { + "epoch": 0.524, + "grad_norm": 0.474609375, + "learning_rate": 0.00010577777777777777, + "loss": 3.7833, + "step": 9956 + }, + { + "epoch": 0.5240526315789473, + "grad_norm": 0.484375, + "learning_rate": 0.00010576608187134502, + "loss": 3.5048, + "step": 9957 + }, + { + "epoch": 0.5241052631578947, + "grad_norm": 0.494140625, + "learning_rate": 0.0001057543859649123, + "loss": 3.0869, + "step": 9958 + }, + { + "epoch": 0.5241578947368422, + "grad_norm": 0.451171875, + "learning_rate": 0.00010574269005847955, + "loss": 3.4684, + "step": 9959 + }, + { + "epoch": 0.5242105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.0001057309941520468, + "loss": 3.1517, + "step": 9960 + }, + { + "epoch": 0.5242631578947369, + "grad_norm": 0.490234375, + "learning_rate": 0.00010571929824561405, + "loss": 3.6313, + "step": 9961 + }, + { + "epoch": 0.5243157894736842, + "grad_norm": 0.53125, + "learning_rate": 0.00010570760233918128, + "loss": 3.5627, + "step": 9962 + }, + { + "epoch": 0.5243684210526316, + "grad_norm": 0.53125, + "learning_rate": 0.00010569590643274853, + "loss": 3.2526, + "step": 9963 + }, + { + "epoch": 0.5244210526315789, + "grad_norm": 0.5078125, + "learning_rate": 0.00010568421052631578, + "loss": 3.2555, + "step": 9964 + }, + { + "epoch": 0.5244736842105263, + "grad_norm": 0.4921875, + "learning_rate": 0.00010567251461988306, + "loss": 3.275, + "step": 9965 + }, + { + "epoch": 0.5245263157894737, + "grad_norm": 0.49609375, + "learning_rate": 0.00010566081871345031, + "loss": 3.714, + "step": 9966 + }, + { + "epoch": 0.524578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.00010564912280701756, + "loss": 3.4904, + "step": 9967 + }, + { + "epoch": 0.5246315789473684, + "grad_norm": 0.478515625, + "learning_rate": 0.0001056374269005848, + "loss": 3.6001, + "step": 9968 + }, + { + "epoch": 0.5246842105263158, + "grad_norm": 0.474609375, + "learning_rate": 0.00010562573099415206, + "loss": 3.5969, + "step": 9969 + }, + { + "epoch": 0.5247368421052632, + "grad_norm": 0.48828125, + "learning_rate": 0.00010561403508771929, + "loss": 3.4938, + "step": 9970 + }, + { + "epoch": 0.5247894736842105, + "grad_norm": 0.494140625, + "learning_rate": 0.00010560233918128654, + "loss": 3.5342, + "step": 9971 + }, + { + "epoch": 0.5248421052631579, + "grad_norm": 0.486328125, + "learning_rate": 0.00010559064327485379, + "loss": 3.305, + "step": 9972 + }, + { + "epoch": 0.5248947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.00010557894736842107, + "loss": 3.0823, + "step": 9973 + }, + { + "epoch": 0.5249473684210526, + "grad_norm": 0.484375, + "learning_rate": 0.00010556725146198832, + "loss": 3.6517, + "step": 9974 + }, + { + "epoch": 0.525, + "grad_norm": 0.515625, + "learning_rate": 0.00010555555555555557, + "loss": 3.4598, + "step": 9975 + }, + { + "epoch": 0.5250526315789473, + "grad_norm": 0.51953125, + "learning_rate": 0.00010554385964912282, + "loss": 3.4057, + "step": 9976 + }, + { + "epoch": 0.5251052631578947, + "grad_norm": 0.4921875, + "learning_rate": 0.00010553216374269007, + "loss": 3.4323, + "step": 9977 + }, + { + "epoch": 0.5251578947368422, + "grad_norm": 0.482421875, + "learning_rate": 0.0001055204678362573, + "loss": 3.2358, + "step": 9978 + }, + { + "epoch": 0.5252105263157895, + "grad_norm": 0.484375, + "learning_rate": 0.00010550877192982455, + "loss": 3.2653, + "step": 9979 + }, + { + "epoch": 0.5252631578947369, + "grad_norm": 0.47265625, + "learning_rate": 0.00010549707602339183, + "loss": 3.1167, + "step": 9980 + }, + { + "epoch": 0.5253157894736842, + "grad_norm": 0.474609375, + "learning_rate": 0.00010548538011695908, + "loss": 3.239, + "step": 9981 + }, + { + "epoch": 0.5253684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00010547368421052633, + "loss": 3.4109, + "step": 9982 + }, + { + "epoch": 0.5254210526315789, + "grad_norm": 0.53125, + "learning_rate": 0.00010546198830409358, + "loss": 3.4958, + "step": 9983 + }, + { + "epoch": 0.5254736842105263, + "grad_norm": 0.53515625, + "learning_rate": 0.00010545029239766083, + "loss": 3.4262, + "step": 9984 + }, + { + "epoch": 0.5255263157894737, + "grad_norm": 0.5390625, + "learning_rate": 0.00010543859649122806, + "loss": 3.4761, + "step": 9985 + }, + { + "epoch": 0.525578947368421, + "grad_norm": 0.5390625, + "learning_rate": 0.00010542690058479531, + "loss": 2.704, + "step": 9986 + }, + { + "epoch": 0.5256315789473684, + "grad_norm": 0.578125, + "learning_rate": 0.00010541520467836259, + "loss": 3.5765, + "step": 9987 + }, + { + "epoch": 0.5256842105263158, + "grad_norm": 0.490234375, + "learning_rate": 0.00010540350877192984, + "loss": 3.5874, + "step": 9988 + }, + { + "epoch": 0.5257368421052632, + "grad_norm": 0.546875, + "learning_rate": 0.00010539181286549709, + "loss": 3.3762, + "step": 9989 + }, + { + "epoch": 0.5257894736842105, + "grad_norm": 0.486328125, + "learning_rate": 0.00010538011695906434, + "loss": 3.265, + "step": 9990 + }, + { + "epoch": 0.5258421052631579, + "grad_norm": 0.494140625, + "learning_rate": 0.00010536842105263159, + "loss": 3.5558, + "step": 9991 + }, + { + "epoch": 0.5258947368421053, + "grad_norm": 0.609375, + "learning_rate": 0.00010535672514619884, + "loss": 3.7366, + "step": 9992 + }, + { + "epoch": 0.5259473684210526, + "grad_norm": 0.5078125, + "learning_rate": 0.00010534502923976607, + "loss": 3.4266, + "step": 9993 + }, + { + "epoch": 0.526, + "grad_norm": 0.48828125, + "learning_rate": 0.00010533333333333332, + "loss": 3.197, + "step": 9994 + }, + { + "epoch": 0.5260526315789473, + "grad_norm": 0.4765625, + "learning_rate": 0.0001053216374269006, + "loss": 3.3499, + "step": 9995 + }, + { + "epoch": 0.5261052631578947, + "grad_norm": 0.86328125, + "learning_rate": 0.00010530994152046785, + "loss": 3.2226, + "step": 9996 + }, + { + "epoch": 0.5261578947368422, + "grad_norm": 0.48046875, + "learning_rate": 0.0001052982456140351, + "loss": 3.647, + "step": 9997 + }, + { + "epoch": 0.5262105263157895, + "grad_norm": 0.46484375, + "learning_rate": 0.00010528654970760235, + "loss": 3.5694, + "step": 9998 + }, + { + "epoch": 0.5262631578947369, + "grad_norm": 0.53515625, + "learning_rate": 0.0001052748538011696, + "loss": 3.5548, + "step": 9999 + }, + { + "epoch": 0.5263157894736842, + "grad_norm": 0.5703125, + "learning_rate": 0.00010526315789473685, + "loss": 3.5596, + "step": 10000 + }, + { + "epoch": 0.5263684210526316, + "grad_norm": 0.458984375, + "learning_rate": 0.00010525146198830408, + "loss": 3.5958, + "step": 10001 + }, + { + "epoch": 0.5264210526315789, + "grad_norm": 0.5390625, + "learning_rate": 0.00010523976608187136, + "loss": 3.6418, + "step": 10002 + }, + { + "epoch": 0.5264736842105263, + "grad_norm": 0.470703125, + "learning_rate": 0.00010522807017543861, + "loss": 3.5544, + "step": 10003 + }, + { + "epoch": 0.5265263157894737, + "grad_norm": 0.51171875, + "learning_rate": 0.00010521637426900586, + "loss": 3.1865, + "step": 10004 + }, + { + "epoch": 0.526578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00010520467836257311, + "loss": 3.1348, + "step": 10005 + }, + { + "epoch": 0.5266315789473685, + "grad_norm": 0.44921875, + "learning_rate": 0.00010519298245614036, + "loss": 3.4446, + "step": 10006 + }, + { + "epoch": 0.5266842105263158, + "grad_norm": 0.470703125, + "learning_rate": 0.00010518128654970761, + "loss": 3.0215, + "step": 10007 + }, + { + "epoch": 0.5267368421052632, + "grad_norm": 0.55859375, + "learning_rate": 0.00010516959064327484, + "loss": 3.8601, + "step": 10008 + }, + { + "epoch": 0.5267894736842105, + "grad_norm": 0.4921875, + "learning_rate": 0.0001051578947368421, + "loss": 3.1595, + "step": 10009 + }, + { + "epoch": 0.5268421052631579, + "grad_norm": 0.486328125, + "learning_rate": 0.00010514619883040937, + "loss": 3.5139, + "step": 10010 + }, + { + "epoch": 0.5268947368421053, + "grad_norm": 0.478515625, + "learning_rate": 0.00010513450292397662, + "loss": 3.2773, + "step": 10011 + }, + { + "epoch": 0.5269473684210526, + "grad_norm": 0.45703125, + "learning_rate": 0.00010512280701754387, + "loss": 3.4927, + "step": 10012 + }, + { + "epoch": 0.527, + "grad_norm": 0.48046875, + "learning_rate": 0.00010511111111111112, + "loss": 2.7884, + "step": 10013 + }, + { + "epoch": 0.5270526315789473, + "grad_norm": 0.55859375, + "learning_rate": 0.00010509941520467837, + "loss": 3.5455, + "step": 10014 + }, + { + "epoch": 0.5271052631578947, + "grad_norm": 0.466796875, + "learning_rate": 0.00010508771929824562, + "loss": 3.3951, + "step": 10015 + }, + { + "epoch": 0.5271578947368422, + "grad_norm": 0.458984375, + "learning_rate": 0.00010507602339181285, + "loss": 2.9866, + "step": 10016 + }, + { + "epoch": 0.5272105263157895, + "grad_norm": 0.466796875, + "learning_rate": 0.00010506432748538013, + "loss": 3.6061, + "step": 10017 + }, + { + "epoch": 0.5272631578947369, + "grad_norm": 0.51953125, + "learning_rate": 0.00010505263157894738, + "loss": 3.4579, + "step": 10018 + }, + { + "epoch": 0.5273157894736842, + "grad_norm": 0.55859375, + "learning_rate": 0.00010504093567251463, + "loss": 3.4586, + "step": 10019 + }, + { + "epoch": 0.5273684210526316, + "grad_norm": 0.474609375, + "learning_rate": 0.00010502923976608188, + "loss": 3.341, + "step": 10020 + }, + { + "epoch": 0.5274210526315789, + "grad_norm": 0.4765625, + "learning_rate": 0.00010501754385964913, + "loss": 3.9236, + "step": 10021 + }, + { + "epoch": 0.5274736842105263, + "grad_norm": 0.515625, + "learning_rate": 0.00010500584795321638, + "loss": 3.4846, + "step": 10022 + }, + { + "epoch": 0.5275263157894737, + "grad_norm": 0.50390625, + "learning_rate": 0.00010499415204678363, + "loss": 3.6708, + "step": 10023 + }, + { + "epoch": 0.527578947368421, + "grad_norm": 0.5625, + "learning_rate": 0.00010498245614035089, + "loss": 3.8115, + "step": 10024 + }, + { + "epoch": 0.5276315789473685, + "grad_norm": 0.609375, + "learning_rate": 0.00010497076023391814, + "loss": 3.4912, + "step": 10025 + }, + { + "epoch": 0.5276842105263158, + "grad_norm": 0.56640625, + "learning_rate": 0.00010495906432748539, + "loss": 3.6371, + "step": 10026 + }, + { + "epoch": 0.5277368421052632, + "grad_norm": 0.55078125, + "learning_rate": 0.00010494736842105264, + "loss": 3.7753, + "step": 10027 + }, + { + "epoch": 0.5277894736842105, + "grad_norm": 0.56640625, + "learning_rate": 0.00010493567251461989, + "loss": 3.3273, + "step": 10028 + }, + { + "epoch": 0.5278421052631579, + "grad_norm": 0.48046875, + "learning_rate": 0.00010492397660818714, + "loss": 3.3676, + "step": 10029 + }, + { + "epoch": 0.5278947368421053, + "grad_norm": 0.46484375, + "learning_rate": 0.00010491228070175439, + "loss": 3.7051, + "step": 10030 + }, + { + "epoch": 0.5279473684210526, + "grad_norm": 0.5, + "learning_rate": 0.00010490058479532163, + "loss": 3.5969, + "step": 10031 + }, + { + "epoch": 0.528, + "grad_norm": 0.498046875, + "learning_rate": 0.0001048888888888889, + "loss": 3.9522, + "step": 10032 + }, + { + "epoch": 0.5280526315789473, + "grad_norm": 0.5234375, + "learning_rate": 0.00010487719298245615, + "loss": 3.6754, + "step": 10033 + }, + { + "epoch": 0.5281052631578947, + "grad_norm": 0.494140625, + "learning_rate": 0.0001048654970760234, + "loss": 3.7699, + "step": 10034 + }, + { + "epoch": 0.5281578947368422, + "grad_norm": 0.486328125, + "learning_rate": 0.00010485380116959065, + "loss": 3.663, + "step": 10035 + }, + { + "epoch": 0.5282105263157895, + "grad_norm": 0.4921875, + "learning_rate": 0.0001048421052631579, + "loss": 3.247, + "step": 10036 + }, + { + "epoch": 0.5282631578947369, + "grad_norm": 0.490234375, + "learning_rate": 0.00010483040935672515, + "loss": 3.494, + "step": 10037 + }, + { + "epoch": 0.5283157894736842, + "grad_norm": 0.47265625, + "learning_rate": 0.0001048187134502924, + "loss": 2.9592, + "step": 10038 + }, + { + "epoch": 0.5283684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00010480701754385966, + "loss": 3.6362, + "step": 10039 + }, + { + "epoch": 0.5284210526315789, + "grad_norm": 0.5078125, + "learning_rate": 0.00010479532163742691, + "loss": 3.021, + "step": 10040 + }, + { + "epoch": 0.5284736842105263, + "grad_norm": 0.4609375, + "learning_rate": 0.00010478362573099416, + "loss": 3.286, + "step": 10041 + }, + { + "epoch": 0.5285263157894737, + "grad_norm": 0.447265625, + "learning_rate": 0.00010477192982456141, + "loss": 3.4955, + "step": 10042 + }, + { + "epoch": 0.528578947368421, + "grad_norm": 0.453125, + "learning_rate": 0.00010476023391812866, + "loss": 3.2947, + "step": 10043 + }, + { + "epoch": 0.5286315789473685, + "grad_norm": 0.515625, + "learning_rate": 0.00010474853801169591, + "loss": 3.6204, + "step": 10044 + }, + { + "epoch": 0.5286842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00010473684210526316, + "loss": 3.4118, + "step": 10045 + }, + { + "epoch": 0.5287368421052632, + "grad_norm": 0.478515625, + "learning_rate": 0.00010472514619883041, + "loss": 3.3994, + "step": 10046 + }, + { + "epoch": 0.5287894736842105, + "grad_norm": 0.474609375, + "learning_rate": 0.00010471345029239767, + "loss": 3.2121, + "step": 10047 + }, + { + "epoch": 0.5288421052631579, + "grad_norm": 0.490234375, + "learning_rate": 0.00010470175438596492, + "loss": 3.5745, + "step": 10048 + }, + { + "epoch": 0.5288947368421053, + "grad_norm": 0.482421875, + "learning_rate": 0.00010469005847953217, + "loss": 3.5889, + "step": 10049 + }, + { + "epoch": 0.5289473684210526, + "grad_norm": 0.484375, + "learning_rate": 0.00010467836257309942, + "loss": 3.933, + "step": 10050 + }, + { + "epoch": 0.529, + "grad_norm": 0.51171875, + "learning_rate": 0.00010466666666666667, + "loss": 3.2958, + "step": 10051 + }, + { + "epoch": 0.5290526315789473, + "grad_norm": 0.484375, + "learning_rate": 0.00010465497076023392, + "loss": 3.4311, + "step": 10052 + }, + { + "epoch": 0.5291052631578947, + "grad_norm": 0.48046875, + "learning_rate": 0.00010464327485380117, + "loss": 3.4293, + "step": 10053 + }, + { + "epoch": 0.5291578947368422, + "grad_norm": 0.5234375, + "learning_rate": 0.00010463157894736843, + "loss": 3.4186, + "step": 10054 + }, + { + "epoch": 0.5292105263157895, + "grad_norm": 0.703125, + "learning_rate": 0.00010461988304093568, + "loss": 3.2067, + "step": 10055 + }, + { + "epoch": 0.5292631578947369, + "grad_norm": 0.55078125, + "learning_rate": 0.00010460818713450293, + "loss": 3.5864, + "step": 10056 + }, + { + "epoch": 0.5293157894736842, + "grad_norm": 0.56640625, + "learning_rate": 0.00010459649122807018, + "loss": 3.1205, + "step": 10057 + }, + { + "epoch": 0.5293684210526316, + "grad_norm": 0.48828125, + "learning_rate": 0.00010458479532163743, + "loss": 3.3086, + "step": 10058 + }, + { + "epoch": 0.5294210526315789, + "grad_norm": 0.45703125, + "learning_rate": 0.00010457309941520468, + "loss": 3.8485, + "step": 10059 + }, + { + "epoch": 0.5294736842105263, + "grad_norm": 0.5, + "learning_rate": 0.00010456140350877193, + "loss": 3.1951, + "step": 10060 + }, + { + "epoch": 0.5295263157894737, + "grad_norm": 0.44921875, + "learning_rate": 0.00010454970760233918, + "loss": 2.9723, + "step": 10061 + }, + { + "epoch": 0.529578947368421, + "grad_norm": 0.458984375, + "learning_rate": 0.00010453801169590644, + "loss": 3.4508, + "step": 10062 + }, + { + "epoch": 0.5296315789473685, + "grad_norm": 0.6953125, + "learning_rate": 0.0001045263157894737, + "loss": 3.6797, + "step": 10063 + }, + { + "epoch": 0.5296842105263158, + "grad_norm": 0.49609375, + "learning_rate": 0.00010451461988304094, + "loss": 3.5145, + "step": 10064 + }, + { + "epoch": 0.5297368421052632, + "grad_norm": 0.70703125, + "learning_rate": 0.0001045029239766082, + "loss": 3.2253, + "step": 10065 + }, + { + "epoch": 0.5297894736842105, + "grad_norm": 0.466796875, + "learning_rate": 0.00010449122807017544, + "loss": 3.8632, + "step": 10066 + }, + { + "epoch": 0.5298421052631579, + "grad_norm": 0.60546875, + "learning_rate": 0.00010447953216374269, + "loss": 3.1684, + "step": 10067 + }, + { + "epoch": 0.5298947368421053, + "grad_norm": 0.482421875, + "learning_rate": 0.00010446783625730994, + "loss": 2.9514, + "step": 10068 + }, + { + "epoch": 0.5299473684210526, + "grad_norm": 0.498046875, + "learning_rate": 0.0001044561403508772, + "loss": 3.3301, + "step": 10069 + }, + { + "epoch": 0.53, + "grad_norm": 0.462890625, + "learning_rate": 0.00010444444444444445, + "loss": 3.2297, + "step": 10070 + }, + { + "epoch": 0.5300526315789473, + "grad_norm": 0.48828125, + "learning_rate": 0.0001044327485380117, + "loss": 3.7653, + "step": 10071 + }, + { + "epoch": 0.5301052631578947, + "grad_norm": 0.484375, + "learning_rate": 0.00010442105263157895, + "loss": 2.9212, + "step": 10072 + }, + { + "epoch": 0.530157894736842, + "grad_norm": 0.51171875, + "learning_rate": 0.0001044093567251462, + "loss": 2.9688, + "step": 10073 + }, + { + "epoch": 0.5302105263157895, + "grad_norm": 0.49609375, + "learning_rate": 0.00010439766081871345, + "loss": 3.233, + "step": 10074 + }, + { + "epoch": 0.5302631578947369, + "grad_norm": 0.478515625, + "learning_rate": 0.0001043859649122807, + "loss": 3.2884, + "step": 10075 + }, + { + "epoch": 0.5303157894736842, + "grad_norm": 0.52734375, + "learning_rate": 0.00010437426900584797, + "loss": 3.5626, + "step": 10076 + }, + { + "epoch": 0.5303684210526316, + "grad_norm": 0.50390625, + "learning_rate": 0.00010436257309941522, + "loss": 3.524, + "step": 10077 + }, + { + "epoch": 0.5304210526315789, + "grad_norm": 0.6796875, + "learning_rate": 0.00010435087719298247, + "loss": 3.2646, + "step": 10078 + }, + { + "epoch": 0.5304736842105263, + "grad_norm": 0.55078125, + "learning_rate": 0.00010433918128654971, + "loss": 3.475, + "step": 10079 + }, + { + "epoch": 0.5305263157894737, + "grad_norm": 0.54296875, + "learning_rate": 0.00010432748538011696, + "loss": 3.1742, + "step": 10080 + }, + { + "epoch": 0.530578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.00010431578947368421, + "loss": 3.3419, + "step": 10081 + }, + { + "epoch": 0.5306315789473685, + "grad_norm": 0.5390625, + "learning_rate": 0.00010430409356725146, + "loss": 2.9232, + "step": 10082 + }, + { + "epoch": 0.5306842105263158, + "grad_norm": 0.498046875, + "learning_rate": 0.00010429239766081871, + "loss": 3.4456, + "step": 10083 + }, + { + "epoch": 0.5307368421052632, + "grad_norm": 0.4765625, + "learning_rate": 0.00010428070175438598, + "loss": 3.6987, + "step": 10084 + }, + { + "epoch": 0.5307894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.00010426900584795323, + "loss": 3.3204, + "step": 10085 + }, + { + "epoch": 0.5308421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00010425730994152048, + "loss": 3.5459, + "step": 10086 + }, + { + "epoch": 0.5308947368421053, + "grad_norm": 0.50390625, + "learning_rate": 0.00010424561403508773, + "loss": 3.5149, + "step": 10087 + }, + { + "epoch": 0.5309473684210526, + "grad_norm": 0.53125, + "learning_rate": 0.00010423391812865497, + "loss": 3.3845, + "step": 10088 + }, + { + "epoch": 0.531, + "grad_norm": 0.515625, + "learning_rate": 0.00010422222222222222, + "loss": 3.3301, + "step": 10089 + }, + { + "epoch": 0.5310526315789473, + "grad_norm": 0.66015625, + "learning_rate": 0.00010421052631578947, + "loss": 3.3377, + "step": 10090 + }, + { + "epoch": 0.5311052631578947, + "grad_norm": 0.46875, + "learning_rate": 0.00010419883040935674, + "loss": 3.7142, + "step": 10091 + }, + { + "epoch": 0.531157894736842, + "grad_norm": 0.494140625, + "learning_rate": 0.00010418713450292399, + "loss": 3.1038, + "step": 10092 + }, + { + "epoch": 0.5312105263157895, + "grad_norm": 0.4609375, + "learning_rate": 0.00010417543859649124, + "loss": 3.8903, + "step": 10093 + }, + { + "epoch": 0.5312631578947369, + "grad_norm": 0.498046875, + "learning_rate": 0.00010416374269005849, + "loss": 3.2835, + "step": 10094 + }, + { + "epoch": 0.5313157894736842, + "grad_norm": 0.4765625, + "learning_rate": 0.00010415204678362574, + "loss": 3.2835, + "step": 10095 + }, + { + "epoch": 0.5313684210526316, + "grad_norm": 0.51171875, + "learning_rate": 0.00010414035087719298, + "loss": 3.002, + "step": 10096 + }, + { + "epoch": 0.5314210526315789, + "grad_norm": 0.48828125, + "learning_rate": 0.00010412865497076023, + "loss": 3.2694, + "step": 10097 + }, + { + "epoch": 0.5314736842105263, + "grad_norm": 0.455078125, + "learning_rate": 0.00010411695906432748, + "loss": 2.6322, + "step": 10098 + }, + { + "epoch": 0.5315263157894737, + "grad_norm": 0.4921875, + "learning_rate": 0.00010410526315789475, + "loss": 3.5409, + "step": 10099 + }, + { + "epoch": 0.531578947368421, + "grad_norm": 0.46875, + "learning_rate": 0.000104093567251462, + "loss": 3.6281, + "step": 10100 + }, + { + "epoch": 0.5316315789473685, + "grad_norm": 0.490234375, + "learning_rate": 0.00010408187134502925, + "loss": 3.557, + "step": 10101 + }, + { + "epoch": 0.5316842105263158, + "grad_norm": 0.515625, + "learning_rate": 0.0001040701754385965, + "loss": 3.6643, + "step": 10102 + }, + { + "epoch": 0.5317368421052632, + "grad_norm": 0.515625, + "learning_rate": 0.00010405847953216375, + "loss": 3.5068, + "step": 10103 + }, + { + "epoch": 0.5317894736842105, + "grad_norm": 0.478515625, + "learning_rate": 0.000104046783625731, + "loss": 3.6154, + "step": 10104 + }, + { + "epoch": 0.5318421052631579, + "grad_norm": 0.462890625, + "learning_rate": 0.00010403508771929824, + "loss": 3.626, + "step": 10105 + }, + { + "epoch": 0.5318947368421053, + "grad_norm": 0.546875, + "learning_rate": 0.00010402339181286551, + "loss": 3.4648, + "step": 10106 + }, + { + "epoch": 0.5319473684210526, + "grad_norm": 0.50390625, + "learning_rate": 0.00010401169590643276, + "loss": 3.1913, + "step": 10107 + }, + { + "epoch": 0.532, + "grad_norm": 0.51171875, + "learning_rate": 0.00010400000000000001, + "loss": 3.3979, + "step": 10108 + }, + { + "epoch": 0.5320526315789473, + "grad_norm": 0.45703125, + "learning_rate": 0.00010398830409356726, + "loss": 3.3905, + "step": 10109 + }, + { + "epoch": 0.5321052631578947, + "grad_norm": 0.478515625, + "learning_rate": 0.0001039766081871345, + "loss": 3.0441, + "step": 10110 + }, + { + "epoch": 0.532157894736842, + "grad_norm": 0.4765625, + "learning_rate": 0.00010396491228070176, + "loss": 3.3376, + "step": 10111 + }, + { + "epoch": 0.5322105263157895, + "grad_norm": 0.478515625, + "learning_rate": 0.000103953216374269, + "loss": 3.363, + "step": 10112 + }, + { + "epoch": 0.5322631578947369, + "grad_norm": 0.451171875, + "learning_rate": 0.00010394152046783627, + "loss": 3.3619, + "step": 10113 + }, + { + "epoch": 0.5323157894736842, + "grad_norm": 0.478515625, + "learning_rate": 0.00010392982456140352, + "loss": 3.8089, + "step": 10114 + }, + { + "epoch": 0.5323684210526316, + "grad_norm": 0.4921875, + "learning_rate": 0.00010391812865497077, + "loss": 3.4698, + "step": 10115 + }, + { + "epoch": 0.5324210526315789, + "grad_norm": 0.470703125, + "learning_rate": 0.00010390643274853802, + "loss": 3.7667, + "step": 10116 + }, + { + "epoch": 0.5324736842105263, + "grad_norm": 0.458984375, + "learning_rate": 0.00010389473684210527, + "loss": 3.2584, + "step": 10117 + }, + { + "epoch": 0.5325263157894737, + "grad_norm": 0.494140625, + "learning_rate": 0.00010388304093567252, + "loss": 3.3644, + "step": 10118 + }, + { + "epoch": 0.532578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.00010387134502923977, + "loss": 2.6931, + "step": 10119 + }, + { + "epoch": 0.5326315789473685, + "grad_norm": 0.49609375, + "learning_rate": 0.00010385964912280702, + "loss": 3.5494, + "step": 10120 + }, + { + "epoch": 0.5326842105263158, + "grad_norm": 0.462890625, + "learning_rate": 0.00010384795321637428, + "loss": 3.4741, + "step": 10121 + }, + { + "epoch": 0.5327368421052632, + "grad_norm": 0.490234375, + "learning_rate": 0.00010383625730994153, + "loss": 3.5807, + "step": 10122 + }, + { + "epoch": 0.5327894736842105, + "grad_norm": 0.5625, + "learning_rate": 0.00010382456140350878, + "loss": 3.4103, + "step": 10123 + }, + { + "epoch": 0.5328421052631579, + "grad_norm": 0.54296875, + "learning_rate": 0.00010381286549707603, + "loss": 3.3275, + "step": 10124 + }, + { + "epoch": 0.5328947368421053, + "grad_norm": 0.63671875, + "learning_rate": 0.00010380116959064328, + "loss": 3.2388, + "step": 10125 + }, + { + "epoch": 0.5329473684210526, + "grad_norm": 0.4765625, + "learning_rate": 0.00010378947368421053, + "loss": 3.7316, + "step": 10126 + }, + { + "epoch": 0.533, + "grad_norm": 0.5078125, + "learning_rate": 0.00010377777777777778, + "loss": 3.4113, + "step": 10127 + }, + { + "epoch": 0.5330526315789473, + "grad_norm": 0.5390625, + "learning_rate": 0.00010376608187134504, + "loss": 3.335, + "step": 10128 + }, + { + "epoch": 0.5331052631578947, + "grad_norm": 0.458984375, + "learning_rate": 0.00010375438596491229, + "loss": 3.8805, + "step": 10129 + }, + { + "epoch": 0.533157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.00010374269005847954, + "loss": 3.5329, + "step": 10130 + }, + { + "epoch": 0.5332105263157895, + "grad_norm": 0.482421875, + "learning_rate": 0.00010373099415204679, + "loss": 3.3473, + "step": 10131 + }, + { + "epoch": 0.5332631578947369, + "grad_norm": 0.5078125, + "learning_rate": 0.00010371929824561404, + "loss": 2.8769, + "step": 10132 + }, + { + "epoch": 0.5333157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.00010370760233918129, + "loss": 2.9598, + "step": 10133 + }, + { + "epoch": 0.5333684210526316, + "grad_norm": 0.5546875, + "learning_rate": 0.00010369590643274854, + "loss": 3.8933, + "step": 10134 + }, + { + "epoch": 0.5334210526315789, + "grad_norm": 0.498046875, + "learning_rate": 0.00010368421052631579, + "loss": 3.1398, + "step": 10135 + }, + { + "epoch": 0.5334736842105263, + "grad_norm": 0.482421875, + "learning_rate": 0.00010367251461988305, + "loss": 3.2251, + "step": 10136 + }, + { + "epoch": 0.5335263157894737, + "grad_norm": 0.4765625, + "learning_rate": 0.0001036608187134503, + "loss": 3.3805, + "step": 10137 + }, + { + "epoch": 0.533578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.00010364912280701755, + "loss": 3.273, + "step": 10138 + }, + { + "epoch": 0.5336315789473685, + "grad_norm": 0.46484375, + "learning_rate": 0.0001036374269005848, + "loss": 3.767, + "step": 10139 + }, + { + "epoch": 0.5336842105263158, + "grad_norm": 0.609375, + "learning_rate": 0.00010362573099415205, + "loss": 3.3639, + "step": 10140 + }, + { + "epoch": 0.5337368421052632, + "grad_norm": 0.515625, + "learning_rate": 0.0001036140350877193, + "loss": 3.2818, + "step": 10141 + }, + { + "epoch": 0.5337894736842105, + "grad_norm": 0.52734375, + "learning_rate": 0.00010360233918128655, + "loss": 3.5677, + "step": 10142 + }, + { + "epoch": 0.5338421052631579, + "grad_norm": 0.466796875, + "learning_rate": 0.00010359064327485381, + "loss": 3.2571, + "step": 10143 + }, + { + "epoch": 0.5338947368421053, + "grad_norm": 0.49609375, + "learning_rate": 0.00010357894736842106, + "loss": 3.086, + "step": 10144 + }, + { + "epoch": 0.5339473684210526, + "grad_norm": 0.453125, + "learning_rate": 0.00010356725146198831, + "loss": 3.3023, + "step": 10145 + }, + { + "epoch": 0.534, + "grad_norm": 0.478515625, + "learning_rate": 0.00010355555555555556, + "loss": 3.8041, + "step": 10146 + }, + { + "epoch": 0.5340526315789473, + "grad_norm": 0.4921875, + "learning_rate": 0.00010354385964912281, + "loss": 3.3942, + "step": 10147 + }, + { + "epoch": 0.5341052631578947, + "grad_norm": 0.462890625, + "learning_rate": 0.00010353216374269006, + "loss": 3.3134, + "step": 10148 + }, + { + "epoch": 0.534157894736842, + "grad_norm": 0.46484375, + "learning_rate": 0.00010352046783625731, + "loss": 3.5196, + "step": 10149 + }, + { + "epoch": 0.5342105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.00010350877192982457, + "loss": 3.2768, + "step": 10150 + }, + { + "epoch": 0.5342631578947369, + "grad_norm": 0.474609375, + "learning_rate": 0.00010349707602339182, + "loss": 3.4676, + "step": 10151 + }, + { + "epoch": 0.5343157894736842, + "grad_norm": 0.46875, + "learning_rate": 0.00010348538011695907, + "loss": 3.5167, + "step": 10152 + }, + { + "epoch": 0.5343684210526316, + "grad_norm": 0.4921875, + "learning_rate": 0.00010347368421052632, + "loss": 3.459, + "step": 10153 + }, + { + "epoch": 0.5344210526315789, + "grad_norm": 0.5, + "learning_rate": 0.00010346198830409357, + "loss": 3.524, + "step": 10154 + }, + { + "epoch": 0.5344736842105263, + "grad_norm": 0.54296875, + "learning_rate": 0.00010345029239766082, + "loss": 3.1662, + "step": 10155 + }, + { + "epoch": 0.5345263157894737, + "grad_norm": 0.4921875, + "learning_rate": 0.00010343859649122807, + "loss": 3.7097, + "step": 10156 + }, + { + "epoch": 0.534578947368421, + "grad_norm": 0.47265625, + "learning_rate": 0.00010342690058479532, + "loss": 3.4286, + "step": 10157 + }, + { + "epoch": 0.5346315789473685, + "grad_norm": 0.484375, + "learning_rate": 0.00010341520467836258, + "loss": 3.1784, + "step": 10158 + }, + { + "epoch": 0.5346842105263158, + "grad_norm": 0.478515625, + "learning_rate": 0.00010340350877192983, + "loss": 3.7533, + "step": 10159 + }, + { + "epoch": 0.5347368421052632, + "grad_norm": 0.515625, + "learning_rate": 0.00010339181286549708, + "loss": 3.6805, + "step": 10160 + }, + { + "epoch": 0.5347894736842105, + "grad_norm": 0.53125, + "learning_rate": 0.00010338011695906433, + "loss": 3.381, + "step": 10161 + }, + { + "epoch": 0.5348421052631579, + "grad_norm": 0.51171875, + "learning_rate": 0.00010336842105263158, + "loss": 3.6659, + "step": 10162 + }, + { + "epoch": 0.5348947368421053, + "grad_norm": 0.474609375, + "learning_rate": 0.00010335672514619883, + "loss": 3.5275, + "step": 10163 + }, + { + "epoch": 0.5349473684210526, + "grad_norm": 0.578125, + "learning_rate": 0.00010334502923976608, + "loss": 3.0879, + "step": 10164 + }, + { + "epoch": 0.535, + "grad_norm": 0.53125, + "learning_rate": 0.00010333333333333334, + "loss": 3.7267, + "step": 10165 + }, + { + "epoch": 0.5350526315789473, + "grad_norm": 0.498046875, + "learning_rate": 0.00010332163742690059, + "loss": 2.9752, + "step": 10166 + }, + { + "epoch": 0.5351052631578948, + "grad_norm": 0.490234375, + "learning_rate": 0.00010330994152046784, + "loss": 3.2489, + "step": 10167 + }, + { + "epoch": 0.535157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00010329824561403509, + "loss": 3.7688, + "step": 10168 + }, + { + "epoch": 0.5352105263157895, + "grad_norm": 0.51171875, + "learning_rate": 0.00010328654970760234, + "loss": 3.1378, + "step": 10169 + }, + { + "epoch": 0.5352631578947369, + "grad_norm": 0.52734375, + "learning_rate": 0.00010327485380116959, + "loss": 3.6101, + "step": 10170 + }, + { + "epoch": 0.5353157894736842, + "grad_norm": 0.494140625, + "learning_rate": 0.00010326315789473684, + "loss": 4.0029, + "step": 10171 + }, + { + "epoch": 0.5353684210526316, + "grad_norm": 0.5625, + "learning_rate": 0.00010325146198830409, + "loss": 3.4491, + "step": 10172 + }, + { + "epoch": 0.5354210526315789, + "grad_norm": 0.486328125, + "learning_rate": 0.00010323976608187135, + "loss": 3.6054, + "step": 10173 + }, + { + "epoch": 0.5354736842105263, + "grad_norm": 0.478515625, + "learning_rate": 0.0001032280701754386, + "loss": 3.5973, + "step": 10174 + }, + { + "epoch": 0.5355263157894737, + "grad_norm": 0.5625, + "learning_rate": 0.00010321637426900585, + "loss": 3.328, + "step": 10175 + }, + { + "epoch": 0.535578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.0001032046783625731, + "loss": 2.9738, + "step": 10176 + }, + { + "epoch": 0.5356315789473685, + "grad_norm": 0.4765625, + "learning_rate": 0.00010319298245614035, + "loss": 2.9122, + "step": 10177 + }, + { + "epoch": 0.5356842105263158, + "grad_norm": 0.48828125, + "learning_rate": 0.0001031812865497076, + "loss": 3.3588, + "step": 10178 + }, + { + "epoch": 0.5357368421052632, + "grad_norm": 0.490234375, + "learning_rate": 0.00010316959064327485, + "loss": 3.7269, + "step": 10179 + }, + { + "epoch": 0.5357894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.00010315789473684211, + "loss": 3.7491, + "step": 10180 + }, + { + "epoch": 0.5358421052631579, + "grad_norm": 0.5078125, + "learning_rate": 0.00010314619883040936, + "loss": 3.3386, + "step": 10181 + }, + { + "epoch": 0.5358947368421053, + "grad_norm": 0.5, + "learning_rate": 0.00010313450292397661, + "loss": 3.2527, + "step": 10182 + }, + { + "epoch": 0.5359473684210526, + "grad_norm": 0.470703125, + "learning_rate": 0.00010312280701754386, + "loss": 3.641, + "step": 10183 + }, + { + "epoch": 0.536, + "grad_norm": 0.5234375, + "learning_rate": 0.00010311111111111111, + "loss": 3.4233, + "step": 10184 + }, + { + "epoch": 0.5360526315789473, + "grad_norm": 0.51953125, + "learning_rate": 0.00010309941520467836, + "loss": 3.5768, + "step": 10185 + }, + { + "epoch": 0.5361052631578948, + "grad_norm": 0.498046875, + "learning_rate": 0.00010308771929824561, + "loss": 3.5506, + "step": 10186 + }, + { + "epoch": 0.5361578947368421, + "grad_norm": 0.57421875, + "learning_rate": 0.00010307602339181289, + "loss": 3.3581, + "step": 10187 + }, + { + "epoch": 0.5362105263157895, + "grad_norm": 0.54296875, + "learning_rate": 0.00010306432748538012, + "loss": 3.2792, + "step": 10188 + }, + { + "epoch": 0.5362631578947369, + "grad_norm": 0.482421875, + "learning_rate": 0.00010305263157894737, + "loss": 3.2747, + "step": 10189 + }, + { + "epoch": 0.5363157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00010304093567251462, + "loss": 3.1193, + "step": 10190 + }, + { + "epoch": 0.5363684210526316, + "grad_norm": 0.46875, + "learning_rate": 0.00010302923976608187, + "loss": 3.2845, + "step": 10191 + }, + { + "epoch": 0.5364210526315789, + "grad_norm": 0.494140625, + "learning_rate": 0.00010301754385964912, + "loss": 3.5162, + "step": 10192 + }, + { + "epoch": 0.5364736842105263, + "grad_norm": 0.458984375, + "learning_rate": 0.00010300584795321637, + "loss": 3.1422, + "step": 10193 + }, + { + "epoch": 0.5365263157894736, + "grad_norm": 0.5078125, + "learning_rate": 0.00010299415204678362, + "loss": 3.337, + "step": 10194 + }, + { + "epoch": 0.536578947368421, + "grad_norm": 0.4765625, + "learning_rate": 0.0001029824561403509, + "loss": 3.5383, + "step": 10195 + }, + { + "epoch": 0.5366315789473685, + "grad_norm": 0.515625, + "learning_rate": 0.00010297076023391813, + "loss": 3.5999, + "step": 10196 + }, + { + "epoch": 0.5366842105263158, + "grad_norm": 0.52734375, + "learning_rate": 0.00010295906432748538, + "loss": 3.2668, + "step": 10197 + }, + { + "epoch": 0.5367368421052632, + "grad_norm": 0.55078125, + "learning_rate": 0.00010294736842105263, + "loss": 3.2284, + "step": 10198 + }, + { + "epoch": 0.5367894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.00010293567251461988, + "loss": 3.2316, + "step": 10199 + }, + { + "epoch": 0.5368421052631579, + "grad_norm": 0.53125, + "learning_rate": 0.00010292397660818713, + "loss": 3.5158, + "step": 10200 + }, + { + "epoch": 0.5368947368421053, + "grad_norm": 0.466796875, + "learning_rate": 0.00010291228070175438, + "loss": 3.6084, + "step": 10201 + }, + { + "epoch": 0.5369473684210526, + "grad_norm": 0.515625, + "learning_rate": 0.00010290058479532166, + "loss": 3.5542, + "step": 10202 + }, + { + "epoch": 0.537, + "grad_norm": 0.470703125, + "learning_rate": 0.0001028888888888889, + "loss": 3.5328, + "step": 10203 + }, + { + "epoch": 0.5370526315789473, + "grad_norm": 0.5078125, + "learning_rate": 0.00010287719298245614, + "loss": 3.4807, + "step": 10204 + }, + { + "epoch": 0.5371052631578948, + "grad_norm": 0.498046875, + "learning_rate": 0.0001028654970760234, + "loss": 3.0167, + "step": 10205 + }, + { + "epoch": 0.5371578947368421, + "grad_norm": 0.5, + "learning_rate": 0.00010285380116959064, + "loss": 3.6527, + "step": 10206 + }, + { + "epoch": 0.5372105263157895, + "grad_norm": 0.4765625, + "learning_rate": 0.00010284210526315789, + "loss": 3.8352, + "step": 10207 + }, + { + "epoch": 0.5372631578947369, + "grad_norm": 0.48828125, + "learning_rate": 0.00010283040935672514, + "loss": 3.439, + "step": 10208 + }, + { + "epoch": 0.5373157894736842, + "grad_norm": 0.7109375, + "learning_rate": 0.00010281871345029239, + "loss": 2.792, + "step": 10209 + }, + { + "epoch": 0.5373684210526316, + "grad_norm": 0.494140625, + "learning_rate": 0.00010280701754385967, + "loss": 3.3164, + "step": 10210 + }, + { + "epoch": 0.5374210526315789, + "grad_norm": 0.515625, + "learning_rate": 0.0001027953216374269, + "loss": 3.7965, + "step": 10211 + }, + { + "epoch": 0.5374736842105263, + "grad_norm": 0.447265625, + "learning_rate": 0.00010278362573099415, + "loss": 3.1573, + "step": 10212 + }, + { + "epoch": 0.5375263157894736, + "grad_norm": 0.46875, + "learning_rate": 0.0001027719298245614, + "loss": 3.7753, + "step": 10213 + }, + { + "epoch": 0.537578947368421, + "grad_norm": 0.44921875, + "learning_rate": 0.00010276023391812865, + "loss": 3.7131, + "step": 10214 + }, + { + "epoch": 0.5376315789473685, + "grad_norm": 0.515625, + "learning_rate": 0.0001027485380116959, + "loss": 3.1247, + "step": 10215 + }, + { + "epoch": 0.5376842105263158, + "grad_norm": 0.53125, + "learning_rate": 0.00010273684210526315, + "loss": 3.2275, + "step": 10216 + }, + { + "epoch": 0.5377368421052632, + "grad_norm": 0.48828125, + "learning_rate": 0.00010272514619883043, + "loss": 3.3391, + "step": 10217 + }, + { + "epoch": 0.5377894736842105, + "grad_norm": 0.466796875, + "learning_rate": 0.00010271345029239768, + "loss": 3.6368, + "step": 10218 + }, + { + "epoch": 0.5378421052631579, + "grad_norm": 0.474609375, + "learning_rate": 0.00010270175438596492, + "loss": 3.3257, + "step": 10219 + }, + { + "epoch": 0.5378947368421053, + "grad_norm": 0.51953125, + "learning_rate": 0.00010269005847953216, + "loss": 3.1988, + "step": 10220 + }, + { + "epoch": 0.5379473684210526, + "grad_norm": 0.451171875, + "learning_rate": 0.00010267836257309941, + "loss": 3.1826, + "step": 10221 + }, + { + "epoch": 0.538, + "grad_norm": 0.451171875, + "learning_rate": 0.00010266666666666666, + "loss": 3.392, + "step": 10222 + }, + { + "epoch": 0.5380526315789473, + "grad_norm": 0.466796875, + "learning_rate": 0.00010265497076023391, + "loss": 3.4293, + "step": 10223 + }, + { + "epoch": 0.5381052631578948, + "grad_norm": 0.52734375, + "learning_rate": 0.00010264327485380116, + "loss": 3.2973, + "step": 10224 + }, + { + "epoch": 0.5381578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.00010263157894736844, + "loss": 3.4674, + "step": 10225 + }, + { + "epoch": 0.5382105263157895, + "grad_norm": 0.484375, + "learning_rate": 0.00010261988304093568, + "loss": 3.2003, + "step": 10226 + }, + { + "epoch": 0.5382631578947369, + "grad_norm": 0.494140625, + "learning_rate": 0.00010260818713450293, + "loss": 3.4087, + "step": 10227 + }, + { + "epoch": 0.5383157894736842, + "grad_norm": 0.482421875, + "learning_rate": 0.00010259649122807018, + "loss": 3.2887, + "step": 10228 + }, + { + "epoch": 0.5383684210526316, + "grad_norm": 0.458984375, + "learning_rate": 0.00010258479532163742, + "loss": 2.8735, + "step": 10229 + }, + { + "epoch": 0.5384210526315789, + "grad_norm": 0.45703125, + "learning_rate": 0.00010257309941520467, + "loss": 3.4502, + "step": 10230 + }, + { + "epoch": 0.5384736842105263, + "grad_norm": 0.546875, + "learning_rate": 0.00010256140350877192, + "loss": 3.3842, + "step": 10231 + }, + { + "epoch": 0.5385263157894736, + "grad_norm": 0.484375, + "learning_rate": 0.0001025497076023392, + "loss": 3.6853, + "step": 10232 + }, + { + "epoch": 0.538578947368421, + "grad_norm": 0.46484375, + "learning_rate": 0.00010253801169590645, + "loss": 3.3699, + "step": 10233 + }, + { + "epoch": 0.5386315789473685, + "grad_norm": 0.53125, + "learning_rate": 0.00010252631578947369, + "loss": 2.7525, + "step": 10234 + }, + { + "epoch": 0.5386842105263158, + "grad_norm": 0.55078125, + "learning_rate": 0.00010251461988304094, + "loss": 3.3797, + "step": 10235 + }, + { + "epoch": 0.5387368421052632, + "grad_norm": 0.474609375, + "learning_rate": 0.00010250292397660819, + "loss": 3.4387, + "step": 10236 + }, + { + "epoch": 0.5387894736842105, + "grad_norm": 0.49609375, + "learning_rate": 0.00010249122807017543, + "loss": 3.4852, + "step": 10237 + }, + { + "epoch": 0.5388421052631579, + "grad_norm": 0.46875, + "learning_rate": 0.00010247953216374268, + "loss": 3.4231, + "step": 10238 + }, + { + "epoch": 0.5388947368421053, + "grad_norm": 0.58203125, + "learning_rate": 0.00010246783625730996, + "loss": 3.4448, + "step": 10239 + }, + { + "epoch": 0.5389473684210526, + "grad_norm": 0.4609375, + "learning_rate": 0.00010245614035087721, + "loss": 3.091, + "step": 10240 + }, + { + "epoch": 0.539, + "grad_norm": 0.515625, + "learning_rate": 0.00010244444444444446, + "loss": 3.3949, + "step": 10241 + }, + { + "epoch": 0.5390526315789473, + "grad_norm": 0.47265625, + "learning_rate": 0.0001024327485380117, + "loss": 3.5135, + "step": 10242 + }, + { + "epoch": 0.5391052631578948, + "grad_norm": 0.54296875, + "learning_rate": 0.00010242105263157895, + "loss": 3.6429, + "step": 10243 + }, + { + "epoch": 0.5391578947368421, + "grad_norm": 0.5, + "learning_rate": 0.0001024093567251462, + "loss": 3.7169, + "step": 10244 + }, + { + "epoch": 0.5392105263157895, + "grad_norm": 0.62109375, + "learning_rate": 0.00010239766081871345, + "loss": 3.1383, + "step": 10245 + }, + { + "epoch": 0.5392631578947369, + "grad_norm": 0.46484375, + "learning_rate": 0.0001023859649122807, + "loss": 3.6168, + "step": 10246 + }, + { + "epoch": 0.5393157894736842, + "grad_norm": 0.470703125, + "learning_rate": 0.00010237426900584797, + "loss": 3.18, + "step": 10247 + }, + { + "epoch": 0.5393684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00010236257309941522, + "loss": 2.9655, + "step": 10248 + }, + { + "epoch": 0.5394210526315789, + "grad_norm": 0.4765625, + "learning_rate": 0.00010235087719298246, + "loss": 3.4554, + "step": 10249 + }, + { + "epoch": 0.5394736842105263, + "grad_norm": 0.458984375, + "learning_rate": 0.00010233918128654971, + "loss": 3.1337, + "step": 10250 + }, + { + "epoch": 0.5395263157894736, + "grad_norm": 0.50390625, + "learning_rate": 0.00010232748538011696, + "loss": 3.0629, + "step": 10251 + }, + { + "epoch": 0.539578947368421, + "grad_norm": 0.5078125, + "learning_rate": 0.0001023157894736842, + "loss": 3.3565, + "step": 10252 + }, + { + "epoch": 0.5396315789473685, + "grad_norm": 0.451171875, + "learning_rate": 0.00010230409356725146, + "loss": 2.8464, + "step": 10253 + }, + { + "epoch": 0.5396842105263158, + "grad_norm": 0.4765625, + "learning_rate": 0.00010229239766081873, + "loss": 3.6822, + "step": 10254 + }, + { + "epoch": 0.5397368421052632, + "grad_norm": 0.4921875, + "learning_rate": 0.00010228070175438598, + "loss": 3.209, + "step": 10255 + }, + { + "epoch": 0.5397894736842105, + "grad_norm": 0.5078125, + "learning_rate": 0.00010226900584795323, + "loss": 3.2828, + "step": 10256 + }, + { + "epoch": 0.5398421052631579, + "grad_norm": 0.47265625, + "learning_rate": 0.00010225730994152047, + "loss": 3.8122, + "step": 10257 + }, + { + "epoch": 0.5398947368421053, + "grad_norm": 0.51171875, + "learning_rate": 0.00010224561403508772, + "loss": 3.2411, + "step": 10258 + }, + { + "epoch": 0.5399473684210526, + "grad_norm": 0.494140625, + "learning_rate": 0.00010223391812865497, + "loss": 3.403, + "step": 10259 + }, + { + "epoch": 0.54, + "grad_norm": 0.470703125, + "learning_rate": 0.00010222222222222222, + "loss": 3.4977, + "step": 10260 + }, + { + "epoch": 0.5400526315789473, + "grad_norm": 0.5, + "learning_rate": 0.00010221052631578947, + "loss": 3.4154, + "step": 10261 + }, + { + "epoch": 0.5401052631578948, + "grad_norm": 0.515625, + "learning_rate": 0.00010219883040935674, + "loss": 3.7995, + "step": 10262 + }, + { + "epoch": 0.5401578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.00010218713450292399, + "loss": 3.5625, + "step": 10263 + }, + { + "epoch": 0.5402105263157895, + "grad_norm": 0.48828125, + "learning_rate": 0.00010217543859649124, + "loss": 3.5957, + "step": 10264 + }, + { + "epoch": 0.5402631578947369, + "grad_norm": 0.4765625, + "learning_rate": 0.00010216374269005848, + "loss": 3.1256, + "step": 10265 + }, + { + "epoch": 0.5403157894736842, + "grad_norm": 0.50390625, + "learning_rate": 0.00010215204678362573, + "loss": 3.8537, + "step": 10266 + }, + { + "epoch": 0.5403684210526316, + "grad_norm": 0.484375, + "learning_rate": 0.00010214035087719298, + "loss": 3.4101, + "step": 10267 + }, + { + "epoch": 0.5404210526315789, + "grad_norm": 0.4921875, + "learning_rate": 0.00010212865497076023, + "loss": 3.8412, + "step": 10268 + }, + { + "epoch": 0.5404736842105263, + "grad_norm": 0.48046875, + "learning_rate": 0.0001021169590643275, + "loss": 3.5056, + "step": 10269 + }, + { + "epoch": 0.5405263157894736, + "grad_norm": 0.5, + "learning_rate": 0.00010210526315789475, + "loss": 3.4448, + "step": 10270 + }, + { + "epoch": 0.540578947368421, + "grad_norm": 0.484375, + "learning_rate": 0.000102093567251462, + "loss": 3.4994, + "step": 10271 + }, + { + "epoch": 0.5406315789473685, + "grad_norm": 0.482421875, + "learning_rate": 0.00010208187134502924, + "loss": 3.3697, + "step": 10272 + }, + { + "epoch": 0.5406842105263158, + "grad_norm": 0.494140625, + "learning_rate": 0.00010207017543859649, + "loss": 3.4888, + "step": 10273 + }, + { + "epoch": 0.5407368421052632, + "grad_norm": 0.484375, + "learning_rate": 0.00010205847953216374, + "loss": 3.1116, + "step": 10274 + }, + { + "epoch": 0.5407894736842105, + "grad_norm": 0.494140625, + "learning_rate": 0.00010204678362573099, + "loss": 3.236, + "step": 10275 + }, + { + "epoch": 0.5408421052631579, + "grad_norm": 0.486328125, + "learning_rate": 0.00010203508771929826, + "loss": 3.4331, + "step": 10276 + }, + { + "epoch": 0.5408947368421053, + "grad_norm": 0.51171875, + "learning_rate": 0.00010202339181286551, + "loss": 3.3747, + "step": 10277 + }, + { + "epoch": 0.5409473684210526, + "grad_norm": 0.53125, + "learning_rate": 0.00010201169590643276, + "loss": 3.1526, + "step": 10278 + }, + { + "epoch": 0.541, + "grad_norm": 0.48046875, + "learning_rate": 0.00010200000000000001, + "loss": 3.3926, + "step": 10279 + }, + { + "epoch": 0.5410526315789473, + "grad_norm": 0.5390625, + "learning_rate": 0.00010198830409356725, + "loss": 3.5951, + "step": 10280 + }, + { + "epoch": 0.5411052631578948, + "grad_norm": 0.486328125, + "learning_rate": 0.0001019766081871345, + "loss": 3.2729, + "step": 10281 + }, + { + "epoch": 0.5411578947368421, + "grad_norm": 0.515625, + "learning_rate": 0.00010196491228070175, + "loss": 3.3532, + "step": 10282 + }, + { + "epoch": 0.5412105263157895, + "grad_norm": 0.56640625, + "learning_rate": 0.000101953216374269, + "loss": 3.0669, + "step": 10283 + }, + { + "epoch": 0.5412631578947369, + "grad_norm": 0.5, + "learning_rate": 0.00010194152046783627, + "loss": 3.4846, + "step": 10284 + }, + { + "epoch": 0.5413157894736842, + "grad_norm": 0.47265625, + "learning_rate": 0.00010192982456140352, + "loss": 2.8307, + "step": 10285 + }, + { + "epoch": 0.5413684210526316, + "grad_norm": 0.5234375, + "learning_rate": 0.00010191812865497077, + "loss": 3.4009, + "step": 10286 + }, + { + "epoch": 0.5414210526315789, + "grad_norm": 0.53125, + "learning_rate": 0.00010190643274853802, + "loss": 3.3457, + "step": 10287 + }, + { + "epoch": 0.5414736842105263, + "grad_norm": 0.470703125, + "learning_rate": 0.00010189473684210526, + "loss": 3.3479, + "step": 10288 + }, + { + "epoch": 0.5415263157894736, + "grad_norm": 0.478515625, + "learning_rate": 0.00010188304093567251, + "loss": 3.3602, + "step": 10289 + }, + { + "epoch": 0.541578947368421, + "grad_norm": 0.470703125, + "learning_rate": 0.00010187134502923976, + "loss": 3.7105, + "step": 10290 + }, + { + "epoch": 0.5416315789473685, + "grad_norm": 0.466796875, + "learning_rate": 0.00010185964912280703, + "loss": 3.3475, + "step": 10291 + }, + { + "epoch": 0.5416842105263158, + "grad_norm": 0.478515625, + "learning_rate": 0.00010184795321637428, + "loss": 2.9721, + "step": 10292 + }, + { + "epoch": 0.5417368421052632, + "grad_norm": 0.53125, + "learning_rate": 0.00010183625730994153, + "loss": 3.1615, + "step": 10293 + }, + { + "epoch": 0.5417894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.00010182456140350878, + "loss": 3.6869, + "step": 10294 + }, + { + "epoch": 0.5418421052631579, + "grad_norm": 0.45703125, + "learning_rate": 0.00010181286549707602, + "loss": 3.3649, + "step": 10295 + }, + { + "epoch": 0.5418947368421053, + "grad_norm": 0.8359375, + "learning_rate": 0.00010180116959064327, + "loss": 3.3558, + "step": 10296 + }, + { + "epoch": 0.5419473684210526, + "grad_norm": 0.46484375, + "learning_rate": 0.00010178947368421052, + "loss": 3.5535, + "step": 10297 + }, + { + "epoch": 0.542, + "grad_norm": 0.4453125, + "learning_rate": 0.00010177777777777777, + "loss": 3.1373, + "step": 10298 + }, + { + "epoch": 0.5420526315789473, + "grad_norm": 0.466796875, + "learning_rate": 0.00010176608187134505, + "loss": 3.6426, + "step": 10299 + }, + { + "epoch": 0.5421052631578948, + "grad_norm": 0.466796875, + "learning_rate": 0.0001017543859649123, + "loss": 3.3049, + "step": 10300 + }, + { + "epoch": 0.5421578947368421, + "grad_norm": 0.51953125, + "learning_rate": 0.00010174269005847954, + "loss": 3.6625, + "step": 10301 + }, + { + "epoch": 0.5422105263157895, + "grad_norm": 0.484375, + "learning_rate": 0.0001017309941520468, + "loss": 3.1119, + "step": 10302 + }, + { + "epoch": 0.5422631578947369, + "grad_norm": 0.4765625, + "learning_rate": 0.00010171929824561403, + "loss": 3.4754, + "step": 10303 + }, + { + "epoch": 0.5423157894736842, + "grad_norm": 0.466796875, + "learning_rate": 0.00010170760233918128, + "loss": 3.5995, + "step": 10304 + }, + { + "epoch": 0.5423684210526316, + "grad_norm": 0.48046875, + "learning_rate": 0.00010169590643274853, + "loss": 3.8203, + "step": 10305 + }, + { + "epoch": 0.5424210526315789, + "grad_norm": 0.53125, + "learning_rate": 0.0001016842105263158, + "loss": 3.4902, + "step": 10306 + }, + { + "epoch": 0.5424736842105263, + "grad_norm": 0.484375, + "learning_rate": 0.00010167251461988306, + "loss": 3.5314, + "step": 10307 + }, + { + "epoch": 0.5425263157894736, + "grad_norm": 0.4921875, + "learning_rate": 0.0001016608187134503, + "loss": 3.8479, + "step": 10308 + }, + { + "epoch": 0.542578947368421, + "grad_norm": 0.4921875, + "learning_rate": 0.00010164912280701755, + "loss": 3.1226, + "step": 10309 + }, + { + "epoch": 0.5426315789473685, + "grad_norm": 0.515625, + "learning_rate": 0.0001016374269005848, + "loss": 2.9884, + "step": 10310 + }, + { + "epoch": 0.5426842105263158, + "grad_norm": 0.5234375, + "learning_rate": 0.00010162573099415204, + "loss": 3.4387, + "step": 10311 + }, + { + "epoch": 0.5427368421052632, + "grad_norm": 0.546875, + "learning_rate": 0.00010161403508771929, + "loss": 3.1963, + "step": 10312 + }, + { + "epoch": 0.5427894736842105, + "grad_norm": 0.51171875, + "learning_rate": 0.00010160233918128657, + "loss": 3.0324, + "step": 10313 + }, + { + "epoch": 0.5428421052631579, + "grad_norm": 0.4609375, + "learning_rate": 0.00010159064327485382, + "loss": 3.3086, + "step": 10314 + }, + { + "epoch": 0.5428947368421052, + "grad_norm": 0.474609375, + "learning_rate": 0.00010157894736842107, + "loss": 3.6755, + "step": 10315 + }, + { + "epoch": 0.5429473684210526, + "grad_norm": 0.478515625, + "learning_rate": 0.00010156725146198832, + "loss": 3.2299, + "step": 10316 + }, + { + "epoch": 0.543, + "grad_norm": 0.498046875, + "learning_rate": 0.00010155555555555557, + "loss": 3.5291, + "step": 10317 + }, + { + "epoch": 0.5430526315789473, + "grad_norm": 0.484375, + "learning_rate": 0.0001015438596491228, + "loss": 3.5284, + "step": 10318 + }, + { + "epoch": 0.5431052631578948, + "grad_norm": 0.52734375, + "learning_rate": 0.00010153216374269005, + "loss": 4.1134, + "step": 10319 + }, + { + "epoch": 0.5431578947368421, + "grad_norm": 0.53515625, + "learning_rate": 0.0001015204678362573, + "loss": 3.5405, + "step": 10320 + }, + { + "epoch": 0.5432105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.00010150877192982458, + "loss": 3.5339, + "step": 10321 + }, + { + "epoch": 0.5432631578947369, + "grad_norm": 0.63671875, + "learning_rate": 0.00010149707602339183, + "loss": 3.1946, + "step": 10322 + }, + { + "epoch": 0.5433157894736842, + "grad_norm": 0.515625, + "learning_rate": 0.00010148538011695908, + "loss": 2.7876, + "step": 10323 + }, + { + "epoch": 0.5433684210526316, + "grad_norm": 0.46875, + "learning_rate": 0.00010147368421052633, + "loss": 3.7079, + "step": 10324 + }, + { + "epoch": 0.5434210526315789, + "grad_norm": 0.5546875, + "learning_rate": 0.00010146198830409358, + "loss": 3.4754, + "step": 10325 + }, + { + "epoch": 0.5434736842105263, + "grad_norm": 0.49609375, + "learning_rate": 0.00010145029239766081, + "loss": 3.7388, + "step": 10326 + }, + { + "epoch": 0.5435263157894736, + "grad_norm": 0.52734375, + "learning_rate": 0.00010143859649122806, + "loss": 3.6854, + "step": 10327 + }, + { + "epoch": 0.543578947368421, + "grad_norm": 0.52734375, + "learning_rate": 0.00010142690058479534, + "loss": 3.8361, + "step": 10328 + }, + { + "epoch": 0.5436315789473685, + "grad_norm": 0.5546875, + "learning_rate": 0.00010141520467836259, + "loss": 3.2171, + "step": 10329 + }, + { + "epoch": 0.5436842105263158, + "grad_norm": 0.490234375, + "learning_rate": 0.00010140350877192984, + "loss": 3.0951, + "step": 10330 + }, + { + "epoch": 0.5437368421052632, + "grad_norm": 0.48046875, + "learning_rate": 0.00010139181286549709, + "loss": 3.1658, + "step": 10331 + }, + { + "epoch": 0.5437894736842105, + "grad_norm": 0.458984375, + "learning_rate": 0.00010138011695906434, + "loss": 3.4443, + "step": 10332 + }, + { + "epoch": 0.5438421052631579, + "grad_norm": 0.494140625, + "learning_rate": 0.00010136842105263159, + "loss": 3.0894, + "step": 10333 + }, + { + "epoch": 0.5438947368421052, + "grad_norm": 0.470703125, + "learning_rate": 0.00010135672514619882, + "loss": 3.3228, + "step": 10334 + }, + { + "epoch": 0.5439473684210526, + "grad_norm": 0.458984375, + "learning_rate": 0.00010134502923976607, + "loss": 3.2723, + "step": 10335 + }, + { + "epoch": 0.544, + "grad_norm": 0.494140625, + "learning_rate": 0.00010133333333333335, + "loss": 3.2085, + "step": 10336 + }, + { + "epoch": 0.5440526315789473, + "grad_norm": 0.498046875, + "learning_rate": 0.0001013216374269006, + "loss": 3.5169, + "step": 10337 + }, + { + "epoch": 0.5441052631578948, + "grad_norm": 0.458984375, + "learning_rate": 0.00010130994152046785, + "loss": 3.8513, + "step": 10338 + }, + { + "epoch": 0.5441578947368421, + "grad_norm": 0.494140625, + "learning_rate": 0.0001012982456140351, + "loss": 3.5179, + "step": 10339 + }, + { + "epoch": 0.5442105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.00010128654970760235, + "loss": 3.5647, + "step": 10340 + }, + { + "epoch": 0.5442631578947369, + "grad_norm": 0.48828125, + "learning_rate": 0.00010127485380116958, + "loss": 3.2915, + "step": 10341 + }, + { + "epoch": 0.5443157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00010126315789473683, + "loss": 3.7464, + "step": 10342 + }, + { + "epoch": 0.5443684210526316, + "grad_norm": 0.51953125, + "learning_rate": 0.00010125146198830411, + "loss": 3.5093, + "step": 10343 + }, + { + "epoch": 0.5444210526315789, + "grad_norm": 0.50390625, + "learning_rate": 0.00010123976608187136, + "loss": 3.505, + "step": 10344 + }, + { + "epoch": 0.5444736842105263, + "grad_norm": 0.56640625, + "learning_rate": 0.00010122807017543861, + "loss": 3.3179, + "step": 10345 + }, + { + "epoch": 0.5445263157894736, + "grad_norm": 0.486328125, + "learning_rate": 0.00010121637426900586, + "loss": 3.307, + "step": 10346 + }, + { + "epoch": 0.544578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00010120467836257311, + "loss": 3.7336, + "step": 10347 + }, + { + "epoch": 0.5446315789473685, + "grad_norm": 0.50390625, + "learning_rate": 0.00010119298245614036, + "loss": 3.3043, + "step": 10348 + }, + { + "epoch": 0.5446842105263158, + "grad_norm": 0.48828125, + "learning_rate": 0.00010118128654970759, + "loss": 3.3873, + "step": 10349 + }, + { + "epoch": 0.5447368421052632, + "grad_norm": 0.99609375, + "learning_rate": 0.00010116959064327487, + "loss": 3.2927, + "step": 10350 + }, + { + "epoch": 0.5447894736842105, + "grad_norm": 0.53125, + "learning_rate": 0.00010115789473684212, + "loss": 3.59, + "step": 10351 + }, + { + "epoch": 0.5448421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00010114619883040937, + "loss": 3.1485, + "step": 10352 + }, + { + "epoch": 0.5448947368421052, + "grad_norm": 0.515625, + "learning_rate": 0.00010113450292397662, + "loss": 3.7576, + "step": 10353 + }, + { + "epoch": 0.5449473684210526, + "grad_norm": 0.5078125, + "learning_rate": 0.00010112280701754387, + "loss": 3.5197, + "step": 10354 + }, + { + "epoch": 0.545, + "grad_norm": 0.5859375, + "learning_rate": 0.00010111111111111112, + "loss": 3.2631, + "step": 10355 + }, + { + "epoch": 0.5450526315789473, + "grad_norm": 0.50390625, + "learning_rate": 0.00010109941520467837, + "loss": 3.7118, + "step": 10356 + }, + { + "epoch": 0.5451052631578948, + "grad_norm": 0.50390625, + "learning_rate": 0.0001010877192982456, + "loss": 3.1329, + "step": 10357 + }, + { + "epoch": 0.5451578947368421, + "grad_norm": 0.53125, + "learning_rate": 0.00010107602339181288, + "loss": 3.2643, + "step": 10358 + }, + { + "epoch": 0.5452105263157895, + "grad_norm": 0.5546875, + "learning_rate": 0.00010106432748538013, + "loss": 3.3149, + "step": 10359 + }, + { + "epoch": 0.5452631578947369, + "grad_norm": 0.474609375, + "learning_rate": 0.00010105263157894738, + "loss": 3.4435, + "step": 10360 + }, + { + "epoch": 0.5453157894736842, + "grad_norm": 0.458984375, + "learning_rate": 0.00010104093567251463, + "loss": 3.1856, + "step": 10361 + }, + { + "epoch": 0.5453684210526316, + "grad_norm": 0.462890625, + "learning_rate": 0.00010102923976608188, + "loss": 3.2787, + "step": 10362 + }, + { + "epoch": 0.5454210526315789, + "grad_norm": 0.5078125, + "learning_rate": 0.00010101754385964913, + "loss": 3.2767, + "step": 10363 + }, + { + "epoch": 0.5454736842105263, + "grad_norm": 0.5546875, + "learning_rate": 0.00010100584795321636, + "loss": 3.4944, + "step": 10364 + }, + { + "epoch": 0.5455263157894736, + "grad_norm": 0.5390625, + "learning_rate": 0.00010099415204678364, + "loss": 3.2638, + "step": 10365 + }, + { + "epoch": 0.545578947368421, + "grad_norm": 0.482421875, + "learning_rate": 0.00010098245614035089, + "loss": 2.9366, + "step": 10366 + }, + { + "epoch": 0.5456315789473685, + "grad_norm": 0.546875, + "learning_rate": 0.00010097076023391814, + "loss": 3.2889, + "step": 10367 + }, + { + "epoch": 0.5456842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.00010095906432748539, + "loss": 3.5817, + "step": 10368 + }, + { + "epoch": 0.5457368421052632, + "grad_norm": 0.546875, + "learning_rate": 0.00010094736842105264, + "loss": 3.6154, + "step": 10369 + }, + { + "epoch": 0.5457894736842105, + "grad_norm": 0.515625, + "learning_rate": 0.00010093567251461989, + "loss": 3.7716, + "step": 10370 + }, + { + "epoch": 0.5458421052631579, + "grad_norm": 0.5, + "learning_rate": 0.00010092397660818714, + "loss": 3.3166, + "step": 10371 + }, + { + "epoch": 0.5458947368421052, + "grad_norm": 0.4609375, + "learning_rate": 0.00010091228070175437, + "loss": 3.3535, + "step": 10372 + }, + { + "epoch": 0.5459473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00010090058479532165, + "loss": 3.3122, + "step": 10373 + }, + { + "epoch": 0.546, + "grad_norm": 0.4921875, + "learning_rate": 0.0001008888888888889, + "loss": 3.5578, + "step": 10374 + }, + { + "epoch": 0.5460526315789473, + "grad_norm": 0.494140625, + "learning_rate": 0.00010087719298245615, + "loss": 3.3392, + "step": 10375 + }, + { + "epoch": 0.5461052631578948, + "grad_norm": 0.53125, + "learning_rate": 0.0001008654970760234, + "loss": 3.3778, + "step": 10376 + }, + { + "epoch": 0.5461578947368421, + "grad_norm": 0.4609375, + "learning_rate": 0.00010085380116959065, + "loss": 3.3482, + "step": 10377 + }, + { + "epoch": 0.5462105263157895, + "grad_norm": 0.5234375, + "learning_rate": 0.0001008421052631579, + "loss": 3.2088, + "step": 10378 + }, + { + "epoch": 0.5462631578947369, + "grad_norm": 0.470703125, + "learning_rate": 0.00010083040935672515, + "loss": 3.7283, + "step": 10379 + }, + { + "epoch": 0.5463157894736842, + "grad_norm": 0.490234375, + "learning_rate": 0.00010081871345029241, + "loss": 3.3262, + "step": 10380 + }, + { + "epoch": 0.5463684210526316, + "grad_norm": 0.49609375, + "learning_rate": 0.00010080701754385966, + "loss": 3.5468, + "step": 10381 + }, + { + "epoch": 0.5464210526315789, + "grad_norm": 0.458984375, + "learning_rate": 0.00010079532163742691, + "loss": 3.6653, + "step": 10382 + }, + { + "epoch": 0.5464736842105263, + "grad_norm": 0.498046875, + "learning_rate": 0.00010078362573099416, + "loss": 2.8948, + "step": 10383 + }, + { + "epoch": 0.5465263157894736, + "grad_norm": 0.478515625, + "learning_rate": 0.00010077192982456141, + "loss": 3.3315, + "step": 10384 + }, + { + "epoch": 0.546578947368421, + "grad_norm": 0.474609375, + "learning_rate": 0.00010076023391812866, + "loss": 3.4881, + "step": 10385 + }, + { + "epoch": 0.5466315789473685, + "grad_norm": 0.5078125, + "learning_rate": 0.00010074853801169591, + "loss": 3.5811, + "step": 10386 + }, + { + "epoch": 0.5466842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.00010073684210526314, + "loss": 3.6729, + "step": 10387 + }, + { + "epoch": 0.5467368421052632, + "grad_norm": 0.5234375, + "learning_rate": 0.00010072514619883042, + "loss": 3.7335, + "step": 10388 + }, + { + "epoch": 0.5467894736842105, + "grad_norm": 0.4921875, + "learning_rate": 0.00010071345029239767, + "loss": 3.4773, + "step": 10389 + }, + { + "epoch": 0.5468421052631579, + "grad_norm": 0.451171875, + "learning_rate": 0.00010070175438596492, + "loss": 3.5766, + "step": 10390 + }, + { + "epoch": 0.5468947368421052, + "grad_norm": 0.482421875, + "learning_rate": 0.00010069005847953217, + "loss": 3.2824, + "step": 10391 + }, + { + "epoch": 0.5469473684210526, + "grad_norm": 0.486328125, + "learning_rate": 0.00010067836257309942, + "loss": 3.3755, + "step": 10392 + }, + { + "epoch": 0.547, + "grad_norm": 0.56640625, + "learning_rate": 0.00010066666666666667, + "loss": 3.4699, + "step": 10393 + }, + { + "epoch": 0.5470526315789473, + "grad_norm": 0.6328125, + "learning_rate": 0.00010065497076023392, + "loss": 3.6289, + "step": 10394 + }, + { + "epoch": 0.5471052631578948, + "grad_norm": 0.484375, + "learning_rate": 0.00010064327485380118, + "loss": 3.6677, + "step": 10395 + }, + { + "epoch": 0.5471578947368421, + "grad_norm": 0.486328125, + "learning_rate": 0.00010063157894736843, + "loss": 3.7594, + "step": 10396 + }, + { + "epoch": 0.5472105263157895, + "grad_norm": 0.546875, + "learning_rate": 0.00010061988304093568, + "loss": 3.1396, + "step": 10397 + }, + { + "epoch": 0.5472631578947369, + "grad_norm": 0.5078125, + "learning_rate": 0.00010060818713450293, + "loss": 3.1552, + "step": 10398 + }, + { + "epoch": 0.5473157894736842, + "grad_norm": 0.470703125, + "learning_rate": 0.00010059649122807018, + "loss": 3.5041, + "step": 10399 + }, + { + "epoch": 0.5473684210526316, + "grad_norm": 0.48828125, + "learning_rate": 0.00010058479532163743, + "loss": 3.373, + "step": 10400 + }, + { + "epoch": 0.5474210526315789, + "grad_norm": 0.5, + "learning_rate": 0.00010057309941520468, + "loss": 3.5338, + "step": 10401 + }, + { + "epoch": 0.5474736842105263, + "grad_norm": 0.53125, + "learning_rate": 0.00010056140350877194, + "loss": 3.1758, + "step": 10402 + }, + { + "epoch": 0.5475263157894736, + "grad_norm": 0.48046875, + "learning_rate": 0.00010054970760233919, + "loss": 3.4745, + "step": 10403 + }, + { + "epoch": 0.547578947368421, + "grad_norm": 0.458984375, + "learning_rate": 0.00010053801169590644, + "loss": 3.9025, + "step": 10404 + }, + { + "epoch": 0.5476315789473685, + "grad_norm": 0.54296875, + "learning_rate": 0.00010052631578947369, + "loss": 3.3404, + "step": 10405 + }, + { + "epoch": 0.5476842105263158, + "grad_norm": 0.458984375, + "learning_rate": 0.00010051461988304094, + "loss": 3.1805, + "step": 10406 + }, + { + "epoch": 0.5477368421052632, + "grad_norm": 0.48046875, + "learning_rate": 0.00010050292397660819, + "loss": 3.3533, + "step": 10407 + }, + { + "epoch": 0.5477894736842105, + "grad_norm": 0.5390625, + "learning_rate": 0.00010049122807017544, + "loss": 3.335, + "step": 10408 + }, + { + "epoch": 0.5478421052631579, + "grad_norm": 0.546875, + "learning_rate": 0.00010047953216374269, + "loss": 3.3705, + "step": 10409 + }, + { + "epoch": 0.5478947368421052, + "grad_norm": 0.5546875, + "learning_rate": 0.00010046783625730995, + "loss": 3.478, + "step": 10410 + }, + { + "epoch": 0.5479473684210526, + "grad_norm": 0.470703125, + "learning_rate": 0.0001004561403508772, + "loss": 3.421, + "step": 10411 + }, + { + "epoch": 0.548, + "grad_norm": 0.5078125, + "learning_rate": 0.00010044444444444445, + "loss": 3.7438, + "step": 10412 + }, + { + "epoch": 0.5480526315789473, + "grad_norm": 0.5, + "learning_rate": 0.0001004327485380117, + "loss": 3.1799, + "step": 10413 + }, + { + "epoch": 0.5481052631578948, + "grad_norm": 0.51171875, + "learning_rate": 0.00010042105263157895, + "loss": 3.3899, + "step": 10414 + }, + { + "epoch": 0.5481578947368421, + "grad_norm": 0.51171875, + "learning_rate": 0.0001004093567251462, + "loss": 3.5485, + "step": 10415 + }, + { + "epoch": 0.5482105263157895, + "grad_norm": 0.6015625, + "learning_rate": 0.00010039766081871345, + "loss": 3.8444, + "step": 10416 + }, + { + "epoch": 0.5482631578947369, + "grad_norm": 0.47265625, + "learning_rate": 0.00010038596491228071, + "loss": 3.3873, + "step": 10417 + }, + { + "epoch": 0.5483157894736842, + "grad_norm": 0.5, + "learning_rate": 0.00010037426900584796, + "loss": 3.277, + "step": 10418 + }, + { + "epoch": 0.5483684210526316, + "grad_norm": 0.486328125, + "learning_rate": 0.00010036257309941521, + "loss": 3.2996, + "step": 10419 + }, + { + "epoch": 0.5484210526315789, + "grad_norm": 0.47265625, + "learning_rate": 0.00010035087719298246, + "loss": 3.7474, + "step": 10420 + }, + { + "epoch": 0.5484736842105263, + "grad_norm": 0.478515625, + "learning_rate": 0.00010033918128654971, + "loss": 2.8538, + "step": 10421 + }, + { + "epoch": 0.5485263157894736, + "grad_norm": 0.4609375, + "learning_rate": 0.00010032748538011696, + "loss": 3.2526, + "step": 10422 + }, + { + "epoch": 0.548578947368421, + "grad_norm": 0.48046875, + "learning_rate": 0.00010031578947368421, + "loss": 3.3901, + "step": 10423 + }, + { + "epoch": 0.5486315789473685, + "grad_norm": 0.51953125, + "learning_rate": 0.00010030409356725146, + "loss": 3.4185, + "step": 10424 + }, + { + "epoch": 0.5486842105263158, + "grad_norm": 0.46484375, + "learning_rate": 0.00010029239766081872, + "loss": 3.4913, + "step": 10425 + }, + { + "epoch": 0.5487368421052632, + "grad_norm": 0.53515625, + "learning_rate": 0.00010028070175438597, + "loss": 2.7941, + "step": 10426 + }, + { + "epoch": 0.5487894736842105, + "grad_norm": 0.48046875, + "learning_rate": 0.00010026900584795322, + "loss": 3.4582, + "step": 10427 + }, + { + "epoch": 0.5488421052631579, + "grad_norm": 0.474609375, + "learning_rate": 0.00010025730994152047, + "loss": 3.3575, + "step": 10428 + }, + { + "epoch": 0.5488947368421052, + "grad_norm": 0.5078125, + "learning_rate": 0.00010024561403508772, + "loss": 3.5443, + "step": 10429 + }, + { + "epoch": 0.5489473684210526, + "grad_norm": 0.482421875, + "learning_rate": 0.00010023391812865497, + "loss": 3.3999, + "step": 10430 + }, + { + "epoch": 0.549, + "grad_norm": 0.48046875, + "learning_rate": 0.00010022222222222222, + "loss": 3.4053, + "step": 10431 + }, + { + "epoch": 0.5490526315789473, + "grad_norm": 0.484375, + "learning_rate": 0.00010021052631578949, + "loss": 3.1026, + "step": 10432 + }, + { + "epoch": 0.5491052631578948, + "grad_norm": 0.46875, + "learning_rate": 0.00010019883040935673, + "loss": 3.3701, + "step": 10433 + }, + { + "epoch": 0.5491578947368421, + "grad_norm": 0.47265625, + "learning_rate": 0.00010018713450292398, + "loss": 3.6613, + "step": 10434 + }, + { + "epoch": 0.5492105263157895, + "grad_norm": 0.51953125, + "learning_rate": 0.00010017543859649123, + "loss": 3.4283, + "step": 10435 + }, + { + "epoch": 0.5492631578947369, + "grad_norm": 0.515625, + "learning_rate": 0.00010016374269005848, + "loss": 3.507, + "step": 10436 + }, + { + "epoch": 0.5493157894736842, + "grad_norm": 0.5390625, + "learning_rate": 0.00010015204678362573, + "loss": 3.7195, + "step": 10437 + }, + { + "epoch": 0.5493684210526316, + "grad_norm": 0.52734375, + "learning_rate": 0.00010014035087719298, + "loss": 3.3459, + "step": 10438 + }, + { + "epoch": 0.5494210526315789, + "grad_norm": 0.50390625, + "learning_rate": 0.00010012865497076025, + "loss": 3.623, + "step": 10439 + }, + { + "epoch": 0.5494736842105263, + "grad_norm": 0.56640625, + "learning_rate": 0.0001001169590643275, + "loss": 3.5782, + "step": 10440 + }, + { + "epoch": 0.5495263157894736, + "grad_norm": 0.494140625, + "learning_rate": 0.00010010526315789474, + "loss": 3.2155, + "step": 10441 + }, + { + "epoch": 0.5495789473684211, + "grad_norm": 0.455078125, + "learning_rate": 0.000100093567251462, + "loss": 3.5941, + "step": 10442 + }, + { + "epoch": 0.5496315789473685, + "grad_norm": 0.5078125, + "learning_rate": 0.00010008187134502924, + "loss": 3.496, + "step": 10443 + }, + { + "epoch": 0.5496842105263158, + "grad_norm": 0.5078125, + "learning_rate": 0.0001000701754385965, + "loss": 3.1176, + "step": 10444 + }, + { + "epoch": 0.5497368421052632, + "grad_norm": 0.5078125, + "learning_rate": 0.00010005847953216374, + "loss": 3.7293, + "step": 10445 + }, + { + "epoch": 0.5497894736842105, + "grad_norm": 0.51953125, + "learning_rate": 0.00010004678362573099, + "loss": 3.5508, + "step": 10446 + }, + { + "epoch": 0.5498421052631579, + "grad_norm": 0.515625, + "learning_rate": 0.00010003508771929826, + "loss": 3.7447, + "step": 10447 + }, + { + "epoch": 0.5498947368421052, + "grad_norm": 0.5390625, + "learning_rate": 0.0001000233918128655, + "loss": 3.2911, + "step": 10448 + }, + { + "epoch": 0.5499473684210526, + "grad_norm": 0.51953125, + "learning_rate": 0.00010001169590643276, + "loss": 3.566, + "step": 10449 + }, + { + "epoch": 0.55, + "grad_norm": 0.50390625, + "learning_rate": 0.0001, + "loss": 3.2822, + "step": 10450 + }, + { + "epoch": 0.5500526315789473, + "grad_norm": 0.466796875, + "learning_rate": 9.998830409356725e-05, + "loss": 3.2791, + "step": 10451 + }, + { + "epoch": 0.5501052631578948, + "grad_norm": 0.470703125, + "learning_rate": 9.997660818713452e-05, + "loss": 3.785, + "step": 10452 + }, + { + "epoch": 0.5501578947368421, + "grad_norm": 0.48046875, + "learning_rate": 9.996491228070177e-05, + "loss": 3.6764, + "step": 10453 + }, + { + "epoch": 0.5502105263157895, + "grad_norm": 0.498046875, + "learning_rate": 9.9953216374269e-05, + "loss": 3.0291, + "step": 10454 + }, + { + "epoch": 0.5502631578947368, + "grad_norm": 0.470703125, + "learning_rate": 9.994152046783625e-05, + "loss": 3.4561, + "step": 10455 + }, + { + "epoch": 0.5503157894736842, + "grad_norm": 0.474609375, + "learning_rate": 9.992982456140352e-05, + "loss": 3.3654, + "step": 10456 + }, + { + "epoch": 0.5503684210526316, + "grad_norm": 0.47265625, + "learning_rate": 9.991812865497077e-05, + "loss": 3.505, + "step": 10457 + }, + { + "epoch": 0.5504210526315789, + "grad_norm": 0.5234375, + "learning_rate": 9.990643274853802e-05, + "loss": 3.2107, + "step": 10458 + }, + { + "epoch": 0.5504736842105263, + "grad_norm": 0.47265625, + "learning_rate": 9.989473684210526e-05, + "loss": 3.801, + "step": 10459 + }, + { + "epoch": 0.5505263157894736, + "grad_norm": 0.4765625, + "learning_rate": 9.988304093567253e-05, + "loss": 3.6913, + "step": 10460 + }, + { + "epoch": 0.5505789473684211, + "grad_norm": 0.458984375, + "learning_rate": 9.987134502923978e-05, + "loss": 3.6728, + "step": 10461 + }, + { + "epoch": 0.5506315789473685, + "grad_norm": 0.470703125, + "learning_rate": 9.985964912280701e-05, + "loss": 3.1441, + "step": 10462 + }, + { + "epoch": 0.5506842105263158, + "grad_norm": 0.46875, + "learning_rate": 9.984795321637428e-05, + "loss": 3.2115, + "step": 10463 + }, + { + "epoch": 0.5507368421052632, + "grad_norm": 0.50390625, + "learning_rate": 9.983625730994153e-05, + "loss": 3.5099, + "step": 10464 + }, + { + "epoch": 0.5507894736842105, + "grad_norm": 0.486328125, + "learning_rate": 9.982456140350878e-05, + "loss": 3.688, + "step": 10465 + }, + { + "epoch": 0.5508421052631579, + "grad_norm": 0.5078125, + "learning_rate": 9.981286549707603e-05, + "loss": 3.5239, + "step": 10466 + }, + { + "epoch": 0.5508947368421052, + "grad_norm": 0.494140625, + "learning_rate": 9.980116959064329e-05, + "loss": 3.1563, + "step": 10467 + }, + { + "epoch": 0.5509473684210526, + "grad_norm": 0.47265625, + "learning_rate": 9.978947368421054e-05, + "loss": 3.5106, + "step": 10468 + }, + { + "epoch": 0.551, + "grad_norm": 0.494140625, + "learning_rate": 9.977777777777779e-05, + "loss": 3.7428, + "step": 10469 + }, + { + "epoch": 0.5510526315789473, + "grad_norm": 0.478515625, + "learning_rate": 9.976608187134502e-05, + "loss": 3.5307, + "step": 10470 + }, + { + "epoch": 0.5511052631578948, + "grad_norm": 0.48046875, + "learning_rate": 9.975438596491229e-05, + "loss": 3.6167, + "step": 10471 + }, + { + "epoch": 0.5511578947368421, + "grad_norm": 0.4765625, + "learning_rate": 9.974269005847954e-05, + "loss": 3.0101, + "step": 10472 + }, + { + "epoch": 0.5512105263157895, + "grad_norm": 0.486328125, + "learning_rate": 9.973099415204679e-05, + "loss": 3.7024, + "step": 10473 + }, + { + "epoch": 0.5512631578947368, + "grad_norm": 0.5703125, + "learning_rate": 9.971929824561404e-05, + "loss": 3.0511, + "step": 10474 + }, + { + "epoch": 0.5513157894736842, + "grad_norm": 0.482421875, + "learning_rate": 9.97076023391813e-05, + "loss": 3.233, + "step": 10475 + }, + { + "epoch": 0.5513684210526316, + "grad_norm": 0.6484375, + "learning_rate": 9.969590643274855e-05, + "loss": 3.6249, + "step": 10476 + }, + { + "epoch": 0.5514210526315789, + "grad_norm": 0.482421875, + "learning_rate": 9.968421052631578e-05, + "loss": 3.48, + "step": 10477 + }, + { + "epoch": 0.5514736842105263, + "grad_norm": 0.484375, + "learning_rate": 9.967251461988305e-05, + "loss": 3.0985, + "step": 10478 + }, + { + "epoch": 0.5515263157894736, + "grad_norm": 0.515625, + "learning_rate": 9.96608187134503e-05, + "loss": 3.065, + "step": 10479 + }, + { + "epoch": 0.5515789473684211, + "grad_norm": 0.4921875, + "learning_rate": 9.964912280701755e-05, + "loss": 3.738, + "step": 10480 + }, + { + "epoch": 0.5516315789473685, + "grad_norm": 0.50390625, + "learning_rate": 9.96374269005848e-05, + "loss": 3.3517, + "step": 10481 + }, + { + "epoch": 0.5516842105263158, + "grad_norm": 0.5, + "learning_rate": 9.962573099415206e-05, + "loss": 3.7401, + "step": 10482 + }, + { + "epoch": 0.5517368421052632, + "grad_norm": 0.5234375, + "learning_rate": 9.961403508771931e-05, + "loss": 3.4465, + "step": 10483 + }, + { + "epoch": 0.5517894736842105, + "grad_norm": 0.66015625, + "learning_rate": 9.960233918128656e-05, + "loss": 3.2169, + "step": 10484 + }, + { + "epoch": 0.5518421052631579, + "grad_norm": 0.50390625, + "learning_rate": 9.95906432748538e-05, + "loss": 2.8537, + "step": 10485 + }, + { + "epoch": 0.5518947368421052, + "grad_norm": 0.5, + "learning_rate": 9.957894736842106e-05, + "loss": 3.5546, + "step": 10486 + }, + { + "epoch": 0.5519473684210526, + "grad_norm": 0.46484375, + "learning_rate": 9.956725146198831e-05, + "loss": 3.2822, + "step": 10487 + }, + { + "epoch": 0.552, + "grad_norm": 0.494140625, + "learning_rate": 9.955555555555556e-05, + "loss": 3.5503, + "step": 10488 + }, + { + "epoch": 0.5520526315789474, + "grad_norm": 0.5078125, + "learning_rate": 9.954385964912282e-05, + "loss": 3.615, + "step": 10489 + }, + { + "epoch": 0.5521052631578948, + "grad_norm": 0.5, + "learning_rate": 9.953216374269007e-05, + "loss": 3.2554, + "step": 10490 + }, + { + "epoch": 0.5521578947368421, + "grad_norm": 0.5, + "learning_rate": 9.952046783625732e-05, + "loss": 3.7935, + "step": 10491 + }, + { + "epoch": 0.5522105263157895, + "grad_norm": 0.486328125, + "learning_rate": 9.950877192982457e-05, + "loss": 3.117, + "step": 10492 + }, + { + "epoch": 0.5522631578947368, + "grad_norm": 0.48046875, + "learning_rate": 9.949707602339182e-05, + "loss": 3.8867, + "step": 10493 + }, + { + "epoch": 0.5523157894736842, + "grad_norm": 0.5234375, + "learning_rate": 9.948538011695907e-05, + "loss": 3.5183, + "step": 10494 + }, + { + "epoch": 0.5523684210526316, + "grad_norm": 0.5234375, + "learning_rate": 9.947368421052632e-05, + "loss": 3.778, + "step": 10495 + }, + { + "epoch": 0.5524210526315789, + "grad_norm": 0.47265625, + "learning_rate": 9.946198830409357e-05, + "loss": 3.7772, + "step": 10496 + }, + { + "epoch": 0.5524736842105263, + "grad_norm": 0.75390625, + "learning_rate": 9.945029239766083e-05, + "loss": 3.0165, + "step": 10497 + }, + { + "epoch": 0.5525263157894736, + "grad_norm": 0.453125, + "learning_rate": 9.943859649122808e-05, + "loss": 3.4234, + "step": 10498 + }, + { + "epoch": 0.5525789473684211, + "grad_norm": 0.484375, + "learning_rate": 9.942690058479533e-05, + "loss": 3.7488, + "step": 10499 + }, + { + "epoch": 0.5526315789473685, + "grad_norm": 0.486328125, + "learning_rate": 9.941520467836257e-05, + "loss": 3.4848, + "step": 10500 + }, + { + "epoch": 0.5526842105263158, + "grad_norm": 0.5, + "learning_rate": 9.940350877192983e-05, + "loss": 3.3917, + "step": 10501 + }, + { + "epoch": 0.5527368421052632, + "grad_norm": 0.5078125, + "learning_rate": 9.939181286549708e-05, + "loss": 3.4507, + "step": 10502 + }, + { + "epoch": 0.5527894736842105, + "grad_norm": 0.53515625, + "learning_rate": 9.938011695906433e-05, + "loss": 3.4948, + "step": 10503 + }, + { + "epoch": 0.5528421052631579, + "grad_norm": 0.478515625, + "learning_rate": 9.936842105263159e-05, + "loss": 3.2439, + "step": 10504 + }, + { + "epoch": 0.5528947368421052, + "grad_norm": 0.5078125, + "learning_rate": 9.935672514619884e-05, + "loss": 3.3566, + "step": 10505 + }, + { + "epoch": 0.5529473684210526, + "grad_norm": 0.48046875, + "learning_rate": 9.934502923976609e-05, + "loss": 3.6197, + "step": 10506 + }, + { + "epoch": 0.553, + "grad_norm": 0.51171875, + "learning_rate": 9.933333333333334e-05, + "loss": 3.2477, + "step": 10507 + }, + { + "epoch": 0.5530526315789474, + "grad_norm": 0.453125, + "learning_rate": 9.932163742690059e-05, + "loss": 2.7735, + "step": 10508 + }, + { + "epoch": 0.5531052631578948, + "grad_norm": 0.625, + "learning_rate": 9.930994152046784e-05, + "loss": 3.2137, + "step": 10509 + }, + { + "epoch": 0.5531578947368421, + "grad_norm": 0.470703125, + "learning_rate": 9.929824561403509e-05, + "loss": 3.3853, + "step": 10510 + }, + { + "epoch": 0.5532105263157895, + "grad_norm": 0.4609375, + "learning_rate": 9.928654970760234e-05, + "loss": 3.3376, + "step": 10511 + }, + { + "epoch": 0.5532631578947368, + "grad_norm": 0.5, + "learning_rate": 9.92748538011696e-05, + "loss": 3.5091, + "step": 10512 + }, + { + "epoch": 0.5533157894736842, + "grad_norm": 0.474609375, + "learning_rate": 9.926315789473685e-05, + "loss": 3.2679, + "step": 10513 + }, + { + "epoch": 0.5533684210526316, + "grad_norm": 0.4921875, + "learning_rate": 9.92514619883041e-05, + "loss": 3.6609, + "step": 10514 + }, + { + "epoch": 0.5534210526315789, + "grad_norm": 0.494140625, + "learning_rate": 9.923976608187135e-05, + "loss": 3.3441, + "step": 10515 + }, + { + "epoch": 0.5534736842105263, + "grad_norm": 0.498046875, + "learning_rate": 9.92280701754386e-05, + "loss": 3.5336, + "step": 10516 + }, + { + "epoch": 0.5535263157894736, + "grad_norm": 0.5, + "learning_rate": 9.921637426900585e-05, + "loss": 3.1832, + "step": 10517 + }, + { + "epoch": 0.5535789473684211, + "grad_norm": 0.451171875, + "learning_rate": 9.92046783625731e-05, + "loss": 3.4167, + "step": 10518 + }, + { + "epoch": 0.5536315789473685, + "grad_norm": 0.474609375, + "learning_rate": 9.919298245614036e-05, + "loss": 3.5141, + "step": 10519 + }, + { + "epoch": 0.5536842105263158, + "grad_norm": 0.5390625, + "learning_rate": 9.918128654970761e-05, + "loss": 3.5358, + "step": 10520 + }, + { + "epoch": 0.5537368421052632, + "grad_norm": 0.515625, + "learning_rate": 9.916959064327486e-05, + "loss": 3.5001, + "step": 10521 + }, + { + "epoch": 0.5537894736842105, + "grad_norm": 0.498046875, + "learning_rate": 9.915789473684211e-05, + "loss": 3.4961, + "step": 10522 + }, + { + "epoch": 0.5538421052631579, + "grad_norm": 0.49609375, + "learning_rate": 9.914619883040936e-05, + "loss": 3.1563, + "step": 10523 + }, + { + "epoch": 0.5538947368421052, + "grad_norm": 0.466796875, + "learning_rate": 9.913450292397661e-05, + "loss": 3.0791, + "step": 10524 + }, + { + "epoch": 0.5539473684210526, + "grad_norm": 0.478515625, + "learning_rate": 9.912280701754386e-05, + "loss": 3.7924, + "step": 10525 + }, + { + "epoch": 0.554, + "grad_norm": 0.58203125, + "learning_rate": 9.911111111111112e-05, + "loss": 3.604, + "step": 10526 + }, + { + "epoch": 0.5540526315789474, + "grad_norm": 0.484375, + "learning_rate": 9.909941520467837e-05, + "loss": 2.8894, + "step": 10527 + }, + { + "epoch": 0.5541052631578948, + "grad_norm": 0.5, + "learning_rate": 9.908771929824562e-05, + "loss": 3.299, + "step": 10528 + }, + { + "epoch": 0.5541578947368421, + "grad_norm": 0.52734375, + "learning_rate": 9.907602339181287e-05, + "loss": 3.7221, + "step": 10529 + }, + { + "epoch": 0.5542105263157895, + "grad_norm": 0.546875, + "learning_rate": 9.906432748538012e-05, + "loss": 3.5422, + "step": 10530 + }, + { + "epoch": 0.5542631578947368, + "grad_norm": 0.48828125, + "learning_rate": 9.905263157894737e-05, + "loss": 3.3457, + "step": 10531 + }, + { + "epoch": 0.5543157894736842, + "grad_norm": 0.59765625, + "learning_rate": 9.904093567251462e-05, + "loss": 3.5775, + "step": 10532 + }, + { + "epoch": 0.5543684210526316, + "grad_norm": 0.5, + "learning_rate": 9.902923976608187e-05, + "loss": 3.5423, + "step": 10533 + }, + { + "epoch": 0.5544210526315789, + "grad_norm": 0.4765625, + "learning_rate": 9.901754385964913e-05, + "loss": 3.3486, + "step": 10534 + }, + { + "epoch": 0.5544736842105263, + "grad_norm": 0.50390625, + "learning_rate": 9.900584795321638e-05, + "loss": 3.6909, + "step": 10535 + }, + { + "epoch": 0.5545263157894736, + "grad_norm": 0.48828125, + "learning_rate": 9.899415204678363e-05, + "loss": 3.3236, + "step": 10536 + }, + { + "epoch": 0.5545789473684211, + "grad_norm": 0.50390625, + "learning_rate": 9.898245614035088e-05, + "loss": 3.4516, + "step": 10537 + }, + { + "epoch": 0.5546315789473685, + "grad_norm": 0.52734375, + "learning_rate": 9.897076023391813e-05, + "loss": 3.2993, + "step": 10538 + }, + { + "epoch": 0.5546842105263158, + "grad_norm": 0.51171875, + "learning_rate": 9.895906432748538e-05, + "loss": 3.2029, + "step": 10539 + }, + { + "epoch": 0.5547368421052632, + "grad_norm": 0.474609375, + "learning_rate": 9.894736842105263e-05, + "loss": 3.3547, + "step": 10540 + }, + { + "epoch": 0.5547894736842105, + "grad_norm": 0.486328125, + "learning_rate": 9.89356725146199e-05, + "loss": 3.7232, + "step": 10541 + }, + { + "epoch": 0.5548421052631579, + "grad_norm": 0.48046875, + "learning_rate": 9.892397660818714e-05, + "loss": 3.0019, + "step": 10542 + }, + { + "epoch": 0.5548947368421052, + "grad_norm": 0.470703125, + "learning_rate": 9.891228070175439e-05, + "loss": 3.7524, + "step": 10543 + }, + { + "epoch": 0.5549473684210526, + "grad_norm": 0.494140625, + "learning_rate": 9.890058479532164e-05, + "loss": 3.4211, + "step": 10544 + }, + { + "epoch": 0.555, + "grad_norm": 0.474609375, + "learning_rate": 9.888888888888889e-05, + "loss": 3.5507, + "step": 10545 + }, + { + "epoch": 0.5550526315789474, + "grad_norm": 0.46875, + "learning_rate": 9.887719298245614e-05, + "loss": 3.9192, + "step": 10546 + }, + { + "epoch": 0.5551052631578948, + "grad_norm": 0.498046875, + "learning_rate": 9.886549707602339e-05, + "loss": 3.306, + "step": 10547 + }, + { + "epoch": 0.5551578947368421, + "grad_norm": 0.53125, + "learning_rate": 9.885380116959064e-05, + "loss": 3.2724, + "step": 10548 + }, + { + "epoch": 0.5552105263157895, + "grad_norm": 0.458984375, + "learning_rate": 9.88421052631579e-05, + "loss": 3.1827, + "step": 10549 + }, + { + "epoch": 0.5552631578947368, + "grad_norm": 0.53515625, + "learning_rate": 9.883040935672515e-05, + "loss": 3.1589, + "step": 10550 + }, + { + "epoch": 0.5553157894736842, + "grad_norm": 0.4921875, + "learning_rate": 9.88187134502924e-05, + "loss": 2.9874, + "step": 10551 + }, + { + "epoch": 0.5553684210526316, + "grad_norm": 0.49609375, + "learning_rate": 9.880701754385965e-05, + "loss": 3.3043, + "step": 10552 + }, + { + "epoch": 0.5554210526315789, + "grad_norm": 0.474609375, + "learning_rate": 9.87953216374269e-05, + "loss": 3.1291, + "step": 10553 + }, + { + "epoch": 0.5554736842105263, + "grad_norm": 0.4765625, + "learning_rate": 9.878362573099415e-05, + "loss": 3.6573, + "step": 10554 + }, + { + "epoch": 0.5555263157894736, + "grad_norm": 0.51171875, + "learning_rate": 9.87719298245614e-05, + "loss": 3.4111, + "step": 10555 + }, + { + "epoch": 0.5555789473684211, + "grad_norm": 0.47265625, + "learning_rate": 9.876023391812866e-05, + "loss": 3.7487, + "step": 10556 + }, + { + "epoch": 0.5556315789473685, + "grad_norm": 0.48046875, + "learning_rate": 9.874853801169591e-05, + "loss": 3.5062, + "step": 10557 + }, + { + "epoch": 0.5556842105263158, + "grad_norm": 0.62890625, + "learning_rate": 9.873684210526316e-05, + "loss": 3.0153, + "step": 10558 + }, + { + "epoch": 0.5557368421052632, + "grad_norm": 0.482421875, + "learning_rate": 9.872514619883041e-05, + "loss": 3.3917, + "step": 10559 + }, + { + "epoch": 0.5557894736842105, + "grad_norm": 0.4765625, + "learning_rate": 9.871345029239766e-05, + "loss": 3.5622, + "step": 10560 + }, + { + "epoch": 0.5558421052631579, + "grad_norm": 0.49609375, + "learning_rate": 9.870175438596491e-05, + "loss": 2.9904, + "step": 10561 + }, + { + "epoch": 0.5558947368421052, + "grad_norm": 0.470703125, + "learning_rate": 9.869005847953216e-05, + "loss": 3.4435, + "step": 10562 + }, + { + "epoch": 0.5559473684210526, + "grad_norm": 0.455078125, + "learning_rate": 9.867836257309943e-05, + "loss": 3.4588, + "step": 10563 + }, + { + "epoch": 0.556, + "grad_norm": 0.466796875, + "learning_rate": 9.866666666666668e-05, + "loss": 3.2311, + "step": 10564 + }, + { + "epoch": 0.5560526315789474, + "grad_norm": 0.48828125, + "learning_rate": 9.865497076023392e-05, + "loss": 3.6626, + "step": 10565 + }, + { + "epoch": 0.5561052631578948, + "grad_norm": 0.52734375, + "learning_rate": 9.864327485380117e-05, + "loss": 3.5743, + "step": 10566 + }, + { + "epoch": 0.5561578947368421, + "grad_norm": 0.46484375, + "learning_rate": 9.863157894736842e-05, + "loss": 3.7117, + "step": 10567 + }, + { + "epoch": 0.5562105263157895, + "grad_norm": 0.490234375, + "learning_rate": 9.861988304093567e-05, + "loss": 3.5579, + "step": 10568 + }, + { + "epoch": 0.5562631578947368, + "grad_norm": 0.470703125, + "learning_rate": 9.860818713450292e-05, + "loss": 2.9361, + "step": 10569 + }, + { + "epoch": 0.5563157894736842, + "grad_norm": 0.48046875, + "learning_rate": 9.859649122807017e-05, + "loss": 3.4466, + "step": 10570 + }, + { + "epoch": 0.5563684210526316, + "grad_norm": 0.51171875, + "learning_rate": 9.858479532163744e-05, + "loss": 3.4158, + "step": 10571 + }, + { + "epoch": 0.5564210526315789, + "grad_norm": 0.470703125, + "learning_rate": 9.857309941520469e-05, + "loss": 3.5982, + "step": 10572 + }, + { + "epoch": 0.5564736842105263, + "grad_norm": 0.52734375, + "learning_rate": 9.856140350877194e-05, + "loss": 3.3992, + "step": 10573 + }, + { + "epoch": 0.5565263157894736, + "grad_norm": 0.478515625, + "learning_rate": 9.854970760233918e-05, + "loss": 3.2614, + "step": 10574 + }, + { + "epoch": 0.5565789473684211, + "grad_norm": 0.482421875, + "learning_rate": 9.853801169590643e-05, + "loss": 3.3142, + "step": 10575 + }, + { + "epoch": 0.5566315789473684, + "grad_norm": 0.9921875, + "learning_rate": 9.852631578947368e-05, + "loss": 3.7236, + "step": 10576 + }, + { + "epoch": 0.5566842105263158, + "grad_norm": 0.466796875, + "learning_rate": 9.851461988304093e-05, + "loss": 3.5482, + "step": 10577 + }, + { + "epoch": 0.5567368421052632, + "grad_norm": 0.5, + "learning_rate": 9.85029239766082e-05, + "loss": 3.5424, + "step": 10578 + }, + { + "epoch": 0.5567894736842105, + "grad_norm": 0.482421875, + "learning_rate": 9.849122807017545e-05, + "loss": 3.3509, + "step": 10579 + }, + { + "epoch": 0.5568421052631579, + "grad_norm": 0.48046875, + "learning_rate": 9.84795321637427e-05, + "loss": 3.9219, + "step": 10580 + }, + { + "epoch": 0.5568947368421052, + "grad_norm": 0.498046875, + "learning_rate": 9.846783625730995e-05, + "loss": 3.3247, + "step": 10581 + }, + { + "epoch": 0.5569473684210526, + "grad_norm": 0.48046875, + "learning_rate": 9.845614035087721e-05, + "loss": 3.1844, + "step": 10582 + }, + { + "epoch": 0.557, + "grad_norm": 0.4765625, + "learning_rate": 9.844444444444444e-05, + "loss": 3.4969, + "step": 10583 + }, + { + "epoch": 0.5570526315789474, + "grad_norm": 0.98828125, + "learning_rate": 9.84327485380117e-05, + "loss": 3.2773, + "step": 10584 + }, + { + "epoch": 0.5571052631578948, + "grad_norm": 0.431640625, + "learning_rate": 9.842105263157894e-05, + "loss": 3.1176, + "step": 10585 + }, + { + "epoch": 0.5571578947368421, + "grad_norm": 0.4453125, + "learning_rate": 9.840935672514621e-05, + "loss": 3.5602, + "step": 10586 + }, + { + "epoch": 0.5572105263157895, + "grad_norm": 0.4921875, + "learning_rate": 9.839766081871346e-05, + "loss": 3.2361, + "step": 10587 + }, + { + "epoch": 0.5572631578947368, + "grad_norm": 0.54296875, + "learning_rate": 9.83859649122807e-05, + "loss": 3.2047, + "step": 10588 + }, + { + "epoch": 0.5573157894736842, + "grad_norm": 0.52734375, + "learning_rate": 9.837426900584797e-05, + "loss": 3.2926, + "step": 10589 + }, + { + "epoch": 0.5573684210526316, + "grad_norm": 0.4921875, + "learning_rate": 9.83625730994152e-05, + "loss": 2.9567, + "step": 10590 + }, + { + "epoch": 0.5574210526315789, + "grad_norm": 0.546875, + "learning_rate": 9.835087719298245e-05, + "loss": 3.1983, + "step": 10591 + }, + { + "epoch": 0.5574736842105263, + "grad_norm": 0.466796875, + "learning_rate": 9.83391812865497e-05, + "loss": 3.3547, + "step": 10592 + }, + { + "epoch": 0.5575263157894736, + "grad_norm": 0.486328125, + "learning_rate": 9.832748538011697e-05, + "loss": 3.2919, + "step": 10593 + }, + { + "epoch": 0.5575789473684211, + "grad_norm": 0.478515625, + "learning_rate": 9.831578947368422e-05, + "loss": 3.1856, + "step": 10594 + }, + { + "epoch": 0.5576315789473684, + "grad_norm": 0.49609375, + "learning_rate": 9.830409356725147e-05, + "loss": 3.6737, + "step": 10595 + }, + { + "epoch": 0.5576842105263158, + "grad_norm": 0.91015625, + "learning_rate": 9.829239766081872e-05, + "loss": 3.6816, + "step": 10596 + }, + { + "epoch": 0.5577368421052632, + "grad_norm": 0.48046875, + "learning_rate": 9.828070175438598e-05, + "loss": 3.2497, + "step": 10597 + }, + { + "epoch": 0.5577894736842105, + "grad_norm": 0.466796875, + "learning_rate": 9.826900584795322e-05, + "loss": 3.7901, + "step": 10598 + }, + { + "epoch": 0.5578421052631579, + "grad_norm": 0.69140625, + "learning_rate": 9.825730994152047e-05, + "loss": 3.5094, + "step": 10599 + }, + { + "epoch": 0.5578947368421052, + "grad_norm": 0.55078125, + "learning_rate": 9.824561403508771e-05, + "loss": 3.3555, + "step": 10600 + }, + { + "epoch": 0.5579473684210526, + "grad_norm": 0.5078125, + "learning_rate": 9.823391812865498e-05, + "loss": 3.214, + "step": 10601 + }, + { + "epoch": 0.558, + "grad_norm": 0.48828125, + "learning_rate": 9.822222222222223e-05, + "loss": 3.5804, + "step": 10602 + }, + { + "epoch": 0.5580526315789474, + "grad_norm": 0.4921875, + "learning_rate": 9.821052631578948e-05, + "loss": 3.3662, + "step": 10603 + }, + { + "epoch": 0.5581052631578948, + "grad_norm": 0.48046875, + "learning_rate": 9.819883040935674e-05, + "loss": 3.5877, + "step": 10604 + }, + { + "epoch": 0.5581578947368421, + "grad_norm": 0.466796875, + "learning_rate": 9.818713450292399e-05, + "loss": 3.1514, + "step": 10605 + }, + { + "epoch": 0.5582105263157895, + "grad_norm": 0.498046875, + "learning_rate": 9.817543859649123e-05, + "loss": 3.3253, + "step": 10606 + }, + { + "epoch": 0.5582631578947368, + "grad_norm": 0.482421875, + "learning_rate": 9.816374269005848e-05, + "loss": 3.1158, + "step": 10607 + }, + { + "epoch": 0.5583157894736842, + "grad_norm": 0.462890625, + "learning_rate": 9.815204678362574e-05, + "loss": 3.5933, + "step": 10608 + }, + { + "epoch": 0.5583684210526316, + "grad_norm": 0.490234375, + "learning_rate": 9.814035087719299e-05, + "loss": 3.4895, + "step": 10609 + }, + { + "epoch": 0.5584210526315789, + "grad_norm": 0.4921875, + "learning_rate": 9.812865497076024e-05, + "loss": 3.2419, + "step": 10610 + }, + { + "epoch": 0.5584736842105263, + "grad_norm": 0.5234375, + "learning_rate": 9.811695906432749e-05, + "loss": 3.7077, + "step": 10611 + }, + { + "epoch": 0.5585263157894736, + "grad_norm": 0.5546875, + "learning_rate": 9.810526315789475e-05, + "loss": 3.4007, + "step": 10612 + }, + { + "epoch": 0.5585789473684211, + "grad_norm": 0.546875, + "learning_rate": 9.809356725146199e-05, + "loss": 3.5581, + "step": 10613 + }, + { + "epoch": 0.5586315789473684, + "grad_norm": 0.482421875, + "learning_rate": 9.808187134502924e-05, + "loss": 3.2952, + "step": 10614 + }, + { + "epoch": 0.5586842105263158, + "grad_norm": 0.59765625, + "learning_rate": 9.80701754385965e-05, + "loss": 3.4686, + "step": 10615 + }, + { + "epoch": 0.5587368421052632, + "grad_norm": 0.50390625, + "learning_rate": 9.805847953216375e-05, + "loss": 3.5906, + "step": 10616 + }, + { + "epoch": 0.5587894736842105, + "grad_norm": 0.5, + "learning_rate": 9.8046783625731e-05, + "loss": 3.1921, + "step": 10617 + }, + { + "epoch": 0.5588421052631579, + "grad_norm": 0.474609375, + "learning_rate": 9.803508771929825e-05, + "loss": 3.732, + "step": 10618 + }, + { + "epoch": 0.5588947368421052, + "grad_norm": 0.470703125, + "learning_rate": 9.802339181286551e-05, + "loss": 3.2197, + "step": 10619 + }, + { + "epoch": 0.5589473684210526, + "grad_norm": 0.53515625, + "learning_rate": 9.801169590643276e-05, + "loss": 3.1086, + "step": 10620 + }, + { + "epoch": 0.559, + "grad_norm": 0.51953125, + "learning_rate": 9.8e-05, + "loss": 3.4595, + "step": 10621 + }, + { + "epoch": 0.5590526315789474, + "grad_norm": 0.515625, + "learning_rate": 9.798830409356725e-05, + "loss": 3.3423, + "step": 10622 + }, + { + "epoch": 0.5591052631578948, + "grad_norm": 0.5625, + "learning_rate": 9.797660818713451e-05, + "loss": 3.2125, + "step": 10623 + }, + { + "epoch": 0.5591578947368421, + "grad_norm": 0.54296875, + "learning_rate": 9.796491228070176e-05, + "loss": 3.5814, + "step": 10624 + }, + { + "epoch": 0.5592105263157895, + "grad_norm": 0.51171875, + "learning_rate": 9.795321637426901e-05, + "loss": 3.6712, + "step": 10625 + }, + { + "epoch": 0.5592631578947368, + "grad_norm": 0.484375, + "learning_rate": 9.794152046783627e-05, + "loss": 3.3499, + "step": 10626 + }, + { + "epoch": 0.5593157894736842, + "grad_norm": 0.48046875, + "learning_rate": 9.792982456140352e-05, + "loss": 3.3587, + "step": 10627 + }, + { + "epoch": 0.5593684210526316, + "grad_norm": 0.5, + "learning_rate": 9.791812865497077e-05, + "loss": 3.4951, + "step": 10628 + }, + { + "epoch": 0.5594210526315789, + "grad_norm": 0.482421875, + "learning_rate": 9.790643274853801e-05, + "loss": 3.6178, + "step": 10629 + }, + { + "epoch": 0.5594736842105263, + "grad_norm": 1.3984375, + "learning_rate": 9.789473684210527e-05, + "loss": 3.5177, + "step": 10630 + }, + { + "epoch": 0.5595263157894736, + "grad_norm": 0.50390625, + "learning_rate": 9.788304093567252e-05, + "loss": 3.3144, + "step": 10631 + }, + { + "epoch": 0.5595789473684211, + "grad_norm": 0.47265625, + "learning_rate": 9.787134502923977e-05, + "loss": 3.5772, + "step": 10632 + }, + { + "epoch": 0.5596315789473684, + "grad_norm": 0.51171875, + "learning_rate": 9.785964912280702e-05, + "loss": 3.0207, + "step": 10633 + }, + { + "epoch": 0.5596842105263158, + "grad_norm": 0.5, + "learning_rate": 9.784795321637428e-05, + "loss": 3.3106, + "step": 10634 + }, + { + "epoch": 0.5597368421052632, + "grad_norm": 0.46484375, + "learning_rate": 9.783625730994153e-05, + "loss": 3.746, + "step": 10635 + }, + { + "epoch": 0.5597894736842105, + "grad_norm": 0.5078125, + "learning_rate": 9.782456140350877e-05, + "loss": 3.485, + "step": 10636 + }, + { + "epoch": 0.5598421052631579, + "grad_norm": 0.48046875, + "learning_rate": 9.781286549707602e-05, + "loss": 3.7617, + "step": 10637 + }, + { + "epoch": 0.5598947368421052, + "grad_norm": 0.51171875, + "learning_rate": 9.780116959064328e-05, + "loss": 3.5482, + "step": 10638 + }, + { + "epoch": 0.5599473684210526, + "grad_norm": 0.486328125, + "learning_rate": 9.778947368421053e-05, + "loss": 3.6293, + "step": 10639 + }, + { + "epoch": 0.56, + "grad_norm": 0.50390625, + "learning_rate": 9.777777777777778e-05, + "loss": 3.6534, + "step": 10640 + }, + { + "epoch": 0.5600526315789474, + "grad_norm": 0.5078125, + "learning_rate": 9.776608187134504e-05, + "loss": 3.5062, + "step": 10641 + }, + { + "epoch": 0.5601052631578948, + "grad_norm": 0.5078125, + "learning_rate": 9.775438596491229e-05, + "loss": 3.225, + "step": 10642 + }, + { + "epoch": 0.5601578947368421, + "grad_norm": 0.474609375, + "learning_rate": 9.774269005847954e-05, + "loss": 3.1295, + "step": 10643 + }, + { + "epoch": 0.5602105263157895, + "grad_norm": 0.482421875, + "learning_rate": 9.773099415204678e-05, + "loss": 3.4246, + "step": 10644 + }, + { + "epoch": 0.5602631578947368, + "grad_norm": 0.482421875, + "learning_rate": 9.771929824561404e-05, + "loss": 3.526, + "step": 10645 + }, + { + "epoch": 0.5603157894736842, + "grad_norm": 0.48046875, + "learning_rate": 9.770760233918129e-05, + "loss": 3.2015, + "step": 10646 + }, + { + "epoch": 0.5603684210526316, + "grad_norm": 0.5078125, + "learning_rate": 9.769590643274854e-05, + "loss": 3.4884, + "step": 10647 + }, + { + "epoch": 0.5604210526315789, + "grad_norm": 0.78125, + "learning_rate": 9.768421052631579e-05, + "loss": 3.3285, + "step": 10648 + }, + { + "epoch": 0.5604736842105263, + "grad_norm": 0.47265625, + "learning_rate": 9.767251461988305e-05, + "loss": 3.8553, + "step": 10649 + }, + { + "epoch": 0.5605263157894737, + "grad_norm": 0.49609375, + "learning_rate": 9.76608187134503e-05, + "loss": 3.4245, + "step": 10650 + }, + { + "epoch": 0.5605789473684211, + "grad_norm": 0.46875, + "learning_rate": 9.764912280701755e-05, + "loss": 3.2667, + "step": 10651 + }, + { + "epoch": 0.5606315789473684, + "grad_norm": 0.470703125, + "learning_rate": 9.76374269005848e-05, + "loss": 3.4174, + "step": 10652 + }, + { + "epoch": 0.5606842105263158, + "grad_norm": 0.50390625, + "learning_rate": 9.762573099415205e-05, + "loss": 3.5055, + "step": 10653 + }, + { + "epoch": 0.5607368421052632, + "grad_norm": 0.546875, + "learning_rate": 9.76140350877193e-05, + "loss": 3.5775, + "step": 10654 + }, + { + "epoch": 0.5607894736842105, + "grad_norm": 0.47265625, + "learning_rate": 9.760233918128655e-05, + "loss": 3.7627, + "step": 10655 + }, + { + "epoch": 0.5608421052631579, + "grad_norm": 0.54296875, + "learning_rate": 9.759064327485381e-05, + "loss": 3.2216, + "step": 10656 + }, + { + "epoch": 0.5608947368421052, + "grad_norm": 0.578125, + "learning_rate": 9.757894736842106e-05, + "loss": 3.2286, + "step": 10657 + }, + { + "epoch": 0.5609473684210526, + "grad_norm": 0.5078125, + "learning_rate": 9.756725146198831e-05, + "loss": 3.1683, + "step": 10658 + }, + { + "epoch": 0.561, + "grad_norm": 0.474609375, + "learning_rate": 9.755555555555555e-05, + "loss": 3.0944, + "step": 10659 + }, + { + "epoch": 0.5610526315789474, + "grad_norm": 0.5, + "learning_rate": 9.754385964912281e-05, + "loss": 3.6114, + "step": 10660 + }, + { + "epoch": 0.5611052631578948, + "grad_norm": 0.51171875, + "learning_rate": 9.753216374269006e-05, + "loss": 3.0925, + "step": 10661 + }, + { + "epoch": 0.5611578947368421, + "grad_norm": 0.498046875, + "learning_rate": 9.752046783625731e-05, + "loss": 3.6857, + "step": 10662 + }, + { + "epoch": 0.5612105263157895, + "grad_norm": 0.484375, + "learning_rate": 9.750877192982456e-05, + "loss": 2.9677, + "step": 10663 + }, + { + "epoch": 0.5612631578947368, + "grad_norm": 0.4921875, + "learning_rate": 9.749707602339182e-05, + "loss": 3.3205, + "step": 10664 + }, + { + "epoch": 0.5613157894736842, + "grad_norm": 0.453125, + "learning_rate": 9.748538011695907e-05, + "loss": 3.4138, + "step": 10665 + }, + { + "epoch": 0.5613684210526316, + "grad_norm": 0.5, + "learning_rate": 9.747368421052632e-05, + "loss": 3.7242, + "step": 10666 + }, + { + "epoch": 0.5614210526315789, + "grad_norm": 0.51171875, + "learning_rate": 9.746198830409357e-05, + "loss": 3.2672, + "step": 10667 + }, + { + "epoch": 0.5614736842105263, + "grad_norm": 0.44921875, + "learning_rate": 9.745029239766082e-05, + "loss": 3.5599, + "step": 10668 + }, + { + "epoch": 0.5615263157894737, + "grad_norm": 0.47265625, + "learning_rate": 9.743859649122807e-05, + "loss": 3.5567, + "step": 10669 + }, + { + "epoch": 0.5615789473684211, + "grad_norm": 0.498046875, + "learning_rate": 9.742690058479532e-05, + "loss": 3.1964, + "step": 10670 + }, + { + "epoch": 0.5616315789473684, + "grad_norm": 0.50390625, + "learning_rate": 9.741520467836258e-05, + "loss": 3.4952, + "step": 10671 + }, + { + "epoch": 0.5616842105263158, + "grad_norm": 0.60546875, + "learning_rate": 9.740350877192983e-05, + "loss": 3.1694, + "step": 10672 + }, + { + "epoch": 0.5617368421052632, + "grad_norm": 1.5078125, + "learning_rate": 9.739181286549708e-05, + "loss": 3.4343, + "step": 10673 + }, + { + "epoch": 0.5617894736842105, + "grad_norm": 0.52734375, + "learning_rate": 9.738011695906433e-05, + "loss": 3.4511, + "step": 10674 + }, + { + "epoch": 0.5618421052631579, + "grad_norm": 0.490234375, + "learning_rate": 9.736842105263158e-05, + "loss": 3.3654, + "step": 10675 + }, + { + "epoch": 0.5618947368421052, + "grad_norm": 0.5625, + "learning_rate": 9.735672514619883e-05, + "loss": 3.1631, + "step": 10676 + }, + { + "epoch": 0.5619473684210526, + "grad_norm": 0.498046875, + "learning_rate": 9.734502923976608e-05, + "loss": 2.9858, + "step": 10677 + }, + { + "epoch": 0.562, + "grad_norm": 0.46875, + "learning_rate": 9.733333333333335e-05, + "loss": 3.6004, + "step": 10678 + }, + { + "epoch": 0.5620526315789474, + "grad_norm": 0.48046875, + "learning_rate": 9.73216374269006e-05, + "loss": 3.0542, + "step": 10679 + }, + { + "epoch": 0.5621052631578948, + "grad_norm": 0.466796875, + "learning_rate": 9.730994152046784e-05, + "loss": 3.4172, + "step": 10680 + }, + { + "epoch": 0.5621578947368421, + "grad_norm": 0.48828125, + "learning_rate": 9.72982456140351e-05, + "loss": 3.5533, + "step": 10681 + }, + { + "epoch": 0.5622105263157895, + "grad_norm": 0.47265625, + "learning_rate": 9.728654970760234e-05, + "loss": 3.4409, + "step": 10682 + }, + { + "epoch": 0.5622631578947368, + "grad_norm": 0.4921875, + "learning_rate": 9.72748538011696e-05, + "loss": 3.2307, + "step": 10683 + }, + { + "epoch": 0.5623157894736842, + "grad_norm": 0.45703125, + "learning_rate": 9.726315789473684e-05, + "loss": 3.436, + "step": 10684 + }, + { + "epoch": 0.5623684210526316, + "grad_norm": 0.45703125, + "learning_rate": 9.725146198830409e-05, + "loss": 3.6553, + "step": 10685 + }, + { + "epoch": 0.5624210526315789, + "grad_norm": 0.50390625, + "learning_rate": 9.723976608187136e-05, + "loss": 3.2536, + "step": 10686 + }, + { + "epoch": 0.5624736842105263, + "grad_norm": 0.45703125, + "learning_rate": 9.72280701754386e-05, + "loss": 3.6358, + "step": 10687 + }, + { + "epoch": 0.5625263157894737, + "grad_norm": 0.478515625, + "learning_rate": 9.721637426900586e-05, + "loss": 3.5078, + "step": 10688 + }, + { + "epoch": 0.5625789473684211, + "grad_norm": 0.470703125, + "learning_rate": 9.72046783625731e-05, + "loss": 3.1612, + "step": 10689 + }, + { + "epoch": 0.5626315789473684, + "grad_norm": 0.51171875, + "learning_rate": 9.719298245614035e-05, + "loss": 3.3426, + "step": 10690 + }, + { + "epoch": 0.5626842105263158, + "grad_norm": 0.490234375, + "learning_rate": 9.71812865497076e-05, + "loss": 3.5271, + "step": 10691 + }, + { + "epoch": 0.5627368421052632, + "grad_norm": 0.49609375, + "learning_rate": 9.716959064327485e-05, + "loss": 3.5943, + "step": 10692 + }, + { + "epoch": 0.5627894736842105, + "grad_norm": 0.50390625, + "learning_rate": 9.715789473684212e-05, + "loss": 3.5057, + "step": 10693 + }, + { + "epoch": 0.5628421052631579, + "grad_norm": 0.490234375, + "learning_rate": 9.714619883040937e-05, + "loss": 3.233, + "step": 10694 + }, + { + "epoch": 0.5628947368421052, + "grad_norm": 0.4765625, + "learning_rate": 9.713450292397662e-05, + "loss": 3.3273, + "step": 10695 + }, + { + "epoch": 0.5629473684210526, + "grad_norm": 0.50390625, + "learning_rate": 9.712280701754387e-05, + "loss": 3.4213, + "step": 10696 + }, + { + "epoch": 0.563, + "grad_norm": 0.466796875, + "learning_rate": 9.711111111111111e-05, + "loss": 3.3015, + "step": 10697 + }, + { + "epoch": 0.5630526315789474, + "grad_norm": 0.486328125, + "learning_rate": 9.709941520467836e-05, + "loss": 3.3202, + "step": 10698 + }, + { + "epoch": 0.5631052631578948, + "grad_norm": 0.48046875, + "learning_rate": 9.708771929824561e-05, + "loss": 3.3895, + "step": 10699 + }, + { + "epoch": 0.5631578947368421, + "grad_norm": 0.48828125, + "learning_rate": 9.707602339181286e-05, + "loss": 3.0751, + "step": 10700 + }, + { + "epoch": 0.5632105263157895, + "grad_norm": 0.48046875, + "learning_rate": 9.706432748538013e-05, + "loss": 3.4334, + "step": 10701 + }, + { + "epoch": 0.5632631578947368, + "grad_norm": 0.47265625, + "learning_rate": 9.705263157894738e-05, + "loss": 3.3354, + "step": 10702 + }, + { + "epoch": 0.5633157894736842, + "grad_norm": 0.453125, + "learning_rate": 9.704093567251463e-05, + "loss": 3.2116, + "step": 10703 + }, + { + "epoch": 0.5633684210526316, + "grad_norm": 0.455078125, + "learning_rate": 9.702923976608188e-05, + "loss": 3.4076, + "step": 10704 + }, + { + "epoch": 0.5634210526315789, + "grad_norm": 0.49609375, + "learning_rate": 9.701754385964913e-05, + "loss": 3.7516, + "step": 10705 + }, + { + "epoch": 0.5634736842105263, + "grad_norm": 0.49609375, + "learning_rate": 9.700584795321637e-05, + "loss": 3.5627, + "step": 10706 + }, + { + "epoch": 0.5635263157894737, + "grad_norm": 0.478515625, + "learning_rate": 9.699415204678362e-05, + "loss": 3.2247, + "step": 10707 + }, + { + "epoch": 0.5635789473684211, + "grad_norm": 0.5, + "learning_rate": 9.698245614035089e-05, + "loss": 2.9017, + "step": 10708 + }, + { + "epoch": 0.5636315789473684, + "grad_norm": 0.466796875, + "learning_rate": 9.697076023391814e-05, + "loss": 3.29, + "step": 10709 + }, + { + "epoch": 0.5636842105263158, + "grad_norm": 0.474609375, + "learning_rate": 9.695906432748539e-05, + "loss": 3.4852, + "step": 10710 + }, + { + "epoch": 0.5637368421052632, + "grad_norm": 0.486328125, + "learning_rate": 9.694736842105264e-05, + "loss": 3.0297, + "step": 10711 + }, + { + "epoch": 0.5637894736842105, + "grad_norm": 0.515625, + "learning_rate": 9.693567251461989e-05, + "loss": 3.462, + "step": 10712 + }, + { + "epoch": 0.5638421052631579, + "grad_norm": 0.486328125, + "learning_rate": 9.692397660818714e-05, + "loss": 3.4636, + "step": 10713 + }, + { + "epoch": 0.5638947368421052, + "grad_norm": 0.478515625, + "learning_rate": 9.691228070175439e-05, + "loss": 3.639, + "step": 10714 + }, + { + "epoch": 0.5639473684210526, + "grad_norm": 0.4921875, + "learning_rate": 9.690058479532165e-05, + "loss": 3.6704, + "step": 10715 + }, + { + "epoch": 0.564, + "grad_norm": 0.453125, + "learning_rate": 9.68888888888889e-05, + "loss": 3.3473, + "step": 10716 + }, + { + "epoch": 0.5640526315789474, + "grad_norm": 0.470703125, + "learning_rate": 9.687719298245615e-05, + "loss": 3.7179, + "step": 10717 + }, + { + "epoch": 0.5641052631578948, + "grad_norm": 0.68359375, + "learning_rate": 9.68654970760234e-05, + "loss": 3.3812, + "step": 10718 + }, + { + "epoch": 0.5641578947368421, + "grad_norm": 0.57421875, + "learning_rate": 9.685380116959065e-05, + "loss": 3.5887, + "step": 10719 + }, + { + "epoch": 0.5642105263157895, + "grad_norm": 0.52734375, + "learning_rate": 9.68421052631579e-05, + "loss": 3.4362, + "step": 10720 + }, + { + "epoch": 0.5642631578947368, + "grad_norm": 0.498046875, + "learning_rate": 9.683040935672515e-05, + "loss": 3.7444, + "step": 10721 + }, + { + "epoch": 0.5643157894736842, + "grad_norm": 0.5390625, + "learning_rate": 9.68187134502924e-05, + "loss": 3.0256, + "step": 10722 + }, + { + "epoch": 0.5643684210526316, + "grad_norm": 0.50390625, + "learning_rate": 9.680701754385966e-05, + "loss": 3.6131, + "step": 10723 + }, + { + "epoch": 0.5644210526315789, + "grad_norm": 0.5859375, + "learning_rate": 9.679532163742691e-05, + "loss": 3.1861, + "step": 10724 + }, + { + "epoch": 0.5644736842105263, + "grad_norm": 0.47265625, + "learning_rate": 9.678362573099416e-05, + "loss": 3.5754, + "step": 10725 + }, + { + "epoch": 0.5645263157894737, + "grad_norm": 0.490234375, + "learning_rate": 9.677192982456141e-05, + "loss": 3.1261, + "step": 10726 + }, + { + "epoch": 0.5645789473684211, + "grad_norm": 0.470703125, + "learning_rate": 9.676023391812866e-05, + "loss": 3.5106, + "step": 10727 + }, + { + "epoch": 0.5646315789473684, + "grad_norm": 0.55859375, + "learning_rate": 9.67485380116959e-05, + "loss": 3.0115, + "step": 10728 + }, + { + "epoch": 0.5646842105263158, + "grad_norm": 0.7890625, + "learning_rate": 9.673684210526316e-05, + "loss": 3.4535, + "step": 10729 + }, + { + "epoch": 0.5647368421052632, + "grad_norm": 0.4921875, + "learning_rate": 9.672514619883042e-05, + "loss": 3.6336, + "step": 10730 + }, + { + "epoch": 0.5647894736842105, + "grad_norm": 0.49609375, + "learning_rate": 9.671345029239767e-05, + "loss": 3.6264, + "step": 10731 + }, + { + "epoch": 0.5648421052631579, + "grad_norm": 0.515625, + "learning_rate": 9.670175438596492e-05, + "loss": 3.3237, + "step": 10732 + }, + { + "epoch": 0.5648947368421052, + "grad_norm": 0.478515625, + "learning_rate": 9.669005847953217e-05, + "loss": 3.4208, + "step": 10733 + }, + { + "epoch": 0.5649473684210526, + "grad_norm": 0.56640625, + "learning_rate": 9.667836257309942e-05, + "loss": 3.57, + "step": 10734 + }, + { + "epoch": 0.565, + "grad_norm": 0.484375, + "learning_rate": 9.666666666666667e-05, + "loss": 3.3931, + "step": 10735 + }, + { + "epoch": 0.5650526315789474, + "grad_norm": 0.47265625, + "learning_rate": 9.665497076023392e-05, + "loss": 3.1919, + "step": 10736 + }, + { + "epoch": 0.5651052631578948, + "grad_norm": 0.478515625, + "learning_rate": 9.664327485380117e-05, + "loss": 3.567, + "step": 10737 + }, + { + "epoch": 0.5651578947368421, + "grad_norm": 0.490234375, + "learning_rate": 9.663157894736843e-05, + "loss": 3.2004, + "step": 10738 + }, + { + "epoch": 0.5652105263157895, + "grad_norm": 0.50390625, + "learning_rate": 9.661988304093568e-05, + "loss": 3.0197, + "step": 10739 + }, + { + "epoch": 0.5652631578947368, + "grad_norm": 0.470703125, + "learning_rate": 9.660818713450293e-05, + "loss": 2.8512, + "step": 10740 + }, + { + "epoch": 0.5653157894736842, + "grad_norm": 0.51953125, + "learning_rate": 9.659649122807018e-05, + "loss": 3.6446, + "step": 10741 + }, + { + "epoch": 0.5653684210526316, + "grad_norm": 0.4765625, + "learning_rate": 9.658479532163743e-05, + "loss": 3.3684, + "step": 10742 + }, + { + "epoch": 0.5654210526315789, + "grad_norm": 0.625, + "learning_rate": 9.657309941520468e-05, + "loss": 3.6115, + "step": 10743 + }, + { + "epoch": 0.5654736842105264, + "grad_norm": 0.486328125, + "learning_rate": 9.656140350877193e-05, + "loss": 3.5248, + "step": 10744 + }, + { + "epoch": 0.5655263157894737, + "grad_norm": 0.453125, + "learning_rate": 9.654970760233919e-05, + "loss": 3.4744, + "step": 10745 + }, + { + "epoch": 0.5655789473684211, + "grad_norm": 0.482421875, + "learning_rate": 9.653801169590644e-05, + "loss": 3.545, + "step": 10746 + }, + { + "epoch": 0.5656315789473684, + "grad_norm": 0.4921875, + "learning_rate": 9.652631578947369e-05, + "loss": 3.6395, + "step": 10747 + }, + { + "epoch": 0.5656842105263158, + "grad_norm": 0.55078125, + "learning_rate": 9.651461988304094e-05, + "loss": 3.2312, + "step": 10748 + }, + { + "epoch": 0.5657368421052632, + "grad_norm": 0.498046875, + "learning_rate": 9.650292397660819e-05, + "loss": 3.0982, + "step": 10749 + }, + { + "epoch": 0.5657894736842105, + "grad_norm": 0.50390625, + "learning_rate": 9.649122807017544e-05, + "loss": 3.3055, + "step": 10750 + }, + { + "epoch": 0.5658421052631579, + "grad_norm": 0.546875, + "learning_rate": 9.647953216374269e-05, + "loss": 3.169, + "step": 10751 + }, + { + "epoch": 0.5658947368421052, + "grad_norm": 0.486328125, + "learning_rate": 9.646783625730995e-05, + "loss": 3.4637, + "step": 10752 + }, + { + "epoch": 0.5659473684210526, + "grad_norm": 0.482421875, + "learning_rate": 9.64561403508772e-05, + "loss": 3.6604, + "step": 10753 + }, + { + "epoch": 0.566, + "grad_norm": 0.48828125, + "learning_rate": 9.644444444444445e-05, + "loss": 3.4924, + "step": 10754 + }, + { + "epoch": 0.5660526315789474, + "grad_norm": 0.5, + "learning_rate": 9.64327485380117e-05, + "loss": 3.116, + "step": 10755 + }, + { + "epoch": 0.5661052631578948, + "grad_norm": 0.54296875, + "learning_rate": 9.642105263157896e-05, + "loss": 3.5226, + "step": 10756 + }, + { + "epoch": 0.5661578947368421, + "grad_norm": 0.494140625, + "learning_rate": 9.64093567251462e-05, + "loss": 3.2871, + "step": 10757 + }, + { + "epoch": 0.5662105263157895, + "grad_norm": 0.5078125, + "learning_rate": 9.639766081871345e-05, + "loss": 3.5298, + "step": 10758 + }, + { + "epoch": 0.5662631578947368, + "grad_norm": 0.455078125, + "learning_rate": 9.63859649122807e-05, + "loss": 3.5995, + "step": 10759 + }, + { + "epoch": 0.5663157894736842, + "grad_norm": 0.458984375, + "learning_rate": 9.637426900584796e-05, + "loss": 2.8794, + "step": 10760 + }, + { + "epoch": 0.5663684210526316, + "grad_norm": 0.50390625, + "learning_rate": 9.636257309941521e-05, + "loss": 3.3108, + "step": 10761 + }, + { + "epoch": 0.5664210526315789, + "grad_norm": 0.5390625, + "learning_rate": 9.635087719298246e-05, + "loss": 3.4201, + "step": 10762 + }, + { + "epoch": 0.5664736842105264, + "grad_norm": 0.44921875, + "learning_rate": 9.633918128654971e-05, + "loss": 3.7393, + "step": 10763 + }, + { + "epoch": 0.5665263157894737, + "grad_norm": 0.51171875, + "learning_rate": 9.632748538011696e-05, + "loss": 3.0689, + "step": 10764 + }, + { + "epoch": 0.5665789473684211, + "grad_norm": 0.5390625, + "learning_rate": 9.631578947368421e-05, + "loss": 3.1266, + "step": 10765 + }, + { + "epoch": 0.5666315789473684, + "grad_norm": 0.4921875, + "learning_rate": 9.630409356725146e-05, + "loss": 3.2063, + "step": 10766 + }, + { + "epoch": 0.5666842105263158, + "grad_norm": 0.46484375, + "learning_rate": 9.629239766081872e-05, + "loss": 3.0013, + "step": 10767 + }, + { + "epoch": 0.5667368421052632, + "grad_norm": 0.52734375, + "learning_rate": 9.628070175438597e-05, + "loss": 3.451, + "step": 10768 + }, + { + "epoch": 0.5667894736842105, + "grad_norm": 0.4921875, + "learning_rate": 9.626900584795322e-05, + "loss": 3.0584, + "step": 10769 + }, + { + "epoch": 0.5668421052631579, + "grad_norm": 0.52734375, + "learning_rate": 9.625730994152047e-05, + "loss": 3.1594, + "step": 10770 + }, + { + "epoch": 0.5668947368421052, + "grad_norm": 1.15625, + "learning_rate": 9.624561403508773e-05, + "loss": 3.4756, + "step": 10771 + }, + { + "epoch": 0.5669473684210526, + "grad_norm": 0.482421875, + "learning_rate": 9.623391812865497e-05, + "loss": 3.6896, + "step": 10772 + }, + { + "epoch": 0.567, + "grad_norm": 0.474609375, + "learning_rate": 9.622222222222222e-05, + "loss": 3.8054, + "step": 10773 + }, + { + "epoch": 0.5670526315789474, + "grad_norm": 0.51171875, + "learning_rate": 9.621052631578947e-05, + "loss": 3.8822, + "step": 10774 + }, + { + "epoch": 0.5671052631578948, + "grad_norm": 0.48046875, + "learning_rate": 9.619883040935673e-05, + "loss": 2.8951, + "step": 10775 + }, + { + "epoch": 0.5671578947368421, + "grad_norm": 0.51953125, + "learning_rate": 9.618713450292398e-05, + "loss": 3.2227, + "step": 10776 + }, + { + "epoch": 0.5672105263157895, + "grad_norm": 0.474609375, + "learning_rate": 9.617543859649123e-05, + "loss": 3.2865, + "step": 10777 + }, + { + "epoch": 0.5672631578947368, + "grad_norm": 0.49609375, + "learning_rate": 9.61637426900585e-05, + "loss": 3.6774, + "step": 10778 + }, + { + "epoch": 0.5673157894736842, + "grad_norm": 0.53125, + "learning_rate": 9.615204678362574e-05, + "loss": 3.4287, + "step": 10779 + }, + { + "epoch": 0.5673684210526316, + "grad_norm": 0.5625, + "learning_rate": 9.614035087719298e-05, + "loss": 3.3785, + "step": 10780 + }, + { + "epoch": 0.5674210526315789, + "grad_norm": 0.52734375, + "learning_rate": 9.612865497076023e-05, + "loss": 2.9996, + "step": 10781 + }, + { + "epoch": 0.5674736842105264, + "grad_norm": 0.47265625, + "learning_rate": 9.611695906432749e-05, + "loss": 3.5983, + "step": 10782 + }, + { + "epoch": 0.5675263157894737, + "grad_norm": 0.484375, + "learning_rate": 9.610526315789474e-05, + "loss": 3.8872, + "step": 10783 + }, + { + "epoch": 0.5675789473684211, + "grad_norm": 0.458984375, + "learning_rate": 9.609356725146199e-05, + "loss": 3.4357, + "step": 10784 + }, + { + "epoch": 0.5676315789473684, + "grad_norm": 0.494140625, + "learning_rate": 9.608187134502924e-05, + "loss": 3.369, + "step": 10785 + }, + { + "epoch": 0.5676842105263158, + "grad_norm": 0.48828125, + "learning_rate": 9.60701754385965e-05, + "loss": 3.0471, + "step": 10786 + }, + { + "epoch": 0.5677368421052632, + "grad_norm": 0.4609375, + "learning_rate": 9.605847953216374e-05, + "loss": 3.5205, + "step": 10787 + }, + { + "epoch": 0.5677894736842105, + "grad_norm": 0.498046875, + "learning_rate": 9.604678362573099e-05, + "loss": 3.3775, + "step": 10788 + }, + { + "epoch": 0.5678421052631579, + "grad_norm": 0.49609375, + "learning_rate": 9.603508771929825e-05, + "loss": 3.4241, + "step": 10789 + }, + { + "epoch": 0.5678947368421052, + "grad_norm": 0.51171875, + "learning_rate": 9.60233918128655e-05, + "loss": 3.5597, + "step": 10790 + }, + { + "epoch": 0.5679473684210526, + "grad_norm": 0.5, + "learning_rate": 9.601169590643275e-05, + "loss": 3.3752, + "step": 10791 + }, + { + "epoch": 0.568, + "grad_norm": 0.515625, + "learning_rate": 9.6e-05, + "loss": 2.6975, + "step": 10792 + }, + { + "epoch": 0.5680526315789474, + "grad_norm": 0.498046875, + "learning_rate": 9.598830409356727e-05, + "loss": 3.7779, + "step": 10793 + }, + { + "epoch": 0.5681052631578948, + "grad_norm": 0.53515625, + "learning_rate": 9.597660818713452e-05, + "loss": 3.8164, + "step": 10794 + }, + { + "epoch": 0.5681578947368421, + "grad_norm": 0.4921875, + "learning_rate": 9.596491228070175e-05, + "loss": 3.5355, + "step": 10795 + }, + { + "epoch": 0.5682105263157895, + "grad_norm": 0.4921875, + "learning_rate": 9.5953216374269e-05, + "loss": 3.5938, + "step": 10796 + }, + { + "epoch": 0.5682631578947368, + "grad_norm": 0.51171875, + "learning_rate": 9.594152046783626e-05, + "loss": 3.4929, + "step": 10797 + }, + { + "epoch": 0.5683157894736842, + "grad_norm": 0.5234375, + "learning_rate": 9.592982456140351e-05, + "loss": 2.8875, + "step": 10798 + }, + { + "epoch": 0.5683684210526316, + "grad_norm": 0.478515625, + "learning_rate": 9.591812865497076e-05, + "loss": 3.6568, + "step": 10799 + }, + { + "epoch": 0.5684210526315789, + "grad_norm": 0.47265625, + "learning_rate": 9.590643274853801e-05, + "loss": 3.085, + "step": 10800 + }, + { + "epoch": 0.5684736842105264, + "grad_norm": 0.50390625, + "learning_rate": 9.589473684210528e-05, + "loss": 3.5301, + "step": 10801 + }, + { + "epoch": 0.5685263157894737, + "grad_norm": 0.474609375, + "learning_rate": 9.588304093567253e-05, + "loss": 3.4265, + "step": 10802 + }, + { + "epoch": 0.5685789473684211, + "grad_norm": 0.48046875, + "learning_rate": 9.587134502923976e-05, + "loss": 3.4942, + "step": 10803 + }, + { + "epoch": 0.5686315789473684, + "grad_norm": 0.482421875, + "learning_rate": 9.585964912280702e-05, + "loss": 3.4849, + "step": 10804 + }, + { + "epoch": 0.5686842105263158, + "grad_norm": 0.515625, + "learning_rate": 9.584795321637427e-05, + "loss": 2.8794, + "step": 10805 + }, + { + "epoch": 0.5687368421052632, + "grad_norm": 0.50390625, + "learning_rate": 9.583625730994152e-05, + "loss": 2.99, + "step": 10806 + }, + { + "epoch": 0.5687894736842105, + "grad_norm": 0.470703125, + "learning_rate": 9.582456140350877e-05, + "loss": 3.4433, + "step": 10807 + }, + { + "epoch": 0.5688421052631579, + "grad_norm": 0.515625, + "learning_rate": 9.581286549707604e-05, + "loss": 3.1304, + "step": 10808 + }, + { + "epoch": 0.5688947368421052, + "grad_norm": 0.5, + "learning_rate": 9.580116959064329e-05, + "loss": 3.7476, + "step": 10809 + }, + { + "epoch": 0.5689473684210526, + "grad_norm": 0.59375, + "learning_rate": 9.578947368421052e-05, + "loss": 3.0438, + "step": 10810 + }, + { + "epoch": 0.569, + "grad_norm": 0.494140625, + "learning_rate": 9.577777777777777e-05, + "loss": 3.3926, + "step": 10811 + }, + { + "epoch": 0.5690526315789474, + "grad_norm": 0.50390625, + "learning_rate": 9.576608187134503e-05, + "loss": 3.7054, + "step": 10812 + }, + { + "epoch": 0.5691052631578948, + "grad_norm": 0.50390625, + "learning_rate": 9.575438596491228e-05, + "loss": 3.0683, + "step": 10813 + }, + { + "epoch": 0.5691578947368421, + "grad_norm": 0.51953125, + "learning_rate": 9.574269005847953e-05, + "loss": 3.1508, + "step": 10814 + }, + { + "epoch": 0.5692105263157895, + "grad_norm": 0.47265625, + "learning_rate": 9.57309941520468e-05, + "loss": 3.3673, + "step": 10815 + }, + { + "epoch": 0.5692631578947368, + "grad_norm": 0.478515625, + "learning_rate": 9.571929824561405e-05, + "loss": 3.6281, + "step": 10816 + }, + { + "epoch": 0.5693157894736842, + "grad_norm": 0.462890625, + "learning_rate": 9.57076023391813e-05, + "loss": 3.673, + "step": 10817 + }, + { + "epoch": 0.5693684210526315, + "grad_norm": 0.5, + "learning_rate": 9.569590643274853e-05, + "loss": 3.3059, + "step": 10818 + }, + { + "epoch": 0.5694210526315789, + "grad_norm": 0.51953125, + "learning_rate": 9.56842105263158e-05, + "loss": 3.1303, + "step": 10819 + }, + { + "epoch": 0.5694736842105264, + "grad_norm": 0.462890625, + "learning_rate": 9.567251461988305e-05, + "loss": 3.3496, + "step": 10820 + }, + { + "epoch": 0.5695263157894737, + "grad_norm": 0.51171875, + "learning_rate": 9.56608187134503e-05, + "loss": 3.7097, + "step": 10821 + }, + { + "epoch": 0.5695789473684211, + "grad_norm": 0.53515625, + "learning_rate": 9.564912280701754e-05, + "loss": 3.8529, + "step": 10822 + }, + { + "epoch": 0.5696315789473684, + "grad_norm": 0.578125, + "learning_rate": 9.563742690058481e-05, + "loss": 3.3395, + "step": 10823 + }, + { + "epoch": 0.5696842105263158, + "grad_norm": 0.5078125, + "learning_rate": 9.562573099415206e-05, + "loss": 3.6213, + "step": 10824 + }, + { + "epoch": 0.5697368421052632, + "grad_norm": 0.57421875, + "learning_rate": 9.56140350877193e-05, + "loss": 3.4222, + "step": 10825 + }, + { + "epoch": 0.5697894736842105, + "grad_norm": 0.55078125, + "learning_rate": 9.560233918128654e-05, + "loss": 3.1262, + "step": 10826 + }, + { + "epoch": 0.5698421052631579, + "grad_norm": 0.490234375, + "learning_rate": 9.55906432748538e-05, + "loss": 3.0046, + "step": 10827 + }, + { + "epoch": 0.5698947368421052, + "grad_norm": 0.470703125, + "learning_rate": 9.557894736842106e-05, + "loss": 3.4016, + "step": 10828 + }, + { + "epoch": 0.5699473684210526, + "grad_norm": 0.4921875, + "learning_rate": 9.55672514619883e-05, + "loss": 3.6013, + "step": 10829 + }, + { + "epoch": 0.57, + "grad_norm": 0.49609375, + "learning_rate": 9.555555555555557e-05, + "loss": 3.2119, + "step": 10830 + }, + { + "epoch": 0.5700526315789474, + "grad_norm": 0.5, + "learning_rate": 9.554385964912282e-05, + "loss": 3.2098, + "step": 10831 + }, + { + "epoch": 0.5701052631578948, + "grad_norm": 0.48828125, + "learning_rate": 9.553216374269007e-05, + "loss": 3.3886, + "step": 10832 + }, + { + "epoch": 0.5701578947368421, + "grad_norm": 0.494140625, + "learning_rate": 9.55204678362573e-05, + "loss": 3.1488, + "step": 10833 + }, + { + "epoch": 0.5702105263157895, + "grad_norm": 0.49609375, + "learning_rate": 9.550877192982457e-05, + "loss": 3.4102, + "step": 10834 + }, + { + "epoch": 0.5702631578947368, + "grad_norm": 0.48046875, + "learning_rate": 9.549707602339182e-05, + "loss": 3.3712, + "step": 10835 + }, + { + "epoch": 0.5703157894736842, + "grad_norm": 0.490234375, + "learning_rate": 9.548538011695907e-05, + "loss": 3.1466, + "step": 10836 + }, + { + "epoch": 0.5703684210526315, + "grad_norm": 0.5078125, + "learning_rate": 9.547368421052632e-05, + "loss": 3.1279, + "step": 10837 + }, + { + "epoch": 0.5704210526315789, + "grad_norm": 0.4609375, + "learning_rate": 9.546198830409358e-05, + "loss": 3.2617, + "step": 10838 + }, + { + "epoch": 0.5704736842105264, + "grad_norm": 0.52734375, + "learning_rate": 9.545029239766083e-05, + "loss": 3.4585, + "step": 10839 + }, + { + "epoch": 0.5705263157894737, + "grad_norm": 0.4921875, + "learning_rate": 9.543859649122808e-05, + "loss": 3.1392, + "step": 10840 + }, + { + "epoch": 0.5705789473684211, + "grad_norm": 0.490234375, + "learning_rate": 9.542690058479533e-05, + "loss": 3.4386, + "step": 10841 + }, + { + "epoch": 0.5706315789473684, + "grad_norm": 0.4921875, + "learning_rate": 9.541520467836258e-05, + "loss": 3.8445, + "step": 10842 + }, + { + "epoch": 0.5706842105263158, + "grad_norm": 0.48828125, + "learning_rate": 9.540350877192983e-05, + "loss": 3.6133, + "step": 10843 + }, + { + "epoch": 0.5707368421052632, + "grad_norm": 0.482421875, + "learning_rate": 9.539181286549708e-05, + "loss": 3.6715, + "step": 10844 + }, + { + "epoch": 0.5707894736842105, + "grad_norm": 0.5234375, + "learning_rate": 9.538011695906434e-05, + "loss": 3.4038, + "step": 10845 + }, + { + "epoch": 0.5708421052631579, + "grad_norm": 0.484375, + "learning_rate": 9.536842105263159e-05, + "loss": 3.5761, + "step": 10846 + }, + { + "epoch": 0.5708947368421052, + "grad_norm": 0.49609375, + "learning_rate": 9.535672514619884e-05, + "loss": 3.2308, + "step": 10847 + }, + { + "epoch": 0.5709473684210526, + "grad_norm": 0.45703125, + "learning_rate": 9.534502923976609e-05, + "loss": 3.4506, + "step": 10848 + }, + { + "epoch": 0.571, + "grad_norm": 0.482421875, + "learning_rate": 9.533333333333334e-05, + "loss": 3.4885, + "step": 10849 + }, + { + "epoch": 0.5710526315789474, + "grad_norm": 0.484375, + "learning_rate": 9.532163742690059e-05, + "loss": 3.739, + "step": 10850 + }, + { + "epoch": 0.5711052631578948, + "grad_norm": 0.453125, + "learning_rate": 9.530994152046784e-05, + "loss": 2.9177, + "step": 10851 + }, + { + "epoch": 0.5711578947368421, + "grad_norm": 0.470703125, + "learning_rate": 9.52982456140351e-05, + "loss": 3.2804, + "step": 10852 + }, + { + "epoch": 0.5712105263157895, + "grad_norm": 0.478515625, + "learning_rate": 9.528654970760235e-05, + "loss": 3.541, + "step": 10853 + }, + { + "epoch": 0.5712631578947368, + "grad_norm": 0.498046875, + "learning_rate": 9.52748538011696e-05, + "loss": 3.3253, + "step": 10854 + }, + { + "epoch": 0.5713157894736842, + "grad_norm": 0.48828125, + "learning_rate": 9.526315789473685e-05, + "loss": 3.2473, + "step": 10855 + }, + { + "epoch": 0.5713684210526315, + "grad_norm": 0.5, + "learning_rate": 9.52514619883041e-05, + "loss": 3.7389, + "step": 10856 + }, + { + "epoch": 0.5714210526315789, + "grad_norm": 0.49609375, + "learning_rate": 9.523976608187135e-05, + "loss": 3.2181, + "step": 10857 + }, + { + "epoch": 0.5714736842105264, + "grad_norm": 0.470703125, + "learning_rate": 9.52280701754386e-05, + "loss": 3.5085, + "step": 10858 + }, + { + "epoch": 0.5715263157894737, + "grad_norm": 0.45703125, + "learning_rate": 9.521637426900585e-05, + "loss": 3.6355, + "step": 10859 + }, + { + "epoch": 0.5715789473684211, + "grad_norm": 0.5078125, + "learning_rate": 9.520467836257311e-05, + "loss": 3.2358, + "step": 10860 + }, + { + "epoch": 0.5716315789473684, + "grad_norm": 0.4921875, + "learning_rate": 9.519298245614036e-05, + "loss": 3.8275, + "step": 10861 + }, + { + "epoch": 0.5716842105263158, + "grad_norm": 0.52734375, + "learning_rate": 9.518128654970761e-05, + "loss": 2.5639, + "step": 10862 + }, + { + "epoch": 0.5717368421052632, + "grad_norm": 0.4609375, + "learning_rate": 9.516959064327486e-05, + "loss": 3.6865, + "step": 10863 + }, + { + "epoch": 0.5717894736842105, + "grad_norm": 0.498046875, + "learning_rate": 9.515789473684211e-05, + "loss": 3.2764, + "step": 10864 + }, + { + "epoch": 0.5718421052631579, + "grad_norm": 0.46875, + "learning_rate": 9.514619883040936e-05, + "loss": 3.6596, + "step": 10865 + }, + { + "epoch": 0.5718947368421052, + "grad_norm": 0.470703125, + "learning_rate": 9.513450292397661e-05, + "loss": 3.4763, + "step": 10866 + }, + { + "epoch": 0.5719473684210526, + "grad_norm": 0.49609375, + "learning_rate": 9.512280701754387e-05, + "loss": 3.1471, + "step": 10867 + }, + { + "epoch": 0.572, + "grad_norm": 0.5390625, + "learning_rate": 9.511111111111112e-05, + "loss": 3.3025, + "step": 10868 + }, + { + "epoch": 0.5720526315789474, + "grad_norm": 0.455078125, + "learning_rate": 9.509941520467837e-05, + "loss": 3.2606, + "step": 10869 + }, + { + "epoch": 0.5721052631578948, + "grad_norm": 0.4609375, + "learning_rate": 9.508771929824562e-05, + "loss": 3.3612, + "step": 10870 + }, + { + "epoch": 0.5721578947368421, + "grad_norm": 0.48046875, + "learning_rate": 9.507602339181287e-05, + "loss": 3.1202, + "step": 10871 + }, + { + "epoch": 0.5722105263157895, + "grad_norm": 0.48046875, + "learning_rate": 9.506432748538012e-05, + "loss": 3.3472, + "step": 10872 + }, + { + "epoch": 0.5722631578947368, + "grad_norm": 0.490234375, + "learning_rate": 9.505263157894737e-05, + "loss": 3.2072, + "step": 10873 + }, + { + "epoch": 0.5723157894736842, + "grad_norm": 0.5078125, + "learning_rate": 9.504093567251462e-05, + "loss": 3.3349, + "step": 10874 + }, + { + "epoch": 0.5723684210526315, + "grad_norm": 0.462890625, + "learning_rate": 9.502923976608188e-05, + "loss": 3.2946, + "step": 10875 + }, + { + "epoch": 0.5724210526315789, + "grad_norm": 0.482421875, + "learning_rate": 9.501754385964913e-05, + "loss": 3.2105, + "step": 10876 + }, + { + "epoch": 0.5724736842105264, + "grad_norm": 0.494140625, + "learning_rate": 9.500584795321638e-05, + "loss": 3.2426, + "step": 10877 + }, + { + "epoch": 0.5725263157894737, + "grad_norm": 0.55078125, + "learning_rate": 9.499415204678363e-05, + "loss": 3.3566, + "step": 10878 + }, + { + "epoch": 0.5725789473684211, + "grad_norm": 0.5078125, + "learning_rate": 9.498245614035088e-05, + "loss": 3.6914, + "step": 10879 + }, + { + "epoch": 0.5726315789473684, + "grad_norm": 0.5078125, + "learning_rate": 9.497076023391813e-05, + "loss": 3.3132, + "step": 10880 + }, + { + "epoch": 0.5726842105263158, + "grad_norm": 0.57421875, + "learning_rate": 9.495906432748538e-05, + "loss": 3.2737, + "step": 10881 + }, + { + "epoch": 0.5727368421052632, + "grad_norm": 0.46875, + "learning_rate": 9.494736842105264e-05, + "loss": 3.2774, + "step": 10882 + }, + { + "epoch": 0.5727894736842105, + "grad_norm": 0.466796875, + "learning_rate": 9.493567251461989e-05, + "loss": 3.5066, + "step": 10883 + }, + { + "epoch": 0.5728421052631579, + "grad_norm": 0.474609375, + "learning_rate": 9.492397660818714e-05, + "loss": 3.3811, + "step": 10884 + }, + { + "epoch": 0.5728947368421052, + "grad_norm": 0.490234375, + "learning_rate": 9.491228070175439e-05, + "loss": 3.3309, + "step": 10885 + }, + { + "epoch": 0.5729473684210526, + "grad_norm": 0.494140625, + "learning_rate": 9.490058479532164e-05, + "loss": 3.4114, + "step": 10886 + }, + { + "epoch": 0.573, + "grad_norm": 0.474609375, + "learning_rate": 9.488888888888889e-05, + "loss": 3.1845, + "step": 10887 + }, + { + "epoch": 0.5730526315789474, + "grad_norm": 0.4921875, + "learning_rate": 9.487719298245614e-05, + "loss": 3.6881, + "step": 10888 + }, + { + "epoch": 0.5731052631578948, + "grad_norm": 0.470703125, + "learning_rate": 9.48654970760234e-05, + "loss": 3.5317, + "step": 10889 + }, + { + "epoch": 0.5731578947368421, + "grad_norm": 0.458984375, + "learning_rate": 9.485380116959065e-05, + "loss": 3.3583, + "step": 10890 + }, + { + "epoch": 0.5732105263157895, + "grad_norm": 0.498046875, + "learning_rate": 9.48421052631579e-05, + "loss": 3.4907, + "step": 10891 + }, + { + "epoch": 0.5732631578947368, + "grad_norm": 0.55859375, + "learning_rate": 9.483040935672515e-05, + "loss": 3.3982, + "step": 10892 + }, + { + "epoch": 0.5733157894736842, + "grad_norm": 0.51171875, + "learning_rate": 9.48187134502924e-05, + "loss": 3.8587, + "step": 10893 + }, + { + "epoch": 0.5733684210526315, + "grad_norm": 0.578125, + "learning_rate": 9.480701754385965e-05, + "loss": 3.3013, + "step": 10894 + }, + { + "epoch": 0.5734210526315789, + "grad_norm": 0.46484375, + "learning_rate": 9.47953216374269e-05, + "loss": 3.5528, + "step": 10895 + }, + { + "epoch": 0.5734736842105264, + "grad_norm": 0.58203125, + "learning_rate": 9.478362573099415e-05, + "loss": 3.65, + "step": 10896 + }, + { + "epoch": 0.5735263157894737, + "grad_norm": 0.50390625, + "learning_rate": 9.477192982456141e-05, + "loss": 3.6339, + "step": 10897 + }, + { + "epoch": 0.5735789473684211, + "grad_norm": 0.51953125, + "learning_rate": 9.476023391812866e-05, + "loss": 3.0899, + "step": 10898 + }, + { + "epoch": 0.5736315789473684, + "grad_norm": 0.48046875, + "learning_rate": 9.474853801169591e-05, + "loss": 3.6236, + "step": 10899 + }, + { + "epoch": 0.5736842105263158, + "grad_norm": 0.515625, + "learning_rate": 9.473684210526316e-05, + "loss": 3.5288, + "step": 10900 + }, + { + "epoch": 0.5737368421052632, + "grad_norm": 0.5078125, + "learning_rate": 9.472514619883041e-05, + "loss": 2.9792, + "step": 10901 + }, + { + "epoch": 0.5737894736842105, + "grad_norm": 0.453125, + "learning_rate": 9.471345029239766e-05, + "loss": 3.5721, + "step": 10902 + }, + { + "epoch": 0.5738421052631579, + "grad_norm": 0.51171875, + "learning_rate": 9.470175438596491e-05, + "loss": 3.5208, + "step": 10903 + }, + { + "epoch": 0.5738947368421052, + "grad_norm": 0.50390625, + "learning_rate": 9.469005847953217e-05, + "loss": 3.4832, + "step": 10904 + }, + { + "epoch": 0.5739473684210527, + "grad_norm": 0.44921875, + "learning_rate": 9.467836257309942e-05, + "loss": 3.2806, + "step": 10905 + }, + { + "epoch": 0.574, + "grad_norm": 0.494140625, + "learning_rate": 9.466666666666667e-05, + "loss": 3.7353, + "step": 10906 + }, + { + "epoch": 0.5740526315789474, + "grad_norm": 0.478515625, + "learning_rate": 9.465497076023392e-05, + "loss": 3.2904, + "step": 10907 + }, + { + "epoch": 0.5741052631578948, + "grad_norm": 0.48046875, + "learning_rate": 9.464327485380117e-05, + "loss": 3.2863, + "step": 10908 + }, + { + "epoch": 0.5741578947368421, + "grad_norm": 0.47265625, + "learning_rate": 9.463157894736842e-05, + "loss": 3.4702, + "step": 10909 + }, + { + "epoch": 0.5742105263157895, + "grad_norm": 0.4921875, + "learning_rate": 9.461988304093567e-05, + "loss": 3.4986, + "step": 10910 + }, + { + "epoch": 0.5742631578947368, + "grad_norm": 0.5, + "learning_rate": 9.460818713450292e-05, + "loss": 3.3368, + "step": 10911 + }, + { + "epoch": 0.5743157894736842, + "grad_norm": 0.51953125, + "learning_rate": 9.459649122807018e-05, + "loss": 2.6265, + "step": 10912 + }, + { + "epoch": 0.5743684210526315, + "grad_norm": 0.494140625, + "learning_rate": 9.458479532163743e-05, + "loss": 3.2356, + "step": 10913 + }, + { + "epoch": 0.5744210526315789, + "grad_norm": 0.46484375, + "learning_rate": 9.457309941520468e-05, + "loss": 3.4028, + "step": 10914 + }, + { + "epoch": 0.5744736842105264, + "grad_norm": 0.498046875, + "learning_rate": 9.456140350877195e-05, + "loss": 3.2857, + "step": 10915 + }, + { + "epoch": 0.5745263157894737, + "grad_norm": 0.47265625, + "learning_rate": 9.454970760233918e-05, + "loss": 3.0506, + "step": 10916 + }, + { + "epoch": 0.5745789473684211, + "grad_norm": 0.466796875, + "learning_rate": 9.453801169590643e-05, + "loss": 3.8462, + "step": 10917 + }, + { + "epoch": 0.5746315789473684, + "grad_norm": 0.482421875, + "learning_rate": 9.452631578947368e-05, + "loss": 3.0617, + "step": 10918 + }, + { + "epoch": 0.5746842105263158, + "grad_norm": 0.50390625, + "learning_rate": 9.451461988304094e-05, + "loss": 3.2998, + "step": 10919 + }, + { + "epoch": 0.5747368421052632, + "grad_norm": 0.83203125, + "learning_rate": 9.45029239766082e-05, + "loss": 3.3771, + "step": 10920 + }, + { + "epoch": 0.5747894736842105, + "grad_norm": 0.4453125, + "learning_rate": 9.449122807017544e-05, + "loss": 3.8772, + "step": 10921 + }, + { + "epoch": 0.5748421052631579, + "grad_norm": 0.51171875, + "learning_rate": 9.44795321637427e-05, + "loss": 3.5281, + "step": 10922 + }, + { + "epoch": 0.5748947368421052, + "grad_norm": 0.53515625, + "learning_rate": 9.446783625730994e-05, + "loss": 3.5686, + "step": 10923 + }, + { + "epoch": 0.5749473684210527, + "grad_norm": 0.5546875, + "learning_rate": 9.445614035087719e-05, + "loss": 3.2537, + "step": 10924 + }, + { + "epoch": 0.575, + "grad_norm": 0.49609375, + "learning_rate": 9.444444444444444e-05, + "loss": 3.369, + "step": 10925 + }, + { + "epoch": 0.5750526315789474, + "grad_norm": 0.478515625, + "learning_rate": 9.443274853801169e-05, + "loss": 3.2342, + "step": 10926 + }, + { + "epoch": 0.5751052631578948, + "grad_norm": 0.478515625, + "learning_rate": 9.442105263157895e-05, + "loss": 3.7983, + "step": 10927 + }, + { + "epoch": 0.5751578947368421, + "grad_norm": 0.5078125, + "learning_rate": 9.44093567251462e-05, + "loss": 3.4523, + "step": 10928 + }, + { + "epoch": 0.5752105263157895, + "grad_norm": 0.51171875, + "learning_rate": 9.439766081871345e-05, + "loss": 3.1771, + "step": 10929 + }, + { + "epoch": 0.5752631578947368, + "grad_norm": 0.51953125, + "learning_rate": 9.438596491228072e-05, + "loss": 3.2404, + "step": 10930 + }, + { + "epoch": 0.5753157894736842, + "grad_norm": 0.478515625, + "learning_rate": 9.437426900584795e-05, + "loss": 3.8173, + "step": 10931 + }, + { + "epoch": 0.5753684210526315, + "grad_norm": 0.478515625, + "learning_rate": 9.43625730994152e-05, + "loss": 3.3649, + "step": 10932 + }, + { + "epoch": 0.5754210526315789, + "grad_norm": 0.48828125, + "learning_rate": 9.435087719298245e-05, + "loss": 3.0736, + "step": 10933 + }, + { + "epoch": 0.5754736842105264, + "grad_norm": 0.486328125, + "learning_rate": 9.433918128654972e-05, + "loss": 3.318, + "step": 10934 + }, + { + "epoch": 0.5755263157894737, + "grad_norm": 0.466796875, + "learning_rate": 9.432748538011697e-05, + "loss": 3.8265, + "step": 10935 + }, + { + "epoch": 0.5755789473684211, + "grad_norm": 0.5078125, + "learning_rate": 9.431578947368421e-05, + "loss": 2.8959, + "step": 10936 + }, + { + "epoch": 0.5756315789473684, + "grad_norm": 0.4921875, + "learning_rate": 9.430409356725146e-05, + "loss": 3.3836, + "step": 10937 + }, + { + "epoch": 0.5756842105263158, + "grad_norm": 0.5078125, + "learning_rate": 9.429239766081873e-05, + "loss": 3.3344, + "step": 10938 + }, + { + "epoch": 0.5757368421052631, + "grad_norm": 0.4453125, + "learning_rate": 9.428070175438596e-05, + "loss": 3.4243, + "step": 10939 + }, + { + "epoch": 0.5757894736842105, + "grad_norm": 0.5234375, + "learning_rate": 9.426900584795321e-05, + "loss": 3.0916, + "step": 10940 + }, + { + "epoch": 0.5758421052631579, + "grad_norm": 0.546875, + "learning_rate": 9.425730994152048e-05, + "loss": 3.5577, + "step": 10941 + }, + { + "epoch": 0.5758947368421052, + "grad_norm": 0.515625, + "learning_rate": 9.424561403508773e-05, + "loss": 3.3203, + "step": 10942 + }, + { + "epoch": 0.5759473684210527, + "grad_norm": 0.466796875, + "learning_rate": 9.423391812865498e-05, + "loss": 3.168, + "step": 10943 + }, + { + "epoch": 0.576, + "grad_norm": 0.484375, + "learning_rate": 9.422222222222223e-05, + "loss": 3.5491, + "step": 10944 + }, + { + "epoch": 0.5760526315789474, + "grad_norm": 0.52734375, + "learning_rate": 9.421052631578949e-05, + "loss": 3.1889, + "step": 10945 + }, + { + "epoch": 0.5761052631578948, + "grad_norm": 0.478515625, + "learning_rate": 9.419883040935672e-05, + "loss": 3.1786, + "step": 10946 + }, + { + "epoch": 0.5761578947368421, + "grad_norm": 0.4765625, + "learning_rate": 9.418713450292397e-05, + "loss": 3.1818, + "step": 10947 + }, + { + "epoch": 0.5762105263157895, + "grad_norm": 0.478515625, + "learning_rate": 9.417543859649122e-05, + "loss": 3.2926, + "step": 10948 + }, + { + "epoch": 0.5762631578947368, + "grad_norm": 0.494140625, + "learning_rate": 9.416374269005849e-05, + "loss": 3.5493, + "step": 10949 + }, + { + "epoch": 0.5763157894736842, + "grad_norm": 0.55859375, + "learning_rate": 9.415204678362574e-05, + "loss": 3.4382, + "step": 10950 + }, + { + "epoch": 0.5763684210526315, + "grad_norm": 0.486328125, + "learning_rate": 9.414035087719299e-05, + "loss": 3.5869, + "step": 10951 + }, + { + "epoch": 0.576421052631579, + "grad_norm": 0.482421875, + "learning_rate": 9.412865497076025e-05, + "loss": 3.3252, + "step": 10952 + }, + { + "epoch": 0.5764736842105264, + "grad_norm": 0.494140625, + "learning_rate": 9.41169590643275e-05, + "loss": 3.6772, + "step": 10953 + }, + { + "epoch": 0.5765263157894737, + "grad_norm": 0.48828125, + "learning_rate": 9.410526315789473e-05, + "loss": 3.6532, + "step": 10954 + }, + { + "epoch": 0.5765789473684211, + "grad_norm": 0.48046875, + "learning_rate": 9.409356725146198e-05, + "loss": 3.173, + "step": 10955 + }, + { + "epoch": 0.5766315789473684, + "grad_norm": 0.478515625, + "learning_rate": 9.408187134502925e-05, + "loss": 3.1778, + "step": 10956 + }, + { + "epoch": 0.5766842105263158, + "grad_norm": 0.53125, + "learning_rate": 9.40701754385965e-05, + "loss": 3.4818, + "step": 10957 + }, + { + "epoch": 0.5767368421052631, + "grad_norm": 0.49609375, + "learning_rate": 9.405847953216375e-05, + "loss": 3.0173, + "step": 10958 + }, + { + "epoch": 0.5767894736842105, + "grad_norm": 0.4765625, + "learning_rate": 9.4046783625731e-05, + "loss": 3.7482, + "step": 10959 + }, + { + "epoch": 0.5768421052631579, + "grad_norm": 0.51953125, + "learning_rate": 9.403508771929826e-05, + "loss": 3.4256, + "step": 10960 + }, + { + "epoch": 0.5768947368421052, + "grad_norm": 0.49609375, + "learning_rate": 9.402339181286551e-05, + "loss": 3.2196, + "step": 10961 + }, + { + "epoch": 0.5769473684210527, + "grad_norm": 0.50390625, + "learning_rate": 9.401169590643274e-05, + "loss": 3.2518, + "step": 10962 + }, + { + "epoch": 0.577, + "grad_norm": 0.4921875, + "learning_rate": 9.4e-05, + "loss": 3.2303, + "step": 10963 + }, + { + "epoch": 0.5770526315789474, + "grad_norm": 0.53515625, + "learning_rate": 9.398830409356726e-05, + "loss": 3.2085, + "step": 10964 + }, + { + "epoch": 0.5771052631578948, + "grad_norm": 0.50390625, + "learning_rate": 9.397660818713451e-05, + "loss": 3.1938, + "step": 10965 + }, + { + "epoch": 0.5771578947368421, + "grad_norm": 0.47265625, + "learning_rate": 9.396491228070176e-05, + "loss": 3.5042, + "step": 10966 + }, + { + "epoch": 0.5772105263157895, + "grad_norm": 0.50390625, + "learning_rate": 9.395321637426902e-05, + "loss": 3.0992, + "step": 10967 + }, + { + "epoch": 0.5772631578947368, + "grad_norm": 0.47265625, + "learning_rate": 9.394152046783627e-05, + "loss": 3.2276, + "step": 10968 + }, + { + "epoch": 0.5773157894736842, + "grad_norm": 0.46484375, + "learning_rate": 9.39298245614035e-05, + "loss": 3.4834, + "step": 10969 + }, + { + "epoch": 0.5773684210526315, + "grad_norm": 0.47265625, + "learning_rate": 9.391812865497076e-05, + "loss": 3.3024, + "step": 10970 + }, + { + "epoch": 0.577421052631579, + "grad_norm": 0.482421875, + "learning_rate": 9.390643274853802e-05, + "loss": 3.4957, + "step": 10971 + }, + { + "epoch": 0.5774736842105264, + "grad_norm": 0.515625, + "learning_rate": 9.389473684210527e-05, + "loss": 3.4202, + "step": 10972 + }, + { + "epoch": 0.5775263157894737, + "grad_norm": 0.455078125, + "learning_rate": 9.388304093567252e-05, + "loss": 3.5431, + "step": 10973 + }, + { + "epoch": 0.5775789473684211, + "grad_norm": 0.474609375, + "learning_rate": 9.387134502923977e-05, + "loss": 3.6504, + "step": 10974 + }, + { + "epoch": 0.5776315789473684, + "grad_norm": 0.515625, + "learning_rate": 9.385964912280703e-05, + "loss": 3.4491, + "step": 10975 + }, + { + "epoch": 0.5776842105263158, + "grad_norm": 0.5234375, + "learning_rate": 9.384795321637428e-05, + "loss": 3.3787, + "step": 10976 + }, + { + "epoch": 0.5777368421052631, + "grad_norm": 0.515625, + "learning_rate": 9.383625730994152e-05, + "loss": 3.1225, + "step": 10977 + }, + { + "epoch": 0.5777894736842105, + "grad_norm": 0.47265625, + "learning_rate": 9.382456140350878e-05, + "loss": 3.2068, + "step": 10978 + }, + { + "epoch": 0.5778421052631579, + "grad_norm": 0.486328125, + "learning_rate": 9.381286549707603e-05, + "loss": 2.9234, + "step": 10979 + }, + { + "epoch": 0.5778947368421052, + "grad_norm": 0.462890625, + "learning_rate": 9.380116959064328e-05, + "loss": 3.7073, + "step": 10980 + }, + { + "epoch": 0.5779473684210527, + "grad_norm": 0.486328125, + "learning_rate": 9.378947368421053e-05, + "loss": 3.4024, + "step": 10981 + }, + { + "epoch": 0.578, + "grad_norm": 0.453125, + "learning_rate": 9.377777777777779e-05, + "loss": 3.4499, + "step": 10982 + }, + { + "epoch": 0.5780526315789474, + "grad_norm": 0.4765625, + "learning_rate": 9.376608187134504e-05, + "loss": 3.4614, + "step": 10983 + }, + { + "epoch": 0.5781052631578948, + "grad_norm": 0.5703125, + "learning_rate": 9.375438596491229e-05, + "loss": 3.4522, + "step": 10984 + }, + { + "epoch": 0.5781578947368421, + "grad_norm": 0.5, + "learning_rate": 9.374269005847953e-05, + "loss": 3.3244, + "step": 10985 + }, + { + "epoch": 0.5782105263157895, + "grad_norm": 0.474609375, + "learning_rate": 9.373099415204679e-05, + "loss": 3.2732, + "step": 10986 + }, + { + "epoch": 0.5782631578947368, + "grad_norm": 0.46875, + "learning_rate": 9.371929824561404e-05, + "loss": 3.7328, + "step": 10987 + }, + { + "epoch": 0.5783157894736842, + "grad_norm": 0.4921875, + "learning_rate": 9.370760233918129e-05, + "loss": 3.4778, + "step": 10988 + }, + { + "epoch": 0.5783684210526315, + "grad_norm": 0.5703125, + "learning_rate": 9.369590643274854e-05, + "loss": 3.2723, + "step": 10989 + }, + { + "epoch": 0.578421052631579, + "grad_norm": 0.53515625, + "learning_rate": 9.36842105263158e-05, + "loss": 3.4297, + "step": 10990 + }, + { + "epoch": 0.5784736842105264, + "grad_norm": 0.48828125, + "learning_rate": 9.367251461988305e-05, + "loss": 3.6355, + "step": 10991 + }, + { + "epoch": 0.5785263157894737, + "grad_norm": 0.478515625, + "learning_rate": 9.366081871345029e-05, + "loss": 3.3716, + "step": 10992 + }, + { + "epoch": 0.5785789473684211, + "grad_norm": 0.45703125, + "learning_rate": 9.364912280701755e-05, + "loss": 3.4186, + "step": 10993 + }, + { + "epoch": 0.5786315789473684, + "grad_norm": 0.48828125, + "learning_rate": 9.36374269005848e-05, + "loss": 3.0926, + "step": 10994 + }, + { + "epoch": 0.5786842105263158, + "grad_norm": 0.48046875, + "learning_rate": 9.362573099415205e-05, + "loss": 3.3834, + "step": 10995 + }, + { + "epoch": 0.5787368421052631, + "grad_norm": 0.5078125, + "learning_rate": 9.36140350877193e-05, + "loss": 3.2214, + "step": 10996 + }, + { + "epoch": 0.5787894736842105, + "grad_norm": 0.57421875, + "learning_rate": 9.360233918128656e-05, + "loss": 3.4216, + "step": 10997 + }, + { + "epoch": 0.5788421052631579, + "grad_norm": 0.48046875, + "learning_rate": 9.359064327485381e-05, + "loss": 3.5601, + "step": 10998 + }, + { + "epoch": 0.5788947368421052, + "grad_norm": 0.490234375, + "learning_rate": 9.357894736842106e-05, + "loss": 3.7349, + "step": 10999 + }, + { + "epoch": 0.5789473684210527, + "grad_norm": 0.57421875, + "learning_rate": 9.35672514619883e-05, + "loss": 3.3804, + "step": 11000 + }, + { + "epoch": 0.579, + "grad_norm": 0.55078125, + "learning_rate": 9.355555555555556e-05, + "loss": 3.1042, + "step": 11001 + }, + { + "epoch": 0.5790526315789474, + "grad_norm": 0.46875, + "learning_rate": 9.354385964912281e-05, + "loss": 3.6202, + "step": 11002 + }, + { + "epoch": 0.5791052631578948, + "grad_norm": 0.490234375, + "learning_rate": 9.353216374269006e-05, + "loss": 3.2888, + "step": 11003 + }, + { + "epoch": 0.5791578947368421, + "grad_norm": 0.46875, + "learning_rate": 9.352046783625732e-05, + "loss": 3.2652, + "step": 11004 + }, + { + "epoch": 0.5792105263157895, + "grad_norm": 0.494140625, + "learning_rate": 9.350877192982457e-05, + "loss": 3.3569, + "step": 11005 + }, + { + "epoch": 0.5792631578947368, + "grad_norm": 0.490234375, + "learning_rate": 9.349707602339182e-05, + "loss": 3.3856, + "step": 11006 + }, + { + "epoch": 0.5793157894736842, + "grad_norm": 0.47265625, + "learning_rate": 9.348538011695907e-05, + "loss": 3.2337, + "step": 11007 + }, + { + "epoch": 0.5793684210526315, + "grad_norm": 0.478515625, + "learning_rate": 9.347368421052632e-05, + "loss": 3.3591, + "step": 11008 + }, + { + "epoch": 0.579421052631579, + "grad_norm": 0.494140625, + "learning_rate": 9.346198830409357e-05, + "loss": 3.1051, + "step": 11009 + }, + { + "epoch": 0.5794736842105264, + "grad_norm": 0.474609375, + "learning_rate": 9.345029239766082e-05, + "loss": 3.6334, + "step": 11010 + }, + { + "epoch": 0.5795263157894737, + "grad_norm": 0.51171875, + "learning_rate": 9.343859649122807e-05, + "loss": 3.2804, + "step": 11011 + }, + { + "epoch": 0.5795789473684211, + "grad_norm": 0.5078125, + "learning_rate": 9.342690058479533e-05, + "loss": 3.6489, + "step": 11012 + }, + { + "epoch": 0.5796315789473684, + "grad_norm": 0.58203125, + "learning_rate": 9.341520467836258e-05, + "loss": 3.3899, + "step": 11013 + }, + { + "epoch": 0.5796842105263158, + "grad_norm": 0.478515625, + "learning_rate": 9.340350877192983e-05, + "loss": 3.201, + "step": 11014 + }, + { + "epoch": 0.5797368421052631, + "grad_norm": 0.4609375, + "learning_rate": 9.339181286549708e-05, + "loss": 3.6039, + "step": 11015 + }, + { + "epoch": 0.5797894736842105, + "grad_norm": 0.482421875, + "learning_rate": 9.338011695906433e-05, + "loss": 3.2507, + "step": 11016 + }, + { + "epoch": 0.5798421052631579, + "grad_norm": 0.453125, + "learning_rate": 9.336842105263158e-05, + "loss": 3.4377, + "step": 11017 + }, + { + "epoch": 0.5798947368421052, + "grad_norm": 0.474609375, + "learning_rate": 9.335672514619883e-05, + "loss": 2.9641, + "step": 11018 + }, + { + "epoch": 0.5799473684210527, + "grad_norm": 0.453125, + "learning_rate": 9.33450292397661e-05, + "loss": 3.1998, + "step": 11019 + }, + { + "epoch": 0.58, + "grad_norm": 0.47265625, + "learning_rate": 9.333333333333334e-05, + "loss": 3.1751, + "step": 11020 + }, + { + "epoch": 0.5800526315789474, + "grad_norm": 0.4609375, + "learning_rate": 9.332163742690059e-05, + "loss": 3.5155, + "step": 11021 + }, + { + "epoch": 0.5801052631578948, + "grad_norm": 0.51171875, + "learning_rate": 9.330994152046784e-05, + "loss": 3.4285, + "step": 11022 + }, + { + "epoch": 0.5801578947368421, + "grad_norm": 0.470703125, + "learning_rate": 9.329824561403509e-05, + "loss": 3.1833, + "step": 11023 + }, + { + "epoch": 0.5802105263157895, + "grad_norm": 0.46875, + "learning_rate": 9.328654970760234e-05, + "loss": 3.6294, + "step": 11024 + }, + { + "epoch": 0.5802631578947368, + "grad_norm": 0.5, + "learning_rate": 9.327485380116959e-05, + "loss": 3.2561, + "step": 11025 + }, + { + "epoch": 0.5803157894736842, + "grad_norm": 0.51171875, + "learning_rate": 9.326315789473684e-05, + "loss": 3.0925, + "step": 11026 + }, + { + "epoch": 0.5803684210526315, + "grad_norm": 0.4921875, + "learning_rate": 9.32514619883041e-05, + "loss": 3.3634, + "step": 11027 + }, + { + "epoch": 0.580421052631579, + "grad_norm": 0.48828125, + "learning_rate": 9.323976608187135e-05, + "loss": 3.1287, + "step": 11028 + }, + { + "epoch": 0.5804736842105264, + "grad_norm": 0.486328125, + "learning_rate": 9.32280701754386e-05, + "loss": 3.1919, + "step": 11029 + }, + { + "epoch": 0.5805263157894737, + "grad_norm": 0.46875, + "learning_rate": 9.321637426900585e-05, + "loss": 3.2473, + "step": 11030 + }, + { + "epoch": 0.5805789473684211, + "grad_norm": 0.50390625, + "learning_rate": 9.32046783625731e-05, + "loss": 3.5882, + "step": 11031 + }, + { + "epoch": 0.5806315789473684, + "grad_norm": 0.498046875, + "learning_rate": 9.319298245614035e-05, + "loss": 3.5506, + "step": 11032 + }, + { + "epoch": 0.5806842105263158, + "grad_norm": 0.474609375, + "learning_rate": 9.31812865497076e-05, + "loss": 3.8787, + "step": 11033 + }, + { + "epoch": 0.5807368421052631, + "grad_norm": 0.484375, + "learning_rate": 9.316959064327486e-05, + "loss": 3.3356, + "step": 11034 + }, + { + "epoch": 0.5807894736842105, + "grad_norm": 0.47265625, + "learning_rate": 9.315789473684211e-05, + "loss": 3.3667, + "step": 11035 + }, + { + "epoch": 0.5808421052631579, + "grad_norm": 0.51953125, + "learning_rate": 9.314619883040936e-05, + "loss": 3.4444, + "step": 11036 + }, + { + "epoch": 0.5808947368421052, + "grad_norm": 0.51171875, + "learning_rate": 9.313450292397661e-05, + "loss": 3.2727, + "step": 11037 + }, + { + "epoch": 0.5809473684210527, + "grad_norm": 0.498046875, + "learning_rate": 9.312280701754386e-05, + "loss": 3.2478, + "step": 11038 + }, + { + "epoch": 0.581, + "grad_norm": 0.48828125, + "learning_rate": 9.311111111111111e-05, + "loss": 2.9444, + "step": 11039 + }, + { + "epoch": 0.5810526315789474, + "grad_norm": 0.455078125, + "learning_rate": 9.309941520467836e-05, + "loss": 3.659, + "step": 11040 + }, + { + "epoch": 0.5811052631578948, + "grad_norm": 0.5546875, + "learning_rate": 9.308771929824563e-05, + "loss": 3.3744, + "step": 11041 + }, + { + "epoch": 0.5811578947368421, + "grad_norm": 0.482421875, + "learning_rate": 9.307602339181287e-05, + "loss": 3.4495, + "step": 11042 + }, + { + "epoch": 0.5812105263157895, + "grad_norm": 0.578125, + "learning_rate": 9.306432748538012e-05, + "loss": 3.5217, + "step": 11043 + }, + { + "epoch": 0.5812631578947368, + "grad_norm": 0.5078125, + "learning_rate": 9.305263157894737e-05, + "loss": 2.9419, + "step": 11044 + }, + { + "epoch": 0.5813157894736842, + "grad_norm": 0.51171875, + "learning_rate": 9.304093567251462e-05, + "loss": 3.2423, + "step": 11045 + }, + { + "epoch": 0.5813684210526315, + "grad_norm": 0.51171875, + "learning_rate": 9.302923976608187e-05, + "loss": 3.4252, + "step": 11046 + }, + { + "epoch": 0.581421052631579, + "grad_norm": 0.55078125, + "learning_rate": 9.301754385964912e-05, + "loss": 3.3496, + "step": 11047 + }, + { + "epoch": 0.5814736842105264, + "grad_norm": 0.515625, + "learning_rate": 9.300584795321637e-05, + "loss": 2.9956, + "step": 11048 + }, + { + "epoch": 0.5815263157894737, + "grad_norm": 0.5, + "learning_rate": 9.299415204678364e-05, + "loss": 3.5571, + "step": 11049 + }, + { + "epoch": 0.5815789473684211, + "grad_norm": 0.5078125, + "learning_rate": 9.298245614035089e-05, + "loss": 3.2982, + "step": 11050 + }, + { + "epoch": 0.5816315789473684, + "grad_norm": 0.4765625, + "learning_rate": 9.297076023391813e-05, + "loss": 3.2641, + "step": 11051 + }, + { + "epoch": 0.5816842105263158, + "grad_norm": 0.46875, + "learning_rate": 9.295906432748538e-05, + "loss": 3.7631, + "step": 11052 + }, + { + "epoch": 0.5817368421052631, + "grad_norm": 0.470703125, + "learning_rate": 9.294736842105263e-05, + "loss": 3.4511, + "step": 11053 + }, + { + "epoch": 0.5817894736842105, + "grad_norm": 0.515625, + "learning_rate": 9.293567251461988e-05, + "loss": 2.8478, + "step": 11054 + }, + { + "epoch": 0.5818421052631579, + "grad_norm": 0.48046875, + "learning_rate": 9.292397660818713e-05, + "loss": 3.7079, + "step": 11055 + }, + { + "epoch": 0.5818947368421052, + "grad_norm": 0.546875, + "learning_rate": 9.29122807017544e-05, + "loss": 3.4401, + "step": 11056 + }, + { + "epoch": 0.5819473684210527, + "grad_norm": 0.50390625, + "learning_rate": 9.290058479532165e-05, + "loss": 3.3213, + "step": 11057 + }, + { + "epoch": 0.582, + "grad_norm": 0.5546875, + "learning_rate": 9.28888888888889e-05, + "loss": 3.2913, + "step": 11058 + }, + { + "epoch": 0.5820526315789474, + "grad_norm": 0.490234375, + "learning_rate": 9.287719298245615e-05, + "loss": 3.4537, + "step": 11059 + }, + { + "epoch": 0.5821052631578948, + "grad_norm": 0.5234375, + "learning_rate": 9.28654970760234e-05, + "loss": 2.9672, + "step": 11060 + }, + { + "epoch": 0.5821578947368421, + "grad_norm": 0.47265625, + "learning_rate": 9.285380116959064e-05, + "loss": 3.6593, + "step": 11061 + }, + { + "epoch": 0.5822105263157895, + "grad_norm": 0.470703125, + "learning_rate": 9.28421052631579e-05, + "loss": 3.1391, + "step": 11062 + }, + { + "epoch": 0.5822631578947368, + "grad_norm": 0.494140625, + "learning_rate": 9.283040935672514e-05, + "loss": 3.2533, + "step": 11063 + }, + { + "epoch": 0.5823157894736842, + "grad_norm": 0.5, + "learning_rate": 9.28187134502924e-05, + "loss": 3.6467, + "step": 11064 + }, + { + "epoch": 0.5823684210526315, + "grad_norm": 0.4765625, + "learning_rate": 9.280701754385966e-05, + "loss": 3.1269, + "step": 11065 + }, + { + "epoch": 0.582421052631579, + "grad_norm": 0.4609375, + "learning_rate": 9.27953216374269e-05, + "loss": 3.5534, + "step": 11066 + }, + { + "epoch": 0.5824736842105264, + "grad_norm": 0.482421875, + "learning_rate": 9.278362573099416e-05, + "loss": 3.2965, + "step": 11067 + }, + { + "epoch": 0.5825263157894737, + "grad_norm": 0.51171875, + "learning_rate": 9.27719298245614e-05, + "loss": 3.3557, + "step": 11068 + }, + { + "epoch": 0.5825789473684211, + "grad_norm": 0.486328125, + "learning_rate": 9.276023391812865e-05, + "loss": 3.3298, + "step": 11069 + }, + { + "epoch": 0.5826315789473684, + "grad_norm": 0.484375, + "learning_rate": 9.27485380116959e-05, + "loss": 3.1258, + "step": 11070 + }, + { + "epoch": 0.5826842105263158, + "grad_norm": 0.4921875, + "learning_rate": 9.273684210526317e-05, + "loss": 3.4084, + "step": 11071 + }, + { + "epoch": 0.5827368421052631, + "grad_norm": 0.486328125, + "learning_rate": 9.272514619883042e-05, + "loss": 3.371, + "step": 11072 + }, + { + "epoch": 0.5827894736842105, + "grad_norm": 0.51171875, + "learning_rate": 9.271345029239767e-05, + "loss": 3.6359, + "step": 11073 + }, + { + "epoch": 0.5828421052631579, + "grad_norm": 0.4921875, + "learning_rate": 9.270175438596492e-05, + "loss": 3.415, + "step": 11074 + }, + { + "epoch": 0.5828947368421052, + "grad_norm": 0.498046875, + "learning_rate": 9.269005847953217e-05, + "loss": 3.3526, + "step": 11075 + }, + { + "epoch": 0.5829473684210527, + "grad_norm": 0.5390625, + "learning_rate": 9.267836257309942e-05, + "loss": 3.6957, + "step": 11076 + }, + { + "epoch": 0.583, + "grad_norm": 0.47265625, + "learning_rate": 9.266666666666666e-05, + "loss": 3.7849, + "step": 11077 + }, + { + "epoch": 0.5830526315789474, + "grad_norm": 0.484375, + "learning_rate": 9.265497076023393e-05, + "loss": 3.2946, + "step": 11078 + }, + { + "epoch": 0.5831052631578947, + "grad_norm": 0.50390625, + "learning_rate": 9.264327485380118e-05, + "loss": 3.5428, + "step": 11079 + }, + { + "epoch": 0.5831578947368421, + "grad_norm": 0.50390625, + "learning_rate": 9.263157894736843e-05, + "loss": 3.1688, + "step": 11080 + }, + { + "epoch": 0.5832105263157895, + "grad_norm": 0.51171875, + "learning_rate": 9.261988304093568e-05, + "loss": 3.6845, + "step": 11081 + }, + { + "epoch": 0.5832631578947368, + "grad_norm": 0.494140625, + "learning_rate": 9.260818713450293e-05, + "loss": 3.3231, + "step": 11082 + }, + { + "epoch": 0.5833157894736842, + "grad_norm": 0.484375, + "learning_rate": 9.259649122807018e-05, + "loss": 2.9341, + "step": 11083 + }, + { + "epoch": 0.5833684210526315, + "grad_norm": 0.466796875, + "learning_rate": 9.258479532163743e-05, + "loss": 3.0156, + "step": 11084 + }, + { + "epoch": 0.583421052631579, + "grad_norm": 0.490234375, + "learning_rate": 9.257309941520468e-05, + "loss": 3.495, + "step": 11085 + }, + { + "epoch": 0.5834736842105264, + "grad_norm": 0.5078125, + "learning_rate": 9.256140350877194e-05, + "loss": 3.2109, + "step": 11086 + }, + { + "epoch": 0.5835263157894737, + "grad_norm": 0.478515625, + "learning_rate": 9.254970760233919e-05, + "loss": 3.3892, + "step": 11087 + }, + { + "epoch": 0.5835789473684211, + "grad_norm": 0.474609375, + "learning_rate": 9.253801169590644e-05, + "loss": 3.6459, + "step": 11088 + }, + { + "epoch": 0.5836315789473684, + "grad_norm": 0.478515625, + "learning_rate": 9.252631578947369e-05, + "loss": 3.1657, + "step": 11089 + }, + { + "epoch": 0.5836842105263158, + "grad_norm": 0.5, + "learning_rate": 9.251461988304094e-05, + "loss": 3.5868, + "step": 11090 + }, + { + "epoch": 0.5837368421052631, + "grad_norm": 0.48046875, + "learning_rate": 9.250292397660819e-05, + "loss": 3.6698, + "step": 11091 + }, + { + "epoch": 0.5837894736842105, + "grad_norm": 0.486328125, + "learning_rate": 9.249122807017544e-05, + "loss": 3.2851, + "step": 11092 + }, + { + "epoch": 0.5838421052631579, + "grad_norm": 0.486328125, + "learning_rate": 9.24795321637427e-05, + "loss": 3.9551, + "step": 11093 + }, + { + "epoch": 0.5838947368421052, + "grad_norm": 0.50390625, + "learning_rate": 9.246783625730995e-05, + "loss": 3.4003, + "step": 11094 + }, + { + "epoch": 0.5839473684210527, + "grad_norm": 0.484375, + "learning_rate": 9.24561403508772e-05, + "loss": 2.9956, + "step": 11095 + }, + { + "epoch": 0.584, + "grad_norm": 0.515625, + "learning_rate": 9.244444444444445e-05, + "loss": 3.2811, + "step": 11096 + }, + { + "epoch": 0.5840526315789474, + "grad_norm": 0.50390625, + "learning_rate": 9.243274853801171e-05, + "loss": 3.5068, + "step": 11097 + }, + { + "epoch": 0.5841052631578947, + "grad_norm": 0.48046875, + "learning_rate": 9.242105263157895e-05, + "loss": 3.4306, + "step": 11098 + }, + { + "epoch": 0.5841578947368421, + "grad_norm": 0.4921875, + "learning_rate": 9.24093567251462e-05, + "loss": 3.5857, + "step": 11099 + }, + { + "epoch": 0.5842105263157895, + "grad_norm": 0.48828125, + "learning_rate": 9.239766081871345e-05, + "loss": 3.6228, + "step": 11100 + }, + { + "epoch": 0.5842631578947368, + "grad_norm": 0.48046875, + "learning_rate": 9.238596491228071e-05, + "loss": 3.4517, + "step": 11101 + }, + { + "epoch": 0.5843157894736842, + "grad_norm": 0.50390625, + "learning_rate": 9.237426900584796e-05, + "loss": 3.4061, + "step": 11102 + }, + { + "epoch": 0.5843684210526315, + "grad_norm": 0.5234375, + "learning_rate": 9.236257309941521e-05, + "loss": 3.4013, + "step": 11103 + }, + { + "epoch": 0.584421052631579, + "grad_norm": 0.478515625, + "learning_rate": 9.235087719298247e-05, + "loss": 3.538, + "step": 11104 + }, + { + "epoch": 0.5844736842105264, + "grad_norm": 0.55859375, + "learning_rate": 9.233918128654971e-05, + "loss": 3.4835, + "step": 11105 + }, + { + "epoch": 0.5845263157894737, + "grad_norm": 0.474609375, + "learning_rate": 9.232748538011696e-05, + "loss": 3.1862, + "step": 11106 + }, + { + "epoch": 0.5845789473684211, + "grad_norm": 0.46875, + "learning_rate": 9.231578947368421e-05, + "loss": 3.3993, + "step": 11107 + }, + { + "epoch": 0.5846315789473684, + "grad_norm": 0.478515625, + "learning_rate": 9.230409356725147e-05, + "loss": 3.3646, + "step": 11108 + }, + { + "epoch": 0.5846842105263158, + "grad_norm": 0.5078125, + "learning_rate": 9.229239766081872e-05, + "loss": 3.465, + "step": 11109 + }, + { + "epoch": 0.5847368421052631, + "grad_norm": 0.46875, + "learning_rate": 9.228070175438597e-05, + "loss": 3.7489, + "step": 11110 + }, + { + "epoch": 0.5847894736842105, + "grad_norm": 0.50390625, + "learning_rate": 9.226900584795322e-05, + "loss": 3.1902, + "step": 11111 + }, + { + "epoch": 0.5848421052631579, + "grad_norm": 0.484375, + "learning_rate": 9.225730994152048e-05, + "loss": 3.6028, + "step": 11112 + }, + { + "epoch": 0.5848947368421052, + "grad_norm": 0.470703125, + "learning_rate": 9.224561403508772e-05, + "loss": 3.4429, + "step": 11113 + }, + { + "epoch": 0.5849473684210527, + "grad_norm": 0.484375, + "learning_rate": 9.223391812865497e-05, + "loss": 3.2108, + "step": 11114 + }, + { + "epoch": 0.585, + "grad_norm": 0.5625, + "learning_rate": 9.222222222222223e-05, + "loss": 3.102, + "step": 11115 + }, + { + "epoch": 0.5850526315789474, + "grad_norm": 0.6015625, + "learning_rate": 9.221052631578948e-05, + "loss": 3.3715, + "step": 11116 + }, + { + "epoch": 0.5851052631578947, + "grad_norm": 0.498046875, + "learning_rate": 9.219883040935673e-05, + "loss": 3.6199, + "step": 11117 + }, + { + "epoch": 0.5851578947368421, + "grad_norm": 0.5390625, + "learning_rate": 9.218713450292398e-05, + "loss": 3.6406, + "step": 11118 + }, + { + "epoch": 0.5852105263157895, + "grad_norm": 0.50390625, + "learning_rate": 9.217543859649124e-05, + "loss": 3.5607, + "step": 11119 + }, + { + "epoch": 0.5852631578947368, + "grad_norm": 0.5078125, + "learning_rate": 9.216374269005849e-05, + "loss": 3.1386, + "step": 11120 + }, + { + "epoch": 0.5853157894736842, + "grad_norm": 0.466796875, + "learning_rate": 9.215204678362573e-05, + "loss": 3.272, + "step": 11121 + }, + { + "epoch": 0.5853684210526315, + "grad_norm": 0.478515625, + "learning_rate": 9.214035087719298e-05, + "loss": 2.7204, + "step": 11122 + }, + { + "epoch": 0.585421052631579, + "grad_norm": 0.52734375, + "learning_rate": 9.212865497076024e-05, + "loss": 3.4326, + "step": 11123 + }, + { + "epoch": 0.5854736842105264, + "grad_norm": 0.5078125, + "learning_rate": 9.211695906432749e-05, + "loss": 3.5288, + "step": 11124 + }, + { + "epoch": 0.5855263157894737, + "grad_norm": 0.6875, + "learning_rate": 9.210526315789474e-05, + "loss": 3.2879, + "step": 11125 + }, + { + "epoch": 0.5855789473684211, + "grad_norm": 0.4765625, + "learning_rate": 9.209356725146199e-05, + "loss": 3.3131, + "step": 11126 + }, + { + "epoch": 0.5856315789473684, + "grad_norm": 0.5234375, + "learning_rate": 9.208187134502925e-05, + "loss": 3.0547, + "step": 11127 + }, + { + "epoch": 0.5856842105263158, + "grad_norm": 0.48046875, + "learning_rate": 9.207017543859649e-05, + "loss": 3.5307, + "step": 11128 + }, + { + "epoch": 0.5857368421052631, + "grad_norm": 0.498046875, + "learning_rate": 9.205847953216374e-05, + "loss": 3.6577, + "step": 11129 + }, + { + "epoch": 0.5857894736842105, + "grad_norm": 0.515625, + "learning_rate": 9.2046783625731e-05, + "loss": 3.4149, + "step": 11130 + }, + { + "epoch": 0.5858421052631579, + "grad_norm": 0.494140625, + "learning_rate": 9.203508771929825e-05, + "loss": 3.6101, + "step": 11131 + }, + { + "epoch": 0.5858947368421052, + "grad_norm": 0.50390625, + "learning_rate": 9.20233918128655e-05, + "loss": 3.991, + "step": 11132 + }, + { + "epoch": 0.5859473684210527, + "grad_norm": 0.494140625, + "learning_rate": 9.201169590643275e-05, + "loss": 3.449, + "step": 11133 + }, + { + "epoch": 0.586, + "grad_norm": 0.474609375, + "learning_rate": 9.200000000000001e-05, + "loss": 3.5422, + "step": 11134 + }, + { + "epoch": 0.5860526315789474, + "grad_norm": 0.482421875, + "learning_rate": 9.198830409356726e-05, + "loss": 3.2596, + "step": 11135 + }, + { + "epoch": 0.5861052631578947, + "grad_norm": 0.55078125, + "learning_rate": 9.19766081871345e-05, + "loss": 3.385, + "step": 11136 + }, + { + "epoch": 0.5861578947368421, + "grad_norm": 0.4609375, + "learning_rate": 9.196491228070175e-05, + "loss": 3.8321, + "step": 11137 + }, + { + "epoch": 0.5862105263157895, + "grad_norm": 0.482421875, + "learning_rate": 9.195321637426901e-05, + "loss": 3.5872, + "step": 11138 + }, + { + "epoch": 0.5862631578947368, + "grad_norm": 0.490234375, + "learning_rate": 9.194152046783626e-05, + "loss": 3.0957, + "step": 11139 + }, + { + "epoch": 0.5863157894736842, + "grad_norm": 0.470703125, + "learning_rate": 9.192982456140351e-05, + "loss": 3.4018, + "step": 11140 + }, + { + "epoch": 0.5863684210526315, + "grad_norm": 0.48046875, + "learning_rate": 9.191812865497077e-05, + "loss": 3.2723, + "step": 11141 + }, + { + "epoch": 0.586421052631579, + "grad_norm": 0.482421875, + "learning_rate": 9.190643274853802e-05, + "loss": 2.9707, + "step": 11142 + }, + { + "epoch": 0.5864736842105264, + "grad_norm": 0.5078125, + "learning_rate": 9.189473684210527e-05, + "loss": 3.7546, + "step": 11143 + }, + { + "epoch": 0.5865263157894737, + "grad_norm": 0.466796875, + "learning_rate": 9.188304093567251e-05, + "loss": 3.3442, + "step": 11144 + }, + { + "epoch": 0.5865789473684211, + "grad_norm": 0.5234375, + "learning_rate": 9.187134502923977e-05, + "loss": 3.27, + "step": 11145 + }, + { + "epoch": 0.5866315789473684, + "grad_norm": 0.5703125, + "learning_rate": 9.185964912280702e-05, + "loss": 2.9279, + "step": 11146 + }, + { + "epoch": 0.5866842105263158, + "grad_norm": 0.46875, + "learning_rate": 9.184795321637427e-05, + "loss": 3.1339, + "step": 11147 + }, + { + "epoch": 0.5867368421052631, + "grad_norm": 0.4609375, + "learning_rate": 9.183625730994152e-05, + "loss": 2.9577, + "step": 11148 + }, + { + "epoch": 0.5867894736842105, + "grad_norm": 0.482421875, + "learning_rate": 9.182456140350878e-05, + "loss": 2.9872, + "step": 11149 + }, + { + "epoch": 0.5868421052631579, + "grad_norm": 0.64453125, + "learning_rate": 9.181286549707603e-05, + "loss": 3.5245, + "step": 11150 + }, + { + "epoch": 0.5868947368421052, + "grad_norm": 0.55078125, + "learning_rate": 9.180116959064327e-05, + "loss": 3.2005, + "step": 11151 + }, + { + "epoch": 0.5869473684210527, + "grad_norm": 0.4765625, + "learning_rate": 9.178947368421052e-05, + "loss": 2.8478, + "step": 11152 + }, + { + "epoch": 0.587, + "grad_norm": 0.462890625, + "learning_rate": 9.177777777777778e-05, + "loss": 3.2933, + "step": 11153 + }, + { + "epoch": 0.5870526315789474, + "grad_norm": 0.44921875, + "learning_rate": 9.176608187134503e-05, + "loss": 3.5642, + "step": 11154 + }, + { + "epoch": 0.5871052631578947, + "grad_norm": 0.478515625, + "learning_rate": 9.175438596491228e-05, + "loss": 3.2411, + "step": 11155 + }, + { + "epoch": 0.5871578947368421, + "grad_norm": 0.5078125, + "learning_rate": 9.174269005847955e-05, + "loss": 3.1145, + "step": 11156 + }, + { + "epoch": 0.5872105263157895, + "grad_norm": 0.490234375, + "learning_rate": 9.17309941520468e-05, + "loss": 3.4344, + "step": 11157 + }, + { + "epoch": 0.5872631578947368, + "grad_norm": 0.470703125, + "learning_rate": 9.171929824561404e-05, + "loss": 3.8349, + "step": 11158 + }, + { + "epoch": 0.5873157894736842, + "grad_norm": 0.47265625, + "learning_rate": 9.170760233918128e-05, + "loss": 3.5438, + "step": 11159 + }, + { + "epoch": 0.5873684210526315, + "grad_norm": 0.453125, + "learning_rate": 9.169590643274854e-05, + "loss": 3.5629, + "step": 11160 + }, + { + "epoch": 0.587421052631579, + "grad_norm": 0.478515625, + "learning_rate": 9.16842105263158e-05, + "loss": 3.491, + "step": 11161 + }, + { + "epoch": 0.5874736842105264, + "grad_norm": 0.482421875, + "learning_rate": 9.167251461988304e-05, + "loss": 3.6654, + "step": 11162 + }, + { + "epoch": 0.5875263157894737, + "grad_norm": 0.4921875, + "learning_rate": 9.166081871345029e-05, + "loss": 3.4986, + "step": 11163 + }, + { + "epoch": 0.5875789473684211, + "grad_norm": 0.51171875, + "learning_rate": 9.164912280701756e-05, + "loss": 3.2107, + "step": 11164 + }, + { + "epoch": 0.5876315789473684, + "grad_norm": 0.46484375, + "learning_rate": 9.16374269005848e-05, + "loss": 3.125, + "step": 11165 + }, + { + "epoch": 0.5876842105263158, + "grad_norm": 0.447265625, + "learning_rate": 9.162573099415205e-05, + "loss": 3.3556, + "step": 11166 + }, + { + "epoch": 0.5877368421052631, + "grad_norm": 0.6640625, + "learning_rate": 9.16140350877193e-05, + "loss": 3.338, + "step": 11167 + }, + { + "epoch": 0.5877894736842105, + "grad_norm": 0.51953125, + "learning_rate": 9.160233918128655e-05, + "loss": 3.1744, + "step": 11168 + }, + { + "epoch": 0.5878421052631579, + "grad_norm": 0.48046875, + "learning_rate": 9.15906432748538e-05, + "loss": 3.35, + "step": 11169 + }, + { + "epoch": 0.5878947368421052, + "grad_norm": 0.4765625, + "learning_rate": 9.157894736842105e-05, + "loss": 3.1882, + "step": 11170 + }, + { + "epoch": 0.5879473684210527, + "grad_norm": 1.3515625, + "learning_rate": 9.156725146198832e-05, + "loss": 3.4491, + "step": 11171 + }, + { + "epoch": 0.588, + "grad_norm": 0.5546875, + "learning_rate": 9.155555555555557e-05, + "loss": 3.6722, + "step": 11172 + }, + { + "epoch": 0.5880526315789474, + "grad_norm": 0.494140625, + "learning_rate": 9.154385964912282e-05, + "loss": 3.376, + "step": 11173 + }, + { + "epoch": 0.5881052631578947, + "grad_norm": 0.5078125, + "learning_rate": 9.153216374269005e-05, + "loss": 3.781, + "step": 11174 + }, + { + "epoch": 0.5881578947368421, + "grad_norm": 0.5, + "learning_rate": 9.152046783625731e-05, + "loss": 3.1533, + "step": 11175 + }, + { + "epoch": 0.5882105263157895, + "grad_norm": 0.447265625, + "learning_rate": 9.150877192982456e-05, + "loss": 3.0677, + "step": 11176 + }, + { + "epoch": 0.5882631578947368, + "grad_norm": 0.490234375, + "learning_rate": 9.149707602339181e-05, + "loss": 3.4055, + "step": 11177 + }, + { + "epoch": 0.5883157894736842, + "grad_norm": 0.5078125, + "learning_rate": 9.148538011695908e-05, + "loss": 3.1135, + "step": 11178 + }, + { + "epoch": 0.5883684210526315, + "grad_norm": 0.474609375, + "learning_rate": 9.147368421052633e-05, + "loss": 3.5512, + "step": 11179 + }, + { + "epoch": 0.588421052631579, + "grad_norm": 0.49609375, + "learning_rate": 9.146198830409358e-05, + "loss": 3.1125, + "step": 11180 + }, + { + "epoch": 0.5884736842105264, + "grad_norm": 0.5078125, + "learning_rate": 9.145029239766083e-05, + "loss": 3.2601, + "step": 11181 + }, + { + "epoch": 0.5885263157894737, + "grad_norm": 0.478515625, + "learning_rate": 9.143859649122808e-05, + "loss": 3.4999, + "step": 11182 + }, + { + "epoch": 0.5885789473684211, + "grad_norm": 0.5078125, + "learning_rate": 9.142690058479532e-05, + "loss": 3.6182, + "step": 11183 + }, + { + "epoch": 0.5886315789473684, + "grad_norm": 0.48046875, + "learning_rate": 9.141520467836257e-05, + "loss": 3.0308, + "step": 11184 + }, + { + "epoch": 0.5886842105263158, + "grad_norm": 0.4453125, + "learning_rate": 9.140350877192982e-05, + "loss": 3.4185, + "step": 11185 + }, + { + "epoch": 0.5887368421052631, + "grad_norm": 0.470703125, + "learning_rate": 9.139181286549709e-05, + "loss": 3.6832, + "step": 11186 + }, + { + "epoch": 0.5887894736842105, + "grad_norm": 0.462890625, + "learning_rate": 9.138011695906434e-05, + "loss": 3.5643, + "step": 11187 + }, + { + "epoch": 0.5888421052631579, + "grad_norm": 0.5234375, + "learning_rate": 9.136842105263159e-05, + "loss": 4.0254, + "step": 11188 + }, + { + "epoch": 0.5888947368421052, + "grad_norm": 0.51953125, + "learning_rate": 9.135672514619884e-05, + "loss": 3.5164, + "step": 11189 + }, + { + "epoch": 0.5889473684210527, + "grad_norm": 0.5546875, + "learning_rate": 9.134502923976609e-05, + "loss": 3.5291, + "step": 11190 + }, + { + "epoch": 0.589, + "grad_norm": 0.546875, + "learning_rate": 9.133333333333334e-05, + "loss": 3.3513, + "step": 11191 + }, + { + "epoch": 0.5890526315789474, + "grad_norm": 0.51171875, + "learning_rate": 9.132163742690058e-05, + "loss": 3.3302, + "step": 11192 + }, + { + "epoch": 0.5891052631578947, + "grad_norm": 0.53125, + "learning_rate": 9.130994152046785e-05, + "loss": 3.392, + "step": 11193 + }, + { + "epoch": 0.5891578947368421, + "grad_norm": 0.5546875, + "learning_rate": 9.12982456140351e-05, + "loss": 3.3506, + "step": 11194 + }, + { + "epoch": 0.5892105263157895, + "grad_norm": 0.494140625, + "learning_rate": 9.128654970760235e-05, + "loss": 3.5047, + "step": 11195 + }, + { + "epoch": 0.5892631578947368, + "grad_norm": 0.54296875, + "learning_rate": 9.12748538011696e-05, + "loss": 3.1097, + "step": 11196 + }, + { + "epoch": 0.5893157894736842, + "grad_norm": 0.478515625, + "learning_rate": 9.126315789473685e-05, + "loss": 3.5027, + "step": 11197 + }, + { + "epoch": 0.5893684210526315, + "grad_norm": 0.494140625, + "learning_rate": 9.12514619883041e-05, + "loss": 3.7941, + "step": 11198 + }, + { + "epoch": 0.589421052631579, + "grad_norm": 0.48046875, + "learning_rate": 9.123976608187135e-05, + "loss": 3.4856, + "step": 11199 + }, + { + "epoch": 0.5894736842105263, + "grad_norm": 0.515625, + "learning_rate": 9.12280701754386e-05, + "loss": 3.7176, + "step": 11200 + }, + { + "epoch": 0.5895263157894737, + "grad_norm": 0.4921875, + "learning_rate": 9.121637426900586e-05, + "loss": 3.2078, + "step": 11201 + }, + { + "epoch": 0.5895789473684211, + "grad_norm": 0.478515625, + "learning_rate": 9.120467836257311e-05, + "loss": 3.7992, + "step": 11202 + }, + { + "epoch": 0.5896315789473684, + "grad_norm": 0.462890625, + "learning_rate": 9.119298245614036e-05, + "loss": 3.2237, + "step": 11203 + }, + { + "epoch": 0.5896842105263158, + "grad_norm": 0.470703125, + "learning_rate": 9.118128654970761e-05, + "loss": 3.2295, + "step": 11204 + }, + { + "epoch": 0.5897368421052631, + "grad_norm": 0.482421875, + "learning_rate": 9.116959064327486e-05, + "loss": 3.4002, + "step": 11205 + }, + { + "epoch": 0.5897894736842105, + "grad_norm": 0.5078125, + "learning_rate": 9.11578947368421e-05, + "loss": 3.2014, + "step": 11206 + }, + { + "epoch": 0.589842105263158, + "grad_norm": 0.55859375, + "learning_rate": 9.114619883040936e-05, + "loss": 3.1386, + "step": 11207 + }, + { + "epoch": 0.5898947368421052, + "grad_norm": 0.51953125, + "learning_rate": 9.113450292397662e-05, + "loss": 3.7332, + "step": 11208 + }, + { + "epoch": 0.5899473684210527, + "grad_norm": 0.498046875, + "learning_rate": 9.112280701754387e-05, + "loss": 3.3596, + "step": 11209 + }, + { + "epoch": 0.59, + "grad_norm": 0.5234375, + "learning_rate": 9.111111111111112e-05, + "loss": 2.903, + "step": 11210 + }, + { + "epoch": 0.5900526315789474, + "grad_norm": 0.49609375, + "learning_rate": 9.109941520467837e-05, + "loss": 3.6217, + "step": 11211 + }, + { + "epoch": 0.5901052631578947, + "grad_norm": 0.458984375, + "learning_rate": 9.108771929824562e-05, + "loss": 3.1585, + "step": 11212 + }, + { + "epoch": 0.5901578947368421, + "grad_norm": 0.46484375, + "learning_rate": 9.107602339181287e-05, + "loss": 2.9854, + "step": 11213 + }, + { + "epoch": 0.5902105263157895, + "grad_norm": 0.498046875, + "learning_rate": 9.106432748538012e-05, + "loss": 3.1498, + "step": 11214 + }, + { + "epoch": 0.5902631578947368, + "grad_norm": 0.4921875, + "learning_rate": 9.105263157894738e-05, + "loss": 3.1168, + "step": 11215 + }, + { + "epoch": 0.5903157894736842, + "grad_norm": 0.486328125, + "learning_rate": 9.104093567251463e-05, + "loss": 3.5132, + "step": 11216 + }, + { + "epoch": 0.5903684210526315, + "grad_norm": 0.50390625, + "learning_rate": 9.102923976608188e-05, + "loss": 3.2937, + "step": 11217 + }, + { + "epoch": 0.590421052631579, + "grad_norm": 0.46875, + "learning_rate": 9.101754385964913e-05, + "loss": 3.7185, + "step": 11218 + }, + { + "epoch": 0.5904736842105263, + "grad_norm": 0.455078125, + "learning_rate": 9.100584795321638e-05, + "loss": 2.6227, + "step": 11219 + }, + { + "epoch": 0.5905263157894737, + "grad_norm": 0.474609375, + "learning_rate": 9.099415204678363e-05, + "loss": 3.3397, + "step": 11220 + }, + { + "epoch": 0.5905789473684211, + "grad_norm": 0.46484375, + "learning_rate": 9.098245614035088e-05, + "loss": 3.3791, + "step": 11221 + }, + { + "epoch": 0.5906315789473684, + "grad_norm": 0.49609375, + "learning_rate": 9.097076023391813e-05, + "loss": 3.5983, + "step": 11222 + }, + { + "epoch": 0.5906842105263158, + "grad_norm": 0.46875, + "learning_rate": 9.095906432748539e-05, + "loss": 3.3685, + "step": 11223 + }, + { + "epoch": 0.5907368421052631, + "grad_norm": 0.5078125, + "learning_rate": 9.094736842105264e-05, + "loss": 3.4174, + "step": 11224 + }, + { + "epoch": 0.5907894736842105, + "grad_norm": 0.478515625, + "learning_rate": 9.093567251461989e-05, + "loss": 3.6044, + "step": 11225 + }, + { + "epoch": 0.590842105263158, + "grad_norm": 0.486328125, + "learning_rate": 9.092397660818714e-05, + "loss": 3.4463, + "step": 11226 + }, + { + "epoch": 0.5908947368421053, + "grad_norm": 0.462890625, + "learning_rate": 9.091228070175439e-05, + "loss": 3.071, + "step": 11227 + }, + { + "epoch": 0.5909473684210527, + "grad_norm": 0.58203125, + "learning_rate": 9.090058479532164e-05, + "loss": 3.5256, + "step": 11228 + }, + { + "epoch": 0.591, + "grad_norm": 0.482421875, + "learning_rate": 9.088888888888889e-05, + "loss": 3.2607, + "step": 11229 + }, + { + "epoch": 0.5910526315789474, + "grad_norm": 0.484375, + "learning_rate": 9.087719298245615e-05, + "loss": 3.355, + "step": 11230 + }, + { + "epoch": 0.5911052631578947, + "grad_norm": 0.4921875, + "learning_rate": 9.08654970760234e-05, + "loss": 3.384, + "step": 11231 + }, + { + "epoch": 0.5911578947368421, + "grad_norm": 0.51953125, + "learning_rate": 9.085380116959065e-05, + "loss": 3.6616, + "step": 11232 + }, + { + "epoch": 0.5912105263157895, + "grad_norm": 0.478515625, + "learning_rate": 9.08421052631579e-05, + "loss": 3.315, + "step": 11233 + }, + { + "epoch": 0.5912631578947368, + "grad_norm": 0.45703125, + "learning_rate": 9.083040935672515e-05, + "loss": 3.6835, + "step": 11234 + }, + { + "epoch": 0.5913157894736842, + "grad_norm": 0.5859375, + "learning_rate": 9.08187134502924e-05, + "loss": 3.4355, + "step": 11235 + }, + { + "epoch": 0.5913684210526315, + "grad_norm": 0.515625, + "learning_rate": 9.080701754385965e-05, + "loss": 2.9566, + "step": 11236 + }, + { + "epoch": 0.591421052631579, + "grad_norm": 0.478515625, + "learning_rate": 9.07953216374269e-05, + "loss": 3.2724, + "step": 11237 + }, + { + "epoch": 0.5914736842105263, + "grad_norm": 0.455078125, + "learning_rate": 9.078362573099416e-05, + "loss": 3.0379, + "step": 11238 + }, + { + "epoch": 0.5915263157894737, + "grad_norm": 0.474609375, + "learning_rate": 9.077192982456141e-05, + "loss": 3.9166, + "step": 11239 + }, + { + "epoch": 0.5915789473684211, + "grad_norm": 0.53515625, + "learning_rate": 9.076023391812866e-05, + "loss": 3.1022, + "step": 11240 + }, + { + "epoch": 0.5916315789473684, + "grad_norm": 0.4765625, + "learning_rate": 9.074853801169591e-05, + "loss": 3.6678, + "step": 11241 + }, + { + "epoch": 0.5916842105263158, + "grad_norm": 0.451171875, + "learning_rate": 9.073684210526316e-05, + "loss": 3.3924, + "step": 11242 + }, + { + "epoch": 0.5917368421052631, + "grad_norm": 0.52734375, + "learning_rate": 9.072514619883041e-05, + "loss": 4.0352, + "step": 11243 + }, + { + "epoch": 0.5917894736842105, + "grad_norm": 0.5234375, + "learning_rate": 9.071345029239766e-05, + "loss": 3.7435, + "step": 11244 + }, + { + "epoch": 0.591842105263158, + "grad_norm": 0.515625, + "learning_rate": 9.070175438596492e-05, + "loss": 3.1023, + "step": 11245 + }, + { + "epoch": 0.5918947368421053, + "grad_norm": 0.54296875, + "learning_rate": 9.069005847953217e-05, + "loss": 3.4411, + "step": 11246 + }, + { + "epoch": 0.5919473684210527, + "grad_norm": 0.515625, + "learning_rate": 9.067836257309942e-05, + "loss": 3.5685, + "step": 11247 + }, + { + "epoch": 0.592, + "grad_norm": 0.5234375, + "learning_rate": 9.066666666666667e-05, + "loss": 3.4136, + "step": 11248 + }, + { + "epoch": 0.5920526315789474, + "grad_norm": 0.5234375, + "learning_rate": 9.065497076023392e-05, + "loss": 3.4466, + "step": 11249 + }, + { + "epoch": 0.5921052631578947, + "grad_norm": 0.515625, + "learning_rate": 9.064327485380117e-05, + "loss": 2.9803, + "step": 11250 + }, + { + "epoch": 0.5921578947368421, + "grad_norm": 0.53125, + "learning_rate": 9.063157894736842e-05, + "loss": 3.1306, + "step": 11251 + }, + { + "epoch": 0.5922105263157895, + "grad_norm": 0.466796875, + "learning_rate": 9.061988304093567e-05, + "loss": 3.3085, + "step": 11252 + }, + { + "epoch": 0.5922631578947368, + "grad_norm": 0.46484375, + "learning_rate": 9.060818713450293e-05, + "loss": 3.3439, + "step": 11253 + }, + { + "epoch": 0.5923157894736842, + "grad_norm": 0.47265625, + "learning_rate": 9.059649122807018e-05, + "loss": 3.6845, + "step": 11254 + }, + { + "epoch": 0.5923684210526315, + "grad_norm": 0.494140625, + "learning_rate": 9.058479532163743e-05, + "loss": 3.6779, + "step": 11255 + }, + { + "epoch": 0.592421052631579, + "grad_norm": 0.5390625, + "learning_rate": 9.057309941520468e-05, + "loss": 3.4555, + "step": 11256 + }, + { + "epoch": 0.5924736842105263, + "grad_norm": 0.515625, + "learning_rate": 9.056140350877193e-05, + "loss": 3.2512, + "step": 11257 + }, + { + "epoch": 0.5925263157894737, + "grad_norm": 0.478515625, + "learning_rate": 9.054970760233918e-05, + "loss": 3.6076, + "step": 11258 + }, + { + "epoch": 0.5925789473684211, + "grad_norm": 0.5078125, + "learning_rate": 9.053801169590643e-05, + "loss": 3.4317, + "step": 11259 + }, + { + "epoch": 0.5926315789473684, + "grad_norm": 0.482421875, + "learning_rate": 9.052631578947369e-05, + "loss": 3.3303, + "step": 11260 + }, + { + "epoch": 0.5926842105263158, + "grad_norm": 0.47265625, + "learning_rate": 9.051461988304094e-05, + "loss": 3.4843, + "step": 11261 + }, + { + "epoch": 0.5927368421052631, + "grad_norm": 0.482421875, + "learning_rate": 9.050292397660819e-05, + "loss": 3.6924, + "step": 11262 + }, + { + "epoch": 0.5927894736842105, + "grad_norm": 0.490234375, + "learning_rate": 9.049122807017544e-05, + "loss": 2.9682, + "step": 11263 + }, + { + "epoch": 0.592842105263158, + "grad_norm": 0.462890625, + "learning_rate": 9.047953216374269e-05, + "loss": 3.6194, + "step": 11264 + }, + { + "epoch": 0.5928947368421053, + "grad_norm": 0.546875, + "learning_rate": 9.046783625730994e-05, + "loss": 3.221, + "step": 11265 + }, + { + "epoch": 0.5929473684210527, + "grad_norm": 0.45703125, + "learning_rate": 9.045614035087719e-05, + "loss": 3.0636, + "step": 11266 + }, + { + "epoch": 0.593, + "grad_norm": 0.5078125, + "learning_rate": 9.044444444444445e-05, + "loss": 3.3112, + "step": 11267 + }, + { + "epoch": 0.5930526315789474, + "grad_norm": 0.51953125, + "learning_rate": 9.04327485380117e-05, + "loss": 3.3503, + "step": 11268 + }, + { + "epoch": 0.5931052631578947, + "grad_norm": 0.5, + "learning_rate": 9.042105263157895e-05, + "loss": 2.6886, + "step": 11269 + }, + { + "epoch": 0.5931578947368421, + "grad_norm": 0.59375, + "learning_rate": 9.04093567251462e-05, + "loss": 3.8804, + "step": 11270 + }, + { + "epoch": 0.5932105263157895, + "grad_norm": 0.474609375, + "learning_rate": 9.039766081871347e-05, + "loss": 3.4397, + "step": 11271 + }, + { + "epoch": 0.5932631578947368, + "grad_norm": 0.50390625, + "learning_rate": 9.03859649122807e-05, + "loss": 3.154, + "step": 11272 + }, + { + "epoch": 0.5933157894736842, + "grad_norm": 0.498046875, + "learning_rate": 9.037426900584795e-05, + "loss": 3.8363, + "step": 11273 + }, + { + "epoch": 0.5933684210526315, + "grad_norm": 0.53125, + "learning_rate": 9.03625730994152e-05, + "loss": 3.62, + "step": 11274 + }, + { + "epoch": 0.593421052631579, + "grad_norm": 0.61328125, + "learning_rate": 9.035087719298246e-05, + "loss": 3.6093, + "step": 11275 + }, + { + "epoch": 0.5934736842105263, + "grad_norm": 0.466796875, + "learning_rate": 9.033918128654971e-05, + "loss": 3.1803, + "step": 11276 + }, + { + "epoch": 0.5935263157894737, + "grad_norm": 0.498046875, + "learning_rate": 9.032748538011696e-05, + "loss": 2.9839, + "step": 11277 + }, + { + "epoch": 0.5935789473684211, + "grad_norm": 0.48828125, + "learning_rate": 9.031578947368423e-05, + "loss": 3.0297, + "step": 11278 + }, + { + "epoch": 0.5936315789473684, + "grad_norm": 0.494140625, + "learning_rate": 9.030409356725146e-05, + "loss": 3.1627, + "step": 11279 + }, + { + "epoch": 0.5936842105263158, + "grad_norm": 0.47265625, + "learning_rate": 9.029239766081871e-05, + "loss": 3.0821, + "step": 11280 + }, + { + "epoch": 0.5937368421052631, + "grad_norm": 0.478515625, + "learning_rate": 9.028070175438596e-05, + "loss": 3.2695, + "step": 11281 + }, + { + "epoch": 0.5937894736842105, + "grad_norm": 0.47265625, + "learning_rate": 9.026900584795322e-05, + "loss": 3.7114, + "step": 11282 + }, + { + "epoch": 0.593842105263158, + "grad_norm": 0.474609375, + "learning_rate": 9.025730994152047e-05, + "loss": 3.6704, + "step": 11283 + }, + { + "epoch": 0.5938947368421053, + "grad_norm": 0.43359375, + "learning_rate": 9.024561403508772e-05, + "loss": 3.4976, + "step": 11284 + }, + { + "epoch": 0.5939473684210527, + "grad_norm": 0.482421875, + "learning_rate": 9.023391812865497e-05, + "loss": 3.4662, + "step": 11285 + }, + { + "epoch": 0.594, + "grad_norm": 0.494140625, + "learning_rate": 9.022222222222224e-05, + "loss": 3.7393, + "step": 11286 + }, + { + "epoch": 0.5940526315789474, + "grad_norm": 0.482421875, + "learning_rate": 9.021052631578947e-05, + "loss": 3.2095, + "step": 11287 + }, + { + "epoch": 0.5941052631578947, + "grad_norm": 0.482421875, + "learning_rate": 9.019883040935672e-05, + "loss": 3.5066, + "step": 11288 + }, + { + "epoch": 0.5941578947368421, + "grad_norm": 0.4765625, + "learning_rate": 9.018713450292397e-05, + "loss": 3.6732, + "step": 11289 + }, + { + "epoch": 0.5942105263157895, + "grad_norm": 0.4921875, + "learning_rate": 9.017543859649123e-05, + "loss": 3.2969, + "step": 11290 + }, + { + "epoch": 0.5942631578947368, + "grad_norm": 0.5234375, + "learning_rate": 9.016374269005848e-05, + "loss": 3.319, + "step": 11291 + }, + { + "epoch": 0.5943157894736842, + "grad_norm": 0.49609375, + "learning_rate": 9.015204678362573e-05, + "loss": 3.4195, + "step": 11292 + }, + { + "epoch": 0.5943684210526315, + "grad_norm": 0.5390625, + "learning_rate": 9.0140350877193e-05, + "loss": 3.1144, + "step": 11293 + }, + { + "epoch": 0.594421052631579, + "grad_norm": 0.56640625, + "learning_rate": 9.012865497076025e-05, + "loss": 3.3862, + "step": 11294 + }, + { + "epoch": 0.5944736842105263, + "grad_norm": 0.498046875, + "learning_rate": 9.011695906432748e-05, + "loss": 3.3301, + "step": 11295 + }, + { + "epoch": 0.5945263157894737, + "grad_norm": 0.50390625, + "learning_rate": 9.010526315789473e-05, + "loss": 3.1673, + "step": 11296 + }, + { + "epoch": 0.5945789473684211, + "grad_norm": 0.53515625, + "learning_rate": 9.0093567251462e-05, + "loss": 3.1111, + "step": 11297 + }, + { + "epoch": 0.5946315789473684, + "grad_norm": 0.46875, + "learning_rate": 9.008187134502924e-05, + "loss": 3.0939, + "step": 11298 + }, + { + "epoch": 0.5946842105263158, + "grad_norm": 0.4921875, + "learning_rate": 9.00701754385965e-05, + "loss": 3.0048, + "step": 11299 + }, + { + "epoch": 0.5947368421052631, + "grad_norm": 0.484375, + "learning_rate": 9.005847953216374e-05, + "loss": 3.4247, + "step": 11300 + }, + { + "epoch": 0.5947894736842105, + "grad_norm": 0.52734375, + "learning_rate": 9.004678362573101e-05, + "loss": 3.6996, + "step": 11301 + }, + { + "epoch": 0.594842105263158, + "grad_norm": 0.54296875, + "learning_rate": 9.003508771929824e-05, + "loss": 2.8738, + "step": 11302 + }, + { + "epoch": 0.5948947368421053, + "grad_norm": 0.77734375, + "learning_rate": 9.002339181286549e-05, + "loss": 3.3058, + "step": 11303 + }, + { + "epoch": 0.5949473684210527, + "grad_norm": 0.46875, + "learning_rate": 9.001169590643276e-05, + "loss": 3.2109, + "step": 11304 + }, + { + "epoch": 0.595, + "grad_norm": 0.478515625, + "learning_rate": 9e-05, + "loss": 3.2061, + "step": 11305 + }, + { + "epoch": 0.5950526315789474, + "grad_norm": 0.46875, + "learning_rate": 8.998830409356726e-05, + "loss": 3.277, + "step": 11306 + }, + { + "epoch": 0.5951052631578947, + "grad_norm": 0.482421875, + "learning_rate": 8.99766081871345e-05, + "loss": 3.5545, + "step": 11307 + }, + { + "epoch": 0.5951578947368421, + "grad_norm": 0.53125, + "learning_rate": 8.996491228070177e-05, + "loss": 2.6614, + "step": 11308 + }, + { + "epoch": 0.5952105263157895, + "grad_norm": 0.474609375, + "learning_rate": 8.995321637426902e-05, + "loss": 3.6391, + "step": 11309 + }, + { + "epoch": 0.5952631578947368, + "grad_norm": 0.494140625, + "learning_rate": 8.994152046783625e-05, + "loss": 2.9665, + "step": 11310 + }, + { + "epoch": 0.5953157894736842, + "grad_norm": 0.474609375, + "learning_rate": 8.99298245614035e-05, + "loss": 3.8648, + "step": 11311 + }, + { + "epoch": 0.5953684210526315, + "grad_norm": 0.50390625, + "learning_rate": 8.991812865497077e-05, + "loss": 3.398, + "step": 11312 + }, + { + "epoch": 0.595421052631579, + "grad_norm": 0.470703125, + "learning_rate": 8.990643274853802e-05, + "loss": 3.7447, + "step": 11313 + }, + { + "epoch": 0.5954736842105263, + "grad_norm": 0.490234375, + "learning_rate": 8.989473684210527e-05, + "loss": 3.5004, + "step": 11314 + }, + { + "epoch": 0.5955263157894737, + "grad_norm": 0.50390625, + "learning_rate": 8.988304093567252e-05, + "loss": 3.5913, + "step": 11315 + }, + { + "epoch": 0.5955789473684211, + "grad_norm": 0.51953125, + "learning_rate": 8.987134502923978e-05, + "loss": 3.1488, + "step": 11316 + }, + { + "epoch": 0.5956315789473684, + "grad_norm": 0.69140625, + "learning_rate": 8.985964912280703e-05, + "loss": 3.3262, + "step": 11317 + }, + { + "epoch": 0.5956842105263158, + "grad_norm": 0.494140625, + "learning_rate": 8.984795321637426e-05, + "loss": 3.0067, + "step": 11318 + }, + { + "epoch": 0.5957368421052631, + "grad_norm": 0.50390625, + "learning_rate": 8.983625730994153e-05, + "loss": 3.4219, + "step": 11319 + }, + { + "epoch": 0.5957894736842105, + "grad_norm": 0.453125, + "learning_rate": 8.982456140350878e-05, + "loss": 3.063, + "step": 11320 + }, + { + "epoch": 0.5958421052631578, + "grad_norm": 0.494140625, + "learning_rate": 8.981286549707603e-05, + "loss": 3.3406, + "step": 11321 + }, + { + "epoch": 0.5958947368421053, + "grad_norm": 0.470703125, + "learning_rate": 8.980116959064328e-05, + "loss": 3.6262, + "step": 11322 + }, + { + "epoch": 0.5959473684210527, + "grad_norm": 0.5625, + "learning_rate": 8.978947368421054e-05, + "loss": 3.314, + "step": 11323 + }, + { + "epoch": 0.596, + "grad_norm": 0.482421875, + "learning_rate": 8.977777777777779e-05, + "loss": 3.2762, + "step": 11324 + }, + { + "epoch": 0.5960526315789474, + "grad_norm": 0.486328125, + "learning_rate": 8.976608187134502e-05, + "loss": 3.5029, + "step": 11325 + }, + { + "epoch": 0.5961052631578947, + "grad_norm": 0.490234375, + "learning_rate": 8.975438596491227e-05, + "loss": 3.3102, + "step": 11326 + }, + { + "epoch": 0.5961578947368421, + "grad_norm": 0.6484375, + "learning_rate": 8.974269005847954e-05, + "loss": 3.4869, + "step": 11327 + }, + { + "epoch": 0.5962105263157895, + "grad_norm": 0.5, + "learning_rate": 8.973099415204679e-05, + "loss": 3.3253, + "step": 11328 + }, + { + "epoch": 0.5962631578947368, + "grad_norm": 0.455078125, + "learning_rate": 8.971929824561404e-05, + "loss": 3.1519, + "step": 11329 + }, + { + "epoch": 0.5963157894736842, + "grad_norm": 0.515625, + "learning_rate": 8.97076023391813e-05, + "loss": 3.8157, + "step": 11330 + }, + { + "epoch": 0.5963684210526315, + "grad_norm": 0.48046875, + "learning_rate": 8.969590643274855e-05, + "loss": 3.5981, + "step": 11331 + }, + { + "epoch": 0.596421052631579, + "grad_norm": 0.4765625, + "learning_rate": 8.96842105263158e-05, + "loss": 3.3609, + "step": 11332 + }, + { + "epoch": 0.5964736842105263, + "grad_norm": 0.5078125, + "learning_rate": 8.967251461988303e-05, + "loss": 3.4015, + "step": 11333 + }, + { + "epoch": 0.5965263157894737, + "grad_norm": 0.482421875, + "learning_rate": 8.96608187134503e-05, + "loss": 3.4275, + "step": 11334 + }, + { + "epoch": 0.5965789473684211, + "grad_norm": 0.482421875, + "learning_rate": 8.964912280701755e-05, + "loss": 3.4499, + "step": 11335 + }, + { + "epoch": 0.5966315789473684, + "grad_norm": 0.486328125, + "learning_rate": 8.96374269005848e-05, + "loss": 3.2958, + "step": 11336 + }, + { + "epoch": 0.5966842105263158, + "grad_norm": 0.5078125, + "learning_rate": 8.962573099415205e-05, + "loss": 3.4733, + "step": 11337 + }, + { + "epoch": 0.5967368421052631, + "grad_norm": 0.490234375, + "learning_rate": 8.961403508771931e-05, + "loss": 3.5086, + "step": 11338 + }, + { + "epoch": 0.5967894736842105, + "grad_norm": 0.494140625, + "learning_rate": 8.960233918128656e-05, + "loss": 3.3229, + "step": 11339 + }, + { + "epoch": 0.5968421052631578, + "grad_norm": 0.4765625, + "learning_rate": 8.959064327485381e-05, + "loss": 3.5361, + "step": 11340 + }, + { + "epoch": 0.5968947368421053, + "grad_norm": 0.56640625, + "learning_rate": 8.957894736842106e-05, + "loss": 3.4768, + "step": 11341 + }, + { + "epoch": 0.5969473684210527, + "grad_norm": 0.52734375, + "learning_rate": 8.956725146198831e-05, + "loss": 3.0315, + "step": 11342 + }, + { + "epoch": 0.597, + "grad_norm": 0.52734375, + "learning_rate": 8.955555555555556e-05, + "loss": 3.4504, + "step": 11343 + }, + { + "epoch": 0.5970526315789474, + "grad_norm": 0.494140625, + "learning_rate": 8.954385964912281e-05, + "loss": 3.4744, + "step": 11344 + }, + { + "epoch": 0.5971052631578947, + "grad_norm": 0.50390625, + "learning_rate": 8.953216374269007e-05, + "loss": 3.3279, + "step": 11345 + }, + { + "epoch": 0.5971578947368421, + "grad_norm": 0.48828125, + "learning_rate": 8.952046783625732e-05, + "loss": 3.6529, + "step": 11346 + }, + { + "epoch": 0.5972105263157895, + "grad_norm": 0.494140625, + "learning_rate": 8.950877192982457e-05, + "loss": 3.3396, + "step": 11347 + }, + { + "epoch": 0.5972631578947368, + "grad_norm": 0.498046875, + "learning_rate": 8.94970760233918e-05, + "loss": 3.204, + "step": 11348 + }, + { + "epoch": 0.5973157894736842, + "grad_norm": 0.53515625, + "learning_rate": 8.948538011695907e-05, + "loss": 3.5988, + "step": 11349 + }, + { + "epoch": 0.5973684210526315, + "grad_norm": 0.5078125, + "learning_rate": 8.947368421052632e-05, + "loss": 3.3969, + "step": 11350 + }, + { + "epoch": 0.597421052631579, + "grad_norm": 0.8671875, + "learning_rate": 8.946198830409357e-05, + "loss": 3.5544, + "step": 11351 + }, + { + "epoch": 0.5974736842105263, + "grad_norm": 0.458984375, + "learning_rate": 8.945029239766082e-05, + "loss": 3.5542, + "step": 11352 + }, + { + "epoch": 0.5975263157894737, + "grad_norm": 0.482421875, + "learning_rate": 8.943859649122808e-05, + "loss": 3.5835, + "step": 11353 + }, + { + "epoch": 0.5975789473684211, + "grad_norm": 0.51171875, + "learning_rate": 8.942690058479533e-05, + "loss": 3.3824, + "step": 11354 + }, + { + "epoch": 0.5976315789473684, + "grad_norm": 0.46484375, + "learning_rate": 8.941520467836258e-05, + "loss": 3.4292, + "step": 11355 + }, + { + "epoch": 0.5976842105263158, + "grad_norm": 0.51171875, + "learning_rate": 8.940350877192983e-05, + "loss": 3.5878, + "step": 11356 + }, + { + "epoch": 0.5977368421052631, + "grad_norm": 0.474609375, + "learning_rate": 8.939181286549708e-05, + "loss": 3.4099, + "step": 11357 + }, + { + "epoch": 0.5977894736842105, + "grad_norm": 0.490234375, + "learning_rate": 8.938011695906433e-05, + "loss": 3.2925, + "step": 11358 + }, + { + "epoch": 0.5978421052631578, + "grad_norm": 0.48828125, + "learning_rate": 8.936842105263158e-05, + "loss": 3.4847, + "step": 11359 + }, + { + "epoch": 0.5978947368421053, + "grad_norm": 0.515625, + "learning_rate": 8.935672514619884e-05, + "loss": 3.5488, + "step": 11360 + }, + { + "epoch": 0.5979473684210527, + "grad_norm": 0.51171875, + "learning_rate": 8.934502923976609e-05, + "loss": 3.4951, + "step": 11361 + }, + { + "epoch": 0.598, + "grad_norm": 0.5, + "learning_rate": 8.933333333333334e-05, + "loss": 3.4447, + "step": 11362 + }, + { + "epoch": 0.5980526315789474, + "grad_norm": 0.48828125, + "learning_rate": 8.932163742690059e-05, + "loss": 3.3833, + "step": 11363 + }, + { + "epoch": 0.5981052631578947, + "grad_norm": 0.4765625, + "learning_rate": 8.930994152046784e-05, + "loss": 3.3785, + "step": 11364 + }, + { + "epoch": 0.5981578947368421, + "grad_norm": 0.490234375, + "learning_rate": 8.929824561403509e-05, + "loss": 3.3028, + "step": 11365 + }, + { + "epoch": 0.5982105263157895, + "grad_norm": 0.494140625, + "learning_rate": 8.928654970760234e-05, + "loss": 3.1696, + "step": 11366 + }, + { + "epoch": 0.5982631578947368, + "grad_norm": 0.49609375, + "learning_rate": 8.92748538011696e-05, + "loss": 3.3933, + "step": 11367 + }, + { + "epoch": 0.5983157894736842, + "grad_norm": 0.5390625, + "learning_rate": 8.926315789473685e-05, + "loss": 2.8961, + "step": 11368 + }, + { + "epoch": 0.5983684210526315, + "grad_norm": 0.484375, + "learning_rate": 8.92514619883041e-05, + "loss": 3.1683, + "step": 11369 + }, + { + "epoch": 0.598421052631579, + "grad_norm": 0.48046875, + "learning_rate": 8.923976608187135e-05, + "loss": 3.4987, + "step": 11370 + }, + { + "epoch": 0.5984736842105263, + "grad_norm": 0.5078125, + "learning_rate": 8.92280701754386e-05, + "loss": 3.4294, + "step": 11371 + }, + { + "epoch": 0.5985263157894737, + "grad_norm": 0.494140625, + "learning_rate": 8.921637426900585e-05, + "loss": 3.4325, + "step": 11372 + }, + { + "epoch": 0.5985789473684211, + "grad_norm": 0.51171875, + "learning_rate": 8.92046783625731e-05, + "loss": 3.3505, + "step": 11373 + }, + { + "epoch": 0.5986315789473684, + "grad_norm": 0.474609375, + "learning_rate": 8.919298245614035e-05, + "loss": 3.1257, + "step": 11374 + }, + { + "epoch": 0.5986842105263158, + "grad_norm": 0.484375, + "learning_rate": 8.918128654970761e-05, + "loss": 3.577, + "step": 11375 + }, + { + "epoch": 0.5987368421052631, + "grad_norm": 0.515625, + "learning_rate": 8.916959064327486e-05, + "loss": 3.4983, + "step": 11376 + }, + { + "epoch": 0.5987894736842105, + "grad_norm": 0.5078125, + "learning_rate": 8.915789473684211e-05, + "loss": 3.6747, + "step": 11377 + }, + { + "epoch": 0.5988421052631578, + "grad_norm": 0.51953125, + "learning_rate": 8.914619883040936e-05, + "loss": 3.4803, + "step": 11378 + }, + { + "epoch": 0.5988947368421053, + "grad_norm": 0.4765625, + "learning_rate": 8.913450292397661e-05, + "loss": 3.8002, + "step": 11379 + }, + { + "epoch": 0.5989473684210527, + "grad_norm": 0.609375, + "learning_rate": 8.912280701754386e-05, + "loss": 3.466, + "step": 11380 + }, + { + "epoch": 0.599, + "grad_norm": 0.5078125, + "learning_rate": 8.911111111111111e-05, + "loss": 3.3216, + "step": 11381 + }, + { + "epoch": 0.5990526315789474, + "grad_norm": 0.51953125, + "learning_rate": 8.909941520467837e-05, + "loss": 3.2741, + "step": 11382 + }, + { + "epoch": 0.5991052631578947, + "grad_norm": 0.5078125, + "learning_rate": 8.908771929824562e-05, + "loss": 2.9544, + "step": 11383 + }, + { + "epoch": 0.5991578947368421, + "grad_norm": 0.5, + "learning_rate": 8.907602339181287e-05, + "loss": 3.3074, + "step": 11384 + }, + { + "epoch": 0.5992105263157895, + "grad_norm": 0.46875, + "learning_rate": 8.906432748538012e-05, + "loss": 3.3446, + "step": 11385 + }, + { + "epoch": 0.5992631578947368, + "grad_norm": 0.4765625, + "learning_rate": 8.905263157894737e-05, + "loss": 3.5347, + "step": 11386 + }, + { + "epoch": 0.5993157894736842, + "grad_norm": 0.478515625, + "learning_rate": 8.904093567251462e-05, + "loss": 3.3515, + "step": 11387 + }, + { + "epoch": 0.5993684210526316, + "grad_norm": 0.4453125, + "learning_rate": 8.902923976608187e-05, + "loss": 3.609, + "step": 11388 + }, + { + "epoch": 0.599421052631579, + "grad_norm": 0.458984375, + "learning_rate": 8.901754385964912e-05, + "loss": 3.2645, + "step": 11389 + }, + { + "epoch": 0.5994736842105263, + "grad_norm": 0.51171875, + "learning_rate": 8.900584795321638e-05, + "loss": 3.188, + "step": 11390 + }, + { + "epoch": 0.5995263157894737, + "grad_norm": 0.51953125, + "learning_rate": 8.899415204678363e-05, + "loss": 3.4064, + "step": 11391 + }, + { + "epoch": 0.5995789473684211, + "grad_norm": 0.458984375, + "learning_rate": 8.898245614035088e-05, + "loss": 3.2984, + "step": 11392 + }, + { + "epoch": 0.5996315789473684, + "grad_norm": 0.494140625, + "learning_rate": 8.897076023391813e-05, + "loss": 3.2525, + "step": 11393 + }, + { + "epoch": 0.5996842105263158, + "grad_norm": 0.498046875, + "learning_rate": 8.895906432748538e-05, + "loss": 3.2079, + "step": 11394 + }, + { + "epoch": 0.5997368421052631, + "grad_norm": 0.478515625, + "learning_rate": 8.894736842105263e-05, + "loss": 3.7638, + "step": 11395 + }, + { + "epoch": 0.5997894736842105, + "grad_norm": 0.5078125, + "learning_rate": 8.893567251461988e-05, + "loss": 3.2015, + "step": 11396 + }, + { + "epoch": 0.5998421052631578, + "grad_norm": 0.515625, + "learning_rate": 8.892397660818714e-05, + "loss": 3.6026, + "step": 11397 + }, + { + "epoch": 0.5998947368421053, + "grad_norm": 0.48828125, + "learning_rate": 8.89122807017544e-05, + "loss": 3.5972, + "step": 11398 + }, + { + "epoch": 0.5999473684210527, + "grad_norm": 0.52734375, + "learning_rate": 8.890058479532164e-05, + "loss": 3.0909, + "step": 11399 + }, + { + "epoch": 0.6, + "grad_norm": 0.53125, + "learning_rate": 8.888888888888889e-05, + "loss": 3.0247, + "step": 11400 + }, + { + "epoch": 0.6000526315789474, + "grad_norm": 0.486328125, + "learning_rate": 8.887719298245614e-05, + "loss": 3.5798, + "step": 11401 + }, + { + "epoch": 0.6001052631578947, + "grad_norm": 0.48046875, + "learning_rate": 8.886549707602339e-05, + "loss": 3.4503, + "step": 11402 + }, + { + "epoch": 0.6001578947368421, + "grad_norm": 0.48046875, + "learning_rate": 8.885380116959064e-05, + "loss": 3.5274, + "step": 11403 + }, + { + "epoch": 0.6002105263157895, + "grad_norm": 0.494140625, + "learning_rate": 8.88421052631579e-05, + "loss": 3.5885, + "step": 11404 + }, + { + "epoch": 0.6002631578947368, + "grad_norm": 0.447265625, + "learning_rate": 8.883040935672515e-05, + "loss": 3.1259, + "step": 11405 + }, + { + "epoch": 0.6003157894736842, + "grad_norm": 0.49609375, + "learning_rate": 8.88187134502924e-05, + "loss": 3.4083, + "step": 11406 + }, + { + "epoch": 0.6003684210526316, + "grad_norm": 0.515625, + "learning_rate": 8.880701754385965e-05, + "loss": 3.4577, + "step": 11407 + }, + { + "epoch": 0.600421052631579, + "grad_norm": 0.49609375, + "learning_rate": 8.87953216374269e-05, + "loss": 3.3806, + "step": 11408 + }, + { + "epoch": 0.6004736842105263, + "grad_norm": 0.49609375, + "learning_rate": 8.878362573099415e-05, + "loss": 3.279, + "step": 11409 + }, + { + "epoch": 0.6005263157894737, + "grad_norm": 0.490234375, + "learning_rate": 8.87719298245614e-05, + "loss": 3.4492, + "step": 11410 + }, + { + "epoch": 0.6005789473684211, + "grad_norm": 0.51171875, + "learning_rate": 8.876023391812865e-05, + "loss": 3.6047, + "step": 11411 + }, + { + "epoch": 0.6006315789473684, + "grad_norm": 0.5078125, + "learning_rate": 8.874853801169592e-05, + "loss": 3.4005, + "step": 11412 + }, + { + "epoch": 0.6006842105263158, + "grad_norm": 0.50390625, + "learning_rate": 8.873684210526316e-05, + "loss": 3.2938, + "step": 11413 + }, + { + "epoch": 0.6007368421052631, + "grad_norm": 0.53515625, + "learning_rate": 8.872514619883041e-05, + "loss": 3.7283, + "step": 11414 + }, + { + "epoch": 0.6007894736842105, + "grad_norm": 0.5390625, + "learning_rate": 8.871345029239766e-05, + "loss": 3.1781, + "step": 11415 + }, + { + "epoch": 0.6008421052631578, + "grad_norm": 0.5859375, + "learning_rate": 8.870175438596491e-05, + "loss": 3.348, + "step": 11416 + }, + { + "epoch": 0.6008947368421053, + "grad_norm": 0.5, + "learning_rate": 8.869005847953216e-05, + "loss": 3.0722, + "step": 11417 + }, + { + "epoch": 0.6009473684210527, + "grad_norm": 0.4921875, + "learning_rate": 8.867836257309941e-05, + "loss": 3.4611, + "step": 11418 + }, + { + "epoch": 0.601, + "grad_norm": 0.486328125, + "learning_rate": 8.866666666666668e-05, + "loss": 3.2253, + "step": 11419 + }, + { + "epoch": 0.6010526315789474, + "grad_norm": 0.54296875, + "learning_rate": 8.865497076023393e-05, + "loss": 3.6114, + "step": 11420 + }, + { + "epoch": 0.6011052631578947, + "grad_norm": 0.5078125, + "learning_rate": 8.864327485380118e-05, + "loss": 3.239, + "step": 11421 + }, + { + "epoch": 0.6011578947368421, + "grad_norm": 0.494140625, + "learning_rate": 8.863157894736842e-05, + "loss": 3.1039, + "step": 11422 + }, + { + "epoch": 0.6012105263157895, + "grad_norm": 0.50390625, + "learning_rate": 8.861988304093567e-05, + "loss": 3.3002, + "step": 11423 + }, + { + "epoch": 0.6012631578947368, + "grad_norm": 0.52734375, + "learning_rate": 8.860818713450292e-05, + "loss": 3.3671, + "step": 11424 + }, + { + "epoch": 0.6013157894736842, + "grad_norm": 0.45703125, + "learning_rate": 8.859649122807017e-05, + "loss": 3.1465, + "step": 11425 + }, + { + "epoch": 0.6013684210526316, + "grad_norm": 0.48828125, + "learning_rate": 8.858479532163742e-05, + "loss": 2.9524, + "step": 11426 + }, + { + "epoch": 0.601421052631579, + "grad_norm": 0.458984375, + "learning_rate": 8.857309941520469e-05, + "loss": 3.5449, + "step": 11427 + }, + { + "epoch": 0.6014736842105263, + "grad_norm": 0.52734375, + "learning_rate": 8.856140350877194e-05, + "loss": 3.2923, + "step": 11428 + }, + { + "epoch": 0.6015263157894737, + "grad_norm": 0.498046875, + "learning_rate": 8.854970760233919e-05, + "loss": 3.6344, + "step": 11429 + }, + { + "epoch": 0.6015789473684211, + "grad_norm": 0.490234375, + "learning_rate": 8.853801169590645e-05, + "loss": 3.3403, + "step": 11430 + }, + { + "epoch": 0.6016315789473684, + "grad_norm": 0.486328125, + "learning_rate": 8.852631578947368e-05, + "loss": 3.6929, + "step": 11431 + }, + { + "epoch": 0.6016842105263158, + "grad_norm": 0.4921875, + "learning_rate": 8.851461988304093e-05, + "loss": 3.2366, + "step": 11432 + }, + { + "epoch": 0.6017368421052631, + "grad_norm": 0.4609375, + "learning_rate": 8.850292397660818e-05, + "loss": 3.8394, + "step": 11433 + }, + { + "epoch": 0.6017894736842105, + "grad_norm": 0.49609375, + "learning_rate": 8.849122807017545e-05, + "loss": 3.1945, + "step": 11434 + }, + { + "epoch": 0.6018421052631578, + "grad_norm": 0.498046875, + "learning_rate": 8.84795321637427e-05, + "loss": 3.3294, + "step": 11435 + }, + { + "epoch": 0.6018947368421053, + "grad_norm": 0.484375, + "learning_rate": 8.846783625730995e-05, + "loss": 3.5165, + "step": 11436 + }, + { + "epoch": 0.6019473684210527, + "grad_norm": 0.458984375, + "learning_rate": 8.84561403508772e-05, + "loss": 3.8288, + "step": 11437 + }, + { + "epoch": 0.602, + "grad_norm": 0.474609375, + "learning_rate": 8.844444444444445e-05, + "loss": 3.534, + "step": 11438 + }, + { + "epoch": 0.6020526315789474, + "grad_norm": 0.462890625, + "learning_rate": 8.84327485380117e-05, + "loss": 3.3478, + "step": 11439 + }, + { + "epoch": 0.6021052631578947, + "grad_norm": 0.494140625, + "learning_rate": 8.842105263157894e-05, + "loss": 3.1163, + "step": 11440 + }, + { + "epoch": 0.6021578947368421, + "grad_norm": 0.490234375, + "learning_rate": 8.840935672514621e-05, + "loss": 3.5194, + "step": 11441 + }, + { + "epoch": 0.6022105263157894, + "grad_norm": 0.453125, + "learning_rate": 8.839766081871346e-05, + "loss": 3.4495, + "step": 11442 + }, + { + "epoch": 0.6022631578947368, + "grad_norm": 0.48828125, + "learning_rate": 8.838596491228071e-05, + "loss": 3.0607, + "step": 11443 + }, + { + "epoch": 0.6023157894736842, + "grad_norm": 0.47265625, + "learning_rate": 8.837426900584796e-05, + "loss": 3.683, + "step": 11444 + }, + { + "epoch": 0.6023684210526316, + "grad_norm": 0.482421875, + "learning_rate": 8.836257309941522e-05, + "loss": 3.6229, + "step": 11445 + }, + { + "epoch": 0.602421052631579, + "grad_norm": 0.478515625, + "learning_rate": 8.835087719298246e-05, + "loss": 3.2545, + "step": 11446 + }, + { + "epoch": 0.6024736842105263, + "grad_norm": 0.470703125, + "learning_rate": 8.83391812865497e-05, + "loss": 4.052, + "step": 11447 + }, + { + "epoch": 0.6025263157894737, + "grad_norm": 0.47265625, + "learning_rate": 8.832748538011695e-05, + "loss": 3.0716, + "step": 11448 + }, + { + "epoch": 0.6025789473684211, + "grad_norm": 0.490234375, + "learning_rate": 8.831578947368422e-05, + "loss": 3.3066, + "step": 11449 + }, + { + "epoch": 0.6026315789473684, + "grad_norm": 0.625, + "learning_rate": 8.830409356725147e-05, + "loss": 3.5095, + "step": 11450 + }, + { + "epoch": 0.6026842105263158, + "grad_norm": 0.5234375, + "learning_rate": 8.829239766081872e-05, + "loss": 3.4221, + "step": 11451 + }, + { + "epoch": 0.6027368421052631, + "grad_norm": 0.515625, + "learning_rate": 8.828070175438597e-05, + "loss": 3.4971, + "step": 11452 + }, + { + "epoch": 0.6027894736842105, + "grad_norm": 0.4921875, + "learning_rate": 8.826900584795323e-05, + "loss": 3.7466, + "step": 11453 + }, + { + "epoch": 0.6028421052631578, + "grad_norm": 0.62109375, + "learning_rate": 8.825730994152047e-05, + "loss": 3.5739, + "step": 11454 + }, + { + "epoch": 0.6028947368421053, + "grad_norm": 0.5390625, + "learning_rate": 8.824561403508772e-05, + "loss": 3.6295, + "step": 11455 + }, + { + "epoch": 0.6029473684210527, + "grad_norm": 0.490234375, + "learning_rate": 8.823391812865498e-05, + "loss": 3.5649, + "step": 11456 + }, + { + "epoch": 0.603, + "grad_norm": 0.486328125, + "learning_rate": 8.822222222222223e-05, + "loss": 3.4258, + "step": 11457 + }, + { + "epoch": 0.6030526315789474, + "grad_norm": 0.486328125, + "learning_rate": 8.821052631578948e-05, + "loss": 3.6175, + "step": 11458 + }, + { + "epoch": 0.6031052631578947, + "grad_norm": 0.49609375, + "learning_rate": 8.819883040935673e-05, + "loss": 3.4898, + "step": 11459 + }, + { + "epoch": 0.6031578947368421, + "grad_norm": 0.4921875, + "learning_rate": 8.818713450292399e-05, + "loss": 3.4537, + "step": 11460 + }, + { + "epoch": 0.6032105263157894, + "grad_norm": 0.490234375, + "learning_rate": 8.817543859649123e-05, + "loss": 3.298, + "step": 11461 + }, + { + "epoch": 0.6032631578947368, + "grad_norm": 0.54296875, + "learning_rate": 8.816374269005848e-05, + "loss": 3.6325, + "step": 11462 + }, + { + "epoch": 0.6033157894736842, + "grad_norm": 0.470703125, + "learning_rate": 8.815204678362573e-05, + "loss": 3.6367, + "step": 11463 + }, + { + "epoch": 0.6033684210526316, + "grad_norm": 0.498046875, + "learning_rate": 8.814035087719299e-05, + "loss": 3.6322, + "step": 11464 + }, + { + "epoch": 0.603421052631579, + "grad_norm": 0.51953125, + "learning_rate": 8.812865497076024e-05, + "loss": 3.4343, + "step": 11465 + }, + { + "epoch": 0.6034736842105263, + "grad_norm": 0.5390625, + "learning_rate": 8.811695906432749e-05, + "loss": 3.74, + "step": 11466 + }, + { + "epoch": 0.6035263157894737, + "grad_norm": 0.47265625, + "learning_rate": 8.810526315789475e-05, + "loss": 3.3612, + "step": 11467 + }, + { + "epoch": 0.6035789473684211, + "grad_norm": 0.515625, + "learning_rate": 8.8093567251462e-05, + "loss": 3.0685, + "step": 11468 + }, + { + "epoch": 0.6036315789473684, + "grad_norm": 0.5234375, + "learning_rate": 8.808187134502924e-05, + "loss": 3.3674, + "step": 11469 + }, + { + "epoch": 0.6036842105263158, + "grad_norm": 0.462890625, + "learning_rate": 8.807017543859649e-05, + "loss": 3.5322, + "step": 11470 + }, + { + "epoch": 0.6037368421052631, + "grad_norm": 0.46484375, + "learning_rate": 8.805847953216375e-05, + "loss": 3.5839, + "step": 11471 + }, + { + "epoch": 0.6037894736842105, + "grad_norm": 0.5, + "learning_rate": 8.8046783625731e-05, + "loss": 3.5935, + "step": 11472 + }, + { + "epoch": 0.6038421052631578, + "grad_norm": 0.52734375, + "learning_rate": 8.803508771929825e-05, + "loss": 3.2592, + "step": 11473 + }, + { + "epoch": 0.6038947368421053, + "grad_norm": 0.50390625, + "learning_rate": 8.80233918128655e-05, + "loss": 3.3288, + "step": 11474 + }, + { + "epoch": 0.6039473684210527, + "grad_norm": 0.515625, + "learning_rate": 8.801169590643276e-05, + "loss": 3.4237, + "step": 11475 + }, + { + "epoch": 0.604, + "grad_norm": 0.5078125, + "learning_rate": 8.800000000000001e-05, + "loss": 3.4669, + "step": 11476 + }, + { + "epoch": 0.6040526315789474, + "grad_norm": 0.50390625, + "learning_rate": 8.798830409356725e-05, + "loss": 2.8747, + "step": 11477 + }, + { + "epoch": 0.6041052631578947, + "grad_norm": 0.470703125, + "learning_rate": 8.79766081871345e-05, + "loss": 3.6076, + "step": 11478 + }, + { + "epoch": 0.6041578947368421, + "grad_norm": 0.5390625, + "learning_rate": 8.796491228070176e-05, + "loss": 3.1608, + "step": 11479 + }, + { + "epoch": 0.6042105263157894, + "grad_norm": 0.490234375, + "learning_rate": 8.795321637426901e-05, + "loss": 3.464, + "step": 11480 + }, + { + "epoch": 0.6042631578947368, + "grad_norm": 0.482421875, + "learning_rate": 8.794152046783626e-05, + "loss": 3.4795, + "step": 11481 + }, + { + "epoch": 0.6043157894736843, + "grad_norm": 0.484375, + "learning_rate": 8.792982456140352e-05, + "loss": 3.198, + "step": 11482 + }, + { + "epoch": 0.6043684210526316, + "grad_norm": 0.478515625, + "learning_rate": 8.791812865497077e-05, + "loss": 3.3601, + "step": 11483 + }, + { + "epoch": 0.604421052631579, + "grad_norm": 0.5, + "learning_rate": 8.790643274853801e-05, + "loss": 3.2691, + "step": 11484 + }, + { + "epoch": 0.6044736842105263, + "grad_norm": 0.478515625, + "learning_rate": 8.789473684210526e-05, + "loss": 3.4764, + "step": 11485 + }, + { + "epoch": 0.6045263157894737, + "grad_norm": 0.46484375, + "learning_rate": 8.788304093567252e-05, + "loss": 3.4721, + "step": 11486 + }, + { + "epoch": 0.6045789473684211, + "grad_norm": 0.478515625, + "learning_rate": 8.787134502923977e-05, + "loss": 3.2891, + "step": 11487 + }, + { + "epoch": 0.6046315789473684, + "grad_norm": 0.474609375, + "learning_rate": 8.785964912280702e-05, + "loss": 3.1192, + "step": 11488 + }, + { + "epoch": 0.6046842105263158, + "grad_norm": 0.49609375, + "learning_rate": 8.784795321637427e-05, + "loss": 3.4537, + "step": 11489 + }, + { + "epoch": 0.6047368421052631, + "grad_norm": 0.482421875, + "learning_rate": 8.783625730994153e-05, + "loss": 3.2957, + "step": 11490 + }, + { + "epoch": 0.6047894736842105, + "grad_norm": 0.4765625, + "learning_rate": 8.782456140350878e-05, + "loss": 3.373, + "step": 11491 + }, + { + "epoch": 0.6048421052631578, + "grad_norm": 0.490234375, + "learning_rate": 8.781286549707602e-05, + "loss": 3.2592, + "step": 11492 + }, + { + "epoch": 0.6048947368421053, + "grad_norm": 0.49609375, + "learning_rate": 8.780116959064328e-05, + "loss": 2.9474, + "step": 11493 + }, + { + "epoch": 0.6049473684210527, + "grad_norm": 0.462890625, + "learning_rate": 8.778947368421053e-05, + "loss": 3.4094, + "step": 11494 + }, + { + "epoch": 0.605, + "grad_norm": 0.482421875, + "learning_rate": 8.777777777777778e-05, + "loss": 3.5183, + "step": 11495 + }, + { + "epoch": 0.6050526315789474, + "grad_norm": 0.5546875, + "learning_rate": 8.776608187134503e-05, + "loss": 3.0557, + "step": 11496 + }, + { + "epoch": 0.6051052631578947, + "grad_norm": 0.4609375, + "learning_rate": 8.77543859649123e-05, + "loss": 3.4765, + "step": 11497 + }, + { + "epoch": 0.6051578947368421, + "grad_norm": 0.5078125, + "learning_rate": 8.774269005847954e-05, + "loss": 3.8889, + "step": 11498 + }, + { + "epoch": 0.6052105263157894, + "grad_norm": 0.4609375, + "learning_rate": 8.773099415204679e-05, + "loss": 3.1958, + "step": 11499 + }, + { + "epoch": 0.6052631578947368, + "grad_norm": 0.515625, + "learning_rate": 8.771929824561403e-05, + "loss": 3.2932, + "step": 11500 + }, + { + "epoch": 0.6053157894736843, + "grad_norm": 0.51953125, + "learning_rate": 8.770760233918129e-05, + "loss": 3.2992, + "step": 11501 + }, + { + "epoch": 0.6053684210526316, + "grad_norm": 0.48828125, + "learning_rate": 8.769590643274854e-05, + "loss": 3.7795, + "step": 11502 + }, + { + "epoch": 0.605421052631579, + "grad_norm": 0.482421875, + "learning_rate": 8.768421052631579e-05, + "loss": 2.9925, + "step": 11503 + }, + { + "epoch": 0.6054736842105263, + "grad_norm": 0.49609375, + "learning_rate": 8.767251461988305e-05, + "loss": 3.428, + "step": 11504 + }, + { + "epoch": 0.6055263157894737, + "grad_norm": 0.515625, + "learning_rate": 8.76608187134503e-05, + "loss": 3.2833, + "step": 11505 + }, + { + "epoch": 0.6055789473684211, + "grad_norm": 0.4921875, + "learning_rate": 8.764912280701755e-05, + "loss": 3.339, + "step": 11506 + }, + { + "epoch": 0.6056315789473684, + "grad_norm": 0.498046875, + "learning_rate": 8.763742690058479e-05, + "loss": 3.6097, + "step": 11507 + }, + { + "epoch": 0.6056842105263158, + "grad_norm": 0.5, + "learning_rate": 8.762573099415205e-05, + "loss": 3.3626, + "step": 11508 + }, + { + "epoch": 0.6057368421052631, + "grad_norm": 0.53515625, + "learning_rate": 8.76140350877193e-05, + "loss": 3.1157, + "step": 11509 + }, + { + "epoch": 0.6057894736842105, + "grad_norm": 0.482421875, + "learning_rate": 8.760233918128655e-05, + "loss": 3.4123, + "step": 11510 + }, + { + "epoch": 0.6058421052631578, + "grad_norm": 0.498046875, + "learning_rate": 8.75906432748538e-05, + "loss": 3.0503, + "step": 11511 + }, + { + "epoch": 0.6058947368421053, + "grad_norm": 0.51171875, + "learning_rate": 8.757894736842106e-05, + "loss": 3.3841, + "step": 11512 + }, + { + "epoch": 0.6059473684210527, + "grad_norm": 0.49609375, + "learning_rate": 8.756725146198831e-05, + "loss": 3.4276, + "step": 11513 + }, + { + "epoch": 0.606, + "grad_norm": 0.53125, + "learning_rate": 8.755555555555556e-05, + "loss": 3.1818, + "step": 11514 + }, + { + "epoch": 0.6060526315789474, + "grad_norm": 0.4609375, + "learning_rate": 8.75438596491228e-05, + "loss": 3.0587, + "step": 11515 + }, + { + "epoch": 0.6061052631578947, + "grad_norm": 0.478515625, + "learning_rate": 8.753216374269006e-05, + "loss": 3.146, + "step": 11516 + }, + { + "epoch": 0.6061578947368421, + "grad_norm": 0.47265625, + "learning_rate": 8.752046783625731e-05, + "loss": 3.477, + "step": 11517 + }, + { + "epoch": 0.6062105263157894, + "grad_norm": 0.50390625, + "learning_rate": 8.750877192982456e-05, + "loss": 3.5044, + "step": 11518 + }, + { + "epoch": 0.6062631578947368, + "grad_norm": 0.482421875, + "learning_rate": 8.749707602339183e-05, + "loss": 3.5732, + "step": 11519 + }, + { + "epoch": 0.6063157894736843, + "grad_norm": 0.48046875, + "learning_rate": 8.748538011695907e-05, + "loss": 3.3977, + "step": 11520 + }, + { + "epoch": 0.6063684210526316, + "grad_norm": 0.5, + "learning_rate": 8.747368421052632e-05, + "loss": 3.5536, + "step": 11521 + }, + { + "epoch": 0.606421052631579, + "grad_norm": 0.466796875, + "learning_rate": 8.746198830409357e-05, + "loss": 3.8382, + "step": 11522 + }, + { + "epoch": 0.6064736842105263, + "grad_norm": 0.498046875, + "learning_rate": 8.745029239766082e-05, + "loss": 3.6518, + "step": 11523 + }, + { + "epoch": 0.6065263157894737, + "grad_norm": 0.466796875, + "learning_rate": 8.743859649122807e-05, + "loss": 3.1443, + "step": 11524 + }, + { + "epoch": 0.6065789473684211, + "grad_norm": 0.462890625, + "learning_rate": 8.742690058479532e-05, + "loss": 3.5058, + "step": 11525 + }, + { + "epoch": 0.6066315789473684, + "grad_norm": 0.5, + "learning_rate": 8.741520467836257e-05, + "loss": 2.9442, + "step": 11526 + }, + { + "epoch": 0.6066842105263158, + "grad_norm": 0.5, + "learning_rate": 8.740350877192984e-05, + "loss": 3.528, + "step": 11527 + }, + { + "epoch": 0.6067368421052631, + "grad_norm": 0.4765625, + "learning_rate": 8.739181286549708e-05, + "loss": 3.0959, + "step": 11528 + }, + { + "epoch": 0.6067894736842105, + "grad_norm": 0.5078125, + "learning_rate": 8.738011695906433e-05, + "loss": 3.4219, + "step": 11529 + }, + { + "epoch": 0.6068421052631578, + "grad_norm": 0.50390625, + "learning_rate": 8.736842105263158e-05, + "loss": 3.335, + "step": 11530 + }, + { + "epoch": 0.6068947368421053, + "grad_norm": 0.5078125, + "learning_rate": 8.735672514619883e-05, + "loss": 3.1428, + "step": 11531 + }, + { + "epoch": 0.6069473684210527, + "grad_norm": 0.50390625, + "learning_rate": 8.734502923976608e-05, + "loss": 3.3514, + "step": 11532 + }, + { + "epoch": 0.607, + "grad_norm": 0.486328125, + "learning_rate": 8.733333333333333e-05, + "loss": 3.1442, + "step": 11533 + }, + { + "epoch": 0.6070526315789474, + "grad_norm": 0.455078125, + "learning_rate": 8.73216374269006e-05, + "loss": 3.25, + "step": 11534 + }, + { + "epoch": 0.6071052631578947, + "grad_norm": 0.478515625, + "learning_rate": 8.730994152046785e-05, + "loss": 3.8065, + "step": 11535 + }, + { + "epoch": 0.6071578947368421, + "grad_norm": 0.515625, + "learning_rate": 8.72982456140351e-05, + "loss": 3.5613, + "step": 11536 + }, + { + "epoch": 0.6072105263157894, + "grad_norm": 0.515625, + "learning_rate": 8.728654970760234e-05, + "loss": 2.6705, + "step": 11537 + }, + { + "epoch": 0.6072631578947368, + "grad_norm": 0.474609375, + "learning_rate": 8.72748538011696e-05, + "loss": 3.135, + "step": 11538 + }, + { + "epoch": 0.6073157894736843, + "grad_norm": 0.46875, + "learning_rate": 8.726315789473684e-05, + "loss": 3.3537, + "step": 11539 + }, + { + "epoch": 0.6073684210526316, + "grad_norm": 0.478515625, + "learning_rate": 8.72514619883041e-05, + "loss": 3.1792, + "step": 11540 + }, + { + "epoch": 0.607421052631579, + "grad_norm": 0.46484375, + "learning_rate": 8.723976608187136e-05, + "loss": 3.0873, + "step": 11541 + }, + { + "epoch": 0.6074736842105263, + "grad_norm": 0.466796875, + "learning_rate": 8.72280701754386e-05, + "loss": 3.3484, + "step": 11542 + }, + { + "epoch": 0.6075263157894737, + "grad_norm": 0.515625, + "learning_rate": 8.721637426900586e-05, + "loss": 3.4185, + "step": 11543 + }, + { + "epoch": 0.6075789473684211, + "grad_norm": 0.515625, + "learning_rate": 8.72046783625731e-05, + "loss": 3.7122, + "step": 11544 + }, + { + "epoch": 0.6076315789473684, + "grad_norm": 0.55078125, + "learning_rate": 8.719298245614036e-05, + "loss": 3.466, + "step": 11545 + }, + { + "epoch": 0.6076842105263158, + "grad_norm": 0.47265625, + "learning_rate": 8.71812865497076e-05, + "loss": 3.0847, + "step": 11546 + }, + { + "epoch": 0.6077368421052631, + "grad_norm": 0.625, + "learning_rate": 8.716959064327485e-05, + "loss": 3.5225, + "step": 11547 + }, + { + "epoch": 0.6077894736842105, + "grad_norm": 0.5390625, + "learning_rate": 8.71578947368421e-05, + "loss": 3.4108, + "step": 11548 + }, + { + "epoch": 0.6078421052631579, + "grad_norm": 0.52734375, + "learning_rate": 8.714619883040937e-05, + "loss": 3.4546, + "step": 11549 + }, + { + "epoch": 0.6078947368421053, + "grad_norm": 0.5, + "learning_rate": 8.713450292397662e-05, + "loss": 3.5952, + "step": 11550 + }, + { + "epoch": 0.6079473684210527, + "grad_norm": 0.50390625, + "learning_rate": 8.712280701754387e-05, + "loss": 3.014, + "step": 11551 + }, + { + "epoch": 0.608, + "grad_norm": 0.474609375, + "learning_rate": 8.711111111111112e-05, + "loss": 3.3277, + "step": 11552 + }, + { + "epoch": 0.6080526315789474, + "grad_norm": 0.47265625, + "learning_rate": 8.709941520467837e-05, + "loss": 3.2783, + "step": 11553 + }, + { + "epoch": 0.6081052631578947, + "grad_norm": 0.48828125, + "learning_rate": 8.708771929824562e-05, + "loss": 3.3032, + "step": 11554 + }, + { + "epoch": 0.6081578947368421, + "grad_norm": 0.474609375, + "learning_rate": 8.707602339181286e-05, + "loss": 3.414, + "step": 11555 + }, + { + "epoch": 0.6082105263157894, + "grad_norm": 0.4921875, + "learning_rate": 8.706432748538013e-05, + "loss": 3.822, + "step": 11556 + }, + { + "epoch": 0.6082631578947368, + "grad_norm": 0.482421875, + "learning_rate": 8.705263157894738e-05, + "loss": 2.841, + "step": 11557 + }, + { + "epoch": 0.6083157894736843, + "grad_norm": 0.455078125, + "learning_rate": 8.704093567251463e-05, + "loss": 3.1851, + "step": 11558 + }, + { + "epoch": 0.6083684210526316, + "grad_norm": 0.515625, + "learning_rate": 8.702923976608188e-05, + "loss": 3.6389, + "step": 11559 + }, + { + "epoch": 0.608421052631579, + "grad_norm": 0.55078125, + "learning_rate": 8.701754385964913e-05, + "loss": 3.4218, + "step": 11560 + }, + { + "epoch": 0.6084736842105263, + "grad_norm": 0.4921875, + "learning_rate": 8.700584795321638e-05, + "loss": 3.328, + "step": 11561 + }, + { + "epoch": 0.6085263157894737, + "grad_norm": 0.494140625, + "learning_rate": 8.699415204678363e-05, + "loss": 3.7641, + "step": 11562 + }, + { + "epoch": 0.608578947368421, + "grad_norm": 0.494140625, + "learning_rate": 8.698245614035087e-05, + "loss": 3.7211, + "step": 11563 + }, + { + "epoch": 0.6086315789473684, + "grad_norm": 0.515625, + "learning_rate": 8.697076023391814e-05, + "loss": 3.6284, + "step": 11564 + }, + { + "epoch": 0.6086842105263158, + "grad_norm": 0.5859375, + "learning_rate": 8.695906432748539e-05, + "loss": 3.6083, + "step": 11565 + }, + { + "epoch": 0.6087368421052631, + "grad_norm": 0.486328125, + "learning_rate": 8.694736842105264e-05, + "loss": 2.9685, + "step": 11566 + }, + { + "epoch": 0.6087894736842105, + "grad_norm": 0.46484375, + "learning_rate": 8.693567251461989e-05, + "loss": 3.2415, + "step": 11567 + }, + { + "epoch": 0.6088421052631579, + "grad_norm": 0.4921875, + "learning_rate": 8.692397660818714e-05, + "loss": 3.6462, + "step": 11568 + }, + { + "epoch": 0.6088947368421053, + "grad_norm": 0.4765625, + "learning_rate": 8.691228070175439e-05, + "loss": 3.0788, + "step": 11569 + }, + { + "epoch": 0.6089473684210527, + "grad_norm": 0.4921875, + "learning_rate": 8.690058479532164e-05, + "loss": 3.293, + "step": 11570 + }, + { + "epoch": 0.609, + "grad_norm": 0.4609375, + "learning_rate": 8.68888888888889e-05, + "loss": 3.4204, + "step": 11571 + }, + { + "epoch": 0.6090526315789474, + "grad_norm": 0.484375, + "learning_rate": 8.687719298245615e-05, + "loss": 3.2613, + "step": 11572 + }, + { + "epoch": 0.6091052631578947, + "grad_norm": 0.52734375, + "learning_rate": 8.68654970760234e-05, + "loss": 3.6006, + "step": 11573 + }, + { + "epoch": 0.6091578947368421, + "grad_norm": 0.5078125, + "learning_rate": 8.685380116959065e-05, + "loss": 2.9962, + "step": 11574 + }, + { + "epoch": 0.6092105263157894, + "grad_norm": 0.47265625, + "learning_rate": 8.68421052631579e-05, + "loss": 3.0588, + "step": 11575 + }, + { + "epoch": 0.6092631578947368, + "grad_norm": 0.48046875, + "learning_rate": 8.683040935672515e-05, + "loss": 3.26, + "step": 11576 + }, + { + "epoch": 0.6093157894736843, + "grad_norm": 0.515625, + "learning_rate": 8.68187134502924e-05, + "loss": 3.5457, + "step": 11577 + }, + { + "epoch": 0.6093684210526316, + "grad_norm": 0.515625, + "learning_rate": 8.680701754385965e-05, + "loss": 3.4533, + "step": 11578 + }, + { + "epoch": 0.609421052631579, + "grad_norm": 0.46484375, + "learning_rate": 8.679532163742691e-05, + "loss": 2.8795, + "step": 11579 + }, + { + "epoch": 0.6094736842105263, + "grad_norm": 0.46875, + "learning_rate": 8.678362573099416e-05, + "loss": 3.6396, + "step": 11580 + }, + { + "epoch": 0.6095263157894737, + "grad_norm": 0.50390625, + "learning_rate": 8.677192982456141e-05, + "loss": 3.6238, + "step": 11581 + }, + { + "epoch": 0.609578947368421, + "grad_norm": 0.50390625, + "learning_rate": 8.676023391812866e-05, + "loss": 3.244, + "step": 11582 + }, + { + "epoch": 0.6096315789473684, + "grad_norm": 0.515625, + "learning_rate": 8.674853801169591e-05, + "loss": 3.5437, + "step": 11583 + }, + { + "epoch": 0.6096842105263158, + "grad_norm": 0.47265625, + "learning_rate": 8.673684210526316e-05, + "loss": 3.3824, + "step": 11584 + }, + { + "epoch": 0.6097368421052631, + "grad_norm": 0.484375, + "learning_rate": 8.67251461988304e-05, + "loss": 3.1078, + "step": 11585 + }, + { + "epoch": 0.6097894736842105, + "grad_norm": 0.47265625, + "learning_rate": 8.671345029239767e-05, + "loss": 3.0728, + "step": 11586 + }, + { + "epoch": 0.6098421052631579, + "grad_norm": 0.47265625, + "learning_rate": 8.670175438596492e-05, + "loss": 3.2603, + "step": 11587 + }, + { + "epoch": 0.6098947368421053, + "grad_norm": 0.466796875, + "learning_rate": 8.669005847953217e-05, + "loss": 3.2183, + "step": 11588 + }, + { + "epoch": 0.6099473684210527, + "grad_norm": 0.4765625, + "learning_rate": 8.667836257309942e-05, + "loss": 3.1874, + "step": 11589 + }, + { + "epoch": 0.61, + "grad_norm": 0.462890625, + "learning_rate": 8.666666666666667e-05, + "loss": 2.9472, + "step": 11590 + }, + { + "epoch": 0.6100526315789474, + "grad_norm": 0.50390625, + "learning_rate": 8.665497076023392e-05, + "loss": 2.8852, + "step": 11591 + }, + { + "epoch": 0.6101052631578947, + "grad_norm": 0.470703125, + "learning_rate": 8.664327485380117e-05, + "loss": 3.5086, + "step": 11592 + }, + { + "epoch": 0.6101578947368421, + "grad_norm": 1.2421875, + "learning_rate": 8.663157894736843e-05, + "loss": 3.0827, + "step": 11593 + }, + { + "epoch": 0.6102105263157894, + "grad_norm": 0.47265625, + "learning_rate": 8.661988304093568e-05, + "loss": 3.3762, + "step": 11594 + }, + { + "epoch": 0.6102631578947368, + "grad_norm": 0.474609375, + "learning_rate": 8.660818713450293e-05, + "loss": 3.5644, + "step": 11595 + }, + { + "epoch": 0.6103157894736843, + "grad_norm": 0.498046875, + "learning_rate": 8.659649122807018e-05, + "loss": 3.3289, + "step": 11596 + }, + { + "epoch": 0.6103684210526316, + "grad_norm": 0.494140625, + "learning_rate": 8.658479532163743e-05, + "loss": 3.0267, + "step": 11597 + }, + { + "epoch": 0.610421052631579, + "grad_norm": 0.6015625, + "learning_rate": 8.657309941520468e-05, + "loss": 3.463, + "step": 11598 + }, + { + "epoch": 0.6104736842105263, + "grad_norm": 0.4921875, + "learning_rate": 8.656140350877193e-05, + "loss": 3.4617, + "step": 11599 + }, + { + "epoch": 0.6105263157894737, + "grad_norm": 0.486328125, + "learning_rate": 8.654970760233918e-05, + "loss": 3.2139, + "step": 11600 + }, + { + "epoch": 0.610578947368421, + "grad_norm": 0.494140625, + "learning_rate": 8.653801169590644e-05, + "loss": 3.15, + "step": 11601 + }, + { + "epoch": 0.6106315789473684, + "grad_norm": 0.52734375, + "learning_rate": 8.652631578947369e-05, + "loss": 3.4067, + "step": 11602 + }, + { + "epoch": 0.6106842105263158, + "grad_norm": 0.48828125, + "learning_rate": 8.651461988304094e-05, + "loss": 3.129, + "step": 11603 + }, + { + "epoch": 0.6107368421052631, + "grad_norm": 0.490234375, + "learning_rate": 8.65029239766082e-05, + "loss": 3.5168, + "step": 11604 + }, + { + "epoch": 0.6107894736842105, + "grad_norm": 0.494140625, + "learning_rate": 8.649122807017544e-05, + "loss": 3.3593, + "step": 11605 + }, + { + "epoch": 0.6108421052631579, + "grad_norm": 0.53515625, + "learning_rate": 8.647953216374269e-05, + "loss": 3.402, + "step": 11606 + }, + { + "epoch": 0.6108947368421053, + "grad_norm": 0.53515625, + "learning_rate": 8.646783625730994e-05, + "loss": 3.5082, + "step": 11607 + }, + { + "epoch": 0.6109473684210527, + "grad_norm": 0.466796875, + "learning_rate": 8.64561403508772e-05, + "loss": 3.6092, + "step": 11608 + }, + { + "epoch": 0.611, + "grad_norm": 0.490234375, + "learning_rate": 8.644444444444445e-05, + "loss": 3.4683, + "step": 11609 + }, + { + "epoch": 0.6110526315789474, + "grad_norm": 0.484375, + "learning_rate": 8.64327485380117e-05, + "loss": 3.2547, + "step": 11610 + }, + { + "epoch": 0.6111052631578947, + "grad_norm": 0.5625, + "learning_rate": 8.642105263157895e-05, + "loss": 3.3491, + "step": 11611 + }, + { + "epoch": 0.6111578947368421, + "grad_norm": 0.474609375, + "learning_rate": 8.640935672514621e-05, + "loss": 3.7966, + "step": 11612 + }, + { + "epoch": 0.6112105263157894, + "grad_norm": 0.47265625, + "learning_rate": 8.639766081871345e-05, + "loss": 2.8911, + "step": 11613 + }, + { + "epoch": 0.6112631578947368, + "grad_norm": 0.50390625, + "learning_rate": 8.63859649122807e-05, + "loss": 3.1977, + "step": 11614 + }, + { + "epoch": 0.6113157894736843, + "grad_norm": 0.482421875, + "learning_rate": 8.637426900584795e-05, + "loss": 3.4767, + "step": 11615 + }, + { + "epoch": 0.6113684210526316, + "grad_norm": 0.478515625, + "learning_rate": 8.636257309941521e-05, + "loss": 3.237, + "step": 11616 + }, + { + "epoch": 0.611421052631579, + "grad_norm": 0.462890625, + "learning_rate": 8.635087719298246e-05, + "loss": 3.1211, + "step": 11617 + }, + { + "epoch": 0.6114736842105263, + "grad_norm": 0.46484375, + "learning_rate": 8.633918128654971e-05, + "loss": 3.1581, + "step": 11618 + }, + { + "epoch": 0.6115263157894737, + "grad_norm": 0.5625, + "learning_rate": 8.632748538011697e-05, + "loss": 3.054, + "step": 11619 + }, + { + "epoch": 0.611578947368421, + "grad_norm": 0.453125, + "learning_rate": 8.631578947368421e-05, + "loss": 3.2103, + "step": 11620 + }, + { + "epoch": 0.6116315789473684, + "grad_norm": 0.46484375, + "learning_rate": 8.630409356725146e-05, + "loss": 3.4276, + "step": 11621 + }, + { + "epoch": 0.6116842105263158, + "grad_norm": 0.5078125, + "learning_rate": 8.629239766081871e-05, + "loss": 3.2363, + "step": 11622 + }, + { + "epoch": 0.6117368421052631, + "grad_norm": 0.474609375, + "learning_rate": 8.628070175438597e-05, + "loss": 3.4415, + "step": 11623 + }, + { + "epoch": 0.6117894736842105, + "grad_norm": 0.50390625, + "learning_rate": 8.626900584795322e-05, + "loss": 3.1938, + "step": 11624 + }, + { + "epoch": 0.6118421052631579, + "grad_norm": 0.55859375, + "learning_rate": 8.625730994152047e-05, + "loss": 3.1619, + "step": 11625 + }, + { + "epoch": 0.6118947368421053, + "grad_norm": 0.466796875, + "learning_rate": 8.624561403508772e-05, + "loss": 3.3683, + "step": 11626 + }, + { + "epoch": 0.6119473684210527, + "grad_norm": 0.4921875, + "learning_rate": 8.623391812865498e-05, + "loss": 3.5567, + "step": 11627 + }, + { + "epoch": 0.612, + "grad_norm": 0.515625, + "learning_rate": 8.622222222222222e-05, + "loss": 3.6218, + "step": 11628 + }, + { + "epoch": 0.6120526315789474, + "grad_norm": 0.494140625, + "learning_rate": 8.621052631578947e-05, + "loss": 3.2696, + "step": 11629 + }, + { + "epoch": 0.6121052631578947, + "grad_norm": 0.486328125, + "learning_rate": 8.619883040935673e-05, + "loss": 3.1177, + "step": 11630 + }, + { + "epoch": 0.6121578947368421, + "grad_norm": 0.498046875, + "learning_rate": 8.618713450292398e-05, + "loss": 3.2385, + "step": 11631 + }, + { + "epoch": 0.6122105263157894, + "grad_norm": 0.5, + "learning_rate": 8.617543859649123e-05, + "loss": 3.0798, + "step": 11632 + }, + { + "epoch": 0.6122631578947368, + "grad_norm": 0.478515625, + "learning_rate": 8.616374269005848e-05, + "loss": 3.2477, + "step": 11633 + }, + { + "epoch": 0.6123157894736843, + "grad_norm": 0.46484375, + "learning_rate": 8.615204678362575e-05, + "loss": 3.2714, + "step": 11634 + }, + { + "epoch": 0.6123684210526316, + "grad_norm": 0.4765625, + "learning_rate": 8.6140350877193e-05, + "loss": 3.3646, + "step": 11635 + }, + { + "epoch": 0.612421052631579, + "grad_norm": 0.494140625, + "learning_rate": 8.612865497076023e-05, + "loss": 3.228, + "step": 11636 + }, + { + "epoch": 0.6124736842105263, + "grad_norm": 0.51171875, + "learning_rate": 8.611695906432748e-05, + "loss": 3.6157, + "step": 11637 + }, + { + "epoch": 0.6125263157894737, + "grad_norm": 0.478515625, + "learning_rate": 8.610526315789474e-05, + "loss": 3.4524, + "step": 11638 + }, + { + "epoch": 0.612578947368421, + "grad_norm": 0.4609375, + "learning_rate": 8.609356725146199e-05, + "loss": 3.3043, + "step": 11639 + }, + { + "epoch": 0.6126315789473684, + "grad_norm": 0.447265625, + "learning_rate": 8.608187134502924e-05, + "loss": 3.6275, + "step": 11640 + }, + { + "epoch": 0.6126842105263158, + "grad_norm": 0.484375, + "learning_rate": 8.60701754385965e-05, + "loss": 3.3612, + "step": 11641 + }, + { + "epoch": 0.6127368421052631, + "grad_norm": 0.443359375, + "learning_rate": 8.605847953216376e-05, + "loss": 3.1256, + "step": 11642 + }, + { + "epoch": 0.6127894736842106, + "grad_norm": 0.462890625, + "learning_rate": 8.604678362573099e-05, + "loss": 3.3126, + "step": 11643 + }, + { + "epoch": 0.6128421052631579, + "grad_norm": 0.494140625, + "learning_rate": 8.603508771929824e-05, + "loss": 3.1826, + "step": 11644 + }, + { + "epoch": 0.6128947368421053, + "grad_norm": 0.482421875, + "learning_rate": 8.60233918128655e-05, + "loss": 3.2381, + "step": 11645 + }, + { + "epoch": 0.6129473684210527, + "grad_norm": 0.4765625, + "learning_rate": 8.601169590643275e-05, + "loss": 3.2737, + "step": 11646 + }, + { + "epoch": 0.613, + "grad_norm": 0.51171875, + "learning_rate": 8.6e-05, + "loss": 3.636, + "step": 11647 + }, + { + "epoch": 0.6130526315789474, + "grad_norm": 0.5078125, + "learning_rate": 8.598830409356725e-05, + "loss": 3.2426, + "step": 11648 + }, + { + "epoch": 0.6131052631578947, + "grad_norm": 0.474609375, + "learning_rate": 8.597660818713452e-05, + "loss": 3.4927, + "step": 11649 + }, + { + "epoch": 0.6131578947368421, + "grad_norm": 0.6015625, + "learning_rate": 8.596491228070177e-05, + "loss": 2.9806, + "step": 11650 + }, + { + "epoch": 0.6132105263157894, + "grad_norm": 0.50390625, + "learning_rate": 8.5953216374269e-05, + "loss": 3.2581, + "step": 11651 + }, + { + "epoch": 0.6132631578947368, + "grad_norm": 0.490234375, + "learning_rate": 8.594152046783625e-05, + "loss": 3.4125, + "step": 11652 + }, + { + "epoch": 0.6133157894736843, + "grad_norm": 0.5, + "learning_rate": 8.592982456140351e-05, + "loss": 3.7616, + "step": 11653 + }, + { + "epoch": 0.6133684210526316, + "grad_norm": 0.5390625, + "learning_rate": 8.591812865497076e-05, + "loss": 3.7012, + "step": 11654 + }, + { + "epoch": 0.613421052631579, + "grad_norm": 0.490234375, + "learning_rate": 8.590643274853801e-05, + "loss": 3.3594, + "step": 11655 + }, + { + "epoch": 0.6134736842105263, + "grad_norm": 0.5234375, + "learning_rate": 8.589473684210528e-05, + "loss": 3.3942, + "step": 11656 + }, + { + "epoch": 0.6135263157894737, + "grad_norm": 0.47265625, + "learning_rate": 8.588304093567253e-05, + "loss": 3.145, + "step": 11657 + }, + { + "epoch": 0.613578947368421, + "grad_norm": 0.625, + "learning_rate": 8.587134502923978e-05, + "loss": 3.6896, + "step": 11658 + }, + { + "epoch": 0.6136315789473684, + "grad_norm": 0.482421875, + "learning_rate": 8.585964912280701e-05, + "loss": 3.724, + "step": 11659 + }, + { + "epoch": 0.6136842105263158, + "grad_norm": 0.4921875, + "learning_rate": 8.584795321637428e-05, + "loss": 3.2556, + "step": 11660 + }, + { + "epoch": 0.6137368421052631, + "grad_norm": 0.53515625, + "learning_rate": 8.583625730994152e-05, + "loss": 3.2803, + "step": 11661 + }, + { + "epoch": 0.6137894736842106, + "grad_norm": 0.4921875, + "learning_rate": 8.582456140350877e-05, + "loss": 3.2296, + "step": 11662 + }, + { + "epoch": 0.6138421052631579, + "grad_norm": 0.47265625, + "learning_rate": 8.581286549707602e-05, + "loss": 3.7297, + "step": 11663 + }, + { + "epoch": 0.6138947368421053, + "grad_norm": 0.5078125, + "learning_rate": 8.580116959064329e-05, + "loss": 2.7421, + "step": 11664 + }, + { + "epoch": 0.6139473684210527, + "grad_norm": 0.48828125, + "learning_rate": 8.578947368421054e-05, + "loss": 2.9836, + "step": 11665 + }, + { + "epoch": 0.614, + "grad_norm": 0.70703125, + "learning_rate": 8.577777777777777e-05, + "loss": 3.1752, + "step": 11666 + }, + { + "epoch": 0.6140526315789474, + "grad_norm": 0.5546875, + "learning_rate": 8.576608187134504e-05, + "loss": 3.4407, + "step": 11667 + }, + { + "epoch": 0.6141052631578947, + "grad_norm": 0.5390625, + "learning_rate": 8.575438596491229e-05, + "loss": 3.4105, + "step": 11668 + }, + { + "epoch": 0.6141578947368421, + "grad_norm": 0.48828125, + "learning_rate": 8.574269005847954e-05, + "loss": 3.1392, + "step": 11669 + }, + { + "epoch": 0.6142105263157894, + "grad_norm": 0.4921875, + "learning_rate": 8.573099415204678e-05, + "loss": 3.3498, + "step": 11670 + }, + { + "epoch": 0.6142631578947368, + "grad_norm": 0.498046875, + "learning_rate": 8.571929824561405e-05, + "loss": 3.3376, + "step": 11671 + }, + { + "epoch": 0.6143157894736843, + "grad_norm": 0.484375, + "learning_rate": 8.57076023391813e-05, + "loss": 3.4143, + "step": 11672 + }, + { + "epoch": 0.6143684210526316, + "grad_norm": 0.5078125, + "learning_rate": 8.569590643274855e-05, + "loss": 3.3017, + "step": 11673 + }, + { + "epoch": 0.614421052631579, + "grad_norm": 0.46484375, + "learning_rate": 8.568421052631578e-05, + "loss": 3.5448, + "step": 11674 + }, + { + "epoch": 0.6144736842105263, + "grad_norm": 0.482421875, + "learning_rate": 8.567251461988305e-05, + "loss": 2.9429, + "step": 11675 + }, + { + "epoch": 0.6145263157894737, + "grad_norm": 0.4765625, + "learning_rate": 8.56608187134503e-05, + "loss": 3.5305, + "step": 11676 + }, + { + "epoch": 0.614578947368421, + "grad_norm": 0.466796875, + "learning_rate": 8.564912280701755e-05, + "loss": 3.2693, + "step": 11677 + }, + { + "epoch": 0.6146315789473684, + "grad_norm": 0.478515625, + "learning_rate": 8.56374269005848e-05, + "loss": 3.5118, + "step": 11678 + }, + { + "epoch": 0.6146842105263158, + "grad_norm": 0.5234375, + "learning_rate": 8.562573099415206e-05, + "loss": 3.2339, + "step": 11679 + }, + { + "epoch": 0.6147368421052631, + "grad_norm": 0.58203125, + "learning_rate": 8.561403508771931e-05, + "loss": 3.4434, + "step": 11680 + }, + { + "epoch": 0.6147894736842106, + "grad_norm": 0.48828125, + "learning_rate": 8.560233918128656e-05, + "loss": 3.6145, + "step": 11681 + }, + { + "epoch": 0.6148421052631579, + "grad_norm": 0.48046875, + "learning_rate": 8.559064327485381e-05, + "loss": 3.3872, + "step": 11682 + }, + { + "epoch": 0.6148947368421053, + "grad_norm": 0.51171875, + "learning_rate": 8.557894736842106e-05, + "loss": 3.8301, + "step": 11683 + }, + { + "epoch": 0.6149473684210527, + "grad_norm": 0.5390625, + "learning_rate": 8.55672514619883e-05, + "loss": 3.4851, + "step": 11684 + }, + { + "epoch": 0.615, + "grad_norm": 0.51171875, + "learning_rate": 8.555555555555556e-05, + "loss": 3.61, + "step": 11685 + }, + { + "epoch": 0.6150526315789474, + "grad_norm": 0.48828125, + "learning_rate": 8.554385964912282e-05, + "loss": 3.2882, + "step": 11686 + }, + { + "epoch": 0.6151052631578947, + "grad_norm": 0.490234375, + "learning_rate": 8.553216374269007e-05, + "loss": 3.235, + "step": 11687 + }, + { + "epoch": 0.6151578947368421, + "grad_norm": 0.46484375, + "learning_rate": 8.552046783625732e-05, + "loss": 2.9447, + "step": 11688 + }, + { + "epoch": 0.6152105263157894, + "grad_norm": 0.486328125, + "learning_rate": 8.550877192982455e-05, + "loss": 3.8607, + "step": 11689 + }, + { + "epoch": 0.6152631578947368, + "grad_norm": 0.48828125, + "learning_rate": 8.549707602339182e-05, + "loss": 3.3335, + "step": 11690 + }, + { + "epoch": 0.6153157894736843, + "grad_norm": 0.478515625, + "learning_rate": 8.548538011695907e-05, + "loss": 3.3806, + "step": 11691 + }, + { + "epoch": 0.6153684210526316, + "grad_norm": 0.4921875, + "learning_rate": 8.547368421052632e-05, + "loss": 3.0245, + "step": 11692 + }, + { + "epoch": 0.615421052631579, + "grad_norm": 0.5, + "learning_rate": 8.546198830409358e-05, + "loss": 3.2072, + "step": 11693 + }, + { + "epoch": 0.6154736842105263, + "grad_norm": 0.73828125, + "learning_rate": 8.545029239766083e-05, + "loss": 3.1148, + "step": 11694 + }, + { + "epoch": 0.6155263157894737, + "grad_norm": 0.5078125, + "learning_rate": 8.543859649122808e-05, + "loss": 2.9436, + "step": 11695 + }, + { + "epoch": 0.615578947368421, + "grad_norm": 0.46875, + "learning_rate": 8.542690058479533e-05, + "loss": 3.022, + "step": 11696 + }, + { + "epoch": 0.6156315789473684, + "grad_norm": 0.5078125, + "learning_rate": 8.541520467836258e-05, + "loss": 3.1369, + "step": 11697 + }, + { + "epoch": 0.6156842105263158, + "grad_norm": 0.494140625, + "learning_rate": 8.540350877192983e-05, + "loss": 3.6179, + "step": 11698 + }, + { + "epoch": 0.6157368421052631, + "grad_norm": 0.50390625, + "learning_rate": 8.539181286549708e-05, + "loss": 3.4594, + "step": 11699 + }, + { + "epoch": 0.6157894736842106, + "grad_norm": 0.515625, + "learning_rate": 8.538011695906433e-05, + "loss": 3.4355, + "step": 11700 + }, + { + "epoch": 0.6158421052631579, + "grad_norm": 0.5078125, + "learning_rate": 8.536842105263159e-05, + "loss": 3.2053, + "step": 11701 + }, + { + "epoch": 0.6158947368421053, + "grad_norm": 0.490234375, + "learning_rate": 8.535672514619884e-05, + "loss": 3.3506, + "step": 11702 + }, + { + "epoch": 0.6159473684210526, + "grad_norm": 0.50390625, + "learning_rate": 8.534502923976609e-05, + "loss": 3.3825, + "step": 11703 + }, + { + "epoch": 0.616, + "grad_norm": 0.5078125, + "learning_rate": 8.533333333333334e-05, + "loss": 3.6081, + "step": 11704 + }, + { + "epoch": 0.6160526315789474, + "grad_norm": 0.5234375, + "learning_rate": 8.532163742690059e-05, + "loss": 3.3277, + "step": 11705 + }, + { + "epoch": 0.6161052631578947, + "grad_norm": 0.486328125, + "learning_rate": 8.530994152046784e-05, + "loss": 3.3068, + "step": 11706 + }, + { + "epoch": 0.6161578947368421, + "grad_norm": 0.51171875, + "learning_rate": 8.529824561403509e-05, + "loss": 3.8603, + "step": 11707 + }, + { + "epoch": 0.6162105263157894, + "grad_norm": 0.490234375, + "learning_rate": 8.528654970760235e-05, + "loss": 3.2064, + "step": 11708 + }, + { + "epoch": 0.6162631578947368, + "grad_norm": 0.5, + "learning_rate": 8.52748538011696e-05, + "loss": 3.5572, + "step": 11709 + }, + { + "epoch": 0.6163157894736843, + "grad_norm": 0.486328125, + "learning_rate": 8.526315789473685e-05, + "loss": 3.177, + "step": 11710 + }, + { + "epoch": 0.6163684210526316, + "grad_norm": 0.4765625, + "learning_rate": 8.52514619883041e-05, + "loss": 3.2073, + "step": 11711 + }, + { + "epoch": 0.616421052631579, + "grad_norm": 0.49609375, + "learning_rate": 8.523976608187135e-05, + "loss": 3.664, + "step": 11712 + }, + { + "epoch": 0.6164736842105263, + "grad_norm": 0.455078125, + "learning_rate": 8.52280701754386e-05, + "loss": 3.5154, + "step": 11713 + }, + { + "epoch": 0.6165263157894737, + "grad_norm": 0.47265625, + "learning_rate": 8.521637426900585e-05, + "loss": 3.2938, + "step": 11714 + }, + { + "epoch": 0.616578947368421, + "grad_norm": 0.466796875, + "learning_rate": 8.52046783625731e-05, + "loss": 3.419, + "step": 11715 + }, + { + "epoch": 0.6166315789473684, + "grad_norm": 0.494140625, + "learning_rate": 8.519298245614036e-05, + "loss": 3.2259, + "step": 11716 + }, + { + "epoch": 0.6166842105263158, + "grad_norm": 0.50390625, + "learning_rate": 8.518128654970761e-05, + "loss": 3.562, + "step": 11717 + }, + { + "epoch": 0.6167368421052631, + "grad_norm": 0.478515625, + "learning_rate": 8.516959064327486e-05, + "loss": 3.5233, + "step": 11718 + }, + { + "epoch": 0.6167894736842106, + "grad_norm": 0.515625, + "learning_rate": 8.515789473684211e-05, + "loss": 3.4912, + "step": 11719 + }, + { + "epoch": 0.6168421052631579, + "grad_norm": 0.46875, + "learning_rate": 8.514619883040936e-05, + "loss": 3.4569, + "step": 11720 + }, + { + "epoch": 0.6168947368421053, + "grad_norm": 0.5, + "learning_rate": 8.513450292397661e-05, + "loss": 3.7474, + "step": 11721 + }, + { + "epoch": 0.6169473684210526, + "grad_norm": 0.482421875, + "learning_rate": 8.512280701754386e-05, + "loss": 3.5267, + "step": 11722 + }, + { + "epoch": 0.617, + "grad_norm": 0.48828125, + "learning_rate": 8.511111111111112e-05, + "loss": 3.6296, + "step": 11723 + }, + { + "epoch": 0.6170526315789474, + "grad_norm": 0.52734375, + "learning_rate": 8.509941520467837e-05, + "loss": 3.2662, + "step": 11724 + }, + { + "epoch": 0.6171052631578947, + "grad_norm": 0.5234375, + "learning_rate": 8.508771929824562e-05, + "loss": 2.9798, + "step": 11725 + }, + { + "epoch": 0.6171578947368421, + "grad_norm": 0.515625, + "learning_rate": 8.507602339181287e-05, + "loss": 3.4168, + "step": 11726 + }, + { + "epoch": 0.6172105263157894, + "grad_norm": 0.515625, + "learning_rate": 8.506432748538012e-05, + "loss": 3.369, + "step": 11727 + }, + { + "epoch": 0.6172631578947368, + "grad_norm": 0.478515625, + "learning_rate": 8.505263157894737e-05, + "loss": 3.2352, + "step": 11728 + }, + { + "epoch": 0.6173157894736843, + "grad_norm": 0.50390625, + "learning_rate": 8.504093567251462e-05, + "loss": 3.4919, + "step": 11729 + }, + { + "epoch": 0.6173684210526316, + "grad_norm": 0.451171875, + "learning_rate": 8.502923976608188e-05, + "loss": 3.4158, + "step": 11730 + }, + { + "epoch": 0.617421052631579, + "grad_norm": 0.45703125, + "learning_rate": 8.501754385964913e-05, + "loss": 3.4578, + "step": 11731 + }, + { + "epoch": 0.6174736842105263, + "grad_norm": 0.490234375, + "learning_rate": 8.500584795321638e-05, + "loss": 3.5439, + "step": 11732 + }, + { + "epoch": 0.6175263157894737, + "grad_norm": 0.494140625, + "learning_rate": 8.499415204678363e-05, + "loss": 3.8236, + "step": 11733 + }, + { + "epoch": 0.617578947368421, + "grad_norm": 0.50390625, + "learning_rate": 8.498245614035088e-05, + "loss": 3.1953, + "step": 11734 + }, + { + "epoch": 0.6176315789473684, + "grad_norm": 0.4765625, + "learning_rate": 8.497076023391813e-05, + "loss": 3.5691, + "step": 11735 + }, + { + "epoch": 0.6176842105263158, + "grad_norm": 0.46875, + "learning_rate": 8.495906432748538e-05, + "loss": 2.8108, + "step": 11736 + }, + { + "epoch": 0.6177368421052631, + "grad_norm": 0.474609375, + "learning_rate": 8.494736842105263e-05, + "loss": 3.3344, + "step": 11737 + }, + { + "epoch": 0.6177894736842106, + "grad_norm": 0.58984375, + "learning_rate": 8.493567251461989e-05, + "loss": 3.407, + "step": 11738 + }, + { + "epoch": 0.6178421052631579, + "grad_norm": 0.49609375, + "learning_rate": 8.492397660818714e-05, + "loss": 3.7239, + "step": 11739 + }, + { + "epoch": 0.6178947368421053, + "grad_norm": 0.478515625, + "learning_rate": 8.491228070175439e-05, + "loss": 3.1831, + "step": 11740 + }, + { + "epoch": 0.6179473684210526, + "grad_norm": 0.474609375, + "learning_rate": 8.490058479532164e-05, + "loss": 3.4268, + "step": 11741 + }, + { + "epoch": 0.618, + "grad_norm": 0.55078125, + "learning_rate": 8.488888888888889e-05, + "loss": 3.0936, + "step": 11742 + }, + { + "epoch": 0.6180526315789474, + "grad_norm": 0.609375, + "learning_rate": 8.487719298245614e-05, + "loss": 3.1635, + "step": 11743 + }, + { + "epoch": 0.6181052631578947, + "grad_norm": 0.4765625, + "learning_rate": 8.486549707602339e-05, + "loss": 3.4639, + "step": 11744 + }, + { + "epoch": 0.6181578947368421, + "grad_norm": 0.515625, + "learning_rate": 8.485380116959065e-05, + "loss": 4.0727, + "step": 11745 + }, + { + "epoch": 0.6182105263157894, + "grad_norm": 0.5390625, + "learning_rate": 8.48421052631579e-05, + "loss": 3.8456, + "step": 11746 + }, + { + "epoch": 0.6182631578947368, + "grad_norm": 0.5234375, + "learning_rate": 8.483040935672515e-05, + "loss": 3.5629, + "step": 11747 + }, + { + "epoch": 0.6183157894736843, + "grad_norm": 0.46875, + "learning_rate": 8.48187134502924e-05, + "loss": 2.9908, + "step": 11748 + }, + { + "epoch": 0.6183684210526316, + "grad_norm": 0.578125, + "learning_rate": 8.480701754385965e-05, + "loss": 3.0986, + "step": 11749 + }, + { + "epoch": 0.618421052631579, + "grad_norm": 0.486328125, + "learning_rate": 8.47953216374269e-05, + "loss": 3.5711, + "step": 11750 + }, + { + "epoch": 0.6184736842105263, + "grad_norm": 0.51171875, + "learning_rate": 8.478362573099415e-05, + "loss": 3.3339, + "step": 11751 + }, + { + "epoch": 0.6185263157894737, + "grad_norm": 0.50390625, + "learning_rate": 8.47719298245614e-05, + "loss": 3.4805, + "step": 11752 + }, + { + "epoch": 0.618578947368421, + "grad_norm": 0.494140625, + "learning_rate": 8.476023391812866e-05, + "loss": 2.9924, + "step": 11753 + }, + { + "epoch": 0.6186315789473684, + "grad_norm": 0.494140625, + "learning_rate": 8.474853801169591e-05, + "loss": 3.5893, + "step": 11754 + }, + { + "epoch": 0.6186842105263158, + "grad_norm": 0.478515625, + "learning_rate": 8.473684210526316e-05, + "loss": 3.3538, + "step": 11755 + }, + { + "epoch": 0.6187368421052631, + "grad_norm": 0.484375, + "learning_rate": 8.472514619883041e-05, + "loss": 3.1026, + "step": 11756 + }, + { + "epoch": 0.6187894736842106, + "grad_norm": 0.4765625, + "learning_rate": 8.471345029239766e-05, + "loss": 3.17, + "step": 11757 + }, + { + "epoch": 0.6188421052631579, + "grad_norm": 0.5, + "learning_rate": 8.470175438596491e-05, + "loss": 3.8009, + "step": 11758 + }, + { + "epoch": 0.6188947368421053, + "grad_norm": 0.478515625, + "learning_rate": 8.469005847953216e-05, + "loss": 3.1963, + "step": 11759 + }, + { + "epoch": 0.6189473684210526, + "grad_norm": 0.47265625, + "learning_rate": 8.467836257309942e-05, + "loss": 3.2559, + "step": 11760 + }, + { + "epoch": 0.619, + "grad_norm": 0.50390625, + "learning_rate": 8.466666666666667e-05, + "loss": 3.3923, + "step": 11761 + }, + { + "epoch": 0.6190526315789474, + "grad_norm": 0.48828125, + "learning_rate": 8.465497076023392e-05, + "loss": 3.5068, + "step": 11762 + }, + { + "epoch": 0.6191052631578947, + "grad_norm": 0.4921875, + "learning_rate": 8.464327485380117e-05, + "loss": 3.0108, + "step": 11763 + }, + { + "epoch": 0.6191578947368421, + "grad_norm": 0.4765625, + "learning_rate": 8.463157894736842e-05, + "loss": 2.9549, + "step": 11764 + }, + { + "epoch": 0.6192105263157894, + "grad_norm": 0.53125, + "learning_rate": 8.461988304093567e-05, + "loss": 3.0891, + "step": 11765 + }, + { + "epoch": 0.6192631578947368, + "grad_norm": 0.482421875, + "learning_rate": 8.460818713450292e-05, + "loss": 3.5672, + "step": 11766 + }, + { + "epoch": 0.6193157894736843, + "grad_norm": 0.52734375, + "learning_rate": 8.459649122807018e-05, + "loss": 3.172, + "step": 11767 + }, + { + "epoch": 0.6193684210526316, + "grad_norm": 0.50390625, + "learning_rate": 8.458479532163743e-05, + "loss": 3.5546, + "step": 11768 + }, + { + "epoch": 0.619421052631579, + "grad_norm": 0.515625, + "learning_rate": 8.457309941520468e-05, + "loss": 3.5547, + "step": 11769 + }, + { + "epoch": 0.6194736842105263, + "grad_norm": 0.4921875, + "learning_rate": 8.456140350877193e-05, + "loss": 3.2913, + "step": 11770 + }, + { + "epoch": 0.6195263157894737, + "grad_norm": 0.474609375, + "learning_rate": 8.454970760233918e-05, + "loss": 3.441, + "step": 11771 + }, + { + "epoch": 0.619578947368421, + "grad_norm": 0.50390625, + "learning_rate": 8.453801169590643e-05, + "loss": 3.441, + "step": 11772 + }, + { + "epoch": 0.6196315789473684, + "grad_norm": 0.484375, + "learning_rate": 8.452631578947368e-05, + "loss": 3.2708, + "step": 11773 + }, + { + "epoch": 0.6196842105263158, + "grad_norm": 0.50390625, + "learning_rate": 8.451461988304093e-05, + "loss": 2.7933, + "step": 11774 + }, + { + "epoch": 0.6197368421052631, + "grad_norm": 0.484375, + "learning_rate": 8.45029239766082e-05, + "loss": 3.8316, + "step": 11775 + }, + { + "epoch": 0.6197894736842106, + "grad_norm": 0.482421875, + "learning_rate": 8.449122807017544e-05, + "loss": 3.4028, + "step": 11776 + }, + { + "epoch": 0.6198421052631579, + "grad_norm": 0.49609375, + "learning_rate": 8.44795321637427e-05, + "loss": 3.4479, + "step": 11777 + }, + { + "epoch": 0.6198947368421053, + "grad_norm": 0.484375, + "learning_rate": 8.446783625730994e-05, + "loss": 3.4764, + "step": 11778 + }, + { + "epoch": 0.6199473684210526, + "grad_norm": 0.54296875, + "learning_rate": 8.44561403508772e-05, + "loss": 3.2854, + "step": 11779 + }, + { + "epoch": 0.62, + "grad_norm": 0.466796875, + "learning_rate": 8.444444444444444e-05, + "loss": 3.1558, + "step": 11780 + }, + { + "epoch": 0.6200526315789474, + "grad_norm": 0.48046875, + "learning_rate": 8.443274853801169e-05, + "loss": 3.4836, + "step": 11781 + }, + { + "epoch": 0.6201052631578947, + "grad_norm": 0.484375, + "learning_rate": 8.442105263157896e-05, + "loss": 3.5781, + "step": 11782 + }, + { + "epoch": 0.6201578947368421, + "grad_norm": 0.5078125, + "learning_rate": 8.44093567251462e-05, + "loss": 3.3854, + "step": 11783 + }, + { + "epoch": 0.6202105263157894, + "grad_norm": 0.46875, + "learning_rate": 8.439766081871345e-05, + "loss": 3.5261, + "step": 11784 + }, + { + "epoch": 0.6202631578947368, + "grad_norm": 0.515625, + "learning_rate": 8.43859649122807e-05, + "loss": 3.3834, + "step": 11785 + }, + { + "epoch": 0.6203157894736843, + "grad_norm": 0.490234375, + "learning_rate": 8.437426900584797e-05, + "loss": 3.4287, + "step": 11786 + }, + { + "epoch": 0.6203684210526316, + "grad_norm": 0.470703125, + "learning_rate": 8.43625730994152e-05, + "loss": 3.0818, + "step": 11787 + }, + { + "epoch": 0.620421052631579, + "grad_norm": 0.45703125, + "learning_rate": 8.435087719298245e-05, + "loss": 3.1243, + "step": 11788 + }, + { + "epoch": 0.6204736842105263, + "grad_norm": 0.47265625, + "learning_rate": 8.43391812865497e-05, + "loss": 3.5483, + "step": 11789 + }, + { + "epoch": 0.6205263157894737, + "grad_norm": 0.5078125, + "learning_rate": 8.432748538011697e-05, + "loss": 3.3027, + "step": 11790 + }, + { + "epoch": 0.620578947368421, + "grad_norm": 0.53125, + "learning_rate": 8.431578947368422e-05, + "loss": 3.4185, + "step": 11791 + }, + { + "epoch": 0.6206315789473684, + "grad_norm": 0.52734375, + "learning_rate": 8.430409356725147e-05, + "loss": 3.0127, + "step": 11792 + }, + { + "epoch": 0.6206842105263158, + "grad_norm": 0.50390625, + "learning_rate": 8.429239766081873e-05, + "loss": 3.3884, + "step": 11793 + }, + { + "epoch": 0.6207368421052631, + "grad_norm": 0.51171875, + "learning_rate": 8.428070175438596e-05, + "loss": 3.3964, + "step": 11794 + }, + { + "epoch": 0.6207894736842106, + "grad_norm": 0.49609375, + "learning_rate": 8.426900584795321e-05, + "loss": 3.3404, + "step": 11795 + }, + { + "epoch": 0.6208421052631579, + "grad_norm": 0.474609375, + "learning_rate": 8.425730994152046e-05, + "loss": 3.2686, + "step": 11796 + }, + { + "epoch": 0.6208947368421053, + "grad_norm": 0.50390625, + "learning_rate": 8.424561403508773e-05, + "loss": 3.649, + "step": 11797 + }, + { + "epoch": 0.6209473684210526, + "grad_norm": 0.50390625, + "learning_rate": 8.423391812865498e-05, + "loss": 3.5663, + "step": 11798 + }, + { + "epoch": 0.621, + "grad_norm": 0.48828125, + "learning_rate": 8.422222222222223e-05, + "loss": 3.5098, + "step": 11799 + }, + { + "epoch": 0.6210526315789474, + "grad_norm": 0.48046875, + "learning_rate": 8.421052631578948e-05, + "loss": 3.3291, + "step": 11800 + }, + { + "epoch": 0.6211052631578947, + "grad_norm": 0.462890625, + "learning_rate": 8.419883040935674e-05, + "loss": 3.5712, + "step": 11801 + }, + { + "epoch": 0.6211578947368421, + "grad_norm": 0.498046875, + "learning_rate": 8.418713450292397e-05, + "loss": 3.4935, + "step": 11802 + }, + { + "epoch": 0.6212105263157894, + "grad_norm": 0.51171875, + "learning_rate": 8.417543859649122e-05, + "loss": 3.4078, + "step": 11803 + }, + { + "epoch": 0.6212631578947369, + "grad_norm": 0.50390625, + "learning_rate": 8.416374269005849e-05, + "loss": 2.8619, + "step": 11804 + }, + { + "epoch": 0.6213157894736843, + "grad_norm": 0.54296875, + "learning_rate": 8.415204678362574e-05, + "loss": 3.2085, + "step": 11805 + }, + { + "epoch": 0.6213684210526316, + "grad_norm": 0.484375, + "learning_rate": 8.414035087719299e-05, + "loss": 3.3305, + "step": 11806 + }, + { + "epoch": 0.621421052631579, + "grad_norm": 0.546875, + "learning_rate": 8.412865497076024e-05, + "loss": 3.3242, + "step": 11807 + }, + { + "epoch": 0.6214736842105263, + "grad_norm": 0.51171875, + "learning_rate": 8.41169590643275e-05, + "loss": 3.4751, + "step": 11808 + }, + { + "epoch": 0.6215263157894737, + "grad_norm": 0.52734375, + "learning_rate": 8.410526315789475e-05, + "loss": 3.3617, + "step": 11809 + }, + { + "epoch": 0.621578947368421, + "grad_norm": 0.515625, + "learning_rate": 8.409356725146199e-05, + "loss": 3.4911, + "step": 11810 + }, + { + "epoch": 0.6216315789473684, + "grad_norm": 0.48828125, + "learning_rate": 8.408187134502923e-05, + "loss": 3.0914, + "step": 11811 + }, + { + "epoch": 0.6216842105263158, + "grad_norm": 0.5, + "learning_rate": 8.40701754385965e-05, + "loss": 3.5615, + "step": 11812 + }, + { + "epoch": 0.6217368421052631, + "grad_norm": 0.49609375, + "learning_rate": 8.405847953216375e-05, + "loss": 3.1558, + "step": 11813 + }, + { + "epoch": 0.6217894736842106, + "grad_norm": 0.474609375, + "learning_rate": 8.4046783625731e-05, + "loss": 3.5124, + "step": 11814 + }, + { + "epoch": 0.6218421052631579, + "grad_norm": 0.4921875, + "learning_rate": 8.403508771929825e-05, + "loss": 3.0694, + "step": 11815 + }, + { + "epoch": 0.6218947368421053, + "grad_norm": 0.54296875, + "learning_rate": 8.402339181286551e-05, + "loss": 3.3849, + "step": 11816 + }, + { + "epoch": 0.6219473684210526, + "grad_norm": 0.51953125, + "learning_rate": 8.401169590643275e-05, + "loss": 3.1413, + "step": 11817 + }, + { + "epoch": 0.622, + "grad_norm": 0.486328125, + "learning_rate": 8.4e-05, + "loss": 3.6963, + "step": 11818 + }, + { + "epoch": 0.6220526315789474, + "grad_norm": 0.46875, + "learning_rate": 8.398830409356726e-05, + "loss": 3.7374, + "step": 11819 + }, + { + "epoch": 0.6221052631578947, + "grad_norm": 0.4765625, + "learning_rate": 8.397660818713451e-05, + "loss": 3.2758, + "step": 11820 + }, + { + "epoch": 0.6221578947368421, + "grad_norm": 0.470703125, + "learning_rate": 8.396491228070176e-05, + "loss": 3.3372, + "step": 11821 + }, + { + "epoch": 0.6222105263157894, + "grad_norm": 0.58203125, + "learning_rate": 8.395321637426901e-05, + "loss": 3.5819, + "step": 11822 + }, + { + "epoch": 0.6222631578947369, + "grad_norm": 0.51171875, + "learning_rate": 8.394152046783627e-05, + "loss": 3.3172, + "step": 11823 + }, + { + "epoch": 0.6223157894736842, + "grad_norm": 0.625, + "learning_rate": 8.392982456140352e-05, + "loss": 3.8666, + "step": 11824 + }, + { + "epoch": 0.6223684210526316, + "grad_norm": 0.47265625, + "learning_rate": 8.391812865497076e-05, + "loss": 2.9774, + "step": 11825 + }, + { + "epoch": 0.622421052631579, + "grad_norm": 0.5859375, + "learning_rate": 8.3906432748538e-05, + "loss": 3.1437, + "step": 11826 + }, + { + "epoch": 0.6224736842105263, + "grad_norm": 0.486328125, + "learning_rate": 8.389473684210527e-05, + "loss": 3.3703, + "step": 11827 + }, + { + "epoch": 0.6225263157894737, + "grad_norm": 0.46484375, + "learning_rate": 8.388304093567252e-05, + "loss": 3.2093, + "step": 11828 + }, + { + "epoch": 0.622578947368421, + "grad_norm": 0.482421875, + "learning_rate": 8.387134502923977e-05, + "loss": 3.3728, + "step": 11829 + }, + { + "epoch": 0.6226315789473684, + "grad_norm": 0.48046875, + "learning_rate": 8.385964912280703e-05, + "loss": 3.3241, + "step": 11830 + }, + { + "epoch": 0.6226842105263158, + "grad_norm": 0.53515625, + "learning_rate": 8.384795321637428e-05, + "loss": 3.252, + "step": 11831 + }, + { + "epoch": 0.6227368421052631, + "grad_norm": 0.470703125, + "learning_rate": 8.383625730994153e-05, + "loss": 3.8983, + "step": 11832 + }, + { + "epoch": 0.6227894736842106, + "grad_norm": 0.5390625, + "learning_rate": 8.382456140350877e-05, + "loss": 3.4105, + "step": 11833 + }, + { + "epoch": 0.6228421052631579, + "grad_norm": 0.46484375, + "learning_rate": 8.381286549707603e-05, + "loss": 3.288, + "step": 11834 + }, + { + "epoch": 0.6228947368421053, + "grad_norm": 0.48046875, + "learning_rate": 8.380116959064328e-05, + "loss": 3.1464, + "step": 11835 + }, + { + "epoch": 0.6229473684210526, + "grad_norm": 0.48046875, + "learning_rate": 8.378947368421053e-05, + "loss": 3.347, + "step": 11836 + }, + { + "epoch": 0.623, + "grad_norm": 0.5234375, + "learning_rate": 8.377777777777778e-05, + "loss": 2.9525, + "step": 11837 + }, + { + "epoch": 0.6230526315789474, + "grad_norm": 0.52734375, + "learning_rate": 8.376608187134504e-05, + "loss": 3.5062, + "step": 11838 + }, + { + "epoch": 0.6231052631578947, + "grad_norm": 0.51171875, + "learning_rate": 8.375438596491229e-05, + "loss": 3.1718, + "step": 11839 + }, + { + "epoch": 0.6231578947368421, + "grad_norm": 0.53515625, + "learning_rate": 8.374269005847953e-05, + "loss": 3.222, + "step": 11840 + }, + { + "epoch": 0.6232105263157894, + "grad_norm": 0.474609375, + "learning_rate": 8.373099415204678e-05, + "loss": 2.9101, + "step": 11841 + }, + { + "epoch": 0.6232631578947369, + "grad_norm": 0.45703125, + "learning_rate": 8.371929824561404e-05, + "loss": 3.3185, + "step": 11842 + }, + { + "epoch": 0.6233157894736842, + "grad_norm": 0.498046875, + "learning_rate": 8.370760233918129e-05, + "loss": 3.6741, + "step": 11843 + }, + { + "epoch": 0.6233684210526316, + "grad_norm": 0.47265625, + "learning_rate": 8.369590643274854e-05, + "loss": 3.8495, + "step": 11844 + }, + { + "epoch": 0.623421052631579, + "grad_norm": 0.8046875, + "learning_rate": 8.36842105263158e-05, + "loss": 3.4275, + "step": 11845 + }, + { + "epoch": 0.6234736842105263, + "grad_norm": 0.53515625, + "learning_rate": 8.367251461988305e-05, + "loss": 3.2452, + "step": 11846 + }, + { + "epoch": 0.6235263157894737, + "grad_norm": 0.490234375, + "learning_rate": 8.36608187134503e-05, + "loss": 3.0396, + "step": 11847 + }, + { + "epoch": 0.623578947368421, + "grad_norm": 0.52734375, + "learning_rate": 8.364912280701754e-05, + "loss": 3.1922, + "step": 11848 + }, + { + "epoch": 0.6236315789473684, + "grad_norm": 0.52734375, + "learning_rate": 8.36374269005848e-05, + "loss": 3.0851, + "step": 11849 + }, + { + "epoch": 0.6236842105263158, + "grad_norm": 0.5078125, + "learning_rate": 8.362573099415205e-05, + "loss": 3.4506, + "step": 11850 + }, + { + "epoch": 0.6237368421052631, + "grad_norm": 0.49609375, + "learning_rate": 8.36140350877193e-05, + "loss": 3.3533, + "step": 11851 + }, + { + "epoch": 0.6237894736842106, + "grad_norm": 0.5234375, + "learning_rate": 8.360233918128655e-05, + "loss": 3.6524, + "step": 11852 + }, + { + "epoch": 0.6238421052631579, + "grad_norm": 0.5, + "learning_rate": 8.359064327485381e-05, + "loss": 3.1128, + "step": 11853 + }, + { + "epoch": 0.6238947368421053, + "grad_norm": 0.494140625, + "learning_rate": 8.357894736842106e-05, + "loss": 3.2754, + "step": 11854 + }, + { + "epoch": 0.6239473684210526, + "grad_norm": 0.5, + "learning_rate": 8.356725146198831e-05, + "loss": 3.385, + "step": 11855 + }, + { + "epoch": 0.624, + "grad_norm": 0.46875, + "learning_rate": 8.355555555555556e-05, + "loss": 3.3152, + "step": 11856 + }, + { + "epoch": 0.6240526315789474, + "grad_norm": 0.50390625, + "learning_rate": 8.354385964912281e-05, + "loss": 3.4498, + "step": 11857 + }, + { + "epoch": 0.6241052631578947, + "grad_norm": 0.494140625, + "learning_rate": 8.353216374269006e-05, + "loss": 3.397, + "step": 11858 + }, + { + "epoch": 0.6241578947368421, + "grad_norm": 0.5, + "learning_rate": 8.352046783625731e-05, + "loss": 3.1145, + "step": 11859 + }, + { + "epoch": 0.6242105263157894, + "grad_norm": 0.466796875, + "learning_rate": 8.350877192982457e-05, + "loss": 3.0321, + "step": 11860 + }, + { + "epoch": 0.6242631578947369, + "grad_norm": 0.486328125, + "learning_rate": 8.349707602339182e-05, + "loss": 3.7335, + "step": 11861 + }, + { + "epoch": 0.6243157894736842, + "grad_norm": 0.5, + "learning_rate": 8.348538011695907e-05, + "loss": 3.0049, + "step": 11862 + }, + { + "epoch": 0.6243684210526316, + "grad_norm": 0.484375, + "learning_rate": 8.347368421052631e-05, + "loss": 3.3113, + "step": 11863 + }, + { + "epoch": 0.624421052631579, + "grad_norm": 0.50390625, + "learning_rate": 8.346198830409357e-05, + "loss": 3.1264, + "step": 11864 + }, + { + "epoch": 0.6244736842105263, + "grad_norm": 0.50390625, + "learning_rate": 8.345029239766082e-05, + "loss": 3.417, + "step": 11865 + }, + { + "epoch": 0.6245263157894737, + "grad_norm": 0.51171875, + "learning_rate": 8.343859649122807e-05, + "loss": 3.251, + "step": 11866 + }, + { + "epoch": 0.624578947368421, + "grad_norm": 0.458984375, + "learning_rate": 8.342690058479533e-05, + "loss": 3.2165, + "step": 11867 + }, + { + "epoch": 0.6246315789473684, + "grad_norm": 0.671875, + "learning_rate": 8.341520467836258e-05, + "loss": 3.1409, + "step": 11868 + }, + { + "epoch": 0.6246842105263158, + "grad_norm": 0.51953125, + "learning_rate": 8.340350877192983e-05, + "loss": 3.8358, + "step": 11869 + }, + { + "epoch": 0.6247368421052631, + "grad_norm": 0.57421875, + "learning_rate": 8.339181286549708e-05, + "loss": 3.4, + "step": 11870 + }, + { + "epoch": 0.6247894736842106, + "grad_norm": 0.52734375, + "learning_rate": 8.338011695906433e-05, + "loss": 3.2945, + "step": 11871 + }, + { + "epoch": 0.6248421052631579, + "grad_norm": 0.4765625, + "learning_rate": 8.336842105263158e-05, + "loss": 3.1155, + "step": 11872 + }, + { + "epoch": 0.6248947368421053, + "grad_norm": 0.5, + "learning_rate": 8.335672514619883e-05, + "loss": 3.6428, + "step": 11873 + }, + { + "epoch": 0.6249473684210526, + "grad_norm": 0.484375, + "learning_rate": 8.334502923976608e-05, + "loss": 3.0883, + "step": 11874 + }, + { + "epoch": 0.625, + "grad_norm": 0.478515625, + "learning_rate": 8.333333333333334e-05, + "loss": 3.4665, + "step": 11875 + }, + { + "epoch": 0.6250526315789474, + "grad_norm": 0.5, + "learning_rate": 8.33216374269006e-05, + "loss": 3.4064, + "step": 11876 + }, + { + "epoch": 0.6251052631578947, + "grad_norm": 0.490234375, + "learning_rate": 8.330994152046784e-05, + "loss": 3.4888, + "step": 11877 + }, + { + "epoch": 0.6251578947368421, + "grad_norm": 0.50390625, + "learning_rate": 8.329824561403509e-05, + "loss": 3.173, + "step": 11878 + }, + { + "epoch": 0.6252105263157894, + "grad_norm": 0.474609375, + "learning_rate": 8.328654970760234e-05, + "loss": 3.502, + "step": 11879 + }, + { + "epoch": 0.6252631578947369, + "grad_norm": 0.53515625, + "learning_rate": 8.327485380116959e-05, + "loss": 3.1953, + "step": 11880 + }, + { + "epoch": 0.6253157894736842, + "grad_norm": 0.466796875, + "learning_rate": 8.326315789473684e-05, + "loss": 3.1392, + "step": 11881 + }, + { + "epoch": 0.6253684210526316, + "grad_norm": 0.462890625, + "learning_rate": 8.32514619883041e-05, + "loss": 3.4276, + "step": 11882 + }, + { + "epoch": 0.625421052631579, + "grad_norm": 0.484375, + "learning_rate": 8.323976608187135e-05, + "loss": 3.2401, + "step": 11883 + }, + { + "epoch": 0.6254736842105263, + "grad_norm": 0.470703125, + "learning_rate": 8.32280701754386e-05, + "loss": 3.5301, + "step": 11884 + }, + { + "epoch": 0.6255263157894737, + "grad_norm": 0.4765625, + "learning_rate": 8.321637426900585e-05, + "loss": 3.4821, + "step": 11885 + }, + { + "epoch": 0.625578947368421, + "grad_norm": 0.51171875, + "learning_rate": 8.32046783625731e-05, + "loss": 3.604, + "step": 11886 + }, + { + "epoch": 0.6256315789473684, + "grad_norm": 0.494140625, + "learning_rate": 8.319298245614035e-05, + "loss": 3.0602, + "step": 11887 + }, + { + "epoch": 0.6256842105263158, + "grad_norm": 0.484375, + "learning_rate": 8.31812865497076e-05, + "loss": 3.7194, + "step": 11888 + }, + { + "epoch": 0.6257368421052631, + "grad_norm": 0.484375, + "learning_rate": 8.316959064327485e-05, + "loss": 3.845, + "step": 11889 + }, + { + "epoch": 0.6257894736842106, + "grad_norm": 0.50390625, + "learning_rate": 8.315789473684212e-05, + "loss": 3.1179, + "step": 11890 + }, + { + "epoch": 0.6258421052631579, + "grad_norm": 0.4765625, + "learning_rate": 8.314619883040936e-05, + "loss": 2.9372, + "step": 11891 + }, + { + "epoch": 0.6258947368421053, + "grad_norm": 0.5078125, + "learning_rate": 8.313450292397661e-05, + "loss": 3.732, + "step": 11892 + }, + { + "epoch": 0.6259473684210526, + "grad_norm": 0.5078125, + "learning_rate": 8.312280701754386e-05, + "loss": 3.4369, + "step": 11893 + }, + { + "epoch": 0.626, + "grad_norm": 0.5078125, + "learning_rate": 8.311111111111111e-05, + "loss": 3.1104, + "step": 11894 + }, + { + "epoch": 0.6260526315789474, + "grad_norm": 0.462890625, + "learning_rate": 8.309941520467836e-05, + "loss": 3.5305, + "step": 11895 + }, + { + "epoch": 0.6261052631578947, + "grad_norm": 0.484375, + "learning_rate": 8.308771929824561e-05, + "loss": 3.5259, + "step": 11896 + }, + { + "epoch": 0.6261578947368421, + "grad_norm": 0.494140625, + "learning_rate": 8.307602339181288e-05, + "loss": 3.3662, + "step": 11897 + }, + { + "epoch": 0.6262105263157894, + "grad_norm": 0.48046875, + "learning_rate": 8.306432748538013e-05, + "loss": 3.0761, + "step": 11898 + }, + { + "epoch": 0.6262631578947369, + "grad_norm": 0.53515625, + "learning_rate": 8.305263157894737e-05, + "loss": 3.0461, + "step": 11899 + }, + { + "epoch": 0.6263157894736842, + "grad_norm": 0.48828125, + "learning_rate": 8.304093567251462e-05, + "loss": 3.5475, + "step": 11900 + }, + { + "epoch": 0.6263684210526316, + "grad_norm": 0.478515625, + "learning_rate": 8.302923976608187e-05, + "loss": 3.22, + "step": 11901 + }, + { + "epoch": 0.626421052631579, + "grad_norm": 0.4609375, + "learning_rate": 8.301754385964912e-05, + "loss": 3.0708, + "step": 11902 + }, + { + "epoch": 0.6264736842105263, + "grad_norm": 0.5234375, + "learning_rate": 8.300584795321637e-05, + "loss": 3.5675, + "step": 11903 + }, + { + "epoch": 0.6265263157894737, + "grad_norm": 0.484375, + "learning_rate": 8.299415204678362e-05, + "loss": 3.4105, + "step": 11904 + }, + { + "epoch": 0.626578947368421, + "grad_norm": 0.482421875, + "learning_rate": 8.298245614035089e-05, + "loss": 3.3789, + "step": 11905 + }, + { + "epoch": 0.6266315789473684, + "grad_norm": 0.490234375, + "learning_rate": 8.297076023391814e-05, + "loss": 3.4003, + "step": 11906 + }, + { + "epoch": 0.6266842105263158, + "grad_norm": 0.57421875, + "learning_rate": 8.295906432748539e-05, + "loss": 3.5024, + "step": 11907 + }, + { + "epoch": 0.6267368421052631, + "grad_norm": 0.6875, + "learning_rate": 8.294736842105263e-05, + "loss": 2.951, + "step": 11908 + }, + { + "epoch": 0.6267894736842106, + "grad_norm": 0.466796875, + "learning_rate": 8.293567251461988e-05, + "loss": 3.3748, + "step": 11909 + }, + { + "epoch": 0.6268421052631579, + "grad_norm": 0.478515625, + "learning_rate": 8.292397660818713e-05, + "loss": 3.8792, + "step": 11910 + }, + { + "epoch": 0.6268947368421053, + "grad_norm": 0.470703125, + "learning_rate": 8.291228070175438e-05, + "loss": 3.4926, + "step": 11911 + }, + { + "epoch": 0.6269473684210526, + "grad_norm": 0.63671875, + "learning_rate": 8.290058479532165e-05, + "loss": 3.3273, + "step": 11912 + }, + { + "epoch": 0.627, + "grad_norm": 0.470703125, + "learning_rate": 8.28888888888889e-05, + "loss": 3.2136, + "step": 11913 + }, + { + "epoch": 0.6270526315789474, + "grad_norm": 0.494140625, + "learning_rate": 8.287719298245615e-05, + "loss": 3.5055, + "step": 11914 + }, + { + "epoch": 0.6271052631578947, + "grad_norm": 0.46484375, + "learning_rate": 8.28654970760234e-05, + "loss": 3.5418, + "step": 11915 + }, + { + "epoch": 0.6271578947368421, + "grad_norm": 0.49609375, + "learning_rate": 8.285380116959065e-05, + "loss": 3.8191, + "step": 11916 + }, + { + "epoch": 0.6272105263157894, + "grad_norm": 0.466796875, + "learning_rate": 8.28421052631579e-05, + "loss": 3.3264, + "step": 11917 + }, + { + "epoch": 0.6272631578947369, + "grad_norm": 0.4921875, + "learning_rate": 8.283040935672514e-05, + "loss": 3.3586, + "step": 11918 + }, + { + "epoch": 0.6273157894736842, + "grad_norm": 0.5, + "learning_rate": 8.281871345029241e-05, + "loss": 3.6156, + "step": 11919 + }, + { + "epoch": 0.6273684210526316, + "grad_norm": 0.63671875, + "learning_rate": 8.280701754385966e-05, + "loss": 3.6634, + "step": 11920 + }, + { + "epoch": 0.627421052631579, + "grad_norm": 0.51171875, + "learning_rate": 8.27953216374269e-05, + "loss": 3.4536, + "step": 11921 + }, + { + "epoch": 0.6274736842105263, + "grad_norm": 0.466796875, + "learning_rate": 8.278362573099416e-05, + "loss": 3.1266, + "step": 11922 + }, + { + "epoch": 0.6275263157894737, + "grad_norm": 0.515625, + "learning_rate": 8.27719298245614e-05, + "loss": 3.5862, + "step": 11923 + }, + { + "epoch": 0.627578947368421, + "grad_norm": 0.51171875, + "learning_rate": 8.276023391812866e-05, + "loss": 3.3881, + "step": 11924 + }, + { + "epoch": 0.6276315789473684, + "grad_norm": 0.87109375, + "learning_rate": 8.27485380116959e-05, + "loss": 3.3392, + "step": 11925 + }, + { + "epoch": 0.6276842105263158, + "grad_norm": 0.4609375, + "learning_rate": 8.273684210526315e-05, + "loss": 3.6004, + "step": 11926 + }, + { + "epoch": 0.6277368421052631, + "grad_norm": 0.4765625, + "learning_rate": 8.272514619883042e-05, + "loss": 3.4657, + "step": 11927 + }, + { + "epoch": 0.6277894736842106, + "grad_norm": 0.50390625, + "learning_rate": 8.271345029239767e-05, + "loss": 3.3024, + "step": 11928 + }, + { + "epoch": 0.6278421052631579, + "grad_norm": 0.4921875, + "learning_rate": 8.270175438596492e-05, + "loss": 3.2301, + "step": 11929 + }, + { + "epoch": 0.6278947368421053, + "grad_norm": 0.478515625, + "learning_rate": 8.269005847953217e-05, + "loss": 3.2735, + "step": 11930 + }, + { + "epoch": 0.6279473684210526, + "grad_norm": 0.4921875, + "learning_rate": 8.267836257309942e-05, + "loss": 3.3089, + "step": 11931 + }, + { + "epoch": 0.628, + "grad_norm": 0.50390625, + "learning_rate": 8.266666666666667e-05, + "loss": 2.7989, + "step": 11932 + }, + { + "epoch": 0.6280526315789474, + "grad_norm": 0.486328125, + "learning_rate": 8.265497076023392e-05, + "loss": 3.3658, + "step": 11933 + }, + { + "epoch": 0.6281052631578947, + "grad_norm": 0.494140625, + "learning_rate": 8.264327485380118e-05, + "loss": 3.6019, + "step": 11934 + }, + { + "epoch": 0.6281578947368421, + "grad_norm": 0.4921875, + "learning_rate": 8.263157894736843e-05, + "loss": 3.6344, + "step": 11935 + }, + { + "epoch": 0.6282105263157894, + "grad_norm": 0.470703125, + "learning_rate": 8.261988304093568e-05, + "loss": 3.2362, + "step": 11936 + }, + { + "epoch": 0.6282631578947369, + "grad_norm": 0.498046875, + "learning_rate": 8.260818713450293e-05, + "loss": 3.1904, + "step": 11937 + }, + { + "epoch": 0.6283157894736842, + "grad_norm": 0.50390625, + "learning_rate": 8.259649122807018e-05, + "loss": 3.2573, + "step": 11938 + }, + { + "epoch": 0.6283684210526316, + "grad_norm": 0.5234375, + "learning_rate": 8.258479532163743e-05, + "loss": 3.1266, + "step": 11939 + }, + { + "epoch": 0.628421052631579, + "grad_norm": 0.484375, + "learning_rate": 8.257309941520468e-05, + "loss": 3.4318, + "step": 11940 + }, + { + "epoch": 0.6284736842105263, + "grad_norm": 0.482421875, + "learning_rate": 8.256140350877193e-05, + "loss": 3.523, + "step": 11941 + }, + { + "epoch": 0.6285263157894737, + "grad_norm": 0.458984375, + "learning_rate": 8.254970760233919e-05, + "loss": 3.2296, + "step": 11942 + }, + { + "epoch": 0.628578947368421, + "grad_norm": 0.494140625, + "learning_rate": 8.253801169590644e-05, + "loss": 3.4046, + "step": 11943 + }, + { + "epoch": 0.6286315789473684, + "grad_norm": 0.462890625, + "learning_rate": 8.252631578947369e-05, + "loss": 3.4026, + "step": 11944 + }, + { + "epoch": 0.6286842105263157, + "grad_norm": 0.47265625, + "learning_rate": 8.251461988304095e-05, + "loss": 3.6943, + "step": 11945 + }, + { + "epoch": 0.6287368421052631, + "grad_norm": 0.46875, + "learning_rate": 8.250292397660819e-05, + "loss": 3.3075, + "step": 11946 + }, + { + "epoch": 0.6287894736842106, + "grad_norm": 0.478515625, + "learning_rate": 8.249122807017544e-05, + "loss": 3.3467, + "step": 11947 + }, + { + "epoch": 0.6288421052631579, + "grad_norm": 0.474609375, + "learning_rate": 8.247953216374269e-05, + "loss": 3.4239, + "step": 11948 + }, + { + "epoch": 0.6288947368421053, + "grad_norm": 0.48828125, + "learning_rate": 8.246783625730995e-05, + "loss": 3.5958, + "step": 11949 + }, + { + "epoch": 0.6289473684210526, + "grad_norm": 0.474609375, + "learning_rate": 8.24561403508772e-05, + "loss": 3.3471, + "step": 11950 + }, + { + "epoch": 0.629, + "grad_norm": 0.490234375, + "learning_rate": 8.244444444444445e-05, + "loss": 3.4279, + "step": 11951 + }, + { + "epoch": 0.6290526315789474, + "grad_norm": 0.546875, + "learning_rate": 8.24327485380117e-05, + "loss": 3.5306, + "step": 11952 + }, + { + "epoch": 0.6291052631578947, + "grad_norm": 0.490234375, + "learning_rate": 8.242105263157895e-05, + "loss": 3.5899, + "step": 11953 + }, + { + "epoch": 0.6291578947368421, + "grad_norm": 0.5078125, + "learning_rate": 8.24093567251462e-05, + "loss": 3.3168, + "step": 11954 + }, + { + "epoch": 0.6292105263157894, + "grad_norm": 0.4765625, + "learning_rate": 8.239766081871345e-05, + "loss": 2.8681, + "step": 11955 + }, + { + "epoch": 0.6292631578947369, + "grad_norm": 0.4453125, + "learning_rate": 8.238596491228071e-05, + "loss": 3.3344, + "step": 11956 + }, + { + "epoch": 0.6293157894736842, + "grad_norm": 0.486328125, + "learning_rate": 8.237426900584796e-05, + "loss": 3.3309, + "step": 11957 + }, + { + "epoch": 0.6293684210526316, + "grad_norm": 0.498046875, + "learning_rate": 8.236257309941521e-05, + "loss": 3.5356, + "step": 11958 + }, + { + "epoch": 0.629421052631579, + "grad_norm": 0.484375, + "learning_rate": 8.235087719298246e-05, + "loss": 3.5948, + "step": 11959 + }, + { + "epoch": 0.6294736842105263, + "grad_norm": 0.482421875, + "learning_rate": 8.233918128654972e-05, + "loss": 3.3655, + "step": 11960 + }, + { + "epoch": 0.6295263157894737, + "grad_norm": 0.5078125, + "learning_rate": 8.232748538011696e-05, + "loss": 3.2457, + "step": 11961 + }, + { + "epoch": 0.629578947368421, + "grad_norm": 0.46484375, + "learning_rate": 8.231578947368421e-05, + "loss": 3.1654, + "step": 11962 + }, + { + "epoch": 0.6296315789473684, + "grad_norm": 0.59765625, + "learning_rate": 8.230409356725146e-05, + "loss": 3.5542, + "step": 11963 + }, + { + "epoch": 0.6296842105263157, + "grad_norm": 0.46875, + "learning_rate": 8.229239766081872e-05, + "loss": 3.3156, + "step": 11964 + }, + { + "epoch": 0.6297368421052632, + "grad_norm": 0.51171875, + "learning_rate": 8.228070175438597e-05, + "loss": 3.4108, + "step": 11965 + }, + { + "epoch": 0.6297894736842106, + "grad_norm": 0.484375, + "learning_rate": 8.226900584795322e-05, + "loss": 3.4793, + "step": 11966 + }, + { + "epoch": 0.6298421052631579, + "grad_norm": 0.484375, + "learning_rate": 8.225730994152048e-05, + "loss": 3.5312, + "step": 11967 + }, + { + "epoch": 0.6298947368421053, + "grad_norm": 0.4765625, + "learning_rate": 8.224561403508773e-05, + "loss": 3.7228, + "step": 11968 + }, + { + "epoch": 0.6299473684210526, + "grad_norm": 0.5234375, + "learning_rate": 8.223391812865497e-05, + "loss": 3.1422, + "step": 11969 + }, + { + "epoch": 0.63, + "grad_norm": 0.478515625, + "learning_rate": 8.222222222222222e-05, + "loss": 3.5733, + "step": 11970 + }, + { + "epoch": 0.6300526315789474, + "grad_norm": 0.50390625, + "learning_rate": 8.221052631578948e-05, + "loss": 3.4563, + "step": 11971 + }, + { + "epoch": 0.6301052631578947, + "grad_norm": 0.53125, + "learning_rate": 8.219883040935673e-05, + "loss": 3.523, + "step": 11972 + }, + { + "epoch": 0.6301578947368421, + "grad_norm": 0.49609375, + "learning_rate": 8.218713450292398e-05, + "loss": 3.3877, + "step": 11973 + }, + { + "epoch": 0.6302105263157894, + "grad_norm": 0.48828125, + "learning_rate": 8.217543859649123e-05, + "loss": 3.3839, + "step": 11974 + }, + { + "epoch": 0.6302631578947369, + "grad_norm": 0.462890625, + "learning_rate": 8.216374269005849e-05, + "loss": 3.9992, + "step": 11975 + }, + { + "epoch": 0.6303157894736842, + "grad_norm": 0.5390625, + "learning_rate": 8.215204678362573e-05, + "loss": 3.5452, + "step": 11976 + }, + { + "epoch": 0.6303684210526316, + "grad_norm": 0.5390625, + "learning_rate": 8.214035087719298e-05, + "loss": 2.958, + "step": 11977 + }, + { + "epoch": 0.630421052631579, + "grad_norm": 0.484375, + "learning_rate": 8.212865497076023e-05, + "loss": 3.2996, + "step": 11978 + }, + { + "epoch": 0.6304736842105263, + "grad_norm": 0.474609375, + "learning_rate": 8.211695906432749e-05, + "loss": 3.3344, + "step": 11979 + }, + { + "epoch": 0.6305263157894737, + "grad_norm": 0.466796875, + "learning_rate": 8.210526315789474e-05, + "loss": 3.4593, + "step": 11980 + }, + { + "epoch": 0.630578947368421, + "grad_norm": 0.50390625, + "learning_rate": 8.209356725146199e-05, + "loss": 2.8137, + "step": 11981 + }, + { + "epoch": 0.6306315789473684, + "grad_norm": 0.498046875, + "learning_rate": 8.208187134502925e-05, + "loss": 2.8994, + "step": 11982 + }, + { + "epoch": 0.6306842105263157, + "grad_norm": 0.48828125, + "learning_rate": 8.20701754385965e-05, + "loss": 3.3413, + "step": 11983 + }, + { + "epoch": 0.6307368421052632, + "grad_norm": 0.494140625, + "learning_rate": 8.205847953216374e-05, + "loss": 3.365, + "step": 11984 + }, + { + "epoch": 0.6307894736842106, + "grad_norm": 0.490234375, + "learning_rate": 8.204678362573099e-05, + "loss": 3.2159, + "step": 11985 + }, + { + "epoch": 0.6308421052631579, + "grad_norm": 0.4765625, + "learning_rate": 8.203508771929825e-05, + "loss": 3.307, + "step": 11986 + }, + { + "epoch": 0.6308947368421053, + "grad_norm": 0.5390625, + "learning_rate": 8.20233918128655e-05, + "loss": 3.17, + "step": 11987 + }, + { + "epoch": 0.6309473684210526, + "grad_norm": 0.5703125, + "learning_rate": 8.201169590643275e-05, + "loss": 3.6758, + "step": 11988 + }, + { + "epoch": 0.631, + "grad_norm": 0.50390625, + "learning_rate": 8.2e-05, + "loss": 3.2834, + "step": 11989 + }, + { + "epoch": 0.6310526315789474, + "grad_norm": 0.51953125, + "learning_rate": 8.198830409356726e-05, + "loss": 2.9422, + "step": 11990 + }, + { + "epoch": 0.6311052631578947, + "grad_norm": 0.474609375, + "learning_rate": 8.197660818713451e-05, + "loss": 3.3218, + "step": 11991 + }, + { + "epoch": 0.6311578947368421, + "grad_norm": 0.5078125, + "learning_rate": 8.196491228070175e-05, + "loss": 3.2539, + "step": 11992 + }, + { + "epoch": 0.6312105263157894, + "grad_norm": 0.50390625, + "learning_rate": 8.195321637426901e-05, + "loss": 3.5661, + "step": 11993 + }, + { + "epoch": 0.6312631578947369, + "grad_norm": 0.466796875, + "learning_rate": 8.194152046783626e-05, + "loss": 3.326, + "step": 11994 + }, + { + "epoch": 0.6313157894736842, + "grad_norm": 0.50390625, + "learning_rate": 8.192982456140351e-05, + "loss": 3.3568, + "step": 11995 + }, + { + "epoch": 0.6313684210526316, + "grad_norm": 0.49609375, + "learning_rate": 8.191812865497076e-05, + "loss": 3.1854, + "step": 11996 + }, + { + "epoch": 0.631421052631579, + "grad_norm": 0.48828125, + "learning_rate": 8.190643274853802e-05, + "loss": 3.5735, + "step": 11997 + }, + { + "epoch": 0.6314736842105263, + "grad_norm": 0.5078125, + "learning_rate": 8.189473684210527e-05, + "loss": 3.7001, + "step": 11998 + }, + { + "epoch": 0.6315263157894737, + "grad_norm": 0.48828125, + "learning_rate": 8.188304093567251e-05, + "loss": 4.0689, + "step": 11999 + }, + { + "epoch": 0.631578947368421, + "grad_norm": 0.75, + "learning_rate": 8.187134502923976e-05, + "loss": 3.4813, + "step": 12000 + }, + { + "epoch": 0.6316315789473684, + "grad_norm": 0.5, + "learning_rate": 8.185964912280702e-05, + "loss": 3.3832, + "step": 12001 + }, + { + "epoch": 0.6316842105263157, + "grad_norm": 0.46484375, + "learning_rate": 8.184795321637427e-05, + "loss": 3.2683, + "step": 12002 + }, + { + "epoch": 0.6317368421052632, + "grad_norm": 0.54296875, + "learning_rate": 8.183625730994152e-05, + "loss": 3.5729, + "step": 12003 + }, + { + "epoch": 0.6317894736842106, + "grad_norm": 0.47265625, + "learning_rate": 8.182456140350877e-05, + "loss": 3.4454, + "step": 12004 + }, + { + "epoch": 0.6318421052631579, + "grad_norm": 0.484375, + "learning_rate": 8.181286549707604e-05, + "loss": 3.2472, + "step": 12005 + }, + { + "epoch": 0.6318947368421053, + "grad_norm": 0.48828125, + "learning_rate": 8.180116959064328e-05, + "loss": 3.3775, + "step": 12006 + }, + { + "epoch": 0.6319473684210526, + "grad_norm": 0.5390625, + "learning_rate": 8.178947368421052e-05, + "loss": 3.5245, + "step": 12007 + }, + { + "epoch": 0.632, + "grad_norm": 0.484375, + "learning_rate": 8.177777777777778e-05, + "loss": 3.2734, + "step": 12008 + }, + { + "epoch": 0.6320526315789474, + "grad_norm": 0.46875, + "learning_rate": 8.176608187134503e-05, + "loss": 3.4552, + "step": 12009 + }, + { + "epoch": 0.6321052631578947, + "grad_norm": 0.52734375, + "learning_rate": 8.175438596491228e-05, + "loss": 3.2628, + "step": 12010 + }, + { + "epoch": 0.6321578947368421, + "grad_norm": 0.51953125, + "learning_rate": 8.174269005847953e-05, + "loss": 3.3201, + "step": 12011 + }, + { + "epoch": 0.6322105263157894, + "grad_norm": 0.4921875, + "learning_rate": 8.17309941520468e-05, + "loss": 3.3685, + "step": 12012 + }, + { + "epoch": 0.6322631578947369, + "grad_norm": 0.5078125, + "learning_rate": 8.171929824561405e-05, + "loss": 3.4996, + "step": 12013 + }, + { + "epoch": 0.6323157894736842, + "grad_norm": 0.4921875, + "learning_rate": 8.17076023391813e-05, + "loss": 3.4783, + "step": 12014 + }, + { + "epoch": 0.6323684210526316, + "grad_norm": 0.482421875, + "learning_rate": 8.169590643274853e-05, + "loss": 3.4323, + "step": 12015 + }, + { + "epoch": 0.632421052631579, + "grad_norm": 0.4921875, + "learning_rate": 8.16842105263158e-05, + "loss": 3.6026, + "step": 12016 + }, + { + "epoch": 0.6324736842105263, + "grad_norm": 0.44921875, + "learning_rate": 8.167251461988304e-05, + "loss": 2.8854, + "step": 12017 + }, + { + "epoch": 0.6325263157894737, + "grad_norm": 0.515625, + "learning_rate": 8.16608187134503e-05, + "loss": 3.3745, + "step": 12018 + }, + { + "epoch": 0.632578947368421, + "grad_norm": 0.49609375, + "learning_rate": 8.164912280701756e-05, + "loss": 3.6288, + "step": 12019 + }, + { + "epoch": 0.6326315789473684, + "grad_norm": 0.51953125, + "learning_rate": 8.16374269005848e-05, + "loss": 3.2573, + "step": 12020 + }, + { + "epoch": 0.6326842105263157, + "grad_norm": 0.478515625, + "learning_rate": 8.162573099415206e-05, + "loss": 3.4136, + "step": 12021 + }, + { + "epoch": 0.6327368421052632, + "grad_norm": 0.50390625, + "learning_rate": 8.161403508771929e-05, + "loss": 3.6262, + "step": 12022 + }, + { + "epoch": 0.6327894736842106, + "grad_norm": 0.77734375, + "learning_rate": 8.160233918128655e-05, + "loss": 3.4627, + "step": 12023 + }, + { + "epoch": 0.6328421052631579, + "grad_norm": 0.474609375, + "learning_rate": 8.15906432748538e-05, + "loss": 3.2676, + "step": 12024 + }, + { + "epoch": 0.6328947368421053, + "grad_norm": 0.484375, + "learning_rate": 8.157894736842105e-05, + "loss": 3.5105, + "step": 12025 + }, + { + "epoch": 0.6329473684210526, + "grad_norm": 0.58203125, + "learning_rate": 8.15672514619883e-05, + "loss": 3.3592, + "step": 12026 + }, + { + "epoch": 0.633, + "grad_norm": 0.482421875, + "learning_rate": 8.155555555555557e-05, + "loss": 4.0966, + "step": 12027 + }, + { + "epoch": 0.6330526315789474, + "grad_norm": 0.54296875, + "learning_rate": 8.154385964912282e-05, + "loss": 3.4089, + "step": 12028 + }, + { + "epoch": 0.6331052631578947, + "grad_norm": 0.484375, + "learning_rate": 8.153216374269007e-05, + "loss": 3.2353, + "step": 12029 + }, + { + "epoch": 0.6331578947368421, + "grad_norm": 0.47265625, + "learning_rate": 8.152046783625732e-05, + "loss": 3.5236, + "step": 12030 + }, + { + "epoch": 0.6332105263157894, + "grad_norm": 0.671875, + "learning_rate": 8.150877192982457e-05, + "loss": 3.6731, + "step": 12031 + }, + { + "epoch": 0.6332631578947369, + "grad_norm": 0.486328125, + "learning_rate": 8.149707602339181e-05, + "loss": 3.2802, + "step": 12032 + }, + { + "epoch": 0.6333157894736842, + "grad_norm": 0.470703125, + "learning_rate": 8.148538011695906e-05, + "loss": 2.5941, + "step": 12033 + }, + { + "epoch": 0.6333684210526316, + "grad_norm": 0.4765625, + "learning_rate": 8.147368421052633e-05, + "loss": 3.1736, + "step": 12034 + }, + { + "epoch": 0.633421052631579, + "grad_norm": 0.47265625, + "learning_rate": 8.146198830409358e-05, + "loss": 3.4678, + "step": 12035 + }, + { + "epoch": 0.6334736842105263, + "grad_norm": 0.462890625, + "learning_rate": 8.145029239766083e-05, + "loss": 2.9374, + "step": 12036 + }, + { + "epoch": 0.6335263157894737, + "grad_norm": 0.484375, + "learning_rate": 8.143859649122808e-05, + "loss": 3.1811, + "step": 12037 + }, + { + "epoch": 0.633578947368421, + "grad_norm": 0.46875, + "learning_rate": 8.142690058479533e-05, + "loss": 3.2172, + "step": 12038 + }, + { + "epoch": 0.6336315789473684, + "grad_norm": 0.474609375, + "learning_rate": 8.141520467836258e-05, + "loss": 3.1044, + "step": 12039 + }, + { + "epoch": 0.6336842105263157, + "grad_norm": 0.490234375, + "learning_rate": 8.140350877192983e-05, + "loss": 3.5058, + "step": 12040 + }, + { + "epoch": 0.6337368421052632, + "grad_norm": 0.494140625, + "learning_rate": 8.139181286549707e-05, + "loss": 3.6116, + "step": 12041 + }, + { + "epoch": 0.6337894736842106, + "grad_norm": 0.48828125, + "learning_rate": 8.138011695906434e-05, + "loss": 3.5471, + "step": 12042 + }, + { + "epoch": 0.6338421052631579, + "grad_norm": 0.46875, + "learning_rate": 8.136842105263159e-05, + "loss": 3.197, + "step": 12043 + }, + { + "epoch": 0.6338947368421053, + "grad_norm": 0.51171875, + "learning_rate": 8.135672514619884e-05, + "loss": 3.4676, + "step": 12044 + }, + { + "epoch": 0.6339473684210526, + "grad_norm": 0.49609375, + "learning_rate": 8.134502923976609e-05, + "loss": 3.1775, + "step": 12045 + }, + { + "epoch": 0.634, + "grad_norm": 0.5703125, + "learning_rate": 8.133333333333334e-05, + "loss": 3.3154, + "step": 12046 + }, + { + "epoch": 0.6340526315789474, + "grad_norm": 0.50390625, + "learning_rate": 8.132163742690059e-05, + "loss": 3.4312, + "step": 12047 + }, + { + "epoch": 0.6341052631578947, + "grad_norm": 0.4921875, + "learning_rate": 8.130994152046784e-05, + "loss": 3.1439, + "step": 12048 + }, + { + "epoch": 0.6341578947368421, + "grad_norm": 0.5390625, + "learning_rate": 8.12982456140351e-05, + "loss": 3.3466, + "step": 12049 + }, + { + "epoch": 0.6342105263157894, + "grad_norm": 0.482421875, + "learning_rate": 8.128654970760235e-05, + "loss": 3.316, + "step": 12050 + }, + { + "epoch": 0.6342631578947369, + "grad_norm": 0.486328125, + "learning_rate": 8.12748538011696e-05, + "loss": 3.0271, + "step": 12051 + }, + { + "epoch": 0.6343157894736842, + "grad_norm": 0.498046875, + "learning_rate": 8.126315789473685e-05, + "loss": 3.2629, + "step": 12052 + }, + { + "epoch": 0.6343684210526316, + "grad_norm": 0.494140625, + "learning_rate": 8.12514619883041e-05, + "loss": 3.0344, + "step": 12053 + }, + { + "epoch": 0.634421052631579, + "grad_norm": 0.5234375, + "learning_rate": 8.123976608187135e-05, + "loss": 3.2439, + "step": 12054 + }, + { + "epoch": 0.6344736842105263, + "grad_norm": 0.5, + "learning_rate": 8.12280701754386e-05, + "loss": 3.6592, + "step": 12055 + }, + { + "epoch": 0.6345263157894737, + "grad_norm": 0.5078125, + "learning_rate": 8.121637426900586e-05, + "loss": 3.4198, + "step": 12056 + }, + { + "epoch": 0.634578947368421, + "grad_norm": 0.490234375, + "learning_rate": 8.120467836257311e-05, + "loss": 3.6361, + "step": 12057 + }, + { + "epoch": 0.6346315789473684, + "grad_norm": 0.53515625, + "learning_rate": 8.119298245614036e-05, + "loss": 3.3638, + "step": 12058 + }, + { + "epoch": 0.6346842105263157, + "grad_norm": 0.482421875, + "learning_rate": 8.118128654970761e-05, + "loss": 3.4329, + "step": 12059 + }, + { + "epoch": 0.6347368421052632, + "grad_norm": 0.46875, + "learning_rate": 8.116959064327486e-05, + "loss": 3.5311, + "step": 12060 + }, + { + "epoch": 0.6347894736842106, + "grad_norm": 0.482421875, + "learning_rate": 8.115789473684211e-05, + "loss": 3.6458, + "step": 12061 + }, + { + "epoch": 0.6348421052631579, + "grad_norm": 0.44140625, + "learning_rate": 8.114619883040936e-05, + "loss": 2.8164, + "step": 12062 + }, + { + "epoch": 0.6348947368421053, + "grad_norm": 0.474609375, + "learning_rate": 8.11345029239766e-05, + "loss": 3.351, + "step": 12063 + }, + { + "epoch": 0.6349473684210526, + "grad_norm": 0.466796875, + "learning_rate": 8.112280701754387e-05, + "loss": 2.8575, + "step": 12064 + }, + { + "epoch": 0.635, + "grad_norm": 0.5625, + "learning_rate": 8.111111111111112e-05, + "loss": 3.8137, + "step": 12065 + }, + { + "epoch": 0.6350526315789473, + "grad_norm": 0.5625, + "learning_rate": 8.109941520467837e-05, + "loss": 3.2388, + "step": 12066 + }, + { + "epoch": 0.6351052631578947, + "grad_norm": 0.466796875, + "learning_rate": 8.108771929824562e-05, + "loss": 3.5181, + "step": 12067 + }, + { + "epoch": 0.6351578947368421, + "grad_norm": 0.60546875, + "learning_rate": 8.107602339181287e-05, + "loss": 2.9592, + "step": 12068 + }, + { + "epoch": 0.6352105263157894, + "grad_norm": 0.50390625, + "learning_rate": 8.106432748538012e-05, + "loss": 3.7482, + "step": 12069 + }, + { + "epoch": 0.6352631578947369, + "grad_norm": 0.47265625, + "learning_rate": 8.105263157894737e-05, + "loss": 3.2027, + "step": 12070 + }, + { + "epoch": 0.6353157894736842, + "grad_norm": 0.470703125, + "learning_rate": 8.104093567251463e-05, + "loss": 3.6989, + "step": 12071 + }, + { + "epoch": 0.6353684210526316, + "grad_norm": 0.55078125, + "learning_rate": 8.102923976608188e-05, + "loss": 3.5047, + "step": 12072 + }, + { + "epoch": 0.635421052631579, + "grad_norm": 0.5546875, + "learning_rate": 8.101754385964913e-05, + "loss": 3.6524, + "step": 12073 + }, + { + "epoch": 0.6354736842105263, + "grad_norm": 0.482421875, + "learning_rate": 8.100584795321638e-05, + "loss": 3.3119, + "step": 12074 + }, + { + "epoch": 0.6355263157894737, + "grad_norm": 0.466796875, + "learning_rate": 8.099415204678363e-05, + "loss": 3.0892, + "step": 12075 + }, + { + "epoch": 0.635578947368421, + "grad_norm": 0.494140625, + "learning_rate": 8.098245614035088e-05, + "loss": 3.1503, + "step": 12076 + }, + { + "epoch": 0.6356315789473684, + "grad_norm": 0.48828125, + "learning_rate": 8.097076023391813e-05, + "loss": 3.1859, + "step": 12077 + }, + { + "epoch": 0.6356842105263157, + "grad_norm": 0.49609375, + "learning_rate": 8.095906432748538e-05, + "loss": 3.0184, + "step": 12078 + }, + { + "epoch": 0.6357368421052632, + "grad_norm": 0.4765625, + "learning_rate": 8.094736842105264e-05, + "loss": 3.7559, + "step": 12079 + }, + { + "epoch": 0.6357894736842106, + "grad_norm": 0.57421875, + "learning_rate": 8.093567251461989e-05, + "loss": 3.2545, + "step": 12080 + }, + { + "epoch": 0.6358421052631579, + "grad_norm": 0.50390625, + "learning_rate": 8.092397660818714e-05, + "loss": 3.251, + "step": 12081 + }, + { + "epoch": 0.6358947368421053, + "grad_norm": 0.546875, + "learning_rate": 8.091228070175439e-05, + "loss": 2.7854, + "step": 12082 + }, + { + "epoch": 0.6359473684210526, + "grad_norm": 0.46484375, + "learning_rate": 8.090058479532164e-05, + "loss": 3.1354, + "step": 12083 + }, + { + "epoch": 0.636, + "grad_norm": 0.484375, + "learning_rate": 8.088888888888889e-05, + "loss": 3.2424, + "step": 12084 + }, + { + "epoch": 0.6360526315789473, + "grad_norm": 0.51171875, + "learning_rate": 8.087719298245614e-05, + "loss": 3.2422, + "step": 12085 + }, + { + "epoch": 0.6361052631578947, + "grad_norm": 0.64453125, + "learning_rate": 8.08654970760234e-05, + "loss": 3.2094, + "step": 12086 + }, + { + "epoch": 0.6361578947368421, + "grad_norm": 0.51171875, + "learning_rate": 8.085380116959065e-05, + "loss": 3.3924, + "step": 12087 + }, + { + "epoch": 0.6362105263157894, + "grad_norm": 0.490234375, + "learning_rate": 8.08421052631579e-05, + "loss": 2.9345, + "step": 12088 + }, + { + "epoch": 0.6362631578947369, + "grad_norm": 0.45703125, + "learning_rate": 8.083040935672515e-05, + "loss": 3.4961, + "step": 12089 + }, + { + "epoch": 0.6363157894736842, + "grad_norm": 0.478515625, + "learning_rate": 8.08187134502924e-05, + "loss": 3.4452, + "step": 12090 + }, + { + "epoch": 0.6363684210526316, + "grad_norm": 0.494140625, + "learning_rate": 8.080701754385965e-05, + "loss": 3.2973, + "step": 12091 + }, + { + "epoch": 0.636421052631579, + "grad_norm": 0.5078125, + "learning_rate": 8.07953216374269e-05, + "loss": 3.5132, + "step": 12092 + }, + { + "epoch": 0.6364736842105263, + "grad_norm": 0.515625, + "learning_rate": 8.078362573099416e-05, + "loss": 3.4413, + "step": 12093 + }, + { + "epoch": 0.6365263157894737, + "grad_norm": 0.5, + "learning_rate": 8.077192982456141e-05, + "loss": 3.3625, + "step": 12094 + }, + { + "epoch": 0.636578947368421, + "grad_norm": 0.498046875, + "learning_rate": 8.076023391812866e-05, + "loss": 3.1061, + "step": 12095 + }, + { + "epoch": 0.6366315789473684, + "grad_norm": 0.51171875, + "learning_rate": 8.074853801169591e-05, + "loss": 3.3086, + "step": 12096 + }, + { + "epoch": 0.6366842105263157, + "grad_norm": 0.498046875, + "learning_rate": 8.073684210526316e-05, + "loss": 3.6376, + "step": 12097 + }, + { + "epoch": 0.6367368421052632, + "grad_norm": 0.53125, + "learning_rate": 8.072514619883041e-05, + "loss": 3.4031, + "step": 12098 + }, + { + "epoch": 0.6367894736842106, + "grad_norm": 0.498046875, + "learning_rate": 8.071345029239766e-05, + "loss": 3.3119, + "step": 12099 + }, + { + "epoch": 0.6368421052631579, + "grad_norm": 0.53125, + "learning_rate": 8.070175438596491e-05, + "loss": 3.353, + "step": 12100 + }, + { + "epoch": 0.6368947368421053, + "grad_norm": 0.54296875, + "learning_rate": 8.069005847953217e-05, + "loss": 3.5302, + "step": 12101 + }, + { + "epoch": 0.6369473684210526, + "grad_norm": 0.48828125, + "learning_rate": 8.067836257309942e-05, + "loss": 3.4142, + "step": 12102 + }, + { + "epoch": 0.637, + "grad_norm": 0.515625, + "learning_rate": 8.066666666666667e-05, + "loss": 3.4498, + "step": 12103 + }, + { + "epoch": 0.6370526315789473, + "grad_norm": 0.51953125, + "learning_rate": 8.065497076023392e-05, + "loss": 3.3853, + "step": 12104 + }, + { + "epoch": 0.6371052631578947, + "grad_norm": 0.4609375, + "learning_rate": 8.064327485380117e-05, + "loss": 3.6304, + "step": 12105 + }, + { + "epoch": 0.6371578947368421, + "grad_norm": 0.51171875, + "learning_rate": 8.063157894736842e-05, + "loss": 3.1031, + "step": 12106 + }, + { + "epoch": 0.6372105263157894, + "grad_norm": 0.4609375, + "learning_rate": 8.061988304093567e-05, + "loss": 3.7363, + "step": 12107 + }, + { + "epoch": 0.6372631578947369, + "grad_norm": 0.53515625, + "learning_rate": 8.060818713450293e-05, + "loss": 3.0927, + "step": 12108 + }, + { + "epoch": 0.6373157894736842, + "grad_norm": 0.498046875, + "learning_rate": 8.059649122807018e-05, + "loss": 3.5097, + "step": 12109 + }, + { + "epoch": 0.6373684210526316, + "grad_norm": 0.47265625, + "learning_rate": 8.058479532163743e-05, + "loss": 3.3806, + "step": 12110 + }, + { + "epoch": 0.637421052631579, + "grad_norm": 0.5390625, + "learning_rate": 8.057309941520468e-05, + "loss": 3.1311, + "step": 12111 + }, + { + "epoch": 0.6374736842105263, + "grad_norm": 0.484375, + "learning_rate": 8.056140350877193e-05, + "loss": 3.2166, + "step": 12112 + }, + { + "epoch": 0.6375263157894737, + "grad_norm": 0.48828125, + "learning_rate": 8.054970760233918e-05, + "loss": 3.5624, + "step": 12113 + }, + { + "epoch": 0.637578947368421, + "grad_norm": 0.486328125, + "learning_rate": 8.053801169590643e-05, + "loss": 3.3847, + "step": 12114 + }, + { + "epoch": 0.6376315789473684, + "grad_norm": 0.45703125, + "learning_rate": 8.052631578947368e-05, + "loss": 3.6037, + "step": 12115 + }, + { + "epoch": 0.6376842105263157, + "grad_norm": 0.46484375, + "learning_rate": 8.051461988304094e-05, + "loss": 3.7228, + "step": 12116 + }, + { + "epoch": 0.6377368421052632, + "grad_norm": 0.482421875, + "learning_rate": 8.050292397660819e-05, + "loss": 3.8266, + "step": 12117 + }, + { + "epoch": 0.6377894736842106, + "grad_norm": 0.51953125, + "learning_rate": 8.049122807017544e-05, + "loss": 3.4977, + "step": 12118 + }, + { + "epoch": 0.6378421052631579, + "grad_norm": 0.5078125, + "learning_rate": 8.04795321637427e-05, + "loss": 3.2487, + "step": 12119 + }, + { + "epoch": 0.6378947368421053, + "grad_norm": 0.53515625, + "learning_rate": 8.046783625730994e-05, + "loss": 3.4688, + "step": 12120 + }, + { + "epoch": 0.6379473684210526, + "grad_norm": 0.609375, + "learning_rate": 8.045614035087719e-05, + "loss": 2.9926, + "step": 12121 + }, + { + "epoch": 0.638, + "grad_norm": 0.4765625, + "learning_rate": 8.044444444444444e-05, + "loss": 3.4679, + "step": 12122 + }, + { + "epoch": 0.6380526315789473, + "grad_norm": 0.484375, + "learning_rate": 8.04327485380117e-05, + "loss": 3.5428, + "step": 12123 + }, + { + "epoch": 0.6381052631578947, + "grad_norm": 0.58203125, + "learning_rate": 8.042105263157895e-05, + "loss": 3.4, + "step": 12124 + }, + { + "epoch": 0.6381578947368421, + "grad_norm": 0.46875, + "learning_rate": 8.04093567251462e-05, + "loss": 3.204, + "step": 12125 + }, + { + "epoch": 0.6382105263157895, + "grad_norm": 0.49609375, + "learning_rate": 8.039766081871345e-05, + "loss": 3.4285, + "step": 12126 + }, + { + "epoch": 0.6382631578947369, + "grad_norm": 0.478515625, + "learning_rate": 8.038596491228072e-05, + "loss": 3.7685, + "step": 12127 + }, + { + "epoch": 0.6383157894736842, + "grad_norm": 0.51171875, + "learning_rate": 8.037426900584795e-05, + "loss": 3.265, + "step": 12128 + }, + { + "epoch": 0.6383684210526316, + "grad_norm": 0.466796875, + "learning_rate": 8.03625730994152e-05, + "loss": 3.357, + "step": 12129 + }, + { + "epoch": 0.638421052631579, + "grad_norm": 0.4765625, + "learning_rate": 8.035087719298246e-05, + "loss": 3.3817, + "step": 12130 + }, + { + "epoch": 0.6384736842105263, + "grad_norm": 0.494140625, + "learning_rate": 8.033918128654971e-05, + "loss": 3.0921, + "step": 12131 + }, + { + "epoch": 0.6385263157894737, + "grad_norm": 0.51171875, + "learning_rate": 8.032748538011696e-05, + "loss": 3.2114, + "step": 12132 + }, + { + "epoch": 0.638578947368421, + "grad_norm": 0.5, + "learning_rate": 8.031578947368421e-05, + "loss": 3.5087, + "step": 12133 + }, + { + "epoch": 0.6386315789473684, + "grad_norm": 0.52734375, + "learning_rate": 8.030409356725148e-05, + "loss": 3.3241, + "step": 12134 + }, + { + "epoch": 0.6386842105263157, + "grad_norm": 0.455078125, + "learning_rate": 8.029239766081871e-05, + "loss": 3.4513, + "step": 12135 + }, + { + "epoch": 0.6387368421052632, + "grad_norm": 0.50390625, + "learning_rate": 8.028070175438596e-05, + "loss": 3.6063, + "step": 12136 + }, + { + "epoch": 0.6387894736842106, + "grad_norm": 0.4921875, + "learning_rate": 8.026900584795321e-05, + "loss": 3.3014, + "step": 12137 + }, + { + "epoch": 0.6388421052631579, + "grad_norm": 0.4765625, + "learning_rate": 8.025730994152047e-05, + "loss": 3.5371, + "step": 12138 + }, + { + "epoch": 0.6388947368421053, + "grad_norm": 0.54296875, + "learning_rate": 8.024561403508772e-05, + "loss": 3.1502, + "step": 12139 + }, + { + "epoch": 0.6389473684210526, + "grad_norm": 0.5, + "learning_rate": 8.023391812865497e-05, + "loss": 3.7783, + "step": 12140 + }, + { + "epoch": 0.639, + "grad_norm": 0.46484375, + "learning_rate": 8.022222222222222e-05, + "loss": 3.2947, + "step": 12141 + }, + { + "epoch": 0.6390526315789473, + "grad_norm": 0.4921875, + "learning_rate": 8.021052631578949e-05, + "loss": 3.8822, + "step": 12142 + }, + { + "epoch": 0.6391052631578947, + "grad_norm": 0.46875, + "learning_rate": 8.019883040935672e-05, + "loss": 3.1138, + "step": 12143 + }, + { + "epoch": 0.6391578947368421, + "grad_norm": 0.671875, + "learning_rate": 8.018713450292397e-05, + "loss": 3.3778, + "step": 12144 + }, + { + "epoch": 0.6392105263157895, + "grad_norm": 0.490234375, + "learning_rate": 8.017543859649124e-05, + "loss": 3.4839, + "step": 12145 + }, + { + "epoch": 0.6392631578947369, + "grad_norm": 0.5, + "learning_rate": 8.016374269005849e-05, + "loss": 3.4313, + "step": 12146 + }, + { + "epoch": 0.6393157894736842, + "grad_norm": 0.494140625, + "learning_rate": 8.015204678362573e-05, + "loss": 3.6745, + "step": 12147 + }, + { + "epoch": 0.6393684210526316, + "grad_norm": 0.470703125, + "learning_rate": 8.014035087719298e-05, + "loss": 3.0508, + "step": 12148 + }, + { + "epoch": 0.639421052631579, + "grad_norm": 0.462890625, + "learning_rate": 8.012865497076025e-05, + "loss": 3.0718, + "step": 12149 + }, + { + "epoch": 0.6394736842105263, + "grad_norm": 0.578125, + "learning_rate": 8.01169590643275e-05, + "loss": 3.3516, + "step": 12150 + }, + { + "epoch": 0.6395263157894737, + "grad_norm": 0.47265625, + "learning_rate": 8.010526315789473e-05, + "loss": 3.3821, + "step": 12151 + }, + { + "epoch": 0.639578947368421, + "grad_norm": 0.474609375, + "learning_rate": 8.009356725146198e-05, + "loss": 3.2918, + "step": 12152 + }, + { + "epoch": 0.6396315789473684, + "grad_norm": 0.478515625, + "learning_rate": 8.008187134502925e-05, + "loss": 3.2283, + "step": 12153 + }, + { + "epoch": 0.6396842105263157, + "grad_norm": 0.498046875, + "learning_rate": 8.00701754385965e-05, + "loss": 3.3964, + "step": 12154 + }, + { + "epoch": 0.6397368421052632, + "grad_norm": 0.498046875, + "learning_rate": 8.005847953216375e-05, + "loss": 3.5364, + "step": 12155 + }, + { + "epoch": 0.6397894736842106, + "grad_norm": 0.474609375, + "learning_rate": 8.004678362573101e-05, + "loss": 3.1858, + "step": 12156 + }, + { + "epoch": 0.6398421052631579, + "grad_norm": 0.51171875, + "learning_rate": 8.003508771929826e-05, + "loss": 3.4506, + "step": 12157 + }, + { + "epoch": 0.6398947368421053, + "grad_norm": 0.4921875, + "learning_rate": 8.00233918128655e-05, + "loss": 3.3663, + "step": 12158 + }, + { + "epoch": 0.6399473684210526, + "grad_norm": 0.462890625, + "learning_rate": 8.001169590643274e-05, + "loss": 3.3062, + "step": 12159 + }, + { + "epoch": 0.64, + "grad_norm": 0.5, + "learning_rate": 8e-05, + "loss": 3.4121, + "step": 12160 + }, + { + "epoch": 0.6400526315789473, + "grad_norm": 0.62890625, + "learning_rate": 7.998830409356726e-05, + "loss": 3.2775, + "step": 12161 + }, + { + "epoch": 0.6401052631578947, + "grad_norm": 0.53125, + "learning_rate": 7.99766081871345e-05, + "loss": 3.6376, + "step": 12162 + }, + { + "epoch": 0.6401578947368421, + "grad_norm": 0.48828125, + "learning_rate": 7.996491228070176e-05, + "loss": 3.5081, + "step": 12163 + }, + { + "epoch": 0.6402105263157895, + "grad_norm": 0.5078125, + "learning_rate": 7.995321637426902e-05, + "loss": 3.3149, + "step": 12164 + }, + { + "epoch": 0.6402631578947369, + "grad_norm": 0.51953125, + "learning_rate": 7.994152046783627e-05, + "loss": 3.7449, + "step": 12165 + }, + { + "epoch": 0.6403157894736842, + "grad_norm": 0.53515625, + "learning_rate": 7.99298245614035e-05, + "loss": 3.4279, + "step": 12166 + }, + { + "epoch": 0.6403684210526316, + "grad_norm": 0.5234375, + "learning_rate": 7.991812865497075e-05, + "loss": 3.359, + "step": 12167 + }, + { + "epoch": 0.640421052631579, + "grad_norm": 0.486328125, + "learning_rate": 7.990643274853802e-05, + "loss": 3.4649, + "step": 12168 + }, + { + "epoch": 0.6404736842105263, + "grad_norm": 0.458984375, + "learning_rate": 7.989473684210527e-05, + "loss": 3.1725, + "step": 12169 + }, + { + "epoch": 0.6405263157894737, + "grad_norm": 0.51171875, + "learning_rate": 7.988304093567252e-05, + "loss": 3.543, + "step": 12170 + }, + { + "epoch": 0.640578947368421, + "grad_norm": 0.478515625, + "learning_rate": 7.987134502923978e-05, + "loss": 3.0598, + "step": 12171 + }, + { + "epoch": 0.6406315789473684, + "grad_norm": 0.51171875, + "learning_rate": 7.985964912280703e-05, + "loss": 3.3548, + "step": 12172 + }, + { + "epoch": 0.6406842105263157, + "grad_norm": 0.5, + "learning_rate": 7.984795321637428e-05, + "loss": 3.5318, + "step": 12173 + }, + { + "epoch": 0.6407368421052632, + "grad_norm": 0.494140625, + "learning_rate": 7.983625730994151e-05, + "loss": 3.2328, + "step": 12174 + }, + { + "epoch": 0.6407894736842106, + "grad_norm": 0.5078125, + "learning_rate": 7.982456140350878e-05, + "loss": 3.1694, + "step": 12175 + }, + { + "epoch": 0.6408421052631579, + "grad_norm": 0.76171875, + "learning_rate": 7.981286549707603e-05, + "loss": 3.3925, + "step": 12176 + }, + { + "epoch": 0.6408947368421053, + "grad_norm": 0.466796875, + "learning_rate": 7.980116959064328e-05, + "loss": 3.4393, + "step": 12177 + }, + { + "epoch": 0.6409473684210526, + "grad_norm": 0.494140625, + "learning_rate": 7.978947368421053e-05, + "loss": 3.3074, + "step": 12178 + }, + { + "epoch": 0.641, + "grad_norm": 0.5, + "learning_rate": 7.977777777777779e-05, + "loss": 3.1428, + "step": 12179 + }, + { + "epoch": 0.6410526315789473, + "grad_norm": 0.494140625, + "learning_rate": 7.976608187134504e-05, + "loss": 3.5187, + "step": 12180 + }, + { + "epoch": 0.6411052631578947, + "grad_norm": 0.5078125, + "learning_rate": 7.975438596491228e-05, + "loss": 3.3824, + "step": 12181 + }, + { + "epoch": 0.6411578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.974269005847954e-05, + "loss": 3.2791, + "step": 12182 + }, + { + "epoch": 0.6412105263157895, + "grad_norm": 0.470703125, + "learning_rate": 7.973099415204679e-05, + "loss": 3.5191, + "step": 12183 + }, + { + "epoch": 0.6412631578947369, + "grad_norm": 0.474609375, + "learning_rate": 7.971929824561404e-05, + "loss": 3.5265, + "step": 12184 + }, + { + "epoch": 0.6413157894736842, + "grad_norm": 0.51171875, + "learning_rate": 7.970760233918129e-05, + "loss": 3.3851, + "step": 12185 + }, + { + "epoch": 0.6413684210526316, + "grad_norm": 0.490234375, + "learning_rate": 7.969590643274855e-05, + "loss": 3.5676, + "step": 12186 + }, + { + "epoch": 0.641421052631579, + "grad_norm": 0.68359375, + "learning_rate": 7.96842105263158e-05, + "loss": 3.4264, + "step": 12187 + }, + { + "epoch": 0.6414736842105263, + "grad_norm": 0.494140625, + "learning_rate": 7.967251461988305e-05, + "loss": 3.7406, + "step": 12188 + }, + { + "epoch": 0.6415263157894737, + "grad_norm": 0.52734375, + "learning_rate": 7.966081871345029e-05, + "loss": 3.3256, + "step": 12189 + }, + { + "epoch": 0.641578947368421, + "grad_norm": 0.5390625, + "learning_rate": 7.964912280701755e-05, + "loss": 3.3312, + "step": 12190 + }, + { + "epoch": 0.6416315789473684, + "grad_norm": 0.51953125, + "learning_rate": 7.96374269005848e-05, + "loss": 3.3414, + "step": 12191 + }, + { + "epoch": 0.6416842105263157, + "grad_norm": 0.490234375, + "learning_rate": 7.962573099415205e-05, + "loss": 3.3203, + "step": 12192 + }, + { + "epoch": 0.6417368421052632, + "grad_norm": 0.52734375, + "learning_rate": 7.961403508771931e-05, + "loss": 3.5168, + "step": 12193 + }, + { + "epoch": 0.6417894736842106, + "grad_norm": 0.53515625, + "learning_rate": 7.960233918128656e-05, + "loss": 3.1874, + "step": 12194 + }, + { + "epoch": 0.6418421052631579, + "grad_norm": 0.49609375, + "learning_rate": 7.959064327485381e-05, + "loss": 3.2852, + "step": 12195 + }, + { + "epoch": 0.6418947368421053, + "grad_norm": 0.76171875, + "learning_rate": 7.957894736842106e-05, + "loss": 3.0636, + "step": 12196 + }, + { + "epoch": 0.6419473684210526, + "grad_norm": 0.51171875, + "learning_rate": 7.956725146198831e-05, + "loss": 3.7104, + "step": 12197 + }, + { + "epoch": 0.642, + "grad_norm": 0.4765625, + "learning_rate": 7.955555555555556e-05, + "loss": 3.2747, + "step": 12198 + }, + { + "epoch": 0.6420526315789473, + "grad_norm": 0.50390625, + "learning_rate": 7.954385964912281e-05, + "loss": 3.3679, + "step": 12199 + }, + { + "epoch": 0.6421052631578947, + "grad_norm": 0.50390625, + "learning_rate": 7.953216374269006e-05, + "loss": 3.524, + "step": 12200 + }, + { + "epoch": 0.6421578947368421, + "grad_norm": 0.484375, + "learning_rate": 7.952046783625732e-05, + "loss": 3.3162, + "step": 12201 + }, + { + "epoch": 0.6422105263157895, + "grad_norm": 0.48046875, + "learning_rate": 7.950877192982457e-05, + "loss": 3.3949, + "step": 12202 + }, + { + "epoch": 0.6422631578947369, + "grad_norm": 0.462890625, + "learning_rate": 7.949707602339182e-05, + "loss": 3.3763, + "step": 12203 + }, + { + "epoch": 0.6423157894736842, + "grad_norm": 0.578125, + "learning_rate": 7.948538011695906e-05, + "loss": 3.1927, + "step": 12204 + }, + { + "epoch": 0.6423684210526316, + "grad_norm": 0.53515625, + "learning_rate": 7.947368421052632e-05, + "loss": 3.5148, + "step": 12205 + }, + { + "epoch": 0.6424210526315789, + "grad_norm": 0.4921875, + "learning_rate": 7.946198830409357e-05, + "loss": 3.0246, + "step": 12206 + }, + { + "epoch": 0.6424736842105263, + "grad_norm": 0.51953125, + "learning_rate": 7.945029239766082e-05, + "loss": 3.2057, + "step": 12207 + }, + { + "epoch": 0.6425263157894737, + "grad_norm": 0.48828125, + "learning_rate": 7.943859649122808e-05, + "loss": 3.6604, + "step": 12208 + }, + { + "epoch": 0.642578947368421, + "grad_norm": 0.51953125, + "learning_rate": 7.942690058479533e-05, + "loss": 3.2005, + "step": 12209 + }, + { + "epoch": 0.6426315789473684, + "grad_norm": 0.466796875, + "learning_rate": 7.941520467836258e-05, + "loss": 3.4227, + "step": 12210 + }, + { + "epoch": 0.6426842105263157, + "grad_norm": 0.478515625, + "learning_rate": 7.940350877192983e-05, + "loss": 3.4078, + "step": 12211 + }, + { + "epoch": 0.6427368421052632, + "grad_norm": 0.478515625, + "learning_rate": 7.939181286549708e-05, + "loss": 3.4133, + "step": 12212 + }, + { + "epoch": 0.6427894736842106, + "grad_norm": 0.484375, + "learning_rate": 7.938011695906433e-05, + "loss": 3.6289, + "step": 12213 + }, + { + "epoch": 0.6428421052631579, + "grad_norm": 0.48828125, + "learning_rate": 7.936842105263158e-05, + "loss": 3.096, + "step": 12214 + }, + { + "epoch": 0.6428947368421053, + "grad_norm": 0.458984375, + "learning_rate": 7.935672514619883e-05, + "loss": 3.3913, + "step": 12215 + }, + { + "epoch": 0.6429473684210526, + "grad_norm": 0.5078125, + "learning_rate": 7.934502923976609e-05, + "loss": 3.416, + "step": 12216 + }, + { + "epoch": 0.643, + "grad_norm": 0.48828125, + "learning_rate": 7.933333333333334e-05, + "loss": 3.2731, + "step": 12217 + }, + { + "epoch": 0.6430526315789473, + "grad_norm": 0.490234375, + "learning_rate": 7.932163742690059e-05, + "loss": 3.6206, + "step": 12218 + }, + { + "epoch": 0.6431052631578947, + "grad_norm": 0.51171875, + "learning_rate": 7.930994152046784e-05, + "loss": 3.3998, + "step": 12219 + }, + { + "epoch": 0.6431578947368422, + "grad_norm": 0.5, + "learning_rate": 7.929824561403509e-05, + "loss": 2.7503, + "step": 12220 + }, + { + "epoch": 0.6432105263157895, + "grad_norm": 0.5, + "learning_rate": 7.928654970760234e-05, + "loss": 3.6516, + "step": 12221 + }, + { + "epoch": 0.6432631578947369, + "grad_norm": 0.50390625, + "learning_rate": 7.927485380116959e-05, + "loss": 3.2566, + "step": 12222 + }, + { + "epoch": 0.6433157894736842, + "grad_norm": 0.51171875, + "learning_rate": 7.926315789473685e-05, + "loss": 3.1781, + "step": 12223 + }, + { + "epoch": 0.6433684210526316, + "grad_norm": 0.474609375, + "learning_rate": 7.92514619883041e-05, + "loss": 3.7174, + "step": 12224 + }, + { + "epoch": 0.6434210526315789, + "grad_norm": 0.46484375, + "learning_rate": 7.923976608187135e-05, + "loss": 3.19, + "step": 12225 + }, + { + "epoch": 0.6434736842105263, + "grad_norm": 0.54296875, + "learning_rate": 7.92280701754386e-05, + "loss": 3.3074, + "step": 12226 + }, + { + "epoch": 0.6435263157894737, + "grad_norm": 0.53515625, + "learning_rate": 7.921637426900585e-05, + "loss": 3.1816, + "step": 12227 + }, + { + "epoch": 0.643578947368421, + "grad_norm": 0.462890625, + "learning_rate": 7.92046783625731e-05, + "loss": 3.3453, + "step": 12228 + }, + { + "epoch": 0.6436315789473684, + "grad_norm": 0.48828125, + "learning_rate": 7.919298245614035e-05, + "loss": 3.3994, + "step": 12229 + }, + { + "epoch": 0.6436842105263157, + "grad_norm": 0.484375, + "learning_rate": 7.91812865497076e-05, + "loss": 3.2358, + "step": 12230 + }, + { + "epoch": 0.6437368421052632, + "grad_norm": 0.5, + "learning_rate": 7.916959064327486e-05, + "loss": 3.4133, + "step": 12231 + }, + { + "epoch": 0.6437894736842106, + "grad_norm": 0.47265625, + "learning_rate": 7.915789473684211e-05, + "loss": 2.9275, + "step": 12232 + }, + { + "epoch": 0.6438421052631579, + "grad_norm": 0.48046875, + "learning_rate": 7.914619883040936e-05, + "loss": 3.5181, + "step": 12233 + }, + { + "epoch": 0.6438947368421053, + "grad_norm": 0.53515625, + "learning_rate": 7.913450292397661e-05, + "loss": 3.8561, + "step": 12234 + }, + { + "epoch": 0.6439473684210526, + "grad_norm": 0.5390625, + "learning_rate": 7.912280701754386e-05, + "loss": 3.3261, + "step": 12235 + }, + { + "epoch": 0.644, + "grad_norm": 0.48046875, + "learning_rate": 7.911111111111111e-05, + "loss": 2.9031, + "step": 12236 + }, + { + "epoch": 0.6440526315789473, + "grad_norm": 0.490234375, + "learning_rate": 7.909941520467836e-05, + "loss": 2.9552, + "step": 12237 + }, + { + "epoch": 0.6441052631578947, + "grad_norm": 0.474609375, + "learning_rate": 7.908771929824562e-05, + "loss": 3.5743, + "step": 12238 + }, + { + "epoch": 0.6441578947368422, + "grad_norm": 0.5625, + "learning_rate": 7.907602339181287e-05, + "loss": 3.7201, + "step": 12239 + }, + { + "epoch": 0.6442105263157895, + "grad_norm": 0.5078125, + "learning_rate": 7.906432748538012e-05, + "loss": 3.68, + "step": 12240 + }, + { + "epoch": 0.6442631578947369, + "grad_norm": 0.51953125, + "learning_rate": 7.905263157894737e-05, + "loss": 3.5863, + "step": 12241 + }, + { + "epoch": 0.6443157894736842, + "grad_norm": 0.494140625, + "learning_rate": 7.904093567251462e-05, + "loss": 3.1196, + "step": 12242 + }, + { + "epoch": 0.6443684210526316, + "grad_norm": 0.48828125, + "learning_rate": 7.902923976608187e-05, + "loss": 3.1961, + "step": 12243 + }, + { + "epoch": 0.6444210526315789, + "grad_norm": 0.50390625, + "learning_rate": 7.901754385964912e-05, + "loss": 3.1268, + "step": 12244 + }, + { + "epoch": 0.6444736842105263, + "grad_norm": 0.455078125, + "learning_rate": 7.900584795321638e-05, + "loss": 2.7757, + "step": 12245 + }, + { + "epoch": 0.6445263157894737, + "grad_norm": 0.498046875, + "learning_rate": 7.899415204678363e-05, + "loss": 3.2291, + "step": 12246 + }, + { + "epoch": 0.644578947368421, + "grad_norm": 0.466796875, + "learning_rate": 7.898245614035088e-05, + "loss": 3.3023, + "step": 12247 + }, + { + "epoch": 0.6446315789473684, + "grad_norm": 0.478515625, + "learning_rate": 7.897076023391813e-05, + "loss": 3.0379, + "step": 12248 + }, + { + "epoch": 0.6446842105263157, + "grad_norm": 0.5390625, + "learning_rate": 7.895906432748538e-05, + "loss": 3.1068, + "step": 12249 + }, + { + "epoch": 0.6447368421052632, + "grad_norm": 0.451171875, + "learning_rate": 7.894736842105263e-05, + "loss": 3.5083, + "step": 12250 + }, + { + "epoch": 0.6447894736842106, + "grad_norm": 0.490234375, + "learning_rate": 7.893567251461988e-05, + "loss": 2.9443, + "step": 12251 + }, + { + "epoch": 0.6448421052631579, + "grad_norm": 0.484375, + "learning_rate": 7.892397660818713e-05, + "loss": 3.8917, + "step": 12252 + }, + { + "epoch": 0.6448947368421053, + "grad_norm": 0.51953125, + "learning_rate": 7.89122807017544e-05, + "loss": 3.017, + "step": 12253 + }, + { + "epoch": 0.6449473684210526, + "grad_norm": 0.458984375, + "learning_rate": 7.890058479532164e-05, + "loss": 3.3794, + "step": 12254 + }, + { + "epoch": 0.645, + "grad_norm": 0.486328125, + "learning_rate": 7.88888888888889e-05, + "loss": 3.7385, + "step": 12255 + }, + { + "epoch": 0.6450526315789473, + "grad_norm": 0.498046875, + "learning_rate": 7.887719298245614e-05, + "loss": 3.6457, + "step": 12256 + }, + { + "epoch": 0.6451052631578947, + "grad_norm": 0.484375, + "learning_rate": 7.886549707602339e-05, + "loss": 3.3364, + "step": 12257 + }, + { + "epoch": 0.6451578947368422, + "grad_norm": 0.53515625, + "learning_rate": 7.885380116959064e-05, + "loss": 2.9533, + "step": 12258 + }, + { + "epoch": 0.6452105263157895, + "grad_norm": 0.490234375, + "learning_rate": 7.884210526315789e-05, + "loss": 3.4188, + "step": 12259 + }, + { + "epoch": 0.6452631578947369, + "grad_norm": 0.494140625, + "learning_rate": 7.883040935672516e-05, + "loss": 3.1417, + "step": 12260 + }, + { + "epoch": 0.6453157894736842, + "grad_norm": 0.4921875, + "learning_rate": 7.88187134502924e-05, + "loss": 3.2311, + "step": 12261 + }, + { + "epoch": 0.6453684210526316, + "grad_norm": 0.455078125, + "learning_rate": 7.880701754385965e-05, + "loss": 3.3961, + "step": 12262 + }, + { + "epoch": 0.6454210526315789, + "grad_norm": 0.5, + "learning_rate": 7.87953216374269e-05, + "loss": 3.284, + "step": 12263 + }, + { + "epoch": 0.6454736842105263, + "grad_norm": 0.4921875, + "learning_rate": 7.878362573099415e-05, + "loss": 3.0593, + "step": 12264 + }, + { + "epoch": 0.6455263157894737, + "grad_norm": 0.46875, + "learning_rate": 7.87719298245614e-05, + "loss": 3.5332, + "step": 12265 + }, + { + "epoch": 0.645578947368421, + "grad_norm": 0.48046875, + "learning_rate": 7.876023391812865e-05, + "loss": 3.3081, + "step": 12266 + }, + { + "epoch": 0.6456315789473684, + "grad_norm": 0.5, + "learning_rate": 7.87485380116959e-05, + "loss": 3.6815, + "step": 12267 + }, + { + "epoch": 0.6456842105263157, + "grad_norm": 0.51171875, + "learning_rate": 7.873684210526317e-05, + "loss": 3.1644, + "step": 12268 + }, + { + "epoch": 0.6457368421052632, + "grad_norm": 0.53125, + "learning_rate": 7.872514619883042e-05, + "loss": 3.8716, + "step": 12269 + }, + { + "epoch": 0.6457894736842106, + "grad_norm": 0.5078125, + "learning_rate": 7.871345029239767e-05, + "loss": 3.0828, + "step": 12270 + }, + { + "epoch": 0.6458421052631579, + "grad_norm": 0.48046875, + "learning_rate": 7.870175438596491e-05, + "loss": 3.2777, + "step": 12271 + }, + { + "epoch": 0.6458947368421053, + "grad_norm": 0.50390625, + "learning_rate": 7.869005847953216e-05, + "loss": 3.5321, + "step": 12272 + }, + { + "epoch": 0.6459473684210526, + "grad_norm": 0.546875, + "learning_rate": 7.867836257309941e-05, + "loss": 3.4524, + "step": 12273 + }, + { + "epoch": 0.646, + "grad_norm": 0.57421875, + "learning_rate": 7.866666666666666e-05, + "loss": 3.1753, + "step": 12274 + }, + { + "epoch": 0.6460526315789473, + "grad_norm": 0.486328125, + "learning_rate": 7.865497076023393e-05, + "loss": 3.6373, + "step": 12275 + }, + { + "epoch": 0.6461052631578947, + "grad_norm": 0.5234375, + "learning_rate": 7.864327485380118e-05, + "loss": 3.4081, + "step": 12276 + }, + { + "epoch": 0.6461578947368422, + "grad_norm": 0.46484375, + "learning_rate": 7.863157894736843e-05, + "loss": 3.0454, + "step": 12277 + }, + { + "epoch": 0.6462105263157895, + "grad_norm": 0.5078125, + "learning_rate": 7.861988304093568e-05, + "loss": 3.3359, + "step": 12278 + }, + { + "epoch": 0.6462631578947369, + "grad_norm": 0.458984375, + "learning_rate": 7.860818713450292e-05, + "loss": 3.1805, + "step": 12279 + }, + { + "epoch": 0.6463157894736842, + "grad_norm": 0.5546875, + "learning_rate": 7.859649122807017e-05, + "loss": 3.5782, + "step": 12280 + }, + { + "epoch": 0.6463684210526316, + "grad_norm": 0.5, + "learning_rate": 7.858479532163742e-05, + "loss": 3.2907, + "step": 12281 + }, + { + "epoch": 0.6464210526315789, + "grad_norm": 0.51171875, + "learning_rate": 7.857309941520469e-05, + "loss": 3.2106, + "step": 12282 + }, + { + "epoch": 0.6464736842105263, + "grad_norm": 0.5546875, + "learning_rate": 7.856140350877194e-05, + "loss": 3.2323, + "step": 12283 + }, + { + "epoch": 0.6465263157894737, + "grad_norm": 0.4921875, + "learning_rate": 7.854970760233919e-05, + "loss": 3.5352, + "step": 12284 + }, + { + "epoch": 0.646578947368421, + "grad_norm": 0.4765625, + "learning_rate": 7.853801169590644e-05, + "loss": 3.2445, + "step": 12285 + }, + { + "epoch": 0.6466315789473684, + "grad_norm": 0.47265625, + "learning_rate": 7.852631578947369e-05, + "loss": 3.3329, + "step": 12286 + }, + { + "epoch": 0.6466842105263158, + "grad_norm": 0.4921875, + "learning_rate": 7.851461988304094e-05, + "loss": 3.5256, + "step": 12287 + }, + { + "epoch": 0.6467368421052632, + "grad_norm": 0.478515625, + "learning_rate": 7.850292397660818e-05, + "loss": 3.373, + "step": 12288 + }, + { + "epoch": 0.6467894736842106, + "grad_norm": 0.55078125, + "learning_rate": 7.849122807017543e-05, + "loss": 3.2387, + "step": 12289 + }, + { + "epoch": 0.6468421052631579, + "grad_norm": 0.470703125, + "learning_rate": 7.84795321637427e-05, + "loss": 3.7126, + "step": 12290 + }, + { + "epoch": 0.6468947368421053, + "grad_norm": 0.51171875, + "learning_rate": 7.846783625730995e-05, + "loss": 3.7124, + "step": 12291 + }, + { + "epoch": 0.6469473684210526, + "grad_norm": 0.515625, + "learning_rate": 7.84561403508772e-05, + "loss": 3.3622, + "step": 12292 + }, + { + "epoch": 0.647, + "grad_norm": 0.478515625, + "learning_rate": 7.844444444444446e-05, + "loss": 3.5335, + "step": 12293 + }, + { + "epoch": 0.6470526315789473, + "grad_norm": 0.466796875, + "learning_rate": 7.84327485380117e-05, + "loss": 3.5457, + "step": 12294 + }, + { + "epoch": 0.6471052631578947, + "grad_norm": 0.54296875, + "learning_rate": 7.842105263157895e-05, + "loss": 3.2049, + "step": 12295 + }, + { + "epoch": 0.6471578947368422, + "grad_norm": 0.46875, + "learning_rate": 7.84093567251462e-05, + "loss": 3.4121, + "step": 12296 + }, + { + "epoch": 0.6472105263157895, + "grad_norm": 0.49609375, + "learning_rate": 7.839766081871346e-05, + "loss": 3.5271, + "step": 12297 + }, + { + "epoch": 0.6472631578947369, + "grad_norm": 0.51953125, + "learning_rate": 7.838596491228071e-05, + "loss": 3.4252, + "step": 12298 + }, + { + "epoch": 0.6473157894736842, + "grad_norm": 0.4921875, + "learning_rate": 7.837426900584796e-05, + "loss": 3.5713, + "step": 12299 + }, + { + "epoch": 0.6473684210526316, + "grad_norm": 0.515625, + "learning_rate": 7.836257309941521e-05, + "loss": 3.1528, + "step": 12300 + }, + { + "epoch": 0.6474210526315789, + "grad_norm": 0.470703125, + "learning_rate": 7.835087719298247e-05, + "loss": 2.5696, + "step": 12301 + }, + { + "epoch": 0.6474736842105263, + "grad_norm": 0.5234375, + "learning_rate": 7.83391812865497e-05, + "loss": 3.6029, + "step": 12302 + }, + { + "epoch": 0.6475263157894737, + "grad_norm": 0.470703125, + "learning_rate": 7.832748538011696e-05, + "loss": 3.2568, + "step": 12303 + }, + { + "epoch": 0.647578947368421, + "grad_norm": 0.5625, + "learning_rate": 7.83157894736842e-05, + "loss": 3.6374, + "step": 12304 + }, + { + "epoch": 0.6476315789473684, + "grad_norm": 0.515625, + "learning_rate": 7.830409356725147e-05, + "loss": 3.3485, + "step": 12305 + }, + { + "epoch": 0.6476842105263158, + "grad_norm": 0.51953125, + "learning_rate": 7.829239766081872e-05, + "loss": 3.2349, + "step": 12306 + }, + { + "epoch": 0.6477368421052632, + "grad_norm": 0.46875, + "learning_rate": 7.828070175438597e-05, + "loss": 3.0923, + "step": 12307 + }, + { + "epoch": 0.6477894736842106, + "grad_norm": 0.484375, + "learning_rate": 7.826900584795323e-05, + "loss": 3.2747, + "step": 12308 + }, + { + "epoch": 0.6478421052631579, + "grad_norm": 0.484375, + "learning_rate": 7.825730994152047e-05, + "loss": 2.9715, + "step": 12309 + }, + { + "epoch": 0.6478947368421053, + "grad_norm": 0.63671875, + "learning_rate": 7.824561403508772e-05, + "loss": 3.4213, + "step": 12310 + }, + { + "epoch": 0.6479473684210526, + "grad_norm": 0.44921875, + "learning_rate": 7.823391812865497e-05, + "loss": 3.8315, + "step": 12311 + }, + { + "epoch": 0.648, + "grad_norm": 0.50390625, + "learning_rate": 7.822222222222223e-05, + "loss": 3.1097, + "step": 12312 + }, + { + "epoch": 0.6480526315789473, + "grad_norm": 0.52734375, + "learning_rate": 7.821052631578948e-05, + "loss": 3.4896, + "step": 12313 + }, + { + "epoch": 0.6481052631578947, + "grad_norm": 0.470703125, + "learning_rate": 7.819883040935673e-05, + "loss": 3.34, + "step": 12314 + }, + { + "epoch": 0.6481578947368422, + "grad_norm": 0.515625, + "learning_rate": 7.818713450292398e-05, + "loss": 3.0941, + "step": 12315 + }, + { + "epoch": 0.6482105263157895, + "grad_norm": 0.46484375, + "learning_rate": 7.817543859649124e-05, + "loss": 3.3134, + "step": 12316 + }, + { + "epoch": 0.6482631578947369, + "grad_norm": 0.50390625, + "learning_rate": 7.816374269005848e-05, + "loss": 3.6084, + "step": 12317 + }, + { + "epoch": 0.6483157894736842, + "grad_norm": 0.5703125, + "learning_rate": 7.815204678362573e-05, + "loss": 3.5116, + "step": 12318 + }, + { + "epoch": 0.6483684210526316, + "grad_norm": 0.49609375, + "learning_rate": 7.814035087719299e-05, + "loss": 3.4091, + "step": 12319 + }, + { + "epoch": 0.6484210526315789, + "grad_norm": 0.484375, + "learning_rate": 7.812865497076024e-05, + "loss": 3.4776, + "step": 12320 + }, + { + "epoch": 0.6484736842105263, + "grad_norm": 0.5234375, + "learning_rate": 7.811695906432749e-05, + "loss": 3.7636, + "step": 12321 + }, + { + "epoch": 0.6485263157894737, + "grad_norm": 0.4765625, + "learning_rate": 7.810526315789474e-05, + "loss": 3.2772, + "step": 12322 + }, + { + "epoch": 0.648578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.8093567251462e-05, + "loss": 3.3381, + "step": 12323 + }, + { + "epoch": 0.6486315789473684, + "grad_norm": 0.5078125, + "learning_rate": 7.808187134502925e-05, + "loss": 3.1764, + "step": 12324 + }, + { + "epoch": 0.6486842105263158, + "grad_norm": 0.5, + "learning_rate": 7.807017543859649e-05, + "loss": 3.4309, + "step": 12325 + }, + { + "epoch": 0.6487368421052632, + "grad_norm": 0.51953125, + "learning_rate": 7.805847953216374e-05, + "loss": 3.4347, + "step": 12326 + }, + { + "epoch": 0.6487894736842105, + "grad_norm": 0.49609375, + "learning_rate": 7.8046783625731e-05, + "loss": 3.1868, + "step": 12327 + }, + { + "epoch": 0.6488421052631579, + "grad_norm": 0.47265625, + "learning_rate": 7.803508771929825e-05, + "loss": 3.5753, + "step": 12328 + }, + { + "epoch": 0.6488947368421053, + "grad_norm": 0.53515625, + "learning_rate": 7.80233918128655e-05, + "loss": 3.1592, + "step": 12329 + }, + { + "epoch": 0.6489473684210526, + "grad_norm": 0.66015625, + "learning_rate": 7.801169590643275e-05, + "loss": 3.6524, + "step": 12330 + }, + { + "epoch": 0.649, + "grad_norm": 0.45703125, + "learning_rate": 7.800000000000001e-05, + "loss": 3.3769, + "step": 12331 + }, + { + "epoch": 0.6490526315789473, + "grad_norm": 0.486328125, + "learning_rate": 7.798830409356725e-05, + "loss": 3.1796, + "step": 12332 + }, + { + "epoch": 0.6491052631578947, + "grad_norm": 0.50390625, + "learning_rate": 7.79766081871345e-05, + "loss": 2.9304, + "step": 12333 + }, + { + "epoch": 0.6491578947368422, + "grad_norm": 0.52734375, + "learning_rate": 7.796491228070176e-05, + "loss": 3.0642, + "step": 12334 + }, + { + "epoch": 0.6492105263157895, + "grad_norm": 0.482421875, + "learning_rate": 7.795321637426901e-05, + "loss": 3.4633, + "step": 12335 + }, + { + "epoch": 0.6492631578947369, + "grad_norm": 0.51171875, + "learning_rate": 7.794152046783626e-05, + "loss": 3.3815, + "step": 12336 + }, + { + "epoch": 0.6493157894736842, + "grad_norm": 0.494140625, + "learning_rate": 7.792982456140351e-05, + "loss": 3.1618, + "step": 12337 + }, + { + "epoch": 0.6493684210526316, + "grad_norm": 0.51171875, + "learning_rate": 7.791812865497077e-05, + "loss": 3.5427, + "step": 12338 + }, + { + "epoch": 0.6494210526315789, + "grad_norm": 0.48828125, + "learning_rate": 7.790643274853802e-05, + "loss": 3.5724, + "step": 12339 + }, + { + "epoch": 0.6494736842105263, + "grad_norm": 0.5078125, + "learning_rate": 7.789473684210526e-05, + "loss": 3.2162, + "step": 12340 + }, + { + "epoch": 0.6495263157894737, + "grad_norm": 0.5390625, + "learning_rate": 7.788304093567251e-05, + "loss": 3.6958, + "step": 12341 + }, + { + "epoch": 0.649578947368421, + "grad_norm": 0.51953125, + "learning_rate": 7.787134502923977e-05, + "loss": 3.3668, + "step": 12342 + }, + { + "epoch": 0.6496315789473684, + "grad_norm": 0.45703125, + "learning_rate": 7.785964912280702e-05, + "loss": 3.5506, + "step": 12343 + }, + { + "epoch": 0.6496842105263158, + "grad_norm": 0.49609375, + "learning_rate": 7.784795321637427e-05, + "loss": 3.4492, + "step": 12344 + }, + { + "epoch": 0.6497368421052632, + "grad_norm": 0.5078125, + "learning_rate": 7.783625730994153e-05, + "loss": 3.6345, + "step": 12345 + }, + { + "epoch": 0.6497894736842105, + "grad_norm": 0.51171875, + "learning_rate": 7.782456140350878e-05, + "loss": 3.6437, + "step": 12346 + }, + { + "epoch": 0.6498421052631579, + "grad_norm": 0.5234375, + "learning_rate": 7.781286549707603e-05, + "loss": 3.3924, + "step": 12347 + }, + { + "epoch": 0.6498947368421053, + "grad_norm": 0.470703125, + "learning_rate": 7.780116959064327e-05, + "loss": 3.5088, + "step": 12348 + }, + { + "epoch": 0.6499473684210526, + "grad_norm": 0.453125, + "learning_rate": 7.778947368421053e-05, + "loss": 3.1085, + "step": 12349 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 7.777777777777778e-05, + "loss": 3.7389, + "step": 12350 + }, + { + "epoch": 0.6500526315789473, + "grad_norm": 0.5078125, + "learning_rate": 7.776608187134503e-05, + "loss": 3.2804, + "step": 12351 + }, + { + "epoch": 0.6501052631578947, + "grad_norm": 0.54296875, + "learning_rate": 7.775438596491228e-05, + "loss": 3.0744, + "step": 12352 + }, + { + "epoch": 0.6501578947368422, + "grad_norm": 0.5, + "learning_rate": 7.774269005847954e-05, + "loss": 3.4047, + "step": 12353 + }, + { + "epoch": 0.6502105263157895, + "grad_norm": 0.5078125, + "learning_rate": 7.77309941520468e-05, + "loss": 3.3534, + "step": 12354 + }, + { + "epoch": 0.6502631578947369, + "grad_norm": 0.55078125, + "learning_rate": 7.771929824561403e-05, + "loss": 3.4859, + "step": 12355 + }, + { + "epoch": 0.6503157894736842, + "grad_norm": 0.490234375, + "learning_rate": 7.770760233918129e-05, + "loss": 2.9414, + "step": 12356 + }, + { + "epoch": 0.6503684210526316, + "grad_norm": 0.48046875, + "learning_rate": 7.769590643274854e-05, + "loss": 3.3331, + "step": 12357 + }, + { + "epoch": 0.6504210526315789, + "grad_norm": 0.490234375, + "learning_rate": 7.768421052631579e-05, + "loss": 3.0877, + "step": 12358 + }, + { + "epoch": 0.6504736842105263, + "grad_norm": 0.498046875, + "learning_rate": 7.767251461988304e-05, + "loss": 3.9239, + "step": 12359 + }, + { + "epoch": 0.6505263157894737, + "grad_norm": 0.494140625, + "learning_rate": 7.76608187134503e-05, + "loss": 3.4105, + "step": 12360 + }, + { + "epoch": 0.650578947368421, + "grad_norm": 0.462890625, + "learning_rate": 7.764912280701755e-05, + "loss": 3.3715, + "step": 12361 + }, + { + "epoch": 0.6506315789473684, + "grad_norm": 0.55859375, + "learning_rate": 7.76374269005848e-05, + "loss": 3.5248, + "step": 12362 + }, + { + "epoch": 0.6506842105263158, + "grad_norm": 0.50390625, + "learning_rate": 7.762573099415204e-05, + "loss": 3.4638, + "step": 12363 + }, + { + "epoch": 0.6507368421052632, + "grad_norm": 0.486328125, + "learning_rate": 7.76140350877193e-05, + "loss": 3.1377, + "step": 12364 + }, + { + "epoch": 0.6507894736842105, + "grad_norm": 0.5234375, + "learning_rate": 7.760233918128655e-05, + "loss": 3.7324, + "step": 12365 + }, + { + "epoch": 0.6508421052631579, + "grad_norm": 0.5, + "learning_rate": 7.75906432748538e-05, + "loss": 3.7359, + "step": 12366 + }, + { + "epoch": 0.6508947368421053, + "grad_norm": 0.466796875, + "learning_rate": 7.757894736842105e-05, + "loss": 3.4207, + "step": 12367 + }, + { + "epoch": 0.6509473684210526, + "grad_norm": 0.474609375, + "learning_rate": 7.756725146198831e-05, + "loss": 3.3233, + "step": 12368 + }, + { + "epoch": 0.651, + "grad_norm": 0.462890625, + "learning_rate": 7.755555555555556e-05, + "loss": 3.7164, + "step": 12369 + }, + { + "epoch": 0.6510526315789473, + "grad_norm": 0.51953125, + "learning_rate": 7.754385964912281e-05, + "loss": 3.3648, + "step": 12370 + }, + { + "epoch": 0.6511052631578947, + "grad_norm": 0.49609375, + "learning_rate": 7.753216374269006e-05, + "loss": 3.4781, + "step": 12371 + }, + { + "epoch": 0.6511578947368422, + "grad_norm": 0.5234375, + "learning_rate": 7.752046783625731e-05, + "loss": 3.7915, + "step": 12372 + }, + { + "epoch": 0.6512105263157895, + "grad_norm": 0.49609375, + "learning_rate": 7.750877192982456e-05, + "loss": 3.1277, + "step": 12373 + }, + { + "epoch": 0.6512631578947369, + "grad_norm": 0.50390625, + "learning_rate": 7.749707602339181e-05, + "loss": 3.3401, + "step": 12374 + }, + { + "epoch": 0.6513157894736842, + "grad_norm": 0.4921875, + "learning_rate": 7.748538011695908e-05, + "loss": 3.3002, + "step": 12375 + }, + { + "epoch": 0.6513684210526316, + "grad_norm": 0.515625, + "learning_rate": 7.747368421052633e-05, + "loss": 3.5185, + "step": 12376 + }, + { + "epoch": 0.6514210526315789, + "grad_norm": 0.486328125, + "learning_rate": 7.746198830409357e-05, + "loss": 3.0983, + "step": 12377 + }, + { + "epoch": 0.6514736842105263, + "grad_norm": 0.5546875, + "learning_rate": 7.745029239766081e-05, + "loss": 2.6962, + "step": 12378 + }, + { + "epoch": 0.6515263157894737, + "grad_norm": 0.49609375, + "learning_rate": 7.743859649122807e-05, + "loss": 3.1374, + "step": 12379 + }, + { + "epoch": 0.651578947368421, + "grad_norm": 0.494140625, + "learning_rate": 7.742690058479532e-05, + "loss": 3.515, + "step": 12380 + }, + { + "epoch": 0.6516315789473685, + "grad_norm": 0.53515625, + "learning_rate": 7.741520467836257e-05, + "loss": 3.3582, + "step": 12381 + }, + { + "epoch": 0.6516842105263158, + "grad_norm": 0.85546875, + "learning_rate": 7.740350877192984e-05, + "loss": 3.4566, + "step": 12382 + }, + { + "epoch": 0.6517368421052632, + "grad_norm": 0.67578125, + "learning_rate": 7.739181286549709e-05, + "loss": 3.3214, + "step": 12383 + }, + { + "epoch": 0.6517894736842105, + "grad_norm": 0.53125, + "learning_rate": 7.738011695906434e-05, + "loss": 3.5203, + "step": 12384 + }, + { + "epoch": 0.6518421052631579, + "grad_norm": 0.4921875, + "learning_rate": 7.736842105263159e-05, + "loss": 3.7012, + "step": 12385 + }, + { + "epoch": 0.6518947368421053, + "grad_norm": 0.46484375, + "learning_rate": 7.735672514619883e-05, + "loss": 3.113, + "step": 12386 + }, + { + "epoch": 0.6519473684210526, + "grad_norm": 0.48828125, + "learning_rate": 7.734502923976608e-05, + "loss": 3.2715, + "step": 12387 + }, + { + "epoch": 0.652, + "grad_norm": 0.486328125, + "learning_rate": 7.733333333333333e-05, + "loss": 3.4056, + "step": 12388 + }, + { + "epoch": 0.6520526315789473, + "grad_norm": 0.474609375, + "learning_rate": 7.732163742690058e-05, + "loss": 3.4122, + "step": 12389 + }, + { + "epoch": 0.6521052631578947, + "grad_norm": 0.50390625, + "learning_rate": 7.730994152046785e-05, + "loss": 3.0656, + "step": 12390 + }, + { + "epoch": 0.6521578947368422, + "grad_norm": 0.474609375, + "learning_rate": 7.72982456140351e-05, + "loss": 3.3582, + "step": 12391 + }, + { + "epoch": 0.6522105263157895, + "grad_norm": 0.48046875, + "learning_rate": 7.728654970760235e-05, + "loss": 3.7404, + "step": 12392 + }, + { + "epoch": 0.6522631578947369, + "grad_norm": 0.5, + "learning_rate": 7.72748538011696e-05, + "loss": 3.0922, + "step": 12393 + }, + { + "epoch": 0.6523157894736842, + "grad_norm": 0.46875, + "learning_rate": 7.726315789473684e-05, + "loss": 3.5123, + "step": 12394 + }, + { + "epoch": 0.6523684210526316, + "grad_norm": 0.466796875, + "learning_rate": 7.72514619883041e-05, + "loss": 3.3546, + "step": 12395 + }, + { + "epoch": 0.6524210526315789, + "grad_norm": 0.5234375, + "learning_rate": 7.723976608187134e-05, + "loss": 3.5116, + "step": 12396 + }, + { + "epoch": 0.6524736842105263, + "grad_norm": 0.494140625, + "learning_rate": 7.722807017543861e-05, + "loss": 3.4631, + "step": 12397 + }, + { + "epoch": 0.6525263157894737, + "grad_norm": 0.53515625, + "learning_rate": 7.721637426900586e-05, + "loss": 3.4281, + "step": 12398 + }, + { + "epoch": 0.652578947368421, + "grad_norm": 0.484375, + "learning_rate": 7.72046783625731e-05, + "loss": 3.4335, + "step": 12399 + }, + { + "epoch": 0.6526315789473685, + "grad_norm": 0.484375, + "learning_rate": 7.719298245614036e-05, + "loss": 3.3862, + "step": 12400 + }, + { + "epoch": 0.6526842105263158, + "grad_norm": 0.49609375, + "learning_rate": 7.71812865497076e-05, + "loss": 3.483, + "step": 12401 + }, + { + "epoch": 0.6527368421052632, + "grad_norm": 0.4765625, + "learning_rate": 7.716959064327486e-05, + "loss": 3.7917, + "step": 12402 + }, + { + "epoch": 0.6527894736842105, + "grad_norm": 0.48046875, + "learning_rate": 7.71578947368421e-05, + "loss": 3.4929, + "step": 12403 + }, + { + "epoch": 0.6528421052631579, + "grad_norm": 0.47265625, + "learning_rate": 7.714619883040935e-05, + "loss": 3.5909, + "step": 12404 + }, + { + "epoch": 0.6528947368421053, + "grad_norm": 0.474609375, + "learning_rate": 7.713450292397662e-05, + "loss": 3.4911, + "step": 12405 + }, + { + "epoch": 0.6529473684210526, + "grad_norm": 0.462890625, + "learning_rate": 7.712280701754387e-05, + "loss": 3.4832, + "step": 12406 + }, + { + "epoch": 0.653, + "grad_norm": 0.4765625, + "learning_rate": 7.711111111111112e-05, + "loss": 3.3412, + "step": 12407 + }, + { + "epoch": 0.6530526315789473, + "grad_norm": 0.46484375, + "learning_rate": 7.709941520467837e-05, + "loss": 3.4311, + "step": 12408 + }, + { + "epoch": 0.6531052631578947, + "grad_norm": 0.5078125, + "learning_rate": 7.708771929824562e-05, + "loss": 3.6475, + "step": 12409 + }, + { + "epoch": 0.6531578947368422, + "grad_norm": 0.48046875, + "learning_rate": 7.707602339181287e-05, + "loss": 3.1816, + "step": 12410 + }, + { + "epoch": 0.6532105263157895, + "grad_norm": 0.4921875, + "learning_rate": 7.706432748538012e-05, + "loss": 3.2795, + "step": 12411 + }, + { + "epoch": 0.6532631578947369, + "grad_norm": 0.470703125, + "learning_rate": 7.705263157894738e-05, + "loss": 3.3419, + "step": 12412 + }, + { + "epoch": 0.6533157894736842, + "grad_norm": 0.482421875, + "learning_rate": 7.704093567251463e-05, + "loss": 3.4433, + "step": 12413 + }, + { + "epoch": 0.6533684210526316, + "grad_norm": 0.515625, + "learning_rate": 7.702923976608188e-05, + "loss": 3.3412, + "step": 12414 + }, + { + "epoch": 0.6534210526315789, + "grad_norm": 0.486328125, + "learning_rate": 7.701754385964913e-05, + "loss": 3.3916, + "step": 12415 + }, + { + "epoch": 0.6534736842105263, + "grad_norm": 0.490234375, + "learning_rate": 7.700584795321638e-05, + "loss": 3.5821, + "step": 12416 + }, + { + "epoch": 0.6535263157894737, + "grad_norm": 0.46875, + "learning_rate": 7.699415204678363e-05, + "loss": 2.9587, + "step": 12417 + }, + { + "epoch": 0.653578947368421, + "grad_norm": 0.46484375, + "learning_rate": 7.698245614035088e-05, + "loss": 3.543, + "step": 12418 + }, + { + "epoch": 0.6536315789473685, + "grad_norm": 0.474609375, + "learning_rate": 7.697076023391814e-05, + "loss": 2.9545, + "step": 12419 + }, + { + "epoch": 0.6536842105263158, + "grad_norm": 0.5234375, + "learning_rate": 7.695906432748539e-05, + "loss": 3.4307, + "step": 12420 + }, + { + "epoch": 0.6537368421052632, + "grad_norm": 0.484375, + "learning_rate": 7.694736842105264e-05, + "loss": 3.3048, + "step": 12421 + }, + { + "epoch": 0.6537894736842105, + "grad_norm": 0.55859375, + "learning_rate": 7.693567251461989e-05, + "loss": 3.2752, + "step": 12422 + }, + { + "epoch": 0.6538421052631579, + "grad_norm": 0.474609375, + "learning_rate": 7.692397660818714e-05, + "loss": 3.6813, + "step": 12423 + }, + { + "epoch": 0.6538947368421053, + "grad_norm": 0.51171875, + "learning_rate": 7.691228070175439e-05, + "loss": 3.5968, + "step": 12424 + }, + { + "epoch": 0.6539473684210526, + "grad_norm": 0.51171875, + "learning_rate": 7.690058479532164e-05, + "loss": 3.2659, + "step": 12425 + }, + { + "epoch": 0.654, + "grad_norm": 0.515625, + "learning_rate": 7.688888888888889e-05, + "loss": 3.3525, + "step": 12426 + }, + { + "epoch": 0.6540526315789473, + "grad_norm": 0.48046875, + "learning_rate": 7.687719298245615e-05, + "loss": 3.1702, + "step": 12427 + }, + { + "epoch": 0.6541052631578947, + "grad_norm": 0.515625, + "learning_rate": 7.68654970760234e-05, + "loss": 3.4747, + "step": 12428 + }, + { + "epoch": 0.6541578947368422, + "grad_norm": 0.51953125, + "learning_rate": 7.685380116959065e-05, + "loss": 3.3907, + "step": 12429 + }, + { + "epoch": 0.6542105263157895, + "grad_norm": 0.458984375, + "learning_rate": 7.68421052631579e-05, + "loss": 3.3721, + "step": 12430 + }, + { + "epoch": 0.6542631578947369, + "grad_norm": 0.4765625, + "learning_rate": 7.683040935672515e-05, + "loss": 3.1564, + "step": 12431 + }, + { + "epoch": 0.6543157894736842, + "grad_norm": 0.48828125, + "learning_rate": 7.68187134502924e-05, + "loss": 3.5753, + "step": 12432 + }, + { + "epoch": 0.6543684210526316, + "grad_norm": 0.48828125, + "learning_rate": 7.680701754385965e-05, + "loss": 3.4994, + "step": 12433 + }, + { + "epoch": 0.6544210526315789, + "grad_norm": 0.494140625, + "learning_rate": 7.679532163742691e-05, + "loss": 2.9734, + "step": 12434 + }, + { + "epoch": 0.6544736842105263, + "grad_norm": 0.51171875, + "learning_rate": 7.678362573099416e-05, + "loss": 3.6874, + "step": 12435 + }, + { + "epoch": 0.6545263157894737, + "grad_norm": 0.48046875, + "learning_rate": 7.677192982456141e-05, + "loss": 3.655, + "step": 12436 + }, + { + "epoch": 0.654578947368421, + "grad_norm": 0.453125, + "learning_rate": 7.676023391812866e-05, + "loss": 2.9627, + "step": 12437 + }, + { + "epoch": 0.6546315789473685, + "grad_norm": 0.5078125, + "learning_rate": 7.674853801169591e-05, + "loss": 3.14, + "step": 12438 + }, + { + "epoch": 0.6546842105263158, + "grad_norm": 0.490234375, + "learning_rate": 7.673684210526316e-05, + "loss": 3.5221, + "step": 12439 + }, + { + "epoch": 0.6547368421052632, + "grad_norm": 0.48046875, + "learning_rate": 7.672514619883041e-05, + "loss": 3.3046, + "step": 12440 + }, + { + "epoch": 0.6547894736842105, + "grad_norm": 0.482421875, + "learning_rate": 7.671345029239766e-05, + "loss": 3.2614, + "step": 12441 + }, + { + "epoch": 0.6548421052631579, + "grad_norm": 0.498046875, + "learning_rate": 7.670175438596492e-05, + "loss": 3.3328, + "step": 12442 + }, + { + "epoch": 0.6548947368421053, + "grad_norm": 0.69921875, + "learning_rate": 7.669005847953217e-05, + "loss": 3.8791, + "step": 12443 + }, + { + "epoch": 0.6549473684210526, + "grad_norm": 0.57421875, + "learning_rate": 7.667836257309942e-05, + "loss": 3.0057, + "step": 12444 + }, + { + "epoch": 0.655, + "grad_norm": 0.49609375, + "learning_rate": 7.666666666666667e-05, + "loss": 2.9347, + "step": 12445 + }, + { + "epoch": 0.6550526315789473, + "grad_norm": 0.47265625, + "learning_rate": 7.665497076023392e-05, + "loss": 3.4093, + "step": 12446 + }, + { + "epoch": 0.6551052631578947, + "grad_norm": 0.5078125, + "learning_rate": 7.664327485380117e-05, + "loss": 2.8434, + "step": 12447 + }, + { + "epoch": 0.655157894736842, + "grad_norm": 0.470703125, + "learning_rate": 7.663157894736842e-05, + "loss": 3.5027, + "step": 12448 + }, + { + "epoch": 0.6552105263157895, + "grad_norm": 0.48828125, + "learning_rate": 7.661988304093568e-05, + "loss": 3.4031, + "step": 12449 + }, + { + "epoch": 0.6552631578947369, + "grad_norm": 0.478515625, + "learning_rate": 7.660818713450293e-05, + "loss": 3.041, + "step": 12450 + }, + { + "epoch": 0.6553157894736842, + "grad_norm": 0.61328125, + "learning_rate": 7.659649122807018e-05, + "loss": 3.6225, + "step": 12451 + }, + { + "epoch": 0.6553684210526316, + "grad_norm": 0.470703125, + "learning_rate": 7.658479532163743e-05, + "loss": 3.2833, + "step": 12452 + }, + { + "epoch": 0.6554210526315789, + "grad_norm": 0.51171875, + "learning_rate": 7.657309941520468e-05, + "loss": 3.5541, + "step": 12453 + }, + { + "epoch": 0.6554736842105263, + "grad_norm": 0.490234375, + "learning_rate": 7.656140350877193e-05, + "loss": 3.4624, + "step": 12454 + }, + { + "epoch": 0.6555263157894737, + "grad_norm": 0.50390625, + "learning_rate": 7.654970760233918e-05, + "loss": 3.6402, + "step": 12455 + }, + { + "epoch": 0.655578947368421, + "grad_norm": 0.56640625, + "learning_rate": 7.653801169590644e-05, + "loss": 3.7469, + "step": 12456 + }, + { + "epoch": 0.6556315789473685, + "grad_norm": 0.51953125, + "learning_rate": 7.652631578947369e-05, + "loss": 3.5263, + "step": 12457 + }, + { + "epoch": 0.6556842105263158, + "grad_norm": 0.51171875, + "learning_rate": 7.651461988304094e-05, + "loss": 2.9958, + "step": 12458 + }, + { + "epoch": 0.6557368421052632, + "grad_norm": 0.52734375, + "learning_rate": 7.650292397660819e-05, + "loss": 3.1479, + "step": 12459 + }, + { + "epoch": 0.6557894736842105, + "grad_norm": 0.46484375, + "learning_rate": 7.649122807017545e-05, + "loss": 3.023, + "step": 12460 + }, + { + "epoch": 0.6558421052631579, + "grad_norm": 0.478515625, + "learning_rate": 7.647953216374269e-05, + "loss": 3.3251, + "step": 12461 + }, + { + "epoch": 0.6558947368421053, + "grad_norm": 0.486328125, + "learning_rate": 7.646783625730994e-05, + "loss": 3.5064, + "step": 12462 + }, + { + "epoch": 0.6559473684210526, + "grad_norm": 0.5078125, + "learning_rate": 7.645614035087719e-05, + "loss": 3.7016, + "step": 12463 + }, + { + "epoch": 0.656, + "grad_norm": 0.4765625, + "learning_rate": 7.644444444444445e-05, + "loss": 3.2573, + "step": 12464 + }, + { + "epoch": 0.6560526315789473, + "grad_norm": 0.51953125, + "learning_rate": 7.64327485380117e-05, + "loss": 3.7447, + "step": 12465 + }, + { + "epoch": 0.6561052631578947, + "grad_norm": 0.46484375, + "learning_rate": 7.642105263157895e-05, + "loss": 3.3377, + "step": 12466 + }, + { + "epoch": 0.656157894736842, + "grad_norm": 0.490234375, + "learning_rate": 7.64093567251462e-05, + "loss": 3.5236, + "step": 12467 + }, + { + "epoch": 0.6562105263157895, + "grad_norm": 0.4921875, + "learning_rate": 7.639766081871345e-05, + "loss": 3.4343, + "step": 12468 + }, + { + "epoch": 0.6562631578947369, + "grad_norm": 0.470703125, + "learning_rate": 7.63859649122807e-05, + "loss": 3.7477, + "step": 12469 + }, + { + "epoch": 0.6563157894736842, + "grad_norm": 0.53125, + "learning_rate": 7.637426900584795e-05, + "loss": 3.479, + "step": 12470 + }, + { + "epoch": 0.6563684210526316, + "grad_norm": 0.498046875, + "learning_rate": 7.636257309941521e-05, + "loss": 3.0769, + "step": 12471 + }, + { + "epoch": 0.6564210526315789, + "grad_norm": 0.49609375, + "learning_rate": 7.635087719298246e-05, + "loss": 3.6135, + "step": 12472 + }, + { + "epoch": 0.6564736842105263, + "grad_norm": 0.53125, + "learning_rate": 7.633918128654971e-05, + "loss": 3.4151, + "step": 12473 + }, + { + "epoch": 0.6565263157894737, + "grad_norm": 0.466796875, + "learning_rate": 7.632748538011696e-05, + "loss": 3.5206, + "step": 12474 + }, + { + "epoch": 0.656578947368421, + "grad_norm": 0.484375, + "learning_rate": 7.631578947368422e-05, + "loss": 3.5447, + "step": 12475 + }, + { + "epoch": 0.6566315789473685, + "grad_norm": 0.474609375, + "learning_rate": 7.630409356725146e-05, + "loss": 3.3865, + "step": 12476 + }, + { + "epoch": 0.6566842105263158, + "grad_norm": 0.63671875, + "learning_rate": 7.629239766081871e-05, + "loss": 3.344, + "step": 12477 + }, + { + "epoch": 0.6567368421052632, + "grad_norm": 0.474609375, + "learning_rate": 7.628070175438596e-05, + "loss": 3.2331, + "step": 12478 + }, + { + "epoch": 0.6567894736842105, + "grad_norm": 0.453125, + "learning_rate": 7.626900584795322e-05, + "loss": 3.4604, + "step": 12479 + }, + { + "epoch": 0.6568421052631579, + "grad_norm": 0.48828125, + "learning_rate": 7.625730994152047e-05, + "loss": 2.9536, + "step": 12480 + }, + { + "epoch": 0.6568947368421053, + "grad_norm": 0.466796875, + "learning_rate": 7.624561403508772e-05, + "loss": 3.3551, + "step": 12481 + }, + { + "epoch": 0.6569473684210526, + "grad_norm": 0.4453125, + "learning_rate": 7.623391812865499e-05, + "loss": 3.3001, + "step": 12482 + }, + { + "epoch": 0.657, + "grad_norm": 0.484375, + "learning_rate": 7.622222222222223e-05, + "loss": 3.6756, + "step": 12483 + }, + { + "epoch": 0.6570526315789473, + "grad_norm": 0.478515625, + "learning_rate": 7.621052631578947e-05, + "loss": 2.8597, + "step": 12484 + }, + { + "epoch": 0.6571052631578947, + "grad_norm": 0.48046875, + "learning_rate": 7.619883040935672e-05, + "loss": 3.5394, + "step": 12485 + }, + { + "epoch": 0.657157894736842, + "grad_norm": 0.49609375, + "learning_rate": 7.618713450292398e-05, + "loss": 3.1456, + "step": 12486 + }, + { + "epoch": 0.6572105263157895, + "grad_norm": 0.5, + "learning_rate": 7.617543859649123e-05, + "loss": 3.178, + "step": 12487 + }, + { + "epoch": 0.6572631578947369, + "grad_norm": 0.5078125, + "learning_rate": 7.616374269005848e-05, + "loss": 3.3499, + "step": 12488 + }, + { + "epoch": 0.6573157894736842, + "grad_norm": 0.490234375, + "learning_rate": 7.615204678362573e-05, + "loss": 2.9381, + "step": 12489 + }, + { + "epoch": 0.6573684210526316, + "grad_norm": 0.498046875, + "learning_rate": 7.6140350877193e-05, + "loss": 3.3018, + "step": 12490 + }, + { + "epoch": 0.6574210526315789, + "grad_norm": 0.5, + "learning_rate": 7.612865497076023e-05, + "loss": 3.3134, + "step": 12491 + }, + { + "epoch": 0.6574736842105263, + "grad_norm": 0.486328125, + "learning_rate": 7.611695906432748e-05, + "loss": 3.3613, + "step": 12492 + }, + { + "epoch": 0.6575263157894737, + "grad_norm": 0.486328125, + "learning_rate": 7.610526315789473e-05, + "loss": 3.4945, + "step": 12493 + }, + { + "epoch": 0.657578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.6093567251462e-05, + "loss": 2.9238, + "step": 12494 + }, + { + "epoch": 0.6576315789473685, + "grad_norm": 0.48046875, + "learning_rate": 7.608187134502924e-05, + "loss": 3.4524, + "step": 12495 + }, + { + "epoch": 0.6576842105263158, + "grad_norm": 0.482421875, + "learning_rate": 7.607017543859649e-05, + "loss": 3.5258, + "step": 12496 + }, + { + "epoch": 0.6577368421052632, + "grad_norm": 0.49609375, + "learning_rate": 7.605847953216376e-05, + "loss": 3.5002, + "step": 12497 + }, + { + "epoch": 0.6577894736842105, + "grad_norm": 0.5078125, + "learning_rate": 7.6046783625731e-05, + "loss": 2.8187, + "step": 12498 + }, + { + "epoch": 0.6578421052631579, + "grad_norm": 0.498046875, + "learning_rate": 7.603508771929824e-05, + "loss": 3.8101, + "step": 12499 + }, + { + "epoch": 0.6578947368421053, + "grad_norm": 0.5078125, + "learning_rate": 7.602339181286549e-05, + "loss": 3.5704, + "step": 12500 + }, + { + "epoch": 0.6579473684210526, + "grad_norm": 0.50390625, + "learning_rate": 7.601169590643275e-05, + "loss": 3.7798, + "step": 12501 + }, + { + "epoch": 0.658, + "grad_norm": 0.498046875, + "learning_rate": 7.6e-05, + "loss": 3.079, + "step": 12502 + }, + { + "epoch": 0.6580526315789473, + "grad_norm": 0.5234375, + "learning_rate": 7.598830409356725e-05, + "loss": 3.0804, + "step": 12503 + }, + { + "epoch": 0.6581052631578947, + "grad_norm": 0.46875, + "learning_rate": 7.59766081871345e-05, + "loss": 3.3215, + "step": 12504 + }, + { + "epoch": 0.658157894736842, + "grad_norm": 0.50390625, + "learning_rate": 7.596491228070177e-05, + "loss": 3.425, + "step": 12505 + }, + { + "epoch": 0.6582105263157895, + "grad_norm": 0.47265625, + "learning_rate": 7.595321637426902e-05, + "loss": 3.3802, + "step": 12506 + }, + { + "epoch": 0.6582631578947369, + "grad_norm": 0.5078125, + "learning_rate": 7.594152046783625e-05, + "loss": 3.3195, + "step": 12507 + }, + { + "epoch": 0.6583157894736842, + "grad_norm": 0.484375, + "learning_rate": 7.592982456140352e-05, + "loss": 3.533, + "step": 12508 + }, + { + "epoch": 0.6583684210526316, + "grad_norm": 0.474609375, + "learning_rate": 7.591812865497076e-05, + "loss": 3.0286, + "step": 12509 + }, + { + "epoch": 0.6584210526315789, + "grad_norm": 0.44921875, + "learning_rate": 7.590643274853801e-05, + "loss": 3.1817, + "step": 12510 + }, + { + "epoch": 0.6584736842105263, + "grad_norm": 0.48828125, + "learning_rate": 7.589473684210526e-05, + "loss": 3.5679, + "step": 12511 + }, + { + "epoch": 0.6585263157894737, + "grad_norm": 0.478515625, + "learning_rate": 7.588304093567253e-05, + "loss": 3.1443, + "step": 12512 + }, + { + "epoch": 0.658578947368421, + "grad_norm": 0.51953125, + "learning_rate": 7.587134502923978e-05, + "loss": 3.3701, + "step": 12513 + }, + { + "epoch": 0.6586315789473685, + "grad_norm": 0.46875, + "learning_rate": 7.585964912280701e-05, + "loss": 2.9132, + "step": 12514 + }, + { + "epoch": 0.6586842105263158, + "grad_norm": 0.458984375, + "learning_rate": 7.584795321637426e-05, + "loss": 3.1166, + "step": 12515 + }, + { + "epoch": 0.6587368421052632, + "grad_norm": 0.462890625, + "learning_rate": 7.583625730994153e-05, + "loss": 3.1601, + "step": 12516 + }, + { + "epoch": 0.6587894736842105, + "grad_norm": 0.478515625, + "learning_rate": 7.582456140350878e-05, + "loss": 3.4411, + "step": 12517 + }, + { + "epoch": 0.6588421052631579, + "grad_norm": 0.490234375, + "learning_rate": 7.581286549707602e-05, + "loss": 2.9451, + "step": 12518 + }, + { + "epoch": 0.6588947368421053, + "grad_norm": 0.4921875, + "learning_rate": 7.580116959064329e-05, + "loss": 3.1485, + "step": 12519 + }, + { + "epoch": 0.6589473684210526, + "grad_norm": 0.515625, + "learning_rate": 7.578947368421054e-05, + "loss": 3.4889, + "step": 12520 + }, + { + "epoch": 0.659, + "grad_norm": 0.5078125, + "learning_rate": 7.577777777777779e-05, + "loss": 3.5999, + "step": 12521 + }, + { + "epoch": 0.6590526315789473, + "grad_norm": 0.51171875, + "learning_rate": 7.576608187134502e-05, + "loss": 3.6128, + "step": 12522 + }, + { + "epoch": 0.6591052631578947, + "grad_norm": 0.486328125, + "learning_rate": 7.575438596491229e-05, + "loss": 3.3043, + "step": 12523 + }, + { + "epoch": 0.659157894736842, + "grad_norm": 0.470703125, + "learning_rate": 7.574269005847954e-05, + "loss": 3.1369, + "step": 12524 + }, + { + "epoch": 0.6592105263157895, + "grad_norm": 0.46875, + "learning_rate": 7.573099415204679e-05, + "loss": 3.2584, + "step": 12525 + }, + { + "epoch": 0.6592631578947369, + "grad_norm": 0.5, + "learning_rate": 7.571929824561404e-05, + "loss": 3.2382, + "step": 12526 + }, + { + "epoch": 0.6593157894736842, + "grad_norm": 0.470703125, + "learning_rate": 7.57076023391813e-05, + "loss": 3.5286, + "step": 12527 + }, + { + "epoch": 0.6593684210526316, + "grad_norm": 0.51171875, + "learning_rate": 7.569590643274855e-05, + "loss": 3.6985, + "step": 12528 + }, + { + "epoch": 0.6594210526315789, + "grad_norm": 0.462890625, + "learning_rate": 7.56842105263158e-05, + "loss": 3.1838, + "step": 12529 + }, + { + "epoch": 0.6594736842105263, + "grad_norm": 0.57421875, + "learning_rate": 7.567251461988303e-05, + "loss": 3.3125, + "step": 12530 + }, + { + "epoch": 0.6595263157894737, + "grad_norm": 0.51953125, + "learning_rate": 7.56608187134503e-05, + "loss": 3.165, + "step": 12531 + }, + { + "epoch": 0.659578947368421, + "grad_norm": 0.5078125, + "learning_rate": 7.564912280701755e-05, + "loss": 3.7146, + "step": 12532 + }, + { + "epoch": 0.6596315789473685, + "grad_norm": 0.53515625, + "learning_rate": 7.56374269005848e-05, + "loss": 2.7411, + "step": 12533 + }, + { + "epoch": 0.6596842105263158, + "grad_norm": 0.5390625, + "learning_rate": 7.562573099415206e-05, + "loss": 3.2688, + "step": 12534 + }, + { + "epoch": 0.6597368421052632, + "grad_norm": 0.4921875, + "learning_rate": 7.561403508771931e-05, + "loss": 3.4035, + "step": 12535 + }, + { + "epoch": 0.6597894736842105, + "grad_norm": 0.546875, + "learning_rate": 7.560233918128656e-05, + "loss": 3.0858, + "step": 12536 + }, + { + "epoch": 0.6598421052631579, + "grad_norm": 0.51171875, + "learning_rate": 7.55906432748538e-05, + "loss": 3.5214, + "step": 12537 + }, + { + "epoch": 0.6598947368421053, + "grad_norm": 0.474609375, + "learning_rate": 7.557894736842106e-05, + "loss": 3.4761, + "step": 12538 + }, + { + "epoch": 0.6599473684210526, + "grad_norm": 0.51171875, + "learning_rate": 7.556725146198831e-05, + "loss": 3.3051, + "step": 12539 + }, + { + "epoch": 0.66, + "grad_norm": 0.484375, + "learning_rate": 7.555555555555556e-05, + "loss": 3.592, + "step": 12540 + }, + { + "epoch": 0.6600526315789473, + "grad_norm": 0.46484375, + "learning_rate": 7.55438596491228e-05, + "loss": 3.7859, + "step": 12541 + }, + { + "epoch": 0.6601052631578948, + "grad_norm": 1.0546875, + "learning_rate": 7.553216374269007e-05, + "loss": 3.2998, + "step": 12542 + }, + { + "epoch": 0.660157894736842, + "grad_norm": 0.47265625, + "learning_rate": 7.552046783625732e-05, + "loss": 3.8588, + "step": 12543 + }, + { + "epoch": 0.6602105263157895, + "grad_norm": 0.5078125, + "learning_rate": 7.550877192982457e-05, + "loss": 3.0084, + "step": 12544 + }, + { + "epoch": 0.6602631578947369, + "grad_norm": 0.51171875, + "learning_rate": 7.549707602339182e-05, + "loss": 3.4067, + "step": 12545 + }, + { + "epoch": 0.6603157894736842, + "grad_norm": 0.478515625, + "learning_rate": 7.548538011695907e-05, + "loss": 3.3588, + "step": 12546 + }, + { + "epoch": 0.6603684210526316, + "grad_norm": 0.482421875, + "learning_rate": 7.547368421052632e-05, + "loss": 3.4672, + "step": 12547 + }, + { + "epoch": 0.6604210526315789, + "grad_norm": 0.50390625, + "learning_rate": 7.546198830409357e-05, + "loss": 3.7526, + "step": 12548 + }, + { + "epoch": 0.6604736842105263, + "grad_norm": 0.49609375, + "learning_rate": 7.545029239766083e-05, + "loss": 3.5299, + "step": 12549 + }, + { + "epoch": 0.6605263157894737, + "grad_norm": 0.49609375, + "learning_rate": 7.543859649122808e-05, + "loss": 3.6178, + "step": 12550 + }, + { + "epoch": 0.660578947368421, + "grad_norm": 0.48828125, + "learning_rate": 7.542690058479533e-05, + "loss": 3.3111, + "step": 12551 + }, + { + "epoch": 0.6606315789473685, + "grad_norm": 0.470703125, + "learning_rate": 7.541520467836258e-05, + "loss": 3.1099, + "step": 12552 + }, + { + "epoch": 0.6606842105263158, + "grad_norm": 0.482421875, + "learning_rate": 7.540350877192983e-05, + "loss": 3.4282, + "step": 12553 + }, + { + "epoch": 0.6607368421052632, + "grad_norm": 0.46484375, + "learning_rate": 7.539181286549708e-05, + "loss": 3.5691, + "step": 12554 + }, + { + "epoch": 0.6607894736842105, + "grad_norm": 0.5, + "learning_rate": 7.538011695906433e-05, + "loss": 3.137, + "step": 12555 + }, + { + "epoch": 0.6608421052631579, + "grad_norm": 0.51171875, + "learning_rate": 7.536842105263158e-05, + "loss": 3.1644, + "step": 12556 + }, + { + "epoch": 0.6608947368421053, + "grad_norm": 0.490234375, + "learning_rate": 7.535672514619884e-05, + "loss": 3.5263, + "step": 12557 + }, + { + "epoch": 0.6609473684210526, + "grad_norm": 0.51171875, + "learning_rate": 7.534502923976609e-05, + "loss": 3.411, + "step": 12558 + }, + { + "epoch": 0.661, + "grad_norm": 0.5, + "learning_rate": 7.533333333333334e-05, + "loss": 3.2169, + "step": 12559 + }, + { + "epoch": 0.6610526315789473, + "grad_norm": 0.515625, + "learning_rate": 7.532163742690059e-05, + "loss": 3.5647, + "step": 12560 + }, + { + "epoch": 0.6611052631578948, + "grad_norm": 0.5390625, + "learning_rate": 7.530994152046784e-05, + "loss": 3.3826, + "step": 12561 + }, + { + "epoch": 0.6611578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.529824561403509e-05, + "loss": 3.8181, + "step": 12562 + }, + { + "epoch": 0.6612105263157895, + "grad_norm": 0.478515625, + "learning_rate": 7.528654970760234e-05, + "loss": 3.0147, + "step": 12563 + }, + { + "epoch": 0.6612631578947369, + "grad_norm": 0.470703125, + "learning_rate": 7.52748538011696e-05, + "loss": 3.2179, + "step": 12564 + }, + { + "epoch": 0.6613157894736842, + "grad_norm": 0.47265625, + "learning_rate": 7.526315789473685e-05, + "loss": 3.3799, + "step": 12565 + }, + { + "epoch": 0.6613684210526316, + "grad_norm": 0.5625, + "learning_rate": 7.52514619883041e-05, + "loss": 3.7748, + "step": 12566 + }, + { + "epoch": 0.6614210526315789, + "grad_norm": 0.51171875, + "learning_rate": 7.523976608187135e-05, + "loss": 3.4006, + "step": 12567 + }, + { + "epoch": 0.6614736842105263, + "grad_norm": 0.49609375, + "learning_rate": 7.52280701754386e-05, + "loss": 3.2136, + "step": 12568 + }, + { + "epoch": 0.6615263157894736, + "grad_norm": 0.50390625, + "learning_rate": 7.521637426900585e-05, + "loss": 3.6644, + "step": 12569 + }, + { + "epoch": 0.661578947368421, + "grad_norm": 0.51171875, + "learning_rate": 7.52046783625731e-05, + "loss": 2.969, + "step": 12570 + }, + { + "epoch": 0.6616315789473685, + "grad_norm": 0.5, + "learning_rate": 7.519298245614036e-05, + "loss": 3.0412, + "step": 12571 + }, + { + "epoch": 0.6616842105263158, + "grad_norm": 0.5078125, + "learning_rate": 7.518128654970761e-05, + "loss": 3.4564, + "step": 12572 + }, + { + "epoch": 0.6617368421052632, + "grad_norm": 0.53125, + "learning_rate": 7.516959064327486e-05, + "loss": 3.2202, + "step": 12573 + }, + { + "epoch": 0.6617894736842105, + "grad_norm": 0.482421875, + "learning_rate": 7.515789473684211e-05, + "loss": 3.1449, + "step": 12574 + }, + { + "epoch": 0.6618421052631579, + "grad_norm": 0.5234375, + "learning_rate": 7.514619883040936e-05, + "loss": 3.1272, + "step": 12575 + }, + { + "epoch": 0.6618947368421053, + "grad_norm": 0.466796875, + "learning_rate": 7.513450292397661e-05, + "loss": 3.5385, + "step": 12576 + }, + { + "epoch": 0.6619473684210526, + "grad_norm": 0.55078125, + "learning_rate": 7.512280701754386e-05, + "loss": 3.6955, + "step": 12577 + }, + { + "epoch": 0.662, + "grad_norm": 0.53125, + "learning_rate": 7.511111111111111e-05, + "loss": 3.0146, + "step": 12578 + }, + { + "epoch": 0.6620526315789473, + "grad_norm": 0.494140625, + "learning_rate": 7.509941520467837e-05, + "loss": 3.3956, + "step": 12579 + }, + { + "epoch": 0.6621052631578948, + "grad_norm": 0.462890625, + "learning_rate": 7.508771929824562e-05, + "loss": 2.9549, + "step": 12580 + }, + { + "epoch": 0.6621578947368421, + "grad_norm": 0.51953125, + "learning_rate": 7.507602339181287e-05, + "loss": 3.1767, + "step": 12581 + }, + { + "epoch": 0.6622105263157895, + "grad_norm": 0.5, + "learning_rate": 7.506432748538012e-05, + "loss": 3.2752, + "step": 12582 + }, + { + "epoch": 0.6622631578947369, + "grad_norm": 0.51953125, + "learning_rate": 7.505263157894737e-05, + "loss": 3.58, + "step": 12583 + }, + { + "epoch": 0.6623157894736842, + "grad_norm": 0.486328125, + "learning_rate": 7.504093567251462e-05, + "loss": 3.4788, + "step": 12584 + }, + { + "epoch": 0.6623684210526316, + "grad_norm": 0.5546875, + "learning_rate": 7.502923976608187e-05, + "loss": 3.3548, + "step": 12585 + }, + { + "epoch": 0.6624210526315789, + "grad_norm": 0.4921875, + "learning_rate": 7.501754385964913e-05, + "loss": 3.3445, + "step": 12586 + }, + { + "epoch": 0.6624736842105263, + "grad_norm": 0.515625, + "learning_rate": 7.500584795321638e-05, + "loss": 3.2618, + "step": 12587 + }, + { + "epoch": 0.6625263157894736, + "grad_norm": 0.48046875, + "learning_rate": 7.499415204678363e-05, + "loss": 3.3523, + "step": 12588 + }, + { + "epoch": 0.662578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.498245614035088e-05, + "loss": 3.2805, + "step": 12589 + }, + { + "epoch": 0.6626315789473685, + "grad_norm": 0.5078125, + "learning_rate": 7.497076023391813e-05, + "loss": 2.9282, + "step": 12590 + }, + { + "epoch": 0.6626842105263158, + "grad_norm": 0.53125, + "learning_rate": 7.495906432748538e-05, + "loss": 3.0342, + "step": 12591 + }, + { + "epoch": 0.6627368421052632, + "grad_norm": 0.54296875, + "learning_rate": 7.494736842105263e-05, + "loss": 3.175, + "step": 12592 + }, + { + "epoch": 0.6627894736842105, + "grad_norm": 0.498046875, + "learning_rate": 7.493567251461988e-05, + "loss": 3.1321, + "step": 12593 + }, + { + "epoch": 0.6628421052631579, + "grad_norm": 0.5, + "learning_rate": 7.492397660818714e-05, + "loss": 3.2048, + "step": 12594 + }, + { + "epoch": 0.6628947368421053, + "grad_norm": 0.50390625, + "learning_rate": 7.491228070175439e-05, + "loss": 3.4197, + "step": 12595 + }, + { + "epoch": 0.6629473684210526, + "grad_norm": 0.52734375, + "learning_rate": 7.490058479532164e-05, + "loss": 3.4143, + "step": 12596 + }, + { + "epoch": 0.663, + "grad_norm": 0.52734375, + "learning_rate": 7.488888888888889e-05, + "loss": 3.5808, + "step": 12597 + }, + { + "epoch": 0.6630526315789473, + "grad_norm": 0.474609375, + "learning_rate": 7.487719298245614e-05, + "loss": 3.5768, + "step": 12598 + }, + { + "epoch": 0.6631052631578948, + "grad_norm": 0.466796875, + "learning_rate": 7.486549707602339e-05, + "loss": 3.6913, + "step": 12599 + }, + { + "epoch": 0.6631578947368421, + "grad_norm": 0.4921875, + "learning_rate": 7.485380116959064e-05, + "loss": 3.0093, + "step": 12600 + }, + { + "epoch": 0.6632105263157895, + "grad_norm": 0.5, + "learning_rate": 7.48421052631579e-05, + "loss": 3.2533, + "step": 12601 + }, + { + "epoch": 0.6632631578947369, + "grad_norm": 0.474609375, + "learning_rate": 7.483040935672515e-05, + "loss": 3.4547, + "step": 12602 + }, + { + "epoch": 0.6633157894736842, + "grad_norm": 0.48046875, + "learning_rate": 7.48187134502924e-05, + "loss": 3.1787, + "step": 12603 + }, + { + "epoch": 0.6633684210526316, + "grad_norm": 0.46484375, + "learning_rate": 7.480701754385965e-05, + "loss": 3.5682, + "step": 12604 + }, + { + "epoch": 0.6634210526315789, + "grad_norm": 0.51953125, + "learning_rate": 7.47953216374269e-05, + "loss": 3.3392, + "step": 12605 + }, + { + "epoch": 0.6634736842105263, + "grad_norm": 0.466796875, + "learning_rate": 7.478362573099415e-05, + "loss": 3.1301, + "step": 12606 + }, + { + "epoch": 0.6635263157894736, + "grad_norm": 0.6640625, + "learning_rate": 7.47719298245614e-05, + "loss": 3.6117, + "step": 12607 + }, + { + "epoch": 0.663578947368421, + "grad_norm": 0.494140625, + "learning_rate": 7.476023391812866e-05, + "loss": 3.59, + "step": 12608 + }, + { + "epoch": 0.6636315789473685, + "grad_norm": 0.50390625, + "learning_rate": 7.474853801169591e-05, + "loss": 3.2142, + "step": 12609 + }, + { + "epoch": 0.6636842105263158, + "grad_norm": 0.50390625, + "learning_rate": 7.473684210526316e-05, + "loss": 3.3317, + "step": 12610 + }, + { + "epoch": 0.6637368421052632, + "grad_norm": 0.46484375, + "learning_rate": 7.472514619883041e-05, + "loss": 3.198, + "step": 12611 + }, + { + "epoch": 0.6637894736842105, + "grad_norm": 0.453125, + "learning_rate": 7.471345029239766e-05, + "loss": 3.7232, + "step": 12612 + }, + { + "epoch": 0.6638421052631579, + "grad_norm": 0.546875, + "learning_rate": 7.470175438596491e-05, + "loss": 3.4374, + "step": 12613 + }, + { + "epoch": 0.6638947368421053, + "grad_norm": 0.49609375, + "learning_rate": 7.469005847953216e-05, + "loss": 3.1995, + "step": 12614 + }, + { + "epoch": 0.6639473684210526, + "grad_norm": 0.4921875, + "learning_rate": 7.467836257309941e-05, + "loss": 3.0145, + "step": 12615 + }, + { + "epoch": 0.664, + "grad_norm": 0.494140625, + "learning_rate": 7.466666666666667e-05, + "loss": 3.3197, + "step": 12616 + }, + { + "epoch": 0.6640526315789473, + "grad_norm": 0.53125, + "learning_rate": 7.465497076023392e-05, + "loss": 3.2049, + "step": 12617 + }, + { + "epoch": 0.6641052631578948, + "grad_norm": 0.48828125, + "learning_rate": 7.464327485380117e-05, + "loss": 3.2726, + "step": 12618 + }, + { + "epoch": 0.6641578947368421, + "grad_norm": 0.51953125, + "learning_rate": 7.463157894736844e-05, + "loss": 2.7533, + "step": 12619 + }, + { + "epoch": 0.6642105263157895, + "grad_norm": 0.546875, + "learning_rate": 7.461988304093567e-05, + "loss": 2.8999, + "step": 12620 + }, + { + "epoch": 0.6642631578947369, + "grad_norm": 0.484375, + "learning_rate": 7.460818713450292e-05, + "loss": 3.2109, + "step": 12621 + }, + { + "epoch": 0.6643157894736842, + "grad_norm": 0.478515625, + "learning_rate": 7.459649122807017e-05, + "loss": 3.1524, + "step": 12622 + }, + { + "epoch": 0.6643684210526316, + "grad_norm": 0.54296875, + "learning_rate": 7.458479532163744e-05, + "loss": 2.9257, + "step": 12623 + }, + { + "epoch": 0.6644210526315789, + "grad_norm": 0.498046875, + "learning_rate": 7.457309941520468e-05, + "loss": 3.3621, + "step": 12624 + }, + { + "epoch": 0.6644736842105263, + "grad_norm": 0.51171875, + "learning_rate": 7.456140350877193e-05, + "loss": 2.9232, + "step": 12625 + }, + { + "epoch": 0.6645263157894736, + "grad_norm": 0.55078125, + "learning_rate": 7.454970760233918e-05, + "loss": 3.1308, + "step": 12626 + }, + { + "epoch": 0.664578947368421, + "grad_norm": 0.54296875, + "learning_rate": 7.453801169590643e-05, + "loss": 3.219, + "step": 12627 + }, + { + "epoch": 0.6646315789473685, + "grad_norm": 0.5, + "learning_rate": 7.452631578947368e-05, + "loss": 3.0618, + "step": 12628 + }, + { + "epoch": 0.6646842105263158, + "grad_norm": 0.51953125, + "learning_rate": 7.451461988304093e-05, + "loss": 3.1449, + "step": 12629 + }, + { + "epoch": 0.6647368421052632, + "grad_norm": 0.546875, + "learning_rate": 7.450292397660818e-05, + "loss": 2.9435, + "step": 12630 + }, + { + "epoch": 0.6647894736842105, + "grad_norm": 0.6171875, + "learning_rate": 7.449122807017545e-05, + "loss": 2.9826, + "step": 12631 + }, + { + "epoch": 0.6648421052631579, + "grad_norm": 0.46875, + "learning_rate": 7.44795321637427e-05, + "loss": 3.1969, + "step": 12632 + }, + { + "epoch": 0.6648947368421053, + "grad_norm": 0.5390625, + "learning_rate": 7.446783625730994e-05, + "loss": 2.8583, + "step": 12633 + }, + { + "epoch": 0.6649473684210526, + "grad_norm": 0.53125, + "learning_rate": 7.445614035087721e-05, + "loss": 3.5204, + "step": 12634 + }, + { + "epoch": 0.665, + "grad_norm": 0.490234375, + "learning_rate": 7.444444444444444e-05, + "loss": 2.9039, + "step": 12635 + }, + { + "epoch": 0.6650526315789473, + "grad_norm": 0.498046875, + "learning_rate": 7.44327485380117e-05, + "loss": 3.0116, + "step": 12636 + }, + { + "epoch": 0.6651052631578948, + "grad_norm": 0.4765625, + "learning_rate": 7.442105263157894e-05, + "loss": 3.7506, + "step": 12637 + }, + { + "epoch": 0.6651578947368421, + "grad_norm": 0.51171875, + "learning_rate": 7.44093567251462e-05, + "loss": 3.3003, + "step": 12638 + }, + { + "epoch": 0.6652105263157895, + "grad_norm": 0.4765625, + "learning_rate": 7.439766081871346e-05, + "loss": 3.0335, + "step": 12639 + }, + { + "epoch": 0.6652631578947369, + "grad_norm": 0.5078125, + "learning_rate": 7.43859649122807e-05, + "loss": 3.4214, + "step": 12640 + }, + { + "epoch": 0.6653157894736842, + "grad_norm": 0.46484375, + "learning_rate": 7.437426900584796e-05, + "loss": 3.6892, + "step": 12641 + }, + { + "epoch": 0.6653684210526316, + "grad_norm": 0.51953125, + "learning_rate": 7.436257309941522e-05, + "loss": 3.4628, + "step": 12642 + }, + { + "epoch": 0.6654210526315789, + "grad_norm": 0.490234375, + "learning_rate": 7.435087719298245e-05, + "loss": 2.9802, + "step": 12643 + }, + { + "epoch": 0.6654736842105263, + "grad_norm": 0.49609375, + "learning_rate": 7.43391812865497e-05, + "loss": 3.4217, + "step": 12644 + }, + { + "epoch": 0.6655263157894736, + "grad_norm": 0.47265625, + "learning_rate": 7.432748538011697e-05, + "loss": 3.334, + "step": 12645 + }, + { + "epoch": 0.665578947368421, + "grad_norm": 0.4765625, + "learning_rate": 7.431578947368422e-05, + "loss": 3.5539, + "step": 12646 + }, + { + "epoch": 0.6656315789473685, + "grad_norm": 0.5, + "learning_rate": 7.430409356725147e-05, + "loss": 3.469, + "step": 12647 + }, + { + "epoch": 0.6656842105263158, + "grad_norm": 0.51953125, + "learning_rate": 7.429239766081872e-05, + "loss": 3.1587, + "step": 12648 + }, + { + "epoch": 0.6657368421052632, + "grad_norm": 0.4921875, + "learning_rate": 7.428070175438598e-05, + "loss": 3.3483, + "step": 12649 + }, + { + "epoch": 0.6657894736842105, + "grad_norm": 0.494140625, + "learning_rate": 7.426900584795321e-05, + "loss": 3.243, + "step": 12650 + }, + { + "epoch": 0.6658421052631579, + "grad_norm": 0.53125, + "learning_rate": 7.425730994152046e-05, + "loss": 3.3861, + "step": 12651 + }, + { + "epoch": 0.6658947368421053, + "grad_norm": 0.46875, + "learning_rate": 7.424561403508771e-05, + "loss": 3.2592, + "step": 12652 + }, + { + "epoch": 0.6659473684210526, + "grad_norm": 0.5078125, + "learning_rate": 7.423391812865498e-05, + "loss": 3.3396, + "step": 12653 + }, + { + "epoch": 0.666, + "grad_norm": 0.54296875, + "learning_rate": 7.422222222222223e-05, + "loss": 3.2407, + "step": 12654 + }, + { + "epoch": 0.6660526315789473, + "grad_norm": 0.48828125, + "learning_rate": 7.421052631578948e-05, + "loss": 3.714, + "step": 12655 + }, + { + "epoch": 0.6661052631578948, + "grad_norm": 0.50390625, + "learning_rate": 7.419883040935673e-05, + "loss": 3.2366, + "step": 12656 + }, + { + "epoch": 0.6661578947368421, + "grad_norm": 0.494140625, + "learning_rate": 7.418713450292399e-05, + "loss": 3.5732, + "step": 12657 + }, + { + "epoch": 0.6662105263157895, + "grad_norm": 0.458984375, + "learning_rate": 7.417543859649123e-05, + "loss": 3.1479, + "step": 12658 + }, + { + "epoch": 0.6662631578947369, + "grad_norm": 0.46484375, + "learning_rate": 7.416374269005847e-05, + "loss": 3.6091, + "step": 12659 + }, + { + "epoch": 0.6663157894736842, + "grad_norm": 0.46484375, + "learning_rate": 7.415204678362574e-05, + "loss": 3.2258, + "step": 12660 + }, + { + "epoch": 0.6663684210526316, + "grad_norm": 0.490234375, + "learning_rate": 7.414035087719299e-05, + "loss": 3.9471, + "step": 12661 + }, + { + "epoch": 0.6664210526315789, + "grad_norm": 0.490234375, + "learning_rate": 7.412865497076024e-05, + "loss": 3.5387, + "step": 12662 + }, + { + "epoch": 0.6664736842105263, + "grad_norm": 0.46875, + "learning_rate": 7.411695906432749e-05, + "loss": 2.9662, + "step": 12663 + }, + { + "epoch": 0.6665263157894736, + "grad_norm": 0.462890625, + "learning_rate": 7.410526315789475e-05, + "loss": 3.3112, + "step": 12664 + }, + { + "epoch": 0.666578947368421, + "grad_norm": 0.5, + "learning_rate": 7.4093567251462e-05, + "loss": 3.2295, + "step": 12665 + }, + { + "epoch": 0.6666315789473685, + "grad_norm": 0.46484375, + "learning_rate": 7.408187134502924e-05, + "loss": 3.2159, + "step": 12666 + }, + { + "epoch": 0.6666842105263158, + "grad_norm": 0.5, + "learning_rate": 7.407017543859649e-05, + "loss": 2.8897, + "step": 12667 + }, + { + "epoch": 0.6667368421052632, + "grad_norm": 0.5546875, + "learning_rate": 7.405847953216375e-05, + "loss": 3.5359, + "step": 12668 + }, + { + "epoch": 0.6667894736842105, + "grad_norm": 0.478515625, + "learning_rate": 7.4046783625731e-05, + "loss": 3.6013, + "step": 12669 + }, + { + "epoch": 0.6668421052631579, + "grad_norm": 0.48828125, + "learning_rate": 7.403508771929825e-05, + "loss": 3.43, + "step": 12670 + }, + { + "epoch": 0.6668947368421053, + "grad_norm": 0.484375, + "learning_rate": 7.402339181286551e-05, + "loss": 2.8138, + "step": 12671 + }, + { + "epoch": 0.6669473684210526, + "grad_norm": 0.48046875, + "learning_rate": 7.401169590643276e-05, + "loss": 3.3797, + "step": 12672 + }, + { + "epoch": 0.667, + "grad_norm": 0.50390625, + "learning_rate": 7.4e-05, + "loss": 3.246, + "step": 12673 + }, + { + "epoch": 0.6670526315789473, + "grad_norm": 0.51171875, + "learning_rate": 7.398830409356725e-05, + "loss": 3.6707, + "step": 12674 + }, + { + "epoch": 0.6671052631578948, + "grad_norm": 0.5390625, + "learning_rate": 7.397660818713451e-05, + "loss": 3.1513, + "step": 12675 + }, + { + "epoch": 0.6671578947368421, + "grad_norm": 0.56640625, + "learning_rate": 7.396491228070176e-05, + "loss": 3.1147, + "step": 12676 + }, + { + "epoch": 0.6672105263157895, + "grad_norm": 0.49609375, + "learning_rate": 7.395321637426901e-05, + "loss": 3.4553, + "step": 12677 + }, + { + "epoch": 0.6672631578947369, + "grad_norm": 0.5, + "learning_rate": 7.394152046783626e-05, + "loss": 3.472, + "step": 12678 + }, + { + "epoch": 0.6673157894736842, + "grad_norm": 0.53515625, + "learning_rate": 7.392982456140352e-05, + "loss": 3.3146, + "step": 12679 + }, + { + "epoch": 0.6673684210526316, + "grad_norm": 0.4921875, + "learning_rate": 7.391812865497077e-05, + "loss": 3.5101, + "step": 12680 + }, + { + "epoch": 0.6674210526315789, + "grad_norm": 0.490234375, + "learning_rate": 7.3906432748538e-05, + "loss": 3.3434, + "step": 12681 + }, + { + "epoch": 0.6674736842105263, + "grad_norm": 0.482421875, + "learning_rate": 7.389473684210527e-05, + "loss": 3.3256, + "step": 12682 + }, + { + "epoch": 0.6675263157894736, + "grad_norm": 0.51953125, + "learning_rate": 7.388304093567252e-05, + "loss": 3.3314, + "step": 12683 + }, + { + "epoch": 0.667578947368421, + "grad_norm": 0.62890625, + "learning_rate": 7.387134502923977e-05, + "loss": 3.0241, + "step": 12684 + }, + { + "epoch": 0.6676315789473685, + "grad_norm": 0.5, + "learning_rate": 7.385964912280702e-05, + "loss": 3.2004, + "step": 12685 + }, + { + "epoch": 0.6676842105263158, + "grad_norm": 0.470703125, + "learning_rate": 7.384795321637428e-05, + "loss": 3.2216, + "step": 12686 + }, + { + "epoch": 0.6677368421052632, + "grad_norm": 0.455078125, + "learning_rate": 7.383625730994153e-05, + "loss": 3.336, + "step": 12687 + }, + { + "epoch": 0.6677894736842105, + "grad_norm": 0.46484375, + "learning_rate": 7.382456140350878e-05, + "loss": 3.2255, + "step": 12688 + }, + { + "epoch": 0.6678421052631579, + "grad_norm": 0.458984375, + "learning_rate": 7.381286549707602e-05, + "loss": 3.237, + "step": 12689 + }, + { + "epoch": 0.6678947368421052, + "grad_norm": 0.49609375, + "learning_rate": 7.380116959064328e-05, + "loss": 2.9539, + "step": 12690 + }, + { + "epoch": 0.6679473684210526, + "grad_norm": 0.50390625, + "learning_rate": 7.378947368421053e-05, + "loss": 3.1021, + "step": 12691 + }, + { + "epoch": 0.668, + "grad_norm": 0.58984375, + "learning_rate": 7.377777777777778e-05, + "loss": 3.2455, + "step": 12692 + }, + { + "epoch": 0.6680526315789473, + "grad_norm": 0.5, + "learning_rate": 7.376608187134503e-05, + "loss": 3.3582, + "step": 12693 + }, + { + "epoch": 0.6681052631578948, + "grad_norm": 0.50390625, + "learning_rate": 7.375438596491229e-05, + "loss": 3.4043, + "step": 12694 + }, + { + "epoch": 0.6681578947368421, + "grad_norm": 0.5, + "learning_rate": 7.374269005847954e-05, + "loss": 3.3283, + "step": 12695 + }, + { + "epoch": 0.6682105263157895, + "grad_norm": 0.48828125, + "learning_rate": 7.373099415204678e-05, + "loss": 2.8146, + "step": 12696 + }, + { + "epoch": 0.6682631578947369, + "grad_norm": 0.482421875, + "learning_rate": 7.371929824561404e-05, + "loss": 2.883, + "step": 12697 + }, + { + "epoch": 0.6683157894736842, + "grad_norm": 0.515625, + "learning_rate": 7.370760233918129e-05, + "loss": 3.2582, + "step": 12698 + }, + { + "epoch": 0.6683684210526316, + "grad_norm": 0.482421875, + "learning_rate": 7.369590643274854e-05, + "loss": 3.3429, + "step": 12699 + }, + { + "epoch": 0.6684210526315789, + "grad_norm": 0.47265625, + "learning_rate": 7.368421052631579e-05, + "loss": 3.6266, + "step": 12700 + }, + { + "epoch": 0.6684736842105263, + "grad_norm": 0.48046875, + "learning_rate": 7.367251461988305e-05, + "loss": 3.4725, + "step": 12701 + }, + { + "epoch": 0.6685263157894736, + "grad_norm": 0.5078125, + "learning_rate": 7.36608187134503e-05, + "loss": 3.7941, + "step": 12702 + }, + { + "epoch": 0.668578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.364912280701755e-05, + "loss": 3.3975, + "step": 12703 + }, + { + "epoch": 0.6686315789473685, + "grad_norm": 0.482421875, + "learning_rate": 7.363742690058479e-05, + "loss": 3.5065, + "step": 12704 + }, + { + "epoch": 0.6686842105263158, + "grad_norm": 0.49609375, + "learning_rate": 7.362573099415205e-05, + "loss": 3.1635, + "step": 12705 + }, + { + "epoch": 0.6687368421052632, + "grad_norm": 0.474609375, + "learning_rate": 7.36140350877193e-05, + "loss": 3.5302, + "step": 12706 + }, + { + "epoch": 0.6687894736842105, + "grad_norm": 0.5078125, + "learning_rate": 7.360233918128655e-05, + "loss": 3.4272, + "step": 12707 + }, + { + "epoch": 0.6688421052631579, + "grad_norm": 0.490234375, + "learning_rate": 7.359064327485381e-05, + "loss": 3.4009, + "step": 12708 + }, + { + "epoch": 0.6688947368421052, + "grad_norm": 0.5, + "learning_rate": 7.357894736842106e-05, + "loss": 3.5481, + "step": 12709 + }, + { + "epoch": 0.6689473684210526, + "grad_norm": 0.515625, + "learning_rate": 7.356725146198831e-05, + "loss": 3.0758, + "step": 12710 + }, + { + "epoch": 0.669, + "grad_norm": 0.52734375, + "learning_rate": 7.355555555555556e-05, + "loss": 3.3853, + "step": 12711 + }, + { + "epoch": 0.6690526315789473, + "grad_norm": 0.62890625, + "learning_rate": 7.354385964912281e-05, + "loss": 3.3399, + "step": 12712 + }, + { + "epoch": 0.6691052631578948, + "grad_norm": 0.48046875, + "learning_rate": 7.353216374269006e-05, + "loss": 3.2512, + "step": 12713 + }, + { + "epoch": 0.6691578947368421, + "grad_norm": 0.462890625, + "learning_rate": 7.352046783625731e-05, + "loss": 3.1678, + "step": 12714 + }, + { + "epoch": 0.6692105263157895, + "grad_norm": 0.48046875, + "learning_rate": 7.350877192982456e-05, + "loss": 3.6756, + "step": 12715 + }, + { + "epoch": 0.6692631578947369, + "grad_norm": 0.49609375, + "learning_rate": 7.349707602339182e-05, + "loss": 3.4532, + "step": 12716 + }, + { + "epoch": 0.6693157894736842, + "grad_norm": 0.474609375, + "learning_rate": 7.348538011695907e-05, + "loss": 3.2422, + "step": 12717 + }, + { + "epoch": 0.6693684210526316, + "grad_norm": 0.44921875, + "learning_rate": 7.347368421052632e-05, + "loss": 3.2639, + "step": 12718 + }, + { + "epoch": 0.6694210526315789, + "grad_norm": 0.5234375, + "learning_rate": 7.346198830409356e-05, + "loss": 3.2466, + "step": 12719 + }, + { + "epoch": 0.6694736842105263, + "grad_norm": 0.478515625, + "learning_rate": 7.345029239766082e-05, + "loss": 2.7633, + "step": 12720 + }, + { + "epoch": 0.6695263157894736, + "grad_norm": 0.486328125, + "learning_rate": 7.343859649122807e-05, + "loss": 3.3951, + "step": 12721 + }, + { + "epoch": 0.669578947368421, + "grad_norm": 0.51953125, + "learning_rate": 7.342690058479532e-05, + "loss": 3.464, + "step": 12722 + }, + { + "epoch": 0.6696315789473685, + "grad_norm": 0.53125, + "learning_rate": 7.341520467836258e-05, + "loss": 3.4836, + "step": 12723 + }, + { + "epoch": 0.6696842105263158, + "grad_norm": 0.55859375, + "learning_rate": 7.340350877192983e-05, + "loss": 2.8666, + "step": 12724 + }, + { + "epoch": 0.6697368421052632, + "grad_norm": 0.49609375, + "learning_rate": 7.339181286549708e-05, + "loss": 3.5983, + "step": 12725 + }, + { + "epoch": 0.6697894736842105, + "grad_norm": 0.53515625, + "learning_rate": 7.338011695906433e-05, + "loss": 2.9994, + "step": 12726 + }, + { + "epoch": 0.6698421052631579, + "grad_norm": 0.5, + "learning_rate": 7.336842105263158e-05, + "loss": 3.5433, + "step": 12727 + }, + { + "epoch": 0.6698947368421052, + "grad_norm": 0.478515625, + "learning_rate": 7.335672514619883e-05, + "loss": 3.5038, + "step": 12728 + }, + { + "epoch": 0.6699473684210526, + "grad_norm": 0.50390625, + "learning_rate": 7.334502923976608e-05, + "loss": 3.2102, + "step": 12729 + }, + { + "epoch": 0.67, + "grad_norm": 0.48046875, + "learning_rate": 7.333333333333333e-05, + "loss": 3.3286, + "step": 12730 + }, + { + "epoch": 0.6700526315789473, + "grad_norm": 0.486328125, + "learning_rate": 7.33216374269006e-05, + "loss": 3.3087, + "step": 12731 + }, + { + "epoch": 0.6701052631578948, + "grad_norm": 0.51171875, + "learning_rate": 7.330994152046784e-05, + "loss": 3.16, + "step": 12732 + }, + { + "epoch": 0.6701578947368421, + "grad_norm": 0.4921875, + "learning_rate": 7.32982456140351e-05, + "loss": 3.3686, + "step": 12733 + }, + { + "epoch": 0.6702105263157895, + "grad_norm": 0.474609375, + "learning_rate": 7.328654970760234e-05, + "loss": 3.3514, + "step": 12734 + }, + { + "epoch": 0.6702631578947369, + "grad_norm": 0.50390625, + "learning_rate": 7.327485380116959e-05, + "loss": 3.0256, + "step": 12735 + }, + { + "epoch": 0.6703157894736842, + "grad_norm": 0.484375, + "learning_rate": 7.326315789473684e-05, + "loss": 3.3494, + "step": 12736 + }, + { + "epoch": 0.6703684210526316, + "grad_norm": 0.5234375, + "learning_rate": 7.325146198830409e-05, + "loss": 3.1839, + "step": 12737 + }, + { + "epoch": 0.6704210526315789, + "grad_norm": 0.50390625, + "learning_rate": 7.323976608187136e-05, + "loss": 3.3009, + "step": 12738 + }, + { + "epoch": 0.6704736842105263, + "grad_norm": 0.53515625, + "learning_rate": 7.32280701754386e-05, + "loss": 2.6431, + "step": 12739 + }, + { + "epoch": 0.6705263157894736, + "grad_norm": 0.5, + "learning_rate": 7.321637426900585e-05, + "loss": 3.2416, + "step": 12740 + }, + { + "epoch": 0.670578947368421, + "grad_norm": 0.484375, + "learning_rate": 7.32046783625731e-05, + "loss": 3.4491, + "step": 12741 + }, + { + "epoch": 0.6706315789473685, + "grad_norm": 0.48828125, + "learning_rate": 7.319298245614035e-05, + "loss": 3.6098, + "step": 12742 + }, + { + "epoch": 0.6706842105263158, + "grad_norm": 0.5234375, + "learning_rate": 7.31812865497076e-05, + "loss": 3.2211, + "step": 12743 + }, + { + "epoch": 0.6707368421052632, + "grad_norm": 0.470703125, + "learning_rate": 7.316959064327485e-05, + "loss": 3.2675, + "step": 12744 + }, + { + "epoch": 0.6707894736842105, + "grad_norm": 0.453125, + "learning_rate": 7.315789473684212e-05, + "loss": 3.4306, + "step": 12745 + }, + { + "epoch": 0.6708421052631579, + "grad_norm": 0.486328125, + "learning_rate": 7.314619883040937e-05, + "loss": 3.3365, + "step": 12746 + }, + { + "epoch": 0.6708947368421052, + "grad_norm": 0.484375, + "learning_rate": 7.313450292397662e-05, + "loss": 3.1603, + "step": 12747 + }, + { + "epoch": 0.6709473684210526, + "grad_norm": 0.498046875, + "learning_rate": 7.312280701754386e-05, + "loss": 3.1631, + "step": 12748 + }, + { + "epoch": 0.671, + "grad_norm": 0.55859375, + "learning_rate": 7.311111111111111e-05, + "loss": 3.3287, + "step": 12749 + }, + { + "epoch": 0.6710526315789473, + "grad_norm": 0.498046875, + "learning_rate": 7.309941520467836e-05, + "loss": 3.1346, + "step": 12750 + }, + { + "epoch": 0.6711052631578948, + "grad_norm": 0.451171875, + "learning_rate": 7.308771929824561e-05, + "loss": 3.0293, + "step": 12751 + }, + { + "epoch": 0.6711578947368421, + "grad_norm": 0.53125, + "learning_rate": 7.307602339181286e-05, + "loss": 2.8511, + "step": 12752 + }, + { + "epoch": 0.6712105263157895, + "grad_norm": 0.484375, + "learning_rate": 7.306432748538013e-05, + "loss": 3.5884, + "step": 12753 + }, + { + "epoch": 0.6712631578947369, + "grad_norm": 0.470703125, + "learning_rate": 7.305263157894738e-05, + "loss": 3.1719, + "step": 12754 + }, + { + "epoch": 0.6713157894736842, + "grad_norm": 0.5078125, + "learning_rate": 7.304093567251463e-05, + "loss": 3.3147, + "step": 12755 + }, + { + "epoch": 0.6713684210526316, + "grad_norm": 0.498046875, + "learning_rate": 7.302923976608188e-05, + "loss": 3.312, + "step": 12756 + }, + { + "epoch": 0.6714210526315789, + "grad_norm": 0.51171875, + "learning_rate": 7.301754385964912e-05, + "loss": 3.4793, + "step": 12757 + }, + { + "epoch": 0.6714736842105263, + "grad_norm": 0.482421875, + "learning_rate": 7.300584795321637e-05, + "loss": 2.8198, + "step": 12758 + }, + { + "epoch": 0.6715263157894736, + "grad_norm": 0.53125, + "learning_rate": 7.299415204678362e-05, + "loss": 3.3589, + "step": 12759 + }, + { + "epoch": 0.671578947368421, + "grad_norm": 0.470703125, + "learning_rate": 7.298245614035089e-05, + "loss": 3.4695, + "step": 12760 + }, + { + "epoch": 0.6716315789473685, + "grad_norm": 0.5078125, + "learning_rate": 7.297076023391814e-05, + "loss": 3.4185, + "step": 12761 + }, + { + "epoch": 0.6716842105263158, + "grad_norm": 0.4921875, + "learning_rate": 7.295906432748539e-05, + "loss": 3.2611, + "step": 12762 + }, + { + "epoch": 0.6717368421052632, + "grad_norm": 0.5078125, + "learning_rate": 7.294736842105264e-05, + "loss": 3.3356, + "step": 12763 + }, + { + "epoch": 0.6717894736842105, + "grad_norm": 0.4765625, + "learning_rate": 7.293567251461989e-05, + "loss": 3.2242, + "step": 12764 + }, + { + "epoch": 0.6718421052631579, + "grad_norm": 0.474609375, + "learning_rate": 7.292397660818713e-05, + "loss": 3.2798, + "step": 12765 + }, + { + "epoch": 0.6718947368421052, + "grad_norm": 0.498046875, + "learning_rate": 7.291228070175438e-05, + "loss": 3.1599, + "step": 12766 + }, + { + "epoch": 0.6719473684210526, + "grad_norm": 0.52734375, + "learning_rate": 7.290058479532163e-05, + "loss": 3.3222, + "step": 12767 + }, + { + "epoch": 0.672, + "grad_norm": 0.51953125, + "learning_rate": 7.28888888888889e-05, + "loss": 3.3767, + "step": 12768 + }, + { + "epoch": 0.6720526315789473, + "grad_norm": 0.5234375, + "learning_rate": 7.287719298245615e-05, + "loss": 3.5177, + "step": 12769 + }, + { + "epoch": 0.6721052631578948, + "grad_norm": 0.52734375, + "learning_rate": 7.28654970760234e-05, + "loss": 3.4382, + "step": 12770 + }, + { + "epoch": 0.6721578947368421, + "grad_norm": 0.546875, + "learning_rate": 7.285380116959065e-05, + "loss": 3.2944, + "step": 12771 + }, + { + "epoch": 0.6722105263157895, + "grad_norm": 0.498046875, + "learning_rate": 7.28421052631579e-05, + "loss": 3.3857, + "step": 12772 + }, + { + "epoch": 0.6722631578947369, + "grad_norm": 0.490234375, + "learning_rate": 7.283040935672515e-05, + "loss": 3.5447, + "step": 12773 + }, + { + "epoch": 0.6723157894736842, + "grad_norm": 0.490234375, + "learning_rate": 7.28187134502924e-05, + "loss": 3.2296, + "step": 12774 + }, + { + "epoch": 0.6723684210526316, + "grad_norm": 0.470703125, + "learning_rate": 7.280701754385966e-05, + "loss": 2.887, + "step": 12775 + }, + { + "epoch": 0.6724210526315789, + "grad_norm": 0.51953125, + "learning_rate": 7.279532163742691e-05, + "loss": 3.9587, + "step": 12776 + }, + { + "epoch": 0.6724736842105263, + "grad_norm": 0.51171875, + "learning_rate": 7.278362573099416e-05, + "loss": 3.5538, + "step": 12777 + }, + { + "epoch": 0.6725263157894736, + "grad_norm": 0.470703125, + "learning_rate": 7.27719298245614e-05, + "loss": 3.0928, + "step": 12778 + }, + { + "epoch": 0.672578947368421, + "grad_norm": 0.48046875, + "learning_rate": 7.276023391812866e-05, + "loss": 3.2506, + "step": 12779 + }, + { + "epoch": 0.6726315789473685, + "grad_norm": 0.49609375, + "learning_rate": 7.27485380116959e-05, + "loss": 3.2152, + "step": 12780 + }, + { + "epoch": 0.6726842105263158, + "grad_norm": 0.4609375, + "learning_rate": 7.273684210526316e-05, + "loss": 3.0237, + "step": 12781 + }, + { + "epoch": 0.6727368421052632, + "grad_norm": 0.474609375, + "learning_rate": 7.272514619883042e-05, + "loss": 3.5013, + "step": 12782 + }, + { + "epoch": 0.6727894736842105, + "grad_norm": 0.5078125, + "learning_rate": 7.271345029239767e-05, + "loss": 3.3754, + "step": 12783 + }, + { + "epoch": 0.6728421052631579, + "grad_norm": 0.5234375, + "learning_rate": 7.270175438596492e-05, + "loss": 3.1265, + "step": 12784 + }, + { + "epoch": 0.6728947368421052, + "grad_norm": 0.515625, + "learning_rate": 7.269005847953217e-05, + "loss": 3.4756, + "step": 12785 + }, + { + "epoch": 0.6729473684210526, + "grad_norm": 0.49609375, + "learning_rate": 7.267836257309942e-05, + "loss": 3.3153, + "step": 12786 + }, + { + "epoch": 0.673, + "grad_norm": 0.48046875, + "learning_rate": 7.266666666666667e-05, + "loss": 3.2577, + "step": 12787 + }, + { + "epoch": 0.6730526315789473, + "grad_norm": 0.5, + "learning_rate": 7.265497076023392e-05, + "loss": 3.2699, + "step": 12788 + }, + { + "epoch": 0.6731052631578948, + "grad_norm": 0.51171875, + "learning_rate": 7.264327485380117e-05, + "loss": 3.561, + "step": 12789 + }, + { + "epoch": 0.6731578947368421, + "grad_norm": 0.5, + "learning_rate": 7.263157894736843e-05, + "loss": 3.8303, + "step": 12790 + }, + { + "epoch": 0.6732105263157895, + "grad_norm": 0.494140625, + "learning_rate": 7.261988304093568e-05, + "loss": 3.184, + "step": 12791 + }, + { + "epoch": 0.6732631578947369, + "grad_norm": 0.4921875, + "learning_rate": 7.260818713450293e-05, + "loss": 3.6184, + "step": 12792 + }, + { + "epoch": 0.6733157894736842, + "grad_norm": 0.515625, + "learning_rate": 7.259649122807018e-05, + "loss": 3.3404, + "step": 12793 + }, + { + "epoch": 0.6733684210526316, + "grad_norm": 0.4921875, + "learning_rate": 7.258479532163743e-05, + "loss": 3.9299, + "step": 12794 + }, + { + "epoch": 0.6734210526315789, + "grad_norm": 0.48828125, + "learning_rate": 7.257309941520468e-05, + "loss": 3.3883, + "step": 12795 + }, + { + "epoch": 0.6734736842105263, + "grad_norm": 0.5078125, + "learning_rate": 7.256140350877193e-05, + "loss": 3.4991, + "step": 12796 + }, + { + "epoch": 0.6735263157894736, + "grad_norm": 0.5078125, + "learning_rate": 7.254970760233919e-05, + "loss": 3.3485, + "step": 12797 + }, + { + "epoch": 0.673578947368421, + "grad_norm": 0.47265625, + "learning_rate": 7.253801169590644e-05, + "loss": 3.6624, + "step": 12798 + }, + { + "epoch": 0.6736315789473685, + "grad_norm": 0.494140625, + "learning_rate": 7.252631578947369e-05, + "loss": 3.6603, + "step": 12799 + }, + { + "epoch": 0.6736842105263158, + "grad_norm": 0.4921875, + "learning_rate": 7.251461988304094e-05, + "loss": 3.6557, + "step": 12800 + }, + { + "epoch": 0.6737368421052632, + "grad_norm": 0.5, + "learning_rate": 7.250292397660819e-05, + "loss": 3.4935, + "step": 12801 + }, + { + "epoch": 0.6737894736842105, + "grad_norm": 0.5, + "learning_rate": 7.249122807017544e-05, + "loss": 3.3916, + "step": 12802 + }, + { + "epoch": 0.6738421052631579, + "grad_norm": 0.482421875, + "learning_rate": 7.247953216374269e-05, + "loss": 2.7422, + "step": 12803 + }, + { + "epoch": 0.6738947368421052, + "grad_norm": 0.49609375, + "learning_rate": 7.246783625730994e-05, + "loss": 3.3547, + "step": 12804 + }, + { + "epoch": 0.6739473684210526, + "grad_norm": 0.466796875, + "learning_rate": 7.24561403508772e-05, + "loss": 3.3455, + "step": 12805 + }, + { + "epoch": 0.674, + "grad_norm": 0.51171875, + "learning_rate": 7.244444444444445e-05, + "loss": 3.5969, + "step": 12806 + }, + { + "epoch": 0.6740526315789473, + "grad_norm": 0.5, + "learning_rate": 7.24327485380117e-05, + "loss": 3.3316, + "step": 12807 + }, + { + "epoch": 0.6741052631578948, + "grad_norm": 0.5, + "learning_rate": 7.242105263157896e-05, + "loss": 2.9588, + "step": 12808 + }, + { + "epoch": 0.6741578947368421, + "grad_norm": 0.4765625, + "learning_rate": 7.24093567251462e-05, + "loss": 3.4562, + "step": 12809 + }, + { + "epoch": 0.6742105263157895, + "grad_norm": 0.54296875, + "learning_rate": 7.239766081871345e-05, + "loss": 3.3448, + "step": 12810 + }, + { + "epoch": 0.6742631578947369, + "grad_norm": 0.484375, + "learning_rate": 7.23859649122807e-05, + "loss": 3.6332, + "step": 12811 + }, + { + "epoch": 0.6743157894736842, + "grad_norm": 0.515625, + "learning_rate": 7.237426900584796e-05, + "loss": 3.1553, + "step": 12812 + }, + { + "epoch": 0.6743684210526316, + "grad_norm": 0.486328125, + "learning_rate": 7.236257309941521e-05, + "loss": 3.7436, + "step": 12813 + }, + { + "epoch": 0.6744210526315789, + "grad_norm": 0.498046875, + "learning_rate": 7.235087719298246e-05, + "loss": 3.2121, + "step": 12814 + }, + { + "epoch": 0.6744736842105263, + "grad_norm": 0.478515625, + "learning_rate": 7.233918128654971e-05, + "loss": 3.219, + "step": 12815 + }, + { + "epoch": 0.6745263157894736, + "grad_norm": 0.470703125, + "learning_rate": 7.232748538011697e-05, + "loss": 3.3046, + "step": 12816 + }, + { + "epoch": 0.6745789473684211, + "grad_norm": 0.474609375, + "learning_rate": 7.231578947368421e-05, + "loss": 3.2673, + "step": 12817 + }, + { + "epoch": 0.6746315789473685, + "grad_norm": 0.5, + "learning_rate": 7.230409356725146e-05, + "loss": 2.7301, + "step": 12818 + }, + { + "epoch": 0.6746842105263158, + "grad_norm": 0.455078125, + "learning_rate": 7.229239766081871e-05, + "loss": 3.2008, + "step": 12819 + }, + { + "epoch": 0.6747368421052632, + "grad_norm": 0.4609375, + "learning_rate": 7.228070175438597e-05, + "loss": 3.4272, + "step": 12820 + }, + { + "epoch": 0.6747894736842105, + "grad_norm": 0.45703125, + "learning_rate": 7.226900584795322e-05, + "loss": 2.838, + "step": 12821 + }, + { + "epoch": 0.6748421052631579, + "grad_norm": 0.46484375, + "learning_rate": 7.225730994152047e-05, + "loss": 3.5067, + "step": 12822 + }, + { + "epoch": 0.6748947368421052, + "grad_norm": 0.46875, + "learning_rate": 7.224561403508773e-05, + "loss": 3.2424, + "step": 12823 + }, + { + "epoch": 0.6749473684210526, + "grad_norm": 0.50390625, + "learning_rate": 7.223391812865497e-05, + "loss": 3.1435, + "step": 12824 + }, + { + "epoch": 0.675, + "grad_norm": 0.48046875, + "learning_rate": 7.222222222222222e-05, + "loss": 3.3358, + "step": 12825 + }, + { + "epoch": 0.6750526315789473, + "grad_norm": 0.7421875, + "learning_rate": 7.221052631578947e-05, + "loss": 3.5299, + "step": 12826 + }, + { + "epoch": 0.6751052631578948, + "grad_norm": 0.5078125, + "learning_rate": 7.219883040935673e-05, + "loss": 3.3657, + "step": 12827 + }, + { + "epoch": 0.6751578947368421, + "grad_norm": 0.494140625, + "learning_rate": 7.218713450292398e-05, + "loss": 3.2899, + "step": 12828 + }, + { + "epoch": 0.6752105263157895, + "grad_norm": 0.466796875, + "learning_rate": 7.217543859649123e-05, + "loss": 3.5137, + "step": 12829 + }, + { + "epoch": 0.6752631578947368, + "grad_norm": 0.52734375, + "learning_rate": 7.216374269005848e-05, + "loss": 3.3179, + "step": 12830 + }, + { + "epoch": 0.6753157894736842, + "grad_norm": 0.61328125, + "learning_rate": 7.215204678362574e-05, + "loss": 3.139, + "step": 12831 + }, + { + "epoch": 0.6753684210526316, + "grad_norm": 0.4765625, + "learning_rate": 7.214035087719298e-05, + "loss": 2.9755, + "step": 12832 + }, + { + "epoch": 0.6754210526315789, + "grad_norm": 0.54296875, + "learning_rate": 7.212865497076023e-05, + "loss": 3.096, + "step": 12833 + }, + { + "epoch": 0.6754736842105263, + "grad_norm": 0.50390625, + "learning_rate": 7.211695906432749e-05, + "loss": 3.381, + "step": 12834 + }, + { + "epoch": 0.6755263157894736, + "grad_norm": 0.51171875, + "learning_rate": 7.210526315789474e-05, + "loss": 3.6983, + "step": 12835 + }, + { + "epoch": 0.6755789473684211, + "grad_norm": 0.490234375, + "learning_rate": 7.209356725146199e-05, + "loss": 3.4566, + "step": 12836 + }, + { + "epoch": 0.6756315789473685, + "grad_norm": 0.5, + "learning_rate": 7.208187134502924e-05, + "loss": 3.2515, + "step": 12837 + }, + { + "epoch": 0.6756842105263158, + "grad_norm": 0.482421875, + "learning_rate": 7.20701754385965e-05, + "loss": 3.5965, + "step": 12838 + }, + { + "epoch": 0.6757368421052632, + "grad_norm": 0.51953125, + "learning_rate": 7.205847953216375e-05, + "loss": 3.823, + "step": 12839 + }, + { + "epoch": 0.6757894736842105, + "grad_norm": 0.50390625, + "learning_rate": 7.204678362573099e-05, + "loss": 3.2594, + "step": 12840 + }, + { + "epoch": 0.6758421052631579, + "grad_norm": 0.470703125, + "learning_rate": 7.203508771929824e-05, + "loss": 2.8082, + "step": 12841 + }, + { + "epoch": 0.6758947368421052, + "grad_norm": 0.498046875, + "learning_rate": 7.20233918128655e-05, + "loss": 3.0428, + "step": 12842 + }, + { + "epoch": 0.6759473684210526, + "grad_norm": 0.498046875, + "learning_rate": 7.201169590643275e-05, + "loss": 3.0478, + "step": 12843 + }, + { + "epoch": 0.676, + "grad_norm": 0.466796875, + "learning_rate": 7.2e-05, + "loss": 3.6195, + "step": 12844 + }, + { + "epoch": 0.6760526315789473, + "grad_norm": 0.5, + "learning_rate": 7.198830409356726e-05, + "loss": 3.3493, + "step": 12845 + }, + { + "epoch": 0.6761052631578948, + "grad_norm": 0.515625, + "learning_rate": 7.197660818713451e-05, + "loss": 3.4248, + "step": 12846 + }, + { + "epoch": 0.6761578947368421, + "grad_norm": 0.515625, + "learning_rate": 7.196491228070175e-05, + "loss": 3.3497, + "step": 12847 + }, + { + "epoch": 0.6762105263157895, + "grad_norm": 0.50390625, + "learning_rate": 7.1953216374269e-05, + "loss": 3.4527, + "step": 12848 + }, + { + "epoch": 0.6762631578947368, + "grad_norm": 0.47265625, + "learning_rate": 7.194152046783626e-05, + "loss": 3.4772, + "step": 12849 + }, + { + "epoch": 0.6763157894736842, + "grad_norm": 0.48046875, + "learning_rate": 7.192982456140351e-05, + "loss": 3.2572, + "step": 12850 + }, + { + "epoch": 0.6763684210526316, + "grad_norm": 0.478515625, + "learning_rate": 7.191812865497076e-05, + "loss": 3.333, + "step": 12851 + }, + { + "epoch": 0.6764210526315789, + "grad_norm": 0.46875, + "learning_rate": 7.190643274853801e-05, + "loss": 3.399, + "step": 12852 + }, + { + "epoch": 0.6764736842105263, + "grad_norm": 0.515625, + "learning_rate": 7.189473684210528e-05, + "loss": 3.7277, + "step": 12853 + }, + { + "epoch": 0.6765263157894736, + "grad_norm": 0.59375, + "learning_rate": 7.188304093567252e-05, + "loss": 2.7089, + "step": 12854 + }, + { + "epoch": 0.6765789473684211, + "grad_norm": 0.482421875, + "learning_rate": 7.187134502923976e-05, + "loss": 3.3794, + "step": 12855 + }, + { + "epoch": 0.6766315789473685, + "grad_norm": 0.470703125, + "learning_rate": 7.185964912280701e-05, + "loss": 3.4369, + "step": 12856 + }, + { + "epoch": 0.6766842105263158, + "grad_norm": 0.470703125, + "learning_rate": 7.184795321637427e-05, + "loss": 3.3297, + "step": 12857 + }, + { + "epoch": 0.6767368421052632, + "grad_norm": 0.47265625, + "learning_rate": 7.183625730994152e-05, + "loss": 3.5109, + "step": 12858 + }, + { + "epoch": 0.6767894736842105, + "grad_norm": 0.482421875, + "learning_rate": 7.182456140350877e-05, + "loss": 3.031, + "step": 12859 + }, + { + "epoch": 0.6768421052631579, + "grad_norm": 0.44921875, + "learning_rate": 7.181286549707604e-05, + "loss": 3.3383, + "step": 12860 + }, + { + "epoch": 0.6768947368421052, + "grad_norm": 0.640625, + "learning_rate": 7.180116959064329e-05, + "loss": 3.3121, + "step": 12861 + }, + { + "epoch": 0.6769473684210526, + "grad_norm": 0.498046875, + "learning_rate": 7.178947368421054e-05, + "loss": 3.3708, + "step": 12862 + }, + { + "epoch": 0.677, + "grad_norm": 0.46875, + "learning_rate": 7.177777777777777e-05, + "loss": 3.2861, + "step": 12863 + }, + { + "epoch": 0.6770526315789474, + "grad_norm": 0.51953125, + "learning_rate": 7.176608187134503e-05, + "loss": 3.5949, + "step": 12864 + }, + { + "epoch": 0.6771052631578948, + "grad_norm": 0.462890625, + "learning_rate": 7.175438596491228e-05, + "loss": 3.4448, + "step": 12865 + }, + { + "epoch": 0.6771578947368421, + "grad_norm": 0.458984375, + "learning_rate": 7.174269005847953e-05, + "loss": 3.0608, + "step": 12866 + }, + { + "epoch": 0.6772105263157895, + "grad_norm": 0.474609375, + "learning_rate": 7.173099415204678e-05, + "loss": 3.3513, + "step": 12867 + }, + { + "epoch": 0.6772631578947368, + "grad_norm": 0.4921875, + "learning_rate": 7.171929824561405e-05, + "loss": 3.2823, + "step": 12868 + }, + { + "epoch": 0.6773157894736842, + "grad_norm": 0.45703125, + "learning_rate": 7.17076023391813e-05, + "loss": 3.5619, + "step": 12869 + }, + { + "epoch": 0.6773684210526316, + "grad_norm": 0.48828125, + "learning_rate": 7.169590643274853e-05, + "loss": 3.148, + "step": 12870 + }, + { + "epoch": 0.6774210526315789, + "grad_norm": 0.54296875, + "learning_rate": 7.16842105263158e-05, + "loss": 3.4798, + "step": 12871 + }, + { + "epoch": 0.6774736842105263, + "grad_norm": 0.51171875, + "learning_rate": 7.167251461988304e-05, + "loss": 3.0982, + "step": 12872 + }, + { + "epoch": 0.6775263157894736, + "grad_norm": 0.478515625, + "learning_rate": 7.16608187134503e-05, + "loss": 3.467, + "step": 12873 + }, + { + "epoch": 0.6775789473684211, + "grad_norm": 0.494140625, + "learning_rate": 7.164912280701754e-05, + "loss": 2.9771, + "step": 12874 + }, + { + "epoch": 0.6776315789473685, + "grad_norm": 0.494140625, + "learning_rate": 7.163742690058481e-05, + "loss": 3.2861, + "step": 12875 + }, + { + "epoch": 0.6776842105263158, + "grad_norm": 0.5234375, + "learning_rate": 7.162573099415206e-05, + "loss": 3.0834, + "step": 12876 + }, + { + "epoch": 0.6777368421052632, + "grad_norm": 0.498046875, + "learning_rate": 7.16140350877193e-05, + "loss": 3.5779, + "step": 12877 + }, + { + "epoch": 0.6777894736842105, + "grad_norm": 0.478515625, + "learning_rate": 7.160233918128654e-05, + "loss": 3.0246, + "step": 12878 + }, + { + "epoch": 0.6778421052631579, + "grad_norm": 0.5078125, + "learning_rate": 7.15906432748538e-05, + "loss": 3.0118, + "step": 12879 + }, + { + "epoch": 0.6778947368421052, + "grad_norm": 0.51171875, + "learning_rate": 7.157894736842105e-05, + "loss": 3.4223, + "step": 12880 + }, + { + "epoch": 0.6779473684210526, + "grad_norm": 0.486328125, + "learning_rate": 7.15672514619883e-05, + "loss": 3.4353, + "step": 12881 + }, + { + "epoch": 0.678, + "grad_norm": 0.5078125, + "learning_rate": 7.155555555555555e-05, + "loss": 3.2307, + "step": 12882 + }, + { + "epoch": 0.6780526315789474, + "grad_norm": 0.48046875, + "learning_rate": 7.154385964912282e-05, + "loss": 3.7056, + "step": 12883 + }, + { + "epoch": 0.6781052631578948, + "grad_norm": 0.4765625, + "learning_rate": 7.153216374269007e-05, + "loss": 3.4291, + "step": 12884 + }, + { + "epoch": 0.6781578947368421, + "grad_norm": 0.474609375, + "learning_rate": 7.152046783625732e-05, + "loss": 3.4585, + "step": 12885 + }, + { + "epoch": 0.6782105263157895, + "grad_norm": 0.57421875, + "learning_rate": 7.150877192982457e-05, + "loss": 3.5953, + "step": 12886 + }, + { + "epoch": 0.6782631578947368, + "grad_norm": 0.482421875, + "learning_rate": 7.149707602339182e-05, + "loss": 3.0594, + "step": 12887 + }, + { + "epoch": 0.6783157894736842, + "grad_norm": 0.49609375, + "learning_rate": 7.148538011695907e-05, + "loss": 3.3635, + "step": 12888 + }, + { + "epoch": 0.6783684210526316, + "grad_norm": 0.484375, + "learning_rate": 7.147368421052631e-05, + "loss": 3.5423, + "step": 12889 + }, + { + "epoch": 0.6784210526315789, + "grad_norm": 0.486328125, + "learning_rate": 7.146198830409358e-05, + "loss": 3.4905, + "step": 12890 + }, + { + "epoch": 0.6784736842105263, + "grad_norm": 0.48828125, + "learning_rate": 7.145029239766083e-05, + "loss": 3.3255, + "step": 12891 + }, + { + "epoch": 0.6785263157894736, + "grad_norm": 0.486328125, + "learning_rate": 7.143859649122808e-05, + "loss": 2.7478, + "step": 12892 + }, + { + "epoch": 0.6785789473684211, + "grad_norm": 0.4765625, + "learning_rate": 7.142690058479531e-05, + "loss": 3.449, + "step": 12893 + }, + { + "epoch": 0.6786315789473685, + "grad_norm": 0.87109375, + "learning_rate": 7.141520467836258e-05, + "loss": 3.373, + "step": 12894 + }, + { + "epoch": 0.6786842105263158, + "grad_norm": 0.49609375, + "learning_rate": 7.140350877192983e-05, + "loss": 3.1523, + "step": 12895 + }, + { + "epoch": 0.6787368421052632, + "grad_norm": 0.51171875, + "learning_rate": 7.139181286549708e-05, + "loss": 3.7308, + "step": 12896 + }, + { + "epoch": 0.6787894736842105, + "grad_norm": 0.59765625, + "learning_rate": 7.138011695906434e-05, + "loss": 3.5469, + "step": 12897 + }, + { + "epoch": 0.6788421052631579, + "grad_norm": 0.48828125, + "learning_rate": 7.136842105263159e-05, + "loss": 3.5913, + "step": 12898 + }, + { + "epoch": 0.6788947368421052, + "grad_norm": 0.486328125, + "learning_rate": 7.135672514619884e-05, + "loss": 3.2519, + "step": 12899 + }, + { + "epoch": 0.6789473684210526, + "grad_norm": 0.48046875, + "learning_rate": 7.134502923976609e-05, + "loss": 3.7806, + "step": 12900 + }, + { + "epoch": 0.679, + "grad_norm": 0.49609375, + "learning_rate": 7.133333333333334e-05, + "loss": 3.3544, + "step": 12901 + }, + { + "epoch": 0.6790526315789474, + "grad_norm": 0.4921875, + "learning_rate": 7.132163742690059e-05, + "loss": 3.1062, + "step": 12902 + }, + { + "epoch": 0.6791052631578948, + "grad_norm": 0.49609375, + "learning_rate": 7.130994152046784e-05, + "loss": 2.8899, + "step": 12903 + }, + { + "epoch": 0.6791578947368421, + "grad_norm": 0.5390625, + "learning_rate": 7.129824561403509e-05, + "loss": 3.7675, + "step": 12904 + }, + { + "epoch": 0.6792105263157895, + "grad_norm": 0.50390625, + "learning_rate": 7.128654970760235e-05, + "loss": 2.9825, + "step": 12905 + }, + { + "epoch": 0.6792631578947368, + "grad_norm": 0.462890625, + "learning_rate": 7.12748538011696e-05, + "loss": 3.2815, + "step": 12906 + }, + { + "epoch": 0.6793157894736842, + "grad_norm": 0.48828125, + "learning_rate": 7.126315789473685e-05, + "loss": 3.715, + "step": 12907 + }, + { + "epoch": 0.6793684210526316, + "grad_norm": 0.494140625, + "learning_rate": 7.12514619883041e-05, + "loss": 3.4022, + "step": 12908 + }, + { + "epoch": 0.6794210526315789, + "grad_norm": 0.51171875, + "learning_rate": 7.123976608187135e-05, + "loss": 3.3167, + "step": 12909 + }, + { + "epoch": 0.6794736842105263, + "grad_norm": 0.490234375, + "learning_rate": 7.12280701754386e-05, + "loss": 3.6778, + "step": 12910 + }, + { + "epoch": 0.6795263157894736, + "grad_norm": 0.5, + "learning_rate": 7.121637426900585e-05, + "loss": 2.8141, + "step": 12911 + }, + { + "epoch": 0.6795789473684211, + "grad_norm": 0.482421875, + "learning_rate": 7.120467836257311e-05, + "loss": 3.7506, + "step": 12912 + }, + { + "epoch": 0.6796315789473685, + "grad_norm": 0.490234375, + "learning_rate": 7.119298245614036e-05, + "loss": 3.2037, + "step": 12913 + }, + { + "epoch": 0.6796842105263158, + "grad_norm": 0.5078125, + "learning_rate": 7.118128654970761e-05, + "loss": 3.2446, + "step": 12914 + }, + { + "epoch": 0.6797368421052632, + "grad_norm": 0.48046875, + "learning_rate": 7.116959064327486e-05, + "loss": 3.0507, + "step": 12915 + }, + { + "epoch": 0.6797894736842105, + "grad_norm": 0.5, + "learning_rate": 7.115789473684211e-05, + "loss": 3.3707, + "step": 12916 + }, + { + "epoch": 0.6798421052631579, + "grad_norm": 0.50390625, + "learning_rate": 7.114619883040936e-05, + "loss": 3.6084, + "step": 12917 + }, + { + "epoch": 0.6798947368421052, + "grad_norm": 0.515625, + "learning_rate": 7.113450292397661e-05, + "loss": 3.1896, + "step": 12918 + }, + { + "epoch": 0.6799473684210526, + "grad_norm": 0.498046875, + "learning_rate": 7.112280701754386e-05, + "loss": 3.5445, + "step": 12919 + }, + { + "epoch": 0.68, + "grad_norm": 0.515625, + "learning_rate": 7.111111111111112e-05, + "loss": 3.0995, + "step": 12920 + }, + { + "epoch": 0.6800526315789474, + "grad_norm": 0.482421875, + "learning_rate": 7.109941520467837e-05, + "loss": 3.4096, + "step": 12921 + }, + { + "epoch": 0.6801052631578948, + "grad_norm": 0.490234375, + "learning_rate": 7.108771929824562e-05, + "loss": 3.7998, + "step": 12922 + }, + { + "epoch": 0.6801578947368421, + "grad_norm": 0.484375, + "learning_rate": 7.107602339181287e-05, + "loss": 3.3247, + "step": 12923 + }, + { + "epoch": 0.6802105263157895, + "grad_norm": 0.5, + "learning_rate": 7.106432748538012e-05, + "loss": 3.2032, + "step": 12924 + }, + { + "epoch": 0.6802631578947368, + "grad_norm": 0.50390625, + "learning_rate": 7.105263157894737e-05, + "loss": 3.0424, + "step": 12925 + }, + { + "epoch": 0.6803157894736842, + "grad_norm": 0.515625, + "learning_rate": 7.104093567251462e-05, + "loss": 3.5036, + "step": 12926 + }, + { + "epoch": 0.6803684210526316, + "grad_norm": 0.47265625, + "learning_rate": 7.102923976608188e-05, + "loss": 3.7694, + "step": 12927 + }, + { + "epoch": 0.6804210526315789, + "grad_norm": 0.470703125, + "learning_rate": 7.101754385964913e-05, + "loss": 3.0181, + "step": 12928 + }, + { + "epoch": 0.6804736842105263, + "grad_norm": 0.50390625, + "learning_rate": 7.100584795321638e-05, + "loss": 3.1143, + "step": 12929 + }, + { + "epoch": 0.6805263157894736, + "grad_norm": 0.5, + "learning_rate": 7.099415204678363e-05, + "loss": 3.5628, + "step": 12930 + }, + { + "epoch": 0.6805789473684211, + "grad_norm": 0.51171875, + "learning_rate": 7.098245614035088e-05, + "loss": 3.5924, + "step": 12931 + }, + { + "epoch": 0.6806315789473685, + "grad_norm": 0.57421875, + "learning_rate": 7.097076023391813e-05, + "loss": 3.4034, + "step": 12932 + }, + { + "epoch": 0.6806842105263158, + "grad_norm": 0.484375, + "learning_rate": 7.095906432748538e-05, + "loss": 3.265, + "step": 12933 + }, + { + "epoch": 0.6807368421052632, + "grad_norm": 0.494140625, + "learning_rate": 7.094736842105264e-05, + "loss": 3.4494, + "step": 12934 + }, + { + "epoch": 0.6807894736842105, + "grad_norm": 0.51953125, + "learning_rate": 7.093567251461989e-05, + "loss": 3.333, + "step": 12935 + }, + { + "epoch": 0.6808421052631579, + "grad_norm": 0.46484375, + "learning_rate": 7.092397660818714e-05, + "loss": 3.4959, + "step": 12936 + }, + { + "epoch": 0.6808947368421052, + "grad_norm": 0.482421875, + "learning_rate": 7.091228070175439e-05, + "loss": 3.2864, + "step": 12937 + }, + { + "epoch": 0.6809473684210526, + "grad_norm": 0.53515625, + "learning_rate": 7.090058479532164e-05, + "loss": 2.7464, + "step": 12938 + }, + { + "epoch": 0.681, + "grad_norm": 0.490234375, + "learning_rate": 7.088888888888889e-05, + "loss": 2.9519, + "step": 12939 + }, + { + "epoch": 0.6810526315789474, + "grad_norm": 0.5, + "learning_rate": 7.087719298245614e-05, + "loss": 3.2859, + "step": 12940 + }, + { + "epoch": 0.6811052631578948, + "grad_norm": 0.4765625, + "learning_rate": 7.086549707602339e-05, + "loss": 3.758, + "step": 12941 + }, + { + "epoch": 0.6811578947368421, + "grad_norm": 0.47265625, + "learning_rate": 7.085380116959065e-05, + "loss": 3.0684, + "step": 12942 + }, + { + "epoch": 0.6812105263157895, + "grad_norm": 0.48046875, + "learning_rate": 7.08421052631579e-05, + "loss": 3.0701, + "step": 12943 + }, + { + "epoch": 0.6812631578947368, + "grad_norm": 0.478515625, + "learning_rate": 7.083040935672515e-05, + "loss": 3.1655, + "step": 12944 + }, + { + "epoch": 0.6813157894736842, + "grad_norm": 0.47265625, + "learning_rate": 7.08187134502924e-05, + "loss": 3.1589, + "step": 12945 + }, + { + "epoch": 0.6813684210526316, + "grad_norm": 0.51953125, + "learning_rate": 7.080701754385965e-05, + "loss": 2.8943, + "step": 12946 + }, + { + "epoch": 0.6814210526315789, + "grad_norm": 0.49609375, + "learning_rate": 7.07953216374269e-05, + "loss": 3.0711, + "step": 12947 + }, + { + "epoch": 0.6814736842105263, + "grad_norm": 0.478515625, + "learning_rate": 7.078362573099415e-05, + "loss": 3.5318, + "step": 12948 + }, + { + "epoch": 0.6815263157894736, + "grad_norm": 0.640625, + "learning_rate": 7.077192982456141e-05, + "loss": 3.5078, + "step": 12949 + }, + { + "epoch": 0.6815789473684211, + "grad_norm": 0.486328125, + "learning_rate": 7.076023391812866e-05, + "loss": 3.3949, + "step": 12950 + }, + { + "epoch": 0.6816315789473684, + "grad_norm": 0.5546875, + "learning_rate": 7.074853801169591e-05, + "loss": 3.5846, + "step": 12951 + }, + { + "epoch": 0.6816842105263158, + "grad_norm": 0.5546875, + "learning_rate": 7.073684210526316e-05, + "loss": 2.8211, + "step": 12952 + }, + { + "epoch": 0.6817368421052632, + "grad_norm": 0.5078125, + "learning_rate": 7.072514619883041e-05, + "loss": 2.9463, + "step": 12953 + }, + { + "epoch": 0.6817894736842105, + "grad_norm": 0.46875, + "learning_rate": 7.071345029239766e-05, + "loss": 3.0143, + "step": 12954 + }, + { + "epoch": 0.6818421052631579, + "grad_norm": 0.51953125, + "learning_rate": 7.070175438596491e-05, + "loss": 3.2076, + "step": 12955 + }, + { + "epoch": 0.6818947368421052, + "grad_norm": 0.458984375, + "learning_rate": 7.069005847953216e-05, + "loss": 3.2571, + "step": 12956 + }, + { + "epoch": 0.6819473684210526, + "grad_norm": 0.49609375, + "learning_rate": 7.067836257309942e-05, + "loss": 3.0475, + "step": 12957 + }, + { + "epoch": 0.682, + "grad_norm": 0.458984375, + "learning_rate": 7.066666666666667e-05, + "loss": 3.1318, + "step": 12958 + }, + { + "epoch": 0.6820526315789474, + "grad_norm": 0.5, + "learning_rate": 7.065497076023392e-05, + "loss": 3.6227, + "step": 12959 + }, + { + "epoch": 0.6821052631578948, + "grad_norm": 0.494140625, + "learning_rate": 7.064327485380117e-05, + "loss": 3.3346, + "step": 12960 + }, + { + "epoch": 0.6821578947368421, + "grad_norm": 0.53125, + "learning_rate": 7.063157894736842e-05, + "loss": 3.0302, + "step": 12961 + }, + { + "epoch": 0.6822105263157895, + "grad_norm": 0.494140625, + "learning_rate": 7.061988304093567e-05, + "loss": 3.6159, + "step": 12962 + }, + { + "epoch": 0.6822631578947368, + "grad_norm": 0.484375, + "learning_rate": 7.060818713450292e-05, + "loss": 3.965, + "step": 12963 + }, + { + "epoch": 0.6823157894736842, + "grad_norm": 0.4921875, + "learning_rate": 7.059649122807018e-05, + "loss": 3.2285, + "step": 12964 + }, + { + "epoch": 0.6823684210526316, + "grad_norm": 0.48828125, + "learning_rate": 7.058479532163743e-05, + "loss": 3.3017, + "step": 12965 + }, + { + "epoch": 0.6824210526315789, + "grad_norm": 0.47265625, + "learning_rate": 7.057309941520468e-05, + "loss": 3.035, + "step": 12966 + }, + { + "epoch": 0.6824736842105263, + "grad_norm": 0.5390625, + "learning_rate": 7.056140350877193e-05, + "loss": 3.4848, + "step": 12967 + }, + { + "epoch": 0.6825263157894736, + "grad_norm": 0.5078125, + "learning_rate": 7.054970760233918e-05, + "loss": 3.7022, + "step": 12968 + }, + { + "epoch": 0.6825789473684211, + "grad_norm": 0.546875, + "learning_rate": 7.053801169590643e-05, + "loss": 3.4616, + "step": 12969 + }, + { + "epoch": 0.6826315789473684, + "grad_norm": 0.51171875, + "learning_rate": 7.052631578947368e-05, + "loss": 3.4301, + "step": 12970 + }, + { + "epoch": 0.6826842105263158, + "grad_norm": 0.498046875, + "learning_rate": 7.051461988304094e-05, + "loss": 3.5829, + "step": 12971 + }, + { + "epoch": 0.6827368421052632, + "grad_norm": 0.5, + "learning_rate": 7.05029239766082e-05, + "loss": 3.357, + "step": 12972 + }, + { + "epoch": 0.6827894736842105, + "grad_norm": 0.49609375, + "learning_rate": 7.049122807017544e-05, + "loss": 3.0862, + "step": 12973 + }, + { + "epoch": 0.6828421052631579, + "grad_norm": 0.484375, + "learning_rate": 7.047953216374269e-05, + "loss": 3.391, + "step": 12974 + }, + { + "epoch": 0.6828947368421052, + "grad_norm": 0.5, + "learning_rate": 7.046783625730996e-05, + "loss": 3.2559, + "step": 12975 + }, + { + "epoch": 0.6829473684210526, + "grad_norm": 0.5, + "learning_rate": 7.045614035087719e-05, + "loss": 3.1222, + "step": 12976 + }, + { + "epoch": 0.683, + "grad_norm": 0.48828125, + "learning_rate": 7.044444444444444e-05, + "loss": 3.476, + "step": 12977 + }, + { + "epoch": 0.6830526315789474, + "grad_norm": 0.546875, + "learning_rate": 7.043274853801169e-05, + "loss": 2.5751, + "step": 12978 + }, + { + "epoch": 0.6831052631578948, + "grad_norm": 0.4921875, + "learning_rate": 7.042105263157895e-05, + "loss": 3.4702, + "step": 12979 + }, + { + "epoch": 0.6831578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.04093567251462e-05, + "loss": 3.362, + "step": 12980 + }, + { + "epoch": 0.6832105263157895, + "grad_norm": 0.484375, + "learning_rate": 7.039766081871345e-05, + "loss": 3.2566, + "step": 12981 + }, + { + "epoch": 0.6832631578947368, + "grad_norm": 0.515625, + "learning_rate": 7.03859649122807e-05, + "loss": 3.441, + "step": 12982 + }, + { + "epoch": 0.6833157894736842, + "grad_norm": 0.482421875, + "learning_rate": 7.037426900584795e-05, + "loss": 3.2732, + "step": 12983 + }, + { + "epoch": 0.6833684210526316, + "grad_norm": 0.53125, + "learning_rate": 7.03625730994152e-05, + "loss": 3.0437, + "step": 12984 + }, + { + "epoch": 0.6834210526315789, + "grad_norm": 0.494140625, + "learning_rate": 7.035087719298245e-05, + "loss": 3.3405, + "step": 12985 + }, + { + "epoch": 0.6834736842105263, + "grad_norm": 0.5078125, + "learning_rate": 7.033918128654972e-05, + "loss": 3.1221, + "step": 12986 + }, + { + "epoch": 0.6835263157894736, + "grad_norm": 0.51171875, + "learning_rate": 7.032748538011696e-05, + "loss": 3.3965, + "step": 12987 + }, + { + "epoch": 0.6835789473684211, + "grad_norm": 0.470703125, + "learning_rate": 7.031578947368421e-05, + "loss": 3.2755, + "step": 12988 + }, + { + "epoch": 0.6836315789473684, + "grad_norm": 0.49609375, + "learning_rate": 7.030409356725146e-05, + "loss": 3.79, + "step": 12989 + }, + { + "epoch": 0.6836842105263158, + "grad_norm": 0.484375, + "learning_rate": 7.029239766081873e-05, + "loss": 3.2364, + "step": 12990 + }, + { + "epoch": 0.6837368421052632, + "grad_norm": 0.5390625, + "learning_rate": 7.028070175438596e-05, + "loss": 3.0976, + "step": 12991 + }, + { + "epoch": 0.6837894736842105, + "grad_norm": 0.50390625, + "learning_rate": 7.026900584795321e-05, + "loss": 3.1442, + "step": 12992 + }, + { + "epoch": 0.6838421052631579, + "grad_norm": 0.50390625, + "learning_rate": 7.025730994152046e-05, + "loss": 3.1859, + "step": 12993 + }, + { + "epoch": 0.6838947368421052, + "grad_norm": 0.48046875, + "learning_rate": 7.024561403508773e-05, + "loss": 3.4525, + "step": 12994 + }, + { + "epoch": 0.6839473684210526, + "grad_norm": 0.490234375, + "learning_rate": 7.023391812865497e-05, + "loss": 3.2295, + "step": 12995 + }, + { + "epoch": 0.684, + "grad_norm": 0.5078125, + "learning_rate": 7.022222222222222e-05, + "loss": 3.0739, + "step": 12996 + }, + { + "epoch": 0.6840526315789474, + "grad_norm": 0.490234375, + "learning_rate": 7.021052631578949e-05, + "loss": 3.1751, + "step": 12997 + }, + { + "epoch": 0.6841052631578948, + "grad_norm": 0.455078125, + "learning_rate": 7.019883040935674e-05, + "loss": 3.2923, + "step": 12998 + }, + { + "epoch": 0.6841578947368421, + "grad_norm": 0.486328125, + "learning_rate": 7.018713450292397e-05, + "loss": 3.6455, + "step": 12999 + }, + { + "epoch": 0.6842105263157895, + "grad_norm": 0.48828125, + "learning_rate": 7.017543859649122e-05, + "loss": 3.0321, + "step": 13000 + }, + { + "epoch": 0.6842631578947368, + "grad_norm": 0.51953125, + "learning_rate": 7.016374269005849e-05, + "loss": 3.4517, + "step": 13001 + }, + { + "epoch": 0.6843157894736842, + "grad_norm": 0.482421875, + "learning_rate": 7.015204678362574e-05, + "loss": 3.7043, + "step": 13002 + }, + { + "epoch": 0.6843684210526316, + "grad_norm": 0.484375, + "learning_rate": 7.014035087719299e-05, + "loss": 3.5055, + "step": 13003 + }, + { + "epoch": 0.6844210526315789, + "grad_norm": 0.482421875, + "learning_rate": 7.012865497076023e-05, + "loss": 3.3335, + "step": 13004 + }, + { + "epoch": 0.6844736842105263, + "grad_norm": 0.68359375, + "learning_rate": 7.01169590643275e-05, + "loss": 3.2879, + "step": 13005 + }, + { + "epoch": 0.6845263157894736, + "grad_norm": 0.578125, + "learning_rate": 7.010526315789473e-05, + "loss": 3.2174, + "step": 13006 + }, + { + "epoch": 0.6845789473684211, + "grad_norm": 0.48046875, + "learning_rate": 7.009356725146198e-05, + "loss": 3.5151, + "step": 13007 + }, + { + "epoch": 0.6846315789473684, + "grad_norm": 0.458984375, + "learning_rate": 7.008187134502925e-05, + "loss": 3.0675, + "step": 13008 + }, + { + "epoch": 0.6846842105263158, + "grad_norm": 0.51171875, + "learning_rate": 7.00701754385965e-05, + "loss": 2.9671, + "step": 13009 + }, + { + "epoch": 0.6847368421052632, + "grad_norm": 0.50390625, + "learning_rate": 7.005847953216375e-05, + "loss": 3.3822, + "step": 13010 + }, + { + "epoch": 0.6847894736842105, + "grad_norm": 0.48046875, + "learning_rate": 7.0046783625731e-05, + "loss": 3.4368, + "step": 13011 + }, + { + "epoch": 0.6848421052631579, + "grad_norm": 0.50390625, + "learning_rate": 7.003508771929826e-05, + "loss": 3.457, + "step": 13012 + }, + { + "epoch": 0.6848947368421052, + "grad_norm": 0.48046875, + "learning_rate": 7.002339181286551e-05, + "loss": 3.5999, + "step": 13013 + }, + { + "epoch": 0.6849473684210526, + "grad_norm": 0.50390625, + "learning_rate": 7.001169590643274e-05, + "loss": 3.6464, + "step": 13014 + }, + { + "epoch": 0.685, + "grad_norm": 0.5703125, + "learning_rate": 7e-05, + "loss": 3.1568, + "step": 13015 + }, + { + "epoch": 0.6850526315789474, + "grad_norm": 0.466796875, + "learning_rate": 6.998830409356726e-05, + "loss": 3.5108, + "step": 13016 + }, + { + "epoch": 0.6851052631578948, + "grad_norm": 0.474609375, + "learning_rate": 6.99766081871345e-05, + "loss": 3.5479, + "step": 13017 + }, + { + "epoch": 0.6851578947368421, + "grad_norm": 0.51953125, + "learning_rate": 6.996491228070176e-05, + "loss": 3.1153, + "step": 13018 + }, + { + "epoch": 0.6852105263157895, + "grad_norm": 0.470703125, + "learning_rate": 6.9953216374269e-05, + "loss": 2.826, + "step": 13019 + }, + { + "epoch": 0.6852631578947368, + "grad_norm": 0.55078125, + "learning_rate": 6.994152046783627e-05, + "loss": 3.4217, + "step": 13020 + }, + { + "epoch": 0.6853157894736842, + "grad_norm": 0.484375, + "learning_rate": 6.992982456140352e-05, + "loss": 3.4944, + "step": 13021 + }, + { + "epoch": 0.6853684210526316, + "grad_norm": 0.4609375, + "learning_rate": 6.991812865497075e-05, + "loss": 3.2493, + "step": 13022 + }, + { + "epoch": 0.6854210526315789, + "grad_norm": 0.466796875, + "learning_rate": 6.990643274853802e-05, + "loss": 3.2211, + "step": 13023 + }, + { + "epoch": 0.6854736842105263, + "grad_norm": 0.486328125, + "learning_rate": 6.989473684210527e-05, + "loss": 3.1308, + "step": 13024 + }, + { + "epoch": 0.6855263157894737, + "grad_norm": 0.466796875, + "learning_rate": 6.988304093567252e-05, + "loss": 3.6027, + "step": 13025 + }, + { + "epoch": 0.6855789473684211, + "grad_norm": 0.48828125, + "learning_rate": 6.987134502923977e-05, + "loss": 3.2278, + "step": 13026 + }, + { + "epoch": 0.6856315789473684, + "grad_norm": 0.53125, + "learning_rate": 6.985964912280703e-05, + "loss": 3.3793, + "step": 13027 + }, + { + "epoch": 0.6856842105263158, + "grad_norm": 0.50390625, + "learning_rate": 6.984795321637428e-05, + "loss": 3.5598, + "step": 13028 + }, + { + "epoch": 0.6857368421052632, + "grad_norm": 0.515625, + "learning_rate": 6.983625730994152e-05, + "loss": 3.4038, + "step": 13029 + }, + { + "epoch": 0.6857894736842105, + "grad_norm": 0.5078125, + "learning_rate": 6.982456140350876e-05, + "loss": 3.0015, + "step": 13030 + }, + { + "epoch": 0.6858421052631579, + "grad_norm": 0.494140625, + "learning_rate": 6.981286549707603e-05, + "loss": 3.3056, + "step": 13031 + }, + { + "epoch": 0.6858947368421052, + "grad_norm": 0.484375, + "learning_rate": 6.980116959064328e-05, + "loss": 3.1044, + "step": 13032 + }, + { + "epoch": 0.6859473684210526, + "grad_norm": 0.82421875, + "learning_rate": 6.978947368421053e-05, + "loss": 3.602, + "step": 13033 + }, + { + "epoch": 0.686, + "grad_norm": 0.50390625, + "learning_rate": 6.977777777777779e-05, + "loss": 3.7259, + "step": 13034 + }, + { + "epoch": 0.6860526315789474, + "grad_norm": 0.48828125, + "learning_rate": 6.976608187134504e-05, + "loss": 3.3383, + "step": 13035 + }, + { + "epoch": 0.6861052631578948, + "grad_norm": 0.47265625, + "learning_rate": 6.975438596491229e-05, + "loss": 3.5158, + "step": 13036 + }, + { + "epoch": 0.6861578947368421, + "grad_norm": 0.462890625, + "learning_rate": 6.974269005847953e-05, + "loss": 3.3928, + "step": 13037 + }, + { + "epoch": 0.6862105263157895, + "grad_norm": 0.484375, + "learning_rate": 6.973099415204679e-05, + "loss": 3.236, + "step": 13038 + }, + { + "epoch": 0.6862631578947368, + "grad_norm": 0.47265625, + "learning_rate": 6.971929824561404e-05, + "loss": 3.5792, + "step": 13039 + }, + { + "epoch": 0.6863157894736842, + "grad_norm": 0.515625, + "learning_rate": 6.970760233918129e-05, + "loss": 3.1882, + "step": 13040 + }, + { + "epoch": 0.6863684210526316, + "grad_norm": 0.486328125, + "learning_rate": 6.969590643274854e-05, + "loss": 2.8789, + "step": 13041 + }, + { + "epoch": 0.6864210526315789, + "grad_norm": 0.53515625, + "learning_rate": 6.96842105263158e-05, + "loss": 3.0242, + "step": 13042 + }, + { + "epoch": 0.6864736842105263, + "grad_norm": 0.48828125, + "learning_rate": 6.967251461988305e-05, + "loss": 3.6443, + "step": 13043 + }, + { + "epoch": 0.6865263157894737, + "grad_norm": 0.47265625, + "learning_rate": 6.96608187134503e-05, + "loss": 3.164, + "step": 13044 + }, + { + "epoch": 0.6865789473684211, + "grad_norm": 0.50390625, + "learning_rate": 6.964912280701754e-05, + "loss": 3.3199, + "step": 13045 + }, + { + "epoch": 0.6866315789473684, + "grad_norm": 0.46875, + "learning_rate": 6.96374269005848e-05, + "loss": 3.0412, + "step": 13046 + }, + { + "epoch": 0.6866842105263158, + "grad_norm": 0.478515625, + "learning_rate": 6.962573099415205e-05, + "loss": 3.5022, + "step": 13047 + }, + { + "epoch": 0.6867368421052632, + "grad_norm": 0.515625, + "learning_rate": 6.96140350877193e-05, + "loss": 3.377, + "step": 13048 + }, + { + "epoch": 0.6867894736842105, + "grad_norm": 0.498046875, + "learning_rate": 6.960233918128656e-05, + "loss": 3.1896, + "step": 13049 + }, + { + "epoch": 0.6868421052631579, + "grad_norm": 0.470703125, + "learning_rate": 6.959064327485381e-05, + "loss": 2.7433, + "step": 13050 + }, + { + "epoch": 0.6868947368421052, + "grad_norm": 0.515625, + "learning_rate": 6.957894736842106e-05, + "loss": 3.1749, + "step": 13051 + }, + { + "epoch": 0.6869473684210526, + "grad_norm": 0.484375, + "learning_rate": 6.95672514619883e-05, + "loss": 3.1513, + "step": 13052 + }, + { + "epoch": 0.687, + "grad_norm": 0.48046875, + "learning_rate": 6.955555555555556e-05, + "loss": 3.0963, + "step": 13053 + }, + { + "epoch": 0.6870526315789474, + "grad_norm": 0.46875, + "learning_rate": 6.954385964912281e-05, + "loss": 3.0507, + "step": 13054 + }, + { + "epoch": 0.6871052631578948, + "grad_norm": 0.5625, + "learning_rate": 6.953216374269006e-05, + "loss": 3.5454, + "step": 13055 + }, + { + "epoch": 0.6871578947368421, + "grad_norm": 0.46875, + "learning_rate": 6.952046783625731e-05, + "loss": 3.5432, + "step": 13056 + }, + { + "epoch": 0.6872105263157895, + "grad_norm": 0.46875, + "learning_rate": 6.950877192982457e-05, + "loss": 2.9992, + "step": 13057 + }, + { + "epoch": 0.6872631578947368, + "grad_norm": 0.462890625, + "learning_rate": 6.949707602339182e-05, + "loss": 3.3785, + "step": 13058 + }, + { + "epoch": 0.6873157894736842, + "grad_norm": 0.458984375, + "learning_rate": 6.948538011695907e-05, + "loss": 3.5377, + "step": 13059 + }, + { + "epoch": 0.6873684210526316, + "grad_norm": 0.5, + "learning_rate": 6.947368421052632e-05, + "loss": 3.3694, + "step": 13060 + }, + { + "epoch": 0.6874210526315789, + "grad_norm": 0.482421875, + "learning_rate": 6.946198830409357e-05, + "loss": 3.278, + "step": 13061 + }, + { + "epoch": 0.6874736842105263, + "grad_norm": 0.494140625, + "learning_rate": 6.945029239766082e-05, + "loss": 3.09, + "step": 13062 + }, + { + "epoch": 0.6875263157894737, + "grad_norm": 0.4921875, + "learning_rate": 6.943859649122807e-05, + "loss": 3.279, + "step": 13063 + }, + { + "epoch": 0.6875789473684211, + "grad_norm": 0.48828125, + "learning_rate": 6.942690058479533e-05, + "loss": 3.0001, + "step": 13064 + }, + { + "epoch": 0.6876315789473684, + "grad_norm": 0.474609375, + "learning_rate": 6.941520467836258e-05, + "loss": 2.9627, + "step": 13065 + }, + { + "epoch": 0.6876842105263158, + "grad_norm": 0.5078125, + "learning_rate": 6.940350877192983e-05, + "loss": 3.4348, + "step": 13066 + }, + { + "epoch": 0.6877368421052632, + "grad_norm": 0.484375, + "learning_rate": 6.939181286549708e-05, + "loss": 3.2837, + "step": 13067 + }, + { + "epoch": 0.6877894736842105, + "grad_norm": 0.51953125, + "learning_rate": 6.938011695906433e-05, + "loss": 3.0194, + "step": 13068 + }, + { + "epoch": 0.6878421052631579, + "grad_norm": 0.49609375, + "learning_rate": 6.936842105263158e-05, + "loss": 3.1187, + "step": 13069 + }, + { + "epoch": 0.6878947368421052, + "grad_norm": 0.50390625, + "learning_rate": 6.935672514619883e-05, + "loss": 3.1323, + "step": 13070 + }, + { + "epoch": 0.6879473684210526, + "grad_norm": 0.494140625, + "learning_rate": 6.934502923976609e-05, + "loss": 3.3177, + "step": 13071 + }, + { + "epoch": 0.688, + "grad_norm": 0.46484375, + "learning_rate": 6.933333333333334e-05, + "loss": 3.4448, + "step": 13072 + }, + { + "epoch": 0.6880526315789474, + "grad_norm": 0.498046875, + "learning_rate": 6.932163742690059e-05, + "loss": 3.416, + "step": 13073 + }, + { + "epoch": 0.6881052631578948, + "grad_norm": 0.478515625, + "learning_rate": 6.930994152046784e-05, + "loss": 3.3022, + "step": 13074 + }, + { + "epoch": 0.6881578947368421, + "grad_norm": 0.49609375, + "learning_rate": 6.929824561403509e-05, + "loss": 3.055, + "step": 13075 + }, + { + "epoch": 0.6882105263157895, + "grad_norm": 0.4921875, + "learning_rate": 6.928654970760234e-05, + "loss": 3.1843, + "step": 13076 + }, + { + "epoch": 0.6882631578947368, + "grad_norm": 0.4921875, + "learning_rate": 6.927485380116959e-05, + "loss": 3.4207, + "step": 13077 + }, + { + "epoch": 0.6883157894736842, + "grad_norm": 0.474609375, + "learning_rate": 6.926315789473684e-05, + "loss": 3.4123, + "step": 13078 + }, + { + "epoch": 0.6883684210526316, + "grad_norm": 0.47265625, + "learning_rate": 6.92514619883041e-05, + "loss": 3.3649, + "step": 13079 + }, + { + "epoch": 0.6884210526315789, + "grad_norm": 0.470703125, + "learning_rate": 6.923976608187135e-05, + "loss": 3.5956, + "step": 13080 + }, + { + "epoch": 0.6884736842105263, + "grad_norm": 0.4609375, + "learning_rate": 6.92280701754386e-05, + "loss": 2.9146, + "step": 13081 + }, + { + "epoch": 0.6885263157894737, + "grad_norm": 0.494140625, + "learning_rate": 6.921637426900585e-05, + "loss": 3.4061, + "step": 13082 + }, + { + "epoch": 0.6885789473684211, + "grad_norm": 0.48046875, + "learning_rate": 6.92046783625731e-05, + "loss": 3.32, + "step": 13083 + }, + { + "epoch": 0.6886315789473684, + "grad_norm": 0.46875, + "learning_rate": 6.919298245614035e-05, + "loss": 3.123, + "step": 13084 + }, + { + "epoch": 0.6886842105263158, + "grad_norm": 0.45703125, + "learning_rate": 6.91812865497076e-05, + "loss": 3.2998, + "step": 13085 + }, + { + "epoch": 0.6887368421052632, + "grad_norm": 0.484375, + "learning_rate": 6.916959064327486e-05, + "loss": 3.9317, + "step": 13086 + }, + { + "epoch": 0.6887894736842105, + "grad_norm": 0.494140625, + "learning_rate": 6.915789473684211e-05, + "loss": 3.5543, + "step": 13087 + }, + { + "epoch": 0.6888421052631579, + "grad_norm": 0.70703125, + "learning_rate": 6.914619883040936e-05, + "loss": 2.7695, + "step": 13088 + }, + { + "epoch": 0.6888947368421052, + "grad_norm": 0.5234375, + "learning_rate": 6.913450292397661e-05, + "loss": 3.3752, + "step": 13089 + }, + { + "epoch": 0.6889473684210526, + "grad_norm": 0.4765625, + "learning_rate": 6.912280701754386e-05, + "loss": 3.5161, + "step": 13090 + }, + { + "epoch": 0.689, + "grad_norm": 0.470703125, + "learning_rate": 6.911111111111111e-05, + "loss": 2.8547, + "step": 13091 + }, + { + "epoch": 0.6890526315789474, + "grad_norm": 0.486328125, + "learning_rate": 6.909941520467836e-05, + "loss": 3.1853, + "step": 13092 + }, + { + "epoch": 0.6891052631578948, + "grad_norm": 0.49609375, + "learning_rate": 6.908771929824561e-05, + "loss": 2.9947, + "step": 13093 + }, + { + "epoch": 0.6891578947368421, + "grad_norm": 0.48828125, + "learning_rate": 6.907602339181287e-05, + "loss": 3.5331, + "step": 13094 + }, + { + "epoch": 0.6892105263157895, + "grad_norm": 0.50390625, + "learning_rate": 6.906432748538012e-05, + "loss": 3.7697, + "step": 13095 + }, + { + "epoch": 0.6892631578947368, + "grad_norm": 0.482421875, + "learning_rate": 6.905263157894737e-05, + "loss": 3.2206, + "step": 13096 + }, + { + "epoch": 0.6893157894736842, + "grad_norm": 0.515625, + "learning_rate": 6.904093567251462e-05, + "loss": 3.4978, + "step": 13097 + }, + { + "epoch": 0.6893684210526316, + "grad_norm": 0.490234375, + "learning_rate": 6.902923976608187e-05, + "loss": 3.3585, + "step": 13098 + }, + { + "epoch": 0.6894210526315789, + "grad_norm": 0.5390625, + "learning_rate": 6.901754385964912e-05, + "loss": 3.4957, + "step": 13099 + }, + { + "epoch": 0.6894736842105263, + "grad_norm": 0.5, + "learning_rate": 6.900584795321637e-05, + "loss": 3.339, + "step": 13100 + }, + { + "epoch": 0.6895263157894737, + "grad_norm": 0.486328125, + "learning_rate": 6.899415204678364e-05, + "loss": 3.2114, + "step": 13101 + }, + { + "epoch": 0.6895789473684211, + "grad_norm": 0.48828125, + "learning_rate": 6.898245614035088e-05, + "loss": 2.9655, + "step": 13102 + }, + { + "epoch": 0.6896315789473684, + "grad_norm": 0.48046875, + "learning_rate": 6.897076023391813e-05, + "loss": 3.1795, + "step": 13103 + }, + { + "epoch": 0.6896842105263158, + "grad_norm": 0.484375, + "learning_rate": 6.895906432748538e-05, + "loss": 3.445, + "step": 13104 + }, + { + "epoch": 0.6897368421052632, + "grad_norm": 0.482421875, + "learning_rate": 6.894736842105263e-05, + "loss": 3.2873, + "step": 13105 + }, + { + "epoch": 0.6897894736842105, + "grad_norm": 0.50390625, + "learning_rate": 6.893567251461988e-05, + "loss": 3.2724, + "step": 13106 + }, + { + "epoch": 0.6898421052631579, + "grad_norm": 0.478515625, + "learning_rate": 6.892397660818713e-05, + "loss": 3.3188, + "step": 13107 + }, + { + "epoch": 0.6898947368421052, + "grad_norm": 0.46875, + "learning_rate": 6.89122807017544e-05, + "loss": 3.1649, + "step": 13108 + }, + { + "epoch": 0.6899473684210526, + "grad_norm": 0.466796875, + "learning_rate": 6.890058479532165e-05, + "loss": 3.3112, + "step": 13109 + }, + { + "epoch": 0.69, + "grad_norm": 0.478515625, + "learning_rate": 6.88888888888889e-05, + "loss": 3.5771, + "step": 13110 + }, + { + "epoch": 0.6900526315789474, + "grad_norm": 0.5390625, + "learning_rate": 6.887719298245614e-05, + "loss": 3.3069, + "step": 13111 + }, + { + "epoch": 0.6901052631578948, + "grad_norm": 0.49609375, + "learning_rate": 6.88654970760234e-05, + "loss": 3.5181, + "step": 13112 + }, + { + "epoch": 0.6901578947368421, + "grad_norm": 0.51171875, + "learning_rate": 6.885380116959064e-05, + "loss": 2.9244, + "step": 13113 + }, + { + "epoch": 0.6902105263157895, + "grad_norm": 0.5078125, + "learning_rate": 6.88421052631579e-05, + "loss": 3.7705, + "step": 13114 + }, + { + "epoch": 0.6902631578947368, + "grad_norm": 0.53125, + "learning_rate": 6.883040935672514e-05, + "loss": 3.2552, + "step": 13115 + }, + { + "epoch": 0.6903157894736842, + "grad_norm": 0.59375, + "learning_rate": 6.88187134502924e-05, + "loss": 3.5357, + "step": 13116 + }, + { + "epoch": 0.6903684210526316, + "grad_norm": 0.48828125, + "learning_rate": 6.880701754385966e-05, + "loss": 3.1965, + "step": 13117 + }, + { + "epoch": 0.6904210526315789, + "grad_norm": 0.486328125, + "learning_rate": 6.87953216374269e-05, + "loss": 3.4801, + "step": 13118 + }, + { + "epoch": 0.6904736842105264, + "grad_norm": 0.50390625, + "learning_rate": 6.878362573099415e-05, + "loss": 3.4894, + "step": 13119 + }, + { + "epoch": 0.6905263157894737, + "grad_norm": 0.578125, + "learning_rate": 6.87719298245614e-05, + "loss": 3.1483, + "step": 13120 + }, + { + "epoch": 0.6905789473684211, + "grad_norm": 0.478515625, + "learning_rate": 6.876023391812865e-05, + "loss": 3.1999, + "step": 13121 + }, + { + "epoch": 0.6906315789473684, + "grad_norm": 0.5234375, + "learning_rate": 6.87485380116959e-05, + "loss": 3.6943, + "step": 13122 + }, + { + "epoch": 0.6906842105263158, + "grad_norm": 0.486328125, + "learning_rate": 6.873684210526317e-05, + "loss": 3.2975, + "step": 13123 + }, + { + "epoch": 0.6907368421052632, + "grad_norm": 0.52734375, + "learning_rate": 6.872514619883042e-05, + "loss": 3.2844, + "step": 13124 + }, + { + "epoch": 0.6907894736842105, + "grad_norm": 0.5078125, + "learning_rate": 6.871345029239767e-05, + "loss": 3.2796, + "step": 13125 + }, + { + "epoch": 0.6908421052631579, + "grad_norm": 0.49609375, + "learning_rate": 6.870175438596492e-05, + "loss": 3.6513, + "step": 13126 + }, + { + "epoch": 0.6908947368421052, + "grad_norm": 0.51171875, + "learning_rate": 6.869005847953217e-05, + "loss": 3.3267, + "step": 13127 + }, + { + "epoch": 0.6909473684210526, + "grad_norm": 0.494140625, + "learning_rate": 6.867836257309941e-05, + "loss": 3.562, + "step": 13128 + }, + { + "epoch": 0.691, + "grad_norm": 0.5, + "learning_rate": 6.866666666666666e-05, + "loss": 3.2411, + "step": 13129 + }, + { + "epoch": 0.6910526315789474, + "grad_norm": 0.546875, + "learning_rate": 6.865497076023391e-05, + "loss": 3.1562, + "step": 13130 + }, + { + "epoch": 0.6911052631578948, + "grad_norm": 0.50390625, + "learning_rate": 6.864327485380118e-05, + "loss": 3.2893, + "step": 13131 + }, + { + "epoch": 0.6911578947368421, + "grad_norm": 0.466796875, + "learning_rate": 6.863157894736843e-05, + "loss": 3.4616, + "step": 13132 + }, + { + "epoch": 0.6912105263157895, + "grad_norm": 0.49609375, + "learning_rate": 6.861988304093568e-05, + "loss": 3.1181, + "step": 13133 + }, + { + "epoch": 0.6912631578947368, + "grad_norm": 0.494140625, + "learning_rate": 6.860818713450294e-05, + "loss": 3.3173, + "step": 13134 + }, + { + "epoch": 0.6913157894736842, + "grad_norm": 0.490234375, + "learning_rate": 6.859649122807018e-05, + "loss": 3.4631, + "step": 13135 + }, + { + "epoch": 0.6913684210526316, + "grad_norm": 0.49609375, + "learning_rate": 6.858479532163742e-05, + "loss": 2.8422, + "step": 13136 + }, + { + "epoch": 0.6914210526315789, + "grad_norm": 0.4921875, + "learning_rate": 6.857309941520467e-05, + "loss": 3.5924, + "step": 13137 + }, + { + "epoch": 0.6914736842105264, + "grad_norm": 0.5, + "learning_rate": 6.856140350877194e-05, + "loss": 3.5325, + "step": 13138 + }, + { + "epoch": 0.6915263157894737, + "grad_norm": 0.47265625, + "learning_rate": 6.854970760233919e-05, + "loss": 2.8355, + "step": 13139 + }, + { + "epoch": 0.6915789473684211, + "grad_norm": 0.54296875, + "learning_rate": 6.853801169590644e-05, + "loss": 3.2068, + "step": 13140 + }, + { + "epoch": 0.6916315789473684, + "grad_norm": 0.51171875, + "learning_rate": 6.852631578947369e-05, + "loss": 3.0514, + "step": 13141 + }, + { + "epoch": 0.6916842105263158, + "grad_norm": 0.47265625, + "learning_rate": 6.851461988304094e-05, + "loss": 3.3631, + "step": 13142 + }, + { + "epoch": 0.6917368421052632, + "grad_norm": 0.5078125, + "learning_rate": 6.850292397660819e-05, + "loss": 3.1113, + "step": 13143 + }, + { + "epoch": 0.6917894736842105, + "grad_norm": 0.4921875, + "learning_rate": 6.849122807017544e-05, + "loss": 3.4121, + "step": 13144 + }, + { + "epoch": 0.6918421052631579, + "grad_norm": 0.498046875, + "learning_rate": 6.847953216374268e-05, + "loss": 3.2656, + "step": 13145 + }, + { + "epoch": 0.6918947368421052, + "grad_norm": 0.5078125, + "learning_rate": 6.846783625730995e-05, + "loss": 3.5527, + "step": 13146 + }, + { + "epoch": 0.6919473684210526, + "grad_norm": 0.5234375, + "learning_rate": 6.84561403508772e-05, + "loss": 3.4498, + "step": 13147 + }, + { + "epoch": 0.692, + "grad_norm": 0.50390625, + "learning_rate": 6.844444444444445e-05, + "loss": 3.5353, + "step": 13148 + }, + { + "epoch": 0.6920526315789474, + "grad_norm": 0.494140625, + "learning_rate": 6.843274853801171e-05, + "loss": 3.4249, + "step": 13149 + }, + { + "epoch": 0.6921052631578948, + "grad_norm": 0.53515625, + "learning_rate": 6.842105263157895e-05, + "loss": 3.1004, + "step": 13150 + }, + { + "epoch": 0.6921578947368421, + "grad_norm": 0.52734375, + "learning_rate": 6.84093567251462e-05, + "loss": 3.4341, + "step": 13151 + }, + { + "epoch": 0.6922105263157895, + "grad_norm": 0.48828125, + "learning_rate": 6.839766081871345e-05, + "loss": 3.2834, + "step": 13152 + }, + { + "epoch": 0.6922631578947368, + "grad_norm": 0.490234375, + "learning_rate": 6.838596491228071e-05, + "loss": 3.2737, + "step": 13153 + }, + { + "epoch": 0.6923157894736842, + "grad_norm": 0.494140625, + "learning_rate": 6.837426900584796e-05, + "loss": 3.4589, + "step": 13154 + }, + { + "epoch": 0.6923684210526316, + "grad_norm": 0.5234375, + "learning_rate": 6.836257309941521e-05, + "loss": 3.0263, + "step": 13155 + }, + { + "epoch": 0.6924210526315789, + "grad_norm": 0.61328125, + "learning_rate": 6.835087719298246e-05, + "loss": 3.302, + "step": 13156 + }, + { + "epoch": 0.6924736842105264, + "grad_norm": 0.484375, + "learning_rate": 6.833918128654972e-05, + "loss": 3.4924, + "step": 13157 + }, + { + "epoch": 0.6925263157894737, + "grad_norm": 0.50390625, + "learning_rate": 6.832748538011696e-05, + "loss": 3.3436, + "step": 13158 + }, + { + "epoch": 0.6925789473684211, + "grad_norm": 0.50390625, + "learning_rate": 6.83157894736842e-05, + "loss": 2.9779, + "step": 13159 + }, + { + "epoch": 0.6926315789473684, + "grad_norm": 0.5078125, + "learning_rate": 6.830409356725147e-05, + "loss": 3.5208, + "step": 13160 + }, + { + "epoch": 0.6926842105263158, + "grad_norm": 0.5703125, + "learning_rate": 6.829239766081872e-05, + "loss": 3.3834, + "step": 13161 + }, + { + "epoch": 0.6927368421052632, + "grad_norm": 0.515625, + "learning_rate": 6.828070175438597e-05, + "loss": 3.0204, + "step": 13162 + }, + { + "epoch": 0.6927894736842105, + "grad_norm": 0.5390625, + "learning_rate": 6.826900584795322e-05, + "loss": 2.9249, + "step": 13163 + }, + { + "epoch": 0.6928421052631579, + "grad_norm": 0.470703125, + "learning_rate": 6.825730994152048e-05, + "loss": 3.4245, + "step": 13164 + }, + { + "epoch": 0.6928947368421052, + "grad_norm": 0.5234375, + "learning_rate": 6.824561403508772e-05, + "loss": 3.2779, + "step": 13165 + }, + { + "epoch": 0.6929473684210526, + "grad_norm": 0.46875, + "learning_rate": 6.823391812865497e-05, + "loss": 2.8839, + "step": 13166 + }, + { + "epoch": 0.693, + "grad_norm": 0.52734375, + "learning_rate": 6.822222222222222e-05, + "loss": 3.4065, + "step": 13167 + }, + { + "epoch": 0.6930526315789474, + "grad_norm": 0.4765625, + "learning_rate": 6.821052631578948e-05, + "loss": 3.361, + "step": 13168 + }, + { + "epoch": 0.6931052631578948, + "grad_norm": 0.48046875, + "learning_rate": 6.819883040935673e-05, + "loss": 3.6292, + "step": 13169 + }, + { + "epoch": 0.6931578947368421, + "grad_norm": 0.494140625, + "learning_rate": 6.818713450292398e-05, + "loss": 3.6327, + "step": 13170 + }, + { + "epoch": 0.6932105263157895, + "grad_norm": 0.48828125, + "learning_rate": 6.817543859649124e-05, + "loss": 3.2273, + "step": 13171 + }, + { + "epoch": 0.6932631578947368, + "grad_norm": 0.5, + "learning_rate": 6.816374269005849e-05, + "loss": 3.3866, + "step": 13172 + }, + { + "epoch": 0.6933157894736842, + "grad_norm": 0.48046875, + "learning_rate": 6.815204678362573e-05, + "loss": 3.4119, + "step": 13173 + }, + { + "epoch": 0.6933684210526316, + "grad_norm": 0.486328125, + "learning_rate": 6.814035087719298e-05, + "loss": 3.4383, + "step": 13174 + }, + { + "epoch": 0.6934210526315789, + "grad_norm": 0.50390625, + "learning_rate": 6.812865497076024e-05, + "loss": 3.1426, + "step": 13175 + }, + { + "epoch": 0.6934736842105264, + "grad_norm": 0.5, + "learning_rate": 6.811695906432749e-05, + "loss": 3.7167, + "step": 13176 + }, + { + "epoch": 0.6935263157894737, + "grad_norm": 0.609375, + "learning_rate": 6.810526315789474e-05, + "loss": 3.4487, + "step": 13177 + }, + { + "epoch": 0.6935789473684211, + "grad_norm": 0.51171875, + "learning_rate": 6.809356725146199e-05, + "loss": 3.1946, + "step": 13178 + }, + { + "epoch": 0.6936315789473684, + "grad_norm": 0.5390625, + "learning_rate": 6.808187134502925e-05, + "loss": 3.2752, + "step": 13179 + }, + { + "epoch": 0.6936842105263158, + "grad_norm": 0.474609375, + "learning_rate": 6.80701754385965e-05, + "loss": 3.0797, + "step": 13180 + }, + { + "epoch": 0.6937368421052632, + "grad_norm": 0.484375, + "learning_rate": 6.805847953216374e-05, + "loss": 3.0711, + "step": 13181 + }, + { + "epoch": 0.6937894736842105, + "grad_norm": 0.49609375, + "learning_rate": 6.804678362573099e-05, + "loss": 3.1996, + "step": 13182 + }, + { + "epoch": 0.6938421052631579, + "grad_norm": 0.48046875, + "learning_rate": 6.803508771929825e-05, + "loss": 2.8139, + "step": 13183 + }, + { + "epoch": 0.6938947368421052, + "grad_norm": 0.54296875, + "learning_rate": 6.80233918128655e-05, + "loss": 3.4261, + "step": 13184 + }, + { + "epoch": 0.6939473684210526, + "grad_norm": 0.494140625, + "learning_rate": 6.801169590643275e-05, + "loss": 3.2432, + "step": 13185 + }, + { + "epoch": 0.694, + "grad_norm": 0.48828125, + "learning_rate": 6.800000000000001e-05, + "loss": 3.4441, + "step": 13186 + }, + { + "epoch": 0.6940526315789474, + "grad_norm": 0.470703125, + "learning_rate": 6.798830409356726e-05, + "loss": 3.2769, + "step": 13187 + }, + { + "epoch": 0.6941052631578948, + "grad_norm": 0.48828125, + "learning_rate": 6.79766081871345e-05, + "loss": 3.4272, + "step": 13188 + }, + { + "epoch": 0.6941578947368421, + "grad_norm": 0.474609375, + "learning_rate": 6.796491228070175e-05, + "loss": 3.2779, + "step": 13189 + }, + { + "epoch": 0.6942105263157895, + "grad_norm": 0.478515625, + "learning_rate": 6.795321637426901e-05, + "loss": 3.1725, + "step": 13190 + }, + { + "epoch": 0.6942631578947368, + "grad_norm": 0.49609375, + "learning_rate": 6.794152046783626e-05, + "loss": 3.75, + "step": 13191 + }, + { + "epoch": 0.6943157894736842, + "grad_norm": 0.5234375, + "learning_rate": 6.792982456140351e-05, + "loss": 2.7853, + "step": 13192 + }, + { + "epoch": 0.6943684210526315, + "grad_norm": 0.494140625, + "learning_rate": 6.791812865497076e-05, + "loss": 3.0442, + "step": 13193 + }, + { + "epoch": 0.6944210526315789, + "grad_norm": 0.5078125, + "learning_rate": 6.790643274853802e-05, + "loss": 2.7933, + "step": 13194 + }, + { + "epoch": 0.6944736842105264, + "grad_norm": 0.478515625, + "learning_rate": 6.789473684210527e-05, + "loss": 3.7487, + "step": 13195 + }, + { + "epoch": 0.6945263157894737, + "grad_norm": 0.4921875, + "learning_rate": 6.788304093567251e-05, + "loss": 3.3435, + "step": 13196 + }, + { + "epoch": 0.6945789473684211, + "grad_norm": 0.484375, + "learning_rate": 6.787134502923977e-05, + "loss": 3.2943, + "step": 13197 + }, + { + "epoch": 0.6946315789473684, + "grad_norm": 0.52734375, + "learning_rate": 6.785964912280702e-05, + "loss": 3.1256, + "step": 13198 + }, + { + "epoch": 0.6946842105263158, + "grad_norm": 0.47265625, + "learning_rate": 6.784795321637427e-05, + "loss": 3.3475, + "step": 13199 + }, + { + "epoch": 0.6947368421052632, + "grad_norm": 0.48828125, + "learning_rate": 6.783625730994152e-05, + "loss": 2.8434, + "step": 13200 + }, + { + "epoch": 0.6947894736842105, + "grad_norm": 0.466796875, + "learning_rate": 6.782456140350878e-05, + "loss": 3.2527, + "step": 13201 + }, + { + "epoch": 0.6948421052631579, + "grad_norm": 0.484375, + "learning_rate": 6.781286549707603e-05, + "loss": 3.3557, + "step": 13202 + }, + { + "epoch": 0.6948947368421052, + "grad_norm": 1.765625, + "learning_rate": 6.780116959064328e-05, + "loss": 3.6004, + "step": 13203 + }, + { + "epoch": 0.6949473684210526, + "grad_norm": 0.515625, + "learning_rate": 6.778947368421052e-05, + "loss": 3.1272, + "step": 13204 + }, + { + "epoch": 0.695, + "grad_norm": 0.51171875, + "learning_rate": 6.777777777777778e-05, + "loss": 3.5928, + "step": 13205 + }, + { + "epoch": 0.6950526315789474, + "grad_norm": 0.4921875, + "learning_rate": 6.776608187134503e-05, + "loss": 3.5033, + "step": 13206 + }, + { + "epoch": 0.6951052631578948, + "grad_norm": 0.51953125, + "learning_rate": 6.775438596491228e-05, + "loss": 3.8083, + "step": 13207 + }, + { + "epoch": 0.6951578947368421, + "grad_norm": 0.5234375, + "learning_rate": 6.774269005847953e-05, + "loss": 3.124, + "step": 13208 + }, + { + "epoch": 0.6952105263157895, + "grad_norm": 0.458984375, + "learning_rate": 6.77309941520468e-05, + "loss": 3.3123, + "step": 13209 + }, + { + "epoch": 0.6952631578947368, + "grad_norm": 0.48046875, + "learning_rate": 6.771929824561404e-05, + "loss": 3.398, + "step": 13210 + }, + { + "epoch": 0.6953157894736842, + "grad_norm": 0.4765625, + "learning_rate": 6.770760233918128e-05, + "loss": 3.2429, + "step": 13211 + }, + { + "epoch": 0.6953684210526315, + "grad_norm": 0.4921875, + "learning_rate": 6.769590643274854e-05, + "loss": 2.6394, + "step": 13212 + }, + { + "epoch": 0.6954210526315789, + "grad_norm": 0.49609375, + "learning_rate": 6.768421052631579e-05, + "loss": 3.1113, + "step": 13213 + }, + { + "epoch": 0.6954736842105264, + "grad_norm": 0.48046875, + "learning_rate": 6.767251461988304e-05, + "loss": 3.3071, + "step": 13214 + }, + { + "epoch": 0.6955263157894737, + "grad_norm": 0.49609375, + "learning_rate": 6.766081871345029e-05, + "loss": 3.348, + "step": 13215 + }, + { + "epoch": 0.6955789473684211, + "grad_norm": 0.51171875, + "learning_rate": 6.764912280701756e-05, + "loss": 3.2024, + "step": 13216 + }, + { + "epoch": 0.6956315789473684, + "grad_norm": 0.46484375, + "learning_rate": 6.76374269005848e-05, + "loss": 3.2128, + "step": 13217 + }, + { + "epoch": 0.6956842105263158, + "grad_norm": 0.4921875, + "learning_rate": 6.762573099415205e-05, + "loss": 3.2872, + "step": 13218 + }, + { + "epoch": 0.6957368421052632, + "grad_norm": 0.4765625, + "learning_rate": 6.761403508771929e-05, + "loss": 3.4297, + "step": 13219 + }, + { + "epoch": 0.6957894736842105, + "grad_norm": 0.498046875, + "learning_rate": 6.760233918128655e-05, + "loss": 2.9646, + "step": 13220 + }, + { + "epoch": 0.6958421052631579, + "grad_norm": 0.4921875, + "learning_rate": 6.75906432748538e-05, + "loss": 3.3063, + "step": 13221 + }, + { + "epoch": 0.6958947368421052, + "grad_norm": 0.48828125, + "learning_rate": 6.757894736842105e-05, + "loss": 3.2811, + "step": 13222 + }, + { + "epoch": 0.6959473684210526, + "grad_norm": 0.52734375, + "learning_rate": 6.756725146198832e-05, + "loss": 3.1707, + "step": 13223 + }, + { + "epoch": 0.696, + "grad_norm": 0.45703125, + "learning_rate": 6.755555555555557e-05, + "loss": 3.4419, + "step": 13224 + }, + { + "epoch": 0.6960526315789474, + "grad_norm": 0.49609375, + "learning_rate": 6.754385964912281e-05, + "loss": 3.5731, + "step": 13225 + }, + { + "epoch": 0.6961052631578948, + "grad_norm": 0.47265625, + "learning_rate": 6.753216374269006e-05, + "loss": 2.8063, + "step": 13226 + }, + { + "epoch": 0.6961578947368421, + "grad_norm": 0.486328125, + "learning_rate": 6.752046783625731e-05, + "loss": 3.0544, + "step": 13227 + }, + { + "epoch": 0.6962105263157895, + "grad_norm": 0.466796875, + "learning_rate": 6.750877192982456e-05, + "loss": 3.112, + "step": 13228 + }, + { + "epoch": 0.6962631578947368, + "grad_norm": 0.46875, + "learning_rate": 6.749707602339181e-05, + "loss": 3.4833, + "step": 13229 + }, + { + "epoch": 0.6963157894736842, + "grad_norm": 0.46875, + "learning_rate": 6.748538011695906e-05, + "loss": 3.6801, + "step": 13230 + }, + { + "epoch": 0.6963684210526315, + "grad_norm": 0.482421875, + "learning_rate": 6.747368421052633e-05, + "loss": 3.1273, + "step": 13231 + }, + { + "epoch": 0.6964210526315789, + "grad_norm": 0.494140625, + "learning_rate": 6.746198830409358e-05, + "loss": 3.5843, + "step": 13232 + }, + { + "epoch": 0.6964736842105264, + "grad_norm": 0.515625, + "learning_rate": 6.745029239766083e-05, + "loss": 3.5299, + "step": 13233 + }, + { + "epoch": 0.6965263157894737, + "grad_norm": 0.50390625, + "learning_rate": 6.743859649122807e-05, + "loss": 3.3597, + "step": 13234 + }, + { + "epoch": 0.6965789473684211, + "grad_norm": 0.484375, + "learning_rate": 6.742690058479532e-05, + "loss": 3.376, + "step": 13235 + }, + { + "epoch": 0.6966315789473684, + "grad_norm": 0.466796875, + "learning_rate": 6.741520467836257e-05, + "loss": 3.2719, + "step": 13236 + }, + { + "epoch": 0.6966842105263158, + "grad_norm": 0.4921875, + "learning_rate": 6.740350877192982e-05, + "loss": 3.2265, + "step": 13237 + }, + { + "epoch": 0.6967368421052632, + "grad_norm": 0.4609375, + "learning_rate": 6.739181286549709e-05, + "loss": 3.3908, + "step": 13238 + }, + { + "epoch": 0.6967894736842105, + "grad_norm": 0.515625, + "learning_rate": 6.738011695906434e-05, + "loss": 3.0223, + "step": 13239 + }, + { + "epoch": 0.6968421052631579, + "grad_norm": 0.55078125, + "learning_rate": 6.736842105263159e-05, + "loss": 3.2502, + "step": 13240 + }, + { + "epoch": 0.6968947368421052, + "grad_norm": 0.486328125, + "learning_rate": 6.735672514619884e-05, + "loss": 3.1988, + "step": 13241 + }, + { + "epoch": 0.6969473684210526, + "grad_norm": 0.474609375, + "learning_rate": 6.734502923976609e-05, + "loss": 3.3895, + "step": 13242 + }, + { + "epoch": 0.697, + "grad_norm": 0.51171875, + "learning_rate": 6.733333333333333e-05, + "loss": 3.2026, + "step": 13243 + }, + { + "epoch": 0.6970526315789474, + "grad_norm": 0.51953125, + "learning_rate": 6.732163742690058e-05, + "loss": 3.6791, + "step": 13244 + }, + { + "epoch": 0.6971052631578948, + "grad_norm": 0.53125, + "learning_rate": 6.730994152046783e-05, + "loss": 3.0981, + "step": 13245 + }, + { + "epoch": 0.6971578947368421, + "grad_norm": 0.482421875, + "learning_rate": 6.72982456140351e-05, + "loss": 3.3523, + "step": 13246 + }, + { + "epoch": 0.6972105263157895, + "grad_norm": 0.49609375, + "learning_rate": 6.728654970760235e-05, + "loss": 3.1079, + "step": 13247 + }, + { + "epoch": 0.6972631578947368, + "grad_norm": 0.462890625, + "learning_rate": 6.72748538011696e-05, + "loss": 3.3231, + "step": 13248 + }, + { + "epoch": 0.6973157894736842, + "grad_norm": 0.53515625, + "learning_rate": 6.726315789473685e-05, + "loss": 3.2555, + "step": 13249 + }, + { + "epoch": 0.6973684210526315, + "grad_norm": 0.4765625, + "learning_rate": 6.72514619883041e-05, + "loss": 3.3315, + "step": 13250 + }, + { + "epoch": 0.6974210526315789, + "grad_norm": 0.490234375, + "learning_rate": 6.723976608187134e-05, + "loss": 2.8542, + "step": 13251 + }, + { + "epoch": 0.6974736842105264, + "grad_norm": 0.64453125, + "learning_rate": 6.72280701754386e-05, + "loss": 3.1677, + "step": 13252 + }, + { + "epoch": 0.6975263157894737, + "grad_norm": 0.59375, + "learning_rate": 6.721637426900586e-05, + "loss": 3.3222, + "step": 13253 + }, + { + "epoch": 0.6975789473684211, + "grad_norm": 0.5703125, + "learning_rate": 6.720467836257311e-05, + "loss": 3.2466, + "step": 13254 + }, + { + "epoch": 0.6976315789473684, + "grad_norm": 0.5078125, + "learning_rate": 6.719298245614036e-05, + "loss": 3.0033, + "step": 13255 + }, + { + "epoch": 0.6976842105263158, + "grad_norm": 0.482421875, + "learning_rate": 6.71812865497076e-05, + "loss": 3.3785, + "step": 13256 + }, + { + "epoch": 0.6977368421052632, + "grad_norm": 0.48828125, + "learning_rate": 6.716959064327486e-05, + "loss": 3.541, + "step": 13257 + }, + { + "epoch": 0.6977894736842105, + "grad_norm": 0.484375, + "learning_rate": 6.71578947368421e-05, + "loss": 3.5034, + "step": 13258 + }, + { + "epoch": 0.6978421052631579, + "grad_norm": 0.53125, + "learning_rate": 6.714619883040936e-05, + "loss": 3.451, + "step": 13259 + }, + { + "epoch": 0.6978947368421052, + "grad_norm": 0.51171875, + "learning_rate": 6.713450292397662e-05, + "loss": 3.1224, + "step": 13260 + }, + { + "epoch": 0.6979473684210526, + "grad_norm": 0.50390625, + "learning_rate": 6.712280701754387e-05, + "loss": 3.5525, + "step": 13261 + }, + { + "epoch": 0.698, + "grad_norm": 0.484375, + "learning_rate": 6.711111111111112e-05, + "loss": 3.3521, + "step": 13262 + }, + { + "epoch": 0.6980526315789474, + "grad_norm": 0.49609375, + "learning_rate": 6.709941520467837e-05, + "loss": 3.4398, + "step": 13263 + }, + { + "epoch": 0.6981052631578948, + "grad_norm": 0.5078125, + "learning_rate": 6.708771929824562e-05, + "loss": 3.0938, + "step": 13264 + }, + { + "epoch": 0.6981578947368421, + "grad_norm": 0.5, + "learning_rate": 6.707602339181287e-05, + "loss": 3.572, + "step": 13265 + }, + { + "epoch": 0.6982105263157895, + "grad_norm": 0.498046875, + "learning_rate": 6.706432748538012e-05, + "loss": 3.6063, + "step": 13266 + }, + { + "epoch": 0.6982631578947368, + "grad_norm": 0.474609375, + "learning_rate": 6.705263157894737e-05, + "loss": 3.473, + "step": 13267 + }, + { + "epoch": 0.6983157894736842, + "grad_norm": 0.46875, + "learning_rate": 6.704093567251463e-05, + "loss": 3.085, + "step": 13268 + }, + { + "epoch": 0.6983684210526315, + "grad_norm": 0.4921875, + "learning_rate": 6.702923976608188e-05, + "loss": 3.5232, + "step": 13269 + }, + { + "epoch": 0.6984210526315789, + "grad_norm": 0.48046875, + "learning_rate": 6.701754385964913e-05, + "loss": 3.3167, + "step": 13270 + }, + { + "epoch": 0.6984736842105264, + "grad_norm": 0.56640625, + "learning_rate": 6.700584795321638e-05, + "loss": 3.3554, + "step": 13271 + }, + { + "epoch": 0.6985263157894737, + "grad_norm": 0.5625, + "learning_rate": 6.699415204678363e-05, + "loss": 3.1477, + "step": 13272 + }, + { + "epoch": 0.6985789473684211, + "grad_norm": 0.5546875, + "learning_rate": 6.698245614035088e-05, + "loss": 3.6094, + "step": 13273 + }, + { + "epoch": 0.6986315789473684, + "grad_norm": 0.494140625, + "learning_rate": 6.697076023391813e-05, + "loss": 3.5764, + "step": 13274 + }, + { + "epoch": 0.6986842105263158, + "grad_norm": 0.5078125, + "learning_rate": 6.695906432748539e-05, + "loss": 3.3052, + "step": 13275 + }, + { + "epoch": 0.6987368421052632, + "grad_norm": 0.50390625, + "learning_rate": 6.694736842105264e-05, + "loss": 3.7002, + "step": 13276 + }, + { + "epoch": 0.6987894736842105, + "grad_norm": 0.462890625, + "learning_rate": 6.693567251461989e-05, + "loss": 3.2204, + "step": 13277 + }, + { + "epoch": 0.6988421052631579, + "grad_norm": 0.46875, + "learning_rate": 6.692397660818714e-05, + "loss": 3.0742, + "step": 13278 + }, + { + "epoch": 0.6988947368421052, + "grad_norm": 0.5234375, + "learning_rate": 6.691228070175439e-05, + "loss": 3.2907, + "step": 13279 + }, + { + "epoch": 0.6989473684210527, + "grad_norm": 0.494140625, + "learning_rate": 6.690058479532164e-05, + "loss": 3.5417, + "step": 13280 + }, + { + "epoch": 0.699, + "grad_norm": 0.5390625, + "learning_rate": 6.688888888888889e-05, + "loss": 3.2446, + "step": 13281 + }, + { + "epoch": 0.6990526315789474, + "grad_norm": 0.478515625, + "learning_rate": 6.687719298245614e-05, + "loss": 3.202, + "step": 13282 + }, + { + "epoch": 0.6991052631578948, + "grad_norm": 0.474609375, + "learning_rate": 6.68654970760234e-05, + "loss": 3.567, + "step": 13283 + }, + { + "epoch": 0.6991578947368421, + "grad_norm": 0.484375, + "learning_rate": 6.685380116959065e-05, + "loss": 3.5411, + "step": 13284 + }, + { + "epoch": 0.6992105263157895, + "grad_norm": 0.482421875, + "learning_rate": 6.68421052631579e-05, + "loss": 3.296, + "step": 13285 + }, + { + "epoch": 0.6992631578947368, + "grad_norm": 0.48046875, + "learning_rate": 6.683040935672515e-05, + "loss": 3.3835, + "step": 13286 + }, + { + "epoch": 0.6993157894736842, + "grad_norm": 0.5, + "learning_rate": 6.68187134502924e-05, + "loss": 3.434, + "step": 13287 + }, + { + "epoch": 0.6993684210526315, + "grad_norm": 0.50390625, + "learning_rate": 6.680701754385965e-05, + "loss": 3.4452, + "step": 13288 + }, + { + "epoch": 0.6994210526315789, + "grad_norm": 0.609375, + "learning_rate": 6.67953216374269e-05, + "loss": 3.5653, + "step": 13289 + }, + { + "epoch": 0.6994736842105264, + "grad_norm": 0.46484375, + "learning_rate": 6.678362573099416e-05, + "loss": 3.2394, + "step": 13290 + }, + { + "epoch": 0.6995263157894737, + "grad_norm": 0.48046875, + "learning_rate": 6.677192982456141e-05, + "loss": 3.2003, + "step": 13291 + }, + { + "epoch": 0.6995789473684211, + "grad_norm": 0.5078125, + "learning_rate": 6.676023391812866e-05, + "loss": 3.3785, + "step": 13292 + }, + { + "epoch": 0.6996315789473684, + "grad_norm": 0.51171875, + "learning_rate": 6.674853801169591e-05, + "loss": 3.7184, + "step": 13293 + }, + { + "epoch": 0.6996842105263158, + "grad_norm": 0.486328125, + "learning_rate": 6.673684210526316e-05, + "loss": 3.3157, + "step": 13294 + }, + { + "epoch": 0.6997368421052632, + "grad_norm": 0.51171875, + "learning_rate": 6.672514619883041e-05, + "loss": 3.5682, + "step": 13295 + }, + { + "epoch": 0.6997894736842105, + "grad_norm": 0.478515625, + "learning_rate": 6.671345029239766e-05, + "loss": 3.0406, + "step": 13296 + }, + { + "epoch": 0.6998421052631579, + "grad_norm": 0.486328125, + "learning_rate": 6.670175438596492e-05, + "loss": 3.6655, + "step": 13297 + }, + { + "epoch": 0.6998947368421052, + "grad_norm": 0.4609375, + "learning_rate": 6.669005847953217e-05, + "loss": 2.7096, + "step": 13298 + }, + { + "epoch": 0.6999473684210527, + "grad_norm": 0.486328125, + "learning_rate": 6.667836257309942e-05, + "loss": 2.8428, + "step": 13299 + }, + { + "epoch": 0.7, + "grad_norm": 0.47265625, + "learning_rate": 6.666666666666667e-05, + "loss": 3.4563, + "step": 13300 + }, + { + "epoch": 0.7000526315789474, + "grad_norm": 0.51171875, + "learning_rate": 6.665497076023392e-05, + "loss": 3.5031, + "step": 13301 + }, + { + "epoch": 0.7001052631578948, + "grad_norm": 0.5390625, + "learning_rate": 6.664327485380117e-05, + "loss": 2.8855, + "step": 13302 + }, + { + "epoch": 0.7001578947368421, + "grad_norm": 0.640625, + "learning_rate": 6.663157894736842e-05, + "loss": 2.876, + "step": 13303 + }, + { + "epoch": 0.7002105263157895, + "grad_norm": 0.50390625, + "learning_rate": 6.661988304093567e-05, + "loss": 3.1144, + "step": 13304 + }, + { + "epoch": 0.7002631578947368, + "grad_norm": 0.47265625, + "learning_rate": 6.660818713450293e-05, + "loss": 3.4529, + "step": 13305 + }, + { + "epoch": 0.7003157894736842, + "grad_norm": 0.48046875, + "learning_rate": 6.659649122807018e-05, + "loss": 3.6455, + "step": 13306 + }, + { + "epoch": 0.7003684210526315, + "grad_norm": 0.48046875, + "learning_rate": 6.658479532163743e-05, + "loss": 3.504, + "step": 13307 + }, + { + "epoch": 0.7004210526315789, + "grad_norm": 0.55078125, + "learning_rate": 6.657309941520468e-05, + "loss": 3.261, + "step": 13308 + }, + { + "epoch": 0.7004736842105264, + "grad_norm": 0.5, + "learning_rate": 6.656140350877193e-05, + "loss": 3.0661, + "step": 13309 + }, + { + "epoch": 0.7005263157894737, + "grad_norm": 0.50390625, + "learning_rate": 6.654970760233918e-05, + "loss": 3.3812, + "step": 13310 + }, + { + "epoch": 0.7005789473684211, + "grad_norm": 0.515625, + "learning_rate": 6.653801169590643e-05, + "loss": 3.3283, + "step": 13311 + }, + { + "epoch": 0.7006315789473684, + "grad_norm": 0.66796875, + "learning_rate": 6.652631578947369e-05, + "loss": 3.2502, + "step": 13312 + }, + { + "epoch": 0.7006842105263158, + "grad_norm": 0.50390625, + "learning_rate": 6.651461988304094e-05, + "loss": 2.8913, + "step": 13313 + }, + { + "epoch": 0.7007368421052631, + "grad_norm": 0.466796875, + "learning_rate": 6.650292397660819e-05, + "loss": 3.2116, + "step": 13314 + }, + { + "epoch": 0.7007894736842105, + "grad_norm": 0.5, + "learning_rate": 6.649122807017544e-05, + "loss": 3.121, + "step": 13315 + }, + { + "epoch": 0.7008421052631579, + "grad_norm": 0.482421875, + "learning_rate": 6.647953216374269e-05, + "loss": 3.133, + "step": 13316 + }, + { + "epoch": 0.7008947368421052, + "grad_norm": 0.5, + "learning_rate": 6.646783625730994e-05, + "loss": 3.0581, + "step": 13317 + }, + { + "epoch": 0.7009473684210527, + "grad_norm": 0.56640625, + "learning_rate": 6.645614035087719e-05, + "loss": 3.0832, + "step": 13318 + }, + { + "epoch": 0.701, + "grad_norm": 0.52734375, + "learning_rate": 6.644444444444444e-05, + "loss": 3.1598, + "step": 13319 + }, + { + "epoch": 0.7010526315789474, + "grad_norm": 0.455078125, + "learning_rate": 6.64327485380117e-05, + "loss": 2.9886, + "step": 13320 + }, + { + "epoch": 0.7011052631578948, + "grad_norm": 0.5, + "learning_rate": 6.642105263157895e-05, + "loss": 3.4944, + "step": 13321 + }, + { + "epoch": 0.7011578947368421, + "grad_norm": 0.5078125, + "learning_rate": 6.64093567251462e-05, + "loss": 3.1788, + "step": 13322 + }, + { + "epoch": 0.7012105263157895, + "grad_norm": 0.4921875, + "learning_rate": 6.639766081871346e-05, + "loss": 3.4203, + "step": 13323 + }, + { + "epoch": 0.7012631578947368, + "grad_norm": 0.52734375, + "learning_rate": 6.63859649122807e-05, + "loss": 3.2711, + "step": 13324 + }, + { + "epoch": 0.7013157894736842, + "grad_norm": 0.466796875, + "learning_rate": 6.637426900584795e-05, + "loss": 3.4004, + "step": 13325 + }, + { + "epoch": 0.7013684210526315, + "grad_norm": 0.4921875, + "learning_rate": 6.63625730994152e-05, + "loss": 3.3217, + "step": 13326 + }, + { + "epoch": 0.701421052631579, + "grad_norm": 0.52734375, + "learning_rate": 6.635087719298246e-05, + "loss": 3.1676, + "step": 13327 + }, + { + "epoch": 0.7014736842105264, + "grad_norm": 0.46875, + "learning_rate": 6.633918128654971e-05, + "loss": 3.2883, + "step": 13328 + }, + { + "epoch": 0.7015263157894737, + "grad_norm": 0.46875, + "learning_rate": 6.632748538011696e-05, + "loss": 3.3165, + "step": 13329 + }, + { + "epoch": 0.7015789473684211, + "grad_norm": 0.486328125, + "learning_rate": 6.631578947368421e-05, + "loss": 3.2255, + "step": 13330 + }, + { + "epoch": 0.7016315789473684, + "grad_norm": 0.494140625, + "learning_rate": 6.630409356725147e-05, + "loss": 2.9364, + "step": 13331 + }, + { + "epoch": 0.7016842105263158, + "grad_norm": 0.49609375, + "learning_rate": 6.629239766081871e-05, + "loss": 3.3268, + "step": 13332 + }, + { + "epoch": 0.7017368421052631, + "grad_norm": 0.5234375, + "learning_rate": 6.628070175438596e-05, + "loss": 3.1416, + "step": 13333 + }, + { + "epoch": 0.7017894736842105, + "grad_norm": 0.53515625, + "learning_rate": 6.626900584795322e-05, + "loss": 3.0013, + "step": 13334 + }, + { + "epoch": 0.7018421052631579, + "grad_norm": 0.48046875, + "learning_rate": 6.625730994152047e-05, + "loss": 3.3365, + "step": 13335 + }, + { + "epoch": 0.7018947368421052, + "grad_norm": 0.47265625, + "learning_rate": 6.624561403508772e-05, + "loss": 3.1776, + "step": 13336 + }, + { + "epoch": 0.7019473684210527, + "grad_norm": 0.48828125, + "learning_rate": 6.623391812865497e-05, + "loss": 2.9899, + "step": 13337 + }, + { + "epoch": 0.702, + "grad_norm": 0.7421875, + "learning_rate": 6.622222222222224e-05, + "loss": 3.5095, + "step": 13338 + }, + { + "epoch": 0.7020526315789474, + "grad_norm": 0.484375, + "learning_rate": 6.621052631578947e-05, + "loss": 3.471, + "step": 13339 + }, + { + "epoch": 0.7021052631578948, + "grad_norm": 0.466796875, + "learning_rate": 6.619883040935672e-05, + "loss": 3.1867, + "step": 13340 + }, + { + "epoch": 0.7021578947368421, + "grad_norm": 0.5078125, + "learning_rate": 6.618713450292397e-05, + "loss": 3.5152, + "step": 13341 + }, + { + "epoch": 0.7022105263157895, + "grad_norm": 0.4765625, + "learning_rate": 6.617543859649123e-05, + "loss": 2.9148, + "step": 13342 + }, + { + "epoch": 0.7022631578947368, + "grad_norm": 0.482421875, + "learning_rate": 6.616374269005848e-05, + "loss": 3.2138, + "step": 13343 + }, + { + "epoch": 0.7023157894736842, + "grad_norm": 0.51953125, + "learning_rate": 6.615204678362573e-05, + "loss": 3.0268, + "step": 13344 + }, + { + "epoch": 0.7023684210526315, + "grad_norm": 0.640625, + "learning_rate": 6.614035087719298e-05, + "loss": 3.709, + "step": 13345 + }, + { + "epoch": 0.702421052631579, + "grad_norm": 0.51171875, + "learning_rate": 6.612865497076025e-05, + "loss": 3.3161, + "step": 13346 + }, + { + "epoch": 0.7024736842105264, + "grad_norm": 0.48046875, + "learning_rate": 6.611695906432748e-05, + "loss": 3.2698, + "step": 13347 + }, + { + "epoch": 0.7025263157894737, + "grad_norm": 0.50390625, + "learning_rate": 6.610526315789473e-05, + "loss": 2.8358, + "step": 13348 + }, + { + "epoch": 0.7025789473684211, + "grad_norm": 0.470703125, + "learning_rate": 6.6093567251462e-05, + "loss": 3.28, + "step": 13349 + }, + { + "epoch": 0.7026315789473684, + "grad_norm": 0.50390625, + "learning_rate": 6.608187134502924e-05, + "loss": 3.2395, + "step": 13350 + }, + { + "epoch": 0.7026842105263158, + "grad_norm": 0.466796875, + "learning_rate": 6.60701754385965e-05, + "loss": 3.5775, + "step": 13351 + }, + { + "epoch": 0.7027368421052631, + "grad_norm": 0.5, + "learning_rate": 6.605847953216374e-05, + "loss": 3.2249, + "step": 13352 + }, + { + "epoch": 0.7027894736842105, + "grad_norm": 0.48828125, + "learning_rate": 6.6046783625731e-05, + "loss": 3.556, + "step": 13353 + }, + { + "epoch": 0.7028421052631579, + "grad_norm": 0.4921875, + "learning_rate": 6.603508771929826e-05, + "loss": 3.3025, + "step": 13354 + }, + { + "epoch": 0.7028947368421052, + "grad_norm": 0.609375, + "learning_rate": 6.602339181286549e-05, + "loss": 3.3414, + "step": 13355 + }, + { + "epoch": 0.7029473684210527, + "grad_norm": 0.4921875, + "learning_rate": 6.601169590643274e-05, + "loss": 3.62, + "step": 13356 + }, + { + "epoch": 0.703, + "grad_norm": 0.5234375, + "learning_rate": 6.6e-05, + "loss": 3.0253, + "step": 13357 + }, + { + "epoch": 0.7030526315789474, + "grad_norm": 0.4921875, + "learning_rate": 6.598830409356725e-05, + "loss": 3.4275, + "step": 13358 + }, + { + "epoch": 0.7031052631578948, + "grad_norm": 0.515625, + "learning_rate": 6.59766081871345e-05, + "loss": 2.796, + "step": 13359 + }, + { + "epoch": 0.7031578947368421, + "grad_norm": 0.5078125, + "learning_rate": 6.596491228070177e-05, + "loss": 3.5027, + "step": 13360 + }, + { + "epoch": 0.7032105263157895, + "grad_norm": 0.494140625, + "learning_rate": 6.595321637426902e-05, + "loss": 3.27, + "step": 13361 + }, + { + "epoch": 0.7032631578947368, + "grad_norm": 0.498046875, + "learning_rate": 6.594152046783625e-05, + "loss": 3.1309, + "step": 13362 + }, + { + "epoch": 0.7033157894736842, + "grad_norm": 0.486328125, + "learning_rate": 6.59298245614035e-05, + "loss": 2.8649, + "step": 13363 + }, + { + "epoch": 0.7033684210526315, + "grad_norm": 0.48046875, + "learning_rate": 6.591812865497077e-05, + "loss": 3.5438, + "step": 13364 + }, + { + "epoch": 0.703421052631579, + "grad_norm": 0.51171875, + "learning_rate": 6.590643274853802e-05, + "loss": 3.2679, + "step": 13365 + }, + { + "epoch": 0.7034736842105264, + "grad_norm": 0.50390625, + "learning_rate": 6.589473684210526e-05, + "loss": 3.355, + "step": 13366 + }, + { + "epoch": 0.7035263157894737, + "grad_norm": 0.494140625, + "learning_rate": 6.588304093567251e-05, + "loss": 3.1426, + "step": 13367 + }, + { + "epoch": 0.7035789473684211, + "grad_norm": 0.5859375, + "learning_rate": 6.587134502923978e-05, + "loss": 3.3678, + "step": 13368 + }, + { + "epoch": 0.7036315789473684, + "grad_norm": 0.486328125, + "learning_rate": 6.585964912280703e-05, + "loss": 3.5581, + "step": 13369 + }, + { + "epoch": 0.7036842105263158, + "grad_norm": 0.50390625, + "learning_rate": 6.584795321637426e-05, + "loss": 3.3522, + "step": 13370 + }, + { + "epoch": 0.7037368421052631, + "grad_norm": 0.4921875, + "learning_rate": 6.583625730994151e-05, + "loss": 3.4253, + "step": 13371 + }, + { + "epoch": 0.7037894736842105, + "grad_norm": 0.482421875, + "learning_rate": 6.582456140350878e-05, + "loss": 3.0325, + "step": 13372 + }, + { + "epoch": 0.7038421052631579, + "grad_norm": 0.51171875, + "learning_rate": 6.581286549707603e-05, + "loss": 2.698, + "step": 13373 + }, + { + "epoch": 0.7038947368421052, + "grad_norm": 0.546875, + "learning_rate": 6.580116959064328e-05, + "loss": 3.1537, + "step": 13374 + }, + { + "epoch": 0.7039473684210527, + "grad_norm": 0.50390625, + "learning_rate": 6.578947368421054e-05, + "loss": 3.5968, + "step": 13375 + }, + { + "epoch": 0.704, + "grad_norm": 0.515625, + "learning_rate": 6.577777777777779e-05, + "loss": 2.935, + "step": 13376 + }, + { + "epoch": 0.7040526315789474, + "grad_norm": 0.486328125, + "learning_rate": 6.576608187134504e-05, + "loss": 2.9886, + "step": 13377 + }, + { + "epoch": 0.7041052631578948, + "grad_norm": 0.48828125, + "learning_rate": 6.575438596491227e-05, + "loss": 3.2862, + "step": 13378 + }, + { + "epoch": 0.7041578947368421, + "grad_norm": 0.482421875, + "learning_rate": 6.574269005847954e-05, + "loss": 3.2038, + "step": 13379 + }, + { + "epoch": 0.7042105263157895, + "grad_norm": 0.48828125, + "learning_rate": 6.573099415204679e-05, + "loss": 3.5701, + "step": 13380 + }, + { + "epoch": 0.7042631578947368, + "grad_norm": 0.484375, + "learning_rate": 6.571929824561404e-05, + "loss": 3.4886, + "step": 13381 + }, + { + "epoch": 0.7043157894736842, + "grad_norm": 0.486328125, + "learning_rate": 6.570760233918129e-05, + "loss": 3.2563, + "step": 13382 + }, + { + "epoch": 0.7043684210526315, + "grad_norm": 0.458984375, + "learning_rate": 6.569590643274855e-05, + "loss": 3.081, + "step": 13383 + }, + { + "epoch": 0.704421052631579, + "grad_norm": 0.490234375, + "learning_rate": 6.56842105263158e-05, + "loss": 2.91, + "step": 13384 + }, + { + "epoch": 0.7044736842105264, + "grad_norm": 0.5703125, + "learning_rate": 6.567251461988303e-05, + "loss": 3.2947, + "step": 13385 + }, + { + "epoch": 0.7045263157894737, + "grad_norm": 0.5078125, + "learning_rate": 6.56608187134503e-05, + "loss": 2.7272, + "step": 13386 + }, + { + "epoch": 0.7045789473684211, + "grad_norm": 0.4921875, + "learning_rate": 6.564912280701755e-05, + "loss": 3.5306, + "step": 13387 + }, + { + "epoch": 0.7046315789473684, + "grad_norm": 0.4765625, + "learning_rate": 6.56374269005848e-05, + "loss": 3.5387, + "step": 13388 + }, + { + "epoch": 0.7046842105263158, + "grad_norm": 0.5234375, + "learning_rate": 6.562573099415205e-05, + "loss": 3.4638, + "step": 13389 + }, + { + "epoch": 0.7047368421052631, + "grad_norm": 0.490234375, + "learning_rate": 6.561403508771931e-05, + "loss": 3.4197, + "step": 13390 + }, + { + "epoch": 0.7047894736842105, + "grad_norm": 0.4921875, + "learning_rate": 6.560233918128656e-05, + "loss": 3.4819, + "step": 13391 + }, + { + "epoch": 0.7048421052631579, + "grad_norm": 0.478515625, + "learning_rate": 6.559064327485381e-05, + "loss": 3.1751, + "step": 13392 + }, + { + "epoch": 0.7048947368421052, + "grad_norm": 0.49609375, + "learning_rate": 6.557894736842104e-05, + "loss": 3.2362, + "step": 13393 + }, + { + "epoch": 0.7049473684210527, + "grad_norm": 0.515625, + "learning_rate": 6.556725146198831e-05, + "loss": 3.5655, + "step": 13394 + }, + { + "epoch": 0.705, + "grad_norm": 0.494140625, + "learning_rate": 6.555555555555556e-05, + "loss": 2.8323, + "step": 13395 + }, + { + "epoch": 0.7050526315789474, + "grad_norm": 0.5390625, + "learning_rate": 6.554385964912281e-05, + "loss": 3.5238, + "step": 13396 + }, + { + "epoch": 0.7051052631578948, + "grad_norm": 0.58203125, + "learning_rate": 6.553216374269007e-05, + "loss": 3.8556, + "step": 13397 + }, + { + "epoch": 0.7051578947368421, + "grad_norm": 0.609375, + "learning_rate": 6.552046783625732e-05, + "loss": 3.4227, + "step": 13398 + }, + { + "epoch": 0.7052105263157895, + "grad_norm": 0.46484375, + "learning_rate": 6.550877192982457e-05, + "loss": 2.9723, + "step": 13399 + }, + { + "epoch": 0.7052631578947368, + "grad_norm": 0.53515625, + "learning_rate": 6.549707602339182e-05, + "loss": 3.3359, + "step": 13400 + }, + { + "epoch": 0.7053157894736842, + "grad_norm": 0.486328125, + "learning_rate": 6.548538011695907e-05, + "loss": 3.2807, + "step": 13401 + }, + { + "epoch": 0.7053684210526315, + "grad_norm": 0.466796875, + "learning_rate": 6.547368421052632e-05, + "loss": 3.109, + "step": 13402 + }, + { + "epoch": 0.705421052631579, + "grad_norm": 0.56640625, + "learning_rate": 6.546198830409357e-05, + "loss": 3.0092, + "step": 13403 + }, + { + "epoch": 0.7054736842105264, + "grad_norm": 0.515625, + "learning_rate": 6.545029239766082e-05, + "loss": 3.4802, + "step": 13404 + }, + { + "epoch": 0.7055263157894737, + "grad_norm": 0.498046875, + "learning_rate": 6.543859649122808e-05, + "loss": 3.1708, + "step": 13405 + }, + { + "epoch": 0.7055789473684211, + "grad_norm": 0.50390625, + "learning_rate": 6.542690058479533e-05, + "loss": 2.759, + "step": 13406 + }, + { + "epoch": 0.7056315789473684, + "grad_norm": 0.48046875, + "learning_rate": 6.541520467836258e-05, + "loss": 3.3767, + "step": 13407 + }, + { + "epoch": 0.7056842105263158, + "grad_norm": 0.55078125, + "learning_rate": 6.540350877192982e-05, + "loss": 3.4392, + "step": 13408 + }, + { + "epoch": 0.7057368421052631, + "grad_norm": 0.47265625, + "learning_rate": 6.539181286549708e-05, + "loss": 3.5531, + "step": 13409 + }, + { + "epoch": 0.7057894736842105, + "grad_norm": 0.515625, + "learning_rate": 6.538011695906433e-05, + "loss": 3.4048, + "step": 13410 + }, + { + "epoch": 0.7058421052631579, + "grad_norm": 0.498046875, + "learning_rate": 6.536842105263158e-05, + "loss": 3.2551, + "step": 13411 + }, + { + "epoch": 0.7058947368421052, + "grad_norm": 0.462890625, + "learning_rate": 6.535672514619884e-05, + "loss": 3.7121, + "step": 13412 + }, + { + "epoch": 0.7059473684210527, + "grad_norm": 0.51171875, + "learning_rate": 6.534502923976609e-05, + "loss": 3.0984, + "step": 13413 + }, + { + "epoch": 0.706, + "grad_norm": 0.4609375, + "learning_rate": 6.533333333333334e-05, + "loss": 3.3831, + "step": 13414 + }, + { + "epoch": 0.7060526315789474, + "grad_norm": 0.47265625, + "learning_rate": 6.532163742690059e-05, + "loss": 3.8576, + "step": 13415 + }, + { + "epoch": 0.7061052631578948, + "grad_norm": 0.46875, + "learning_rate": 6.530994152046784e-05, + "loss": 3.1491, + "step": 13416 + }, + { + "epoch": 0.7061578947368421, + "grad_norm": 0.470703125, + "learning_rate": 6.529824561403509e-05, + "loss": 3.0523, + "step": 13417 + }, + { + "epoch": 0.7062105263157895, + "grad_norm": 0.50390625, + "learning_rate": 6.528654970760234e-05, + "loss": 2.791, + "step": 13418 + }, + { + "epoch": 0.7062631578947368, + "grad_norm": 0.50390625, + "learning_rate": 6.527485380116959e-05, + "loss": 3.3597, + "step": 13419 + }, + { + "epoch": 0.7063157894736842, + "grad_norm": 0.474609375, + "learning_rate": 6.526315789473685e-05, + "loss": 3.183, + "step": 13420 + }, + { + "epoch": 0.7063684210526315, + "grad_norm": 0.5078125, + "learning_rate": 6.52514619883041e-05, + "loss": 3.1818, + "step": 13421 + }, + { + "epoch": 0.706421052631579, + "grad_norm": 0.48828125, + "learning_rate": 6.523976608187135e-05, + "loss": 3.3553, + "step": 13422 + }, + { + "epoch": 0.7064736842105264, + "grad_norm": 0.59765625, + "learning_rate": 6.52280701754386e-05, + "loss": 3.3746, + "step": 13423 + }, + { + "epoch": 0.7065263157894737, + "grad_norm": 0.5546875, + "learning_rate": 6.521637426900585e-05, + "loss": 3.5916, + "step": 13424 + }, + { + "epoch": 0.7065789473684211, + "grad_norm": 0.486328125, + "learning_rate": 6.52046783625731e-05, + "loss": 3.4135, + "step": 13425 + }, + { + "epoch": 0.7066315789473684, + "grad_norm": 0.515625, + "learning_rate": 6.519298245614035e-05, + "loss": 3.129, + "step": 13426 + }, + { + "epoch": 0.7066842105263158, + "grad_norm": 0.490234375, + "learning_rate": 6.518128654970761e-05, + "loss": 3.3307, + "step": 13427 + }, + { + "epoch": 0.7067368421052631, + "grad_norm": 0.58203125, + "learning_rate": 6.516959064327486e-05, + "loss": 2.9752, + "step": 13428 + }, + { + "epoch": 0.7067894736842105, + "grad_norm": 0.482421875, + "learning_rate": 6.515789473684211e-05, + "loss": 2.8108, + "step": 13429 + }, + { + "epoch": 0.7068421052631579, + "grad_norm": 0.55859375, + "learning_rate": 6.514619883040936e-05, + "loss": 2.9042, + "step": 13430 + }, + { + "epoch": 0.7068947368421052, + "grad_norm": 0.455078125, + "learning_rate": 6.513450292397661e-05, + "loss": 3.2698, + "step": 13431 + }, + { + "epoch": 0.7069473684210527, + "grad_norm": 0.4765625, + "learning_rate": 6.512280701754386e-05, + "loss": 3.1273, + "step": 13432 + }, + { + "epoch": 0.707, + "grad_norm": 0.50390625, + "learning_rate": 6.511111111111111e-05, + "loss": 3.7859, + "step": 13433 + }, + { + "epoch": 0.7070526315789474, + "grad_norm": 0.5234375, + "learning_rate": 6.509941520467837e-05, + "loss": 3.4136, + "step": 13434 + }, + { + "epoch": 0.7071052631578948, + "grad_norm": 0.51171875, + "learning_rate": 6.508771929824562e-05, + "loss": 3.415, + "step": 13435 + }, + { + "epoch": 0.7071578947368421, + "grad_norm": 0.484375, + "learning_rate": 6.507602339181287e-05, + "loss": 3.5192, + "step": 13436 + }, + { + "epoch": 0.7072105263157895, + "grad_norm": 0.490234375, + "learning_rate": 6.506432748538012e-05, + "loss": 3.4715, + "step": 13437 + }, + { + "epoch": 0.7072631578947368, + "grad_norm": 0.55078125, + "learning_rate": 6.505263157894737e-05, + "loss": 3.2004, + "step": 13438 + }, + { + "epoch": 0.7073157894736842, + "grad_norm": 0.474609375, + "learning_rate": 6.504093567251462e-05, + "loss": 3.1485, + "step": 13439 + }, + { + "epoch": 0.7073684210526315, + "grad_norm": 0.50390625, + "learning_rate": 6.502923976608187e-05, + "loss": 3.6264, + "step": 13440 + }, + { + "epoch": 0.707421052631579, + "grad_norm": 0.486328125, + "learning_rate": 6.501754385964912e-05, + "loss": 3.1456, + "step": 13441 + }, + { + "epoch": 0.7074736842105264, + "grad_norm": 0.498046875, + "learning_rate": 6.500584795321638e-05, + "loss": 3.3917, + "step": 13442 + }, + { + "epoch": 0.7075263157894737, + "grad_norm": 0.466796875, + "learning_rate": 6.499415204678363e-05, + "loss": 3.1474, + "step": 13443 + }, + { + "epoch": 0.7075789473684211, + "grad_norm": 0.5, + "learning_rate": 6.498245614035088e-05, + "loss": 3.0977, + "step": 13444 + }, + { + "epoch": 0.7076315789473684, + "grad_norm": 0.51171875, + "learning_rate": 6.497076023391813e-05, + "loss": 3.3284, + "step": 13445 + }, + { + "epoch": 0.7076842105263158, + "grad_norm": 0.5, + "learning_rate": 6.495906432748538e-05, + "loss": 3.4702, + "step": 13446 + }, + { + "epoch": 0.7077368421052631, + "grad_norm": 0.46875, + "learning_rate": 6.494736842105263e-05, + "loss": 3.1993, + "step": 13447 + }, + { + "epoch": 0.7077894736842105, + "grad_norm": 0.46875, + "learning_rate": 6.493567251461988e-05, + "loss": 2.8602, + "step": 13448 + }, + { + "epoch": 0.7078421052631579, + "grad_norm": 0.48046875, + "learning_rate": 6.492397660818714e-05, + "loss": 3.3385, + "step": 13449 + }, + { + "epoch": 0.7078947368421052, + "grad_norm": 0.46875, + "learning_rate": 6.49122807017544e-05, + "loss": 3.0923, + "step": 13450 + }, + { + "epoch": 0.7079473684210527, + "grad_norm": 0.49609375, + "learning_rate": 6.490058479532164e-05, + "loss": 2.9798, + "step": 13451 + }, + { + "epoch": 0.708, + "grad_norm": 0.4765625, + "learning_rate": 6.488888888888889e-05, + "loss": 3.5655, + "step": 13452 + }, + { + "epoch": 0.7080526315789474, + "grad_norm": 0.65234375, + "learning_rate": 6.487719298245614e-05, + "loss": 3.0646, + "step": 13453 + }, + { + "epoch": 0.7081052631578947, + "grad_norm": 0.48046875, + "learning_rate": 6.486549707602339e-05, + "loss": 3.6881, + "step": 13454 + }, + { + "epoch": 0.7081578947368421, + "grad_norm": 0.484375, + "learning_rate": 6.485380116959064e-05, + "loss": 3.2104, + "step": 13455 + }, + { + "epoch": 0.7082105263157895, + "grad_norm": 0.490234375, + "learning_rate": 6.484210526315789e-05, + "loss": 3.4781, + "step": 13456 + }, + { + "epoch": 0.7082631578947368, + "grad_norm": 0.484375, + "learning_rate": 6.483040935672515e-05, + "loss": 3.019, + "step": 13457 + }, + { + "epoch": 0.7083157894736842, + "grad_norm": 0.4921875, + "learning_rate": 6.48187134502924e-05, + "loss": 3.1681, + "step": 13458 + }, + { + "epoch": 0.7083684210526315, + "grad_norm": 0.4765625, + "learning_rate": 6.480701754385965e-05, + "loss": 3.2959, + "step": 13459 + }, + { + "epoch": 0.708421052631579, + "grad_norm": 0.486328125, + "learning_rate": 6.47953216374269e-05, + "loss": 3.743, + "step": 13460 + }, + { + "epoch": 0.7084736842105264, + "grad_norm": 0.486328125, + "learning_rate": 6.478362573099415e-05, + "loss": 3.1063, + "step": 13461 + }, + { + "epoch": 0.7085263157894737, + "grad_norm": 0.484375, + "learning_rate": 6.47719298245614e-05, + "loss": 3.6091, + "step": 13462 + }, + { + "epoch": 0.7085789473684211, + "grad_norm": 0.49609375, + "learning_rate": 6.476023391812865e-05, + "loss": 2.5808, + "step": 13463 + }, + { + "epoch": 0.7086315789473684, + "grad_norm": 0.48046875, + "learning_rate": 6.474853801169591e-05, + "loss": 3.2188, + "step": 13464 + }, + { + "epoch": 0.7086842105263158, + "grad_norm": 0.458984375, + "learning_rate": 6.473684210526316e-05, + "loss": 3.448, + "step": 13465 + }, + { + "epoch": 0.7087368421052631, + "grad_norm": 0.470703125, + "learning_rate": 6.472514619883041e-05, + "loss": 3.2502, + "step": 13466 + }, + { + "epoch": 0.7087894736842105, + "grad_norm": 0.48828125, + "learning_rate": 6.471345029239766e-05, + "loss": 3.2872, + "step": 13467 + }, + { + "epoch": 0.7088421052631579, + "grad_norm": 0.47265625, + "learning_rate": 6.470175438596491e-05, + "loss": 3.0833, + "step": 13468 + }, + { + "epoch": 0.7088947368421052, + "grad_norm": 0.4765625, + "learning_rate": 6.469005847953216e-05, + "loss": 3.1898, + "step": 13469 + }, + { + "epoch": 0.7089473684210527, + "grad_norm": 0.51171875, + "learning_rate": 6.467836257309941e-05, + "loss": 3.2707, + "step": 13470 + }, + { + "epoch": 0.709, + "grad_norm": 0.50390625, + "learning_rate": 6.466666666666666e-05, + "loss": 3.1972, + "step": 13471 + }, + { + "epoch": 0.7090526315789474, + "grad_norm": 0.51171875, + "learning_rate": 6.465497076023393e-05, + "loss": 3.1286, + "step": 13472 + }, + { + "epoch": 0.7091052631578947, + "grad_norm": 0.546875, + "learning_rate": 6.464327485380117e-05, + "loss": 2.8747, + "step": 13473 + }, + { + "epoch": 0.7091578947368421, + "grad_norm": 0.494140625, + "learning_rate": 6.463157894736842e-05, + "loss": 3.3511, + "step": 13474 + }, + { + "epoch": 0.7092105263157895, + "grad_norm": 0.515625, + "learning_rate": 6.461988304093567e-05, + "loss": 3.3814, + "step": 13475 + }, + { + "epoch": 0.7092631578947368, + "grad_norm": 0.5, + "learning_rate": 6.460818713450292e-05, + "loss": 3.3316, + "step": 13476 + }, + { + "epoch": 0.7093157894736842, + "grad_norm": 0.5, + "learning_rate": 6.459649122807017e-05, + "loss": 3.251, + "step": 13477 + }, + { + "epoch": 0.7093684210526315, + "grad_norm": 0.50390625, + "learning_rate": 6.458479532163742e-05, + "loss": 3.0874, + "step": 13478 + }, + { + "epoch": 0.709421052631579, + "grad_norm": 0.486328125, + "learning_rate": 6.457309941520469e-05, + "loss": 3.6273, + "step": 13479 + }, + { + "epoch": 0.7094736842105264, + "grad_norm": 0.5234375, + "learning_rate": 6.456140350877194e-05, + "loss": 3.4359, + "step": 13480 + }, + { + "epoch": 0.7095263157894737, + "grad_norm": 0.478515625, + "learning_rate": 6.454970760233918e-05, + "loss": 3.429, + "step": 13481 + }, + { + "epoch": 0.7095789473684211, + "grad_norm": 0.486328125, + "learning_rate": 6.453801169590643e-05, + "loss": 3.449, + "step": 13482 + }, + { + "epoch": 0.7096315789473684, + "grad_norm": 0.490234375, + "learning_rate": 6.452631578947368e-05, + "loss": 3.362, + "step": 13483 + }, + { + "epoch": 0.7096842105263158, + "grad_norm": 0.490234375, + "learning_rate": 6.451461988304093e-05, + "loss": 3.3194, + "step": 13484 + }, + { + "epoch": 0.7097368421052631, + "grad_norm": 0.470703125, + "learning_rate": 6.450292397660818e-05, + "loss": 3.3149, + "step": 13485 + }, + { + "epoch": 0.7097894736842105, + "grad_norm": 0.478515625, + "learning_rate": 6.449122807017545e-05, + "loss": 3.3039, + "step": 13486 + }, + { + "epoch": 0.7098421052631579, + "grad_norm": 0.484375, + "learning_rate": 6.44795321637427e-05, + "loss": 3.5627, + "step": 13487 + }, + { + "epoch": 0.7098947368421052, + "grad_norm": 0.494140625, + "learning_rate": 6.446783625730995e-05, + "loss": 3.4988, + "step": 13488 + }, + { + "epoch": 0.7099473684210527, + "grad_norm": 0.6484375, + "learning_rate": 6.44561403508772e-05, + "loss": 3.7321, + "step": 13489 + }, + { + "epoch": 0.71, + "grad_norm": 0.50390625, + "learning_rate": 6.444444444444446e-05, + "loss": 3.1931, + "step": 13490 + }, + { + "epoch": 0.7100526315789474, + "grad_norm": 0.484375, + "learning_rate": 6.44327485380117e-05, + "loss": 3.3481, + "step": 13491 + }, + { + "epoch": 0.7101052631578947, + "grad_norm": 0.546875, + "learning_rate": 6.442105263157894e-05, + "loss": 3.1911, + "step": 13492 + }, + { + "epoch": 0.7101578947368421, + "grad_norm": 0.470703125, + "learning_rate": 6.44093567251462e-05, + "loss": 2.8432, + "step": 13493 + }, + { + "epoch": 0.7102105263157895, + "grad_norm": 0.484375, + "learning_rate": 6.439766081871346e-05, + "loss": 3.8467, + "step": 13494 + }, + { + "epoch": 0.7102631578947368, + "grad_norm": 0.50390625, + "learning_rate": 6.43859649122807e-05, + "loss": 3.0732, + "step": 13495 + }, + { + "epoch": 0.7103157894736842, + "grad_norm": 0.494140625, + "learning_rate": 6.437426900584796e-05, + "loss": 3.1645, + "step": 13496 + }, + { + "epoch": 0.7103684210526315, + "grad_norm": 0.474609375, + "learning_rate": 6.436257309941522e-05, + "loss": 3.3311, + "step": 13497 + }, + { + "epoch": 0.710421052631579, + "grad_norm": 0.5234375, + "learning_rate": 6.435087719298246e-05, + "loss": 3.2486, + "step": 13498 + }, + { + "epoch": 0.7104736842105264, + "grad_norm": 0.494140625, + "learning_rate": 6.43391812865497e-05, + "loss": 3.1769, + "step": 13499 + }, + { + "epoch": 0.7105263157894737, + "grad_norm": 0.51953125, + "learning_rate": 6.432748538011695e-05, + "loss": 3.4378, + "step": 13500 + }, + { + "epoch": 0.7105789473684211, + "grad_norm": 0.490234375, + "learning_rate": 6.431578947368422e-05, + "loss": 3.349, + "step": 13501 + }, + { + "epoch": 0.7106315789473684, + "grad_norm": 0.47265625, + "learning_rate": 6.430409356725147e-05, + "loss": 3.2804, + "step": 13502 + }, + { + "epoch": 0.7106842105263158, + "grad_norm": 0.474609375, + "learning_rate": 6.429239766081872e-05, + "loss": 3.2543, + "step": 13503 + }, + { + "epoch": 0.7107368421052631, + "grad_norm": 0.5078125, + "learning_rate": 6.428070175438597e-05, + "loss": 3.6822, + "step": 13504 + }, + { + "epoch": 0.7107894736842105, + "grad_norm": 0.486328125, + "learning_rate": 6.426900584795323e-05, + "loss": 3.5416, + "step": 13505 + }, + { + "epoch": 0.7108421052631579, + "grad_norm": 0.46484375, + "learning_rate": 6.425730994152047e-05, + "loss": 3.0791, + "step": 13506 + }, + { + "epoch": 0.7108947368421052, + "grad_norm": 0.51953125, + "learning_rate": 6.424561403508772e-05, + "loss": 3.6336, + "step": 13507 + }, + { + "epoch": 0.7109473684210527, + "grad_norm": 0.5078125, + "learning_rate": 6.423391812865496e-05, + "loss": 3.1578, + "step": 13508 + }, + { + "epoch": 0.711, + "grad_norm": 0.486328125, + "learning_rate": 6.422222222222223e-05, + "loss": 3.1595, + "step": 13509 + }, + { + "epoch": 0.7110526315789474, + "grad_norm": 0.53125, + "learning_rate": 6.421052631578948e-05, + "loss": 3.35, + "step": 13510 + }, + { + "epoch": 0.7111052631578947, + "grad_norm": 0.51171875, + "learning_rate": 6.419883040935673e-05, + "loss": 3.2175, + "step": 13511 + }, + { + "epoch": 0.7111578947368421, + "grad_norm": 0.498046875, + "learning_rate": 6.418713450292399e-05, + "loss": 3.3222, + "step": 13512 + }, + { + "epoch": 0.7112105263157895, + "grad_norm": 0.578125, + "learning_rate": 6.417543859649124e-05, + "loss": 3.0655, + "step": 13513 + }, + { + "epoch": 0.7112631578947368, + "grad_norm": 0.4921875, + "learning_rate": 6.416374269005848e-05, + "loss": 3.2288, + "step": 13514 + }, + { + "epoch": 0.7113157894736842, + "grad_norm": 0.5, + "learning_rate": 6.415204678362573e-05, + "loss": 3.7475, + "step": 13515 + }, + { + "epoch": 0.7113684210526315, + "grad_norm": 0.49609375, + "learning_rate": 6.414035087719299e-05, + "loss": 3.7258, + "step": 13516 + }, + { + "epoch": 0.711421052631579, + "grad_norm": 0.490234375, + "learning_rate": 6.412865497076024e-05, + "loss": 3.4947, + "step": 13517 + }, + { + "epoch": 0.7114736842105264, + "grad_norm": 0.486328125, + "learning_rate": 6.411695906432749e-05, + "loss": 3.1582, + "step": 13518 + }, + { + "epoch": 0.7115263157894737, + "grad_norm": 0.49609375, + "learning_rate": 6.410526315789474e-05, + "loss": 3.5397, + "step": 13519 + }, + { + "epoch": 0.7115789473684211, + "grad_norm": 0.5, + "learning_rate": 6.4093567251462e-05, + "loss": 3.2895, + "step": 13520 + }, + { + "epoch": 0.7116315789473684, + "grad_norm": 0.54296875, + "learning_rate": 6.408187134502924e-05, + "loss": 3.1881, + "step": 13521 + }, + { + "epoch": 0.7116842105263158, + "grad_norm": 0.494140625, + "learning_rate": 6.407017543859649e-05, + "loss": 3.3242, + "step": 13522 + }, + { + "epoch": 0.7117368421052631, + "grad_norm": 0.484375, + "learning_rate": 6.405847953216375e-05, + "loss": 3.2897, + "step": 13523 + }, + { + "epoch": 0.7117894736842105, + "grad_norm": 0.51953125, + "learning_rate": 6.4046783625731e-05, + "loss": 3.4859, + "step": 13524 + }, + { + "epoch": 0.7118421052631579, + "grad_norm": 0.470703125, + "learning_rate": 6.403508771929825e-05, + "loss": 3.4043, + "step": 13525 + }, + { + "epoch": 0.7118947368421052, + "grad_norm": 0.50390625, + "learning_rate": 6.40233918128655e-05, + "loss": 3.2479, + "step": 13526 + }, + { + "epoch": 0.7119473684210527, + "grad_norm": 0.494140625, + "learning_rate": 6.401169590643276e-05, + "loss": 3.604, + "step": 13527 + }, + { + "epoch": 0.712, + "grad_norm": 0.48046875, + "learning_rate": 6.400000000000001e-05, + "loss": 3.2642, + "step": 13528 + }, + { + "epoch": 0.7120526315789474, + "grad_norm": 0.52734375, + "learning_rate": 6.398830409356725e-05, + "loss": 3.4715, + "step": 13529 + }, + { + "epoch": 0.7121052631578947, + "grad_norm": 0.4921875, + "learning_rate": 6.39766081871345e-05, + "loss": 3.4404, + "step": 13530 + }, + { + "epoch": 0.7121578947368421, + "grad_norm": 0.4765625, + "learning_rate": 6.396491228070176e-05, + "loss": 3.0606, + "step": 13531 + }, + { + "epoch": 0.7122105263157895, + "grad_norm": 0.51953125, + "learning_rate": 6.395321637426901e-05, + "loss": 2.9634, + "step": 13532 + }, + { + "epoch": 0.7122631578947368, + "grad_norm": 0.46875, + "learning_rate": 6.394152046783626e-05, + "loss": 3.2892, + "step": 13533 + }, + { + "epoch": 0.7123157894736842, + "grad_norm": 0.451171875, + "learning_rate": 6.392982456140351e-05, + "loss": 2.9983, + "step": 13534 + }, + { + "epoch": 0.7123684210526315, + "grad_norm": 0.55859375, + "learning_rate": 6.391812865497077e-05, + "loss": 3.424, + "step": 13535 + }, + { + "epoch": 0.712421052631579, + "grad_norm": 0.466796875, + "learning_rate": 6.390643274853802e-05, + "loss": 2.697, + "step": 13536 + }, + { + "epoch": 0.7124736842105264, + "grad_norm": 0.515625, + "learning_rate": 6.389473684210526e-05, + "loss": 3.158, + "step": 13537 + }, + { + "epoch": 0.7125263157894737, + "grad_norm": 0.478515625, + "learning_rate": 6.388304093567252e-05, + "loss": 3.0921, + "step": 13538 + }, + { + "epoch": 0.7125789473684211, + "grad_norm": 0.4921875, + "learning_rate": 6.387134502923977e-05, + "loss": 3.1244, + "step": 13539 + }, + { + "epoch": 0.7126315789473684, + "grad_norm": 0.4609375, + "learning_rate": 6.385964912280702e-05, + "loss": 2.8067, + "step": 13540 + }, + { + "epoch": 0.7126842105263158, + "grad_norm": 0.498046875, + "learning_rate": 6.384795321637427e-05, + "loss": 3.195, + "step": 13541 + }, + { + "epoch": 0.7127368421052631, + "grad_norm": 0.490234375, + "learning_rate": 6.383625730994153e-05, + "loss": 3.7521, + "step": 13542 + }, + { + "epoch": 0.7127894736842105, + "grad_norm": 0.46875, + "learning_rate": 6.382456140350878e-05, + "loss": 3.3613, + "step": 13543 + }, + { + "epoch": 0.7128421052631579, + "grad_norm": 0.482421875, + "learning_rate": 6.381286549707602e-05, + "loss": 3.422, + "step": 13544 + }, + { + "epoch": 0.7128947368421052, + "grad_norm": 0.490234375, + "learning_rate": 6.380116959064327e-05, + "loss": 3.4172, + "step": 13545 + }, + { + "epoch": 0.7129473684210527, + "grad_norm": 0.494140625, + "learning_rate": 6.378947368421053e-05, + "loss": 3.4953, + "step": 13546 + }, + { + "epoch": 0.713, + "grad_norm": 0.50390625, + "learning_rate": 6.377777777777778e-05, + "loss": 3.4053, + "step": 13547 + }, + { + "epoch": 0.7130526315789474, + "grad_norm": 0.5078125, + "learning_rate": 6.376608187134503e-05, + "loss": 3.5637, + "step": 13548 + }, + { + "epoch": 0.7131052631578947, + "grad_norm": 0.51171875, + "learning_rate": 6.375438596491229e-05, + "loss": 3.2917, + "step": 13549 + }, + { + "epoch": 0.7131578947368421, + "grad_norm": 0.494140625, + "learning_rate": 6.374269005847954e-05, + "loss": 2.9369, + "step": 13550 + }, + { + "epoch": 0.7132105263157895, + "grad_norm": 0.466796875, + "learning_rate": 6.373099415204679e-05, + "loss": 2.9707, + "step": 13551 + }, + { + "epoch": 0.7132631578947368, + "grad_norm": 0.490234375, + "learning_rate": 6.371929824561403e-05, + "loss": 3.2139, + "step": 13552 + }, + { + "epoch": 0.7133157894736842, + "grad_norm": 0.494140625, + "learning_rate": 6.370760233918129e-05, + "loss": 3.173, + "step": 13553 + }, + { + "epoch": 0.7133684210526315, + "grad_norm": 0.50390625, + "learning_rate": 6.369590643274854e-05, + "loss": 3.163, + "step": 13554 + }, + { + "epoch": 0.713421052631579, + "grad_norm": 0.52734375, + "learning_rate": 6.368421052631579e-05, + "loss": 2.994, + "step": 13555 + }, + { + "epoch": 0.7134736842105264, + "grad_norm": 0.4921875, + "learning_rate": 6.367251461988304e-05, + "loss": 3.4905, + "step": 13556 + }, + { + "epoch": 0.7135263157894737, + "grad_norm": 0.48828125, + "learning_rate": 6.36608187134503e-05, + "loss": 3.2065, + "step": 13557 + }, + { + "epoch": 0.7135789473684211, + "grad_norm": 0.50390625, + "learning_rate": 6.364912280701755e-05, + "loss": 3.3508, + "step": 13558 + }, + { + "epoch": 0.7136315789473684, + "grad_norm": 0.49609375, + "learning_rate": 6.36374269005848e-05, + "loss": 3.1598, + "step": 13559 + }, + { + "epoch": 0.7136842105263158, + "grad_norm": 0.546875, + "learning_rate": 6.362573099415205e-05, + "loss": 3.5252, + "step": 13560 + }, + { + "epoch": 0.7137368421052631, + "grad_norm": 0.494140625, + "learning_rate": 6.36140350877193e-05, + "loss": 3.4501, + "step": 13561 + }, + { + "epoch": 0.7137894736842105, + "grad_norm": 0.58984375, + "learning_rate": 6.360233918128655e-05, + "loss": 2.9224, + "step": 13562 + }, + { + "epoch": 0.7138421052631579, + "grad_norm": 0.494140625, + "learning_rate": 6.35906432748538e-05, + "loss": 3.3479, + "step": 13563 + }, + { + "epoch": 0.7138947368421052, + "grad_norm": 0.486328125, + "learning_rate": 6.357894736842106e-05, + "loss": 3.1942, + "step": 13564 + }, + { + "epoch": 0.7139473684210527, + "grad_norm": 0.490234375, + "learning_rate": 6.356725146198831e-05, + "loss": 3.4894, + "step": 13565 + }, + { + "epoch": 0.714, + "grad_norm": 0.50390625, + "learning_rate": 6.355555555555556e-05, + "loss": 3.1395, + "step": 13566 + }, + { + "epoch": 0.7140526315789474, + "grad_norm": 0.4921875, + "learning_rate": 6.35438596491228e-05, + "loss": 3.5724, + "step": 13567 + }, + { + "epoch": 0.7141052631578947, + "grad_norm": 0.498046875, + "learning_rate": 6.353216374269006e-05, + "loss": 2.9966, + "step": 13568 + }, + { + "epoch": 0.7141578947368421, + "grad_norm": 0.51953125, + "learning_rate": 6.352046783625731e-05, + "loss": 3.2456, + "step": 13569 + }, + { + "epoch": 0.7142105263157895, + "grad_norm": 0.498046875, + "learning_rate": 6.350877192982456e-05, + "loss": 3.6059, + "step": 13570 + }, + { + "epoch": 0.7142631578947368, + "grad_norm": 0.5390625, + "learning_rate": 6.349707602339181e-05, + "loss": 3.0221, + "step": 13571 + }, + { + "epoch": 0.7143157894736842, + "grad_norm": 0.4921875, + "learning_rate": 6.348538011695907e-05, + "loss": 3.3681, + "step": 13572 + }, + { + "epoch": 0.7143684210526315, + "grad_norm": 0.482421875, + "learning_rate": 6.347368421052632e-05, + "loss": 3.5063, + "step": 13573 + }, + { + "epoch": 0.714421052631579, + "grad_norm": 0.51171875, + "learning_rate": 6.346198830409357e-05, + "loss": 3.2182, + "step": 13574 + }, + { + "epoch": 0.7144736842105263, + "grad_norm": 0.52734375, + "learning_rate": 6.345029239766082e-05, + "loss": 3.0086, + "step": 13575 + }, + { + "epoch": 0.7145263157894737, + "grad_norm": 0.4765625, + "learning_rate": 6.343859649122807e-05, + "loss": 3.3333, + "step": 13576 + }, + { + "epoch": 0.7145789473684211, + "grad_norm": 0.515625, + "learning_rate": 6.342690058479532e-05, + "loss": 3.3401, + "step": 13577 + }, + { + "epoch": 0.7146315789473684, + "grad_norm": 0.466796875, + "learning_rate": 6.341520467836257e-05, + "loss": 3.0972, + "step": 13578 + }, + { + "epoch": 0.7146842105263158, + "grad_norm": 0.482421875, + "learning_rate": 6.340350877192983e-05, + "loss": 3.4839, + "step": 13579 + }, + { + "epoch": 0.7147368421052631, + "grad_norm": 0.484375, + "learning_rate": 6.339181286549708e-05, + "loss": 2.8193, + "step": 13580 + }, + { + "epoch": 0.7147894736842105, + "grad_norm": 0.494140625, + "learning_rate": 6.338011695906433e-05, + "loss": 3.3715, + "step": 13581 + }, + { + "epoch": 0.714842105263158, + "grad_norm": 0.498046875, + "learning_rate": 6.336842105263158e-05, + "loss": 3.2199, + "step": 13582 + }, + { + "epoch": 0.7148947368421052, + "grad_norm": 0.49609375, + "learning_rate": 6.335672514619883e-05, + "loss": 3.3058, + "step": 13583 + }, + { + "epoch": 0.7149473684210527, + "grad_norm": 0.515625, + "learning_rate": 6.334502923976608e-05, + "loss": 3.4714, + "step": 13584 + }, + { + "epoch": 0.715, + "grad_norm": 0.5, + "learning_rate": 6.333333333333333e-05, + "loss": 3.3804, + "step": 13585 + }, + { + "epoch": 0.7150526315789474, + "grad_norm": 0.5859375, + "learning_rate": 6.33216374269006e-05, + "loss": 3.5778, + "step": 13586 + }, + { + "epoch": 0.7151052631578947, + "grad_norm": 0.470703125, + "learning_rate": 6.330994152046785e-05, + "loss": 3.2537, + "step": 13587 + }, + { + "epoch": 0.7151578947368421, + "grad_norm": 0.498046875, + "learning_rate": 6.32982456140351e-05, + "loss": 3.2586, + "step": 13588 + }, + { + "epoch": 0.7152105263157895, + "grad_norm": 0.5078125, + "learning_rate": 6.328654970760234e-05, + "loss": 3.3373, + "step": 13589 + }, + { + "epoch": 0.7152631578947368, + "grad_norm": 0.478515625, + "learning_rate": 6.32748538011696e-05, + "loss": 3.1135, + "step": 13590 + }, + { + "epoch": 0.7153157894736842, + "grad_norm": 0.67578125, + "learning_rate": 6.326315789473684e-05, + "loss": 2.9262, + "step": 13591 + }, + { + "epoch": 0.7153684210526315, + "grad_norm": 0.53515625, + "learning_rate": 6.325146198830409e-05, + "loss": 3.0701, + "step": 13592 + }, + { + "epoch": 0.715421052631579, + "grad_norm": 0.478515625, + "learning_rate": 6.323976608187134e-05, + "loss": 3.3667, + "step": 13593 + }, + { + "epoch": 0.7154736842105263, + "grad_norm": 0.4921875, + "learning_rate": 6.32280701754386e-05, + "loss": 3.355, + "step": 13594 + }, + { + "epoch": 0.7155263157894737, + "grad_norm": 0.458984375, + "learning_rate": 6.321637426900586e-05, + "loss": 3.5321, + "step": 13595 + }, + { + "epoch": 0.7155789473684211, + "grad_norm": 0.486328125, + "learning_rate": 6.32046783625731e-05, + "loss": 3.3616, + "step": 13596 + }, + { + "epoch": 0.7156315789473684, + "grad_norm": 0.48046875, + "learning_rate": 6.319298245614035e-05, + "loss": 3.1006, + "step": 13597 + }, + { + "epoch": 0.7156842105263158, + "grad_norm": 0.50390625, + "learning_rate": 6.31812865497076e-05, + "loss": 3.7189, + "step": 13598 + }, + { + "epoch": 0.7157368421052631, + "grad_norm": 0.4921875, + "learning_rate": 6.316959064327485e-05, + "loss": 3.6325, + "step": 13599 + }, + { + "epoch": 0.7157894736842105, + "grad_norm": 0.48828125, + "learning_rate": 6.31578947368421e-05, + "loss": 3.4393, + "step": 13600 + }, + { + "epoch": 0.715842105263158, + "grad_norm": 0.466796875, + "learning_rate": 6.314619883040937e-05, + "loss": 2.867, + "step": 13601 + }, + { + "epoch": 0.7158947368421053, + "grad_norm": 0.48046875, + "learning_rate": 6.313450292397662e-05, + "loss": 3.6808, + "step": 13602 + }, + { + "epoch": 0.7159473684210527, + "grad_norm": 0.46875, + "learning_rate": 6.312280701754387e-05, + "loss": 3.0798, + "step": 13603 + }, + { + "epoch": 0.716, + "grad_norm": 0.4921875, + "learning_rate": 6.311111111111112e-05, + "loss": 3.63, + "step": 13604 + }, + { + "epoch": 0.7160526315789474, + "grad_norm": 0.484375, + "learning_rate": 6.309941520467836e-05, + "loss": 3.0603, + "step": 13605 + }, + { + "epoch": 0.7161052631578947, + "grad_norm": 0.470703125, + "learning_rate": 6.308771929824561e-05, + "loss": 3.2701, + "step": 13606 + }, + { + "epoch": 0.7161578947368421, + "grad_norm": 0.484375, + "learning_rate": 6.307602339181286e-05, + "loss": 3.408, + "step": 13607 + }, + { + "epoch": 0.7162105263157895, + "grad_norm": 0.51953125, + "learning_rate": 6.306432748538011e-05, + "loss": 3.4038, + "step": 13608 + }, + { + "epoch": 0.7162631578947368, + "grad_norm": 0.48046875, + "learning_rate": 6.305263157894738e-05, + "loss": 3.4365, + "step": 13609 + }, + { + "epoch": 0.7163157894736842, + "grad_norm": 0.5, + "learning_rate": 6.304093567251463e-05, + "loss": 3.7307, + "step": 13610 + }, + { + "epoch": 0.7163684210526315, + "grad_norm": 0.5, + "learning_rate": 6.302923976608188e-05, + "loss": 3.3462, + "step": 13611 + }, + { + "epoch": 0.716421052631579, + "grad_norm": 0.4921875, + "learning_rate": 6.301754385964913e-05, + "loss": 3.6747, + "step": 13612 + }, + { + "epoch": 0.7164736842105263, + "grad_norm": 0.455078125, + "learning_rate": 6.300584795321638e-05, + "loss": 2.9958, + "step": 13613 + }, + { + "epoch": 0.7165263157894737, + "grad_norm": 0.478515625, + "learning_rate": 6.299415204678362e-05, + "loss": 3.3997, + "step": 13614 + }, + { + "epoch": 0.7165789473684211, + "grad_norm": 0.482421875, + "learning_rate": 6.298245614035087e-05, + "loss": 3.648, + "step": 13615 + }, + { + "epoch": 0.7166315789473684, + "grad_norm": 0.5390625, + "learning_rate": 6.297076023391814e-05, + "loss": 3.2837, + "step": 13616 + }, + { + "epoch": 0.7166842105263158, + "grad_norm": 0.53125, + "learning_rate": 6.295906432748539e-05, + "loss": 3.6786, + "step": 13617 + }, + { + "epoch": 0.7167368421052631, + "grad_norm": 0.4921875, + "learning_rate": 6.294736842105264e-05, + "loss": 3.2686, + "step": 13618 + }, + { + "epoch": 0.7167894736842105, + "grad_norm": 0.5, + "learning_rate": 6.293567251461989e-05, + "loss": 3.277, + "step": 13619 + }, + { + "epoch": 0.716842105263158, + "grad_norm": 0.50390625, + "learning_rate": 6.292397660818714e-05, + "loss": 3.2082, + "step": 13620 + }, + { + "epoch": 0.7168947368421053, + "grad_norm": 0.478515625, + "learning_rate": 6.291228070175439e-05, + "loss": 3.4053, + "step": 13621 + }, + { + "epoch": 0.7169473684210527, + "grad_norm": 0.50390625, + "learning_rate": 6.290058479532163e-05, + "loss": 3.5029, + "step": 13622 + }, + { + "epoch": 0.717, + "grad_norm": 0.50390625, + "learning_rate": 6.28888888888889e-05, + "loss": 3.3472, + "step": 13623 + }, + { + "epoch": 0.7170526315789474, + "grad_norm": 0.490234375, + "learning_rate": 6.287719298245615e-05, + "loss": 2.9808, + "step": 13624 + }, + { + "epoch": 0.7171052631578947, + "grad_norm": 0.48828125, + "learning_rate": 6.28654970760234e-05, + "loss": 3.6777, + "step": 13625 + }, + { + "epoch": 0.7171578947368421, + "grad_norm": 0.49609375, + "learning_rate": 6.285380116959065e-05, + "loss": 3.5991, + "step": 13626 + }, + { + "epoch": 0.7172105263157895, + "grad_norm": 0.453125, + "learning_rate": 6.28421052631579e-05, + "loss": 2.9407, + "step": 13627 + }, + { + "epoch": 0.7172631578947368, + "grad_norm": 0.46875, + "learning_rate": 6.283040935672515e-05, + "loss": 2.9282, + "step": 13628 + }, + { + "epoch": 0.7173157894736842, + "grad_norm": 0.490234375, + "learning_rate": 6.28187134502924e-05, + "loss": 3.1707, + "step": 13629 + }, + { + "epoch": 0.7173684210526315, + "grad_norm": 0.51953125, + "learning_rate": 6.280701754385965e-05, + "loss": 3.1021, + "step": 13630 + }, + { + "epoch": 0.717421052631579, + "grad_norm": 0.48046875, + "learning_rate": 6.279532163742691e-05, + "loss": 2.8876, + "step": 13631 + }, + { + "epoch": 0.7174736842105263, + "grad_norm": 0.58203125, + "learning_rate": 6.278362573099416e-05, + "loss": 3.4352, + "step": 13632 + }, + { + "epoch": 0.7175263157894737, + "grad_norm": 0.4921875, + "learning_rate": 6.277192982456141e-05, + "loss": 3.4062, + "step": 13633 + }, + { + "epoch": 0.7175789473684211, + "grad_norm": 0.4921875, + "learning_rate": 6.276023391812866e-05, + "loss": 3.577, + "step": 13634 + }, + { + "epoch": 0.7176315789473684, + "grad_norm": 0.490234375, + "learning_rate": 6.274853801169591e-05, + "loss": 3.0894, + "step": 13635 + }, + { + "epoch": 0.7176842105263158, + "grad_norm": 0.486328125, + "learning_rate": 6.273684210526316e-05, + "loss": 3.6601, + "step": 13636 + }, + { + "epoch": 0.7177368421052631, + "grad_norm": 0.59765625, + "learning_rate": 6.27251461988304e-05, + "loss": 3.1026, + "step": 13637 + }, + { + "epoch": 0.7177894736842105, + "grad_norm": 0.49609375, + "learning_rate": 6.271345029239767e-05, + "loss": 3.5186, + "step": 13638 + }, + { + "epoch": 0.717842105263158, + "grad_norm": 0.45703125, + "learning_rate": 6.270175438596492e-05, + "loss": 3.232, + "step": 13639 + }, + { + "epoch": 0.7178947368421053, + "grad_norm": 0.48046875, + "learning_rate": 6.269005847953217e-05, + "loss": 3.3161, + "step": 13640 + }, + { + "epoch": 0.7179473684210527, + "grad_norm": 0.48828125, + "learning_rate": 6.267836257309942e-05, + "loss": 3.1668, + "step": 13641 + }, + { + "epoch": 0.718, + "grad_norm": 0.466796875, + "learning_rate": 6.266666666666667e-05, + "loss": 3.6001, + "step": 13642 + }, + { + "epoch": 0.7180526315789474, + "grad_norm": 0.49609375, + "learning_rate": 6.265497076023392e-05, + "loss": 3.6752, + "step": 13643 + }, + { + "epoch": 0.7181052631578947, + "grad_norm": 0.50390625, + "learning_rate": 6.264327485380117e-05, + "loss": 3.0321, + "step": 13644 + }, + { + "epoch": 0.7181578947368421, + "grad_norm": 0.5078125, + "learning_rate": 6.263157894736842e-05, + "loss": 2.7053, + "step": 13645 + }, + { + "epoch": 0.7182105263157895, + "grad_norm": 0.470703125, + "learning_rate": 6.261988304093568e-05, + "loss": 3.2714, + "step": 13646 + }, + { + "epoch": 0.7182631578947368, + "grad_norm": 0.466796875, + "learning_rate": 6.260818713450293e-05, + "loss": 3.5656, + "step": 13647 + }, + { + "epoch": 0.7183157894736842, + "grad_norm": 0.48828125, + "learning_rate": 6.259649122807018e-05, + "loss": 2.9383, + "step": 13648 + }, + { + "epoch": 0.7183684210526315, + "grad_norm": 0.486328125, + "learning_rate": 6.258479532163743e-05, + "loss": 3.2217, + "step": 13649 + }, + { + "epoch": 0.718421052631579, + "grad_norm": 0.453125, + "learning_rate": 6.257309941520468e-05, + "loss": 3.3393, + "step": 13650 + }, + { + "epoch": 0.7184736842105263, + "grad_norm": 0.486328125, + "learning_rate": 6.256140350877193e-05, + "loss": 3.0244, + "step": 13651 + }, + { + "epoch": 0.7185263157894737, + "grad_norm": 0.5234375, + "learning_rate": 6.254970760233918e-05, + "loss": 3.5337, + "step": 13652 + }, + { + "epoch": 0.7185789473684211, + "grad_norm": 0.5234375, + "learning_rate": 6.253801169590644e-05, + "loss": 3.3684, + "step": 13653 + }, + { + "epoch": 0.7186315789473684, + "grad_norm": 0.498046875, + "learning_rate": 6.252631578947369e-05, + "loss": 2.9466, + "step": 13654 + }, + { + "epoch": 0.7186842105263158, + "grad_norm": 0.48046875, + "learning_rate": 6.251461988304094e-05, + "loss": 3.2928, + "step": 13655 + }, + { + "epoch": 0.7187368421052631, + "grad_norm": 0.50390625, + "learning_rate": 6.250292397660819e-05, + "loss": 3.3601, + "step": 13656 + }, + { + "epoch": 0.7187894736842105, + "grad_norm": 0.5, + "learning_rate": 6.249122807017544e-05, + "loss": 3.6464, + "step": 13657 + }, + { + "epoch": 0.718842105263158, + "grad_norm": 0.49609375, + "learning_rate": 6.247953216374269e-05, + "loss": 3.1706, + "step": 13658 + }, + { + "epoch": 0.7188947368421053, + "grad_norm": 0.5078125, + "learning_rate": 6.246783625730994e-05, + "loss": 3.3226, + "step": 13659 + }, + { + "epoch": 0.7189473684210527, + "grad_norm": 0.48046875, + "learning_rate": 6.24561403508772e-05, + "loss": 3.5381, + "step": 13660 + }, + { + "epoch": 0.719, + "grad_norm": 0.48046875, + "learning_rate": 6.244444444444445e-05, + "loss": 3.2464, + "step": 13661 + }, + { + "epoch": 0.7190526315789474, + "grad_norm": 0.48046875, + "learning_rate": 6.24327485380117e-05, + "loss": 3.4221, + "step": 13662 + }, + { + "epoch": 0.7191052631578947, + "grad_norm": 0.51171875, + "learning_rate": 6.242105263157895e-05, + "loss": 3.1562, + "step": 13663 + }, + { + "epoch": 0.7191578947368421, + "grad_norm": 0.498046875, + "learning_rate": 6.240935672514621e-05, + "loss": 3.074, + "step": 13664 + }, + { + "epoch": 0.7192105263157895, + "grad_norm": 0.515625, + "learning_rate": 6.239766081871345e-05, + "loss": 3.2354, + "step": 13665 + }, + { + "epoch": 0.7192631578947368, + "grad_norm": 0.5234375, + "learning_rate": 6.23859649122807e-05, + "loss": 3.0475, + "step": 13666 + }, + { + "epoch": 0.7193157894736842, + "grad_norm": 0.478515625, + "learning_rate": 6.237426900584795e-05, + "loss": 3.2832, + "step": 13667 + }, + { + "epoch": 0.7193684210526315, + "grad_norm": 0.578125, + "learning_rate": 6.236257309941521e-05, + "loss": 3.3559, + "step": 13668 + }, + { + "epoch": 0.719421052631579, + "grad_norm": 0.8203125, + "learning_rate": 6.235087719298246e-05, + "loss": 3.2431, + "step": 13669 + }, + { + "epoch": 0.7194736842105263, + "grad_norm": 0.50390625, + "learning_rate": 6.233918128654971e-05, + "loss": 3.633, + "step": 13670 + }, + { + "epoch": 0.7195263157894737, + "grad_norm": 0.470703125, + "learning_rate": 6.232748538011696e-05, + "loss": 3.4499, + "step": 13671 + }, + { + "epoch": 0.7195789473684211, + "grad_norm": 0.515625, + "learning_rate": 6.231578947368422e-05, + "loss": 3.8261, + "step": 13672 + }, + { + "epoch": 0.7196315789473684, + "grad_norm": 0.5234375, + "learning_rate": 6.230409356725146e-05, + "loss": 3.2252, + "step": 13673 + }, + { + "epoch": 0.7196842105263158, + "grad_norm": 0.5078125, + "learning_rate": 6.229239766081871e-05, + "loss": 3.4508, + "step": 13674 + }, + { + "epoch": 0.7197368421052631, + "grad_norm": 0.5390625, + "learning_rate": 6.228070175438597e-05, + "loss": 3.4028, + "step": 13675 + }, + { + "epoch": 0.7197894736842105, + "grad_norm": 0.494140625, + "learning_rate": 6.226900584795322e-05, + "loss": 3.5004, + "step": 13676 + }, + { + "epoch": 0.719842105263158, + "grad_norm": 0.48828125, + "learning_rate": 6.225730994152047e-05, + "loss": 2.845, + "step": 13677 + }, + { + "epoch": 0.7198947368421053, + "grad_norm": 0.494140625, + "learning_rate": 6.224561403508772e-05, + "loss": 3.0552, + "step": 13678 + }, + { + "epoch": 0.7199473684210527, + "grad_norm": 0.484375, + "learning_rate": 6.223391812865498e-05, + "loss": 3.1935, + "step": 13679 + }, + { + "epoch": 0.72, + "grad_norm": 0.5234375, + "learning_rate": 6.222222222222222e-05, + "loss": 3.2823, + "step": 13680 + }, + { + "epoch": 0.7200526315789474, + "grad_norm": 0.5234375, + "learning_rate": 6.221052631578947e-05, + "loss": 2.9101, + "step": 13681 + }, + { + "epoch": 0.7201052631578947, + "grad_norm": 0.5, + "learning_rate": 6.219883040935672e-05, + "loss": 3.3467, + "step": 13682 + }, + { + "epoch": 0.7201578947368421, + "grad_norm": 0.494140625, + "learning_rate": 6.218713450292398e-05, + "loss": 3.4985, + "step": 13683 + }, + { + "epoch": 0.7202105263157895, + "grad_norm": 0.53125, + "learning_rate": 6.217543859649123e-05, + "loss": 3.7192, + "step": 13684 + }, + { + "epoch": 0.7202631578947368, + "grad_norm": 0.490234375, + "learning_rate": 6.216374269005848e-05, + "loss": 3.2996, + "step": 13685 + }, + { + "epoch": 0.7203157894736842, + "grad_norm": 0.51953125, + "learning_rate": 6.215204678362574e-05, + "loss": 3.4248, + "step": 13686 + }, + { + "epoch": 0.7203684210526315, + "grad_norm": 0.4921875, + "learning_rate": 6.2140350877193e-05, + "loss": 3.0125, + "step": 13687 + }, + { + "epoch": 0.720421052631579, + "grad_norm": 0.46875, + "learning_rate": 6.212865497076023e-05, + "loss": 3.1686, + "step": 13688 + }, + { + "epoch": 0.7204736842105263, + "grad_norm": 0.53515625, + "learning_rate": 6.211695906432748e-05, + "loss": 3.2552, + "step": 13689 + }, + { + "epoch": 0.7205263157894737, + "grad_norm": 0.51171875, + "learning_rate": 6.210526315789474e-05, + "loss": 3.425, + "step": 13690 + }, + { + "epoch": 0.7205789473684211, + "grad_norm": 0.490234375, + "learning_rate": 6.209356725146199e-05, + "loss": 3.7123, + "step": 13691 + }, + { + "epoch": 0.7206315789473684, + "grad_norm": 0.46875, + "learning_rate": 6.208187134502924e-05, + "loss": 3.2602, + "step": 13692 + }, + { + "epoch": 0.7206842105263158, + "grad_norm": 0.47265625, + "learning_rate": 6.207017543859649e-05, + "loss": 3.6101, + "step": 13693 + }, + { + "epoch": 0.7207368421052631, + "grad_norm": 0.478515625, + "learning_rate": 6.205847953216375e-05, + "loss": 2.9982, + "step": 13694 + }, + { + "epoch": 0.7207894736842105, + "grad_norm": 0.46484375, + "learning_rate": 6.2046783625731e-05, + "loss": 2.9467, + "step": 13695 + }, + { + "epoch": 0.7208421052631578, + "grad_norm": 0.546875, + "learning_rate": 6.203508771929824e-05, + "loss": 3.3595, + "step": 13696 + }, + { + "epoch": 0.7208947368421053, + "grad_norm": 0.4921875, + "learning_rate": 6.20233918128655e-05, + "loss": 3.428, + "step": 13697 + }, + { + "epoch": 0.7209473684210527, + "grad_norm": 0.515625, + "learning_rate": 6.201169590643275e-05, + "loss": 3.3802, + "step": 13698 + }, + { + "epoch": 0.721, + "grad_norm": 0.486328125, + "learning_rate": 6.2e-05, + "loss": 2.9728, + "step": 13699 + }, + { + "epoch": 0.7210526315789474, + "grad_norm": 0.5078125, + "learning_rate": 6.198830409356725e-05, + "loss": 3.5935, + "step": 13700 + }, + { + "epoch": 0.7211052631578947, + "grad_norm": 0.494140625, + "learning_rate": 6.197660818713452e-05, + "loss": 3.2036, + "step": 13701 + }, + { + "epoch": 0.7211578947368421, + "grad_norm": 0.48828125, + "learning_rate": 6.196491228070177e-05, + "loss": 3.5508, + "step": 13702 + }, + { + "epoch": 0.7212105263157895, + "grad_norm": 0.546875, + "learning_rate": 6.1953216374269e-05, + "loss": 2.8091, + "step": 13703 + }, + { + "epoch": 0.7212631578947368, + "grad_norm": 0.53125, + "learning_rate": 6.194152046783625e-05, + "loss": 3.0386, + "step": 13704 + }, + { + "epoch": 0.7213157894736842, + "grad_norm": 0.5625, + "learning_rate": 6.192982456140351e-05, + "loss": 3.6077, + "step": 13705 + }, + { + "epoch": 0.7213684210526315, + "grad_norm": 0.53515625, + "learning_rate": 6.191812865497076e-05, + "loss": 3.3854, + "step": 13706 + }, + { + "epoch": 0.721421052631579, + "grad_norm": 0.515625, + "learning_rate": 6.190643274853801e-05, + "loss": 3.2217, + "step": 13707 + }, + { + "epoch": 0.7214736842105263, + "grad_norm": 0.47265625, + "learning_rate": 6.189473684210526e-05, + "loss": 2.9787, + "step": 13708 + }, + { + "epoch": 0.7215263157894737, + "grad_norm": 0.515625, + "learning_rate": 6.188304093567253e-05, + "loss": 3.1805, + "step": 13709 + }, + { + "epoch": 0.7215789473684211, + "grad_norm": 0.490234375, + "learning_rate": 6.187134502923978e-05, + "loss": 3.1412, + "step": 13710 + }, + { + "epoch": 0.7216315789473684, + "grad_norm": 0.48828125, + "learning_rate": 6.185964912280701e-05, + "loss": 3.3097, + "step": 13711 + }, + { + "epoch": 0.7216842105263158, + "grad_norm": 0.498046875, + "learning_rate": 6.184795321637427e-05, + "loss": 3.5777, + "step": 13712 + }, + { + "epoch": 0.7217368421052631, + "grad_norm": 0.50390625, + "learning_rate": 6.183625730994152e-05, + "loss": 3.2317, + "step": 13713 + }, + { + "epoch": 0.7217894736842105, + "grad_norm": 0.47265625, + "learning_rate": 6.182456140350877e-05, + "loss": 3.2307, + "step": 13714 + }, + { + "epoch": 0.7218421052631578, + "grad_norm": 0.51171875, + "learning_rate": 6.181286549707602e-05, + "loss": 3.29, + "step": 13715 + }, + { + "epoch": 0.7218947368421053, + "grad_norm": 0.5390625, + "learning_rate": 6.180116959064329e-05, + "loss": 3.1277, + "step": 13716 + }, + { + "epoch": 0.7219473684210527, + "grad_norm": 0.50390625, + "learning_rate": 6.178947368421054e-05, + "loss": 3.2433, + "step": 13717 + }, + { + "epoch": 0.722, + "grad_norm": 0.51171875, + "learning_rate": 6.177777777777779e-05, + "loss": 3.3015, + "step": 13718 + }, + { + "epoch": 0.7220526315789474, + "grad_norm": 0.498046875, + "learning_rate": 6.176608187134502e-05, + "loss": 3.3318, + "step": 13719 + }, + { + "epoch": 0.7221052631578947, + "grad_norm": 0.6875, + "learning_rate": 6.175438596491228e-05, + "loss": 3.1615, + "step": 13720 + }, + { + "epoch": 0.7221578947368421, + "grad_norm": 0.4765625, + "learning_rate": 6.174269005847953e-05, + "loss": 3.4421, + "step": 13721 + }, + { + "epoch": 0.7222105263157895, + "grad_norm": 0.55859375, + "learning_rate": 6.173099415204678e-05, + "loss": 3.5366, + "step": 13722 + }, + { + "epoch": 0.7222631578947368, + "grad_norm": 0.490234375, + "learning_rate": 6.171929824561405e-05, + "loss": 3.6489, + "step": 13723 + }, + { + "epoch": 0.7223157894736842, + "grad_norm": 0.50390625, + "learning_rate": 6.17076023391813e-05, + "loss": 3.3628, + "step": 13724 + }, + { + "epoch": 0.7223684210526315, + "grad_norm": 0.494140625, + "learning_rate": 6.169590643274855e-05, + "loss": 3.1437, + "step": 13725 + }, + { + "epoch": 0.722421052631579, + "grad_norm": 0.439453125, + "learning_rate": 6.168421052631578e-05, + "loss": 2.9412, + "step": 13726 + }, + { + "epoch": 0.7224736842105263, + "grad_norm": 0.451171875, + "learning_rate": 6.167251461988305e-05, + "loss": 2.982, + "step": 13727 + }, + { + "epoch": 0.7225263157894737, + "grad_norm": 0.47265625, + "learning_rate": 6.16608187134503e-05, + "loss": 3.5023, + "step": 13728 + }, + { + "epoch": 0.7225789473684211, + "grad_norm": 0.61328125, + "learning_rate": 6.164912280701754e-05, + "loss": 3.2885, + "step": 13729 + }, + { + "epoch": 0.7226315789473684, + "grad_norm": 0.50390625, + "learning_rate": 6.16374269005848e-05, + "loss": 3.3079, + "step": 13730 + }, + { + "epoch": 0.7226842105263158, + "grad_norm": 0.478515625, + "learning_rate": 6.162573099415206e-05, + "loss": 3.3367, + "step": 13731 + }, + { + "epoch": 0.7227368421052631, + "grad_norm": 0.47265625, + "learning_rate": 6.161403508771931e-05, + "loss": 3.2775, + "step": 13732 + }, + { + "epoch": 0.7227894736842105, + "grad_norm": 0.458984375, + "learning_rate": 6.160233918128656e-05, + "loss": 3.4555, + "step": 13733 + }, + { + "epoch": 0.7228421052631578, + "grad_norm": 0.4765625, + "learning_rate": 6.159064327485379e-05, + "loss": 3.6162, + "step": 13734 + }, + { + "epoch": 0.7228947368421053, + "grad_norm": 0.5625, + "learning_rate": 6.157894736842106e-05, + "loss": 3.1799, + "step": 13735 + }, + { + "epoch": 0.7229473684210527, + "grad_norm": 0.4609375, + "learning_rate": 6.15672514619883e-05, + "loss": 3.5963, + "step": 13736 + }, + { + "epoch": 0.723, + "grad_norm": 0.458984375, + "learning_rate": 6.155555555555555e-05, + "loss": 3.3072, + "step": 13737 + }, + { + "epoch": 0.7230526315789474, + "grad_norm": 0.52734375, + "learning_rate": 6.154385964912282e-05, + "loss": 3.5337, + "step": 13738 + }, + { + "epoch": 0.7231052631578947, + "grad_norm": 0.51171875, + "learning_rate": 6.153216374269007e-05, + "loss": 3.3097, + "step": 13739 + }, + { + "epoch": 0.7231578947368421, + "grad_norm": 0.484375, + "learning_rate": 6.152046783625732e-05, + "loss": 3.1583, + "step": 13740 + }, + { + "epoch": 0.7232105263157895, + "grad_norm": 0.48046875, + "learning_rate": 6.150877192982457e-05, + "loss": 3.8971, + "step": 13741 + }, + { + "epoch": 0.7232631578947368, + "grad_norm": 0.484375, + "learning_rate": 6.149707602339182e-05, + "loss": 3.452, + "step": 13742 + }, + { + "epoch": 0.7233157894736842, + "grad_norm": 0.54296875, + "learning_rate": 6.148538011695907e-05, + "loss": 3.1515, + "step": 13743 + }, + { + "epoch": 0.7233684210526315, + "grad_norm": 0.482421875, + "learning_rate": 6.147368421052632e-05, + "loss": 3.0947, + "step": 13744 + }, + { + "epoch": 0.723421052631579, + "grad_norm": 0.50390625, + "learning_rate": 6.146198830409357e-05, + "loss": 3.7472, + "step": 13745 + }, + { + "epoch": 0.7234736842105263, + "grad_norm": 0.52734375, + "learning_rate": 6.145029239766083e-05, + "loss": 3.4977, + "step": 13746 + }, + { + "epoch": 0.7235263157894737, + "grad_norm": 0.48828125, + "learning_rate": 6.143859649122808e-05, + "loss": 3.3536, + "step": 13747 + }, + { + "epoch": 0.7235789473684211, + "grad_norm": 0.5390625, + "learning_rate": 6.142690058479533e-05, + "loss": 3.4756, + "step": 13748 + }, + { + "epoch": 0.7236315789473684, + "grad_norm": 0.5078125, + "learning_rate": 6.141520467836258e-05, + "loss": 3.3198, + "step": 13749 + }, + { + "epoch": 0.7236842105263158, + "grad_norm": 0.4921875, + "learning_rate": 6.140350877192983e-05, + "loss": 3.3269, + "step": 13750 + }, + { + "epoch": 0.7237368421052631, + "grad_norm": 0.51953125, + "learning_rate": 6.139181286549708e-05, + "loss": 3.3615, + "step": 13751 + }, + { + "epoch": 0.7237894736842105, + "grad_norm": 0.4921875, + "learning_rate": 6.138011695906433e-05, + "loss": 3.0747, + "step": 13752 + }, + { + "epoch": 0.7238421052631578, + "grad_norm": 0.46484375, + "learning_rate": 6.136842105263159e-05, + "loss": 3.3521, + "step": 13753 + }, + { + "epoch": 0.7238947368421053, + "grad_norm": 0.48828125, + "learning_rate": 6.135672514619884e-05, + "loss": 3.2959, + "step": 13754 + }, + { + "epoch": 0.7239473684210527, + "grad_norm": 0.453125, + "learning_rate": 6.134502923976609e-05, + "loss": 3.0111, + "step": 13755 + }, + { + "epoch": 0.724, + "grad_norm": 0.46875, + "learning_rate": 6.133333333333334e-05, + "loss": 3.2984, + "step": 13756 + }, + { + "epoch": 0.7240526315789474, + "grad_norm": 0.515625, + "learning_rate": 6.132163742690059e-05, + "loss": 3.3223, + "step": 13757 + }, + { + "epoch": 0.7241052631578947, + "grad_norm": 0.51171875, + "learning_rate": 6.130994152046784e-05, + "loss": 2.9527, + "step": 13758 + }, + { + "epoch": 0.7241578947368421, + "grad_norm": 0.49609375, + "learning_rate": 6.129824561403509e-05, + "loss": 2.7107, + "step": 13759 + }, + { + "epoch": 0.7242105263157895, + "grad_norm": 0.48046875, + "learning_rate": 6.128654970760235e-05, + "loss": 3.334, + "step": 13760 + }, + { + "epoch": 0.7242631578947368, + "grad_norm": 0.50390625, + "learning_rate": 6.12748538011696e-05, + "loss": 2.9445, + "step": 13761 + }, + { + "epoch": 0.7243157894736842, + "grad_norm": 0.50390625, + "learning_rate": 6.126315789473685e-05, + "loss": 3.3428, + "step": 13762 + }, + { + "epoch": 0.7243684210526316, + "grad_norm": 0.5078125, + "learning_rate": 6.12514619883041e-05, + "loss": 2.8012, + "step": 13763 + }, + { + "epoch": 0.724421052631579, + "grad_norm": 0.4921875, + "learning_rate": 6.123976608187135e-05, + "loss": 3.0545, + "step": 13764 + }, + { + "epoch": 0.7244736842105263, + "grad_norm": 0.498046875, + "learning_rate": 6.12280701754386e-05, + "loss": 3.038, + "step": 13765 + }, + { + "epoch": 0.7245263157894737, + "grad_norm": 0.486328125, + "learning_rate": 6.121637426900585e-05, + "loss": 3.2734, + "step": 13766 + }, + { + "epoch": 0.7245789473684211, + "grad_norm": 0.51171875, + "learning_rate": 6.12046783625731e-05, + "loss": 3.2516, + "step": 13767 + }, + { + "epoch": 0.7246315789473684, + "grad_norm": 0.4765625, + "learning_rate": 6.119298245614036e-05, + "loss": 3.1894, + "step": 13768 + }, + { + "epoch": 0.7246842105263158, + "grad_norm": 0.482421875, + "learning_rate": 6.118128654970761e-05, + "loss": 3.3185, + "step": 13769 + }, + { + "epoch": 0.7247368421052631, + "grad_norm": 0.46875, + "learning_rate": 6.116959064327486e-05, + "loss": 3.4615, + "step": 13770 + }, + { + "epoch": 0.7247894736842105, + "grad_norm": 0.486328125, + "learning_rate": 6.115789473684211e-05, + "loss": 3.4695, + "step": 13771 + }, + { + "epoch": 0.7248421052631578, + "grad_norm": 0.486328125, + "learning_rate": 6.114619883040936e-05, + "loss": 3.4505, + "step": 13772 + }, + { + "epoch": 0.7248947368421053, + "grad_norm": 0.50390625, + "learning_rate": 6.113450292397661e-05, + "loss": 3.6281, + "step": 13773 + }, + { + "epoch": 0.7249473684210527, + "grad_norm": 0.5, + "learning_rate": 6.112280701754386e-05, + "loss": 3.0174, + "step": 13774 + }, + { + "epoch": 0.725, + "grad_norm": 0.52734375, + "learning_rate": 6.111111111111112e-05, + "loss": 3.1972, + "step": 13775 + }, + { + "epoch": 0.7250526315789474, + "grad_norm": 0.48046875, + "learning_rate": 6.109941520467837e-05, + "loss": 3.2793, + "step": 13776 + }, + { + "epoch": 0.7251052631578947, + "grad_norm": 0.53515625, + "learning_rate": 6.108771929824562e-05, + "loss": 3.2284, + "step": 13777 + }, + { + "epoch": 0.7251578947368421, + "grad_norm": 0.474609375, + "learning_rate": 6.107602339181287e-05, + "loss": 3.7042, + "step": 13778 + }, + { + "epoch": 0.7252105263157895, + "grad_norm": 0.5546875, + "learning_rate": 6.106432748538012e-05, + "loss": 3.1176, + "step": 13779 + }, + { + "epoch": 0.7252631578947368, + "grad_norm": 0.48828125, + "learning_rate": 6.105263157894737e-05, + "loss": 3.5422, + "step": 13780 + }, + { + "epoch": 0.7253157894736842, + "grad_norm": 0.47265625, + "learning_rate": 6.104093567251462e-05, + "loss": 3.2236, + "step": 13781 + }, + { + "epoch": 0.7253684210526316, + "grad_norm": 0.51953125, + "learning_rate": 6.102923976608187e-05, + "loss": 2.7642, + "step": 13782 + }, + { + "epoch": 0.725421052631579, + "grad_norm": 0.482421875, + "learning_rate": 6.101754385964913e-05, + "loss": 3.3622, + "step": 13783 + }, + { + "epoch": 0.7254736842105263, + "grad_norm": 0.50390625, + "learning_rate": 6.100584795321638e-05, + "loss": 3.1747, + "step": 13784 + }, + { + "epoch": 0.7255263157894737, + "grad_norm": 0.48046875, + "learning_rate": 6.0994152046783624e-05, + "loss": 3.3474, + "step": 13785 + }, + { + "epoch": 0.7255789473684211, + "grad_norm": 0.482421875, + "learning_rate": 6.098245614035089e-05, + "loss": 3.1882, + "step": 13786 + }, + { + "epoch": 0.7256315789473684, + "grad_norm": 0.62890625, + "learning_rate": 6.0970760233918136e-05, + "loss": 2.9189, + "step": 13787 + }, + { + "epoch": 0.7256842105263158, + "grad_norm": 0.482421875, + "learning_rate": 6.095906432748538e-05, + "loss": 3.3207, + "step": 13788 + }, + { + "epoch": 0.7257368421052631, + "grad_norm": 0.55859375, + "learning_rate": 6.094736842105263e-05, + "loss": 2.8984, + "step": 13789 + }, + { + "epoch": 0.7257894736842105, + "grad_norm": 0.494140625, + "learning_rate": 6.093567251461989e-05, + "loss": 3.4668, + "step": 13790 + }, + { + "epoch": 0.7258421052631578, + "grad_norm": 0.4921875, + "learning_rate": 6.092397660818714e-05, + "loss": 3.2741, + "step": 13791 + }, + { + "epoch": 0.7258947368421053, + "grad_norm": 0.51171875, + "learning_rate": 6.0912280701754384e-05, + "loss": 3.5877, + "step": 13792 + }, + { + "epoch": 0.7259473684210527, + "grad_norm": 0.4921875, + "learning_rate": 6.0900584795321634e-05, + "loss": 3.2868, + "step": 13793 + }, + { + "epoch": 0.726, + "grad_norm": 0.5, + "learning_rate": 6.08888888888889e-05, + "loss": 3.0721, + "step": 13794 + }, + { + "epoch": 0.7260526315789474, + "grad_norm": 0.671875, + "learning_rate": 6.0877192982456147e-05, + "loss": 3.1426, + "step": 13795 + }, + { + "epoch": 0.7261052631578947, + "grad_norm": 0.50390625, + "learning_rate": 6.086549707602339e-05, + "loss": 3.5821, + "step": 13796 + }, + { + "epoch": 0.7261578947368421, + "grad_norm": 0.490234375, + "learning_rate": 6.085380116959064e-05, + "loss": 3.5289, + "step": 13797 + }, + { + "epoch": 0.7262105263157895, + "grad_norm": 0.4921875, + "learning_rate": 6.08421052631579e-05, + "loss": 2.7719, + "step": 13798 + }, + { + "epoch": 0.7262631578947368, + "grad_norm": 0.53125, + "learning_rate": 6.083040935672515e-05, + "loss": 3.0094, + "step": 13799 + }, + { + "epoch": 0.7263157894736842, + "grad_norm": 0.4921875, + "learning_rate": 6.0818713450292395e-05, + "loss": 2.967, + "step": 13800 + }, + { + "epoch": 0.7263684210526316, + "grad_norm": 0.6328125, + "learning_rate": 6.080701754385966e-05, + "loss": 3.1682, + "step": 13801 + }, + { + "epoch": 0.726421052631579, + "grad_norm": 0.55859375, + "learning_rate": 6.079532163742691e-05, + "loss": 3.0866, + "step": 13802 + }, + { + "epoch": 0.7264736842105263, + "grad_norm": 0.494140625, + "learning_rate": 6.078362573099416e-05, + "loss": 3.3783, + "step": 13803 + }, + { + "epoch": 0.7265263157894737, + "grad_norm": 0.51953125, + "learning_rate": 6.07719298245614e-05, + "loss": 3.0768, + "step": 13804 + }, + { + "epoch": 0.7265789473684211, + "grad_norm": 0.478515625, + "learning_rate": 6.076023391812866e-05, + "loss": 3.2291, + "step": 13805 + }, + { + "epoch": 0.7266315789473684, + "grad_norm": 0.5078125, + "learning_rate": 6.074853801169591e-05, + "loss": 3.2831, + "step": 13806 + }, + { + "epoch": 0.7266842105263158, + "grad_norm": 0.515625, + "learning_rate": 6.073684210526316e-05, + "loss": 3.7824, + "step": 13807 + }, + { + "epoch": 0.7267368421052631, + "grad_norm": 0.51171875, + "learning_rate": 6.0725146198830405e-05, + "loss": 3.6606, + "step": 13808 + }, + { + "epoch": 0.7267894736842105, + "grad_norm": 0.498046875, + "learning_rate": 6.071345029239767e-05, + "loss": 3.4428, + "step": 13809 + }, + { + "epoch": 0.7268421052631578, + "grad_norm": 0.46484375, + "learning_rate": 6.070175438596492e-05, + "loss": 3.3941, + "step": 13810 + }, + { + "epoch": 0.7268947368421053, + "grad_norm": 0.4765625, + "learning_rate": 6.069005847953216e-05, + "loss": 3.4345, + "step": 13811 + }, + { + "epoch": 0.7269473684210527, + "grad_norm": 0.5703125, + "learning_rate": 6.0678362573099424e-05, + "loss": 3.3503, + "step": 13812 + }, + { + "epoch": 0.727, + "grad_norm": 0.5, + "learning_rate": 6.066666666666667e-05, + "loss": 3.1386, + "step": 13813 + }, + { + "epoch": 0.7270526315789474, + "grad_norm": 0.50390625, + "learning_rate": 6.065497076023392e-05, + "loss": 3.1646, + "step": 13814 + }, + { + "epoch": 0.7271052631578947, + "grad_norm": 0.447265625, + "learning_rate": 6.0643274853801166e-05, + "loss": 3.4477, + "step": 13815 + }, + { + "epoch": 0.7271578947368421, + "grad_norm": 0.474609375, + "learning_rate": 6.063157894736843e-05, + "loss": 3.4147, + "step": 13816 + }, + { + "epoch": 0.7272105263157894, + "grad_norm": 0.50390625, + "learning_rate": 6.061988304093568e-05, + "loss": 2.8343, + "step": 13817 + }, + { + "epoch": 0.7272631578947368, + "grad_norm": 0.5546875, + "learning_rate": 6.060818713450293e-05, + "loss": 3.4175, + "step": 13818 + }, + { + "epoch": 0.7273157894736842, + "grad_norm": 0.5078125, + "learning_rate": 6.059649122807017e-05, + "loss": 3.3923, + "step": 13819 + }, + { + "epoch": 0.7273684210526316, + "grad_norm": 0.48828125, + "learning_rate": 6.0584795321637434e-05, + "loss": 3.274, + "step": 13820 + }, + { + "epoch": 0.727421052631579, + "grad_norm": 0.54296875, + "learning_rate": 6.0573099415204683e-05, + "loss": 2.9661, + "step": 13821 + }, + { + "epoch": 0.7274736842105263, + "grad_norm": 0.478515625, + "learning_rate": 6.056140350877193e-05, + "loss": 2.9143, + "step": 13822 + }, + { + "epoch": 0.7275263157894737, + "grad_norm": 0.4921875, + "learning_rate": 6.054970760233919e-05, + "loss": 3.5281, + "step": 13823 + }, + { + "epoch": 0.7275789473684211, + "grad_norm": 0.498046875, + "learning_rate": 6.053801169590644e-05, + "loss": 3.1834, + "step": 13824 + }, + { + "epoch": 0.7276315789473684, + "grad_norm": 0.458984375, + "learning_rate": 6.052631578947369e-05, + "loss": 3.255, + "step": 13825 + }, + { + "epoch": 0.7276842105263158, + "grad_norm": 0.498046875, + "learning_rate": 6.051461988304094e-05, + "loss": 2.9014, + "step": 13826 + }, + { + "epoch": 0.7277368421052631, + "grad_norm": 0.482421875, + "learning_rate": 6.0502923976608194e-05, + "loss": 3.7193, + "step": 13827 + }, + { + "epoch": 0.7277894736842105, + "grad_norm": 0.484375, + "learning_rate": 6.0491228070175444e-05, + "loss": 3.5585, + "step": 13828 + }, + { + "epoch": 0.7278421052631578, + "grad_norm": 0.50390625, + "learning_rate": 6.0479532163742694e-05, + "loss": 3.2299, + "step": 13829 + }, + { + "epoch": 0.7278947368421053, + "grad_norm": 0.474609375, + "learning_rate": 6.046783625730994e-05, + "loss": 3.2282, + "step": 13830 + }, + { + "epoch": 0.7279473684210527, + "grad_norm": 0.484375, + "learning_rate": 6.04561403508772e-05, + "loss": 2.9219, + "step": 13831 + }, + { + "epoch": 0.728, + "grad_norm": 0.5, + "learning_rate": 6.044444444444445e-05, + "loss": 3.6134, + "step": 13832 + }, + { + "epoch": 0.7280526315789474, + "grad_norm": 0.53515625, + "learning_rate": 6.04327485380117e-05, + "loss": 3.2206, + "step": 13833 + }, + { + "epoch": 0.7281052631578947, + "grad_norm": 0.490234375, + "learning_rate": 6.042105263157894e-05, + "loss": 3.4943, + "step": 13834 + }, + { + "epoch": 0.7281578947368421, + "grad_norm": 0.474609375, + "learning_rate": 6.0409356725146205e-05, + "loss": 3.3507, + "step": 13835 + }, + { + "epoch": 0.7282105263157894, + "grad_norm": 0.5, + "learning_rate": 6.0397660818713454e-05, + "loss": 3.4429, + "step": 13836 + }, + { + "epoch": 0.7282631578947368, + "grad_norm": 0.46875, + "learning_rate": 6.0385964912280704e-05, + "loss": 3.2129, + "step": 13837 + }, + { + "epoch": 0.7283157894736842, + "grad_norm": 0.482421875, + "learning_rate": 6.037426900584796e-05, + "loss": 3.3432, + "step": 13838 + }, + { + "epoch": 0.7283684210526316, + "grad_norm": 0.515625, + "learning_rate": 6.036257309941521e-05, + "loss": 2.8906, + "step": 13839 + }, + { + "epoch": 0.728421052631579, + "grad_norm": 0.470703125, + "learning_rate": 6.035087719298246e-05, + "loss": 3.161, + "step": 13840 + }, + { + "epoch": 0.7284736842105263, + "grad_norm": 0.4921875, + "learning_rate": 6.033918128654971e-05, + "loss": 3.6326, + "step": 13841 + }, + { + "epoch": 0.7285263157894737, + "grad_norm": 0.482421875, + "learning_rate": 6.0327485380116965e-05, + "loss": 3.7212, + "step": 13842 + }, + { + "epoch": 0.7285789473684211, + "grad_norm": 0.51953125, + "learning_rate": 6.0315789473684215e-05, + "loss": 3.6942, + "step": 13843 + }, + { + "epoch": 0.7286315789473684, + "grad_norm": 0.470703125, + "learning_rate": 6.0304093567251465e-05, + "loss": 2.9058, + "step": 13844 + }, + { + "epoch": 0.7286842105263158, + "grad_norm": 0.51171875, + "learning_rate": 6.0292397660818714e-05, + "loss": 3.5614, + "step": 13845 + }, + { + "epoch": 0.7287368421052631, + "grad_norm": 0.4765625, + "learning_rate": 6.028070175438597e-05, + "loss": 3.005, + "step": 13846 + }, + { + "epoch": 0.7287894736842105, + "grad_norm": 0.51171875, + "learning_rate": 6.026900584795322e-05, + "loss": 2.95, + "step": 13847 + }, + { + "epoch": 0.7288421052631578, + "grad_norm": 0.5234375, + "learning_rate": 6.025730994152047e-05, + "loss": 3.2235, + "step": 13848 + }, + { + "epoch": 0.7288947368421053, + "grad_norm": 0.46484375, + "learning_rate": 6.0245614035087726e-05, + "loss": 3.0851, + "step": 13849 + }, + { + "epoch": 0.7289473684210527, + "grad_norm": 0.470703125, + "learning_rate": 6.0233918128654976e-05, + "loss": 3.4996, + "step": 13850 + }, + { + "epoch": 0.729, + "grad_norm": 0.49609375, + "learning_rate": 6.0222222222222225e-05, + "loss": 3.2504, + "step": 13851 + }, + { + "epoch": 0.7290526315789474, + "grad_norm": 0.53515625, + "learning_rate": 6.0210526315789475e-05, + "loss": 3.4044, + "step": 13852 + }, + { + "epoch": 0.7291052631578947, + "grad_norm": 0.546875, + "learning_rate": 6.019883040935673e-05, + "loss": 3.4765, + "step": 13853 + }, + { + "epoch": 0.7291578947368421, + "grad_norm": 0.484375, + "learning_rate": 6.018713450292398e-05, + "loss": 3.345, + "step": 13854 + }, + { + "epoch": 0.7292105263157894, + "grad_norm": 0.51953125, + "learning_rate": 6.017543859649123e-05, + "loss": 3.0436, + "step": 13855 + }, + { + "epoch": 0.7292631578947368, + "grad_norm": 0.494140625, + "learning_rate": 6.016374269005848e-05, + "loss": 3.6614, + "step": 13856 + }, + { + "epoch": 0.7293157894736843, + "grad_norm": 0.49609375, + "learning_rate": 6.0152046783625736e-05, + "loss": 3.0792, + "step": 13857 + }, + { + "epoch": 0.7293684210526316, + "grad_norm": 0.5078125, + "learning_rate": 6.0140350877192986e-05, + "loss": 3.1276, + "step": 13858 + }, + { + "epoch": 0.729421052631579, + "grad_norm": 0.55078125, + "learning_rate": 6.0128654970760236e-05, + "loss": 2.9423, + "step": 13859 + }, + { + "epoch": 0.7294736842105263, + "grad_norm": 0.5, + "learning_rate": 6.011695906432749e-05, + "loss": 3.3371, + "step": 13860 + }, + { + "epoch": 0.7295263157894737, + "grad_norm": 0.486328125, + "learning_rate": 6.010526315789474e-05, + "loss": 3.2789, + "step": 13861 + }, + { + "epoch": 0.7295789473684211, + "grad_norm": 0.462890625, + "learning_rate": 6.009356725146199e-05, + "loss": 3.1639, + "step": 13862 + }, + { + "epoch": 0.7296315789473684, + "grad_norm": 0.484375, + "learning_rate": 6.008187134502924e-05, + "loss": 3.1981, + "step": 13863 + }, + { + "epoch": 0.7296842105263158, + "grad_norm": 0.494140625, + "learning_rate": 6.00701754385965e-05, + "loss": 3.0896, + "step": 13864 + }, + { + "epoch": 0.7297368421052631, + "grad_norm": 0.5, + "learning_rate": 6.005847953216375e-05, + "loss": 3.6298, + "step": 13865 + }, + { + "epoch": 0.7297894736842105, + "grad_norm": 0.486328125, + "learning_rate": 6.0046783625730996e-05, + "loss": 3.2297, + "step": 13866 + }, + { + "epoch": 0.7298421052631578, + "grad_norm": 0.470703125, + "learning_rate": 6.0035087719298246e-05, + "loss": 3.2069, + "step": 13867 + }, + { + "epoch": 0.7298947368421053, + "grad_norm": 0.4921875, + "learning_rate": 6.00233918128655e-05, + "loss": 3.3283, + "step": 13868 + }, + { + "epoch": 0.7299473684210527, + "grad_norm": 0.49609375, + "learning_rate": 6.001169590643275e-05, + "loss": 3.1537, + "step": 13869 + }, + { + "epoch": 0.73, + "grad_norm": 0.498046875, + "learning_rate": 6e-05, + "loss": 3.2803, + "step": 13870 + }, + { + "epoch": 0.7300526315789474, + "grad_norm": 0.498046875, + "learning_rate": 5.998830409356725e-05, + "loss": 3.3182, + "step": 13871 + }, + { + "epoch": 0.7301052631578947, + "grad_norm": 0.50390625, + "learning_rate": 5.997660818713451e-05, + "loss": 3.8027, + "step": 13872 + }, + { + "epoch": 0.7301578947368421, + "grad_norm": 0.4765625, + "learning_rate": 5.996491228070176e-05, + "loss": 3.2721, + "step": 13873 + }, + { + "epoch": 0.7302105263157894, + "grad_norm": 0.515625, + "learning_rate": 5.995321637426901e-05, + "loss": 3.5818, + "step": 13874 + }, + { + "epoch": 0.7302631578947368, + "grad_norm": 0.48828125, + "learning_rate": 5.994152046783626e-05, + "loss": 3.1643, + "step": 13875 + }, + { + "epoch": 0.7303157894736843, + "grad_norm": 0.478515625, + "learning_rate": 5.992982456140351e-05, + "loss": 3.7879, + "step": 13876 + }, + { + "epoch": 0.7303684210526316, + "grad_norm": 0.48828125, + "learning_rate": 5.991812865497076e-05, + "loss": 3.2409, + "step": 13877 + }, + { + "epoch": 0.730421052631579, + "grad_norm": 0.4921875, + "learning_rate": 5.990643274853801e-05, + "loss": 3.6516, + "step": 13878 + }, + { + "epoch": 0.7304736842105263, + "grad_norm": 0.5, + "learning_rate": 5.989473684210527e-05, + "loss": 3.3329, + "step": 13879 + }, + { + "epoch": 0.7305263157894737, + "grad_norm": 0.478515625, + "learning_rate": 5.988304093567252e-05, + "loss": 2.9358, + "step": 13880 + }, + { + "epoch": 0.7305789473684211, + "grad_norm": 0.47265625, + "learning_rate": 5.987134502923977e-05, + "loss": 3.1183, + "step": 13881 + }, + { + "epoch": 0.7306315789473684, + "grad_norm": 0.4765625, + "learning_rate": 5.985964912280702e-05, + "loss": 3.4474, + "step": 13882 + }, + { + "epoch": 0.7306842105263158, + "grad_norm": 0.57421875, + "learning_rate": 5.984795321637427e-05, + "loss": 3.0772, + "step": 13883 + }, + { + "epoch": 0.7307368421052631, + "grad_norm": 0.451171875, + "learning_rate": 5.983625730994152e-05, + "loss": 2.8687, + "step": 13884 + }, + { + "epoch": 0.7307894736842105, + "grad_norm": 0.47265625, + "learning_rate": 5.982456140350877e-05, + "loss": 3.1676, + "step": 13885 + }, + { + "epoch": 0.7308421052631578, + "grad_norm": 0.474609375, + "learning_rate": 5.981286549707603e-05, + "loss": 2.8582, + "step": 13886 + }, + { + "epoch": 0.7308947368421053, + "grad_norm": 0.48828125, + "learning_rate": 5.980116959064328e-05, + "loss": 2.9892, + "step": 13887 + }, + { + "epoch": 0.7309473684210527, + "grad_norm": 0.478515625, + "learning_rate": 5.978947368421053e-05, + "loss": 3.1592, + "step": 13888 + }, + { + "epoch": 0.731, + "grad_norm": 0.494140625, + "learning_rate": 5.977777777777778e-05, + "loss": 3.1221, + "step": 13889 + }, + { + "epoch": 0.7310526315789474, + "grad_norm": 0.49609375, + "learning_rate": 5.9766081871345034e-05, + "loss": 3.413, + "step": 13890 + }, + { + "epoch": 0.7311052631578947, + "grad_norm": 0.486328125, + "learning_rate": 5.9754385964912284e-05, + "loss": 3.1432, + "step": 13891 + }, + { + "epoch": 0.7311578947368421, + "grad_norm": 0.490234375, + "learning_rate": 5.974269005847953e-05, + "loss": 3.582, + "step": 13892 + }, + { + "epoch": 0.7312105263157894, + "grad_norm": 0.45703125, + "learning_rate": 5.973099415204678e-05, + "loss": 3.318, + "step": 13893 + }, + { + "epoch": 0.7312631578947368, + "grad_norm": 0.478515625, + "learning_rate": 5.971929824561404e-05, + "loss": 3.4013, + "step": 13894 + }, + { + "epoch": 0.7313157894736843, + "grad_norm": 0.470703125, + "learning_rate": 5.970760233918129e-05, + "loss": 3.1449, + "step": 13895 + }, + { + "epoch": 0.7313684210526316, + "grad_norm": 0.51171875, + "learning_rate": 5.969590643274854e-05, + "loss": 3.4919, + "step": 13896 + }, + { + "epoch": 0.731421052631579, + "grad_norm": 0.4921875, + "learning_rate": 5.968421052631579e-05, + "loss": 3.6492, + "step": 13897 + }, + { + "epoch": 0.7314736842105263, + "grad_norm": 0.486328125, + "learning_rate": 5.9672514619883044e-05, + "loss": 3.4744, + "step": 13898 + }, + { + "epoch": 0.7315263157894737, + "grad_norm": 0.49609375, + "learning_rate": 5.9660818713450294e-05, + "loss": 3.5143, + "step": 13899 + }, + { + "epoch": 0.7315789473684211, + "grad_norm": 0.5, + "learning_rate": 5.9649122807017544e-05, + "loss": 3.5417, + "step": 13900 + }, + { + "epoch": 0.7316315789473684, + "grad_norm": 0.498046875, + "learning_rate": 5.96374269005848e-05, + "loss": 3.236, + "step": 13901 + }, + { + "epoch": 0.7316842105263158, + "grad_norm": 0.5625, + "learning_rate": 5.962573099415205e-05, + "loss": 3.3845, + "step": 13902 + }, + { + "epoch": 0.7317368421052631, + "grad_norm": 0.486328125, + "learning_rate": 5.96140350877193e-05, + "loss": 3.0772, + "step": 13903 + }, + { + "epoch": 0.7317894736842105, + "grad_norm": 0.494140625, + "learning_rate": 5.960233918128655e-05, + "loss": 3.3369, + "step": 13904 + }, + { + "epoch": 0.7318421052631578, + "grad_norm": 0.50390625, + "learning_rate": 5.9590643274853805e-05, + "loss": 3.018, + "step": 13905 + }, + { + "epoch": 0.7318947368421053, + "grad_norm": 0.478515625, + "learning_rate": 5.9578947368421055e-05, + "loss": 3.3925, + "step": 13906 + }, + { + "epoch": 0.7319473684210527, + "grad_norm": 0.48046875, + "learning_rate": 5.9567251461988304e-05, + "loss": 3.4633, + "step": 13907 + }, + { + "epoch": 0.732, + "grad_norm": 0.5078125, + "learning_rate": 5.9555555555555554e-05, + "loss": 3.326, + "step": 13908 + }, + { + "epoch": 0.7320526315789474, + "grad_norm": 0.498046875, + "learning_rate": 5.954385964912281e-05, + "loss": 3.1823, + "step": 13909 + }, + { + "epoch": 0.7321052631578947, + "grad_norm": 0.486328125, + "learning_rate": 5.953216374269006e-05, + "loss": 3.3019, + "step": 13910 + }, + { + "epoch": 0.7321578947368421, + "grad_norm": 0.5546875, + "learning_rate": 5.952046783625731e-05, + "loss": 3.2472, + "step": 13911 + }, + { + "epoch": 0.7322105263157894, + "grad_norm": 0.50390625, + "learning_rate": 5.950877192982457e-05, + "loss": 3.4309, + "step": 13912 + }, + { + "epoch": 0.7322631578947368, + "grad_norm": 0.54296875, + "learning_rate": 5.9497076023391815e-05, + "loss": 3.1477, + "step": 13913 + }, + { + "epoch": 0.7323157894736843, + "grad_norm": 0.5, + "learning_rate": 5.9485380116959065e-05, + "loss": 3.2603, + "step": 13914 + }, + { + "epoch": 0.7323684210526316, + "grad_norm": 0.52734375, + "learning_rate": 5.9473684210526315e-05, + "loss": 3.2018, + "step": 13915 + }, + { + "epoch": 0.732421052631579, + "grad_norm": 0.478515625, + "learning_rate": 5.946198830409358e-05, + "loss": 3.3748, + "step": 13916 + }, + { + "epoch": 0.7324736842105263, + "grad_norm": 0.546875, + "learning_rate": 5.945029239766082e-05, + "loss": 3.1218, + "step": 13917 + }, + { + "epoch": 0.7325263157894737, + "grad_norm": 0.494140625, + "learning_rate": 5.943859649122807e-05, + "loss": 3.2164, + "step": 13918 + }, + { + "epoch": 0.7325789473684211, + "grad_norm": 0.5546875, + "learning_rate": 5.942690058479532e-05, + "loss": 3.0315, + "step": 13919 + }, + { + "epoch": 0.7326315789473684, + "grad_norm": 0.6015625, + "learning_rate": 5.9415204678362576e-05, + "loss": 3.3055, + "step": 13920 + }, + { + "epoch": 0.7326842105263158, + "grad_norm": 0.66015625, + "learning_rate": 5.9403508771929826e-05, + "loss": 3.3, + "step": 13921 + }, + { + "epoch": 0.7327368421052631, + "grad_norm": 0.46484375, + "learning_rate": 5.9391812865497075e-05, + "loss": 2.7898, + "step": 13922 + }, + { + "epoch": 0.7327894736842105, + "grad_norm": 0.498046875, + "learning_rate": 5.938011695906434e-05, + "loss": 3.4336, + "step": 13923 + }, + { + "epoch": 0.7328421052631579, + "grad_norm": 0.484375, + "learning_rate": 5.936842105263158e-05, + "loss": 3.2896, + "step": 13924 + }, + { + "epoch": 0.7328947368421053, + "grad_norm": 0.46484375, + "learning_rate": 5.935672514619883e-05, + "loss": 3.2451, + "step": 13925 + }, + { + "epoch": 0.7329473684210527, + "grad_norm": 0.4765625, + "learning_rate": 5.934502923976608e-05, + "loss": 2.9886, + "step": 13926 + }, + { + "epoch": 0.733, + "grad_norm": 0.4921875, + "learning_rate": 5.9333333333333343e-05, + "loss": 3.563, + "step": 13927 + }, + { + "epoch": 0.7330526315789474, + "grad_norm": 0.59765625, + "learning_rate": 5.9321637426900586e-05, + "loss": 3.4071, + "step": 13928 + }, + { + "epoch": 0.7331052631578947, + "grad_norm": 0.474609375, + "learning_rate": 5.9309941520467836e-05, + "loss": 3.5667, + "step": 13929 + }, + { + "epoch": 0.7331578947368421, + "grad_norm": 0.486328125, + "learning_rate": 5.9298245614035085e-05, + "loss": 3.5344, + "step": 13930 + }, + { + "epoch": 0.7332105263157894, + "grad_norm": 0.546875, + "learning_rate": 5.928654970760235e-05, + "loss": 3.3124, + "step": 13931 + }, + { + "epoch": 0.7332631578947368, + "grad_norm": 0.515625, + "learning_rate": 5.927485380116959e-05, + "loss": 2.8753, + "step": 13932 + }, + { + "epoch": 0.7333157894736843, + "grad_norm": 0.46484375, + "learning_rate": 5.926315789473684e-05, + "loss": 3.5048, + "step": 13933 + }, + { + "epoch": 0.7333684210526316, + "grad_norm": 0.5078125, + "learning_rate": 5.925146198830409e-05, + "loss": 3.3257, + "step": 13934 + }, + { + "epoch": 0.733421052631579, + "grad_norm": 0.50390625, + "learning_rate": 5.9239766081871354e-05, + "loss": 3.5207, + "step": 13935 + }, + { + "epoch": 0.7334736842105263, + "grad_norm": 0.5234375, + "learning_rate": 5.9228070175438597e-05, + "loss": 3.3173, + "step": 13936 + }, + { + "epoch": 0.7335263157894737, + "grad_norm": 0.482421875, + "learning_rate": 5.9216374269005846e-05, + "loss": 3.3532, + "step": 13937 + }, + { + "epoch": 0.733578947368421, + "grad_norm": 0.51953125, + "learning_rate": 5.920467836257311e-05, + "loss": 3.2606, + "step": 13938 + }, + { + "epoch": 0.7336315789473684, + "grad_norm": 0.51953125, + "learning_rate": 5.919298245614036e-05, + "loss": 3.1146, + "step": 13939 + }, + { + "epoch": 0.7336842105263158, + "grad_norm": 0.48828125, + "learning_rate": 5.91812865497076e-05, + "loss": 3.0631, + "step": 13940 + }, + { + "epoch": 0.7337368421052631, + "grad_norm": 0.51171875, + "learning_rate": 5.916959064327485e-05, + "loss": 3.6999, + "step": 13941 + }, + { + "epoch": 0.7337894736842105, + "grad_norm": 0.48046875, + "learning_rate": 5.9157894736842114e-05, + "loss": 3.2175, + "step": 13942 + }, + { + "epoch": 0.7338421052631579, + "grad_norm": 0.515625, + "learning_rate": 5.914619883040936e-05, + "loss": 3.0426, + "step": 13943 + }, + { + "epoch": 0.7338947368421053, + "grad_norm": 0.5234375, + "learning_rate": 5.913450292397661e-05, + "loss": 3.3288, + "step": 13944 + }, + { + "epoch": 0.7339473684210527, + "grad_norm": 0.494140625, + "learning_rate": 5.9122807017543856e-05, + "loss": 3.5518, + "step": 13945 + }, + { + "epoch": 0.734, + "grad_norm": 0.48828125, + "learning_rate": 5.911111111111112e-05, + "loss": 3.1735, + "step": 13946 + }, + { + "epoch": 0.7340526315789474, + "grad_norm": 0.51171875, + "learning_rate": 5.909941520467836e-05, + "loss": 3.2234, + "step": 13947 + }, + { + "epoch": 0.7341052631578947, + "grad_norm": 0.51171875, + "learning_rate": 5.908771929824561e-05, + "loss": 2.9923, + "step": 13948 + }, + { + "epoch": 0.7341578947368421, + "grad_norm": 0.48046875, + "learning_rate": 5.9076023391812875e-05, + "loss": 3.2127, + "step": 13949 + }, + { + "epoch": 0.7342105263157894, + "grad_norm": 0.6484375, + "learning_rate": 5.9064327485380125e-05, + "loss": 3.1979, + "step": 13950 + }, + { + "epoch": 0.7342631578947368, + "grad_norm": 0.5078125, + "learning_rate": 5.905263157894737e-05, + "loss": 3.2684, + "step": 13951 + }, + { + "epoch": 0.7343157894736843, + "grad_norm": 0.4921875, + "learning_rate": 5.904093567251462e-05, + "loss": 3.488, + "step": 13952 + }, + { + "epoch": 0.7343684210526316, + "grad_norm": 0.51171875, + "learning_rate": 5.902923976608188e-05, + "loss": 3.4092, + "step": 13953 + }, + { + "epoch": 0.734421052631579, + "grad_norm": 0.462890625, + "learning_rate": 5.901754385964913e-05, + "loss": 3.1669, + "step": 13954 + }, + { + "epoch": 0.7344736842105263, + "grad_norm": 0.46875, + "learning_rate": 5.900584795321637e-05, + "loss": 3.3642, + "step": 13955 + }, + { + "epoch": 0.7345263157894737, + "grad_norm": 0.515625, + "learning_rate": 5.899415204678362e-05, + "loss": 3.5077, + "step": 13956 + }, + { + "epoch": 0.734578947368421, + "grad_norm": 0.484375, + "learning_rate": 5.8982456140350885e-05, + "loss": 3.628, + "step": 13957 + }, + { + "epoch": 0.7346315789473684, + "grad_norm": 0.52734375, + "learning_rate": 5.8970760233918135e-05, + "loss": 3.5341, + "step": 13958 + }, + { + "epoch": 0.7346842105263158, + "grad_norm": 0.447265625, + "learning_rate": 5.895906432748538e-05, + "loss": 3.1442, + "step": 13959 + }, + { + "epoch": 0.7347368421052631, + "grad_norm": 0.48046875, + "learning_rate": 5.894736842105263e-05, + "loss": 2.8389, + "step": 13960 + }, + { + "epoch": 0.7347894736842105, + "grad_norm": 0.4921875, + "learning_rate": 5.893567251461989e-05, + "loss": 2.9271, + "step": 13961 + }, + { + "epoch": 0.7348421052631579, + "grad_norm": 0.494140625, + "learning_rate": 5.892397660818714e-05, + "loss": 3.3961, + "step": 13962 + }, + { + "epoch": 0.7348947368421053, + "grad_norm": 0.50390625, + "learning_rate": 5.891228070175438e-05, + "loss": 3.4743, + "step": 13963 + }, + { + "epoch": 0.7349473684210527, + "grad_norm": 0.5078125, + "learning_rate": 5.8900584795321646e-05, + "loss": 3.3871, + "step": 13964 + }, + { + "epoch": 0.735, + "grad_norm": 0.51171875, + "learning_rate": 5.8888888888888896e-05, + "loss": 3.1687, + "step": 13965 + }, + { + "epoch": 0.7350526315789474, + "grad_norm": 0.478515625, + "learning_rate": 5.887719298245614e-05, + "loss": 3.2007, + "step": 13966 + }, + { + "epoch": 0.7351052631578947, + "grad_norm": 0.4921875, + "learning_rate": 5.886549707602339e-05, + "loss": 3.4245, + "step": 13967 + }, + { + "epoch": 0.7351578947368421, + "grad_norm": 0.48828125, + "learning_rate": 5.885380116959065e-05, + "loss": 3.0572, + "step": 13968 + }, + { + "epoch": 0.7352105263157894, + "grad_norm": 0.48828125, + "learning_rate": 5.88421052631579e-05, + "loss": 3.0136, + "step": 13969 + }, + { + "epoch": 0.7352631578947368, + "grad_norm": 0.51171875, + "learning_rate": 5.8830409356725144e-05, + "loss": 3.2364, + "step": 13970 + }, + { + "epoch": 0.7353157894736843, + "grad_norm": 0.498046875, + "learning_rate": 5.881871345029239e-05, + "loss": 3.3103, + "step": 13971 + }, + { + "epoch": 0.7353684210526316, + "grad_norm": 0.478515625, + "learning_rate": 5.8807017543859656e-05, + "loss": 3.4646, + "step": 13972 + }, + { + "epoch": 0.735421052631579, + "grad_norm": 0.51953125, + "learning_rate": 5.8795321637426906e-05, + "loss": 2.8879, + "step": 13973 + }, + { + "epoch": 0.7354736842105263, + "grad_norm": 0.482421875, + "learning_rate": 5.878362573099415e-05, + "loss": 2.8154, + "step": 13974 + }, + { + "epoch": 0.7355263157894737, + "grad_norm": 0.4765625, + "learning_rate": 5.877192982456141e-05, + "loss": 3.2017, + "step": 13975 + }, + { + "epoch": 0.735578947368421, + "grad_norm": 0.486328125, + "learning_rate": 5.876023391812866e-05, + "loss": 3.4081, + "step": 13976 + }, + { + "epoch": 0.7356315789473684, + "grad_norm": 0.53515625, + "learning_rate": 5.874853801169591e-05, + "loss": 3.5716, + "step": 13977 + }, + { + "epoch": 0.7356842105263158, + "grad_norm": 0.466796875, + "learning_rate": 5.8736842105263154e-05, + "loss": 3.2388, + "step": 13978 + }, + { + "epoch": 0.7357368421052631, + "grad_norm": 0.60546875, + "learning_rate": 5.872514619883042e-05, + "loss": 3.2314, + "step": 13979 + }, + { + "epoch": 0.7357894736842105, + "grad_norm": 0.51171875, + "learning_rate": 5.871345029239767e-05, + "loss": 3.5092, + "step": 13980 + }, + { + "epoch": 0.7358421052631579, + "grad_norm": 0.5078125, + "learning_rate": 5.8701754385964916e-05, + "loss": 3.1868, + "step": 13981 + }, + { + "epoch": 0.7358947368421053, + "grad_norm": 0.48828125, + "learning_rate": 5.869005847953216e-05, + "loss": 3.2969, + "step": 13982 + }, + { + "epoch": 0.7359473684210527, + "grad_norm": 0.52734375, + "learning_rate": 5.867836257309942e-05, + "loss": 2.7507, + "step": 13983 + }, + { + "epoch": 0.736, + "grad_norm": 0.494140625, + "learning_rate": 5.866666666666667e-05, + "loss": 3.5181, + "step": 13984 + }, + { + "epoch": 0.7360526315789474, + "grad_norm": 0.51171875, + "learning_rate": 5.865497076023392e-05, + "loss": 3.279, + "step": 13985 + }, + { + "epoch": 0.7361052631578947, + "grad_norm": 0.5078125, + "learning_rate": 5.864327485380118e-05, + "loss": 3.4625, + "step": 13986 + }, + { + "epoch": 0.7361578947368421, + "grad_norm": 0.5, + "learning_rate": 5.863157894736843e-05, + "loss": 3.4841, + "step": 13987 + }, + { + "epoch": 0.7362105263157894, + "grad_norm": 0.486328125, + "learning_rate": 5.861988304093568e-05, + "loss": 3.1304, + "step": 13988 + }, + { + "epoch": 0.7362631578947368, + "grad_norm": 0.50390625, + "learning_rate": 5.860818713450292e-05, + "loss": 3.7091, + "step": 13989 + }, + { + "epoch": 0.7363157894736843, + "grad_norm": 0.52734375, + "learning_rate": 5.859649122807018e-05, + "loss": 3.3374, + "step": 13990 + }, + { + "epoch": 0.7363684210526316, + "grad_norm": 0.5546875, + "learning_rate": 5.858479532163743e-05, + "loss": 3.8291, + "step": 13991 + }, + { + "epoch": 0.736421052631579, + "grad_norm": 0.486328125, + "learning_rate": 5.857309941520468e-05, + "loss": 3.5408, + "step": 13992 + }, + { + "epoch": 0.7364736842105263, + "grad_norm": 0.53125, + "learning_rate": 5.8561403508771925e-05, + "loss": 3.4816, + "step": 13993 + }, + { + "epoch": 0.7365263157894737, + "grad_norm": 0.4921875, + "learning_rate": 5.854970760233919e-05, + "loss": 2.9161, + "step": 13994 + }, + { + "epoch": 0.736578947368421, + "grad_norm": 0.4609375, + "learning_rate": 5.853801169590644e-05, + "loss": 2.6868, + "step": 13995 + }, + { + "epoch": 0.7366315789473684, + "grad_norm": 0.50390625, + "learning_rate": 5.852631578947369e-05, + "loss": 3.1898, + "step": 13996 + }, + { + "epoch": 0.7366842105263158, + "grad_norm": 0.546875, + "learning_rate": 5.851461988304093e-05, + "loss": 3.0584, + "step": 13997 + }, + { + "epoch": 0.7367368421052631, + "grad_norm": 0.48828125, + "learning_rate": 5.850292397660819e-05, + "loss": 2.8816, + "step": 13998 + }, + { + "epoch": 0.7367894736842105, + "grad_norm": 0.50390625, + "learning_rate": 5.849122807017544e-05, + "loss": 3.2639, + "step": 13999 + }, + { + "epoch": 0.7368421052631579, + "grad_norm": 0.5078125, + "learning_rate": 5.847953216374269e-05, + "loss": 3.3618, + "step": 14000 + }, + { + "epoch": 0.7368947368421053, + "grad_norm": 0.49609375, + "learning_rate": 5.846783625730995e-05, + "loss": 3.1413, + "step": 14001 + }, + { + "epoch": 0.7369473684210527, + "grad_norm": 0.484375, + "learning_rate": 5.84561403508772e-05, + "loss": 3.5307, + "step": 14002 + }, + { + "epoch": 0.737, + "grad_norm": 0.484375, + "learning_rate": 5.844444444444445e-05, + "loss": 3.3901, + "step": 14003 + }, + { + "epoch": 0.7370526315789474, + "grad_norm": 0.46484375, + "learning_rate": 5.84327485380117e-05, + "loss": 3.1011, + "step": 14004 + }, + { + "epoch": 0.7371052631578947, + "grad_norm": 0.470703125, + "learning_rate": 5.8421052631578954e-05, + "loss": 2.7507, + "step": 14005 + }, + { + "epoch": 0.7371578947368421, + "grad_norm": 0.484375, + "learning_rate": 5.8409356725146204e-05, + "loss": 3.4021, + "step": 14006 + }, + { + "epoch": 0.7372105263157894, + "grad_norm": 0.484375, + "learning_rate": 5.839766081871345e-05, + "loss": 3.4085, + "step": 14007 + }, + { + "epoch": 0.7372631578947368, + "grad_norm": 0.4921875, + "learning_rate": 5.83859649122807e-05, + "loss": 3.0853, + "step": 14008 + }, + { + "epoch": 0.7373157894736843, + "grad_norm": 0.48046875, + "learning_rate": 5.837426900584796e-05, + "loss": 3.5139, + "step": 14009 + }, + { + "epoch": 0.7373684210526316, + "grad_norm": 0.52734375, + "learning_rate": 5.836257309941521e-05, + "loss": 3.2896, + "step": 14010 + }, + { + "epoch": 0.737421052631579, + "grad_norm": 0.49609375, + "learning_rate": 5.835087719298246e-05, + "loss": 3.4981, + "step": 14011 + }, + { + "epoch": 0.7374736842105263, + "grad_norm": 0.494140625, + "learning_rate": 5.8339181286549715e-05, + "loss": 3.0368, + "step": 14012 + }, + { + "epoch": 0.7375263157894737, + "grad_norm": 0.51171875, + "learning_rate": 5.8327485380116964e-05, + "loss": 3.1507, + "step": 14013 + }, + { + "epoch": 0.737578947368421, + "grad_norm": 0.61328125, + "learning_rate": 5.8315789473684214e-05, + "loss": 2.9778, + "step": 14014 + }, + { + "epoch": 0.7376315789473684, + "grad_norm": 0.5078125, + "learning_rate": 5.8304093567251463e-05, + "loss": 3.3519, + "step": 14015 + }, + { + "epoch": 0.7376842105263158, + "grad_norm": 0.640625, + "learning_rate": 5.829239766081872e-05, + "loss": 3.2456, + "step": 14016 + }, + { + "epoch": 0.7377368421052631, + "grad_norm": 0.46875, + "learning_rate": 5.828070175438597e-05, + "loss": 3.0889, + "step": 14017 + }, + { + "epoch": 0.7377894736842106, + "grad_norm": 0.50390625, + "learning_rate": 5.826900584795322e-05, + "loss": 3.1243, + "step": 14018 + }, + { + "epoch": 0.7378421052631579, + "grad_norm": 0.482421875, + "learning_rate": 5.825730994152047e-05, + "loss": 3.0572, + "step": 14019 + }, + { + "epoch": 0.7378947368421053, + "grad_norm": 0.75390625, + "learning_rate": 5.8245614035087725e-05, + "loss": 3.1669, + "step": 14020 + }, + { + "epoch": 0.7379473684210527, + "grad_norm": 0.5078125, + "learning_rate": 5.8233918128654975e-05, + "loss": 3.2132, + "step": 14021 + }, + { + "epoch": 0.738, + "grad_norm": 0.51953125, + "learning_rate": 5.8222222222222224e-05, + "loss": 3.0558, + "step": 14022 + }, + { + "epoch": 0.7380526315789474, + "grad_norm": 0.48828125, + "learning_rate": 5.821052631578948e-05, + "loss": 3.3354, + "step": 14023 + }, + { + "epoch": 0.7381052631578947, + "grad_norm": 0.494140625, + "learning_rate": 5.819883040935673e-05, + "loss": 3.1401, + "step": 14024 + }, + { + "epoch": 0.7381578947368421, + "grad_norm": 0.49609375, + "learning_rate": 5.818713450292398e-05, + "loss": 3.4201, + "step": 14025 + }, + { + "epoch": 0.7382105263157894, + "grad_norm": 0.484375, + "learning_rate": 5.817543859649123e-05, + "loss": 3.2063, + "step": 14026 + }, + { + "epoch": 0.7382631578947368, + "grad_norm": 0.49609375, + "learning_rate": 5.8163742690058486e-05, + "loss": 3.106, + "step": 14027 + }, + { + "epoch": 0.7383157894736843, + "grad_norm": 0.50390625, + "learning_rate": 5.8152046783625735e-05, + "loss": 3.2069, + "step": 14028 + }, + { + "epoch": 0.7383684210526316, + "grad_norm": 0.5078125, + "learning_rate": 5.8140350877192985e-05, + "loss": 3.5595, + "step": 14029 + }, + { + "epoch": 0.738421052631579, + "grad_norm": 0.5, + "learning_rate": 5.8128654970760234e-05, + "loss": 3.2112, + "step": 14030 + }, + { + "epoch": 0.7384736842105263, + "grad_norm": 0.470703125, + "learning_rate": 5.811695906432749e-05, + "loss": 3.6306, + "step": 14031 + }, + { + "epoch": 0.7385263157894737, + "grad_norm": 0.486328125, + "learning_rate": 5.810526315789474e-05, + "loss": 3.2344, + "step": 14032 + }, + { + "epoch": 0.738578947368421, + "grad_norm": 0.474609375, + "learning_rate": 5.809356725146199e-05, + "loss": 3.4741, + "step": 14033 + }, + { + "epoch": 0.7386315789473684, + "grad_norm": 0.5390625, + "learning_rate": 5.808187134502924e-05, + "loss": 2.9405, + "step": 14034 + }, + { + "epoch": 0.7386842105263158, + "grad_norm": 0.51171875, + "learning_rate": 5.8070175438596496e-05, + "loss": 3.3175, + "step": 14035 + }, + { + "epoch": 0.7387368421052631, + "grad_norm": 0.484375, + "learning_rate": 5.8058479532163746e-05, + "loss": 3.3353, + "step": 14036 + }, + { + "epoch": 0.7387894736842106, + "grad_norm": 0.470703125, + "learning_rate": 5.8046783625730995e-05, + "loss": 2.9526, + "step": 14037 + }, + { + "epoch": 0.7388421052631579, + "grad_norm": 0.48046875, + "learning_rate": 5.803508771929825e-05, + "loss": 3.3311, + "step": 14038 + }, + { + "epoch": 0.7388947368421053, + "grad_norm": 0.4921875, + "learning_rate": 5.80233918128655e-05, + "loss": 3.2221, + "step": 14039 + }, + { + "epoch": 0.7389473684210527, + "grad_norm": 0.5, + "learning_rate": 5.801169590643275e-05, + "loss": 3.318, + "step": 14040 + }, + { + "epoch": 0.739, + "grad_norm": 0.53125, + "learning_rate": 5.8e-05, + "loss": 3.2152, + "step": 14041 + }, + { + "epoch": 0.7390526315789474, + "grad_norm": 0.474609375, + "learning_rate": 5.798830409356726e-05, + "loss": 3.0998, + "step": 14042 + }, + { + "epoch": 0.7391052631578947, + "grad_norm": 0.486328125, + "learning_rate": 5.7976608187134506e-05, + "loss": 3.5207, + "step": 14043 + }, + { + "epoch": 0.7391578947368421, + "grad_norm": 0.498046875, + "learning_rate": 5.7964912280701756e-05, + "loss": 3.227, + "step": 14044 + }, + { + "epoch": 0.7392105263157894, + "grad_norm": 0.515625, + "learning_rate": 5.7953216374269005e-05, + "loss": 2.879, + "step": 14045 + }, + { + "epoch": 0.7392631578947368, + "grad_norm": 0.5234375, + "learning_rate": 5.794152046783626e-05, + "loss": 3.2776, + "step": 14046 + }, + { + "epoch": 0.7393157894736843, + "grad_norm": 0.47265625, + "learning_rate": 5.792982456140351e-05, + "loss": 3.1554, + "step": 14047 + }, + { + "epoch": 0.7393684210526316, + "grad_norm": 0.5078125, + "learning_rate": 5.791812865497076e-05, + "loss": 3.0485, + "step": 14048 + }, + { + "epoch": 0.739421052631579, + "grad_norm": 0.484375, + "learning_rate": 5.790643274853802e-05, + "loss": 3.0638, + "step": 14049 + }, + { + "epoch": 0.7394736842105263, + "grad_norm": 0.474609375, + "learning_rate": 5.789473684210527e-05, + "loss": 2.7249, + "step": 14050 + }, + { + "epoch": 0.7395263157894737, + "grad_norm": 0.48046875, + "learning_rate": 5.7883040935672517e-05, + "loss": 3.6187, + "step": 14051 + }, + { + "epoch": 0.739578947368421, + "grad_norm": 0.478515625, + "learning_rate": 5.7871345029239766e-05, + "loss": 3.0627, + "step": 14052 + }, + { + "epoch": 0.7396315789473684, + "grad_norm": 0.46875, + "learning_rate": 5.785964912280702e-05, + "loss": 2.8139, + "step": 14053 + }, + { + "epoch": 0.7396842105263158, + "grad_norm": 0.45703125, + "learning_rate": 5.784795321637427e-05, + "loss": 2.841, + "step": 14054 + }, + { + "epoch": 0.7397368421052631, + "grad_norm": 0.4921875, + "learning_rate": 5.783625730994152e-05, + "loss": 3.2107, + "step": 14055 + }, + { + "epoch": 0.7397894736842106, + "grad_norm": 0.50390625, + "learning_rate": 5.782456140350877e-05, + "loss": 3.2982, + "step": 14056 + }, + { + "epoch": 0.7398421052631579, + "grad_norm": 0.5234375, + "learning_rate": 5.781286549707603e-05, + "loss": 3.0693, + "step": 14057 + }, + { + "epoch": 0.7398947368421053, + "grad_norm": 0.462890625, + "learning_rate": 5.780116959064328e-05, + "loss": 2.8247, + "step": 14058 + }, + { + "epoch": 0.7399473684210527, + "grad_norm": 0.470703125, + "learning_rate": 5.778947368421053e-05, + "loss": 3.1481, + "step": 14059 + }, + { + "epoch": 0.74, + "grad_norm": 0.46875, + "learning_rate": 5.7777777777777776e-05, + "loss": 2.9572, + "step": 14060 + }, + { + "epoch": 0.7400526315789474, + "grad_norm": 0.5234375, + "learning_rate": 5.776608187134503e-05, + "loss": 3.0325, + "step": 14061 + }, + { + "epoch": 0.7401052631578947, + "grad_norm": 0.49609375, + "learning_rate": 5.775438596491228e-05, + "loss": 3.3709, + "step": 14062 + }, + { + "epoch": 0.7401578947368421, + "grad_norm": 0.478515625, + "learning_rate": 5.774269005847953e-05, + "loss": 3.1338, + "step": 14063 + }, + { + "epoch": 0.7402105263157894, + "grad_norm": 0.494140625, + "learning_rate": 5.773099415204679e-05, + "loss": 3.6602, + "step": 14064 + }, + { + "epoch": 0.7402631578947368, + "grad_norm": 0.5078125, + "learning_rate": 5.771929824561404e-05, + "loss": 3.0011, + "step": 14065 + }, + { + "epoch": 0.7403157894736843, + "grad_norm": 0.490234375, + "learning_rate": 5.770760233918129e-05, + "loss": 3.691, + "step": 14066 + }, + { + "epoch": 0.7403684210526316, + "grad_norm": 0.484375, + "learning_rate": 5.769590643274854e-05, + "loss": 3.157, + "step": 14067 + }, + { + "epoch": 0.740421052631579, + "grad_norm": 0.5, + "learning_rate": 5.7684210526315794e-05, + "loss": 3.2508, + "step": 14068 + }, + { + "epoch": 0.7404736842105263, + "grad_norm": 0.51171875, + "learning_rate": 5.767251461988304e-05, + "loss": 3.4394, + "step": 14069 + }, + { + "epoch": 0.7405263157894737, + "grad_norm": 0.55859375, + "learning_rate": 5.766081871345029e-05, + "loss": 2.8797, + "step": 14070 + }, + { + "epoch": 0.740578947368421, + "grad_norm": 0.5625, + "learning_rate": 5.764912280701754e-05, + "loss": 3.0793, + "step": 14071 + }, + { + "epoch": 0.7406315789473684, + "grad_norm": 0.515625, + "learning_rate": 5.76374269005848e-05, + "loss": 3.0685, + "step": 14072 + }, + { + "epoch": 0.7406842105263158, + "grad_norm": 0.462890625, + "learning_rate": 5.762573099415205e-05, + "loss": 2.6891, + "step": 14073 + }, + { + "epoch": 0.7407368421052631, + "grad_norm": 0.5703125, + "learning_rate": 5.76140350877193e-05, + "loss": 3.5337, + "step": 14074 + }, + { + "epoch": 0.7407894736842106, + "grad_norm": 0.48046875, + "learning_rate": 5.7602339181286554e-05, + "loss": 3.0395, + "step": 14075 + }, + { + "epoch": 0.7408421052631579, + "grad_norm": 0.50390625, + "learning_rate": 5.7590643274853804e-05, + "loss": 3.3122, + "step": 14076 + }, + { + "epoch": 0.7408947368421053, + "grad_norm": 0.498046875, + "learning_rate": 5.7578947368421053e-05, + "loss": 3.6298, + "step": 14077 + }, + { + "epoch": 0.7409473684210526, + "grad_norm": 0.51953125, + "learning_rate": 5.75672514619883e-05, + "loss": 3.3388, + "step": 14078 + }, + { + "epoch": 0.741, + "grad_norm": 0.47265625, + "learning_rate": 5.755555555555556e-05, + "loss": 3.0724, + "step": 14079 + }, + { + "epoch": 0.7410526315789474, + "grad_norm": 0.498046875, + "learning_rate": 5.754385964912281e-05, + "loss": 3.2648, + "step": 14080 + }, + { + "epoch": 0.7411052631578947, + "grad_norm": 0.515625, + "learning_rate": 5.753216374269006e-05, + "loss": 3.3195, + "step": 14081 + }, + { + "epoch": 0.7411578947368421, + "grad_norm": 0.47265625, + "learning_rate": 5.752046783625731e-05, + "loss": 3.1756, + "step": 14082 + }, + { + "epoch": 0.7412105263157894, + "grad_norm": 0.46875, + "learning_rate": 5.7508771929824565e-05, + "loss": 3.3769, + "step": 14083 + }, + { + "epoch": 0.7412631578947368, + "grad_norm": 0.494140625, + "learning_rate": 5.7497076023391814e-05, + "loss": 2.9312, + "step": 14084 + }, + { + "epoch": 0.7413157894736843, + "grad_norm": 0.49609375, + "learning_rate": 5.7485380116959064e-05, + "loss": 3.4241, + "step": 14085 + }, + { + "epoch": 0.7413684210526316, + "grad_norm": 0.5, + "learning_rate": 5.747368421052633e-05, + "loss": 3.9682, + "step": 14086 + }, + { + "epoch": 0.741421052631579, + "grad_norm": 0.50390625, + "learning_rate": 5.746198830409357e-05, + "loss": 3.5143, + "step": 14087 + }, + { + "epoch": 0.7414736842105263, + "grad_norm": 0.49609375, + "learning_rate": 5.745029239766082e-05, + "loss": 3.3573, + "step": 14088 + }, + { + "epoch": 0.7415263157894737, + "grad_norm": 0.5078125, + "learning_rate": 5.743859649122807e-05, + "loss": 3.4657, + "step": 14089 + }, + { + "epoch": 0.741578947368421, + "grad_norm": 0.51953125, + "learning_rate": 5.742690058479533e-05, + "loss": 3.5355, + "step": 14090 + }, + { + "epoch": 0.7416315789473684, + "grad_norm": 0.48828125, + "learning_rate": 5.7415204678362575e-05, + "loss": 3.4975, + "step": 14091 + }, + { + "epoch": 0.7416842105263158, + "grad_norm": 0.4921875, + "learning_rate": 5.7403508771929824e-05, + "loss": 2.9602, + "step": 14092 + }, + { + "epoch": 0.7417368421052631, + "grad_norm": 0.482421875, + "learning_rate": 5.7391812865497074e-05, + "loss": 3.4214, + "step": 14093 + }, + { + "epoch": 0.7417894736842106, + "grad_norm": 0.50390625, + "learning_rate": 5.738011695906434e-05, + "loss": 3.311, + "step": 14094 + }, + { + "epoch": 0.7418421052631579, + "grad_norm": 0.482421875, + "learning_rate": 5.736842105263158e-05, + "loss": 3.4317, + "step": 14095 + }, + { + "epoch": 0.7418947368421053, + "grad_norm": 0.4765625, + "learning_rate": 5.735672514619883e-05, + "loss": 3.5315, + "step": 14096 + }, + { + "epoch": 0.7419473684210526, + "grad_norm": 0.5234375, + "learning_rate": 5.734502923976608e-05, + "loss": 3.2376, + "step": 14097 + }, + { + "epoch": 0.742, + "grad_norm": 0.51171875, + "learning_rate": 5.7333333333333336e-05, + "loss": 3.6718, + "step": 14098 + }, + { + "epoch": 0.7420526315789474, + "grad_norm": 0.484375, + "learning_rate": 5.7321637426900585e-05, + "loss": 3.5099, + "step": 14099 + }, + { + "epoch": 0.7421052631578947, + "grad_norm": 0.494140625, + "learning_rate": 5.7309941520467835e-05, + "loss": 3.6486, + "step": 14100 + }, + { + "epoch": 0.7421578947368421, + "grad_norm": 3.0, + "learning_rate": 5.72982456140351e-05, + "loss": 3.3043, + "step": 14101 + }, + { + "epoch": 0.7422105263157894, + "grad_norm": 0.50390625, + "learning_rate": 5.728654970760234e-05, + "loss": 3.3025, + "step": 14102 + }, + { + "epoch": 0.7422631578947368, + "grad_norm": 0.59765625, + "learning_rate": 5.727485380116959e-05, + "loss": 3.2193, + "step": 14103 + }, + { + "epoch": 0.7423157894736843, + "grad_norm": 0.51171875, + "learning_rate": 5.726315789473684e-05, + "loss": 3.3506, + "step": 14104 + }, + { + "epoch": 0.7423684210526316, + "grad_norm": 0.50390625, + "learning_rate": 5.72514619883041e-05, + "loss": 3.286, + "step": 14105 + }, + { + "epoch": 0.742421052631579, + "grad_norm": 0.498046875, + "learning_rate": 5.7239766081871346e-05, + "loss": 3.2733, + "step": 14106 + }, + { + "epoch": 0.7424736842105263, + "grad_norm": 0.5, + "learning_rate": 5.7228070175438595e-05, + "loss": 3.2508, + "step": 14107 + }, + { + "epoch": 0.7425263157894737, + "grad_norm": 0.484375, + "learning_rate": 5.7216374269005845e-05, + "loss": 3.1962, + "step": 14108 + }, + { + "epoch": 0.742578947368421, + "grad_norm": 0.474609375, + "learning_rate": 5.720467836257311e-05, + "loss": 3.3165, + "step": 14109 + }, + { + "epoch": 0.7426315789473684, + "grad_norm": 0.5390625, + "learning_rate": 5.719298245614035e-05, + "loss": 3.3339, + "step": 14110 + }, + { + "epoch": 0.7426842105263158, + "grad_norm": 0.486328125, + "learning_rate": 5.71812865497076e-05, + "loss": 3.2588, + "step": 14111 + }, + { + "epoch": 0.7427368421052631, + "grad_norm": 0.59375, + "learning_rate": 5.7169590643274864e-05, + "loss": 3.0646, + "step": 14112 + }, + { + "epoch": 0.7427894736842106, + "grad_norm": 0.5, + "learning_rate": 5.715789473684211e-05, + "loss": 3.1569, + "step": 14113 + }, + { + "epoch": 0.7428421052631579, + "grad_norm": 0.48828125, + "learning_rate": 5.7146198830409356e-05, + "loss": 2.9192, + "step": 14114 + }, + { + "epoch": 0.7428947368421053, + "grad_norm": 0.50390625, + "learning_rate": 5.7134502923976606e-05, + "loss": 3.5647, + "step": 14115 + }, + { + "epoch": 0.7429473684210526, + "grad_norm": 0.50390625, + "learning_rate": 5.712280701754387e-05, + "loss": 3.2425, + "step": 14116 + }, + { + "epoch": 0.743, + "grad_norm": 0.486328125, + "learning_rate": 5.711111111111112e-05, + "loss": 3.0829, + "step": 14117 + }, + { + "epoch": 0.7430526315789474, + "grad_norm": 0.5078125, + "learning_rate": 5.709941520467836e-05, + "loss": 3.5883, + "step": 14118 + }, + { + "epoch": 0.7431052631578947, + "grad_norm": 0.48046875, + "learning_rate": 5.708771929824561e-05, + "loss": 3.4024, + "step": 14119 + }, + { + "epoch": 0.7431578947368421, + "grad_norm": 0.49609375, + "learning_rate": 5.7076023391812874e-05, + "loss": 3.5737, + "step": 14120 + }, + { + "epoch": 0.7432105263157894, + "grad_norm": 0.4921875, + "learning_rate": 5.706432748538012e-05, + "loss": 3.5061, + "step": 14121 + }, + { + "epoch": 0.7432631578947368, + "grad_norm": 0.4765625, + "learning_rate": 5.7052631578947366e-05, + "loss": 2.9697, + "step": 14122 + }, + { + "epoch": 0.7433157894736843, + "grad_norm": 0.50390625, + "learning_rate": 5.7040935672514616e-05, + "loss": 2.9156, + "step": 14123 + }, + { + "epoch": 0.7433684210526316, + "grad_norm": 0.48828125, + "learning_rate": 5.702923976608188e-05, + "loss": 3.5207, + "step": 14124 + }, + { + "epoch": 0.743421052631579, + "grad_norm": 0.51953125, + "learning_rate": 5.701754385964912e-05, + "loss": 3.613, + "step": 14125 + }, + { + "epoch": 0.7434736842105263, + "grad_norm": 0.515625, + "learning_rate": 5.700584795321637e-05, + "loss": 3.5178, + "step": 14126 + }, + { + "epoch": 0.7435263157894737, + "grad_norm": 0.482421875, + "learning_rate": 5.6994152046783635e-05, + "loss": 3.4845, + "step": 14127 + }, + { + "epoch": 0.743578947368421, + "grad_norm": 0.4921875, + "learning_rate": 5.6982456140350884e-05, + "loss": 3.3531, + "step": 14128 + }, + { + "epoch": 0.7436315789473684, + "grad_norm": 0.4921875, + "learning_rate": 5.697076023391813e-05, + "loss": 3.2903, + "step": 14129 + }, + { + "epoch": 0.7436842105263158, + "grad_norm": 0.5, + "learning_rate": 5.695906432748538e-05, + "loss": 3.0365, + "step": 14130 + }, + { + "epoch": 0.7437368421052631, + "grad_norm": 0.48046875, + "learning_rate": 5.694736842105264e-05, + "loss": 2.7893, + "step": 14131 + }, + { + "epoch": 0.7437894736842106, + "grad_norm": 0.49609375, + "learning_rate": 5.693567251461989e-05, + "loss": 2.9912, + "step": 14132 + }, + { + "epoch": 0.7438421052631579, + "grad_norm": 0.486328125, + "learning_rate": 5.692397660818713e-05, + "loss": 3.6875, + "step": 14133 + }, + { + "epoch": 0.7438947368421053, + "grad_norm": 0.51171875, + "learning_rate": 5.691228070175438e-05, + "loss": 3.306, + "step": 14134 + }, + { + "epoch": 0.7439473684210526, + "grad_norm": 0.470703125, + "learning_rate": 5.6900584795321645e-05, + "loss": 3.2334, + "step": 14135 + }, + { + "epoch": 0.744, + "grad_norm": 0.494140625, + "learning_rate": 5.6888888888888895e-05, + "loss": 3.3068, + "step": 14136 + }, + { + "epoch": 0.7440526315789474, + "grad_norm": 0.490234375, + "learning_rate": 5.687719298245614e-05, + "loss": 3.2429, + "step": 14137 + }, + { + "epoch": 0.7441052631578947, + "grad_norm": 0.490234375, + "learning_rate": 5.68654970760234e-05, + "loss": 2.9533, + "step": 14138 + }, + { + "epoch": 0.7441578947368421, + "grad_norm": 0.53515625, + "learning_rate": 5.685380116959065e-05, + "loss": 3.1239, + "step": 14139 + }, + { + "epoch": 0.7442105263157894, + "grad_norm": 0.482421875, + "learning_rate": 5.68421052631579e-05, + "loss": 3.4594, + "step": 14140 + }, + { + "epoch": 0.7442631578947368, + "grad_norm": 0.58984375, + "learning_rate": 5.683040935672514e-05, + "loss": 2.8912, + "step": 14141 + }, + { + "epoch": 0.7443157894736843, + "grad_norm": 0.482421875, + "learning_rate": 5.6818713450292406e-05, + "loss": 3.4205, + "step": 14142 + }, + { + "epoch": 0.7443684210526316, + "grad_norm": 0.515625, + "learning_rate": 5.6807017543859655e-05, + "loss": 3.1842, + "step": 14143 + }, + { + "epoch": 0.744421052631579, + "grad_norm": 0.484375, + "learning_rate": 5.67953216374269e-05, + "loss": 3.0861, + "step": 14144 + }, + { + "epoch": 0.7444736842105263, + "grad_norm": 0.51171875, + "learning_rate": 5.678362573099415e-05, + "loss": 3.3916, + "step": 14145 + }, + { + "epoch": 0.7445263157894737, + "grad_norm": 0.490234375, + "learning_rate": 5.677192982456141e-05, + "loss": 3.3217, + "step": 14146 + }, + { + "epoch": 0.744578947368421, + "grad_norm": 0.51171875, + "learning_rate": 5.676023391812866e-05, + "loss": 2.9929, + "step": 14147 + }, + { + "epoch": 0.7446315789473684, + "grad_norm": 0.70703125, + "learning_rate": 5.67485380116959e-05, + "loss": 3.6902, + "step": 14148 + }, + { + "epoch": 0.7446842105263158, + "grad_norm": 0.49609375, + "learning_rate": 5.6736842105263166e-05, + "loss": 3.4198, + "step": 14149 + }, + { + "epoch": 0.7447368421052631, + "grad_norm": 0.48828125, + "learning_rate": 5.6725146198830416e-05, + "loss": 3.4537, + "step": 14150 + }, + { + "epoch": 0.7447894736842106, + "grad_norm": 0.54296875, + "learning_rate": 5.6713450292397666e-05, + "loss": 3.2765, + "step": 14151 + }, + { + "epoch": 0.7448421052631579, + "grad_norm": 0.51953125, + "learning_rate": 5.670175438596491e-05, + "loss": 3.1525, + "step": 14152 + }, + { + "epoch": 0.7448947368421053, + "grad_norm": 0.58203125, + "learning_rate": 5.669005847953217e-05, + "loss": 3.0045, + "step": 14153 + }, + { + "epoch": 0.7449473684210526, + "grad_norm": 0.52734375, + "learning_rate": 5.667836257309942e-05, + "loss": 3.6588, + "step": 14154 + }, + { + "epoch": 0.745, + "grad_norm": 0.5078125, + "learning_rate": 5.666666666666667e-05, + "loss": 3.4353, + "step": 14155 + }, + { + "epoch": 0.7450526315789474, + "grad_norm": 0.478515625, + "learning_rate": 5.6654970760233914e-05, + "loss": 3.4865, + "step": 14156 + }, + { + "epoch": 0.7451052631578947, + "grad_norm": 0.47265625, + "learning_rate": 5.664327485380118e-05, + "loss": 3.415, + "step": 14157 + }, + { + "epoch": 0.7451578947368421, + "grad_norm": 0.484375, + "learning_rate": 5.6631578947368426e-05, + "loss": 3.2961, + "step": 14158 + }, + { + "epoch": 0.7452105263157894, + "grad_norm": 0.484375, + "learning_rate": 5.6619883040935676e-05, + "loss": 3.376, + "step": 14159 + }, + { + "epoch": 0.7452631578947368, + "grad_norm": 0.48046875, + "learning_rate": 5.660818713450292e-05, + "loss": 3.5296, + "step": 14160 + }, + { + "epoch": 0.7453157894736843, + "grad_norm": 0.50390625, + "learning_rate": 5.659649122807018e-05, + "loss": 3.2128, + "step": 14161 + }, + { + "epoch": 0.7453684210526316, + "grad_norm": 0.5, + "learning_rate": 5.658479532163743e-05, + "loss": 3.1922, + "step": 14162 + }, + { + "epoch": 0.745421052631579, + "grad_norm": 0.490234375, + "learning_rate": 5.657309941520468e-05, + "loss": 3.6208, + "step": 14163 + }, + { + "epoch": 0.7454736842105263, + "grad_norm": 0.49609375, + "learning_rate": 5.656140350877194e-05, + "loss": 3.6827, + "step": 14164 + }, + { + "epoch": 0.7455263157894737, + "grad_norm": 0.48828125, + "learning_rate": 5.654970760233919e-05, + "loss": 3.3633, + "step": 14165 + }, + { + "epoch": 0.745578947368421, + "grad_norm": 0.5, + "learning_rate": 5.6538011695906437e-05, + "loss": 3.3858, + "step": 14166 + }, + { + "epoch": 0.7456315789473684, + "grad_norm": 0.5, + "learning_rate": 5.652631578947368e-05, + "loss": 3.3976, + "step": 14167 + }, + { + "epoch": 0.7456842105263158, + "grad_norm": 0.48828125, + "learning_rate": 5.651461988304094e-05, + "loss": 3.3378, + "step": 14168 + }, + { + "epoch": 0.7457368421052631, + "grad_norm": 0.494140625, + "learning_rate": 5.650292397660819e-05, + "loss": 3.3123, + "step": 14169 + }, + { + "epoch": 0.7457894736842106, + "grad_norm": 0.482421875, + "learning_rate": 5.649122807017544e-05, + "loss": 3.501, + "step": 14170 + }, + { + "epoch": 0.7458421052631579, + "grad_norm": 0.71484375, + "learning_rate": 5.6479532163742685e-05, + "loss": 3.049, + "step": 14171 + }, + { + "epoch": 0.7458947368421053, + "grad_norm": 0.5, + "learning_rate": 5.646783625730995e-05, + "loss": 3.2413, + "step": 14172 + }, + { + "epoch": 0.7459473684210526, + "grad_norm": 0.50390625, + "learning_rate": 5.64561403508772e-05, + "loss": 3.2706, + "step": 14173 + }, + { + "epoch": 0.746, + "grad_norm": 0.50390625, + "learning_rate": 5.644444444444445e-05, + "loss": 3.3523, + "step": 14174 + }, + { + "epoch": 0.7460526315789474, + "grad_norm": 0.4765625, + "learning_rate": 5.64327485380117e-05, + "loss": 3.0414, + "step": 14175 + }, + { + "epoch": 0.7461052631578947, + "grad_norm": 0.486328125, + "learning_rate": 5.642105263157895e-05, + "loss": 3.2666, + "step": 14176 + }, + { + "epoch": 0.7461578947368421, + "grad_norm": 0.478515625, + "learning_rate": 5.64093567251462e-05, + "loss": 3.0512, + "step": 14177 + }, + { + "epoch": 0.7462105263157894, + "grad_norm": 0.470703125, + "learning_rate": 5.639766081871345e-05, + "loss": 2.7705, + "step": 14178 + }, + { + "epoch": 0.7462631578947369, + "grad_norm": 0.486328125, + "learning_rate": 5.638596491228071e-05, + "loss": 3.2727, + "step": 14179 + }, + { + "epoch": 0.7463157894736843, + "grad_norm": 0.470703125, + "learning_rate": 5.637426900584796e-05, + "loss": 3.415, + "step": 14180 + }, + { + "epoch": 0.7463684210526316, + "grad_norm": 0.46875, + "learning_rate": 5.636257309941521e-05, + "loss": 3.2321, + "step": 14181 + }, + { + "epoch": 0.746421052631579, + "grad_norm": 0.546875, + "learning_rate": 5.635087719298246e-05, + "loss": 3.2778, + "step": 14182 + }, + { + "epoch": 0.7464736842105263, + "grad_norm": 0.50390625, + "learning_rate": 5.6339181286549714e-05, + "loss": 3.68, + "step": 14183 + }, + { + "epoch": 0.7465263157894737, + "grad_norm": 0.478515625, + "learning_rate": 5.632748538011696e-05, + "loss": 3.615, + "step": 14184 + }, + { + "epoch": 0.746578947368421, + "grad_norm": 0.48046875, + "learning_rate": 5.631578947368421e-05, + "loss": 3.3738, + "step": 14185 + }, + { + "epoch": 0.7466315789473684, + "grad_norm": 0.5, + "learning_rate": 5.630409356725147e-05, + "loss": 3.3881, + "step": 14186 + }, + { + "epoch": 0.7466842105263158, + "grad_norm": 0.490234375, + "learning_rate": 5.629239766081872e-05, + "loss": 3.4729, + "step": 14187 + }, + { + "epoch": 0.7467368421052631, + "grad_norm": 0.486328125, + "learning_rate": 5.628070175438597e-05, + "loss": 3.3009, + "step": 14188 + }, + { + "epoch": 0.7467894736842106, + "grad_norm": 0.470703125, + "learning_rate": 5.626900584795322e-05, + "loss": 3.469, + "step": 14189 + }, + { + "epoch": 0.7468421052631579, + "grad_norm": 0.52734375, + "learning_rate": 5.6257309941520474e-05, + "loss": 3.5417, + "step": 14190 + }, + { + "epoch": 0.7468947368421053, + "grad_norm": 0.50390625, + "learning_rate": 5.6245614035087724e-05, + "loss": 3.5, + "step": 14191 + }, + { + "epoch": 0.7469473684210526, + "grad_norm": 0.462890625, + "learning_rate": 5.6233918128654973e-05, + "loss": 3.3286, + "step": 14192 + }, + { + "epoch": 0.747, + "grad_norm": 0.55078125, + "learning_rate": 5.622222222222222e-05, + "loss": 2.9724, + "step": 14193 + }, + { + "epoch": 0.7470526315789474, + "grad_norm": 0.498046875, + "learning_rate": 5.621052631578948e-05, + "loss": 3.09, + "step": 14194 + }, + { + "epoch": 0.7471052631578947, + "grad_norm": 0.48828125, + "learning_rate": 5.619883040935673e-05, + "loss": 3.3561, + "step": 14195 + }, + { + "epoch": 0.7471578947368421, + "grad_norm": 0.5, + "learning_rate": 5.618713450292398e-05, + "loss": 3.5087, + "step": 14196 + }, + { + "epoch": 0.7472105263157894, + "grad_norm": 0.515625, + "learning_rate": 5.617543859649123e-05, + "loss": 3.1972, + "step": 14197 + }, + { + "epoch": 0.7472631578947369, + "grad_norm": 0.49609375, + "learning_rate": 5.6163742690058485e-05, + "loss": 3.3206, + "step": 14198 + }, + { + "epoch": 0.7473157894736842, + "grad_norm": 0.5, + "learning_rate": 5.6152046783625734e-05, + "loss": 3.3676, + "step": 14199 + }, + { + "epoch": 0.7473684210526316, + "grad_norm": 0.55859375, + "learning_rate": 5.6140350877192984e-05, + "loss": 2.9748, + "step": 14200 + }, + { + "epoch": 0.747421052631579, + "grad_norm": 0.470703125, + "learning_rate": 5.612865497076024e-05, + "loss": 3.6752, + "step": 14201 + }, + { + "epoch": 0.7474736842105263, + "grad_norm": 0.486328125, + "learning_rate": 5.611695906432749e-05, + "loss": 3.3545, + "step": 14202 + }, + { + "epoch": 0.7475263157894737, + "grad_norm": 0.470703125, + "learning_rate": 5.610526315789474e-05, + "loss": 3.6109, + "step": 14203 + }, + { + "epoch": 0.747578947368421, + "grad_norm": 0.52734375, + "learning_rate": 5.609356725146199e-05, + "loss": 3.3743, + "step": 14204 + }, + { + "epoch": 0.7476315789473684, + "grad_norm": 0.51171875, + "learning_rate": 5.6081871345029245e-05, + "loss": 3.3595, + "step": 14205 + }, + { + "epoch": 0.7476842105263158, + "grad_norm": 0.50390625, + "learning_rate": 5.6070175438596495e-05, + "loss": 3.3591, + "step": 14206 + }, + { + "epoch": 0.7477368421052631, + "grad_norm": 0.490234375, + "learning_rate": 5.6058479532163744e-05, + "loss": 3.2013, + "step": 14207 + }, + { + "epoch": 0.7477894736842106, + "grad_norm": 0.484375, + "learning_rate": 5.6046783625730994e-05, + "loss": 3.7311, + "step": 14208 + }, + { + "epoch": 0.7478421052631579, + "grad_norm": 0.490234375, + "learning_rate": 5.603508771929825e-05, + "loss": 3.2672, + "step": 14209 + }, + { + "epoch": 0.7478947368421053, + "grad_norm": 0.50390625, + "learning_rate": 5.60233918128655e-05, + "loss": 3.3928, + "step": 14210 + }, + { + "epoch": 0.7479473684210526, + "grad_norm": 0.7578125, + "learning_rate": 5.601169590643275e-05, + "loss": 3.2622, + "step": 14211 + }, + { + "epoch": 0.748, + "grad_norm": 0.455078125, + "learning_rate": 5.6000000000000006e-05, + "loss": 2.9503, + "step": 14212 + }, + { + "epoch": 0.7480526315789474, + "grad_norm": 0.484375, + "learning_rate": 5.5988304093567256e-05, + "loss": 3.5362, + "step": 14213 + }, + { + "epoch": 0.7481052631578947, + "grad_norm": 0.49609375, + "learning_rate": 5.5976608187134505e-05, + "loss": 3.2007, + "step": 14214 + }, + { + "epoch": 0.7481578947368421, + "grad_norm": 1.28125, + "learning_rate": 5.5964912280701755e-05, + "loss": 3.4951, + "step": 14215 + }, + { + "epoch": 0.7482105263157894, + "grad_norm": 0.490234375, + "learning_rate": 5.595321637426901e-05, + "loss": 3.332, + "step": 14216 + }, + { + "epoch": 0.7482631578947369, + "grad_norm": 0.498046875, + "learning_rate": 5.594152046783626e-05, + "loss": 3.4171, + "step": 14217 + }, + { + "epoch": 0.7483157894736842, + "grad_norm": 0.474609375, + "learning_rate": 5.592982456140351e-05, + "loss": 3.0089, + "step": 14218 + }, + { + "epoch": 0.7483684210526316, + "grad_norm": 0.6484375, + "learning_rate": 5.591812865497076e-05, + "loss": 3.7965, + "step": 14219 + }, + { + "epoch": 0.748421052631579, + "grad_norm": 0.47265625, + "learning_rate": 5.5906432748538016e-05, + "loss": 3.0641, + "step": 14220 + }, + { + "epoch": 0.7484736842105263, + "grad_norm": 0.490234375, + "learning_rate": 5.5894736842105266e-05, + "loss": 2.8036, + "step": 14221 + }, + { + "epoch": 0.7485263157894737, + "grad_norm": 0.54296875, + "learning_rate": 5.5883040935672515e-05, + "loss": 3.4411, + "step": 14222 + }, + { + "epoch": 0.748578947368421, + "grad_norm": 0.50390625, + "learning_rate": 5.5871345029239765e-05, + "loss": 3.1208, + "step": 14223 + }, + { + "epoch": 0.7486315789473684, + "grad_norm": 0.54296875, + "learning_rate": 5.585964912280702e-05, + "loss": 2.9243, + "step": 14224 + }, + { + "epoch": 0.7486842105263158, + "grad_norm": 0.494140625, + "learning_rate": 5.584795321637427e-05, + "loss": 3.4372, + "step": 14225 + }, + { + "epoch": 0.7487368421052631, + "grad_norm": 0.458984375, + "learning_rate": 5.583625730994152e-05, + "loss": 3.305, + "step": 14226 + }, + { + "epoch": 0.7487894736842106, + "grad_norm": 0.494140625, + "learning_rate": 5.582456140350878e-05, + "loss": 3.4274, + "step": 14227 + }, + { + "epoch": 0.7488421052631579, + "grad_norm": 0.48828125, + "learning_rate": 5.5812865497076027e-05, + "loss": 3.1213, + "step": 14228 + }, + { + "epoch": 0.7488947368421053, + "grad_norm": 0.63671875, + "learning_rate": 5.5801169590643276e-05, + "loss": 3.415, + "step": 14229 + }, + { + "epoch": 0.7489473684210526, + "grad_norm": 0.52734375, + "learning_rate": 5.5789473684210526e-05, + "loss": 3.0688, + "step": 14230 + }, + { + "epoch": 0.749, + "grad_norm": 0.482421875, + "learning_rate": 5.577777777777778e-05, + "loss": 3.2613, + "step": 14231 + }, + { + "epoch": 0.7490526315789474, + "grad_norm": 0.4765625, + "learning_rate": 5.576608187134503e-05, + "loss": 3.1165, + "step": 14232 + }, + { + "epoch": 0.7491052631578947, + "grad_norm": 0.48828125, + "learning_rate": 5.575438596491228e-05, + "loss": 3.4283, + "step": 14233 + }, + { + "epoch": 0.7491578947368421, + "grad_norm": 0.51953125, + "learning_rate": 5.574269005847953e-05, + "loss": 3.2945, + "step": 14234 + }, + { + "epoch": 0.7492105263157894, + "grad_norm": 0.486328125, + "learning_rate": 5.573099415204679e-05, + "loss": 3.7652, + "step": 14235 + }, + { + "epoch": 0.7492631578947369, + "grad_norm": 0.5546875, + "learning_rate": 5.571929824561404e-05, + "loss": 3.3355, + "step": 14236 + }, + { + "epoch": 0.7493157894736842, + "grad_norm": 0.478515625, + "learning_rate": 5.5707602339181286e-05, + "loss": 3.3993, + "step": 14237 + }, + { + "epoch": 0.7493684210526316, + "grad_norm": 0.53515625, + "learning_rate": 5.569590643274854e-05, + "loss": 3.4137, + "step": 14238 + }, + { + "epoch": 0.749421052631579, + "grad_norm": 0.52734375, + "learning_rate": 5.568421052631579e-05, + "loss": 3.3423, + "step": 14239 + }, + { + "epoch": 0.7494736842105263, + "grad_norm": 0.50390625, + "learning_rate": 5.567251461988304e-05, + "loss": 3.097, + "step": 14240 + }, + { + "epoch": 0.7495263157894737, + "grad_norm": 0.51953125, + "learning_rate": 5.566081871345029e-05, + "loss": 3.3034, + "step": 14241 + }, + { + "epoch": 0.749578947368421, + "grad_norm": 0.51953125, + "learning_rate": 5.564912280701755e-05, + "loss": 2.9614, + "step": 14242 + }, + { + "epoch": 0.7496315789473684, + "grad_norm": 0.486328125, + "learning_rate": 5.56374269005848e-05, + "loss": 3.34, + "step": 14243 + }, + { + "epoch": 0.7496842105263158, + "grad_norm": 0.50390625, + "learning_rate": 5.562573099415205e-05, + "loss": 3.4639, + "step": 14244 + }, + { + "epoch": 0.7497368421052631, + "grad_norm": 0.484375, + "learning_rate": 5.56140350877193e-05, + "loss": 3.3155, + "step": 14245 + }, + { + "epoch": 0.7497894736842106, + "grad_norm": 0.498046875, + "learning_rate": 5.560233918128655e-05, + "loss": 3.2596, + "step": 14246 + }, + { + "epoch": 0.7498421052631579, + "grad_norm": 0.5390625, + "learning_rate": 5.55906432748538e-05, + "loss": 2.855, + "step": 14247 + }, + { + "epoch": 0.7498947368421053, + "grad_norm": 0.51953125, + "learning_rate": 5.557894736842105e-05, + "loss": 3.2769, + "step": 14248 + }, + { + "epoch": 0.7499473684210526, + "grad_norm": 0.49609375, + "learning_rate": 5.5567251461988315e-05, + "loss": 3.6816, + "step": 14249 + }, + { + "epoch": 0.75, + "grad_norm": 0.4609375, + "learning_rate": 5.555555555555556e-05, + "loss": 3.093, + "step": 14250 + }, + { + "epoch": 0.7500526315789474, + "grad_norm": 0.53515625, + "learning_rate": 5.554385964912281e-05, + "loss": 3.2978, + "step": 14251 + }, + { + "epoch": 0.7501052631578947, + "grad_norm": 0.494140625, + "learning_rate": 5.553216374269006e-05, + "loss": 3.3226, + "step": 14252 + }, + { + "epoch": 0.7501578947368421, + "grad_norm": 0.484375, + "learning_rate": 5.552046783625732e-05, + "loss": 3.3296, + "step": 14253 + }, + { + "epoch": 0.7502105263157894, + "grad_norm": 0.4921875, + "learning_rate": 5.550877192982456e-05, + "loss": 3.6138, + "step": 14254 + }, + { + "epoch": 0.7502631578947369, + "grad_norm": 0.482421875, + "learning_rate": 5.549707602339181e-05, + "loss": 3.4305, + "step": 14255 + }, + { + "epoch": 0.7503157894736842, + "grad_norm": 0.50390625, + "learning_rate": 5.548538011695906e-05, + "loss": 3.2773, + "step": 14256 + }, + { + "epoch": 0.7503684210526316, + "grad_norm": 0.48828125, + "learning_rate": 5.547368421052632e-05, + "loss": 3.5781, + "step": 14257 + }, + { + "epoch": 0.750421052631579, + "grad_norm": 0.87109375, + "learning_rate": 5.546198830409357e-05, + "loss": 3.2118, + "step": 14258 + }, + { + "epoch": 0.7504736842105263, + "grad_norm": 0.50390625, + "learning_rate": 5.545029239766082e-05, + "loss": 3.2644, + "step": 14259 + }, + { + "epoch": 0.7505263157894737, + "grad_norm": 0.490234375, + "learning_rate": 5.543859649122807e-05, + "loss": 3.3697, + "step": 14260 + }, + { + "epoch": 0.750578947368421, + "grad_norm": 0.470703125, + "learning_rate": 5.5426900584795324e-05, + "loss": 3.1141, + "step": 14261 + }, + { + "epoch": 0.7506315789473684, + "grad_norm": 0.51953125, + "learning_rate": 5.5415204678362574e-05, + "loss": 3.4532, + "step": 14262 + }, + { + "epoch": 0.7506842105263158, + "grad_norm": 0.50390625, + "learning_rate": 5.540350877192982e-05, + "loss": 3.7167, + "step": 14263 + }, + { + "epoch": 0.7507368421052631, + "grad_norm": 0.482421875, + "learning_rate": 5.5391812865497086e-05, + "loss": 3.3965, + "step": 14264 + }, + { + "epoch": 0.7507894736842106, + "grad_norm": 0.494140625, + "learning_rate": 5.538011695906433e-05, + "loss": 3.5347, + "step": 14265 + }, + { + "epoch": 0.7508421052631579, + "grad_norm": 0.51171875, + "learning_rate": 5.536842105263158e-05, + "loss": 3.0885, + "step": 14266 + }, + { + "epoch": 0.7508947368421053, + "grad_norm": 0.486328125, + "learning_rate": 5.535672514619883e-05, + "loss": 2.9445, + "step": 14267 + }, + { + "epoch": 0.7509473684210526, + "grad_norm": 0.47265625, + "learning_rate": 5.534502923976609e-05, + "loss": 3.1611, + "step": 14268 + }, + { + "epoch": 0.751, + "grad_norm": 0.51953125, + "learning_rate": 5.5333333333333334e-05, + "loss": 3.1096, + "step": 14269 + }, + { + "epoch": 0.7510526315789474, + "grad_norm": 0.52734375, + "learning_rate": 5.5321637426900584e-05, + "loss": 3.0137, + "step": 14270 + }, + { + "epoch": 0.7511052631578947, + "grad_norm": 0.515625, + "learning_rate": 5.5309941520467834e-05, + "loss": 3.5653, + "step": 14271 + }, + { + "epoch": 0.7511578947368421, + "grad_norm": 0.49609375, + "learning_rate": 5.52982456140351e-05, + "loss": 2.9357, + "step": 14272 + }, + { + "epoch": 0.7512105263157894, + "grad_norm": 0.462890625, + "learning_rate": 5.528654970760234e-05, + "loss": 3.4012, + "step": 14273 + }, + { + "epoch": 0.7512631578947369, + "grad_norm": 0.5078125, + "learning_rate": 5.527485380116959e-05, + "loss": 2.8503, + "step": 14274 + }, + { + "epoch": 0.7513157894736842, + "grad_norm": 0.490234375, + "learning_rate": 5.526315789473685e-05, + "loss": 3.4407, + "step": 14275 + }, + { + "epoch": 0.7513684210526316, + "grad_norm": 0.458984375, + "learning_rate": 5.52514619883041e-05, + "loss": 3.1035, + "step": 14276 + }, + { + "epoch": 0.751421052631579, + "grad_norm": 0.53515625, + "learning_rate": 5.5239766081871345e-05, + "loss": 3.4399, + "step": 14277 + }, + { + "epoch": 0.7514736842105263, + "grad_norm": 0.50390625, + "learning_rate": 5.5228070175438594e-05, + "loss": 3.3188, + "step": 14278 + }, + { + "epoch": 0.7515263157894737, + "grad_norm": 0.50390625, + "learning_rate": 5.521637426900586e-05, + "loss": 3.0862, + "step": 14279 + }, + { + "epoch": 0.751578947368421, + "grad_norm": 0.48828125, + "learning_rate": 5.52046783625731e-05, + "loss": 2.6602, + "step": 14280 + }, + { + "epoch": 0.7516315789473684, + "grad_norm": 0.490234375, + "learning_rate": 5.519298245614035e-05, + "loss": 3.4289, + "step": 14281 + }, + { + "epoch": 0.7516842105263158, + "grad_norm": 0.462890625, + "learning_rate": 5.51812865497076e-05, + "loss": 3.3013, + "step": 14282 + }, + { + "epoch": 0.7517368421052631, + "grad_norm": 0.50390625, + "learning_rate": 5.516959064327486e-05, + "loss": 3.2655, + "step": 14283 + }, + { + "epoch": 0.7517894736842106, + "grad_norm": 0.44921875, + "learning_rate": 5.5157894736842105e-05, + "loss": 3.4006, + "step": 14284 + }, + { + "epoch": 0.7518421052631579, + "grad_norm": 0.51171875, + "learning_rate": 5.5146198830409355e-05, + "loss": 3.2214, + "step": 14285 + }, + { + "epoch": 0.7518947368421053, + "grad_norm": 0.490234375, + "learning_rate": 5.5134502923976605e-05, + "loss": 3.22, + "step": 14286 + }, + { + "epoch": 0.7519473684210526, + "grad_norm": 0.4765625, + "learning_rate": 5.512280701754387e-05, + "loss": 3.4796, + "step": 14287 + }, + { + "epoch": 0.752, + "grad_norm": 0.490234375, + "learning_rate": 5.511111111111111e-05, + "loss": 3.1988, + "step": 14288 + }, + { + "epoch": 0.7520526315789474, + "grad_norm": 0.51953125, + "learning_rate": 5.509941520467836e-05, + "loss": 3.2063, + "step": 14289 + }, + { + "epoch": 0.7521052631578947, + "grad_norm": 0.50390625, + "learning_rate": 5.508771929824562e-05, + "loss": 3.3712, + "step": 14290 + }, + { + "epoch": 0.7521578947368421, + "grad_norm": 0.474609375, + "learning_rate": 5.507602339181287e-05, + "loss": 3.3487, + "step": 14291 + }, + { + "epoch": 0.7522105263157894, + "grad_norm": 0.5, + "learning_rate": 5.5064327485380116e-05, + "loss": 3.4761, + "step": 14292 + }, + { + "epoch": 0.7522631578947369, + "grad_norm": 0.55078125, + "learning_rate": 5.5052631578947365e-05, + "loss": 2.7845, + "step": 14293 + }, + { + "epoch": 0.7523157894736842, + "grad_norm": 0.490234375, + "learning_rate": 5.504093567251463e-05, + "loss": 3.3389, + "step": 14294 + }, + { + "epoch": 0.7523684210526316, + "grad_norm": 0.51171875, + "learning_rate": 5.502923976608188e-05, + "loss": 2.8635, + "step": 14295 + }, + { + "epoch": 0.752421052631579, + "grad_norm": 0.51171875, + "learning_rate": 5.501754385964912e-05, + "loss": 3.0968, + "step": 14296 + }, + { + "epoch": 0.7524736842105263, + "grad_norm": 0.52734375, + "learning_rate": 5.500584795321637e-05, + "loss": 3.0249, + "step": 14297 + }, + { + "epoch": 0.7525263157894737, + "grad_norm": 0.4921875, + "learning_rate": 5.4994152046783634e-05, + "loss": 2.9789, + "step": 14298 + }, + { + "epoch": 0.752578947368421, + "grad_norm": 0.4765625, + "learning_rate": 5.498245614035088e-05, + "loss": 3.0738, + "step": 14299 + }, + { + "epoch": 0.7526315789473684, + "grad_norm": 0.5, + "learning_rate": 5.4970760233918126e-05, + "loss": 3.1838, + "step": 14300 + }, + { + "epoch": 0.7526842105263158, + "grad_norm": 0.466796875, + "learning_rate": 5.495906432748539e-05, + "loss": 3.2711, + "step": 14301 + }, + { + "epoch": 0.7527368421052631, + "grad_norm": 0.515625, + "learning_rate": 5.494736842105264e-05, + "loss": 3.2664, + "step": 14302 + }, + { + "epoch": 0.7527894736842106, + "grad_norm": 0.494140625, + "learning_rate": 5.493567251461988e-05, + "loss": 3.1795, + "step": 14303 + }, + { + "epoch": 0.7528421052631579, + "grad_norm": 0.498046875, + "learning_rate": 5.492397660818713e-05, + "loss": 3.4267, + "step": 14304 + }, + { + "epoch": 0.7528947368421053, + "grad_norm": 0.5078125, + "learning_rate": 5.4912280701754394e-05, + "loss": 3.1346, + "step": 14305 + }, + { + "epoch": 0.7529473684210526, + "grad_norm": 0.48828125, + "learning_rate": 5.4900584795321644e-05, + "loss": 3.1486, + "step": 14306 + }, + { + "epoch": 0.753, + "grad_norm": 0.51953125, + "learning_rate": 5.488888888888889e-05, + "loss": 2.8854, + "step": 14307 + }, + { + "epoch": 0.7530526315789474, + "grad_norm": 0.474609375, + "learning_rate": 5.4877192982456136e-05, + "loss": 3.5192, + "step": 14308 + }, + { + "epoch": 0.7531052631578947, + "grad_norm": 0.5, + "learning_rate": 5.48654970760234e-05, + "loss": 3.279, + "step": 14309 + }, + { + "epoch": 0.7531578947368421, + "grad_norm": 0.53515625, + "learning_rate": 5.485380116959065e-05, + "loss": 3.215, + "step": 14310 + }, + { + "epoch": 0.7532105263157894, + "grad_norm": 0.46875, + "learning_rate": 5.484210526315789e-05, + "loss": 3.2241, + "step": 14311 + }, + { + "epoch": 0.7532631578947369, + "grad_norm": 0.5, + "learning_rate": 5.4830409356725155e-05, + "loss": 3.7035, + "step": 14312 + }, + { + "epoch": 0.7533157894736842, + "grad_norm": 0.490234375, + "learning_rate": 5.4818713450292405e-05, + "loss": 3.3578, + "step": 14313 + }, + { + "epoch": 0.7533684210526316, + "grad_norm": 0.484375, + "learning_rate": 5.4807017543859654e-05, + "loss": 3.2519, + "step": 14314 + }, + { + "epoch": 0.753421052631579, + "grad_norm": 0.486328125, + "learning_rate": 5.47953216374269e-05, + "loss": 3.239, + "step": 14315 + }, + { + "epoch": 0.7534736842105263, + "grad_norm": 0.515625, + "learning_rate": 5.478362573099416e-05, + "loss": 3.3065, + "step": 14316 + }, + { + "epoch": 0.7535263157894737, + "grad_norm": 0.54296875, + "learning_rate": 5.477192982456141e-05, + "loss": 3.2768, + "step": 14317 + }, + { + "epoch": 0.753578947368421, + "grad_norm": 0.5390625, + "learning_rate": 5.476023391812866e-05, + "loss": 3.4087, + "step": 14318 + }, + { + "epoch": 0.7536315789473684, + "grad_norm": 0.490234375, + "learning_rate": 5.47485380116959e-05, + "loss": 3.3396, + "step": 14319 + }, + { + "epoch": 0.7536842105263157, + "grad_norm": 0.46875, + "learning_rate": 5.4736842105263165e-05, + "loss": 3.0164, + "step": 14320 + }, + { + "epoch": 0.7537368421052631, + "grad_norm": 0.490234375, + "learning_rate": 5.4725146198830415e-05, + "loss": 3.2795, + "step": 14321 + }, + { + "epoch": 0.7537894736842106, + "grad_norm": 0.47265625, + "learning_rate": 5.4713450292397664e-05, + "loss": 3.5149, + "step": 14322 + }, + { + "epoch": 0.7538421052631579, + "grad_norm": 0.48046875, + "learning_rate": 5.470175438596491e-05, + "loss": 2.8313, + "step": 14323 + }, + { + "epoch": 0.7538947368421053, + "grad_norm": 0.482421875, + "learning_rate": 5.469005847953217e-05, + "loss": 3.3248, + "step": 14324 + }, + { + "epoch": 0.7539473684210526, + "grad_norm": 0.490234375, + "learning_rate": 5.467836257309942e-05, + "loss": 3.188, + "step": 14325 + }, + { + "epoch": 0.754, + "grad_norm": 0.5, + "learning_rate": 5.466666666666666e-05, + "loss": 3.5065, + "step": 14326 + }, + { + "epoch": 0.7540526315789474, + "grad_norm": 0.494140625, + "learning_rate": 5.4654970760233926e-05, + "loss": 2.9653, + "step": 14327 + }, + { + "epoch": 0.7541052631578947, + "grad_norm": 0.5390625, + "learning_rate": 5.4643274853801176e-05, + "loss": 3.0507, + "step": 14328 + }, + { + "epoch": 0.7541578947368421, + "grad_norm": 0.486328125, + "learning_rate": 5.4631578947368425e-05, + "loss": 3.6373, + "step": 14329 + }, + { + "epoch": 0.7542105263157894, + "grad_norm": 0.482421875, + "learning_rate": 5.461988304093567e-05, + "loss": 2.7529, + "step": 14330 + }, + { + "epoch": 0.7542631578947369, + "grad_norm": 0.4921875, + "learning_rate": 5.460818713450293e-05, + "loss": 3.0251, + "step": 14331 + }, + { + "epoch": 0.7543157894736842, + "grad_norm": 0.466796875, + "learning_rate": 5.459649122807018e-05, + "loss": 3.0242, + "step": 14332 + }, + { + "epoch": 0.7543684210526316, + "grad_norm": 0.49609375, + "learning_rate": 5.458479532163743e-05, + "loss": 2.9543, + "step": 14333 + }, + { + "epoch": 0.754421052631579, + "grad_norm": 0.5703125, + "learning_rate": 5.457309941520467e-05, + "loss": 3.5167, + "step": 14334 + }, + { + "epoch": 0.7544736842105263, + "grad_norm": 0.494140625, + "learning_rate": 5.4561403508771936e-05, + "loss": 2.7576, + "step": 14335 + }, + { + "epoch": 0.7545263157894737, + "grad_norm": 0.46875, + "learning_rate": 5.4549707602339186e-05, + "loss": 3.1522, + "step": 14336 + }, + { + "epoch": 0.754578947368421, + "grad_norm": 0.5234375, + "learning_rate": 5.4538011695906435e-05, + "loss": 3.5534, + "step": 14337 + }, + { + "epoch": 0.7546315789473684, + "grad_norm": 0.51171875, + "learning_rate": 5.452631578947369e-05, + "loss": 3.4536, + "step": 14338 + }, + { + "epoch": 0.7546842105263157, + "grad_norm": 0.50390625, + "learning_rate": 5.451461988304094e-05, + "loss": 3.454, + "step": 14339 + }, + { + "epoch": 0.7547368421052632, + "grad_norm": 0.48046875, + "learning_rate": 5.450292397660819e-05, + "loss": 3.2081, + "step": 14340 + }, + { + "epoch": 0.7547894736842106, + "grad_norm": 0.474609375, + "learning_rate": 5.449122807017544e-05, + "loss": 3.4474, + "step": 14341 + }, + { + "epoch": 0.7548421052631579, + "grad_norm": 0.50390625, + "learning_rate": 5.44795321637427e-05, + "loss": 3.3099, + "step": 14342 + }, + { + "epoch": 0.7548947368421053, + "grad_norm": 0.470703125, + "learning_rate": 5.4467836257309947e-05, + "loss": 3.5764, + "step": 14343 + }, + { + "epoch": 0.7549473684210526, + "grad_norm": 0.5078125, + "learning_rate": 5.4456140350877196e-05, + "loss": 3.2644, + "step": 14344 + }, + { + "epoch": 0.755, + "grad_norm": 0.5, + "learning_rate": 5.4444444444444446e-05, + "loss": 3.5501, + "step": 14345 + }, + { + "epoch": 0.7550526315789474, + "grad_norm": 0.5078125, + "learning_rate": 5.44327485380117e-05, + "loss": 3.5024, + "step": 14346 + }, + { + "epoch": 0.7551052631578947, + "grad_norm": 0.5234375, + "learning_rate": 5.442105263157895e-05, + "loss": 3.4927, + "step": 14347 + }, + { + "epoch": 0.7551578947368421, + "grad_norm": 0.48828125, + "learning_rate": 5.44093567251462e-05, + "loss": 3.3888, + "step": 14348 + }, + { + "epoch": 0.7552105263157894, + "grad_norm": 0.49609375, + "learning_rate": 5.439766081871346e-05, + "loss": 3.3524, + "step": 14349 + }, + { + "epoch": 0.7552631578947369, + "grad_norm": 0.46875, + "learning_rate": 5.438596491228071e-05, + "loss": 3.2214, + "step": 14350 + }, + { + "epoch": 0.7553157894736842, + "grad_norm": 0.4921875, + "learning_rate": 5.437426900584796e-05, + "loss": 3.2612, + "step": 14351 + }, + { + "epoch": 0.7553684210526316, + "grad_norm": 0.50390625, + "learning_rate": 5.4362573099415206e-05, + "loss": 3.2667, + "step": 14352 + }, + { + "epoch": 0.755421052631579, + "grad_norm": 0.490234375, + "learning_rate": 5.435087719298246e-05, + "loss": 3.0373, + "step": 14353 + }, + { + "epoch": 0.7554736842105263, + "grad_norm": 0.46875, + "learning_rate": 5.433918128654971e-05, + "loss": 3.075, + "step": 14354 + }, + { + "epoch": 0.7555263157894737, + "grad_norm": 0.46875, + "learning_rate": 5.432748538011696e-05, + "loss": 3.4683, + "step": 14355 + }, + { + "epoch": 0.755578947368421, + "grad_norm": 0.486328125, + "learning_rate": 5.431578947368421e-05, + "loss": 3.5451, + "step": 14356 + }, + { + "epoch": 0.7556315789473684, + "grad_norm": 0.48828125, + "learning_rate": 5.430409356725147e-05, + "loss": 3.0191, + "step": 14357 + }, + { + "epoch": 0.7556842105263157, + "grad_norm": 0.474609375, + "learning_rate": 5.429239766081872e-05, + "loss": 3.0451, + "step": 14358 + }, + { + "epoch": 0.7557368421052632, + "grad_norm": 0.51171875, + "learning_rate": 5.428070175438597e-05, + "loss": 3.5129, + "step": 14359 + }, + { + "epoch": 0.7557894736842106, + "grad_norm": 0.53515625, + "learning_rate": 5.426900584795322e-05, + "loss": 3.5928, + "step": 14360 + }, + { + "epoch": 0.7558421052631579, + "grad_norm": 0.494140625, + "learning_rate": 5.425730994152047e-05, + "loss": 3.0932, + "step": 14361 + }, + { + "epoch": 0.7558947368421053, + "grad_norm": 1.328125, + "learning_rate": 5.424561403508772e-05, + "loss": 3.002, + "step": 14362 + }, + { + "epoch": 0.7559473684210526, + "grad_norm": 0.52734375, + "learning_rate": 5.423391812865497e-05, + "loss": 3.1786, + "step": 14363 + }, + { + "epoch": 0.756, + "grad_norm": 0.484375, + "learning_rate": 5.422222222222223e-05, + "loss": 3.4519, + "step": 14364 + }, + { + "epoch": 0.7560526315789474, + "grad_norm": 0.515625, + "learning_rate": 5.421052631578948e-05, + "loss": 3.1561, + "step": 14365 + }, + { + "epoch": 0.7561052631578947, + "grad_norm": 0.51171875, + "learning_rate": 5.419883040935673e-05, + "loss": 3.4671, + "step": 14366 + }, + { + "epoch": 0.7561578947368421, + "grad_norm": 0.50390625, + "learning_rate": 5.418713450292398e-05, + "loss": 3.2805, + "step": 14367 + }, + { + "epoch": 0.7562105263157894, + "grad_norm": 0.546875, + "learning_rate": 5.4175438596491234e-05, + "loss": 3.5529, + "step": 14368 + }, + { + "epoch": 0.7562631578947369, + "grad_norm": 0.47265625, + "learning_rate": 5.416374269005848e-05, + "loss": 3.3563, + "step": 14369 + }, + { + "epoch": 0.7563157894736842, + "grad_norm": 0.4765625, + "learning_rate": 5.415204678362573e-05, + "loss": 3.131, + "step": 14370 + }, + { + "epoch": 0.7563684210526316, + "grad_norm": 0.482421875, + "learning_rate": 5.414035087719298e-05, + "loss": 3.5312, + "step": 14371 + }, + { + "epoch": 0.756421052631579, + "grad_norm": 0.5390625, + "learning_rate": 5.412865497076024e-05, + "loss": 3.0745, + "step": 14372 + }, + { + "epoch": 0.7564736842105263, + "grad_norm": 0.498046875, + "learning_rate": 5.411695906432749e-05, + "loss": 3.2376, + "step": 14373 + }, + { + "epoch": 0.7565263157894737, + "grad_norm": 0.57421875, + "learning_rate": 5.410526315789474e-05, + "loss": 3.2016, + "step": 14374 + }, + { + "epoch": 0.756578947368421, + "grad_norm": 0.484375, + "learning_rate": 5.4093567251461994e-05, + "loss": 3.2967, + "step": 14375 + }, + { + "epoch": 0.7566315789473684, + "grad_norm": 0.4921875, + "learning_rate": 5.4081871345029244e-05, + "loss": 3.4707, + "step": 14376 + }, + { + "epoch": 0.7566842105263157, + "grad_norm": 0.5078125, + "learning_rate": 5.4070175438596494e-05, + "loss": 3.4147, + "step": 14377 + }, + { + "epoch": 0.7567368421052632, + "grad_norm": 0.49609375, + "learning_rate": 5.405847953216374e-05, + "loss": 3.3636, + "step": 14378 + }, + { + "epoch": 0.7567894736842106, + "grad_norm": 0.474609375, + "learning_rate": 5.4046783625731e-05, + "loss": 3.5411, + "step": 14379 + }, + { + "epoch": 0.7568421052631579, + "grad_norm": 0.5078125, + "learning_rate": 5.403508771929825e-05, + "loss": 2.9674, + "step": 14380 + }, + { + "epoch": 0.7568947368421053, + "grad_norm": 0.734375, + "learning_rate": 5.40233918128655e-05, + "loss": 3.5246, + "step": 14381 + }, + { + "epoch": 0.7569473684210526, + "grad_norm": 0.48046875, + "learning_rate": 5.401169590643275e-05, + "loss": 3.4263, + "step": 14382 + }, + { + "epoch": 0.757, + "grad_norm": 0.5078125, + "learning_rate": 5.4000000000000005e-05, + "loss": 3.3342, + "step": 14383 + }, + { + "epoch": 0.7570526315789474, + "grad_norm": 0.4765625, + "learning_rate": 5.3988304093567254e-05, + "loss": 3.2508, + "step": 14384 + }, + { + "epoch": 0.7571052631578947, + "grad_norm": 0.51171875, + "learning_rate": 5.3976608187134504e-05, + "loss": 3.1817, + "step": 14385 + }, + { + "epoch": 0.7571578947368421, + "grad_norm": 0.474609375, + "learning_rate": 5.3964912280701754e-05, + "loss": 3.4004, + "step": 14386 + }, + { + "epoch": 0.7572105263157894, + "grad_norm": 0.47265625, + "learning_rate": 5.395321637426901e-05, + "loss": 3.3435, + "step": 14387 + }, + { + "epoch": 0.7572631578947369, + "grad_norm": 0.486328125, + "learning_rate": 5.394152046783626e-05, + "loss": 2.9832, + "step": 14388 + }, + { + "epoch": 0.7573157894736842, + "grad_norm": 0.5234375, + "learning_rate": 5.392982456140351e-05, + "loss": 3.1807, + "step": 14389 + }, + { + "epoch": 0.7573684210526316, + "grad_norm": 0.5, + "learning_rate": 5.3918128654970765e-05, + "loss": 3.5215, + "step": 14390 + }, + { + "epoch": 0.757421052631579, + "grad_norm": 0.5, + "learning_rate": 5.3906432748538015e-05, + "loss": 3.555, + "step": 14391 + }, + { + "epoch": 0.7574736842105263, + "grad_norm": 0.515625, + "learning_rate": 5.3894736842105265e-05, + "loss": 3.3934, + "step": 14392 + }, + { + "epoch": 0.7575263157894737, + "grad_norm": 0.5078125, + "learning_rate": 5.3883040935672514e-05, + "loss": 3.2302, + "step": 14393 + }, + { + "epoch": 0.757578947368421, + "grad_norm": 0.46875, + "learning_rate": 5.387134502923977e-05, + "loss": 3.2041, + "step": 14394 + }, + { + "epoch": 0.7576315789473684, + "grad_norm": 0.474609375, + "learning_rate": 5.385964912280702e-05, + "loss": 3.0303, + "step": 14395 + }, + { + "epoch": 0.7576842105263157, + "grad_norm": 0.474609375, + "learning_rate": 5.384795321637427e-05, + "loss": 3.2236, + "step": 14396 + }, + { + "epoch": 0.7577368421052632, + "grad_norm": 0.494140625, + "learning_rate": 5.383625730994152e-05, + "loss": 3.4481, + "step": 14397 + }, + { + "epoch": 0.7577894736842106, + "grad_norm": 0.5078125, + "learning_rate": 5.3824561403508776e-05, + "loss": 3.24, + "step": 14398 + }, + { + "epoch": 0.7578421052631579, + "grad_norm": 0.4765625, + "learning_rate": 5.3812865497076025e-05, + "loss": 3.3924, + "step": 14399 + }, + { + "epoch": 0.7578947368421053, + "grad_norm": 0.4765625, + "learning_rate": 5.3801169590643275e-05, + "loss": 3.402, + "step": 14400 + }, + { + "epoch": 0.7579473684210526, + "grad_norm": 0.484375, + "learning_rate": 5.378947368421053e-05, + "loss": 3.0695, + "step": 14401 + }, + { + "epoch": 0.758, + "grad_norm": 0.498046875, + "learning_rate": 5.377777777777778e-05, + "loss": 2.9849, + "step": 14402 + }, + { + "epoch": 0.7580526315789474, + "grad_norm": 0.478515625, + "learning_rate": 5.376608187134503e-05, + "loss": 3.7291, + "step": 14403 + }, + { + "epoch": 0.7581052631578947, + "grad_norm": 0.515625, + "learning_rate": 5.375438596491228e-05, + "loss": 3.1284, + "step": 14404 + }, + { + "epoch": 0.7581578947368421, + "grad_norm": 0.5, + "learning_rate": 5.3742690058479536e-05, + "loss": 3.1262, + "step": 14405 + }, + { + "epoch": 0.7582105263157894, + "grad_norm": 0.50390625, + "learning_rate": 5.3730994152046786e-05, + "loss": 3.2976, + "step": 14406 + }, + { + "epoch": 0.7582631578947369, + "grad_norm": 0.5234375, + "learning_rate": 5.3719298245614036e-05, + "loss": 3.1199, + "step": 14407 + }, + { + "epoch": 0.7583157894736842, + "grad_norm": 0.49609375, + "learning_rate": 5.3707602339181285e-05, + "loss": 3.3775, + "step": 14408 + }, + { + "epoch": 0.7583684210526316, + "grad_norm": 0.51171875, + "learning_rate": 5.369590643274854e-05, + "loss": 3.4502, + "step": 14409 + }, + { + "epoch": 0.758421052631579, + "grad_norm": 0.48046875, + "learning_rate": 5.368421052631579e-05, + "loss": 3.3741, + "step": 14410 + }, + { + "epoch": 0.7584736842105263, + "grad_norm": 0.490234375, + "learning_rate": 5.367251461988304e-05, + "loss": 3.4013, + "step": 14411 + }, + { + "epoch": 0.7585263157894737, + "grad_norm": 0.482421875, + "learning_rate": 5.36608187134503e-05, + "loss": 3.2759, + "step": 14412 + }, + { + "epoch": 0.758578947368421, + "grad_norm": 0.50390625, + "learning_rate": 5.364912280701755e-05, + "loss": 3.4869, + "step": 14413 + }, + { + "epoch": 0.7586315789473684, + "grad_norm": 0.498046875, + "learning_rate": 5.3637426900584796e-05, + "loss": 3.1832, + "step": 14414 + }, + { + "epoch": 0.7586842105263157, + "grad_norm": 0.5703125, + "learning_rate": 5.3625730994152046e-05, + "loss": 3.2836, + "step": 14415 + }, + { + "epoch": 0.7587368421052632, + "grad_norm": 0.5, + "learning_rate": 5.36140350877193e-05, + "loss": 3.2962, + "step": 14416 + }, + { + "epoch": 0.7587894736842106, + "grad_norm": 0.53125, + "learning_rate": 5.360233918128655e-05, + "loss": 3.2798, + "step": 14417 + }, + { + "epoch": 0.7588421052631579, + "grad_norm": 0.49609375, + "learning_rate": 5.35906432748538e-05, + "loss": 3.2981, + "step": 14418 + }, + { + "epoch": 0.7588947368421053, + "grad_norm": 0.578125, + "learning_rate": 5.357894736842105e-05, + "loss": 2.9471, + "step": 14419 + }, + { + "epoch": 0.7589473684210526, + "grad_norm": 0.490234375, + "learning_rate": 5.356725146198831e-05, + "loss": 3.1712, + "step": 14420 + }, + { + "epoch": 0.759, + "grad_norm": 0.494140625, + "learning_rate": 5.355555555555556e-05, + "loss": 3.319, + "step": 14421 + }, + { + "epoch": 0.7590526315789474, + "grad_norm": 0.486328125, + "learning_rate": 5.354385964912281e-05, + "loss": 3.4099, + "step": 14422 + }, + { + "epoch": 0.7591052631578947, + "grad_norm": 0.490234375, + "learning_rate": 5.3532163742690056e-05, + "loss": 2.9053, + "step": 14423 + }, + { + "epoch": 0.7591578947368421, + "grad_norm": 0.498046875, + "learning_rate": 5.352046783625731e-05, + "loss": 3.4138, + "step": 14424 + }, + { + "epoch": 0.7592105263157894, + "grad_norm": 0.478515625, + "learning_rate": 5.350877192982456e-05, + "loss": 3.6702, + "step": 14425 + }, + { + "epoch": 0.7592631578947369, + "grad_norm": 0.52734375, + "learning_rate": 5.349707602339181e-05, + "loss": 3.2644, + "step": 14426 + }, + { + "epoch": 0.7593157894736842, + "grad_norm": 0.515625, + "learning_rate": 5.3485380116959075e-05, + "loss": 3.2142, + "step": 14427 + }, + { + "epoch": 0.7593684210526316, + "grad_norm": 0.494140625, + "learning_rate": 5.347368421052632e-05, + "loss": 3.2266, + "step": 14428 + }, + { + "epoch": 0.759421052631579, + "grad_norm": 0.5234375, + "learning_rate": 5.346198830409357e-05, + "loss": 3.7555, + "step": 14429 + }, + { + "epoch": 0.7594736842105263, + "grad_norm": 0.72265625, + "learning_rate": 5.345029239766082e-05, + "loss": 3.2849, + "step": 14430 + }, + { + "epoch": 0.7595263157894737, + "grad_norm": 0.546875, + "learning_rate": 5.343859649122808e-05, + "loss": 3.4348, + "step": 14431 + }, + { + "epoch": 0.759578947368421, + "grad_norm": 0.51953125, + "learning_rate": 5.342690058479532e-05, + "loss": 3.5272, + "step": 14432 + }, + { + "epoch": 0.7596315789473684, + "grad_norm": 0.53515625, + "learning_rate": 5.341520467836257e-05, + "loss": 3.468, + "step": 14433 + }, + { + "epoch": 0.7596842105263157, + "grad_norm": 0.5234375, + "learning_rate": 5.340350877192982e-05, + "loss": 3.6728, + "step": 14434 + }, + { + "epoch": 0.7597368421052632, + "grad_norm": 0.5078125, + "learning_rate": 5.339181286549708e-05, + "loss": 2.8903, + "step": 14435 + }, + { + "epoch": 0.7597894736842106, + "grad_norm": 0.7734375, + "learning_rate": 5.338011695906433e-05, + "loss": 3.076, + "step": 14436 + }, + { + "epoch": 0.7598421052631579, + "grad_norm": 0.515625, + "learning_rate": 5.336842105263158e-05, + "loss": 3.0872, + "step": 14437 + }, + { + "epoch": 0.7598947368421053, + "grad_norm": 0.458984375, + "learning_rate": 5.335672514619884e-05, + "loss": 3.2898, + "step": 14438 + }, + { + "epoch": 0.7599473684210526, + "grad_norm": 0.484375, + "learning_rate": 5.3345029239766084e-05, + "loss": 3.4724, + "step": 14439 + }, + { + "epoch": 0.76, + "grad_norm": 0.5078125, + "learning_rate": 5.333333333333333e-05, + "loss": 3.1063, + "step": 14440 + }, + { + "epoch": 0.7600526315789473, + "grad_norm": 0.486328125, + "learning_rate": 5.332163742690058e-05, + "loss": 3.1156, + "step": 14441 + }, + { + "epoch": 0.7601052631578947, + "grad_norm": 0.484375, + "learning_rate": 5.3309941520467846e-05, + "loss": 2.6666, + "step": 14442 + }, + { + "epoch": 0.7601578947368421, + "grad_norm": 0.50390625, + "learning_rate": 5.329824561403509e-05, + "loss": 3.4626, + "step": 14443 + }, + { + "epoch": 0.7602105263157894, + "grad_norm": 0.4765625, + "learning_rate": 5.328654970760234e-05, + "loss": 3.4124, + "step": 14444 + }, + { + "epoch": 0.7602631578947369, + "grad_norm": 0.53125, + "learning_rate": 5.327485380116959e-05, + "loss": 3.1757, + "step": 14445 + }, + { + "epoch": 0.7603157894736842, + "grad_norm": 0.494140625, + "learning_rate": 5.326315789473685e-05, + "loss": 3.4032, + "step": 14446 + }, + { + "epoch": 0.7603684210526316, + "grad_norm": 0.5078125, + "learning_rate": 5.3251461988304094e-05, + "loss": 3.0896, + "step": 14447 + }, + { + "epoch": 0.760421052631579, + "grad_norm": 0.5234375, + "learning_rate": 5.3239766081871343e-05, + "loss": 3.4525, + "step": 14448 + }, + { + "epoch": 0.7604736842105263, + "grad_norm": 0.50390625, + "learning_rate": 5.322807017543859e-05, + "loss": 3.1731, + "step": 14449 + }, + { + "epoch": 0.7605263157894737, + "grad_norm": 0.5, + "learning_rate": 5.3216374269005856e-05, + "loss": 3.1539, + "step": 14450 + }, + { + "epoch": 0.760578947368421, + "grad_norm": 0.46484375, + "learning_rate": 5.32046783625731e-05, + "loss": 3.1895, + "step": 14451 + }, + { + "epoch": 0.7606315789473684, + "grad_norm": 0.81640625, + "learning_rate": 5.319298245614035e-05, + "loss": 3.171, + "step": 14452 + }, + { + "epoch": 0.7606842105263157, + "grad_norm": 0.5, + "learning_rate": 5.318128654970761e-05, + "loss": 3.0821, + "step": 14453 + }, + { + "epoch": 0.7607368421052632, + "grad_norm": 0.50390625, + "learning_rate": 5.316959064327486e-05, + "loss": 3.0676, + "step": 14454 + }, + { + "epoch": 0.7607894736842106, + "grad_norm": 0.482421875, + "learning_rate": 5.3157894736842104e-05, + "loss": 3.791, + "step": 14455 + }, + { + "epoch": 0.7608421052631579, + "grad_norm": 0.498046875, + "learning_rate": 5.3146198830409354e-05, + "loss": 3.1863, + "step": 14456 + }, + { + "epoch": 0.7608947368421053, + "grad_norm": 0.48046875, + "learning_rate": 5.313450292397662e-05, + "loss": 2.8098, + "step": 14457 + }, + { + "epoch": 0.7609473684210526, + "grad_norm": 0.5078125, + "learning_rate": 5.312280701754386e-05, + "loss": 3.3142, + "step": 14458 + }, + { + "epoch": 0.761, + "grad_norm": 0.498046875, + "learning_rate": 5.311111111111111e-05, + "loss": 3.415, + "step": 14459 + }, + { + "epoch": 0.7610526315789473, + "grad_norm": 0.55859375, + "learning_rate": 5.309941520467836e-05, + "loss": 3.4135, + "step": 14460 + }, + { + "epoch": 0.7611052631578947, + "grad_norm": 0.51171875, + "learning_rate": 5.308771929824562e-05, + "loss": 3.2971, + "step": 14461 + }, + { + "epoch": 0.7611578947368421, + "grad_norm": 0.515625, + "learning_rate": 5.3076023391812865e-05, + "loss": 3.4511, + "step": 14462 + }, + { + "epoch": 0.7612105263157894, + "grad_norm": 0.56640625, + "learning_rate": 5.3064327485380114e-05, + "loss": 3.3415, + "step": 14463 + }, + { + "epoch": 0.7612631578947369, + "grad_norm": 0.486328125, + "learning_rate": 5.305263157894738e-05, + "loss": 3.2558, + "step": 14464 + }, + { + "epoch": 0.7613157894736842, + "grad_norm": 0.48046875, + "learning_rate": 5.304093567251463e-05, + "loss": 3.0005, + "step": 14465 + }, + { + "epoch": 0.7613684210526316, + "grad_norm": 0.490234375, + "learning_rate": 5.302923976608187e-05, + "loss": 3.0252, + "step": 14466 + }, + { + "epoch": 0.761421052631579, + "grad_norm": 0.4921875, + "learning_rate": 5.301754385964912e-05, + "loss": 3.4394, + "step": 14467 + }, + { + "epoch": 0.7614736842105263, + "grad_norm": 0.52734375, + "learning_rate": 5.300584795321638e-05, + "loss": 3.3573, + "step": 14468 + }, + { + "epoch": 0.7615263157894737, + "grad_norm": 0.48046875, + "learning_rate": 5.299415204678363e-05, + "loss": 3.357, + "step": 14469 + }, + { + "epoch": 0.761578947368421, + "grad_norm": 0.50390625, + "learning_rate": 5.2982456140350875e-05, + "loss": 3.6568, + "step": 14470 + }, + { + "epoch": 0.7616315789473684, + "grad_norm": 0.51171875, + "learning_rate": 5.2970760233918125e-05, + "loss": 3.2579, + "step": 14471 + }, + { + "epoch": 0.7616842105263157, + "grad_norm": 0.49609375, + "learning_rate": 5.295906432748539e-05, + "loss": 3.4731, + "step": 14472 + }, + { + "epoch": 0.7617368421052632, + "grad_norm": 0.494140625, + "learning_rate": 5.294736842105264e-05, + "loss": 2.9577, + "step": 14473 + }, + { + "epoch": 0.7617894736842106, + "grad_norm": 0.52734375, + "learning_rate": 5.293567251461988e-05, + "loss": 2.9856, + "step": 14474 + }, + { + "epoch": 0.7618421052631579, + "grad_norm": 0.486328125, + "learning_rate": 5.2923976608187143e-05, + "loss": 3.6109, + "step": 14475 + }, + { + "epoch": 0.7618947368421053, + "grad_norm": 0.455078125, + "learning_rate": 5.291228070175439e-05, + "loss": 3.4701, + "step": 14476 + }, + { + "epoch": 0.7619473684210526, + "grad_norm": 0.53125, + "learning_rate": 5.290058479532164e-05, + "loss": 3.125, + "step": 14477 + }, + { + "epoch": 0.762, + "grad_norm": 0.4765625, + "learning_rate": 5.2888888888888885e-05, + "loss": 3.3806, + "step": 14478 + }, + { + "epoch": 0.7620526315789473, + "grad_norm": 0.49609375, + "learning_rate": 5.287719298245615e-05, + "loss": 3.3273, + "step": 14479 + }, + { + "epoch": 0.7621052631578947, + "grad_norm": 0.462890625, + "learning_rate": 5.28654970760234e-05, + "loss": 3.2018, + "step": 14480 + }, + { + "epoch": 0.7621578947368421, + "grad_norm": 0.484375, + "learning_rate": 5.285380116959064e-05, + "loss": 3.3736, + "step": 14481 + }, + { + "epoch": 0.7622105263157894, + "grad_norm": 0.478515625, + "learning_rate": 5.284210526315789e-05, + "loss": 3.2243, + "step": 14482 + }, + { + "epoch": 0.7622631578947369, + "grad_norm": 0.50390625, + "learning_rate": 5.2830409356725154e-05, + "loss": 3.156, + "step": 14483 + }, + { + "epoch": 0.7623157894736842, + "grad_norm": 0.55859375, + "learning_rate": 5.28187134502924e-05, + "loss": 3.4341, + "step": 14484 + }, + { + "epoch": 0.7623684210526316, + "grad_norm": 0.5078125, + "learning_rate": 5.2807017543859646e-05, + "loss": 3.5466, + "step": 14485 + }, + { + "epoch": 0.762421052631579, + "grad_norm": 0.494140625, + "learning_rate": 5.2795321637426896e-05, + "loss": 3.2039, + "step": 14486 + }, + { + "epoch": 0.7624736842105263, + "grad_norm": 0.5, + "learning_rate": 5.278362573099416e-05, + "loss": 3.2936, + "step": 14487 + }, + { + "epoch": 0.7625263157894737, + "grad_norm": 0.498046875, + "learning_rate": 5.277192982456141e-05, + "loss": 3.2889, + "step": 14488 + }, + { + "epoch": 0.762578947368421, + "grad_norm": 0.53125, + "learning_rate": 5.276023391812865e-05, + "loss": 3.3714, + "step": 14489 + }, + { + "epoch": 0.7626315789473684, + "grad_norm": 0.49609375, + "learning_rate": 5.2748538011695914e-05, + "loss": 3.3291, + "step": 14490 + }, + { + "epoch": 0.7626842105263157, + "grad_norm": 0.4921875, + "learning_rate": 5.2736842105263164e-05, + "loss": 3.3548, + "step": 14491 + }, + { + "epoch": 0.7627368421052632, + "grad_norm": 0.52734375, + "learning_rate": 5.2725146198830414e-05, + "loss": 3.2463, + "step": 14492 + }, + { + "epoch": 0.7627894736842106, + "grad_norm": 0.50390625, + "learning_rate": 5.2713450292397656e-05, + "loss": 3.0593, + "step": 14493 + }, + { + "epoch": 0.7628421052631579, + "grad_norm": 0.484375, + "learning_rate": 5.270175438596492e-05, + "loss": 3.7175, + "step": 14494 + }, + { + "epoch": 0.7628947368421053, + "grad_norm": 0.53515625, + "learning_rate": 5.269005847953217e-05, + "loss": 3.149, + "step": 14495 + }, + { + "epoch": 0.7629473684210526, + "grad_norm": 0.490234375, + "learning_rate": 5.267836257309942e-05, + "loss": 3.4432, + "step": 14496 + }, + { + "epoch": 0.763, + "grad_norm": 0.50390625, + "learning_rate": 5.266666666666666e-05, + "loss": 2.7602, + "step": 14497 + }, + { + "epoch": 0.7630526315789473, + "grad_norm": 0.53515625, + "learning_rate": 5.2654970760233925e-05, + "loss": 3.4815, + "step": 14498 + }, + { + "epoch": 0.7631052631578947, + "grad_norm": 0.51171875, + "learning_rate": 5.2643274853801174e-05, + "loss": 2.9821, + "step": 14499 + }, + { + "epoch": 0.7631578947368421, + "grad_norm": 0.478515625, + "learning_rate": 5.2631578947368424e-05, + "loss": 3.0958, + "step": 14500 + }, + { + "epoch": 0.7632105263157895, + "grad_norm": 0.490234375, + "learning_rate": 5.261988304093568e-05, + "loss": 2.7624, + "step": 14501 + }, + { + "epoch": 0.7632631578947369, + "grad_norm": 0.4765625, + "learning_rate": 5.260818713450293e-05, + "loss": 3.1451, + "step": 14502 + }, + { + "epoch": 0.7633157894736842, + "grad_norm": 0.490234375, + "learning_rate": 5.259649122807018e-05, + "loss": 3.065, + "step": 14503 + }, + { + "epoch": 0.7633684210526316, + "grad_norm": 0.484375, + "learning_rate": 5.258479532163742e-05, + "loss": 3.4707, + "step": 14504 + }, + { + "epoch": 0.763421052631579, + "grad_norm": 0.48046875, + "learning_rate": 5.2573099415204685e-05, + "loss": 3.3365, + "step": 14505 + }, + { + "epoch": 0.7634736842105263, + "grad_norm": 0.5, + "learning_rate": 5.2561403508771935e-05, + "loss": 3.0718, + "step": 14506 + }, + { + "epoch": 0.7635263157894737, + "grad_norm": 0.5078125, + "learning_rate": 5.2549707602339185e-05, + "loss": 3.0245, + "step": 14507 + }, + { + "epoch": 0.763578947368421, + "grad_norm": 0.498046875, + "learning_rate": 5.253801169590643e-05, + "loss": 3.441, + "step": 14508 + }, + { + "epoch": 0.7636315789473684, + "grad_norm": 0.484375, + "learning_rate": 5.252631578947369e-05, + "loss": 3.3928, + "step": 14509 + }, + { + "epoch": 0.7636842105263157, + "grad_norm": 0.4921875, + "learning_rate": 5.251461988304094e-05, + "loss": 3.2924, + "step": 14510 + }, + { + "epoch": 0.7637368421052632, + "grad_norm": 0.484375, + "learning_rate": 5.250292397660819e-05, + "loss": 3.1489, + "step": 14511 + }, + { + "epoch": 0.7637894736842106, + "grad_norm": 0.640625, + "learning_rate": 5.2491228070175446e-05, + "loss": 3.476, + "step": 14512 + }, + { + "epoch": 0.7638421052631579, + "grad_norm": 0.51953125, + "learning_rate": 5.2479532163742696e-05, + "loss": 3.3449, + "step": 14513 + }, + { + "epoch": 0.7638947368421053, + "grad_norm": 0.490234375, + "learning_rate": 5.2467836257309945e-05, + "loss": 3.3648, + "step": 14514 + }, + { + "epoch": 0.7639473684210526, + "grad_norm": 0.50390625, + "learning_rate": 5.2456140350877195e-05, + "loss": 3.4249, + "step": 14515 + }, + { + "epoch": 0.764, + "grad_norm": 0.482421875, + "learning_rate": 5.244444444444445e-05, + "loss": 2.6957, + "step": 14516 + }, + { + "epoch": 0.7640526315789473, + "grad_norm": 0.51171875, + "learning_rate": 5.24327485380117e-05, + "loss": 3.7365, + "step": 14517 + }, + { + "epoch": 0.7641052631578947, + "grad_norm": 0.5390625, + "learning_rate": 5.242105263157895e-05, + "loss": 3.4469, + "step": 14518 + }, + { + "epoch": 0.7641578947368421, + "grad_norm": 0.447265625, + "learning_rate": 5.24093567251462e-05, + "loss": 3.3095, + "step": 14519 + }, + { + "epoch": 0.7642105263157895, + "grad_norm": 0.49609375, + "learning_rate": 5.2397660818713456e-05, + "loss": 3.8084, + "step": 14520 + }, + { + "epoch": 0.7642631578947369, + "grad_norm": 0.48828125, + "learning_rate": 5.2385964912280706e-05, + "loss": 3.3702, + "step": 14521 + }, + { + "epoch": 0.7643157894736842, + "grad_norm": 0.474609375, + "learning_rate": 5.2374269005847956e-05, + "loss": 3.1014, + "step": 14522 + }, + { + "epoch": 0.7643684210526316, + "grad_norm": 0.462890625, + "learning_rate": 5.2362573099415205e-05, + "loss": 3.1177, + "step": 14523 + }, + { + "epoch": 0.764421052631579, + "grad_norm": 0.4765625, + "learning_rate": 5.235087719298246e-05, + "loss": 3.2754, + "step": 14524 + }, + { + "epoch": 0.7644736842105263, + "grad_norm": 0.482421875, + "learning_rate": 5.233918128654971e-05, + "loss": 3.3733, + "step": 14525 + }, + { + "epoch": 0.7645263157894737, + "grad_norm": 1.125, + "learning_rate": 5.232748538011696e-05, + "loss": 2.9868, + "step": 14526 + }, + { + "epoch": 0.764578947368421, + "grad_norm": 0.490234375, + "learning_rate": 5.231578947368422e-05, + "loss": 3.3651, + "step": 14527 + }, + { + "epoch": 0.7646315789473684, + "grad_norm": 0.51953125, + "learning_rate": 5.230409356725147e-05, + "loss": 3.4019, + "step": 14528 + }, + { + "epoch": 0.7646842105263157, + "grad_norm": 0.4765625, + "learning_rate": 5.2292397660818716e-05, + "loss": 3.438, + "step": 14529 + }, + { + "epoch": 0.7647368421052632, + "grad_norm": 0.462890625, + "learning_rate": 5.2280701754385966e-05, + "loss": 3.2429, + "step": 14530 + }, + { + "epoch": 0.7647894736842106, + "grad_norm": 0.50390625, + "learning_rate": 5.226900584795322e-05, + "loss": 3.3207, + "step": 14531 + }, + { + "epoch": 0.7648421052631579, + "grad_norm": 0.48046875, + "learning_rate": 5.225730994152047e-05, + "loss": 3.2491, + "step": 14532 + }, + { + "epoch": 0.7648947368421053, + "grad_norm": 0.47265625, + "learning_rate": 5.224561403508772e-05, + "loss": 3.3818, + "step": 14533 + }, + { + "epoch": 0.7649473684210526, + "grad_norm": 0.490234375, + "learning_rate": 5.223391812865497e-05, + "loss": 3.4785, + "step": 14534 + }, + { + "epoch": 0.765, + "grad_norm": 0.5234375, + "learning_rate": 5.222222222222223e-05, + "loss": 3.1873, + "step": 14535 + }, + { + "epoch": 0.7650526315789473, + "grad_norm": 0.49609375, + "learning_rate": 5.221052631578948e-05, + "loss": 3.2834, + "step": 14536 + }, + { + "epoch": 0.7651052631578947, + "grad_norm": 0.5078125, + "learning_rate": 5.219883040935673e-05, + "loss": 3.5674, + "step": 14537 + }, + { + "epoch": 0.7651578947368421, + "grad_norm": 0.50390625, + "learning_rate": 5.218713450292398e-05, + "loss": 3.3374, + "step": 14538 + }, + { + "epoch": 0.7652105263157895, + "grad_norm": 0.5078125, + "learning_rate": 5.217543859649123e-05, + "loss": 3.6147, + "step": 14539 + }, + { + "epoch": 0.7652631578947369, + "grad_norm": 0.494140625, + "learning_rate": 5.216374269005848e-05, + "loss": 2.8289, + "step": 14540 + }, + { + "epoch": 0.7653157894736842, + "grad_norm": 0.47265625, + "learning_rate": 5.215204678362573e-05, + "loss": 2.8431, + "step": 14541 + }, + { + "epoch": 0.7653684210526316, + "grad_norm": 0.490234375, + "learning_rate": 5.214035087719299e-05, + "loss": 3.2698, + "step": 14542 + }, + { + "epoch": 0.765421052631579, + "grad_norm": 0.462890625, + "learning_rate": 5.212865497076024e-05, + "loss": 3.2867, + "step": 14543 + }, + { + "epoch": 0.7654736842105263, + "grad_norm": 0.5, + "learning_rate": 5.211695906432749e-05, + "loss": 3.105, + "step": 14544 + }, + { + "epoch": 0.7655263157894737, + "grad_norm": 0.48828125, + "learning_rate": 5.210526315789474e-05, + "loss": 3.6109, + "step": 14545 + }, + { + "epoch": 0.765578947368421, + "grad_norm": 0.474609375, + "learning_rate": 5.209356725146199e-05, + "loss": 2.9126, + "step": 14546 + }, + { + "epoch": 0.7656315789473684, + "grad_norm": 0.51171875, + "learning_rate": 5.208187134502924e-05, + "loss": 3.4058, + "step": 14547 + }, + { + "epoch": 0.7656842105263157, + "grad_norm": 0.515625, + "learning_rate": 5.207017543859649e-05, + "loss": 2.8963, + "step": 14548 + }, + { + "epoch": 0.7657368421052632, + "grad_norm": 0.455078125, + "learning_rate": 5.205847953216374e-05, + "loss": 2.932, + "step": 14549 + }, + { + "epoch": 0.7657894736842106, + "grad_norm": 0.474609375, + "learning_rate": 5.2046783625731e-05, + "loss": 3.3661, + "step": 14550 + }, + { + "epoch": 0.7658421052631579, + "grad_norm": 0.49609375, + "learning_rate": 5.203508771929825e-05, + "loss": 3.3263, + "step": 14551 + }, + { + "epoch": 0.7658947368421053, + "grad_norm": 0.4609375, + "learning_rate": 5.20233918128655e-05, + "loss": 3.1382, + "step": 14552 + }, + { + "epoch": 0.7659473684210526, + "grad_norm": 0.470703125, + "learning_rate": 5.2011695906432754e-05, + "loss": 3.3106, + "step": 14553 + }, + { + "epoch": 0.766, + "grad_norm": 0.482421875, + "learning_rate": 5.2000000000000004e-05, + "loss": 3.2752, + "step": 14554 + }, + { + "epoch": 0.7660526315789473, + "grad_norm": 0.51953125, + "learning_rate": 5.198830409356725e-05, + "loss": 3.0351, + "step": 14555 + }, + { + "epoch": 0.7661052631578947, + "grad_norm": 0.494140625, + "learning_rate": 5.19766081871345e-05, + "loss": 3.1072, + "step": 14556 + }, + { + "epoch": 0.7661578947368421, + "grad_norm": 0.486328125, + "learning_rate": 5.196491228070176e-05, + "loss": 3.6337, + "step": 14557 + }, + { + "epoch": 0.7662105263157895, + "grad_norm": 0.50390625, + "learning_rate": 5.195321637426901e-05, + "loss": 3.0851, + "step": 14558 + }, + { + "epoch": 0.7662631578947369, + "grad_norm": 0.51171875, + "learning_rate": 5.194152046783626e-05, + "loss": 3.0831, + "step": 14559 + }, + { + "epoch": 0.7663157894736842, + "grad_norm": 0.498046875, + "learning_rate": 5.192982456140351e-05, + "loss": 3.6117, + "step": 14560 + }, + { + "epoch": 0.7663684210526316, + "grad_norm": 0.494140625, + "learning_rate": 5.1918128654970764e-05, + "loss": 3.1144, + "step": 14561 + }, + { + "epoch": 0.766421052631579, + "grad_norm": 0.6015625, + "learning_rate": 5.1906432748538014e-05, + "loss": 3.3527, + "step": 14562 + }, + { + "epoch": 0.7664736842105263, + "grad_norm": 0.51953125, + "learning_rate": 5.1894736842105263e-05, + "loss": 3.657, + "step": 14563 + }, + { + "epoch": 0.7665263157894737, + "grad_norm": 0.484375, + "learning_rate": 5.188304093567252e-05, + "loss": 3.3179, + "step": 14564 + }, + { + "epoch": 0.766578947368421, + "grad_norm": 0.5, + "learning_rate": 5.187134502923977e-05, + "loss": 3.1472, + "step": 14565 + }, + { + "epoch": 0.7666315789473684, + "grad_norm": 0.625, + "learning_rate": 5.185964912280702e-05, + "loss": 3.3638, + "step": 14566 + }, + { + "epoch": 0.7666842105263157, + "grad_norm": 0.46875, + "learning_rate": 5.184795321637427e-05, + "loss": 3.1603, + "step": 14567 + }, + { + "epoch": 0.7667368421052632, + "grad_norm": 0.52734375, + "learning_rate": 5.1836257309941525e-05, + "loss": 2.9621, + "step": 14568 + }, + { + "epoch": 0.7667894736842106, + "grad_norm": 0.51953125, + "learning_rate": 5.1824561403508775e-05, + "loss": 3.4127, + "step": 14569 + }, + { + "epoch": 0.7668421052631579, + "grad_norm": 0.474609375, + "learning_rate": 5.1812865497076024e-05, + "loss": 3.3835, + "step": 14570 + }, + { + "epoch": 0.7668947368421053, + "grad_norm": 0.494140625, + "learning_rate": 5.1801169590643274e-05, + "loss": 3.3259, + "step": 14571 + }, + { + "epoch": 0.7669473684210526, + "grad_norm": 0.484375, + "learning_rate": 5.178947368421053e-05, + "loss": 3.3313, + "step": 14572 + }, + { + "epoch": 0.767, + "grad_norm": 0.56640625, + "learning_rate": 5.177777777777778e-05, + "loss": 3.3568, + "step": 14573 + }, + { + "epoch": 0.7670526315789473, + "grad_norm": 0.498046875, + "learning_rate": 5.176608187134503e-05, + "loss": 3.3039, + "step": 14574 + }, + { + "epoch": 0.7671052631578947, + "grad_norm": 0.48828125, + "learning_rate": 5.1754385964912286e-05, + "loss": 3.2551, + "step": 14575 + }, + { + "epoch": 0.7671578947368421, + "grad_norm": 0.51171875, + "learning_rate": 5.1742690058479535e-05, + "loss": 3.6304, + "step": 14576 + }, + { + "epoch": 0.7672105263157895, + "grad_norm": 0.484375, + "learning_rate": 5.1730994152046785e-05, + "loss": 3.3228, + "step": 14577 + }, + { + "epoch": 0.7672631578947369, + "grad_norm": 0.55078125, + "learning_rate": 5.1719298245614034e-05, + "loss": 3.3757, + "step": 14578 + }, + { + "epoch": 0.7673157894736842, + "grad_norm": 0.490234375, + "learning_rate": 5.170760233918129e-05, + "loss": 3.0219, + "step": 14579 + }, + { + "epoch": 0.7673684210526316, + "grad_norm": 0.52734375, + "learning_rate": 5.169590643274854e-05, + "loss": 3.2396, + "step": 14580 + }, + { + "epoch": 0.7674210526315789, + "grad_norm": 0.5078125, + "learning_rate": 5.168421052631579e-05, + "loss": 3.2802, + "step": 14581 + }, + { + "epoch": 0.7674736842105263, + "grad_norm": 0.5234375, + "learning_rate": 5.167251461988304e-05, + "loss": 3.0701, + "step": 14582 + }, + { + "epoch": 0.7675263157894737, + "grad_norm": 0.50390625, + "learning_rate": 5.1660818713450296e-05, + "loss": 3.0546, + "step": 14583 + }, + { + "epoch": 0.767578947368421, + "grad_norm": 0.494140625, + "learning_rate": 5.1649122807017546e-05, + "loss": 3.3906, + "step": 14584 + }, + { + "epoch": 0.7676315789473684, + "grad_norm": 0.470703125, + "learning_rate": 5.1637426900584795e-05, + "loss": 3.3313, + "step": 14585 + }, + { + "epoch": 0.7676842105263157, + "grad_norm": 0.515625, + "learning_rate": 5.1625730994152045e-05, + "loss": 3.0828, + "step": 14586 + }, + { + "epoch": 0.7677368421052632, + "grad_norm": 0.51171875, + "learning_rate": 5.16140350877193e-05, + "loss": 2.7547, + "step": 14587 + }, + { + "epoch": 0.7677894736842106, + "grad_norm": 0.5, + "learning_rate": 5.160233918128655e-05, + "loss": 3.1985, + "step": 14588 + }, + { + "epoch": 0.7678421052631579, + "grad_norm": 0.50390625, + "learning_rate": 5.15906432748538e-05, + "loss": 3.1042, + "step": 14589 + }, + { + "epoch": 0.7678947368421053, + "grad_norm": 0.482421875, + "learning_rate": 5.157894736842106e-05, + "loss": 3.3492, + "step": 14590 + }, + { + "epoch": 0.7679473684210526, + "grad_norm": 0.48828125, + "learning_rate": 5.1567251461988306e-05, + "loss": 3.3937, + "step": 14591 + }, + { + "epoch": 0.768, + "grad_norm": 0.4921875, + "learning_rate": 5.1555555555555556e-05, + "loss": 3.2745, + "step": 14592 + }, + { + "epoch": 0.7680526315789473, + "grad_norm": 0.478515625, + "learning_rate": 5.1543859649122805e-05, + "loss": 3.0711, + "step": 14593 + }, + { + "epoch": 0.7681052631578947, + "grad_norm": 0.5390625, + "learning_rate": 5.153216374269006e-05, + "loss": 3.0235, + "step": 14594 + }, + { + "epoch": 0.7681578947368422, + "grad_norm": 0.474609375, + "learning_rate": 5.152046783625731e-05, + "loss": 3.05, + "step": 14595 + }, + { + "epoch": 0.7682105263157895, + "grad_norm": 0.60546875, + "learning_rate": 5.150877192982456e-05, + "loss": 3.2808, + "step": 14596 + }, + { + "epoch": 0.7682631578947369, + "grad_norm": 0.51171875, + "learning_rate": 5.149707602339181e-05, + "loss": 3.985, + "step": 14597 + }, + { + "epoch": 0.7683157894736842, + "grad_norm": 0.494140625, + "learning_rate": 5.148538011695907e-05, + "loss": 3.3294, + "step": 14598 + }, + { + "epoch": 0.7683684210526316, + "grad_norm": 0.5390625, + "learning_rate": 5.1473684210526317e-05, + "loss": 3.3643, + "step": 14599 + }, + { + "epoch": 0.7684210526315789, + "grad_norm": 0.486328125, + "learning_rate": 5.1461988304093566e-05, + "loss": 2.9096, + "step": 14600 + }, + { + "epoch": 0.7684736842105263, + "grad_norm": 0.484375, + "learning_rate": 5.145029239766083e-05, + "loss": 3.3162, + "step": 14601 + }, + { + "epoch": 0.7685263157894737, + "grad_norm": 0.494140625, + "learning_rate": 5.143859649122807e-05, + "loss": 3.2117, + "step": 14602 + }, + { + "epoch": 0.768578947368421, + "grad_norm": 0.51953125, + "learning_rate": 5.142690058479532e-05, + "loss": 3.551, + "step": 14603 + }, + { + "epoch": 0.7686315789473684, + "grad_norm": 0.494140625, + "learning_rate": 5.141520467836257e-05, + "loss": 3.1892, + "step": 14604 + }, + { + "epoch": 0.7686842105263157, + "grad_norm": 0.486328125, + "learning_rate": 5.1403508771929834e-05, + "loss": 3.0937, + "step": 14605 + }, + { + "epoch": 0.7687368421052632, + "grad_norm": 0.474609375, + "learning_rate": 5.139181286549708e-05, + "loss": 3.3357, + "step": 14606 + }, + { + "epoch": 0.7687894736842106, + "grad_norm": 0.494140625, + "learning_rate": 5.138011695906433e-05, + "loss": 3.4838, + "step": 14607 + }, + { + "epoch": 0.7688421052631579, + "grad_norm": 0.52734375, + "learning_rate": 5.1368421052631576e-05, + "loss": 2.8964, + "step": 14608 + }, + { + "epoch": 0.7688947368421053, + "grad_norm": 0.482421875, + "learning_rate": 5.135672514619884e-05, + "loss": 3.4525, + "step": 14609 + }, + { + "epoch": 0.7689473684210526, + "grad_norm": 0.490234375, + "learning_rate": 5.134502923976608e-05, + "loss": 3.6659, + "step": 14610 + }, + { + "epoch": 0.769, + "grad_norm": 0.48828125, + "learning_rate": 5.133333333333333e-05, + "loss": 3.3315, + "step": 14611 + }, + { + "epoch": 0.7690526315789473, + "grad_norm": 0.498046875, + "learning_rate": 5.132163742690058e-05, + "loss": 3.1312, + "step": 14612 + }, + { + "epoch": 0.7691052631578947, + "grad_norm": 0.5234375, + "learning_rate": 5.130994152046784e-05, + "loss": 3.4626, + "step": 14613 + }, + { + "epoch": 0.7691578947368422, + "grad_norm": 0.51171875, + "learning_rate": 5.129824561403509e-05, + "loss": 3.4863, + "step": 14614 + }, + { + "epoch": 0.7692105263157895, + "grad_norm": 0.4921875, + "learning_rate": 5.128654970760234e-05, + "loss": 3.5124, + "step": 14615 + }, + { + "epoch": 0.7692631578947369, + "grad_norm": 0.478515625, + "learning_rate": 5.12748538011696e-05, + "loss": 3.1076, + "step": 14616 + }, + { + "epoch": 0.7693157894736842, + "grad_norm": 0.5, + "learning_rate": 5.126315789473684e-05, + "loss": 3.6073, + "step": 14617 + }, + { + "epoch": 0.7693684210526316, + "grad_norm": 0.5078125, + "learning_rate": 5.125146198830409e-05, + "loss": 3.3475, + "step": 14618 + }, + { + "epoch": 0.7694210526315789, + "grad_norm": 0.47265625, + "learning_rate": 5.123976608187134e-05, + "loss": 3.0113, + "step": 14619 + }, + { + "epoch": 0.7694736842105263, + "grad_norm": 0.490234375, + "learning_rate": 5.1228070175438605e-05, + "loss": 3.6495, + "step": 14620 + }, + { + "epoch": 0.7695263157894737, + "grad_norm": 0.51171875, + "learning_rate": 5.121637426900585e-05, + "loss": 3.5522, + "step": 14621 + }, + { + "epoch": 0.769578947368421, + "grad_norm": 0.494140625, + "learning_rate": 5.12046783625731e-05, + "loss": 3.282, + "step": 14622 + }, + { + "epoch": 0.7696315789473684, + "grad_norm": 0.484375, + "learning_rate": 5.119298245614035e-05, + "loss": 3.5614, + "step": 14623 + }, + { + "epoch": 0.7696842105263157, + "grad_norm": 0.490234375, + "learning_rate": 5.118128654970761e-05, + "loss": 3.228, + "step": 14624 + }, + { + "epoch": 0.7697368421052632, + "grad_norm": 0.54296875, + "learning_rate": 5.1169590643274853e-05, + "loss": 3.5718, + "step": 14625 + }, + { + "epoch": 0.7697894736842106, + "grad_norm": 0.478515625, + "learning_rate": 5.11578947368421e-05, + "loss": 3.1295, + "step": 14626 + }, + { + "epoch": 0.7698421052631579, + "grad_norm": 0.52734375, + "learning_rate": 5.1146198830409366e-05, + "loss": 3.1929, + "step": 14627 + }, + { + "epoch": 0.7698947368421053, + "grad_norm": 0.5390625, + "learning_rate": 5.1134502923976616e-05, + "loss": 3.5496, + "step": 14628 + }, + { + "epoch": 0.7699473684210526, + "grad_norm": 0.49609375, + "learning_rate": 5.112280701754386e-05, + "loss": 3.0847, + "step": 14629 + }, + { + "epoch": 0.77, + "grad_norm": 0.48828125, + "learning_rate": 5.111111111111111e-05, + "loss": 3.2392, + "step": 14630 + }, + { + "epoch": 0.7700526315789473, + "grad_norm": 0.54296875, + "learning_rate": 5.109941520467837e-05, + "loss": 3.1598, + "step": 14631 + }, + { + "epoch": 0.7701052631578947, + "grad_norm": 0.51953125, + "learning_rate": 5.108771929824562e-05, + "loss": 3.1079, + "step": 14632 + }, + { + "epoch": 0.7701578947368422, + "grad_norm": 0.51953125, + "learning_rate": 5.1076023391812864e-05, + "loss": 3.4275, + "step": 14633 + }, + { + "epoch": 0.7702105263157895, + "grad_norm": 0.48046875, + "learning_rate": 5.106432748538011e-05, + "loss": 3.4616, + "step": 14634 + }, + { + "epoch": 0.7702631578947369, + "grad_norm": 0.51171875, + "learning_rate": 5.1052631578947376e-05, + "loss": 3.5215, + "step": 14635 + }, + { + "epoch": 0.7703157894736842, + "grad_norm": 0.46875, + "learning_rate": 5.104093567251462e-05, + "loss": 3.3864, + "step": 14636 + }, + { + "epoch": 0.7703684210526316, + "grad_norm": 0.5078125, + "learning_rate": 5.102923976608187e-05, + "loss": 3.1109, + "step": 14637 + }, + { + "epoch": 0.7704210526315789, + "grad_norm": 0.5, + "learning_rate": 5.101754385964913e-05, + "loss": 3.1382, + "step": 14638 + }, + { + "epoch": 0.7704736842105263, + "grad_norm": 0.498046875, + "learning_rate": 5.100584795321638e-05, + "loss": 3.1383, + "step": 14639 + }, + { + "epoch": 0.7705263157894737, + "grad_norm": 0.478515625, + "learning_rate": 5.0994152046783624e-05, + "loss": 3.1348, + "step": 14640 + }, + { + "epoch": 0.770578947368421, + "grad_norm": 0.50390625, + "learning_rate": 5.0982456140350874e-05, + "loss": 3.135, + "step": 14641 + }, + { + "epoch": 0.7706315789473684, + "grad_norm": 0.50390625, + "learning_rate": 5.097076023391814e-05, + "loss": 3.1918, + "step": 14642 + }, + { + "epoch": 0.7706842105263157, + "grad_norm": 0.48828125, + "learning_rate": 5.095906432748539e-05, + "loss": 3.1394, + "step": 14643 + }, + { + "epoch": 0.7707368421052632, + "grad_norm": 0.5078125, + "learning_rate": 5.094736842105263e-05, + "loss": 3.1158, + "step": 14644 + }, + { + "epoch": 0.7707894736842106, + "grad_norm": 0.51171875, + "learning_rate": 5.093567251461988e-05, + "loss": 3.42, + "step": 14645 + }, + { + "epoch": 0.7708421052631579, + "grad_norm": 0.490234375, + "learning_rate": 5.092397660818714e-05, + "loss": 3.4161, + "step": 14646 + }, + { + "epoch": 0.7708947368421053, + "grad_norm": 0.50390625, + "learning_rate": 5.091228070175439e-05, + "loss": 3.3051, + "step": 14647 + }, + { + "epoch": 0.7709473684210526, + "grad_norm": 0.53515625, + "learning_rate": 5.0900584795321635e-05, + "loss": 3.4311, + "step": 14648 + }, + { + "epoch": 0.771, + "grad_norm": 0.515625, + "learning_rate": 5.0888888888888884e-05, + "loss": 2.9011, + "step": 14649 + }, + { + "epoch": 0.7710526315789473, + "grad_norm": 0.4921875, + "learning_rate": 5.087719298245615e-05, + "loss": 3.3207, + "step": 14650 + }, + { + "epoch": 0.7711052631578947, + "grad_norm": 0.5234375, + "learning_rate": 5.08654970760234e-05, + "loss": 3.041, + "step": 14651 + }, + { + "epoch": 0.7711578947368422, + "grad_norm": 0.56640625, + "learning_rate": 5.085380116959064e-05, + "loss": 3.1462, + "step": 14652 + }, + { + "epoch": 0.7712105263157895, + "grad_norm": 0.5078125, + "learning_rate": 5.08421052631579e-05, + "loss": 3.1357, + "step": 14653 + }, + { + "epoch": 0.7712631578947369, + "grad_norm": 0.5078125, + "learning_rate": 5.083040935672515e-05, + "loss": 3.2188, + "step": 14654 + }, + { + "epoch": 0.7713157894736842, + "grad_norm": 0.49609375, + "learning_rate": 5.08187134502924e-05, + "loss": 3.6384, + "step": 14655 + }, + { + "epoch": 0.7713684210526316, + "grad_norm": 0.4921875, + "learning_rate": 5.0807017543859645e-05, + "loss": 3.1502, + "step": 14656 + }, + { + "epoch": 0.7714210526315789, + "grad_norm": 0.58984375, + "learning_rate": 5.079532163742691e-05, + "loss": 3.079, + "step": 14657 + }, + { + "epoch": 0.7714736842105263, + "grad_norm": 0.478515625, + "learning_rate": 5.078362573099416e-05, + "loss": 3.1814, + "step": 14658 + }, + { + "epoch": 0.7715263157894737, + "grad_norm": 0.490234375, + "learning_rate": 5.07719298245614e-05, + "loss": 3.4378, + "step": 14659 + }, + { + "epoch": 0.771578947368421, + "grad_norm": 0.48046875, + "learning_rate": 5.076023391812865e-05, + "loss": 3.302, + "step": 14660 + }, + { + "epoch": 0.7716315789473684, + "grad_norm": 1.3359375, + "learning_rate": 5.074853801169591e-05, + "loss": 3.3638, + "step": 14661 + }, + { + "epoch": 0.7716842105263158, + "grad_norm": 0.486328125, + "learning_rate": 5.073684210526316e-05, + "loss": 3.2473, + "step": 14662 + }, + { + "epoch": 0.7717368421052632, + "grad_norm": 0.51171875, + "learning_rate": 5.0725146198830406e-05, + "loss": 3.4033, + "step": 14663 + }, + { + "epoch": 0.7717894736842106, + "grad_norm": 0.5234375, + "learning_rate": 5.071345029239767e-05, + "loss": 3.0196, + "step": 14664 + }, + { + "epoch": 0.7718421052631579, + "grad_norm": 0.49609375, + "learning_rate": 5.070175438596492e-05, + "loss": 3.2845, + "step": 14665 + }, + { + "epoch": 0.7718947368421053, + "grad_norm": 0.4765625, + "learning_rate": 5.069005847953217e-05, + "loss": 3.1145, + "step": 14666 + }, + { + "epoch": 0.7719473684210526, + "grad_norm": 0.53515625, + "learning_rate": 5.067836257309941e-05, + "loss": 3.0639, + "step": 14667 + }, + { + "epoch": 0.772, + "grad_norm": 0.4921875, + "learning_rate": 5.0666666666666674e-05, + "loss": 3.3493, + "step": 14668 + }, + { + "epoch": 0.7720526315789473, + "grad_norm": 0.462890625, + "learning_rate": 5.0654970760233924e-05, + "loss": 3.0249, + "step": 14669 + }, + { + "epoch": 0.7721052631578947, + "grad_norm": 0.5078125, + "learning_rate": 5.064327485380117e-05, + "loss": 3.438, + "step": 14670 + }, + { + "epoch": 0.7721578947368422, + "grad_norm": 0.5234375, + "learning_rate": 5.0631578947368416e-05, + "loss": 3.1757, + "step": 14671 + }, + { + "epoch": 0.7722105263157895, + "grad_norm": 0.470703125, + "learning_rate": 5.061988304093568e-05, + "loss": 3.0052, + "step": 14672 + }, + { + "epoch": 0.7722631578947369, + "grad_norm": 0.5234375, + "learning_rate": 5.060818713450293e-05, + "loss": 3.5069, + "step": 14673 + }, + { + "epoch": 0.7723157894736842, + "grad_norm": 0.474609375, + "learning_rate": 5.059649122807018e-05, + "loss": 2.6761, + "step": 14674 + }, + { + "epoch": 0.7723684210526316, + "grad_norm": 0.486328125, + "learning_rate": 5.0584795321637435e-05, + "loss": 2.8018, + "step": 14675 + }, + { + "epoch": 0.7724210526315789, + "grad_norm": 0.515625, + "learning_rate": 5.0573099415204684e-05, + "loss": 3.3432, + "step": 14676 + }, + { + "epoch": 0.7724736842105263, + "grad_norm": 0.466796875, + "learning_rate": 5.0561403508771934e-05, + "loss": 2.9787, + "step": 14677 + }, + { + "epoch": 0.7725263157894737, + "grad_norm": 0.466796875, + "learning_rate": 5.0549707602339183e-05, + "loss": 3.4754, + "step": 14678 + }, + { + "epoch": 0.772578947368421, + "grad_norm": 0.46875, + "learning_rate": 5.053801169590644e-05, + "loss": 3.1539, + "step": 14679 + }, + { + "epoch": 0.7726315789473684, + "grad_norm": 0.49609375, + "learning_rate": 5.052631578947369e-05, + "loss": 3.4173, + "step": 14680 + }, + { + "epoch": 0.7726842105263158, + "grad_norm": 0.490234375, + "learning_rate": 5.051461988304094e-05, + "loss": 3.0113, + "step": 14681 + }, + { + "epoch": 0.7727368421052632, + "grad_norm": 0.5625, + "learning_rate": 5.050292397660818e-05, + "loss": 2.9827, + "step": 14682 + }, + { + "epoch": 0.7727894736842106, + "grad_norm": 0.484375, + "learning_rate": 5.0491228070175445e-05, + "loss": 3.1496, + "step": 14683 + }, + { + "epoch": 0.7728421052631579, + "grad_norm": 0.51171875, + "learning_rate": 5.0479532163742695e-05, + "loss": 3.1984, + "step": 14684 + }, + { + "epoch": 0.7728947368421053, + "grad_norm": 0.494140625, + "learning_rate": 5.0467836257309944e-05, + "loss": 3.0102, + "step": 14685 + }, + { + "epoch": 0.7729473684210526, + "grad_norm": 0.4921875, + "learning_rate": 5.045614035087719e-05, + "loss": 3.3484, + "step": 14686 + }, + { + "epoch": 0.773, + "grad_norm": 0.498046875, + "learning_rate": 5.044444444444445e-05, + "loss": 3.6468, + "step": 14687 + }, + { + "epoch": 0.7730526315789473, + "grad_norm": 0.490234375, + "learning_rate": 5.04327485380117e-05, + "loss": 3.4419, + "step": 14688 + }, + { + "epoch": 0.7731052631578947, + "grad_norm": 0.478515625, + "learning_rate": 5.042105263157895e-05, + "loss": 3.1876, + "step": 14689 + }, + { + "epoch": 0.7731578947368422, + "grad_norm": 0.498046875, + "learning_rate": 5.0409356725146206e-05, + "loss": 3.4445, + "step": 14690 + }, + { + "epoch": 0.7732105263157895, + "grad_norm": 0.5, + "learning_rate": 5.0397660818713455e-05, + "loss": 3.3486, + "step": 14691 + }, + { + "epoch": 0.7732631578947369, + "grad_norm": 0.498046875, + "learning_rate": 5.0385964912280705e-05, + "loss": 3.0862, + "step": 14692 + }, + { + "epoch": 0.7733157894736842, + "grad_norm": 0.51171875, + "learning_rate": 5.0374269005847954e-05, + "loss": 3.321, + "step": 14693 + }, + { + "epoch": 0.7733684210526316, + "grad_norm": 0.50390625, + "learning_rate": 5.036257309941521e-05, + "loss": 3.3248, + "step": 14694 + }, + { + "epoch": 0.7734210526315789, + "grad_norm": 0.48828125, + "learning_rate": 5.035087719298246e-05, + "loss": 3.4135, + "step": 14695 + }, + { + "epoch": 0.7734736842105263, + "grad_norm": 0.48046875, + "learning_rate": 5.033918128654971e-05, + "loss": 3.0472, + "step": 14696 + }, + { + "epoch": 0.7735263157894737, + "grad_norm": 0.478515625, + "learning_rate": 5.032748538011696e-05, + "loss": 2.745, + "step": 14697 + }, + { + "epoch": 0.773578947368421, + "grad_norm": 0.47265625, + "learning_rate": 5.0315789473684216e-05, + "loss": 3.1244, + "step": 14698 + }, + { + "epoch": 0.7736315789473684, + "grad_norm": 0.5078125, + "learning_rate": 5.0304093567251466e-05, + "loss": 3.0607, + "step": 14699 + }, + { + "epoch": 0.7736842105263158, + "grad_norm": 0.5078125, + "learning_rate": 5.0292397660818715e-05, + "loss": 3.4797, + "step": 14700 + }, + { + "epoch": 0.7737368421052632, + "grad_norm": 0.47265625, + "learning_rate": 5.028070175438597e-05, + "loss": 3.3425, + "step": 14701 + }, + { + "epoch": 0.7737894736842105, + "grad_norm": 0.482421875, + "learning_rate": 5.026900584795322e-05, + "loss": 3.1683, + "step": 14702 + }, + { + "epoch": 0.7738421052631579, + "grad_norm": 0.48828125, + "learning_rate": 5.025730994152047e-05, + "loss": 3.2677, + "step": 14703 + }, + { + "epoch": 0.7738947368421053, + "grad_norm": 0.50390625, + "learning_rate": 5.024561403508772e-05, + "loss": 3.1751, + "step": 14704 + }, + { + "epoch": 0.7739473684210526, + "grad_norm": 0.484375, + "learning_rate": 5.023391812865498e-05, + "loss": 3.4076, + "step": 14705 + }, + { + "epoch": 0.774, + "grad_norm": 0.478515625, + "learning_rate": 5.0222222222222226e-05, + "loss": 3.0938, + "step": 14706 + }, + { + "epoch": 0.7740526315789473, + "grad_norm": 0.48828125, + "learning_rate": 5.0210526315789476e-05, + "loss": 2.8247, + "step": 14707 + }, + { + "epoch": 0.7741052631578947, + "grad_norm": 0.482421875, + "learning_rate": 5.0198830409356725e-05, + "loss": 3.0412, + "step": 14708 + }, + { + "epoch": 0.7741578947368422, + "grad_norm": 0.5703125, + "learning_rate": 5.018713450292398e-05, + "loss": 3.1796, + "step": 14709 + }, + { + "epoch": 0.7742105263157895, + "grad_norm": 0.5078125, + "learning_rate": 5.017543859649123e-05, + "loss": 3.252, + "step": 14710 + }, + { + "epoch": 0.7742631578947369, + "grad_norm": 0.46484375, + "learning_rate": 5.016374269005848e-05, + "loss": 3.5814, + "step": 14711 + }, + { + "epoch": 0.7743157894736842, + "grad_norm": 0.48046875, + "learning_rate": 5.015204678362573e-05, + "loss": 2.9683, + "step": 14712 + }, + { + "epoch": 0.7743684210526316, + "grad_norm": 0.5078125, + "learning_rate": 5.014035087719299e-05, + "loss": 3.1295, + "step": 14713 + }, + { + "epoch": 0.7744210526315789, + "grad_norm": 0.48828125, + "learning_rate": 5.0128654970760237e-05, + "loss": 3.624, + "step": 14714 + }, + { + "epoch": 0.7744736842105263, + "grad_norm": 0.515625, + "learning_rate": 5.0116959064327486e-05, + "loss": 3.2266, + "step": 14715 + }, + { + "epoch": 0.7745263157894737, + "grad_norm": 0.498046875, + "learning_rate": 5.010526315789474e-05, + "loss": 3.4499, + "step": 14716 + }, + { + "epoch": 0.774578947368421, + "grad_norm": 0.50390625, + "learning_rate": 5.009356725146199e-05, + "loss": 3.8217, + "step": 14717 + }, + { + "epoch": 0.7746315789473684, + "grad_norm": 0.494140625, + "learning_rate": 5.008187134502924e-05, + "loss": 3.5096, + "step": 14718 + }, + { + "epoch": 0.7746842105263158, + "grad_norm": 0.5078125, + "learning_rate": 5.007017543859649e-05, + "loss": 3.572, + "step": 14719 + }, + { + "epoch": 0.7747368421052632, + "grad_norm": 0.515625, + "learning_rate": 5.005847953216375e-05, + "loss": 3.279, + "step": 14720 + }, + { + "epoch": 0.7747894736842105, + "grad_norm": 0.53515625, + "learning_rate": 5.0046783625731e-05, + "loss": 3.3903, + "step": 14721 + }, + { + "epoch": 0.7748421052631579, + "grad_norm": 0.5, + "learning_rate": 5.003508771929825e-05, + "loss": 3.269, + "step": 14722 + }, + { + "epoch": 0.7748947368421053, + "grad_norm": 0.50390625, + "learning_rate": 5.0023391812865496e-05, + "loss": 3.3218, + "step": 14723 + }, + { + "epoch": 0.7749473684210526, + "grad_norm": 0.494140625, + "learning_rate": 5.001169590643275e-05, + "loss": 3.369, + "step": 14724 + }, + { + "epoch": 0.775, + "grad_norm": 0.5546875, + "learning_rate": 5e-05, + "loss": 3.8426, + "step": 14725 + }, + { + "epoch": 0.7750526315789473, + "grad_norm": 0.51953125, + "learning_rate": 4.998830409356726e-05, + "loss": 3.329, + "step": 14726 + }, + { + "epoch": 0.7751052631578947, + "grad_norm": 0.5078125, + "learning_rate": 4.99766081871345e-05, + "loss": 3.0674, + "step": 14727 + }, + { + "epoch": 0.7751578947368422, + "grad_norm": 0.49609375, + "learning_rate": 4.996491228070176e-05, + "loss": 3.5201, + "step": 14728 + }, + { + "epoch": 0.7752105263157895, + "grad_norm": 0.48046875, + "learning_rate": 4.995321637426901e-05, + "loss": 3.5461, + "step": 14729 + }, + { + "epoch": 0.7752631578947369, + "grad_norm": 0.498046875, + "learning_rate": 4.9941520467836264e-05, + "loss": 3.4651, + "step": 14730 + }, + { + "epoch": 0.7753157894736842, + "grad_norm": 0.58203125, + "learning_rate": 4.992982456140351e-05, + "loss": 2.917, + "step": 14731 + }, + { + "epoch": 0.7753684210526316, + "grad_norm": 0.498046875, + "learning_rate": 4.991812865497076e-05, + "loss": 3.7936, + "step": 14732 + }, + { + "epoch": 0.7754210526315789, + "grad_norm": 0.482421875, + "learning_rate": 4.990643274853801e-05, + "loss": 3.5201, + "step": 14733 + }, + { + "epoch": 0.7754736842105263, + "grad_norm": 0.490234375, + "learning_rate": 4.989473684210527e-05, + "loss": 3.3534, + "step": 14734 + }, + { + "epoch": 0.7755263157894737, + "grad_norm": 0.455078125, + "learning_rate": 4.988304093567251e-05, + "loss": 3.1706, + "step": 14735 + }, + { + "epoch": 0.775578947368421, + "grad_norm": 0.47265625, + "learning_rate": 4.987134502923977e-05, + "loss": 3.3413, + "step": 14736 + }, + { + "epoch": 0.7756315789473684, + "grad_norm": 0.60546875, + "learning_rate": 4.985964912280702e-05, + "loss": 3.0876, + "step": 14737 + }, + { + "epoch": 0.7756842105263158, + "grad_norm": 0.462890625, + "learning_rate": 4.9847953216374274e-05, + "loss": 2.8703, + "step": 14738 + }, + { + "epoch": 0.7757368421052632, + "grad_norm": 0.5078125, + "learning_rate": 4.9836257309941524e-05, + "loss": 3.3346, + "step": 14739 + }, + { + "epoch": 0.7757894736842105, + "grad_norm": 0.51953125, + "learning_rate": 4.9824561403508773e-05, + "loss": 3.1812, + "step": 14740 + }, + { + "epoch": 0.7758421052631579, + "grad_norm": 0.5, + "learning_rate": 4.981286549707603e-05, + "loss": 3.722, + "step": 14741 + }, + { + "epoch": 0.7758947368421053, + "grad_norm": 0.490234375, + "learning_rate": 4.980116959064328e-05, + "loss": 3.0742, + "step": 14742 + }, + { + "epoch": 0.7759473684210526, + "grad_norm": 0.5546875, + "learning_rate": 4.978947368421053e-05, + "loss": 3.3265, + "step": 14743 + }, + { + "epoch": 0.776, + "grad_norm": 0.5, + "learning_rate": 4.977777777777778e-05, + "loss": 3.0901, + "step": 14744 + }, + { + "epoch": 0.7760526315789473, + "grad_norm": 0.515625, + "learning_rate": 4.9766081871345035e-05, + "loss": 3.0484, + "step": 14745 + }, + { + "epoch": 0.7761052631578947, + "grad_norm": 0.5234375, + "learning_rate": 4.9754385964912284e-05, + "loss": 3.54, + "step": 14746 + }, + { + "epoch": 0.7761578947368422, + "grad_norm": 0.515625, + "learning_rate": 4.9742690058479534e-05, + "loss": 3.3355, + "step": 14747 + }, + { + "epoch": 0.7762105263157895, + "grad_norm": 0.5546875, + "learning_rate": 4.9730994152046784e-05, + "loss": 2.8804, + "step": 14748 + }, + { + "epoch": 0.7762631578947369, + "grad_norm": 0.52734375, + "learning_rate": 4.971929824561404e-05, + "loss": 3.2086, + "step": 14749 + }, + { + "epoch": 0.7763157894736842, + "grad_norm": 0.515625, + "learning_rate": 4.970760233918128e-05, + "loss": 2.9619, + "step": 14750 + }, + { + "epoch": 0.7763684210526316, + "grad_norm": 0.470703125, + "learning_rate": 4.969590643274854e-05, + "loss": 3.4162, + "step": 14751 + }, + { + "epoch": 0.7764210526315789, + "grad_norm": 0.48046875, + "learning_rate": 4.9684210526315796e-05, + "loss": 3.1911, + "step": 14752 + }, + { + "epoch": 0.7764736842105263, + "grad_norm": 0.50390625, + "learning_rate": 4.9672514619883045e-05, + "loss": 3.4253, + "step": 14753 + }, + { + "epoch": 0.7765263157894737, + "grad_norm": 0.53515625, + "learning_rate": 4.9660818713450295e-05, + "loss": 2.8223, + "step": 14754 + }, + { + "epoch": 0.776578947368421, + "grad_norm": 0.48046875, + "learning_rate": 4.9649122807017544e-05, + "loss": 3.5745, + "step": 14755 + }, + { + "epoch": 0.7766315789473685, + "grad_norm": 0.51953125, + "learning_rate": 4.96374269005848e-05, + "loss": 3.3902, + "step": 14756 + }, + { + "epoch": 0.7766842105263158, + "grad_norm": 0.51953125, + "learning_rate": 4.962573099415205e-05, + "loss": 3.0501, + "step": 14757 + }, + { + "epoch": 0.7767368421052632, + "grad_norm": 0.48828125, + "learning_rate": 4.96140350877193e-05, + "loss": 2.9364, + "step": 14758 + }, + { + "epoch": 0.7767894736842105, + "grad_norm": 0.64453125, + "learning_rate": 4.960233918128655e-05, + "loss": 3.4364, + "step": 14759 + }, + { + "epoch": 0.7768421052631579, + "grad_norm": 0.51171875, + "learning_rate": 4.9590643274853806e-05, + "loss": 3.7318, + "step": 14760 + }, + { + "epoch": 0.7768947368421053, + "grad_norm": 0.47265625, + "learning_rate": 4.9578947368421055e-05, + "loss": 2.9675, + "step": 14761 + }, + { + "epoch": 0.7769473684210526, + "grad_norm": 0.5546875, + "learning_rate": 4.9567251461988305e-05, + "loss": 3.2167, + "step": 14762 + }, + { + "epoch": 0.777, + "grad_norm": 0.51953125, + "learning_rate": 4.955555555555556e-05, + "loss": 3.4729, + "step": 14763 + }, + { + "epoch": 0.7770526315789473, + "grad_norm": 0.486328125, + "learning_rate": 4.954385964912281e-05, + "loss": 3.0779, + "step": 14764 + }, + { + "epoch": 0.7771052631578947, + "grad_norm": 0.48828125, + "learning_rate": 4.953216374269006e-05, + "loss": 3.994, + "step": 14765 + }, + { + "epoch": 0.7771578947368422, + "grad_norm": 0.498046875, + "learning_rate": 4.952046783625731e-05, + "loss": 3.2985, + "step": 14766 + }, + { + "epoch": 0.7772105263157895, + "grad_norm": 0.48046875, + "learning_rate": 4.9508771929824567e-05, + "loss": 3.3408, + "step": 14767 + }, + { + "epoch": 0.7772631578947369, + "grad_norm": 0.490234375, + "learning_rate": 4.9497076023391816e-05, + "loss": 3.3616, + "step": 14768 + }, + { + "epoch": 0.7773157894736842, + "grad_norm": 0.474609375, + "learning_rate": 4.9485380116959066e-05, + "loss": 3.4348, + "step": 14769 + }, + { + "epoch": 0.7773684210526316, + "grad_norm": 0.5390625, + "learning_rate": 4.9473684210526315e-05, + "loss": 3.0051, + "step": 14770 + }, + { + "epoch": 0.7774210526315789, + "grad_norm": 0.5234375, + "learning_rate": 4.946198830409357e-05, + "loss": 3.5984, + "step": 14771 + }, + { + "epoch": 0.7774736842105263, + "grad_norm": 0.5234375, + "learning_rate": 4.945029239766082e-05, + "loss": 3.3939, + "step": 14772 + }, + { + "epoch": 0.7775263157894737, + "grad_norm": 0.4765625, + "learning_rate": 4.943859649122807e-05, + "loss": 3.3272, + "step": 14773 + }, + { + "epoch": 0.777578947368421, + "grad_norm": 0.4921875, + "learning_rate": 4.942690058479532e-05, + "loss": 3.8402, + "step": 14774 + }, + { + "epoch": 0.7776315789473685, + "grad_norm": 0.5078125, + "learning_rate": 4.941520467836258e-05, + "loss": 3.1004, + "step": 14775 + }, + { + "epoch": 0.7776842105263158, + "grad_norm": 0.486328125, + "learning_rate": 4.9403508771929826e-05, + "loss": 3.3407, + "step": 14776 + }, + { + "epoch": 0.7777368421052632, + "grad_norm": 0.515625, + "learning_rate": 4.9391812865497076e-05, + "loss": 3.0555, + "step": 14777 + }, + { + "epoch": 0.7777894736842105, + "grad_norm": 0.5078125, + "learning_rate": 4.938011695906433e-05, + "loss": 3.0659, + "step": 14778 + }, + { + "epoch": 0.7778421052631579, + "grad_norm": 0.498046875, + "learning_rate": 4.936842105263158e-05, + "loss": 3.1075, + "step": 14779 + }, + { + "epoch": 0.7778947368421053, + "grad_norm": 0.48046875, + "learning_rate": 4.935672514619883e-05, + "loss": 3.1917, + "step": 14780 + }, + { + "epoch": 0.7779473684210526, + "grad_norm": 0.5, + "learning_rate": 4.934502923976608e-05, + "loss": 3.7123, + "step": 14781 + }, + { + "epoch": 0.778, + "grad_norm": 0.486328125, + "learning_rate": 4.933333333333334e-05, + "loss": 3.1882, + "step": 14782 + }, + { + "epoch": 0.7780526315789473, + "grad_norm": 0.51171875, + "learning_rate": 4.932163742690059e-05, + "loss": 3.2405, + "step": 14783 + }, + { + "epoch": 0.7781052631578947, + "grad_norm": 0.50390625, + "learning_rate": 4.930994152046784e-05, + "loss": 3.0279, + "step": 14784 + }, + { + "epoch": 0.7781578947368422, + "grad_norm": 0.52734375, + "learning_rate": 4.9298245614035086e-05, + "loss": 2.9914, + "step": 14785 + }, + { + "epoch": 0.7782105263157895, + "grad_norm": 0.515625, + "learning_rate": 4.928654970760234e-05, + "loss": 3.7563, + "step": 14786 + }, + { + "epoch": 0.7782631578947369, + "grad_norm": 0.50390625, + "learning_rate": 4.927485380116959e-05, + "loss": 3.0846, + "step": 14787 + }, + { + "epoch": 0.7783157894736842, + "grad_norm": 0.48828125, + "learning_rate": 4.926315789473684e-05, + "loss": 3.3488, + "step": 14788 + }, + { + "epoch": 0.7783684210526316, + "grad_norm": 0.484375, + "learning_rate": 4.92514619883041e-05, + "loss": 3.2297, + "step": 14789 + }, + { + "epoch": 0.7784210526315789, + "grad_norm": 0.48828125, + "learning_rate": 4.923976608187135e-05, + "loss": 3.0511, + "step": 14790 + }, + { + "epoch": 0.7784736842105263, + "grad_norm": 0.53125, + "learning_rate": 4.9228070175438604e-05, + "loss": 3.4229, + "step": 14791 + }, + { + "epoch": 0.7785263157894737, + "grad_norm": 0.515625, + "learning_rate": 4.921637426900585e-05, + "loss": 3.4892, + "step": 14792 + }, + { + "epoch": 0.778578947368421, + "grad_norm": 0.52734375, + "learning_rate": 4.9204678362573103e-05, + "loss": 3.2504, + "step": 14793 + }, + { + "epoch": 0.7786315789473685, + "grad_norm": 0.5078125, + "learning_rate": 4.919298245614035e-05, + "loss": 3.7004, + "step": 14794 + }, + { + "epoch": 0.7786842105263158, + "grad_norm": 0.490234375, + "learning_rate": 4.91812865497076e-05, + "loss": 3.7306, + "step": 14795 + }, + { + "epoch": 0.7787368421052632, + "grad_norm": 0.515625, + "learning_rate": 4.916959064327485e-05, + "loss": 3.4438, + "step": 14796 + }, + { + "epoch": 0.7787894736842105, + "grad_norm": 0.5546875, + "learning_rate": 4.915789473684211e-05, + "loss": 3.2681, + "step": 14797 + }, + { + "epoch": 0.7788421052631579, + "grad_norm": 0.51171875, + "learning_rate": 4.914619883040936e-05, + "loss": 3.372, + "step": 14798 + }, + { + "epoch": 0.7788947368421053, + "grad_norm": 0.484375, + "learning_rate": 4.913450292397661e-05, + "loss": 3.5883, + "step": 14799 + }, + { + "epoch": 0.7789473684210526, + "grad_norm": 0.490234375, + "learning_rate": 4.912280701754386e-05, + "loss": 2.9173, + "step": 14800 + }, + { + "epoch": 0.779, + "grad_norm": 0.56640625, + "learning_rate": 4.9111111111111114e-05, + "loss": 3.1254, + "step": 14801 + }, + { + "epoch": 0.7790526315789473, + "grad_norm": 0.494140625, + "learning_rate": 4.909941520467837e-05, + "loss": 3.1371, + "step": 14802 + }, + { + "epoch": 0.7791052631578947, + "grad_norm": 0.5, + "learning_rate": 4.908771929824561e-05, + "loss": 3.2947, + "step": 14803 + }, + { + "epoch": 0.7791578947368422, + "grad_norm": 0.51171875, + "learning_rate": 4.907602339181287e-05, + "loss": 3.2247, + "step": 14804 + }, + { + "epoch": 0.7792105263157895, + "grad_norm": 0.474609375, + "learning_rate": 4.906432748538012e-05, + "loss": 3.6015, + "step": 14805 + }, + { + "epoch": 0.7792631578947369, + "grad_norm": 0.51171875, + "learning_rate": 4.9052631578947375e-05, + "loss": 3.0025, + "step": 14806 + }, + { + "epoch": 0.7793157894736842, + "grad_norm": 0.50390625, + "learning_rate": 4.904093567251462e-05, + "loss": 3.3961, + "step": 14807 + }, + { + "epoch": 0.7793684210526316, + "grad_norm": 0.4921875, + "learning_rate": 4.9029239766081874e-05, + "loss": 3.5191, + "step": 14808 + }, + { + "epoch": 0.7794210526315789, + "grad_norm": 0.55078125, + "learning_rate": 4.9017543859649124e-05, + "loss": 3.4521, + "step": 14809 + }, + { + "epoch": 0.7794736842105263, + "grad_norm": 0.51953125, + "learning_rate": 4.900584795321638e-05, + "loss": 3.3356, + "step": 14810 + }, + { + "epoch": 0.7795263157894737, + "grad_norm": 0.4921875, + "learning_rate": 4.899415204678362e-05, + "loss": 3.5708, + "step": 14811 + }, + { + "epoch": 0.779578947368421, + "grad_norm": 0.49609375, + "learning_rate": 4.898245614035088e-05, + "loss": 3.3377, + "step": 14812 + }, + { + "epoch": 0.7796315789473685, + "grad_norm": 0.486328125, + "learning_rate": 4.8970760233918136e-05, + "loss": 3.5229, + "step": 14813 + }, + { + "epoch": 0.7796842105263158, + "grad_norm": 0.498046875, + "learning_rate": 4.8959064327485386e-05, + "loss": 3.5139, + "step": 14814 + }, + { + "epoch": 0.7797368421052632, + "grad_norm": 0.4609375, + "learning_rate": 4.8947368421052635e-05, + "loss": 3.3234, + "step": 14815 + }, + { + "epoch": 0.7797894736842105, + "grad_norm": 0.470703125, + "learning_rate": 4.8935672514619885e-05, + "loss": 3.3942, + "step": 14816 + }, + { + "epoch": 0.7798421052631579, + "grad_norm": 0.546875, + "learning_rate": 4.892397660818714e-05, + "loss": 3.2076, + "step": 14817 + }, + { + "epoch": 0.7798947368421053, + "grad_norm": 0.5078125, + "learning_rate": 4.8912280701754384e-05, + "loss": 3.4347, + "step": 14818 + }, + { + "epoch": 0.7799473684210526, + "grad_norm": 0.50390625, + "learning_rate": 4.890058479532164e-05, + "loss": 3.8577, + "step": 14819 + }, + { + "epoch": 0.78, + "grad_norm": 0.55078125, + "learning_rate": 4.888888888888889e-05, + "loss": 3.2979, + "step": 14820 + }, + { + "epoch": 0.7800526315789473, + "grad_norm": 0.490234375, + "learning_rate": 4.8877192982456146e-05, + "loss": 3.3362, + "step": 14821 + }, + { + "epoch": 0.7801052631578947, + "grad_norm": 0.51953125, + "learning_rate": 4.886549707602339e-05, + "loss": 2.9027, + "step": 14822 + }, + { + "epoch": 0.780157894736842, + "grad_norm": 0.51171875, + "learning_rate": 4.8853801169590645e-05, + "loss": 2.7632, + "step": 14823 + }, + { + "epoch": 0.7802105263157895, + "grad_norm": 0.50390625, + "learning_rate": 4.8842105263157895e-05, + "loss": 3.5569, + "step": 14824 + }, + { + "epoch": 0.7802631578947369, + "grad_norm": 0.52734375, + "learning_rate": 4.883040935672515e-05, + "loss": 2.9838, + "step": 14825 + }, + { + "epoch": 0.7803157894736842, + "grad_norm": 0.59765625, + "learning_rate": 4.88187134502924e-05, + "loss": 3.1207, + "step": 14826 + }, + { + "epoch": 0.7803684210526316, + "grad_norm": 0.5078125, + "learning_rate": 4.880701754385965e-05, + "loss": 3.5817, + "step": 14827 + }, + { + "epoch": 0.7804210526315789, + "grad_norm": 0.49609375, + "learning_rate": 4.879532163742691e-05, + "loss": 3.2481, + "step": 14828 + }, + { + "epoch": 0.7804736842105263, + "grad_norm": 0.49609375, + "learning_rate": 4.8783625730994157e-05, + "loss": 3.5902, + "step": 14829 + }, + { + "epoch": 0.7805263157894737, + "grad_norm": 0.5390625, + "learning_rate": 4.8771929824561406e-05, + "loss": 2.8637, + "step": 14830 + }, + { + "epoch": 0.780578947368421, + "grad_norm": 0.484375, + "learning_rate": 4.8760233918128656e-05, + "loss": 2.8334, + "step": 14831 + }, + { + "epoch": 0.7806315789473685, + "grad_norm": 0.478515625, + "learning_rate": 4.874853801169591e-05, + "loss": 3.1594, + "step": 14832 + }, + { + "epoch": 0.7806842105263158, + "grad_norm": 0.48828125, + "learning_rate": 4.873684210526316e-05, + "loss": 3.5846, + "step": 14833 + }, + { + "epoch": 0.7807368421052632, + "grad_norm": 0.494140625, + "learning_rate": 4.872514619883041e-05, + "loss": 2.839, + "step": 14834 + }, + { + "epoch": 0.7807894736842105, + "grad_norm": 0.515625, + "learning_rate": 4.871345029239766e-05, + "loss": 3.2142, + "step": 14835 + }, + { + "epoch": 0.7808421052631579, + "grad_norm": 0.48046875, + "learning_rate": 4.870175438596492e-05, + "loss": 3.2361, + "step": 14836 + }, + { + "epoch": 0.7808947368421053, + "grad_norm": 0.482421875, + "learning_rate": 4.869005847953217e-05, + "loss": 3.2438, + "step": 14837 + }, + { + "epoch": 0.7809473684210526, + "grad_norm": 0.65234375, + "learning_rate": 4.8678362573099416e-05, + "loss": 2.8507, + "step": 14838 + }, + { + "epoch": 0.781, + "grad_norm": 0.498046875, + "learning_rate": 4.866666666666667e-05, + "loss": 2.9445, + "step": 14839 + }, + { + "epoch": 0.7810526315789473, + "grad_norm": 0.482421875, + "learning_rate": 4.865497076023392e-05, + "loss": 3.1398, + "step": 14840 + }, + { + "epoch": 0.7811052631578947, + "grad_norm": 0.47265625, + "learning_rate": 4.864327485380117e-05, + "loss": 2.9942, + "step": 14841 + }, + { + "epoch": 0.781157894736842, + "grad_norm": 0.47265625, + "learning_rate": 4.863157894736842e-05, + "loss": 3.2283, + "step": 14842 + }, + { + "epoch": 0.7812105263157895, + "grad_norm": 0.49609375, + "learning_rate": 4.861988304093568e-05, + "loss": 3.3897, + "step": 14843 + }, + { + "epoch": 0.7812631578947369, + "grad_norm": 0.46484375, + "learning_rate": 4.860818713450293e-05, + "loss": 3.0788, + "step": 14844 + }, + { + "epoch": 0.7813157894736842, + "grad_norm": 0.5078125, + "learning_rate": 4.859649122807018e-05, + "loss": 3.1985, + "step": 14845 + }, + { + "epoch": 0.7813684210526316, + "grad_norm": 0.49609375, + "learning_rate": 4.858479532163743e-05, + "loss": 3.401, + "step": 14846 + }, + { + "epoch": 0.7814210526315789, + "grad_norm": 0.494140625, + "learning_rate": 4.857309941520468e-05, + "loss": 3.2218, + "step": 14847 + }, + { + "epoch": 0.7814736842105263, + "grad_norm": 0.4921875, + "learning_rate": 4.856140350877193e-05, + "loss": 3.5213, + "step": 14848 + }, + { + "epoch": 0.7815263157894737, + "grad_norm": 0.494140625, + "learning_rate": 4.854970760233918e-05, + "loss": 3.2817, + "step": 14849 + }, + { + "epoch": 0.781578947368421, + "grad_norm": 0.46875, + "learning_rate": 4.853801169590643e-05, + "loss": 3.1764, + "step": 14850 + }, + { + "epoch": 0.7816315789473685, + "grad_norm": 0.48828125, + "learning_rate": 4.852631578947369e-05, + "loss": 3.0789, + "step": 14851 + }, + { + "epoch": 0.7816842105263158, + "grad_norm": 0.55859375, + "learning_rate": 4.851461988304094e-05, + "loss": 3.1982, + "step": 14852 + }, + { + "epoch": 0.7817368421052632, + "grad_norm": 0.48828125, + "learning_rate": 4.850292397660819e-05, + "loss": 3.5453, + "step": 14853 + }, + { + "epoch": 0.7817894736842105, + "grad_norm": 0.466796875, + "learning_rate": 4.8491228070175444e-05, + "loss": 3.1987, + "step": 14854 + }, + { + "epoch": 0.7818421052631579, + "grad_norm": 0.50390625, + "learning_rate": 4.847953216374269e-05, + "loss": 3.2959, + "step": 14855 + }, + { + "epoch": 0.7818947368421053, + "grad_norm": 0.4765625, + "learning_rate": 4.846783625730994e-05, + "loss": 3.4908, + "step": 14856 + }, + { + "epoch": 0.7819473684210526, + "grad_norm": 0.88671875, + "learning_rate": 4.845614035087719e-05, + "loss": 3.2842, + "step": 14857 + }, + { + "epoch": 0.782, + "grad_norm": 0.50390625, + "learning_rate": 4.844444444444445e-05, + "loss": 3.3404, + "step": 14858 + }, + { + "epoch": 0.7820526315789473, + "grad_norm": 0.498046875, + "learning_rate": 4.84327485380117e-05, + "loss": 3.7142, + "step": 14859 + }, + { + "epoch": 0.7821052631578947, + "grad_norm": 0.49609375, + "learning_rate": 4.842105263157895e-05, + "loss": 3.4971, + "step": 14860 + }, + { + "epoch": 0.782157894736842, + "grad_norm": 0.55078125, + "learning_rate": 4.84093567251462e-05, + "loss": 3.1016, + "step": 14861 + }, + { + "epoch": 0.7822105263157895, + "grad_norm": 0.53515625, + "learning_rate": 4.8397660818713454e-05, + "loss": 3.4423, + "step": 14862 + }, + { + "epoch": 0.7822631578947369, + "grad_norm": 0.498046875, + "learning_rate": 4.8385964912280704e-05, + "loss": 3.2268, + "step": 14863 + }, + { + "epoch": 0.7823157894736842, + "grad_norm": 0.49609375, + "learning_rate": 4.837426900584795e-05, + "loss": 3.3264, + "step": 14864 + }, + { + "epoch": 0.7823684210526316, + "grad_norm": 0.486328125, + "learning_rate": 4.836257309941521e-05, + "loss": 3.3731, + "step": 14865 + }, + { + "epoch": 0.7824210526315789, + "grad_norm": 0.49609375, + "learning_rate": 4.835087719298246e-05, + "loss": 3.3194, + "step": 14866 + }, + { + "epoch": 0.7824736842105263, + "grad_norm": 0.546875, + "learning_rate": 4.833918128654971e-05, + "loss": 4.0063, + "step": 14867 + }, + { + "epoch": 0.7825263157894737, + "grad_norm": 0.49609375, + "learning_rate": 4.832748538011696e-05, + "loss": 3.2836, + "step": 14868 + }, + { + "epoch": 0.782578947368421, + "grad_norm": 0.51171875, + "learning_rate": 4.8315789473684215e-05, + "loss": 3.3072, + "step": 14869 + }, + { + "epoch": 0.7826315789473685, + "grad_norm": 0.462890625, + "learning_rate": 4.8304093567251464e-05, + "loss": 3.1061, + "step": 14870 + }, + { + "epoch": 0.7826842105263158, + "grad_norm": 0.486328125, + "learning_rate": 4.8292397660818714e-05, + "loss": 3.6353, + "step": 14871 + }, + { + "epoch": 0.7827368421052632, + "grad_norm": 0.5234375, + "learning_rate": 4.8280701754385964e-05, + "loss": 3.2163, + "step": 14872 + }, + { + "epoch": 0.7827894736842105, + "grad_norm": 0.498046875, + "learning_rate": 4.826900584795322e-05, + "loss": 3.6622, + "step": 14873 + }, + { + "epoch": 0.7828421052631579, + "grad_norm": 0.50390625, + "learning_rate": 4.825730994152047e-05, + "loss": 3.2928, + "step": 14874 + }, + { + "epoch": 0.7828947368421053, + "grad_norm": 0.494140625, + "learning_rate": 4.824561403508772e-05, + "loss": 3.2197, + "step": 14875 + }, + { + "epoch": 0.7829473684210526, + "grad_norm": 0.69140625, + "learning_rate": 4.8233918128654975e-05, + "loss": 3.2617, + "step": 14876 + }, + { + "epoch": 0.783, + "grad_norm": 0.51171875, + "learning_rate": 4.8222222222222225e-05, + "loss": 3.1006, + "step": 14877 + }, + { + "epoch": 0.7830526315789473, + "grad_norm": 0.5078125, + "learning_rate": 4.821052631578948e-05, + "loss": 3.3164, + "step": 14878 + }, + { + "epoch": 0.7831052631578947, + "grad_norm": 0.478515625, + "learning_rate": 4.8198830409356724e-05, + "loss": 3.2516, + "step": 14879 + }, + { + "epoch": 0.783157894736842, + "grad_norm": 0.48046875, + "learning_rate": 4.818713450292398e-05, + "loss": 2.9554, + "step": 14880 + }, + { + "epoch": 0.7832105263157895, + "grad_norm": 0.484375, + "learning_rate": 4.817543859649123e-05, + "loss": 3.1724, + "step": 14881 + }, + { + "epoch": 0.7832631578947369, + "grad_norm": 0.478515625, + "learning_rate": 4.816374269005848e-05, + "loss": 3.0612, + "step": 14882 + }, + { + "epoch": 0.7833157894736842, + "grad_norm": 0.51953125, + "learning_rate": 4.815204678362573e-05, + "loss": 3.1629, + "step": 14883 + }, + { + "epoch": 0.7833684210526316, + "grad_norm": 0.5, + "learning_rate": 4.8140350877192986e-05, + "loss": 3.5468, + "step": 14884 + }, + { + "epoch": 0.7834210526315789, + "grad_norm": 0.5078125, + "learning_rate": 4.8128654970760235e-05, + "loss": 3.4027, + "step": 14885 + }, + { + "epoch": 0.7834736842105263, + "grad_norm": 0.59765625, + "learning_rate": 4.8116959064327485e-05, + "loss": 3.2677, + "step": 14886 + }, + { + "epoch": 0.7835263157894737, + "grad_norm": 0.48046875, + "learning_rate": 4.8105263157894735e-05, + "loss": 3.2464, + "step": 14887 + }, + { + "epoch": 0.783578947368421, + "grad_norm": 0.490234375, + "learning_rate": 4.809356725146199e-05, + "loss": 3.4071, + "step": 14888 + }, + { + "epoch": 0.7836315789473685, + "grad_norm": 0.51953125, + "learning_rate": 4.808187134502925e-05, + "loss": 3.3037, + "step": 14889 + }, + { + "epoch": 0.7836842105263158, + "grad_norm": 0.48046875, + "learning_rate": 4.807017543859649e-05, + "loss": 3.14, + "step": 14890 + }, + { + "epoch": 0.7837368421052632, + "grad_norm": 0.52734375, + "learning_rate": 4.8058479532163746e-05, + "loss": 3.4224, + "step": 14891 + }, + { + "epoch": 0.7837894736842105, + "grad_norm": 0.486328125, + "learning_rate": 4.8046783625730996e-05, + "loss": 3.6258, + "step": 14892 + }, + { + "epoch": 0.7838421052631579, + "grad_norm": 0.5, + "learning_rate": 4.803508771929825e-05, + "loss": 3.1938, + "step": 14893 + }, + { + "epoch": 0.7838947368421053, + "grad_norm": 0.50390625, + "learning_rate": 4.8023391812865495e-05, + "loss": 3.2487, + "step": 14894 + }, + { + "epoch": 0.7839473684210526, + "grad_norm": 0.53125, + "learning_rate": 4.801169590643275e-05, + "loss": 3.7489, + "step": 14895 + }, + { + "epoch": 0.784, + "grad_norm": 0.50390625, + "learning_rate": 4.8e-05, + "loss": 3.1673, + "step": 14896 + }, + { + "epoch": 0.7840526315789473, + "grad_norm": 0.53515625, + "learning_rate": 4.798830409356726e-05, + "loss": 2.9753, + "step": 14897 + }, + { + "epoch": 0.7841052631578947, + "grad_norm": 0.50390625, + "learning_rate": 4.79766081871345e-05, + "loss": 2.9763, + "step": 14898 + }, + { + "epoch": 0.784157894736842, + "grad_norm": 0.48046875, + "learning_rate": 4.796491228070176e-05, + "loss": 3.1888, + "step": 14899 + }, + { + "epoch": 0.7842105263157895, + "grad_norm": 0.50390625, + "learning_rate": 4.7953216374269006e-05, + "loss": 3.795, + "step": 14900 + }, + { + "epoch": 0.7842631578947369, + "grad_norm": 0.5234375, + "learning_rate": 4.794152046783626e-05, + "loss": 3.4472, + "step": 14901 + }, + { + "epoch": 0.7843157894736842, + "grad_norm": 0.515625, + "learning_rate": 4.792982456140351e-05, + "loss": 3.3411, + "step": 14902 + }, + { + "epoch": 0.7843684210526316, + "grad_norm": 0.4765625, + "learning_rate": 4.791812865497076e-05, + "loss": 3.3826, + "step": 14903 + }, + { + "epoch": 0.7844210526315789, + "grad_norm": 0.49609375, + "learning_rate": 4.790643274853802e-05, + "loss": 3.6825, + "step": 14904 + }, + { + "epoch": 0.7844736842105263, + "grad_norm": 0.51171875, + "learning_rate": 4.789473684210526e-05, + "loss": 3.2971, + "step": 14905 + }, + { + "epoch": 0.7845263157894737, + "grad_norm": 0.47265625, + "learning_rate": 4.788304093567252e-05, + "loss": 3.1073, + "step": 14906 + }, + { + "epoch": 0.784578947368421, + "grad_norm": 0.50390625, + "learning_rate": 4.787134502923977e-05, + "loss": 3.3234, + "step": 14907 + }, + { + "epoch": 0.7846315789473685, + "grad_norm": 0.50390625, + "learning_rate": 4.7859649122807023e-05, + "loss": 3.4021, + "step": 14908 + }, + { + "epoch": 0.7846842105263158, + "grad_norm": 0.49609375, + "learning_rate": 4.7847953216374266e-05, + "loss": 3.2862, + "step": 14909 + }, + { + "epoch": 0.7847368421052632, + "grad_norm": 0.5, + "learning_rate": 4.783625730994152e-05, + "loss": 2.8764, + "step": 14910 + }, + { + "epoch": 0.7847894736842105, + "grad_norm": 0.478515625, + "learning_rate": 4.782456140350877e-05, + "loss": 3.3506, + "step": 14911 + }, + { + "epoch": 0.7848421052631579, + "grad_norm": 0.5, + "learning_rate": 4.781286549707603e-05, + "loss": 2.999, + "step": 14912 + }, + { + "epoch": 0.7848947368421053, + "grad_norm": 0.5078125, + "learning_rate": 4.780116959064327e-05, + "loss": 3.5326, + "step": 14913 + }, + { + "epoch": 0.7849473684210526, + "grad_norm": 0.486328125, + "learning_rate": 4.778947368421053e-05, + "loss": 2.3838, + "step": 14914 + }, + { + "epoch": 0.785, + "grad_norm": 0.484375, + "learning_rate": 4.7777777777777784e-05, + "loss": 3.51, + "step": 14915 + }, + { + "epoch": 0.7850526315789473, + "grad_norm": 0.47265625, + "learning_rate": 4.7766081871345034e-05, + "loss": 3.3938, + "step": 14916 + }, + { + "epoch": 0.7851052631578948, + "grad_norm": 0.4921875, + "learning_rate": 4.775438596491228e-05, + "loss": 3.3594, + "step": 14917 + }, + { + "epoch": 0.785157894736842, + "grad_norm": 0.484375, + "learning_rate": 4.774269005847953e-05, + "loss": 3.0129, + "step": 14918 + }, + { + "epoch": 0.7852105263157895, + "grad_norm": 0.486328125, + "learning_rate": 4.773099415204679e-05, + "loss": 3.5456, + "step": 14919 + }, + { + "epoch": 0.7852631578947369, + "grad_norm": 0.478515625, + "learning_rate": 4.771929824561404e-05, + "loss": 3.2184, + "step": 14920 + }, + { + "epoch": 0.7853157894736842, + "grad_norm": 0.48828125, + "learning_rate": 4.770760233918129e-05, + "loss": 3.6363, + "step": 14921 + }, + { + "epoch": 0.7853684210526316, + "grad_norm": 0.5, + "learning_rate": 4.769590643274854e-05, + "loss": 3.0566, + "step": 14922 + }, + { + "epoch": 0.7854210526315789, + "grad_norm": 0.478515625, + "learning_rate": 4.7684210526315794e-05, + "loss": 3.0285, + "step": 14923 + }, + { + "epoch": 0.7854736842105263, + "grad_norm": 0.53125, + "learning_rate": 4.7672514619883044e-05, + "loss": 3.191, + "step": 14924 + }, + { + "epoch": 0.7855263157894737, + "grad_norm": 0.5078125, + "learning_rate": 4.7660818713450294e-05, + "loss": 3.0793, + "step": 14925 + }, + { + "epoch": 0.785578947368421, + "grad_norm": 0.474609375, + "learning_rate": 4.764912280701755e-05, + "loss": 3.3376, + "step": 14926 + }, + { + "epoch": 0.7856315789473685, + "grad_norm": 0.47265625, + "learning_rate": 4.76374269005848e-05, + "loss": 3.537, + "step": 14927 + }, + { + "epoch": 0.7856842105263158, + "grad_norm": 0.60546875, + "learning_rate": 4.762573099415205e-05, + "loss": 3.2142, + "step": 14928 + }, + { + "epoch": 0.7857368421052632, + "grad_norm": 0.455078125, + "learning_rate": 4.76140350877193e-05, + "loss": 3.4487, + "step": 14929 + }, + { + "epoch": 0.7857894736842105, + "grad_norm": 0.55859375, + "learning_rate": 4.7602339181286555e-05, + "loss": 3.3783, + "step": 14930 + }, + { + "epoch": 0.7858421052631579, + "grad_norm": 0.48828125, + "learning_rate": 4.7590643274853805e-05, + "loss": 3.4617, + "step": 14931 + }, + { + "epoch": 0.7858947368421053, + "grad_norm": 0.474609375, + "learning_rate": 4.7578947368421054e-05, + "loss": 3.3576, + "step": 14932 + }, + { + "epoch": 0.7859473684210526, + "grad_norm": 0.5, + "learning_rate": 4.7567251461988304e-05, + "loss": 3.0522, + "step": 14933 + }, + { + "epoch": 0.786, + "grad_norm": 0.498046875, + "learning_rate": 4.755555555555556e-05, + "loss": 3.1307, + "step": 14934 + }, + { + "epoch": 0.7860526315789473, + "grad_norm": 0.494140625, + "learning_rate": 4.754385964912281e-05, + "loss": 3.4523, + "step": 14935 + }, + { + "epoch": 0.7861052631578948, + "grad_norm": 0.48046875, + "learning_rate": 4.753216374269006e-05, + "loss": 3.5625, + "step": 14936 + }, + { + "epoch": 0.7861578947368421, + "grad_norm": 0.53125, + "learning_rate": 4.752046783625731e-05, + "loss": 3.2335, + "step": 14937 + }, + { + "epoch": 0.7862105263157895, + "grad_norm": 0.48828125, + "learning_rate": 4.7508771929824565e-05, + "loss": 3.2238, + "step": 14938 + }, + { + "epoch": 0.7862631578947369, + "grad_norm": 0.48046875, + "learning_rate": 4.7497076023391815e-05, + "loss": 3.4475, + "step": 14939 + }, + { + "epoch": 0.7863157894736842, + "grad_norm": 0.6171875, + "learning_rate": 4.7485380116959065e-05, + "loss": 3.1807, + "step": 14940 + }, + { + "epoch": 0.7863684210526316, + "grad_norm": 0.49609375, + "learning_rate": 4.747368421052632e-05, + "loss": 3.3488, + "step": 14941 + }, + { + "epoch": 0.7864210526315789, + "grad_norm": 0.50390625, + "learning_rate": 4.746198830409357e-05, + "loss": 3.315, + "step": 14942 + }, + { + "epoch": 0.7864736842105263, + "grad_norm": 0.482421875, + "learning_rate": 4.745029239766082e-05, + "loss": 3.2671, + "step": 14943 + }, + { + "epoch": 0.7865263157894736, + "grad_norm": 0.494140625, + "learning_rate": 4.743859649122807e-05, + "loss": 3.4812, + "step": 14944 + }, + { + "epoch": 0.786578947368421, + "grad_norm": 0.5234375, + "learning_rate": 4.7426900584795326e-05, + "loss": 3.0005, + "step": 14945 + }, + { + "epoch": 0.7866315789473685, + "grad_norm": 0.50390625, + "learning_rate": 4.7415204678362576e-05, + "loss": 3.3751, + "step": 14946 + }, + { + "epoch": 0.7866842105263158, + "grad_norm": 0.486328125, + "learning_rate": 4.7403508771929825e-05, + "loss": 3.121, + "step": 14947 + }, + { + "epoch": 0.7867368421052632, + "grad_norm": 0.478515625, + "learning_rate": 4.7391812865497075e-05, + "loss": 3.1523, + "step": 14948 + }, + { + "epoch": 0.7867894736842105, + "grad_norm": 0.5546875, + "learning_rate": 4.738011695906433e-05, + "loss": 3.4228, + "step": 14949 + }, + { + "epoch": 0.7868421052631579, + "grad_norm": 0.5, + "learning_rate": 4.736842105263158e-05, + "loss": 3.2812, + "step": 14950 + }, + { + "epoch": 0.7868947368421053, + "grad_norm": 0.50390625, + "learning_rate": 4.735672514619883e-05, + "loss": 3.4509, + "step": 14951 + }, + { + "epoch": 0.7869473684210526, + "grad_norm": 0.5, + "learning_rate": 4.734502923976609e-05, + "loss": 3.3322, + "step": 14952 + }, + { + "epoch": 0.787, + "grad_norm": 0.478515625, + "learning_rate": 4.7333333333333336e-05, + "loss": 3.2668, + "step": 14953 + }, + { + "epoch": 0.7870526315789473, + "grad_norm": 0.5078125, + "learning_rate": 4.7321637426900586e-05, + "loss": 3.2428, + "step": 14954 + }, + { + "epoch": 0.7871052631578948, + "grad_norm": 0.46484375, + "learning_rate": 4.7309941520467836e-05, + "loss": 2.9873, + "step": 14955 + }, + { + "epoch": 0.7871578947368421, + "grad_norm": 0.56640625, + "learning_rate": 4.729824561403509e-05, + "loss": 3.3504, + "step": 14956 + }, + { + "epoch": 0.7872105263157895, + "grad_norm": 0.5078125, + "learning_rate": 4.728654970760234e-05, + "loss": 3.4996, + "step": 14957 + }, + { + "epoch": 0.7872631578947369, + "grad_norm": 0.4765625, + "learning_rate": 4.727485380116959e-05, + "loss": 3.2262, + "step": 14958 + }, + { + "epoch": 0.7873157894736842, + "grad_norm": 0.47265625, + "learning_rate": 4.726315789473684e-05, + "loss": 3.0854, + "step": 14959 + }, + { + "epoch": 0.7873684210526316, + "grad_norm": 0.50390625, + "learning_rate": 4.72514619883041e-05, + "loss": 3.0478, + "step": 14960 + }, + { + "epoch": 0.7874210526315789, + "grad_norm": 0.515625, + "learning_rate": 4.723976608187135e-05, + "loss": 3.3473, + "step": 14961 + }, + { + "epoch": 0.7874736842105263, + "grad_norm": 0.48046875, + "learning_rate": 4.7228070175438596e-05, + "loss": 3.1128, + "step": 14962 + }, + { + "epoch": 0.7875263157894736, + "grad_norm": 0.546875, + "learning_rate": 4.7216374269005846e-05, + "loss": 3.0813, + "step": 14963 + }, + { + "epoch": 0.787578947368421, + "grad_norm": 0.50390625, + "learning_rate": 4.72046783625731e-05, + "loss": 3.4796, + "step": 14964 + }, + { + "epoch": 0.7876315789473685, + "grad_norm": 0.55078125, + "learning_rate": 4.719298245614036e-05, + "loss": 2.8952, + "step": 14965 + }, + { + "epoch": 0.7876842105263158, + "grad_norm": 0.478515625, + "learning_rate": 4.71812865497076e-05, + "loss": 3.1811, + "step": 14966 + }, + { + "epoch": 0.7877368421052632, + "grad_norm": 0.52734375, + "learning_rate": 4.716959064327486e-05, + "loss": 3.1685, + "step": 14967 + }, + { + "epoch": 0.7877894736842105, + "grad_norm": 0.46875, + "learning_rate": 4.715789473684211e-05, + "loss": 3.2705, + "step": 14968 + }, + { + "epoch": 0.7878421052631579, + "grad_norm": 0.4921875, + "learning_rate": 4.7146198830409364e-05, + "loss": 3.1428, + "step": 14969 + }, + { + "epoch": 0.7878947368421053, + "grad_norm": 0.4921875, + "learning_rate": 4.7134502923976607e-05, + "loss": 3.3285, + "step": 14970 + }, + { + "epoch": 0.7879473684210526, + "grad_norm": 0.48046875, + "learning_rate": 4.712280701754386e-05, + "loss": 3.0734, + "step": 14971 + }, + { + "epoch": 0.788, + "grad_norm": 0.498046875, + "learning_rate": 4.711111111111111e-05, + "loss": 2.8262, + "step": 14972 + }, + { + "epoch": 0.7880526315789473, + "grad_norm": 0.484375, + "learning_rate": 4.709941520467836e-05, + "loss": 3.3207, + "step": 14973 + }, + { + "epoch": 0.7881052631578948, + "grad_norm": 0.50390625, + "learning_rate": 4.708771929824561e-05, + "loss": 3.3231, + "step": 14974 + }, + { + "epoch": 0.7881578947368421, + "grad_norm": 0.470703125, + "learning_rate": 4.707602339181287e-05, + "loss": 2.9718, + "step": 14975 + }, + { + "epoch": 0.7882105263157895, + "grad_norm": 0.765625, + "learning_rate": 4.7064327485380124e-05, + "loss": 3.103, + "step": 14976 + }, + { + "epoch": 0.7882631578947369, + "grad_norm": 0.48828125, + "learning_rate": 4.705263157894737e-05, + "loss": 3.0866, + "step": 14977 + }, + { + "epoch": 0.7883157894736842, + "grad_norm": 0.494140625, + "learning_rate": 4.7040935672514624e-05, + "loss": 3.1846, + "step": 14978 + }, + { + "epoch": 0.7883684210526316, + "grad_norm": 0.482421875, + "learning_rate": 4.702923976608187e-05, + "loss": 3.5408, + "step": 14979 + }, + { + "epoch": 0.7884210526315789, + "grad_norm": 0.490234375, + "learning_rate": 4.701754385964913e-05, + "loss": 3.1762, + "step": 14980 + }, + { + "epoch": 0.7884736842105263, + "grad_norm": 0.494140625, + "learning_rate": 4.700584795321637e-05, + "loss": 3.3874, + "step": 14981 + }, + { + "epoch": 0.7885263157894736, + "grad_norm": 0.47265625, + "learning_rate": 4.699415204678363e-05, + "loss": 3.5459, + "step": 14982 + }, + { + "epoch": 0.788578947368421, + "grad_norm": 0.49609375, + "learning_rate": 4.698245614035088e-05, + "loss": 3.1038, + "step": 14983 + }, + { + "epoch": 0.7886315789473685, + "grad_norm": 0.50390625, + "learning_rate": 4.6970760233918135e-05, + "loss": 3.2682, + "step": 14984 + }, + { + "epoch": 0.7886842105263158, + "grad_norm": 0.49609375, + "learning_rate": 4.695906432748538e-05, + "loss": 2.9736, + "step": 14985 + }, + { + "epoch": 0.7887368421052632, + "grad_norm": 0.5078125, + "learning_rate": 4.6947368421052634e-05, + "loss": 3.3521, + "step": 14986 + }, + { + "epoch": 0.7887894736842105, + "grad_norm": 0.5, + "learning_rate": 4.6935672514619884e-05, + "loss": 2.8411, + "step": 14987 + }, + { + "epoch": 0.7888421052631579, + "grad_norm": 0.49609375, + "learning_rate": 4.692397660818714e-05, + "loss": 3.1744, + "step": 14988 + }, + { + "epoch": 0.7888947368421053, + "grad_norm": 0.466796875, + "learning_rate": 4.691228070175439e-05, + "loss": 3.4036, + "step": 14989 + }, + { + "epoch": 0.7889473684210526, + "grad_norm": 0.498046875, + "learning_rate": 4.690058479532164e-05, + "loss": 3.1795, + "step": 14990 + }, + { + "epoch": 0.789, + "grad_norm": 0.49609375, + "learning_rate": 4.6888888888888895e-05, + "loss": 2.9261, + "step": 14991 + }, + { + "epoch": 0.7890526315789473, + "grad_norm": 0.56640625, + "learning_rate": 4.6877192982456145e-05, + "loss": 3.0001, + "step": 14992 + }, + { + "epoch": 0.7891052631578948, + "grad_norm": 0.5078125, + "learning_rate": 4.6865497076023395e-05, + "loss": 3.1068, + "step": 14993 + }, + { + "epoch": 0.7891578947368421, + "grad_norm": 0.48046875, + "learning_rate": 4.6853801169590644e-05, + "loss": 3.5338, + "step": 14994 + }, + { + "epoch": 0.7892105263157895, + "grad_norm": 0.5, + "learning_rate": 4.68421052631579e-05, + "loss": 2.9518, + "step": 14995 + }, + { + "epoch": 0.7892631578947369, + "grad_norm": 0.49609375, + "learning_rate": 4.6830409356725143e-05, + "loss": 3.5491, + "step": 14996 + }, + { + "epoch": 0.7893157894736842, + "grad_norm": 0.486328125, + "learning_rate": 4.68187134502924e-05, + "loss": 3.1814, + "step": 14997 + }, + { + "epoch": 0.7893684210526316, + "grad_norm": 0.5, + "learning_rate": 4.680701754385965e-05, + "loss": 3.1783, + "step": 14998 + }, + { + "epoch": 0.7894210526315789, + "grad_norm": 0.490234375, + "learning_rate": 4.6795321637426906e-05, + "loss": 3.4033, + "step": 14999 + }, + { + "epoch": 0.7894736842105263, + "grad_norm": 0.453125, + "learning_rate": 4.678362573099415e-05, + "loss": 3.5236, + "step": 15000 + }, + { + "epoch": 0.7895263157894736, + "grad_norm": 0.49609375, + "learning_rate": 4.6771929824561405e-05, + "loss": 3.3711, + "step": 15001 + }, + { + "epoch": 0.789578947368421, + "grad_norm": 0.53125, + "learning_rate": 4.676023391812866e-05, + "loss": 3.8062, + "step": 15002 + }, + { + "epoch": 0.7896315789473685, + "grad_norm": 0.48046875, + "learning_rate": 4.674853801169591e-05, + "loss": 2.9262, + "step": 15003 + }, + { + "epoch": 0.7896842105263158, + "grad_norm": 0.484375, + "learning_rate": 4.673684210526316e-05, + "loss": 3.1154, + "step": 15004 + }, + { + "epoch": 0.7897368421052632, + "grad_norm": 0.48046875, + "learning_rate": 4.672514619883041e-05, + "loss": 3.2468, + "step": 15005 + }, + { + "epoch": 0.7897894736842105, + "grad_norm": 0.4921875, + "learning_rate": 4.6713450292397666e-05, + "loss": 2.8455, + "step": 15006 + }, + { + "epoch": 0.7898421052631579, + "grad_norm": 0.51171875, + "learning_rate": 4.6701754385964916e-05, + "loss": 3.2301, + "step": 15007 + }, + { + "epoch": 0.7898947368421053, + "grad_norm": 0.482421875, + "learning_rate": 4.6690058479532166e-05, + "loss": 3.3862, + "step": 15008 + }, + { + "epoch": 0.7899473684210526, + "grad_norm": 0.5234375, + "learning_rate": 4.6678362573099415e-05, + "loss": 3.5012, + "step": 15009 + }, + { + "epoch": 0.79, + "grad_norm": 0.54296875, + "learning_rate": 4.666666666666667e-05, + "loss": 3.118, + "step": 15010 + }, + { + "epoch": 0.7900526315789473, + "grad_norm": 0.5234375, + "learning_rate": 4.665497076023392e-05, + "loss": 3.3664, + "step": 15011 + }, + { + "epoch": 0.7901052631578948, + "grad_norm": 0.490234375, + "learning_rate": 4.664327485380117e-05, + "loss": 2.81, + "step": 15012 + }, + { + "epoch": 0.7901578947368421, + "grad_norm": 0.51953125, + "learning_rate": 4.663157894736842e-05, + "loss": 2.9927, + "step": 15013 + }, + { + "epoch": 0.7902105263157895, + "grad_norm": 0.5234375, + "learning_rate": 4.661988304093568e-05, + "loss": 3.2189, + "step": 15014 + }, + { + "epoch": 0.7902631578947369, + "grad_norm": 0.50390625, + "learning_rate": 4.6608187134502926e-05, + "loss": 3.3551, + "step": 15015 + }, + { + "epoch": 0.7903157894736842, + "grad_norm": 0.498046875, + "learning_rate": 4.6596491228070176e-05, + "loss": 3.5079, + "step": 15016 + }, + { + "epoch": 0.7903684210526316, + "grad_norm": 0.5, + "learning_rate": 4.658479532163743e-05, + "loss": 3.3926, + "step": 15017 + }, + { + "epoch": 0.7904210526315789, + "grad_norm": 0.4765625, + "learning_rate": 4.657309941520468e-05, + "loss": 3.212, + "step": 15018 + }, + { + "epoch": 0.7904736842105263, + "grad_norm": 0.494140625, + "learning_rate": 4.656140350877193e-05, + "loss": 3.427, + "step": 15019 + }, + { + "epoch": 0.7905263157894736, + "grad_norm": 0.482421875, + "learning_rate": 4.654970760233918e-05, + "loss": 2.9399, + "step": 15020 + }, + { + "epoch": 0.790578947368421, + "grad_norm": 0.4921875, + "learning_rate": 4.653801169590644e-05, + "loss": 2.7539, + "step": 15021 + }, + { + "epoch": 0.7906315789473685, + "grad_norm": 0.55078125, + "learning_rate": 4.652631578947369e-05, + "loss": 3.2252, + "step": 15022 + }, + { + "epoch": 0.7906842105263158, + "grad_norm": 0.51171875, + "learning_rate": 4.651461988304094e-05, + "loss": 2.9153, + "step": 15023 + }, + { + "epoch": 0.7907368421052632, + "grad_norm": 0.4921875, + "learning_rate": 4.6502923976608186e-05, + "loss": 3.4384, + "step": 15024 + }, + { + "epoch": 0.7907894736842105, + "grad_norm": 0.486328125, + "learning_rate": 4.649122807017544e-05, + "loss": 3.5308, + "step": 15025 + }, + { + "epoch": 0.7908421052631579, + "grad_norm": 0.4921875, + "learning_rate": 4.647953216374269e-05, + "loss": 3.1835, + "step": 15026 + }, + { + "epoch": 0.7908947368421053, + "grad_norm": 0.50390625, + "learning_rate": 4.646783625730994e-05, + "loss": 3.4662, + "step": 15027 + }, + { + "epoch": 0.7909473684210526, + "grad_norm": 0.50390625, + "learning_rate": 4.64561403508772e-05, + "loss": 3.4607, + "step": 15028 + }, + { + "epoch": 0.791, + "grad_norm": 0.4765625, + "learning_rate": 4.644444444444445e-05, + "loss": 3.4721, + "step": 15029 + }, + { + "epoch": 0.7910526315789473, + "grad_norm": 0.51953125, + "learning_rate": 4.64327485380117e-05, + "loss": 3.3519, + "step": 15030 + }, + { + "epoch": 0.7911052631578948, + "grad_norm": 0.478515625, + "learning_rate": 4.642105263157895e-05, + "loss": 3.2769, + "step": 15031 + }, + { + "epoch": 0.7911578947368421, + "grad_norm": 0.5859375, + "learning_rate": 4.64093567251462e-05, + "loss": 3.0392, + "step": 15032 + }, + { + "epoch": 0.7912105263157895, + "grad_norm": 0.5078125, + "learning_rate": 4.639766081871345e-05, + "loss": 3.9011, + "step": 15033 + }, + { + "epoch": 0.7912631578947369, + "grad_norm": 0.54296875, + "learning_rate": 4.63859649122807e-05, + "loss": 3.4371, + "step": 15034 + }, + { + "epoch": 0.7913157894736842, + "grad_norm": 0.51171875, + "learning_rate": 4.637426900584795e-05, + "loss": 3.518, + "step": 15035 + }, + { + "epoch": 0.7913684210526316, + "grad_norm": 0.486328125, + "learning_rate": 4.636257309941521e-05, + "loss": 3.3762, + "step": 15036 + }, + { + "epoch": 0.7914210526315789, + "grad_norm": 0.58203125, + "learning_rate": 4.635087719298246e-05, + "loss": 3.4691, + "step": 15037 + }, + { + "epoch": 0.7914736842105263, + "grad_norm": 0.53125, + "learning_rate": 4.633918128654971e-05, + "loss": 3.4259, + "step": 15038 + }, + { + "epoch": 0.7915263157894736, + "grad_norm": 0.486328125, + "learning_rate": 4.6327485380116964e-05, + "loss": 3.246, + "step": 15039 + }, + { + "epoch": 0.791578947368421, + "grad_norm": 0.48828125, + "learning_rate": 4.6315789473684214e-05, + "loss": 3.3023, + "step": 15040 + }, + { + "epoch": 0.7916315789473685, + "grad_norm": 0.5234375, + "learning_rate": 4.630409356725146e-05, + "loss": 2.843, + "step": 15041 + }, + { + "epoch": 0.7916842105263158, + "grad_norm": 0.474609375, + "learning_rate": 4.629239766081871e-05, + "loss": 2.9994, + "step": 15042 + }, + { + "epoch": 0.7917368421052632, + "grad_norm": 0.498046875, + "learning_rate": 4.628070175438597e-05, + "loss": 3.3402, + "step": 15043 + }, + { + "epoch": 0.7917894736842105, + "grad_norm": 0.5, + "learning_rate": 4.626900584795322e-05, + "loss": 3.218, + "step": 15044 + }, + { + "epoch": 0.7918421052631579, + "grad_norm": 0.5625, + "learning_rate": 4.625730994152047e-05, + "loss": 2.7392, + "step": 15045 + }, + { + "epoch": 0.7918947368421053, + "grad_norm": 0.50390625, + "learning_rate": 4.624561403508772e-05, + "loss": 3.2714, + "step": 15046 + }, + { + "epoch": 0.7919473684210526, + "grad_norm": 0.5, + "learning_rate": 4.6233918128654974e-05, + "loss": 3.4198, + "step": 15047 + }, + { + "epoch": 0.792, + "grad_norm": 0.5, + "learning_rate": 4.6222222222222224e-05, + "loss": 3.3111, + "step": 15048 + }, + { + "epoch": 0.7920526315789473, + "grad_norm": 0.494140625, + "learning_rate": 4.6210526315789473e-05, + "loss": 3.4143, + "step": 15049 + }, + { + "epoch": 0.7921052631578948, + "grad_norm": 0.4765625, + "learning_rate": 4.619883040935672e-05, + "loss": 3.4452, + "step": 15050 + }, + { + "epoch": 0.7921578947368421, + "grad_norm": 0.50390625, + "learning_rate": 4.618713450292398e-05, + "loss": 3.3309, + "step": 15051 + }, + { + "epoch": 0.7922105263157895, + "grad_norm": 0.490234375, + "learning_rate": 4.6175438596491236e-05, + "loss": 3.5522, + "step": 15052 + }, + { + "epoch": 0.7922631578947369, + "grad_norm": 0.52734375, + "learning_rate": 4.616374269005848e-05, + "loss": 3.4268, + "step": 15053 + }, + { + "epoch": 0.7923157894736842, + "grad_norm": 0.5, + "learning_rate": 4.6152046783625735e-05, + "loss": 3.0574, + "step": 15054 + }, + { + "epoch": 0.7923684210526316, + "grad_norm": 0.49609375, + "learning_rate": 4.6140350877192985e-05, + "loss": 3.3904, + "step": 15055 + }, + { + "epoch": 0.7924210526315789, + "grad_norm": 0.4921875, + "learning_rate": 4.612865497076024e-05, + "loss": 3.3025, + "step": 15056 + }, + { + "epoch": 0.7924736842105263, + "grad_norm": 0.484375, + "learning_rate": 4.6116959064327484e-05, + "loss": 2.9321, + "step": 15057 + }, + { + "epoch": 0.7925263157894736, + "grad_norm": 0.5234375, + "learning_rate": 4.610526315789474e-05, + "loss": 3.2975, + "step": 15058 + }, + { + "epoch": 0.792578947368421, + "grad_norm": 0.48046875, + "learning_rate": 4.609356725146199e-05, + "loss": 3.1803, + "step": 15059 + }, + { + "epoch": 0.7926315789473685, + "grad_norm": 0.486328125, + "learning_rate": 4.6081871345029246e-05, + "loss": 3.1339, + "step": 15060 + }, + { + "epoch": 0.7926842105263158, + "grad_norm": 0.498046875, + "learning_rate": 4.607017543859649e-05, + "loss": 3.2855, + "step": 15061 + }, + { + "epoch": 0.7927368421052632, + "grad_norm": 0.498046875, + "learning_rate": 4.6058479532163745e-05, + "loss": 3.4182, + "step": 15062 + }, + { + "epoch": 0.7927894736842105, + "grad_norm": 0.47265625, + "learning_rate": 4.6046783625730995e-05, + "loss": 3.1112, + "step": 15063 + }, + { + "epoch": 0.7928421052631579, + "grad_norm": 0.482421875, + "learning_rate": 4.6035087719298244e-05, + "loss": 3.0401, + "step": 15064 + }, + { + "epoch": 0.7928947368421052, + "grad_norm": 0.50390625, + "learning_rate": 4.60233918128655e-05, + "loss": 2.7573, + "step": 15065 + }, + { + "epoch": 0.7929473684210526, + "grad_norm": 0.482421875, + "learning_rate": 4.601169590643275e-05, + "loss": 3.5179, + "step": 15066 + }, + { + "epoch": 0.793, + "grad_norm": 0.5390625, + "learning_rate": 4.600000000000001e-05, + "loss": 3.094, + "step": 15067 + }, + { + "epoch": 0.7930526315789473, + "grad_norm": 0.5, + "learning_rate": 4.598830409356725e-05, + "loss": 3.4319, + "step": 15068 + }, + { + "epoch": 0.7931052631578948, + "grad_norm": 0.50390625, + "learning_rate": 4.5976608187134506e-05, + "loss": 3.3829, + "step": 15069 + }, + { + "epoch": 0.7931578947368421, + "grad_norm": 0.474609375, + "learning_rate": 4.5964912280701756e-05, + "loss": 3.2695, + "step": 15070 + }, + { + "epoch": 0.7932105263157895, + "grad_norm": 0.484375, + "learning_rate": 4.595321637426901e-05, + "loss": 3.7273, + "step": 15071 + }, + { + "epoch": 0.7932631578947369, + "grad_norm": 0.486328125, + "learning_rate": 4.5941520467836255e-05, + "loss": 3.312, + "step": 15072 + }, + { + "epoch": 0.7933157894736842, + "grad_norm": 0.494140625, + "learning_rate": 4.592982456140351e-05, + "loss": 2.8004, + "step": 15073 + }, + { + "epoch": 0.7933684210526316, + "grad_norm": 0.546875, + "learning_rate": 4.591812865497076e-05, + "loss": 2.8788, + "step": 15074 + }, + { + "epoch": 0.7934210526315789, + "grad_norm": 0.55859375, + "learning_rate": 4.590643274853802e-05, + "loss": 3.4073, + "step": 15075 + }, + { + "epoch": 0.7934736842105263, + "grad_norm": 0.515625, + "learning_rate": 4.589473684210526e-05, + "loss": 3.4878, + "step": 15076 + }, + { + "epoch": 0.7935263157894736, + "grad_norm": 0.5390625, + "learning_rate": 4.5883040935672516e-05, + "loss": 3.1969, + "step": 15077 + }, + { + "epoch": 0.793578947368421, + "grad_norm": 0.482421875, + "learning_rate": 4.587134502923977e-05, + "loss": 3.3215, + "step": 15078 + }, + { + "epoch": 0.7936315789473685, + "grad_norm": 0.51171875, + "learning_rate": 4.585964912280702e-05, + "loss": 3.2696, + "step": 15079 + }, + { + "epoch": 0.7936842105263158, + "grad_norm": 0.5, + "learning_rate": 4.584795321637427e-05, + "loss": 3.4544, + "step": 15080 + }, + { + "epoch": 0.7937368421052632, + "grad_norm": 0.474609375, + "learning_rate": 4.583625730994152e-05, + "loss": 3.0995, + "step": 15081 + }, + { + "epoch": 0.7937894736842105, + "grad_norm": 0.50390625, + "learning_rate": 4.582456140350878e-05, + "loss": 3.1717, + "step": 15082 + }, + { + "epoch": 0.7938421052631579, + "grad_norm": 0.53125, + "learning_rate": 4.581286549707603e-05, + "loss": 3.5006, + "step": 15083 + }, + { + "epoch": 0.7938947368421052, + "grad_norm": 0.50390625, + "learning_rate": 4.580116959064328e-05, + "loss": 3.3291, + "step": 15084 + }, + { + "epoch": 0.7939473684210526, + "grad_norm": 0.494140625, + "learning_rate": 4.5789473684210527e-05, + "loss": 3.2055, + "step": 15085 + }, + { + "epoch": 0.794, + "grad_norm": 0.498046875, + "learning_rate": 4.577777777777778e-05, + "loss": 2.942, + "step": 15086 + }, + { + "epoch": 0.7940526315789473, + "grad_norm": 0.494140625, + "learning_rate": 4.5766081871345026e-05, + "loss": 3.4735, + "step": 15087 + }, + { + "epoch": 0.7941052631578948, + "grad_norm": 0.51171875, + "learning_rate": 4.575438596491228e-05, + "loss": 3.5656, + "step": 15088 + }, + { + "epoch": 0.7941578947368421, + "grad_norm": 0.486328125, + "learning_rate": 4.574269005847954e-05, + "loss": 3.5342, + "step": 15089 + }, + { + "epoch": 0.7942105263157895, + "grad_norm": 0.51171875, + "learning_rate": 4.573099415204679e-05, + "loss": 3.3652, + "step": 15090 + }, + { + "epoch": 0.7942631578947369, + "grad_norm": 0.48046875, + "learning_rate": 4.571929824561404e-05, + "loss": 3.4231, + "step": 15091 + }, + { + "epoch": 0.7943157894736842, + "grad_norm": 0.515625, + "learning_rate": 4.570760233918129e-05, + "loss": 3.07, + "step": 15092 + }, + { + "epoch": 0.7943684210526316, + "grad_norm": 0.50390625, + "learning_rate": 4.5695906432748544e-05, + "loss": 3.007, + "step": 15093 + }, + { + "epoch": 0.7944210526315789, + "grad_norm": 0.494140625, + "learning_rate": 4.568421052631579e-05, + "loss": 3.5852, + "step": 15094 + }, + { + "epoch": 0.7944736842105263, + "grad_norm": 0.48046875, + "learning_rate": 4.567251461988304e-05, + "loss": 3.4215, + "step": 15095 + }, + { + "epoch": 0.7945263157894736, + "grad_norm": 0.490234375, + "learning_rate": 4.566081871345029e-05, + "loss": 3.3644, + "step": 15096 + }, + { + "epoch": 0.794578947368421, + "grad_norm": 0.482421875, + "learning_rate": 4.564912280701755e-05, + "loss": 3.2597, + "step": 15097 + }, + { + "epoch": 0.7946315789473685, + "grad_norm": 0.484375, + "learning_rate": 4.56374269005848e-05, + "loss": 3.3319, + "step": 15098 + }, + { + "epoch": 0.7946842105263158, + "grad_norm": 0.4765625, + "learning_rate": 4.562573099415205e-05, + "loss": 3.3638, + "step": 15099 + }, + { + "epoch": 0.7947368421052632, + "grad_norm": 0.50390625, + "learning_rate": 4.56140350877193e-05, + "loss": 3.1446, + "step": 15100 + }, + { + "epoch": 0.7947894736842105, + "grad_norm": 0.49609375, + "learning_rate": 4.5602339181286554e-05, + "loss": 3.3912, + "step": 15101 + }, + { + "epoch": 0.7948421052631579, + "grad_norm": 0.4765625, + "learning_rate": 4.5590643274853804e-05, + "loss": 3.2959, + "step": 15102 + }, + { + "epoch": 0.7948947368421052, + "grad_norm": 0.515625, + "learning_rate": 4.557894736842105e-05, + "loss": 3.3858, + "step": 15103 + }, + { + "epoch": 0.7949473684210526, + "grad_norm": 0.51953125, + "learning_rate": 4.556725146198831e-05, + "loss": 3.0529, + "step": 15104 + }, + { + "epoch": 0.795, + "grad_norm": 0.5, + "learning_rate": 4.555555555555556e-05, + "loss": 3.0534, + "step": 15105 + }, + { + "epoch": 0.7950526315789473, + "grad_norm": 0.5703125, + "learning_rate": 4.554385964912281e-05, + "loss": 3.173, + "step": 15106 + }, + { + "epoch": 0.7951052631578948, + "grad_norm": 0.53515625, + "learning_rate": 4.553216374269006e-05, + "loss": 3.3133, + "step": 15107 + }, + { + "epoch": 0.7951578947368421, + "grad_norm": 0.48046875, + "learning_rate": 4.5520467836257315e-05, + "loss": 2.9602, + "step": 15108 + }, + { + "epoch": 0.7952105263157895, + "grad_norm": 0.47265625, + "learning_rate": 4.5508771929824564e-05, + "loss": 3.6128, + "step": 15109 + }, + { + "epoch": 0.7952631578947369, + "grad_norm": 0.5, + "learning_rate": 4.5497076023391814e-05, + "loss": 3.2654, + "step": 15110 + }, + { + "epoch": 0.7953157894736842, + "grad_norm": 0.50390625, + "learning_rate": 4.5485380116959063e-05, + "loss": 3.0102, + "step": 15111 + }, + { + "epoch": 0.7953684210526316, + "grad_norm": 0.51953125, + "learning_rate": 4.547368421052632e-05, + "loss": 3.1456, + "step": 15112 + }, + { + "epoch": 0.7954210526315789, + "grad_norm": 0.51171875, + "learning_rate": 4.546198830409357e-05, + "loss": 3.4119, + "step": 15113 + }, + { + "epoch": 0.7954736842105263, + "grad_norm": 0.51171875, + "learning_rate": 4.545029239766082e-05, + "loss": 3.1502, + "step": 15114 + }, + { + "epoch": 0.7955263157894736, + "grad_norm": 0.50390625, + "learning_rate": 4.5438596491228075e-05, + "loss": 3.3281, + "step": 15115 + }, + { + "epoch": 0.795578947368421, + "grad_norm": 0.5078125, + "learning_rate": 4.5426900584795325e-05, + "loss": 3.5043, + "step": 15116 + }, + { + "epoch": 0.7956315789473685, + "grad_norm": 0.49609375, + "learning_rate": 4.5415204678362575e-05, + "loss": 3.6422, + "step": 15117 + }, + { + "epoch": 0.7956842105263158, + "grad_norm": 0.5234375, + "learning_rate": 4.5403508771929824e-05, + "loss": 3.6708, + "step": 15118 + }, + { + "epoch": 0.7957368421052632, + "grad_norm": 0.486328125, + "learning_rate": 4.539181286549708e-05, + "loss": 3.6493, + "step": 15119 + }, + { + "epoch": 0.7957894736842105, + "grad_norm": 0.4921875, + "learning_rate": 4.538011695906433e-05, + "loss": 3.3591, + "step": 15120 + }, + { + "epoch": 0.7958421052631579, + "grad_norm": 0.48828125, + "learning_rate": 4.536842105263158e-05, + "loss": 3.2953, + "step": 15121 + }, + { + "epoch": 0.7958947368421052, + "grad_norm": 0.51953125, + "learning_rate": 4.535672514619883e-05, + "loss": 3.3588, + "step": 15122 + }, + { + "epoch": 0.7959473684210526, + "grad_norm": 0.478515625, + "learning_rate": 4.5345029239766086e-05, + "loss": 3.4033, + "step": 15123 + }, + { + "epoch": 0.796, + "grad_norm": 0.48828125, + "learning_rate": 4.5333333333333335e-05, + "loss": 3.2591, + "step": 15124 + }, + { + "epoch": 0.7960526315789473, + "grad_norm": 0.51953125, + "learning_rate": 4.5321637426900585e-05, + "loss": 3.4367, + "step": 15125 + }, + { + "epoch": 0.7961052631578948, + "grad_norm": 0.5, + "learning_rate": 4.5309941520467834e-05, + "loss": 3.5176, + "step": 15126 + }, + { + "epoch": 0.7961578947368421, + "grad_norm": 0.486328125, + "learning_rate": 4.529824561403509e-05, + "loss": 3.4369, + "step": 15127 + }, + { + "epoch": 0.7962105263157895, + "grad_norm": 0.49609375, + "learning_rate": 4.528654970760234e-05, + "loss": 3.197, + "step": 15128 + }, + { + "epoch": 0.7962631578947369, + "grad_norm": 0.51171875, + "learning_rate": 4.527485380116959e-05, + "loss": 3.2547, + "step": 15129 + }, + { + "epoch": 0.7963157894736842, + "grad_norm": 0.5, + "learning_rate": 4.5263157894736846e-05, + "loss": 3.0466, + "step": 15130 + }, + { + "epoch": 0.7963684210526316, + "grad_norm": 0.47265625, + "learning_rate": 4.5251461988304096e-05, + "loss": 2.8708, + "step": 15131 + }, + { + "epoch": 0.7964210526315789, + "grad_norm": 0.49609375, + "learning_rate": 4.5239766081871346e-05, + "loss": 3.2338, + "step": 15132 + }, + { + "epoch": 0.7964736842105263, + "grad_norm": 0.47265625, + "learning_rate": 4.5228070175438595e-05, + "loss": 3.5795, + "step": 15133 + }, + { + "epoch": 0.7965263157894736, + "grad_norm": 0.52734375, + "learning_rate": 4.521637426900585e-05, + "loss": 3.5043, + "step": 15134 + }, + { + "epoch": 0.796578947368421, + "grad_norm": 0.490234375, + "learning_rate": 4.52046783625731e-05, + "loss": 3.2174, + "step": 15135 + }, + { + "epoch": 0.7966315789473685, + "grad_norm": 0.53125, + "learning_rate": 4.519298245614035e-05, + "loss": 3.1197, + "step": 15136 + }, + { + "epoch": 0.7966842105263158, + "grad_norm": 0.5078125, + "learning_rate": 4.51812865497076e-05, + "loss": 3.1161, + "step": 15137 + }, + { + "epoch": 0.7967368421052632, + "grad_norm": 0.4921875, + "learning_rate": 4.516959064327486e-05, + "loss": 3.435, + "step": 15138 + }, + { + "epoch": 0.7967894736842105, + "grad_norm": 0.5, + "learning_rate": 4.515789473684211e-05, + "loss": 3.2593, + "step": 15139 + }, + { + "epoch": 0.7968421052631579, + "grad_norm": 0.498046875, + "learning_rate": 4.5146198830409356e-05, + "loss": 3.1171, + "step": 15140 + }, + { + "epoch": 0.7968947368421052, + "grad_norm": 0.48046875, + "learning_rate": 4.513450292397661e-05, + "loss": 3.5475, + "step": 15141 + }, + { + "epoch": 0.7969473684210526, + "grad_norm": 0.49609375, + "learning_rate": 4.512280701754386e-05, + "loss": 3.148, + "step": 15142 + }, + { + "epoch": 0.797, + "grad_norm": 0.482421875, + "learning_rate": 4.511111111111112e-05, + "loss": 3.4871, + "step": 15143 + }, + { + "epoch": 0.7970526315789473, + "grad_norm": 0.47265625, + "learning_rate": 4.509941520467836e-05, + "loss": 3.1642, + "step": 15144 + }, + { + "epoch": 0.7971052631578948, + "grad_norm": 0.494140625, + "learning_rate": 4.508771929824562e-05, + "loss": 2.9723, + "step": 15145 + }, + { + "epoch": 0.7971578947368421, + "grad_norm": 0.51171875, + "learning_rate": 4.507602339181287e-05, + "loss": 3.5181, + "step": 15146 + }, + { + "epoch": 0.7972105263157895, + "grad_norm": 0.470703125, + "learning_rate": 4.506432748538012e-05, + "loss": 3.3446, + "step": 15147 + }, + { + "epoch": 0.7972631578947369, + "grad_norm": 0.478515625, + "learning_rate": 4.5052631578947366e-05, + "loss": 3.2836, + "step": 15148 + }, + { + "epoch": 0.7973157894736842, + "grad_norm": 0.494140625, + "learning_rate": 4.504093567251462e-05, + "loss": 3.3958, + "step": 15149 + }, + { + "epoch": 0.7973684210526316, + "grad_norm": 0.466796875, + "learning_rate": 4.502923976608187e-05, + "loss": 3.2395, + "step": 15150 + }, + { + "epoch": 0.7974210526315789, + "grad_norm": 0.49609375, + "learning_rate": 4.501754385964912e-05, + "loss": 3.5972, + "step": 15151 + }, + { + "epoch": 0.7974736842105263, + "grad_norm": 0.515625, + "learning_rate": 4.500584795321638e-05, + "loss": 3.3662, + "step": 15152 + }, + { + "epoch": 0.7975263157894736, + "grad_norm": 0.515625, + "learning_rate": 4.499415204678363e-05, + "loss": 3.0402, + "step": 15153 + }, + { + "epoch": 0.797578947368421, + "grad_norm": 0.546875, + "learning_rate": 4.4982456140350884e-05, + "loss": 3.5624, + "step": 15154 + }, + { + "epoch": 0.7976315789473685, + "grad_norm": 0.48828125, + "learning_rate": 4.497076023391813e-05, + "loss": 2.65, + "step": 15155 + }, + { + "epoch": 0.7976842105263158, + "grad_norm": 0.5, + "learning_rate": 4.495906432748538e-05, + "loss": 3.4454, + "step": 15156 + }, + { + "epoch": 0.7977368421052632, + "grad_norm": 0.490234375, + "learning_rate": 4.494736842105263e-05, + "loss": 3.2035, + "step": 15157 + }, + { + "epoch": 0.7977894736842105, + "grad_norm": 0.5234375, + "learning_rate": 4.493567251461989e-05, + "loss": 3.232, + "step": 15158 + }, + { + "epoch": 0.7978421052631579, + "grad_norm": 0.52734375, + "learning_rate": 4.492397660818713e-05, + "loss": 3.3451, + "step": 15159 + }, + { + "epoch": 0.7978947368421052, + "grad_norm": 0.48046875, + "learning_rate": 4.491228070175439e-05, + "loss": 2.7292, + "step": 15160 + }, + { + "epoch": 0.7979473684210526, + "grad_norm": 0.53515625, + "learning_rate": 4.490058479532164e-05, + "loss": 3.4523, + "step": 15161 + }, + { + "epoch": 0.798, + "grad_norm": 0.5078125, + "learning_rate": 4.4888888888888894e-05, + "loss": 3.1569, + "step": 15162 + }, + { + "epoch": 0.7980526315789473, + "grad_norm": 0.48046875, + "learning_rate": 4.487719298245614e-05, + "loss": 2.9191, + "step": 15163 + }, + { + "epoch": 0.7981052631578948, + "grad_norm": 0.48046875, + "learning_rate": 4.4865497076023393e-05, + "loss": 3.2238, + "step": 15164 + }, + { + "epoch": 0.7981578947368421, + "grad_norm": 0.52734375, + "learning_rate": 4.485380116959065e-05, + "loss": 2.9926, + "step": 15165 + }, + { + "epoch": 0.7982105263157895, + "grad_norm": 0.49609375, + "learning_rate": 4.48421052631579e-05, + "loss": 3.3339, + "step": 15166 + }, + { + "epoch": 0.7982631578947369, + "grad_norm": 0.54296875, + "learning_rate": 4.483040935672515e-05, + "loss": 3.3803, + "step": 15167 + }, + { + "epoch": 0.7983157894736842, + "grad_norm": 0.546875, + "learning_rate": 4.48187134502924e-05, + "loss": 3.06, + "step": 15168 + }, + { + "epoch": 0.7983684210526316, + "grad_norm": 0.490234375, + "learning_rate": 4.4807017543859655e-05, + "loss": 3.2155, + "step": 15169 + }, + { + "epoch": 0.7984210526315789, + "grad_norm": 0.5, + "learning_rate": 4.4795321637426905e-05, + "loss": 2.97, + "step": 15170 + }, + { + "epoch": 0.7984736842105263, + "grad_norm": 0.498046875, + "learning_rate": 4.4783625730994154e-05, + "loss": 3.3204, + "step": 15171 + }, + { + "epoch": 0.7985263157894736, + "grad_norm": 0.5234375, + "learning_rate": 4.4771929824561404e-05, + "loss": 3.195, + "step": 15172 + }, + { + "epoch": 0.798578947368421, + "grad_norm": 0.486328125, + "learning_rate": 4.476023391812866e-05, + "loss": 2.8722, + "step": 15173 + }, + { + "epoch": 0.7986315789473685, + "grad_norm": 0.498046875, + "learning_rate": 4.47485380116959e-05, + "loss": 3.4965, + "step": 15174 + }, + { + "epoch": 0.7986842105263158, + "grad_norm": 0.4921875, + "learning_rate": 4.473684210526316e-05, + "loss": 3.3707, + "step": 15175 + }, + { + "epoch": 0.7987368421052632, + "grad_norm": 0.5, + "learning_rate": 4.472514619883041e-05, + "loss": 3.4464, + "step": 15176 + }, + { + "epoch": 0.7987894736842105, + "grad_norm": 0.58203125, + "learning_rate": 4.4713450292397665e-05, + "loss": 3.448, + "step": 15177 + }, + { + "epoch": 0.7988421052631579, + "grad_norm": 0.484375, + "learning_rate": 4.4701754385964915e-05, + "loss": 3.0976, + "step": 15178 + }, + { + "epoch": 0.7988947368421052, + "grad_norm": 0.4921875, + "learning_rate": 4.4690058479532164e-05, + "loss": 2.8804, + "step": 15179 + }, + { + "epoch": 0.7989473684210526, + "grad_norm": 0.498046875, + "learning_rate": 4.467836257309942e-05, + "loss": 3.7291, + "step": 15180 + }, + { + "epoch": 0.799, + "grad_norm": 0.515625, + "learning_rate": 4.466666666666667e-05, + "loss": 3.3283, + "step": 15181 + }, + { + "epoch": 0.7990526315789473, + "grad_norm": 0.484375, + "learning_rate": 4.465497076023392e-05, + "loss": 3.2208, + "step": 15182 + }, + { + "epoch": 0.7991052631578948, + "grad_norm": 0.484375, + "learning_rate": 4.464327485380117e-05, + "loss": 2.8033, + "step": 15183 + }, + { + "epoch": 0.7991578947368421, + "grad_norm": 0.5078125, + "learning_rate": 4.4631578947368426e-05, + "loss": 3.4178, + "step": 15184 + }, + { + "epoch": 0.7992105263157895, + "grad_norm": 0.50390625, + "learning_rate": 4.4619883040935676e-05, + "loss": 3.207, + "step": 15185 + }, + { + "epoch": 0.7992631578947369, + "grad_norm": 0.486328125, + "learning_rate": 4.4608187134502925e-05, + "loss": 2.9864, + "step": 15186 + }, + { + "epoch": 0.7993157894736842, + "grad_norm": 0.5078125, + "learning_rate": 4.4596491228070175e-05, + "loss": 3.4394, + "step": 15187 + }, + { + "epoch": 0.7993684210526316, + "grad_norm": 0.515625, + "learning_rate": 4.458479532163743e-05, + "loss": 3.3754, + "step": 15188 + }, + { + "epoch": 0.7994210526315789, + "grad_norm": 0.5390625, + "learning_rate": 4.457309941520468e-05, + "loss": 3.3633, + "step": 15189 + }, + { + "epoch": 0.7994736842105263, + "grad_norm": 0.478515625, + "learning_rate": 4.456140350877193e-05, + "loss": 2.9536, + "step": 15190 + }, + { + "epoch": 0.7995263157894736, + "grad_norm": 0.49609375, + "learning_rate": 4.454970760233919e-05, + "loss": 3.3007, + "step": 15191 + }, + { + "epoch": 0.7995789473684211, + "grad_norm": 0.49609375, + "learning_rate": 4.4538011695906436e-05, + "loss": 3.0303, + "step": 15192 + }, + { + "epoch": 0.7996315789473685, + "grad_norm": 0.4765625, + "learning_rate": 4.4526315789473686e-05, + "loss": 3.0745, + "step": 15193 + }, + { + "epoch": 0.7996842105263158, + "grad_norm": 0.51953125, + "learning_rate": 4.4514619883040935e-05, + "loss": 3.5427, + "step": 15194 + }, + { + "epoch": 0.7997368421052632, + "grad_norm": 0.546875, + "learning_rate": 4.450292397660819e-05, + "loss": 3.1335, + "step": 15195 + }, + { + "epoch": 0.7997894736842105, + "grad_norm": 0.52734375, + "learning_rate": 4.449122807017544e-05, + "loss": 2.9849, + "step": 15196 + }, + { + "epoch": 0.7998421052631579, + "grad_norm": 0.494140625, + "learning_rate": 4.447953216374269e-05, + "loss": 3.09, + "step": 15197 + }, + { + "epoch": 0.7998947368421052, + "grad_norm": 0.498046875, + "learning_rate": 4.446783625730994e-05, + "loss": 3.3468, + "step": 15198 + }, + { + "epoch": 0.7999473684210526, + "grad_norm": 0.494140625, + "learning_rate": 4.44561403508772e-05, + "loss": 3.1561, + "step": 15199 + }, + { + "epoch": 0.8, + "grad_norm": 0.5, + "learning_rate": 4.4444444444444447e-05, + "loss": 3.0352, + "step": 15200 + }, + { + "epoch": 0.8000526315789473, + "grad_norm": 0.47265625, + "learning_rate": 4.4432748538011696e-05, + "loss": 2.7042, + "step": 15201 + }, + { + "epoch": 0.8001052631578948, + "grad_norm": 0.51953125, + "learning_rate": 4.442105263157895e-05, + "loss": 3.6098, + "step": 15202 + }, + { + "epoch": 0.8001578947368421, + "grad_norm": 0.48828125, + "learning_rate": 4.44093567251462e-05, + "loss": 3.3085, + "step": 15203 + }, + { + "epoch": 0.8002105263157895, + "grad_norm": 0.48828125, + "learning_rate": 4.439766081871345e-05, + "loss": 3.0108, + "step": 15204 + }, + { + "epoch": 0.8002631578947368, + "grad_norm": 0.51171875, + "learning_rate": 4.43859649122807e-05, + "loss": 3.4514, + "step": 15205 + }, + { + "epoch": 0.8003157894736842, + "grad_norm": 0.50390625, + "learning_rate": 4.437426900584796e-05, + "loss": 3.5839, + "step": 15206 + }, + { + "epoch": 0.8003684210526316, + "grad_norm": 0.474609375, + "learning_rate": 4.436257309941521e-05, + "loss": 3.1821, + "step": 15207 + }, + { + "epoch": 0.8004210526315789, + "grad_norm": 0.482421875, + "learning_rate": 4.435087719298246e-05, + "loss": 2.9036, + "step": 15208 + }, + { + "epoch": 0.8004736842105263, + "grad_norm": 0.490234375, + "learning_rate": 4.4339181286549706e-05, + "loss": 3.3777, + "step": 15209 + }, + { + "epoch": 0.8005263157894736, + "grad_norm": 0.51171875, + "learning_rate": 4.432748538011696e-05, + "loss": 3.313, + "step": 15210 + }, + { + "epoch": 0.8005789473684211, + "grad_norm": 0.48046875, + "learning_rate": 4.431578947368421e-05, + "loss": 3.5289, + "step": 15211 + }, + { + "epoch": 0.8006315789473685, + "grad_norm": 0.484375, + "learning_rate": 4.430409356725146e-05, + "loss": 3.2104, + "step": 15212 + }, + { + "epoch": 0.8006842105263158, + "grad_norm": 0.515625, + "learning_rate": 4.429239766081871e-05, + "loss": 3.0287, + "step": 15213 + }, + { + "epoch": 0.8007368421052632, + "grad_norm": 0.5, + "learning_rate": 4.428070175438597e-05, + "loss": 3.4382, + "step": 15214 + }, + { + "epoch": 0.8007894736842105, + "grad_norm": 0.490234375, + "learning_rate": 4.4269005847953224e-05, + "loss": 3.5557, + "step": 15215 + }, + { + "epoch": 0.8008421052631579, + "grad_norm": 0.486328125, + "learning_rate": 4.425730994152047e-05, + "loss": 3.5298, + "step": 15216 + }, + { + "epoch": 0.8008947368421052, + "grad_norm": 0.494140625, + "learning_rate": 4.4245614035087724e-05, + "loss": 3.3158, + "step": 15217 + }, + { + "epoch": 0.8009473684210526, + "grad_norm": 0.484375, + "learning_rate": 4.423391812865497e-05, + "loss": 3.1092, + "step": 15218 + }, + { + "epoch": 0.801, + "grad_norm": 0.50390625, + "learning_rate": 4.422222222222222e-05, + "loss": 3.5622, + "step": 15219 + }, + { + "epoch": 0.8010526315789473, + "grad_norm": 0.4765625, + "learning_rate": 4.421052631578947e-05, + "loss": 3.397, + "step": 15220 + }, + { + "epoch": 0.8011052631578948, + "grad_norm": 0.60546875, + "learning_rate": 4.419883040935673e-05, + "loss": 2.696, + "step": 15221 + }, + { + "epoch": 0.8011578947368421, + "grad_norm": 0.5078125, + "learning_rate": 4.418713450292398e-05, + "loss": 3.1179, + "step": 15222 + }, + { + "epoch": 0.8012105263157895, + "grad_norm": 0.58203125, + "learning_rate": 4.417543859649123e-05, + "loss": 3.2987, + "step": 15223 + }, + { + "epoch": 0.8012631578947368, + "grad_norm": 0.7734375, + "learning_rate": 4.416374269005848e-05, + "loss": 3.5702, + "step": 15224 + }, + { + "epoch": 0.8013157894736842, + "grad_norm": 0.51171875, + "learning_rate": 4.4152046783625734e-05, + "loss": 3.2236, + "step": 15225 + }, + { + "epoch": 0.8013684210526316, + "grad_norm": 0.478515625, + "learning_rate": 4.4140350877192983e-05, + "loss": 3.1021, + "step": 15226 + }, + { + "epoch": 0.8014210526315789, + "grad_norm": 0.55078125, + "learning_rate": 4.412865497076023e-05, + "loss": 3.3527, + "step": 15227 + }, + { + "epoch": 0.8014736842105263, + "grad_norm": 0.5390625, + "learning_rate": 4.411695906432749e-05, + "loss": 3.5294, + "step": 15228 + }, + { + "epoch": 0.8015263157894736, + "grad_norm": 0.515625, + "learning_rate": 4.410526315789474e-05, + "loss": 3.3159, + "step": 15229 + }, + { + "epoch": 0.8015789473684211, + "grad_norm": 0.47265625, + "learning_rate": 4.4093567251461995e-05, + "loss": 3.4122, + "step": 15230 + }, + { + "epoch": 0.8016315789473685, + "grad_norm": 0.52734375, + "learning_rate": 4.408187134502924e-05, + "loss": 3.2846, + "step": 15231 + }, + { + "epoch": 0.8016842105263158, + "grad_norm": 0.5234375, + "learning_rate": 4.4070175438596495e-05, + "loss": 3.232, + "step": 15232 + }, + { + "epoch": 0.8017368421052632, + "grad_norm": 0.51171875, + "learning_rate": 4.4058479532163744e-05, + "loss": 3.4739, + "step": 15233 + }, + { + "epoch": 0.8017894736842105, + "grad_norm": 0.490234375, + "learning_rate": 4.4046783625731e-05, + "loss": 3.6677, + "step": 15234 + }, + { + "epoch": 0.8018421052631579, + "grad_norm": 0.486328125, + "learning_rate": 4.403508771929824e-05, + "loss": 3.4444, + "step": 15235 + }, + { + "epoch": 0.8018947368421052, + "grad_norm": 0.478515625, + "learning_rate": 4.40233918128655e-05, + "loss": 3.113, + "step": 15236 + }, + { + "epoch": 0.8019473684210526, + "grad_norm": 0.494140625, + "learning_rate": 4.401169590643275e-05, + "loss": 3.1411, + "step": 15237 + }, + { + "epoch": 0.802, + "grad_norm": 0.48828125, + "learning_rate": 4.4000000000000006e-05, + "loss": 3.5071, + "step": 15238 + }, + { + "epoch": 0.8020526315789474, + "grad_norm": 0.4921875, + "learning_rate": 4.398830409356725e-05, + "loss": 2.9106, + "step": 15239 + }, + { + "epoch": 0.8021052631578948, + "grad_norm": 0.490234375, + "learning_rate": 4.3976608187134505e-05, + "loss": 3.5168, + "step": 15240 + }, + { + "epoch": 0.8021578947368421, + "grad_norm": 0.7109375, + "learning_rate": 4.396491228070176e-05, + "loss": 3.3483, + "step": 15241 + }, + { + "epoch": 0.8022105263157895, + "grad_norm": 0.46875, + "learning_rate": 4.3953216374269004e-05, + "loss": 3.1354, + "step": 15242 + }, + { + "epoch": 0.8022631578947368, + "grad_norm": 0.52734375, + "learning_rate": 4.394152046783626e-05, + "loss": 2.665, + "step": 15243 + }, + { + "epoch": 0.8023157894736842, + "grad_norm": 0.490234375, + "learning_rate": 4.392982456140351e-05, + "loss": 3.1244, + "step": 15244 + }, + { + "epoch": 0.8023684210526316, + "grad_norm": 0.51171875, + "learning_rate": 4.3918128654970766e-05, + "loss": 3.2824, + "step": 15245 + }, + { + "epoch": 0.8024210526315789, + "grad_norm": 0.49609375, + "learning_rate": 4.390643274853801e-05, + "loss": 3.3915, + "step": 15246 + }, + { + "epoch": 0.8024736842105263, + "grad_norm": 0.48046875, + "learning_rate": 4.3894736842105266e-05, + "loss": 3.387, + "step": 15247 + }, + { + "epoch": 0.8025263157894736, + "grad_norm": 0.498046875, + "learning_rate": 4.3883040935672515e-05, + "loss": 3.1987, + "step": 15248 + }, + { + "epoch": 0.8025789473684211, + "grad_norm": 0.48046875, + "learning_rate": 4.387134502923977e-05, + "loss": 3.216, + "step": 15249 + }, + { + "epoch": 0.8026315789473685, + "grad_norm": 0.5078125, + "learning_rate": 4.3859649122807014e-05, + "loss": 3.1043, + "step": 15250 + }, + { + "epoch": 0.8026842105263158, + "grad_norm": 0.4921875, + "learning_rate": 4.384795321637427e-05, + "loss": 3.5569, + "step": 15251 + }, + { + "epoch": 0.8027368421052632, + "grad_norm": 0.484375, + "learning_rate": 4.383625730994153e-05, + "loss": 3.6924, + "step": 15252 + }, + { + "epoch": 0.8027894736842105, + "grad_norm": 0.48046875, + "learning_rate": 4.382456140350878e-05, + "loss": 2.7662, + "step": 15253 + }, + { + "epoch": 0.8028421052631579, + "grad_norm": 0.498046875, + "learning_rate": 4.3812865497076026e-05, + "loss": 2.9478, + "step": 15254 + }, + { + "epoch": 0.8028947368421052, + "grad_norm": 0.46875, + "learning_rate": 4.3801169590643276e-05, + "loss": 2.9745, + "step": 15255 + }, + { + "epoch": 0.8029473684210526, + "grad_norm": 0.515625, + "learning_rate": 4.378947368421053e-05, + "loss": 3.1273, + "step": 15256 + }, + { + "epoch": 0.803, + "grad_norm": 0.4921875, + "learning_rate": 4.377777777777778e-05, + "loss": 3.3366, + "step": 15257 + }, + { + "epoch": 0.8030526315789474, + "grad_norm": 0.482421875, + "learning_rate": 4.376608187134503e-05, + "loss": 3.2418, + "step": 15258 + }, + { + "epoch": 0.8031052631578948, + "grad_norm": 0.50390625, + "learning_rate": 4.375438596491228e-05, + "loss": 3.0113, + "step": 15259 + }, + { + "epoch": 0.8031578947368421, + "grad_norm": 0.50390625, + "learning_rate": 4.374269005847954e-05, + "loss": 3.5645, + "step": 15260 + }, + { + "epoch": 0.8032105263157895, + "grad_norm": 0.4921875, + "learning_rate": 4.373099415204679e-05, + "loss": 3.2444, + "step": 15261 + }, + { + "epoch": 0.8032631578947368, + "grad_norm": 0.484375, + "learning_rate": 4.3719298245614037e-05, + "loss": 3.4144, + "step": 15262 + }, + { + "epoch": 0.8033157894736842, + "grad_norm": 0.53125, + "learning_rate": 4.3707602339181286e-05, + "loss": 3.1418, + "step": 15263 + }, + { + "epoch": 0.8033684210526316, + "grad_norm": 0.4921875, + "learning_rate": 4.369590643274854e-05, + "loss": 3.8912, + "step": 15264 + }, + { + "epoch": 0.8034210526315789, + "grad_norm": 0.546875, + "learning_rate": 4.368421052631579e-05, + "loss": 3.0722, + "step": 15265 + }, + { + "epoch": 0.8034736842105263, + "grad_norm": 0.55078125, + "learning_rate": 4.367251461988304e-05, + "loss": 2.892, + "step": 15266 + }, + { + "epoch": 0.8035263157894736, + "grad_norm": 0.494140625, + "learning_rate": 4.36608187134503e-05, + "loss": 3.4049, + "step": 15267 + }, + { + "epoch": 0.8035789473684211, + "grad_norm": 0.494140625, + "learning_rate": 4.364912280701755e-05, + "loss": 3.1415, + "step": 15268 + }, + { + "epoch": 0.8036315789473685, + "grad_norm": 0.47265625, + "learning_rate": 4.36374269005848e-05, + "loss": 3.6301, + "step": 15269 + }, + { + "epoch": 0.8036842105263158, + "grad_norm": 0.50390625, + "learning_rate": 4.362573099415205e-05, + "loss": 3.5188, + "step": 15270 + }, + { + "epoch": 0.8037368421052632, + "grad_norm": 0.482421875, + "learning_rate": 4.36140350877193e-05, + "loss": 3.461, + "step": 15271 + }, + { + "epoch": 0.8037894736842105, + "grad_norm": 0.515625, + "learning_rate": 4.360233918128655e-05, + "loss": 3.2108, + "step": 15272 + }, + { + "epoch": 0.8038421052631579, + "grad_norm": 0.4921875, + "learning_rate": 4.35906432748538e-05, + "loss": 3.6396, + "step": 15273 + }, + { + "epoch": 0.8038947368421052, + "grad_norm": 0.515625, + "learning_rate": 4.357894736842105e-05, + "loss": 3.3675, + "step": 15274 + }, + { + "epoch": 0.8039473684210526, + "grad_norm": 0.484375, + "learning_rate": 4.356725146198831e-05, + "loss": 2.8632, + "step": 15275 + }, + { + "epoch": 0.804, + "grad_norm": 0.50390625, + "learning_rate": 4.355555555555556e-05, + "loss": 3.3168, + "step": 15276 + }, + { + "epoch": 0.8040526315789474, + "grad_norm": 0.490234375, + "learning_rate": 4.354385964912281e-05, + "loss": 3.4704, + "step": 15277 + }, + { + "epoch": 0.8041052631578948, + "grad_norm": 0.50390625, + "learning_rate": 4.3532163742690064e-05, + "loss": 3.2828, + "step": 15278 + }, + { + "epoch": 0.8041578947368421, + "grad_norm": 0.482421875, + "learning_rate": 4.3520467836257313e-05, + "loss": 3.072, + "step": 15279 + }, + { + "epoch": 0.8042105263157895, + "grad_norm": 0.478515625, + "learning_rate": 4.350877192982456e-05, + "loss": 3.4263, + "step": 15280 + }, + { + "epoch": 0.8042631578947368, + "grad_norm": 0.4921875, + "learning_rate": 4.349707602339181e-05, + "loss": 2.9586, + "step": 15281 + }, + { + "epoch": 0.8043157894736842, + "grad_norm": 0.49609375, + "learning_rate": 4.348538011695907e-05, + "loss": 3.1038, + "step": 15282 + }, + { + "epoch": 0.8043684210526316, + "grad_norm": 0.5078125, + "learning_rate": 4.347368421052632e-05, + "loss": 3.0864, + "step": 15283 + }, + { + "epoch": 0.8044210526315789, + "grad_norm": 0.53125, + "learning_rate": 4.346198830409357e-05, + "loss": 3.3484, + "step": 15284 + }, + { + "epoch": 0.8044736842105263, + "grad_norm": 0.478515625, + "learning_rate": 4.345029239766082e-05, + "loss": 3.1435, + "step": 15285 + }, + { + "epoch": 0.8045263157894736, + "grad_norm": 0.498046875, + "learning_rate": 4.3438596491228074e-05, + "loss": 3.0834, + "step": 15286 + }, + { + "epoch": 0.8045789473684211, + "grad_norm": 0.61328125, + "learning_rate": 4.3426900584795324e-05, + "loss": 3.4009, + "step": 15287 + }, + { + "epoch": 0.8046315789473685, + "grad_norm": 0.515625, + "learning_rate": 4.341520467836257e-05, + "loss": 3.6534, + "step": 15288 + }, + { + "epoch": 0.8046842105263158, + "grad_norm": 0.58984375, + "learning_rate": 4.340350877192982e-05, + "loss": 2.7123, + "step": 15289 + }, + { + "epoch": 0.8047368421052632, + "grad_norm": 0.5546875, + "learning_rate": 4.339181286549708e-05, + "loss": 3.3522, + "step": 15290 + }, + { + "epoch": 0.8047894736842105, + "grad_norm": 0.5078125, + "learning_rate": 4.338011695906433e-05, + "loss": 3.1224, + "step": 15291 + }, + { + "epoch": 0.8048421052631579, + "grad_norm": 0.5, + "learning_rate": 4.336842105263158e-05, + "loss": 2.8792, + "step": 15292 + }, + { + "epoch": 0.8048947368421052, + "grad_norm": 0.47265625, + "learning_rate": 4.3356725146198835e-05, + "loss": 3.2329, + "step": 15293 + }, + { + "epoch": 0.8049473684210526, + "grad_norm": 0.48828125, + "learning_rate": 4.3345029239766084e-05, + "loss": 3.4008, + "step": 15294 + }, + { + "epoch": 0.805, + "grad_norm": 0.51953125, + "learning_rate": 4.3333333333333334e-05, + "loss": 3.2723, + "step": 15295 + }, + { + "epoch": 0.8050526315789474, + "grad_norm": 0.486328125, + "learning_rate": 4.3321637426900584e-05, + "loss": 3.0831, + "step": 15296 + }, + { + "epoch": 0.8051052631578948, + "grad_norm": 0.490234375, + "learning_rate": 4.330994152046784e-05, + "loss": 3.4148, + "step": 15297 + }, + { + "epoch": 0.8051578947368421, + "grad_norm": 0.546875, + "learning_rate": 4.329824561403509e-05, + "loss": 3.1582, + "step": 15298 + }, + { + "epoch": 0.8052105263157895, + "grad_norm": 0.484375, + "learning_rate": 4.328654970760234e-05, + "loss": 3.5279, + "step": 15299 + }, + { + "epoch": 0.8052631578947368, + "grad_norm": 0.48046875, + "learning_rate": 4.327485380116959e-05, + "loss": 3.1382, + "step": 15300 + }, + { + "epoch": 0.8053157894736842, + "grad_norm": 0.46875, + "learning_rate": 4.3263157894736845e-05, + "loss": 2.8118, + "step": 15301 + }, + { + "epoch": 0.8053684210526316, + "grad_norm": 0.5078125, + "learning_rate": 4.32514619883041e-05, + "loss": 3.0657, + "step": 15302 + }, + { + "epoch": 0.8054210526315789, + "grad_norm": 0.51953125, + "learning_rate": 4.3239766081871344e-05, + "loss": 3.7671, + "step": 15303 + }, + { + "epoch": 0.8054736842105263, + "grad_norm": 0.478515625, + "learning_rate": 4.32280701754386e-05, + "loss": 3.3959, + "step": 15304 + }, + { + "epoch": 0.8055263157894736, + "grad_norm": 0.51171875, + "learning_rate": 4.321637426900585e-05, + "loss": 3.3585, + "step": 15305 + }, + { + "epoch": 0.8055789473684211, + "grad_norm": 0.498046875, + "learning_rate": 4.320467836257311e-05, + "loss": 3.2587, + "step": 15306 + }, + { + "epoch": 0.8056315789473685, + "grad_norm": 0.48828125, + "learning_rate": 4.319298245614035e-05, + "loss": 3.0436, + "step": 15307 + }, + { + "epoch": 0.8056842105263158, + "grad_norm": 0.482421875, + "learning_rate": 4.3181286549707606e-05, + "loss": 3.0508, + "step": 15308 + }, + { + "epoch": 0.8057368421052632, + "grad_norm": 0.486328125, + "learning_rate": 4.3169590643274855e-05, + "loss": 3.1846, + "step": 15309 + }, + { + "epoch": 0.8057894736842105, + "grad_norm": 0.494140625, + "learning_rate": 4.3157894736842105e-05, + "loss": 3.044, + "step": 15310 + }, + { + "epoch": 0.8058421052631579, + "grad_norm": 0.48828125, + "learning_rate": 4.3146198830409355e-05, + "loss": 3.2815, + "step": 15311 + }, + { + "epoch": 0.8058947368421052, + "grad_norm": 0.48828125, + "learning_rate": 4.313450292397661e-05, + "loss": 3.7976, + "step": 15312 + }, + { + "epoch": 0.8059473684210526, + "grad_norm": 0.51171875, + "learning_rate": 4.312280701754386e-05, + "loss": 3.4811, + "step": 15313 + }, + { + "epoch": 0.806, + "grad_norm": 0.5078125, + "learning_rate": 4.311111111111111e-05, + "loss": 3.5444, + "step": 15314 + }, + { + "epoch": 0.8060526315789474, + "grad_norm": 0.494140625, + "learning_rate": 4.3099415204678367e-05, + "loss": 2.7406, + "step": 15315 + }, + { + "epoch": 0.8061052631578948, + "grad_norm": 0.484375, + "learning_rate": 4.3087719298245616e-05, + "loss": 3.5212, + "step": 15316 + }, + { + "epoch": 0.8061578947368421, + "grad_norm": 0.515625, + "learning_rate": 4.307602339181287e-05, + "loss": 3.2928, + "step": 15317 + }, + { + "epoch": 0.8062105263157895, + "grad_norm": 0.5234375, + "learning_rate": 4.3064327485380115e-05, + "loss": 3.3583, + "step": 15318 + }, + { + "epoch": 0.8062631578947368, + "grad_norm": 0.51171875, + "learning_rate": 4.305263157894737e-05, + "loss": 3.4701, + "step": 15319 + }, + { + "epoch": 0.8063157894736842, + "grad_norm": 0.51953125, + "learning_rate": 4.304093567251462e-05, + "loss": 3.1391, + "step": 15320 + }, + { + "epoch": 0.8063684210526316, + "grad_norm": 0.466796875, + "learning_rate": 4.302923976608188e-05, + "loss": 2.9362, + "step": 15321 + }, + { + "epoch": 0.8064210526315789, + "grad_norm": 0.51953125, + "learning_rate": 4.301754385964912e-05, + "loss": 2.8526, + "step": 15322 + }, + { + "epoch": 0.8064736842105263, + "grad_norm": 0.52734375, + "learning_rate": 4.300584795321638e-05, + "loss": 3.5402, + "step": 15323 + }, + { + "epoch": 0.8065263157894736, + "grad_norm": 0.4765625, + "learning_rate": 4.2994152046783626e-05, + "loss": 3.2604, + "step": 15324 + }, + { + "epoch": 0.8065789473684211, + "grad_norm": 0.4921875, + "learning_rate": 4.298245614035088e-05, + "loss": 3.148, + "step": 15325 + }, + { + "epoch": 0.8066315789473684, + "grad_norm": 0.53515625, + "learning_rate": 4.2970760233918126e-05, + "loss": 3.0895, + "step": 15326 + }, + { + "epoch": 0.8066842105263158, + "grad_norm": 0.490234375, + "learning_rate": 4.295906432748538e-05, + "loss": 2.8686, + "step": 15327 + }, + { + "epoch": 0.8067368421052632, + "grad_norm": 0.48828125, + "learning_rate": 4.294736842105264e-05, + "loss": 3.3051, + "step": 15328 + }, + { + "epoch": 0.8067894736842105, + "grad_norm": 0.5, + "learning_rate": 4.293567251461989e-05, + "loss": 3.6747, + "step": 15329 + }, + { + "epoch": 0.8068421052631579, + "grad_norm": 0.51953125, + "learning_rate": 4.292397660818714e-05, + "loss": 3.4357, + "step": 15330 + }, + { + "epoch": 0.8068947368421052, + "grad_norm": 0.4765625, + "learning_rate": 4.291228070175439e-05, + "loss": 3.0918, + "step": 15331 + }, + { + "epoch": 0.8069473684210526, + "grad_norm": 0.478515625, + "learning_rate": 4.2900584795321644e-05, + "loss": 3.3346, + "step": 15332 + }, + { + "epoch": 0.807, + "grad_norm": 0.470703125, + "learning_rate": 4.2888888888888886e-05, + "loss": 3.1253, + "step": 15333 + }, + { + "epoch": 0.8070526315789474, + "grad_norm": 0.4765625, + "learning_rate": 4.287719298245614e-05, + "loss": 3.1885, + "step": 15334 + }, + { + "epoch": 0.8071052631578948, + "grad_norm": 0.55078125, + "learning_rate": 4.286549707602339e-05, + "loss": 3.0195, + "step": 15335 + }, + { + "epoch": 0.8071578947368421, + "grad_norm": 0.486328125, + "learning_rate": 4.285380116959065e-05, + "loss": 3.2534, + "step": 15336 + }, + { + "epoch": 0.8072105263157895, + "grad_norm": 0.498046875, + "learning_rate": 4.284210526315789e-05, + "loss": 3.443, + "step": 15337 + }, + { + "epoch": 0.8072631578947368, + "grad_norm": 0.48046875, + "learning_rate": 4.283040935672515e-05, + "loss": 3.3027, + "step": 15338 + }, + { + "epoch": 0.8073157894736842, + "grad_norm": 0.490234375, + "learning_rate": 4.28187134502924e-05, + "loss": 3.1417, + "step": 15339 + }, + { + "epoch": 0.8073684210526316, + "grad_norm": 0.486328125, + "learning_rate": 4.2807017543859654e-05, + "loss": 3.1226, + "step": 15340 + }, + { + "epoch": 0.8074210526315789, + "grad_norm": 0.48046875, + "learning_rate": 4.2795321637426903e-05, + "loss": 3.2336, + "step": 15341 + }, + { + "epoch": 0.8074736842105263, + "grad_norm": 0.494140625, + "learning_rate": 4.278362573099415e-05, + "loss": 3.3772, + "step": 15342 + }, + { + "epoch": 0.8075263157894736, + "grad_norm": 0.498046875, + "learning_rate": 4.277192982456141e-05, + "loss": 3.2098, + "step": 15343 + }, + { + "epoch": 0.8075789473684211, + "grad_norm": 0.51953125, + "learning_rate": 4.276023391812866e-05, + "loss": 2.8078, + "step": 15344 + }, + { + "epoch": 0.8076315789473684, + "grad_norm": 0.51171875, + "learning_rate": 4.274853801169591e-05, + "loss": 3.6981, + "step": 15345 + }, + { + "epoch": 0.8076842105263158, + "grad_norm": 0.5234375, + "learning_rate": 4.273684210526316e-05, + "loss": 2.8803, + "step": 15346 + }, + { + "epoch": 0.8077368421052632, + "grad_norm": 0.50390625, + "learning_rate": 4.2725146198830415e-05, + "loss": 3.7448, + "step": 15347 + }, + { + "epoch": 0.8077894736842105, + "grad_norm": 0.486328125, + "learning_rate": 4.2713450292397664e-05, + "loss": 3.5447, + "step": 15348 + }, + { + "epoch": 0.8078421052631579, + "grad_norm": 0.51953125, + "learning_rate": 4.2701754385964914e-05, + "loss": 3.5065, + "step": 15349 + }, + { + "epoch": 0.8078947368421052, + "grad_norm": 0.5, + "learning_rate": 4.269005847953216e-05, + "loss": 3.2586, + "step": 15350 + }, + { + "epoch": 0.8079473684210526, + "grad_norm": 0.5703125, + "learning_rate": 4.267836257309942e-05, + "loss": 3.0057, + "step": 15351 + }, + { + "epoch": 0.808, + "grad_norm": 0.515625, + "learning_rate": 4.266666666666667e-05, + "loss": 2.7712, + "step": 15352 + }, + { + "epoch": 0.8080526315789474, + "grad_norm": 0.4609375, + "learning_rate": 4.265497076023392e-05, + "loss": 3.1088, + "step": 15353 + }, + { + "epoch": 0.8081052631578948, + "grad_norm": 0.50390625, + "learning_rate": 4.2643274853801175e-05, + "loss": 3.4855, + "step": 15354 + }, + { + "epoch": 0.8081578947368421, + "grad_norm": 0.47265625, + "learning_rate": 4.2631578947368425e-05, + "loss": 3.058, + "step": 15355 + }, + { + "epoch": 0.8082105263157895, + "grad_norm": 0.484375, + "learning_rate": 4.2619883040935674e-05, + "loss": 3.6134, + "step": 15356 + }, + { + "epoch": 0.8082631578947368, + "grad_norm": 0.4765625, + "learning_rate": 4.2608187134502924e-05, + "loss": 3.0833, + "step": 15357 + }, + { + "epoch": 0.8083157894736842, + "grad_norm": 0.46875, + "learning_rate": 4.259649122807018e-05, + "loss": 2.9812, + "step": 15358 + }, + { + "epoch": 0.8083684210526316, + "grad_norm": 0.64453125, + "learning_rate": 4.258479532163743e-05, + "loss": 3.5535, + "step": 15359 + }, + { + "epoch": 0.8084210526315789, + "grad_norm": 0.53515625, + "learning_rate": 4.257309941520468e-05, + "loss": 3.054, + "step": 15360 + }, + { + "epoch": 0.8084736842105263, + "grad_norm": 0.5, + "learning_rate": 4.256140350877193e-05, + "loss": 3.0884, + "step": 15361 + }, + { + "epoch": 0.8085263157894736, + "grad_norm": 0.4921875, + "learning_rate": 4.2549707602339186e-05, + "loss": 2.9142, + "step": 15362 + }, + { + "epoch": 0.8085789473684211, + "grad_norm": 0.53125, + "learning_rate": 4.2538011695906435e-05, + "loss": 3.4086, + "step": 15363 + }, + { + "epoch": 0.8086315789473684, + "grad_norm": 0.515625, + "learning_rate": 4.2526315789473685e-05, + "loss": 3.5619, + "step": 15364 + }, + { + "epoch": 0.8086842105263158, + "grad_norm": 0.486328125, + "learning_rate": 4.251461988304094e-05, + "loss": 3.833, + "step": 15365 + }, + { + "epoch": 0.8087368421052632, + "grad_norm": 0.5, + "learning_rate": 4.250292397660819e-05, + "loss": 3.4167, + "step": 15366 + }, + { + "epoch": 0.8087894736842105, + "grad_norm": 0.482421875, + "learning_rate": 4.249122807017544e-05, + "loss": 2.7482, + "step": 15367 + }, + { + "epoch": 0.8088421052631579, + "grad_norm": 0.5390625, + "learning_rate": 4.247953216374269e-05, + "loss": 3.0305, + "step": 15368 + }, + { + "epoch": 0.8088947368421052, + "grad_norm": 0.51171875, + "learning_rate": 4.2467836257309946e-05, + "loss": 3.2433, + "step": 15369 + }, + { + "epoch": 0.8089473684210526, + "grad_norm": 0.47265625, + "learning_rate": 4.2456140350877196e-05, + "loss": 3.1205, + "step": 15370 + }, + { + "epoch": 0.809, + "grad_norm": 0.490234375, + "learning_rate": 4.2444444444444445e-05, + "loss": 3.158, + "step": 15371 + }, + { + "epoch": 0.8090526315789474, + "grad_norm": 0.478515625, + "learning_rate": 4.2432748538011695e-05, + "loss": 3.3088, + "step": 15372 + }, + { + "epoch": 0.8091052631578948, + "grad_norm": 0.5546875, + "learning_rate": 4.242105263157895e-05, + "loss": 2.91, + "step": 15373 + }, + { + "epoch": 0.8091578947368421, + "grad_norm": 0.50390625, + "learning_rate": 4.24093567251462e-05, + "loss": 3.3473, + "step": 15374 + }, + { + "epoch": 0.8092105263157895, + "grad_norm": 0.4921875, + "learning_rate": 4.239766081871345e-05, + "loss": 3.6555, + "step": 15375 + }, + { + "epoch": 0.8092631578947368, + "grad_norm": 0.50390625, + "learning_rate": 4.23859649122807e-05, + "loss": 3.402, + "step": 15376 + }, + { + "epoch": 0.8093157894736842, + "grad_norm": 0.5, + "learning_rate": 4.2374269005847957e-05, + "loss": 2.9978, + "step": 15377 + }, + { + "epoch": 0.8093684210526316, + "grad_norm": 0.51953125, + "learning_rate": 4.2362573099415206e-05, + "loss": 3.2565, + "step": 15378 + }, + { + "epoch": 0.8094210526315789, + "grad_norm": 0.458984375, + "learning_rate": 4.2350877192982456e-05, + "loss": 3.0289, + "step": 15379 + }, + { + "epoch": 0.8094736842105263, + "grad_norm": 0.4765625, + "learning_rate": 4.233918128654971e-05, + "loss": 3.4596, + "step": 15380 + }, + { + "epoch": 0.8095263157894736, + "grad_norm": 0.4921875, + "learning_rate": 4.232748538011696e-05, + "loss": 3.0937, + "step": 15381 + }, + { + "epoch": 0.8095789473684211, + "grad_norm": 0.498046875, + "learning_rate": 4.231578947368421e-05, + "loss": 3.1149, + "step": 15382 + }, + { + "epoch": 0.8096315789473684, + "grad_norm": 0.515625, + "learning_rate": 4.230409356725146e-05, + "loss": 3.3264, + "step": 15383 + }, + { + "epoch": 0.8096842105263158, + "grad_norm": 0.4921875, + "learning_rate": 4.229239766081872e-05, + "loss": 3.0576, + "step": 15384 + }, + { + "epoch": 0.8097368421052632, + "grad_norm": 0.486328125, + "learning_rate": 4.228070175438597e-05, + "loss": 3.1003, + "step": 15385 + }, + { + "epoch": 0.8097894736842105, + "grad_norm": 0.49609375, + "learning_rate": 4.2269005847953216e-05, + "loss": 3.5121, + "step": 15386 + }, + { + "epoch": 0.8098421052631579, + "grad_norm": 0.4921875, + "learning_rate": 4.2257309941520466e-05, + "loss": 3.2804, + "step": 15387 + }, + { + "epoch": 0.8098947368421052, + "grad_norm": 0.51953125, + "learning_rate": 4.224561403508772e-05, + "loss": 2.9864, + "step": 15388 + }, + { + "epoch": 0.8099473684210526, + "grad_norm": 0.486328125, + "learning_rate": 4.223391812865497e-05, + "loss": 3.0611, + "step": 15389 + }, + { + "epoch": 0.81, + "grad_norm": 0.490234375, + "learning_rate": 4.222222222222222e-05, + "loss": 3.4148, + "step": 15390 + }, + { + "epoch": 0.8100526315789474, + "grad_norm": 0.484375, + "learning_rate": 4.221052631578948e-05, + "loss": 3.2506, + "step": 15391 + }, + { + "epoch": 0.8101052631578948, + "grad_norm": 0.486328125, + "learning_rate": 4.219883040935673e-05, + "loss": 3.4702, + "step": 15392 + }, + { + "epoch": 0.8101578947368421, + "grad_norm": 0.466796875, + "learning_rate": 4.2187134502923984e-05, + "loss": 2.9908, + "step": 15393 + }, + { + "epoch": 0.8102105263157895, + "grad_norm": 0.51953125, + "learning_rate": 4.217543859649123e-05, + "loss": 3.3851, + "step": 15394 + }, + { + "epoch": 0.8102631578947368, + "grad_norm": 0.484375, + "learning_rate": 4.216374269005848e-05, + "loss": 3.3924, + "step": 15395 + }, + { + "epoch": 0.8103157894736842, + "grad_norm": 0.51953125, + "learning_rate": 4.215204678362573e-05, + "loss": 3.449, + "step": 15396 + }, + { + "epoch": 0.8103684210526316, + "grad_norm": 0.49609375, + "learning_rate": 4.214035087719298e-05, + "loss": 3.0033, + "step": 15397 + }, + { + "epoch": 0.8104210526315789, + "grad_norm": 0.5234375, + "learning_rate": 4.212865497076023e-05, + "loss": 3.3508, + "step": 15398 + }, + { + "epoch": 0.8104736842105263, + "grad_norm": 0.5234375, + "learning_rate": 4.211695906432749e-05, + "loss": 3.2437, + "step": 15399 + }, + { + "epoch": 0.8105263157894737, + "grad_norm": 0.490234375, + "learning_rate": 4.210526315789474e-05, + "loss": 3.4334, + "step": 15400 + }, + { + "epoch": 0.8105789473684211, + "grad_norm": 0.46875, + "learning_rate": 4.209356725146199e-05, + "loss": 3.7211, + "step": 15401 + }, + { + "epoch": 0.8106315789473684, + "grad_norm": 0.47265625, + "learning_rate": 4.2081871345029244e-05, + "loss": 3.3629, + "step": 15402 + }, + { + "epoch": 0.8106842105263158, + "grad_norm": 0.51953125, + "learning_rate": 4.207017543859649e-05, + "loss": 3.5849, + "step": 15403 + }, + { + "epoch": 0.8107368421052632, + "grad_norm": 0.51953125, + "learning_rate": 4.205847953216375e-05, + "loss": 3.7707, + "step": 15404 + }, + { + "epoch": 0.8107894736842105, + "grad_norm": 0.5, + "learning_rate": 4.204678362573099e-05, + "loss": 3.5707, + "step": 15405 + }, + { + "epoch": 0.8108421052631579, + "grad_norm": 0.48828125, + "learning_rate": 4.203508771929825e-05, + "loss": 3.4464, + "step": 15406 + }, + { + "epoch": 0.8108947368421052, + "grad_norm": 0.5390625, + "learning_rate": 4.20233918128655e-05, + "loss": 3.1508, + "step": 15407 + }, + { + "epoch": 0.8109473684210526, + "grad_norm": 0.48046875, + "learning_rate": 4.2011695906432755e-05, + "loss": 3.0637, + "step": 15408 + }, + { + "epoch": 0.811, + "grad_norm": 0.478515625, + "learning_rate": 4.2e-05, + "loss": 3.2085, + "step": 15409 + }, + { + "epoch": 0.8110526315789474, + "grad_norm": 0.486328125, + "learning_rate": 4.1988304093567254e-05, + "loss": 3.0907, + "step": 15410 + }, + { + "epoch": 0.8111052631578948, + "grad_norm": 0.48046875, + "learning_rate": 4.1976608187134504e-05, + "loss": 3.3992, + "step": 15411 + }, + { + "epoch": 0.8111578947368421, + "grad_norm": 0.5078125, + "learning_rate": 4.196491228070176e-05, + "loss": 3.4009, + "step": 15412 + }, + { + "epoch": 0.8112105263157895, + "grad_norm": 0.515625, + "learning_rate": 4.1953216374269e-05, + "loss": 3.2609, + "step": 15413 + }, + { + "epoch": 0.8112631578947368, + "grad_norm": 0.4765625, + "learning_rate": 4.194152046783626e-05, + "loss": 3.1951, + "step": 15414 + }, + { + "epoch": 0.8113157894736842, + "grad_norm": 0.5, + "learning_rate": 4.1929824561403516e-05, + "loss": 2.6402, + "step": 15415 + }, + { + "epoch": 0.8113684210526316, + "grad_norm": 0.55078125, + "learning_rate": 4.1918128654970765e-05, + "loss": 3.0749, + "step": 15416 + }, + { + "epoch": 0.8114210526315789, + "grad_norm": 0.478515625, + "learning_rate": 4.1906432748538015e-05, + "loss": 3.5193, + "step": 15417 + }, + { + "epoch": 0.8114736842105263, + "grad_norm": 0.486328125, + "learning_rate": 4.1894736842105264e-05, + "loss": 2.901, + "step": 15418 + }, + { + "epoch": 0.8115263157894737, + "grad_norm": 0.49609375, + "learning_rate": 4.188304093567252e-05, + "loss": 3.2126, + "step": 15419 + }, + { + "epoch": 0.8115789473684211, + "grad_norm": 0.48828125, + "learning_rate": 4.1871345029239764e-05, + "loss": 3.4226, + "step": 15420 + }, + { + "epoch": 0.8116315789473684, + "grad_norm": 0.482421875, + "learning_rate": 4.185964912280702e-05, + "loss": 3.065, + "step": 15421 + }, + { + "epoch": 0.8116842105263158, + "grad_norm": 0.515625, + "learning_rate": 4.184795321637427e-05, + "loss": 3.4617, + "step": 15422 + }, + { + "epoch": 0.8117368421052632, + "grad_norm": 0.5078125, + "learning_rate": 4.1836257309941526e-05, + "loss": 3.6947, + "step": 15423 + }, + { + "epoch": 0.8117894736842105, + "grad_norm": 0.51953125, + "learning_rate": 4.182456140350877e-05, + "loss": 3.5517, + "step": 15424 + }, + { + "epoch": 0.8118421052631579, + "grad_norm": 0.5078125, + "learning_rate": 4.1812865497076025e-05, + "loss": 3.5836, + "step": 15425 + }, + { + "epoch": 0.8118947368421052, + "grad_norm": 0.62109375, + "learning_rate": 4.1801169590643275e-05, + "loss": 2.9602, + "step": 15426 + }, + { + "epoch": 0.8119473684210526, + "grad_norm": 0.5078125, + "learning_rate": 4.178947368421053e-05, + "loss": 3.4924, + "step": 15427 + }, + { + "epoch": 0.812, + "grad_norm": 0.48046875, + "learning_rate": 4.177777777777778e-05, + "loss": 3.2299, + "step": 15428 + }, + { + "epoch": 0.8120526315789474, + "grad_norm": 0.51171875, + "learning_rate": 4.176608187134503e-05, + "loss": 3.6055, + "step": 15429 + }, + { + "epoch": 0.8121052631578948, + "grad_norm": 0.490234375, + "learning_rate": 4.1754385964912287e-05, + "loss": 3.1786, + "step": 15430 + }, + { + "epoch": 0.8121578947368421, + "grad_norm": 0.4921875, + "learning_rate": 4.1742690058479536e-05, + "loss": 3.6357, + "step": 15431 + }, + { + "epoch": 0.8122105263157895, + "grad_norm": 0.466796875, + "learning_rate": 4.1730994152046786e-05, + "loss": 3.3343, + "step": 15432 + }, + { + "epoch": 0.8122631578947368, + "grad_norm": 0.59765625, + "learning_rate": 4.1719298245614035e-05, + "loss": 2.919, + "step": 15433 + }, + { + "epoch": 0.8123157894736842, + "grad_norm": 0.51171875, + "learning_rate": 4.170760233918129e-05, + "loss": 3.1909, + "step": 15434 + }, + { + "epoch": 0.8123684210526316, + "grad_norm": 0.498046875, + "learning_rate": 4.169590643274854e-05, + "loss": 3.2392, + "step": 15435 + }, + { + "epoch": 0.8124210526315789, + "grad_norm": 0.515625, + "learning_rate": 4.168421052631579e-05, + "loss": 3.2699, + "step": 15436 + }, + { + "epoch": 0.8124736842105263, + "grad_norm": 0.490234375, + "learning_rate": 4.167251461988304e-05, + "loss": 3.3268, + "step": 15437 + }, + { + "epoch": 0.8125263157894737, + "grad_norm": 0.50390625, + "learning_rate": 4.16608187134503e-05, + "loss": 2.8973, + "step": 15438 + }, + { + "epoch": 0.8125789473684211, + "grad_norm": 0.51953125, + "learning_rate": 4.1649122807017546e-05, + "loss": 3.1262, + "step": 15439 + }, + { + "epoch": 0.8126315789473684, + "grad_norm": 0.55078125, + "learning_rate": 4.1637426900584796e-05, + "loss": 3.8381, + "step": 15440 + }, + { + "epoch": 0.8126842105263158, + "grad_norm": 0.48828125, + "learning_rate": 4.162573099415205e-05, + "loss": 3.6197, + "step": 15441 + }, + { + "epoch": 0.8127368421052632, + "grad_norm": 0.486328125, + "learning_rate": 4.16140350877193e-05, + "loss": 3.4568, + "step": 15442 + }, + { + "epoch": 0.8127894736842105, + "grad_norm": 0.4921875, + "learning_rate": 4.160233918128655e-05, + "loss": 3.2178, + "step": 15443 + }, + { + "epoch": 0.8128421052631579, + "grad_norm": 0.47265625, + "learning_rate": 4.15906432748538e-05, + "loss": 2.941, + "step": 15444 + }, + { + "epoch": 0.8128947368421052, + "grad_norm": 0.50390625, + "learning_rate": 4.157894736842106e-05, + "loss": 3.2892, + "step": 15445 + }, + { + "epoch": 0.8129473684210526, + "grad_norm": 0.482421875, + "learning_rate": 4.156725146198831e-05, + "loss": 3.4316, + "step": 15446 + }, + { + "epoch": 0.813, + "grad_norm": 0.51171875, + "learning_rate": 4.155555555555556e-05, + "loss": 3.2045, + "step": 15447 + }, + { + "epoch": 0.8130526315789474, + "grad_norm": 0.486328125, + "learning_rate": 4.1543859649122806e-05, + "loss": 3.6489, + "step": 15448 + }, + { + "epoch": 0.8131052631578948, + "grad_norm": 0.51953125, + "learning_rate": 4.153216374269006e-05, + "loss": 3.1534, + "step": 15449 + }, + { + "epoch": 0.8131578947368421, + "grad_norm": 0.5, + "learning_rate": 4.152046783625731e-05, + "loss": 3.2134, + "step": 15450 + }, + { + "epoch": 0.8132105263157895, + "grad_norm": 0.53515625, + "learning_rate": 4.150877192982456e-05, + "loss": 3.1431, + "step": 15451 + }, + { + "epoch": 0.8132631578947368, + "grad_norm": 0.51171875, + "learning_rate": 4.149707602339181e-05, + "loss": 3.5933, + "step": 15452 + }, + { + "epoch": 0.8133157894736842, + "grad_norm": 0.5, + "learning_rate": 4.148538011695907e-05, + "loss": 3.3429, + "step": 15453 + }, + { + "epoch": 0.8133684210526316, + "grad_norm": 0.5078125, + "learning_rate": 4.147368421052632e-05, + "loss": 3.1433, + "step": 15454 + }, + { + "epoch": 0.8134210526315789, + "grad_norm": 0.494140625, + "learning_rate": 4.146198830409357e-05, + "loss": 2.6596, + "step": 15455 + }, + { + "epoch": 0.8134736842105263, + "grad_norm": 0.52734375, + "learning_rate": 4.1450292397660823e-05, + "loss": 3.5598, + "step": 15456 + }, + { + "epoch": 0.8135263157894737, + "grad_norm": 0.4765625, + "learning_rate": 4.143859649122807e-05, + "loss": 3.2734, + "step": 15457 + }, + { + "epoch": 0.8135789473684211, + "grad_norm": 0.515625, + "learning_rate": 4.142690058479532e-05, + "loss": 3.2104, + "step": 15458 + }, + { + "epoch": 0.8136315789473684, + "grad_norm": 0.478515625, + "learning_rate": 4.141520467836257e-05, + "loss": 3.4205, + "step": 15459 + }, + { + "epoch": 0.8136842105263158, + "grad_norm": 0.57421875, + "learning_rate": 4.140350877192983e-05, + "loss": 3.2602, + "step": 15460 + }, + { + "epoch": 0.8137368421052632, + "grad_norm": 0.484375, + "learning_rate": 4.139181286549708e-05, + "loss": 3.4469, + "step": 15461 + }, + { + "epoch": 0.8137894736842105, + "grad_norm": 0.484375, + "learning_rate": 4.138011695906433e-05, + "loss": 3.2591, + "step": 15462 + }, + { + "epoch": 0.8138421052631579, + "grad_norm": 0.494140625, + "learning_rate": 4.136842105263158e-05, + "loss": 3.1676, + "step": 15463 + }, + { + "epoch": 0.8138947368421052, + "grad_norm": 0.5, + "learning_rate": 4.1356725146198834e-05, + "loss": 3.5738, + "step": 15464 + }, + { + "epoch": 0.8139473684210526, + "grad_norm": 0.48828125, + "learning_rate": 4.134502923976608e-05, + "loss": 3.5962, + "step": 15465 + }, + { + "epoch": 0.814, + "grad_norm": 0.482421875, + "learning_rate": 4.133333333333333e-05, + "loss": 3.4354, + "step": 15466 + }, + { + "epoch": 0.8140526315789474, + "grad_norm": 0.48828125, + "learning_rate": 4.132163742690059e-05, + "loss": 3.3519, + "step": 15467 + }, + { + "epoch": 0.8141052631578948, + "grad_norm": 0.5078125, + "learning_rate": 4.130994152046784e-05, + "loss": 3.3468, + "step": 15468 + }, + { + "epoch": 0.8141578947368421, + "grad_norm": 0.5390625, + "learning_rate": 4.129824561403509e-05, + "loss": 3.2663, + "step": 15469 + }, + { + "epoch": 0.8142105263157895, + "grad_norm": 0.5, + "learning_rate": 4.128654970760234e-05, + "loss": 3.2115, + "step": 15470 + }, + { + "epoch": 0.8142631578947368, + "grad_norm": 0.486328125, + "learning_rate": 4.1274853801169594e-05, + "loss": 3.2853, + "step": 15471 + }, + { + "epoch": 0.8143157894736842, + "grad_norm": 0.486328125, + "learning_rate": 4.1263157894736844e-05, + "loss": 3.5585, + "step": 15472 + }, + { + "epoch": 0.8143684210526316, + "grad_norm": 0.49609375, + "learning_rate": 4.1251461988304094e-05, + "loss": 3.3871, + "step": 15473 + }, + { + "epoch": 0.8144210526315789, + "grad_norm": 0.5078125, + "learning_rate": 4.123976608187134e-05, + "loss": 2.8251, + "step": 15474 + }, + { + "epoch": 0.8144736842105263, + "grad_norm": 0.5078125, + "learning_rate": 4.12280701754386e-05, + "loss": 3.3751, + "step": 15475 + }, + { + "epoch": 0.8145263157894737, + "grad_norm": 0.494140625, + "learning_rate": 4.121637426900585e-05, + "loss": 3.479, + "step": 15476 + }, + { + "epoch": 0.8145789473684211, + "grad_norm": 0.7890625, + "learning_rate": 4.12046783625731e-05, + "loss": 3.4348, + "step": 15477 + }, + { + "epoch": 0.8146315789473684, + "grad_norm": 0.484375, + "learning_rate": 4.1192982456140355e-05, + "loss": 3.2965, + "step": 15478 + }, + { + "epoch": 0.8146842105263158, + "grad_norm": 0.5390625, + "learning_rate": 4.1181286549707605e-05, + "loss": 2.6354, + "step": 15479 + }, + { + "epoch": 0.8147368421052632, + "grad_norm": 0.50390625, + "learning_rate": 4.116959064327486e-05, + "loss": 3.2923, + "step": 15480 + }, + { + "epoch": 0.8147894736842105, + "grad_norm": 0.50390625, + "learning_rate": 4.1157894736842104e-05, + "loss": 3.441, + "step": 15481 + }, + { + "epoch": 0.8148421052631579, + "grad_norm": 0.50390625, + "learning_rate": 4.114619883040936e-05, + "loss": 3.3303, + "step": 15482 + }, + { + "epoch": 0.8148947368421052, + "grad_norm": 0.494140625, + "learning_rate": 4.113450292397661e-05, + "loss": 3.2095, + "step": 15483 + }, + { + "epoch": 0.8149473684210526, + "grad_norm": 0.5, + "learning_rate": 4.1122807017543866e-05, + "loss": 2.9797, + "step": 15484 + }, + { + "epoch": 0.815, + "grad_norm": 0.50390625, + "learning_rate": 4.111111111111111e-05, + "loss": 3.2108, + "step": 15485 + }, + { + "epoch": 0.8150526315789474, + "grad_norm": 0.50390625, + "learning_rate": 4.1099415204678365e-05, + "loss": 3.3793, + "step": 15486 + }, + { + "epoch": 0.8151052631578948, + "grad_norm": 0.49609375, + "learning_rate": 4.1087719298245615e-05, + "loss": 3.5225, + "step": 15487 + }, + { + "epoch": 0.8151578947368421, + "grad_norm": 0.478515625, + "learning_rate": 4.1076023391812865e-05, + "loss": 3.5271, + "step": 15488 + }, + { + "epoch": 0.8152105263157895, + "grad_norm": 0.462890625, + "learning_rate": 4.1064327485380114e-05, + "loss": 3.1566, + "step": 15489 + }, + { + "epoch": 0.8152631578947368, + "grad_norm": 0.478515625, + "learning_rate": 4.105263157894737e-05, + "loss": 3.3407, + "step": 15490 + }, + { + "epoch": 0.8153157894736842, + "grad_norm": 0.494140625, + "learning_rate": 4.104093567251463e-05, + "loss": 2.7333, + "step": 15491 + }, + { + "epoch": 0.8153684210526316, + "grad_norm": 0.48046875, + "learning_rate": 4.102923976608187e-05, + "loss": 3.1525, + "step": 15492 + }, + { + "epoch": 0.8154210526315789, + "grad_norm": 0.484375, + "learning_rate": 4.1017543859649126e-05, + "loss": 3.2932, + "step": 15493 + }, + { + "epoch": 0.8154736842105264, + "grad_norm": 0.478515625, + "learning_rate": 4.1005847953216376e-05, + "loss": 3.0962, + "step": 15494 + }, + { + "epoch": 0.8155263157894737, + "grad_norm": 0.5078125, + "learning_rate": 4.099415204678363e-05, + "loss": 3.532, + "step": 15495 + }, + { + "epoch": 0.8155789473684211, + "grad_norm": 0.486328125, + "learning_rate": 4.0982456140350875e-05, + "loss": 3.4403, + "step": 15496 + }, + { + "epoch": 0.8156315789473684, + "grad_norm": 0.4921875, + "learning_rate": 4.097076023391813e-05, + "loss": 2.9864, + "step": 15497 + }, + { + "epoch": 0.8156842105263158, + "grad_norm": 0.5078125, + "learning_rate": 4.095906432748538e-05, + "loss": 3.0098, + "step": 15498 + }, + { + "epoch": 0.8157368421052632, + "grad_norm": 0.49609375, + "learning_rate": 4.094736842105264e-05, + "loss": 3.2082, + "step": 15499 + }, + { + "epoch": 0.8157894736842105, + "grad_norm": 0.5078125, + "learning_rate": 4.093567251461988e-05, + "loss": 2.7441, + "step": 15500 + }, + { + "epoch": 0.8158421052631579, + "grad_norm": 0.50390625, + "learning_rate": 4.0923976608187136e-05, + "loss": 3.364, + "step": 15501 + }, + { + "epoch": 0.8158947368421052, + "grad_norm": 0.50390625, + "learning_rate": 4.0912280701754386e-05, + "loss": 3.4696, + "step": 15502 + }, + { + "epoch": 0.8159473684210526, + "grad_norm": 0.51171875, + "learning_rate": 4.090058479532164e-05, + "loss": 3.2131, + "step": 15503 + }, + { + "epoch": 0.816, + "grad_norm": 0.490234375, + "learning_rate": 4.088888888888889e-05, + "loss": 3.345, + "step": 15504 + }, + { + "epoch": 0.8160526315789474, + "grad_norm": 0.5, + "learning_rate": 4.087719298245614e-05, + "loss": 3.3753, + "step": 15505 + }, + { + "epoch": 0.8161052631578948, + "grad_norm": 0.498046875, + "learning_rate": 4.08654970760234e-05, + "loss": 3.1964, + "step": 15506 + }, + { + "epoch": 0.8161578947368421, + "grad_norm": 0.4765625, + "learning_rate": 4.085380116959065e-05, + "loss": 3.3547, + "step": 15507 + }, + { + "epoch": 0.8162105263157895, + "grad_norm": 0.5, + "learning_rate": 4.08421052631579e-05, + "loss": 3.2707, + "step": 15508 + }, + { + "epoch": 0.8162631578947368, + "grad_norm": 0.515625, + "learning_rate": 4.083040935672515e-05, + "loss": 2.586, + "step": 15509 + }, + { + "epoch": 0.8163157894736842, + "grad_norm": 0.50390625, + "learning_rate": 4.08187134502924e-05, + "loss": 3.2693, + "step": 15510 + }, + { + "epoch": 0.8163684210526316, + "grad_norm": 0.5703125, + "learning_rate": 4.0807017543859646e-05, + "loss": 3.0363, + "step": 15511 + }, + { + "epoch": 0.8164210526315789, + "grad_norm": 0.490234375, + "learning_rate": 4.07953216374269e-05, + "loss": 3.0712, + "step": 15512 + }, + { + "epoch": 0.8164736842105264, + "grad_norm": 0.46484375, + "learning_rate": 4.078362573099415e-05, + "loss": 2.9525, + "step": 15513 + }, + { + "epoch": 0.8165263157894737, + "grad_norm": 0.53125, + "learning_rate": 4.077192982456141e-05, + "loss": 3.4154, + "step": 15514 + }, + { + "epoch": 0.8165789473684211, + "grad_norm": 0.5078125, + "learning_rate": 4.076023391812866e-05, + "loss": 3.2428, + "step": 15515 + }, + { + "epoch": 0.8166315789473684, + "grad_norm": 0.4921875, + "learning_rate": 4.074853801169591e-05, + "loss": 3.7207, + "step": 15516 + }, + { + "epoch": 0.8166842105263158, + "grad_norm": 0.490234375, + "learning_rate": 4.0736842105263164e-05, + "loss": 3.2329, + "step": 15517 + }, + { + "epoch": 0.8167368421052632, + "grad_norm": 0.466796875, + "learning_rate": 4.072514619883041e-05, + "loss": 3.0354, + "step": 15518 + }, + { + "epoch": 0.8167894736842105, + "grad_norm": 0.51171875, + "learning_rate": 4.071345029239766e-05, + "loss": 3.449, + "step": 15519 + }, + { + "epoch": 0.8168421052631579, + "grad_norm": 0.52734375, + "learning_rate": 4.070175438596491e-05, + "loss": 2.9399, + "step": 15520 + }, + { + "epoch": 0.8168947368421052, + "grad_norm": 0.47265625, + "learning_rate": 4.069005847953217e-05, + "loss": 3.14, + "step": 15521 + }, + { + "epoch": 0.8169473684210526, + "grad_norm": 0.484375, + "learning_rate": 4.067836257309942e-05, + "loss": 3.2797, + "step": 15522 + }, + { + "epoch": 0.817, + "grad_norm": 0.5078125, + "learning_rate": 4.066666666666667e-05, + "loss": 3.5857, + "step": 15523 + }, + { + "epoch": 0.8170526315789474, + "grad_norm": 0.486328125, + "learning_rate": 4.065497076023392e-05, + "loss": 3.3526, + "step": 15524 + }, + { + "epoch": 0.8171052631578948, + "grad_norm": 0.494140625, + "learning_rate": 4.0643274853801174e-05, + "loss": 3.6738, + "step": 15525 + }, + { + "epoch": 0.8171578947368421, + "grad_norm": 0.6171875, + "learning_rate": 4.0631578947368424e-05, + "loss": 3.4706, + "step": 15526 + }, + { + "epoch": 0.8172105263157895, + "grad_norm": 0.51953125, + "learning_rate": 4.061988304093567e-05, + "loss": 3.0508, + "step": 15527 + }, + { + "epoch": 0.8172631578947368, + "grad_norm": 0.5, + "learning_rate": 4.060818713450293e-05, + "loss": 2.943, + "step": 15528 + }, + { + "epoch": 0.8173157894736842, + "grad_norm": 0.498046875, + "learning_rate": 4.059649122807018e-05, + "loss": 3.117, + "step": 15529 + }, + { + "epoch": 0.8173684210526316, + "grad_norm": 0.49609375, + "learning_rate": 4.058479532163743e-05, + "loss": 3.3429, + "step": 15530 + }, + { + "epoch": 0.8174210526315789, + "grad_norm": 0.47265625, + "learning_rate": 4.057309941520468e-05, + "loss": 3.3364, + "step": 15531 + }, + { + "epoch": 0.8174736842105264, + "grad_norm": 0.48046875, + "learning_rate": 4.0561403508771935e-05, + "loss": 3.2474, + "step": 15532 + }, + { + "epoch": 0.8175263157894737, + "grad_norm": 0.498046875, + "learning_rate": 4.0549707602339184e-05, + "loss": 3.6838, + "step": 15533 + }, + { + "epoch": 0.8175789473684211, + "grad_norm": 0.51171875, + "learning_rate": 4.0538011695906434e-05, + "loss": 2.9605, + "step": 15534 + }, + { + "epoch": 0.8176315789473684, + "grad_norm": 0.55859375, + "learning_rate": 4.0526315789473684e-05, + "loss": 2.7441, + "step": 15535 + }, + { + "epoch": 0.8176842105263158, + "grad_norm": 0.51171875, + "learning_rate": 4.051461988304094e-05, + "loss": 2.9342, + "step": 15536 + }, + { + "epoch": 0.8177368421052632, + "grad_norm": 0.50390625, + "learning_rate": 4.050292397660819e-05, + "loss": 3.1269, + "step": 15537 + }, + { + "epoch": 0.8177894736842105, + "grad_norm": 0.49609375, + "learning_rate": 4.049122807017544e-05, + "loss": 3.3831, + "step": 15538 + }, + { + "epoch": 0.8178421052631579, + "grad_norm": 0.50390625, + "learning_rate": 4.047953216374269e-05, + "loss": 3.2773, + "step": 15539 + }, + { + "epoch": 0.8178947368421052, + "grad_norm": 0.484375, + "learning_rate": 4.0467836257309945e-05, + "loss": 3.8647, + "step": 15540 + }, + { + "epoch": 0.8179473684210526, + "grad_norm": 0.50390625, + "learning_rate": 4.0456140350877195e-05, + "loss": 3.3076, + "step": 15541 + }, + { + "epoch": 0.818, + "grad_norm": 0.49609375, + "learning_rate": 4.0444444444444444e-05, + "loss": 3.5861, + "step": 15542 + }, + { + "epoch": 0.8180526315789474, + "grad_norm": 0.494140625, + "learning_rate": 4.04327485380117e-05, + "loss": 3.1072, + "step": 15543 + }, + { + "epoch": 0.8181052631578948, + "grad_norm": 0.462890625, + "learning_rate": 4.042105263157895e-05, + "loss": 2.974, + "step": 15544 + }, + { + "epoch": 0.8181578947368421, + "grad_norm": 0.49609375, + "learning_rate": 4.04093567251462e-05, + "loss": 3.2566, + "step": 15545 + }, + { + "epoch": 0.8182105263157895, + "grad_norm": 0.486328125, + "learning_rate": 4.039766081871345e-05, + "loss": 3.3458, + "step": 15546 + }, + { + "epoch": 0.8182631578947368, + "grad_norm": 0.5, + "learning_rate": 4.0385964912280706e-05, + "loss": 3.2766, + "step": 15547 + }, + { + "epoch": 0.8183157894736842, + "grad_norm": 0.60546875, + "learning_rate": 4.0374269005847955e-05, + "loss": 2.9287, + "step": 15548 + }, + { + "epoch": 0.8183684210526316, + "grad_norm": 0.4765625, + "learning_rate": 4.0362573099415205e-05, + "loss": 3.2987, + "step": 15549 + }, + { + "epoch": 0.8184210526315789, + "grad_norm": 0.5078125, + "learning_rate": 4.0350877192982455e-05, + "loss": 3.4703, + "step": 15550 + }, + { + "epoch": 0.8184736842105264, + "grad_norm": 0.484375, + "learning_rate": 4.033918128654971e-05, + "loss": 3.0091, + "step": 15551 + }, + { + "epoch": 0.8185263157894737, + "grad_norm": 0.486328125, + "learning_rate": 4.032748538011696e-05, + "loss": 3.1802, + "step": 15552 + }, + { + "epoch": 0.8185789473684211, + "grad_norm": 0.470703125, + "learning_rate": 4.031578947368421e-05, + "loss": 3.125, + "step": 15553 + }, + { + "epoch": 0.8186315789473684, + "grad_norm": 0.470703125, + "learning_rate": 4.0304093567251466e-05, + "loss": 2.9855, + "step": 15554 + }, + { + "epoch": 0.8186842105263158, + "grad_norm": 0.484375, + "learning_rate": 4.0292397660818716e-05, + "loss": 3.556, + "step": 15555 + }, + { + "epoch": 0.8187368421052632, + "grad_norm": 0.5078125, + "learning_rate": 4.0280701754385966e-05, + "loss": 3.2227, + "step": 15556 + }, + { + "epoch": 0.8187894736842105, + "grad_norm": 0.48046875, + "learning_rate": 4.0269005847953215e-05, + "loss": 3.6697, + "step": 15557 + }, + { + "epoch": 0.8188421052631579, + "grad_norm": 0.5859375, + "learning_rate": 4.025730994152047e-05, + "loss": 3.152, + "step": 15558 + }, + { + "epoch": 0.8188947368421052, + "grad_norm": 0.69140625, + "learning_rate": 4.024561403508772e-05, + "loss": 3.1231, + "step": 15559 + }, + { + "epoch": 0.8189473684210526, + "grad_norm": 0.49609375, + "learning_rate": 4.023391812865497e-05, + "loss": 3.4674, + "step": 15560 + }, + { + "epoch": 0.819, + "grad_norm": 0.494140625, + "learning_rate": 4.022222222222222e-05, + "loss": 3.5506, + "step": 15561 + }, + { + "epoch": 0.8190526315789474, + "grad_norm": 0.4765625, + "learning_rate": 4.021052631578948e-05, + "loss": 3.1821, + "step": 15562 + }, + { + "epoch": 0.8191052631578948, + "grad_norm": 0.48828125, + "learning_rate": 4.0198830409356726e-05, + "loss": 3.7546, + "step": 15563 + }, + { + "epoch": 0.8191578947368421, + "grad_norm": 0.5078125, + "learning_rate": 4.0187134502923976e-05, + "loss": 3.2544, + "step": 15564 + }, + { + "epoch": 0.8192105263157895, + "grad_norm": 0.5078125, + "learning_rate": 4.017543859649123e-05, + "loss": 3.3965, + "step": 15565 + }, + { + "epoch": 0.8192631578947368, + "grad_norm": 0.61328125, + "learning_rate": 4.016374269005848e-05, + "loss": 3.4037, + "step": 15566 + }, + { + "epoch": 0.8193157894736842, + "grad_norm": 0.74609375, + "learning_rate": 4.015204678362574e-05, + "loss": 3.1372, + "step": 15567 + }, + { + "epoch": 0.8193684210526315, + "grad_norm": 0.5, + "learning_rate": 4.014035087719298e-05, + "loss": 3.4074, + "step": 15568 + }, + { + "epoch": 0.8194210526315789, + "grad_norm": 0.49609375, + "learning_rate": 4.012865497076024e-05, + "loss": 3.3662, + "step": 15569 + }, + { + "epoch": 0.8194736842105264, + "grad_norm": 0.5, + "learning_rate": 4.011695906432749e-05, + "loss": 3.6201, + "step": 15570 + }, + { + "epoch": 0.8195263157894737, + "grad_norm": 0.48828125, + "learning_rate": 4.010526315789474e-05, + "loss": 3.2189, + "step": 15571 + }, + { + "epoch": 0.8195789473684211, + "grad_norm": 0.5, + "learning_rate": 4.0093567251461986e-05, + "loss": 3.3901, + "step": 15572 + }, + { + "epoch": 0.8196315789473684, + "grad_norm": 0.48828125, + "learning_rate": 4.008187134502924e-05, + "loss": 3.2766, + "step": 15573 + }, + { + "epoch": 0.8196842105263158, + "grad_norm": 0.53125, + "learning_rate": 4.007017543859649e-05, + "loss": 3.4578, + "step": 15574 + }, + { + "epoch": 0.8197368421052632, + "grad_norm": 0.5625, + "learning_rate": 4.005847953216375e-05, + "loss": 3.4031, + "step": 15575 + }, + { + "epoch": 0.8197894736842105, + "grad_norm": 0.5, + "learning_rate": 4.004678362573099e-05, + "loss": 3.5464, + "step": 15576 + }, + { + "epoch": 0.8198421052631579, + "grad_norm": 0.474609375, + "learning_rate": 4.003508771929825e-05, + "loss": 3.1022, + "step": 15577 + }, + { + "epoch": 0.8198947368421052, + "grad_norm": 0.50390625, + "learning_rate": 4.0023391812865504e-05, + "loss": 3.3399, + "step": 15578 + }, + { + "epoch": 0.8199473684210526, + "grad_norm": 0.48828125, + "learning_rate": 4.001169590643275e-05, + "loss": 3.1957, + "step": 15579 + }, + { + "epoch": 0.82, + "grad_norm": 0.52734375, + "learning_rate": 4e-05, + "loss": 3.0353, + "step": 15580 + }, + { + "epoch": 0.8200526315789474, + "grad_norm": 0.482421875, + "learning_rate": 3.998830409356725e-05, + "loss": 3.2216, + "step": 15581 + }, + { + "epoch": 0.8201052631578948, + "grad_norm": 0.494140625, + "learning_rate": 3.997660818713451e-05, + "loss": 3.4329, + "step": 15582 + }, + { + "epoch": 0.8201578947368421, + "grad_norm": 0.482421875, + "learning_rate": 3.996491228070175e-05, + "loss": 2.9142, + "step": 15583 + }, + { + "epoch": 0.8202105263157895, + "grad_norm": 0.4921875, + "learning_rate": 3.995321637426901e-05, + "loss": 3.5197, + "step": 15584 + }, + { + "epoch": 0.8202631578947368, + "grad_norm": 0.51953125, + "learning_rate": 3.994152046783626e-05, + "loss": 3.2721, + "step": 15585 + }, + { + "epoch": 0.8203157894736842, + "grad_norm": 0.47265625, + "learning_rate": 3.9929824561403514e-05, + "loss": 3.4338, + "step": 15586 + }, + { + "epoch": 0.8203684210526315, + "grad_norm": 0.50390625, + "learning_rate": 3.991812865497076e-05, + "loss": 3.0204, + "step": 15587 + }, + { + "epoch": 0.8204210526315789, + "grad_norm": 0.48828125, + "learning_rate": 3.9906432748538014e-05, + "loss": 3.2114, + "step": 15588 + }, + { + "epoch": 0.8204736842105264, + "grad_norm": 0.50390625, + "learning_rate": 3.989473684210526e-05, + "loss": 3.6334, + "step": 15589 + }, + { + "epoch": 0.8205263157894737, + "grad_norm": 0.50390625, + "learning_rate": 3.988304093567252e-05, + "loss": 3.1741, + "step": 15590 + }, + { + "epoch": 0.8205789473684211, + "grad_norm": 0.51953125, + "learning_rate": 3.987134502923977e-05, + "loss": 3.4316, + "step": 15591 + }, + { + "epoch": 0.8206315789473684, + "grad_norm": 0.515625, + "learning_rate": 3.985964912280702e-05, + "loss": 2.7528, + "step": 15592 + }, + { + "epoch": 0.8206842105263158, + "grad_norm": 0.4921875, + "learning_rate": 3.9847953216374275e-05, + "loss": 3.4136, + "step": 15593 + }, + { + "epoch": 0.8207368421052632, + "grad_norm": 0.5, + "learning_rate": 3.9836257309941525e-05, + "loss": 3.3606, + "step": 15594 + }, + { + "epoch": 0.8207894736842105, + "grad_norm": 0.515625, + "learning_rate": 3.9824561403508774e-05, + "loss": 3.4056, + "step": 15595 + }, + { + "epoch": 0.8208421052631579, + "grad_norm": 0.498046875, + "learning_rate": 3.9812865497076024e-05, + "loss": 3.2587, + "step": 15596 + }, + { + "epoch": 0.8208947368421052, + "grad_norm": 0.52734375, + "learning_rate": 3.980116959064328e-05, + "loss": 3.2848, + "step": 15597 + }, + { + "epoch": 0.8209473684210526, + "grad_norm": 0.470703125, + "learning_rate": 3.978947368421053e-05, + "loss": 3.1629, + "step": 15598 + }, + { + "epoch": 0.821, + "grad_norm": 0.5078125, + "learning_rate": 3.977777777777778e-05, + "loss": 3.2162, + "step": 15599 + }, + { + "epoch": 0.8210526315789474, + "grad_norm": 0.4921875, + "learning_rate": 3.976608187134503e-05, + "loss": 3.3351, + "step": 15600 + }, + { + "epoch": 0.8211052631578948, + "grad_norm": 0.51953125, + "learning_rate": 3.9754385964912285e-05, + "loss": 3.3544, + "step": 15601 + }, + { + "epoch": 0.8211578947368421, + "grad_norm": 0.53125, + "learning_rate": 3.974269005847953e-05, + "loss": 3.2365, + "step": 15602 + }, + { + "epoch": 0.8212105263157895, + "grad_norm": 0.52734375, + "learning_rate": 3.9730994152046785e-05, + "loss": 3.421, + "step": 15603 + }, + { + "epoch": 0.8212631578947368, + "grad_norm": 0.498046875, + "learning_rate": 3.971929824561404e-05, + "loss": 3.0553, + "step": 15604 + }, + { + "epoch": 0.8213157894736842, + "grad_norm": 0.5078125, + "learning_rate": 3.970760233918129e-05, + "loss": 3.3812, + "step": 15605 + }, + { + "epoch": 0.8213684210526315, + "grad_norm": 0.494140625, + "learning_rate": 3.969590643274854e-05, + "loss": 3.3594, + "step": 15606 + }, + { + "epoch": 0.8214210526315789, + "grad_norm": 0.5, + "learning_rate": 3.968421052631579e-05, + "loss": 3.1277, + "step": 15607 + }, + { + "epoch": 0.8214736842105264, + "grad_norm": 0.494140625, + "learning_rate": 3.9672514619883046e-05, + "loss": 3.4426, + "step": 15608 + }, + { + "epoch": 0.8215263157894737, + "grad_norm": 0.482421875, + "learning_rate": 3.9660818713450296e-05, + "loss": 3.5313, + "step": 15609 + }, + { + "epoch": 0.8215789473684211, + "grad_norm": 0.5078125, + "learning_rate": 3.9649122807017545e-05, + "loss": 3.058, + "step": 15610 + }, + { + "epoch": 0.8216315789473684, + "grad_norm": 0.4765625, + "learning_rate": 3.9637426900584795e-05, + "loss": 3.3267, + "step": 15611 + }, + { + "epoch": 0.8216842105263158, + "grad_norm": 0.5, + "learning_rate": 3.962573099415205e-05, + "loss": 3.6273, + "step": 15612 + }, + { + "epoch": 0.8217368421052632, + "grad_norm": 0.5078125, + "learning_rate": 3.96140350877193e-05, + "loss": 3.3505, + "step": 15613 + }, + { + "epoch": 0.8217894736842105, + "grad_norm": 0.482421875, + "learning_rate": 3.960233918128655e-05, + "loss": 3.0405, + "step": 15614 + }, + { + "epoch": 0.8218421052631579, + "grad_norm": 0.4765625, + "learning_rate": 3.95906432748538e-05, + "loss": 3.2328, + "step": 15615 + }, + { + "epoch": 0.8218947368421052, + "grad_norm": 0.4921875, + "learning_rate": 3.9578947368421056e-05, + "loss": 3.7397, + "step": 15616 + }, + { + "epoch": 0.8219473684210526, + "grad_norm": 0.50390625, + "learning_rate": 3.9567251461988306e-05, + "loss": 3.5105, + "step": 15617 + }, + { + "epoch": 0.822, + "grad_norm": 0.48046875, + "learning_rate": 3.9555555555555556e-05, + "loss": 3.0352, + "step": 15618 + }, + { + "epoch": 0.8220526315789474, + "grad_norm": 0.5234375, + "learning_rate": 3.954385964912281e-05, + "loss": 3.4456, + "step": 15619 + }, + { + "epoch": 0.8221052631578948, + "grad_norm": 0.50390625, + "learning_rate": 3.953216374269006e-05, + "loss": 3.0513, + "step": 15620 + }, + { + "epoch": 0.8221578947368421, + "grad_norm": 0.4765625, + "learning_rate": 3.952046783625731e-05, + "loss": 3.1212, + "step": 15621 + }, + { + "epoch": 0.8222105263157895, + "grad_norm": 0.51171875, + "learning_rate": 3.950877192982456e-05, + "loss": 3.1937, + "step": 15622 + }, + { + "epoch": 0.8222631578947368, + "grad_norm": 0.48046875, + "learning_rate": 3.949707602339182e-05, + "loss": 3.0215, + "step": 15623 + }, + { + "epoch": 0.8223157894736842, + "grad_norm": 0.49609375, + "learning_rate": 3.948538011695907e-05, + "loss": 3.135, + "step": 15624 + }, + { + "epoch": 0.8223684210526315, + "grad_norm": 0.51953125, + "learning_rate": 3.9473684210526316e-05, + "loss": 3.3419, + "step": 15625 + }, + { + "epoch": 0.8224210526315789, + "grad_norm": 0.5078125, + "learning_rate": 3.9461988304093566e-05, + "loss": 3.1381, + "step": 15626 + }, + { + "epoch": 0.8224736842105264, + "grad_norm": 0.490234375, + "learning_rate": 3.945029239766082e-05, + "loss": 3.1277, + "step": 15627 + }, + { + "epoch": 0.8225263157894737, + "grad_norm": 0.55078125, + "learning_rate": 3.943859649122807e-05, + "loss": 2.8018, + "step": 15628 + }, + { + "epoch": 0.8225789473684211, + "grad_norm": 0.5, + "learning_rate": 3.942690058479532e-05, + "loss": 2.903, + "step": 15629 + }, + { + "epoch": 0.8226315789473684, + "grad_norm": 0.484375, + "learning_rate": 3.941520467836258e-05, + "loss": 3.2523, + "step": 15630 + }, + { + "epoch": 0.8226842105263158, + "grad_norm": 0.494140625, + "learning_rate": 3.940350877192983e-05, + "loss": 3.3505, + "step": 15631 + }, + { + "epoch": 0.8227368421052632, + "grad_norm": 0.486328125, + "learning_rate": 3.939181286549708e-05, + "loss": 3.0485, + "step": 15632 + }, + { + "epoch": 0.8227894736842105, + "grad_norm": 0.48046875, + "learning_rate": 3.9380116959064327e-05, + "loss": 2.956, + "step": 15633 + }, + { + "epoch": 0.8228421052631579, + "grad_norm": 0.482421875, + "learning_rate": 3.936842105263158e-05, + "loss": 3.3186, + "step": 15634 + }, + { + "epoch": 0.8228947368421052, + "grad_norm": 0.490234375, + "learning_rate": 3.935672514619883e-05, + "loss": 3.0242, + "step": 15635 + }, + { + "epoch": 0.8229473684210526, + "grad_norm": 0.45703125, + "learning_rate": 3.934502923976608e-05, + "loss": 2.9175, + "step": 15636 + }, + { + "epoch": 0.823, + "grad_norm": 0.48046875, + "learning_rate": 3.933333333333333e-05, + "loss": 3.2502, + "step": 15637 + }, + { + "epoch": 0.8230526315789474, + "grad_norm": 0.5, + "learning_rate": 3.932163742690059e-05, + "loss": 3.6368, + "step": 15638 + }, + { + "epoch": 0.8231052631578948, + "grad_norm": 0.5703125, + "learning_rate": 3.930994152046784e-05, + "loss": 3.3582, + "step": 15639 + }, + { + "epoch": 0.8231578947368421, + "grad_norm": 0.52734375, + "learning_rate": 3.929824561403509e-05, + "loss": 3.4714, + "step": 15640 + }, + { + "epoch": 0.8232105263157895, + "grad_norm": 0.5703125, + "learning_rate": 3.9286549707602344e-05, + "loss": 3.1502, + "step": 15641 + }, + { + "epoch": 0.8232631578947368, + "grad_norm": 0.48046875, + "learning_rate": 3.927485380116959e-05, + "loss": 3.0916, + "step": 15642 + }, + { + "epoch": 0.8233157894736842, + "grad_norm": 0.5, + "learning_rate": 3.926315789473684e-05, + "loss": 3.5106, + "step": 15643 + }, + { + "epoch": 0.8233684210526315, + "grad_norm": 0.5, + "learning_rate": 3.925146198830409e-05, + "loss": 2.953, + "step": 15644 + }, + { + "epoch": 0.8234210526315789, + "grad_norm": 0.4765625, + "learning_rate": 3.923976608187135e-05, + "loss": 2.9638, + "step": 15645 + }, + { + "epoch": 0.8234736842105264, + "grad_norm": 0.49609375, + "learning_rate": 3.92280701754386e-05, + "loss": 3.4692, + "step": 15646 + }, + { + "epoch": 0.8235263157894737, + "grad_norm": 0.49609375, + "learning_rate": 3.921637426900585e-05, + "loss": 2.6529, + "step": 15647 + }, + { + "epoch": 0.8235789473684211, + "grad_norm": 0.490234375, + "learning_rate": 3.92046783625731e-05, + "loss": 3.5113, + "step": 15648 + }, + { + "epoch": 0.8236315789473684, + "grad_norm": 0.53125, + "learning_rate": 3.9192982456140354e-05, + "loss": 3.6038, + "step": 15649 + }, + { + "epoch": 0.8236842105263158, + "grad_norm": 0.515625, + "learning_rate": 3.9181286549707604e-05, + "loss": 2.5744, + "step": 15650 + }, + { + "epoch": 0.8237368421052632, + "grad_norm": 0.447265625, + "learning_rate": 3.916959064327485e-05, + "loss": 2.9784, + "step": 15651 + }, + { + "epoch": 0.8237894736842105, + "grad_norm": 0.48046875, + "learning_rate": 3.91578947368421e-05, + "loss": 3.4687, + "step": 15652 + }, + { + "epoch": 0.8238421052631579, + "grad_norm": 0.470703125, + "learning_rate": 3.914619883040936e-05, + "loss": 3.2769, + "step": 15653 + }, + { + "epoch": 0.8238947368421052, + "grad_norm": 0.5, + "learning_rate": 3.9134502923976615e-05, + "loss": 2.5174, + "step": 15654 + }, + { + "epoch": 0.8239473684210527, + "grad_norm": 0.546875, + "learning_rate": 3.912280701754386e-05, + "loss": 3.0804, + "step": 15655 + }, + { + "epoch": 0.824, + "grad_norm": 0.498046875, + "learning_rate": 3.9111111111111115e-05, + "loss": 3.2318, + "step": 15656 + }, + { + "epoch": 0.8240526315789474, + "grad_norm": 0.53125, + "learning_rate": 3.9099415204678364e-05, + "loss": 3.2278, + "step": 15657 + }, + { + "epoch": 0.8241052631578948, + "grad_norm": 0.5078125, + "learning_rate": 3.908771929824562e-05, + "loss": 2.8102, + "step": 15658 + }, + { + "epoch": 0.8241578947368421, + "grad_norm": 0.490234375, + "learning_rate": 3.9076023391812863e-05, + "loss": 2.9567, + "step": 15659 + }, + { + "epoch": 0.8242105263157895, + "grad_norm": 0.5078125, + "learning_rate": 3.906432748538012e-05, + "loss": 3.7267, + "step": 15660 + }, + { + "epoch": 0.8242631578947368, + "grad_norm": 0.4765625, + "learning_rate": 3.905263157894737e-05, + "loss": 3.3482, + "step": 15661 + }, + { + "epoch": 0.8243157894736842, + "grad_norm": 0.515625, + "learning_rate": 3.9040935672514626e-05, + "loss": 2.923, + "step": 15662 + }, + { + "epoch": 0.8243684210526315, + "grad_norm": 0.5234375, + "learning_rate": 3.902923976608187e-05, + "loss": 3.1121, + "step": 15663 + }, + { + "epoch": 0.8244210526315789, + "grad_norm": 0.49609375, + "learning_rate": 3.9017543859649125e-05, + "loss": 3.4372, + "step": 15664 + }, + { + "epoch": 0.8244736842105264, + "grad_norm": 0.470703125, + "learning_rate": 3.9005847953216374e-05, + "loss": 2.7633, + "step": 15665 + }, + { + "epoch": 0.8245263157894737, + "grad_norm": 0.486328125, + "learning_rate": 3.8994152046783624e-05, + "loss": 2.8699, + "step": 15666 + }, + { + "epoch": 0.8245789473684211, + "grad_norm": 0.494140625, + "learning_rate": 3.898245614035088e-05, + "loss": 3.2984, + "step": 15667 + }, + { + "epoch": 0.8246315789473684, + "grad_norm": 0.51171875, + "learning_rate": 3.897076023391813e-05, + "loss": 3.176, + "step": 15668 + }, + { + "epoch": 0.8246842105263158, + "grad_norm": 0.4921875, + "learning_rate": 3.8959064327485386e-05, + "loss": 3.2748, + "step": 15669 + }, + { + "epoch": 0.8247368421052632, + "grad_norm": 0.484375, + "learning_rate": 3.894736842105263e-05, + "loss": 3.3325, + "step": 15670 + }, + { + "epoch": 0.8247894736842105, + "grad_norm": 0.49609375, + "learning_rate": 3.8935672514619886e-05, + "loss": 3.0073, + "step": 15671 + }, + { + "epoch": 0.8248421052631579, + "grad_norm": 0.5078125, + "learning_rate": 3.8923976608187135e-05, + "loss": 3.418, + "step": 15672 + }, + { + "epoch": 0.8248947368421052, + "grad_norm": 0.50390625, + "learning_rate": 3.891228070175439e-05, + "loss": 3.3139, + "step": 15673 + }, + { + "epoch": 0.8249473684210527, + "grad_norm": 0.4921875, + "learning_rate": 3.8900584795321634e-05, + "loss": 3.1121, + "step": 15674 + }, + { + "epoch": 0.825, + "grad_norm": 0.51171875, + "learning_rate": 3.888888888888889e-05, + "loss": 3.3026, + "step": 15675 + }, + { + "epoch": 0.8250526315789474, + "grad_norm": 0.5234375, + "learning_rate": 3.887719298245614e-05, + "loss": 3.6297, + "step": 15676 + }, + { + "epoch": 0.8251052631578948, + "grad_norm": 0.51171875, + "learning_rate": 3.88654970760234e-05, + "loss": 2.6278, + "step": 15677 + }, + { + "epoch": 0.8251578947368421, + "grad_norm": 0.470703125, + "learning_rate": 3.8853801169590646e-05, + "loss": 3.1122, + "step": 15678 + }, + { + "epoch": 0.8252105263157895, + "grad_norm": 0.51953125, + "learning_rate": 3.8842105263157896e-05, + "loss": 3.3147, + "step": 15679 + }, + { + "epoch": 0.8252631578947368, + "grad_norm": 0.515625, + "learning_rate": 3.883040935672515e-05, + "loss": 3.0849, + "step": 15680 + }, + { + "epoch": 0.8253157894736842, + "grad_norm": 0.50390625, + "learning_rate": 3.88187134502924e-05, + "loss": 3.5981, + "step": 15681 + }, + { + "epoch": 0.8253684210526315, + "grad_norm": 0.484375, + "learning_rate": 3.880701754385965e-05, + "loss": 3.445, + "step": 15682 + }, + { + "epoch": 0.8254210526315789, + "grad_norm": 0.5234375, + "learning_rate": 3.87953216374269e-05, + "loss": 3.6487, + "step": 15683 + }, + { + "epoch": 0.8254736842105264, + "grad_norm": 0.46875, + "learning_rate": 3.878362573099416e-05, + "loss": 3.4396, + "step": 15684 + }, + { + "epoch": 0.8255263157894737, + "grad_norm": 0.474609375, + "learning_rate": 3.877192982456141e-05, + "loss": 2.4302, + "step": 15685 + }, + { + "epoch": 0.8255789473684211, + "grad_norm": 0.4921875, + "learning_rate": 3.8760233918128657e-05, + "loss": 3.2603, + "step": 15686 + }, + { + "epoch": 0.8256315789473684, + "grad_norm": 0.515625, + "learning_rate": 3.8748538011695906e-05, + "loss": 3.1992, + "step": 15687 + }, + { + "epoch": 0.8256842105263158, + "grad_norm": 0.51171875, + "learning_rate": 3.873684210526316e-05, + "loss": 3.2787, + "step": 15688 + }, + { + "epoch": 0.8257368421052631, + "grad_norm": 0.515625, + "learning_rate": 3.8725146198830405e-05, + "loss": 3.4383, + "step": 15689 + }, + { + "epoch": 0.8257894736842105, + "grad_norm": 0.51171875, + "learning_rate": 3.871345029239766e-05, + "loss": 3.2835, + "step": 15690 + }, + { + "epoch": 0.8258421052631579, + "grad_norm": 0.494140625, + "learning_rate": 3.870175438596492e-05, + "loss": 3.1638, + "step": 15691 + }, + { + "epoch": 0.8258947368421052, + "grad_norm": 0.490234375, + "learning_rate": 3.869005847953217e-05, + "loss": 3.4966, + "step": 15692 + }, + { + "epoch": 0.8259473684210527, + "grad_norm": 0.53515625, + "learning_rate": 3.867836257309942e-05, + "loss": 3.2498, + "step": 15693 + }, + { + "epoch": 0.826, + "grad_norm": 0.51171875, + "learning_rate": 3.866666666666667e-05, + "loss": 3.0512, + "step": 15694 + }, + { + "epoch": 0.8260526315789474, + "grad_norm": 0.50390625, + "learning_rate": 3.865497076023392e-05, + "loss": 3.5772, + "step": 15695 + }, + { + "epoch": 0.8261052631578948, + "grad_norm": 0.50390625, + "learning_rate": 3.864327485380117e-05, + "loss": 3.7423, + "step": 15696 + }, + { + "epoch": 0.8261578947368421, + "grad_norm": 0.5, + "learning_rate": 3.863157894736842e-05, + "loss": 2.7783, + "step": 15697 + }, + { + "epoch": 0.8262105263157895, + "grad_norm": 0.53125, + "learning_rate": 3.861988304093567e-05, + "loss": 3.5005, + "step": 15698 + }, + { + "epoch": 0.8262631578947368, + "grad_norm": 0.5234375, + "learning_rate": 3.860818713450293e-05, + "loss": 3.3759, + "step": 15699 + }, + { + "epoch": 0.8263157894736842, + "grad_norm": 0.4609375, + "learning_rate": 3.859649122807018e-05, + "loss": 3.5266, + "step": 15700 + }, + { + "epoch": 0.8263684210526315, + "grad_norm": 0.53125, + "learning_rate": 3.858479532163743e-05, + "loss": 2.9157, + "step": 15701 + }, + { + "epoch": 0.826421052631579, + "grad_norm": 0.474609375, + "learning_rate": 3.857309941520468e-05, + "loss": 3.1671, + "step": 15702 + }, + { + "epoch": 0.8264736842105264, + "grad_norm": 0.490234375, + "learning_rate": 3.8561403508771934e-05, + "loss": 3.3899, + "step": 15703 + }, + { + "epoch": 0.8265263157894737, + "grad_norm": 0.47265625, + "learning_rate": 3.854970760233918e-05, + "loss": 3.2246, + "step": 15704 + }, + { + "epoch": 0.8265789473684211, + "grad_norm": 0.51171875, + "learning_rate": 3.853801169590643e-05, + "loss": 2.8968, + "step": 15705 + }, + { + "epoch": 0.8266315789473684, + "grad_norm": 0.482421875, + "learning_rate": 3.852631578947369e-05, + "loss": 3.0398, + "step": 15706 + }, + { + "epoch": 0.8266842105263158, + "grad_norm": 0.494140625, + "learning_rate": 3.851461988304094e-05, + "loss": 3.3601, + "step": 15707 + }, + { + "epoch": 0.8267368421052631, + "grad_norm": 0.52734375, + "learning_rate": 3.850292397660819e-05, + "loss": 3.1581, + "step": 15708 + }, + { + "epoch": 0.8267894736842105, + "grad_norm": 0.49609375, + "learning_rate": 3.849122807017544e-05, + "loss": 3.3611, + "step": 15709 + }, + { + "epoch": 0.8268421052631579, + "grad_norm": 0.46875, + "learning_rate": 3.8479532163742694e-05, + "loss": 3.3568, + "step": 15710 + }, + { + "epoch": 0.8268947368421052, + "grad_norm": 0.49609375, + "learning_rate": 3.8467836257309944e-05, + "loss": 3.4821, + "step": 15711 + }, + { + "epoch": 0.8269473684210527, + "grad_norm": 0.482421875, + "learning_rate": 3.8456140350877193e-05, + "loss": 3.4835, + "step": 15712 + }, + { + "epoch": 0.827, + "grad_norm": 0.50390625, + "learning_rate": 3.844444444444444e-05, + "loss": 3.2683, + "step": 15713 + }, + { + "epoch": 0.8270526315789474, + "grad_norm": 0.515625, + "learning_rate": 3.84327485380117e-05, + "loss": 3.5393, + "step": 15714 + }, + { + "epoch": 0.8271052631578948, + "grad_norm": 0.5546875, + "learning_rate": 3.842105263157895e-05, + "loss": 3.4527, + "step": 15715 + }, + { + "epoch": 0.8271578947368421, + "grad_norm": 0.5625, + "learning_rate": 3.84093567251462e-05, + "loss": 3.2658, + "step": 15716 + }, + { + "epoch": 0.8272105263157895, + "grad_norm": 0.48828125, + "learning_rate": 3.8397660818713455e-05, + "loss": 3.5438, + "step": 15717 + }, + { + "epoch": 0.8272631578947368, + "grad_norm": 0.51171875, + "learning_rate": 3.8385964912280705e-05, + "loss": 3.1436, + "step": 15718 + }, + { + "epoch": 0.8273157894736842, + "grad_norm": 0.5, + "learning_rate": 3.8374269005847954e-05, + "loss": 3.7464, + "step": 15719 + }, + { + "epoch": 0.8273684210526315, + "grad_norm": 0.494140625, + "learning_rate": 3.8362573099415204e-05, + "loss": 2.8959, + "step": 15720 + }, + { + "epoch": 0.827421052631579, + "grad_norm": 0.474609375, + "learning_rate": 3.835087719298246e-05, + "loss": 3.3812, + "step": 15721 + }, + { + "epoch": 0.8274736842105264, + "grad_norm": 0.482421875, + "learning_rate": 3.833918128654971e-05, + "loss": 3.5811, + "step": 15722 + }, + { + "epoch": 0.8275263157894737, + "grad_norm": 0.484375, + "learning_rate": 3.832748538011696e-05, + "loss": 3.2833, + "step": 15723 + }, + { + "epoch": 0.8275789473684211, + "grad_norm": 0.49609375, + "learning_rate": 3.831578947368421e-05, + "loss": 3.6075, + "step": 15724 + }, + { + "epoch": 0.8276315789473684, + "grad_norm": 0.498046875, + "learning_rate": 3.8304093567251465e-05, + "loss": 3.163, + "step": 15725 + }, + { + "epoch": 0.8276842105263158, + "grad_norm": 0.498046875, + "learning_rate": 3.8292397660818715e-05, + "loss": 2.9481, + "step": 15726 + }, + { + "epoch": 0.8277368421052631, + "grad_norm": 0.482421875, + "learning_rate": 3.8280701754385964e-05, + "loss": 3.2591, + "step": 15727 + }, + { + "epoch": 0.8277894736842105, + "grad_norm": 0.47265625, + "learning_rate": 3.826900584795322e-05, + "loss": 2.9681, + "step": 15728 + }, + { + "epoch": 0.8278421052631579, + "grad_norm": 0.515625, + "learning_rate": 3.825730994152047e-05, + "loss": 3.0998, + "step": 15729 + }, + { + "epoch": 0.8278947368421052, + "grad_norm": 0.5078125, + "learning_rate": 3.824561403508773e-05, + "loss": 3.4335, + "step": 15730 + }, + { + "epoch": 0.8279473684210527, + "grad_norm": 0.5, + "learning_rate": 3.823391812865497e-05, + "loss": 3.3499, + "step": 15731 + }, + { + "epoch": 0.828, + "grad_norm": 0.5, + "learning_rate": 3.8222222222222226e-05, + "loss": 3.5918, + "step": 15732 + }, + { + "epoch": 0.8280526315789474, + "grad_norm": 0.5078125, + "learning_rate": 3.8210526315789476e-05, + "loss": 3.1035, + "step": 15733 + }, + { + "epoch": 0.8281052631578948, + "grad_norm": 0.47265625, + "learning_rate": 3.8198830409356725e-05, + "loss": 2.9921, + "step": 15734 + }, + { + "epoch": 0.8281578947368421, + "grad_norm": 0.486328125, + "learning_rate": 3.8187134502923975e-05, + "loss": 3.5838, + "step": 15735 + }, + { + "epoch": 0.8282105263157895, + "grad_norm": 0.53515625, + "learning_rate": 3.817543859649123e-05, + "loss": 3.2453, + "step": 15736 + }, + { + "epoch": 0.8282631578947368, + "grad_norm": 0.498046875, + "learning_rate": 3.816374269005848e-05, + "loss": 3.3131, + "step": 15737 + }, + { + "epoch": 0.8283157894736842, + "grad_norm": 0.4921875, + "learning_rate": 3.815204678362573e-05, + "loss": 3.0076, + "step": 15738 + }, + { + "epoch": 0.8283684210526315, + "grad_norm": 0.494140625, + "learning_rate": 3.814035087719298e-05, + "loss": 3.2032, + "step": 15739 + }, + { + "epoch": 0.828421052631579, + "grad_norm": 0.482421875, + "learning_rate": 3.8128654970760236e-05, + "loss": 3.1414, + "step": 15740 + }, + { + "epoch": 0.8284736842105264, + "grad_norm": 0.484375, + "learning_rate": 3.811695906432749e-05, + "loss": 3.4643, + "step": 15741 + }, + { + "epoch": 0.8285263157894737, + "grad_norm": 0.55859375, + "learning_rate": 3.8105263157894735e-05, + "loss": 3.5994, + "step": 15742 + }, + { + "epoch": 0.8285789473684211, + "grad_norm": 0.498046875, + "learning_rate": 3.809356725146199e-05, + "loss": 3.2481, + "step": 15743 + }, + { + "epoch": 0.8286315789473684, + "grad_norm": 0.5546875, + "learning_rate": 3.808187134502924e-05, + "loss": 3.1788, + "step": 15744 + }, + { + "epoch": 0.8286842105263158, + "grad_norm": 0.53125, + "learning_rate": 3.80701754385965e-05, + "loss": 2.8737, + "step": 15745 + }, + { + "epoch": 0.8287368421052631, + "grad_norm": 0.4609375, + "learning_rate": 3.805847953216374e-05, + "loss": 3.2151, + "step": 15746 + }, + { + "epoch": 0.8287894736842105, + "grad_norm": 0.54296875, + "learning_rate": 3.8046783625731e-05, + "loss": 3.3655, + "step": 15747 + }, + { + "epoch": 0.8288421052631579, + "grad_norm": 0.5078125, + "learning_rate": 3.8035087719298247e-05, + "loss": 3.407, + "step": 15748 + }, + { + "epoch": 0.8288947368421052, + "grad_norm": 0.5078125, + "learning_rate": 3.80233918128655e-05, + "loss": 3.462, + "step": 15749 + }, + { + "epoch": 0.8289473684210527, + "grad_norm": 0.52734375, + "learning_rate": 3.8011695906432746e-05, + "loss": 3.5903, + "step": 15750 + }, + { + "epoch": 0.829, + "grad_norm": 0.5078125, + "learning_rate": 3.8e-05, + "loss": 3.2784, + "step": 15751 + }, + { + "epoch": 0.8290526315789474, + "grad_norm": 0.50390625, + "learning_rate": 3.798830409356725e-05, + "loss": 3.3357, + "step": 15752 + }, + { + "epoch": 0.8291052631578948, + "grad_norm": 0.5234375, + "learning_rate": 3.797660818713451e-05, + "loss": 3.2233, + "step": 15753 + }, + { + "epoch": 0.8291578947368421, + "grad_norm": 0.494140625, + "learning_rate": 3.796491228070176e-05, + "loss": 3.0787, + "step": 15754 + }, + { + "epoch": 0.8292105263157895, + "grad_norm": 0.49609375, + "learning_rate": 3.795321637426901e-05, + "loss": 3.319, + "step": 15755 + }, + { + "epoch": 0.8292631578947368, + "grad_norm": 0.48046875, + "learning_rate": 3.7941520467836264e-05, + "loss": 3.1776, + "step": 15756 + }, + { + "epoch": 0.8293157894736842, + "grad_norm": 0.49609375, + "learning_rate": 3.7929824561403506e-05, + "loss": 3.2472, + "step": 15757 + }, + { + "epoch": 0.8293684210526315, + "grad_norm": 0.49609375, + "learning_rate": 3.791812865497076e-05, + "loss": 3.4208, + "step": 15758 + }, + { + "epoch": 0.829421052631579, + "grad_norm": 0.52734375, + "learning_rate": 3.790643274853801e-05, + "loss": 2.9533, + "step": 15759 + }, + { + "epoch": 0.8294736842105264, + "grad_norm": 0.546875, + "learning_rate": 3.789473684210527e-05, + "loss": 3.0619, + "step": 15760 + }, + { + "epoch": 0.8295263157894737, + "grad_norm": 0.5078125, + "learning_rate": 3.788304093567251e-05, + "loss": 3.5908, + "step": 15761 + }, + { + "epoch": 0.8295789473684211, + "grad_norm": 0.51953125, + "learning_rate": 3.787134502923977e-05, + "loss": 3.2729, + "step": 15762 + }, + { + "epoch": 0.8296315789473684, + "grad_norm": 0.50390625, + "learning_rate": 3.785964912280702e-05, + "loss": 3.1622, + "step": 15763 + }, + { + "epoch": 0.8296842105263158, + "grad_norm": 0.48828125, + "learning_rate": 3.7847953216374274e-05, + "loss": 3.5331, + "step": 15764 + }, + { + "epoch": 0.8297368421052631, + "grad_norm": 0.55078125, + "learning_rate": 3.783625730994152e-05, + "loss": 3.1428, + "step": 15765 + }, + { + "epoch": 0.8297894736842105, + "grad_norm": 0.6015625, + "learning_rate": 3.782456140350877e-05, + "loss": 3.0134, + "step": 15766 + }, + { + "epoch": 0.8298421052631579, + "grad_norm": 0.482421875, + "learning_rate": 3.781286549707603e-05, + "loss": 3.0046, + "step": 15767 + }, + { + "epoch": 0.8298947368421052, + "grad_norm": 0.53125, + "learning_rate": 3.780116959064328e-05, + "loss": 3.1393, + "step": 15768 + }, + { + "epoch": 0.8299473684210527, + "grad_norm": 0.5234375, + "learning_rate": 3.778947368421053e-05, + "loss": 3.4424, + "step": 15769 + }, + { + "epoch": 0.83, + "grad_norm": 0.4921875, + "learning_rate": 3.777777777777778e-05, + "loss": 3.9683, + "step": 15770 + }, + { + "epoch": 0.8300526315789474, + "grad_norm": 0.5, + "learning_rate": 3.7766081871345035e-05, + "loss": 3.3794, + "step": 15771 + }, + { + "epoch": 0.8301052631578948, + "grad_norm": 0.51171875, + "learning_rate": 3.7754385964912284e-05, + "loss": 2.9927, + "step": 15772 + }, + { + "epoch": 0.8301578947368421, + "grad_norm": 0.52734375, + "learning_rate": 3.7742690058479534e-05, + "loss": 3.3444, + "step": 15773 + }, + { + "epoch": 0.8302105263157895, + "grad_norm": 0.515625, + "learning_rate": 3.773099415204678e-05, + "loss": 3.4192, + "step": 15774 + }, + { + "epoch": 0.8302631578947368, + "grad_norm": 0.486328125, + "learning_rate": 3.771929824561404e-05, + "loss": 3.412, + "step": 15775 + }, + { + "epoch": 0.8303157894736842, + "grad_norm": 0.50390625, + "learning_rate": 3.770760233918129e-05, + "loss": 3.2372, + "step": 15776 + }, + { + "epoch": 0.8303684210526315, + "grad_norm": 0.4921875, + "learning_rate": 3.769590643274854e-05, + "loss": 3.6301, + "step": 15777 + }, + { + "epoch": 0.830421052631579, + "grad_norm": 0.498046875, + "learning_rate": 3.768421052631579e-05, + "loss": 3.0576, + "step": 15778 + }, + { + "epoch": 0.8304736842105264, + "grad_norm": 0.498046875, + "learning_rate": 3.7672514619883045e-05, + "loss": 3.3756, + "step": 15779 + }, + { + "epoch": 0.8305263157894737, + "grad_norm": 0.49609375, + "learning_rate": 3.7660818713450294e-05, + "loss": 3.403, + "step": 15780 + }, + { + "epoch": 0.8305789473684211, + "grad_norm": 0.498046875, + "learning_rate": 3.7649122807017544e-05, + "loss": 3.5968, + "step": 15781 + }, + { + "epoch": 0.8306315789473684, + "grad_norm": 0.494140625, + "learning_rate": 3.76374269005848e-05, + "loss": 3.0757, + "step": 15782 + }, + { + "epoch": 0.8306842105263158, + "grad_norm": 0.515625, + "learning_rate": 3.762573099415205e-05, + "loss": 3.4777, + "step": 15783 + }, + { + "epoch": 0.8307368421052631, + "grad_norm": 0.5, + "learning_rate": 3.76140350877193e-05, + "loss": 3.3062, + "step": 15784 + }, + { + "epoch": 0.8307894736842105, + "grad_norm": 0.498046875, + "learning_rate": 3.760233918128655e-05, + "loss": 2.7378, + "step": 15785 + }, + { + "epoch": 0.8308421052631579, + "grad_norm": 0.49609375, + "learning_rate": 3.7590643274853806e-05, + "loss": 3.6294, + "step": 15786 + }, + { + "epoch": 0.8308947368421052, + "grad_norm": 0.5390625, + "learning_rate": 3.7578947368421055e-05, + "loss": 3.3824, + "step": 15787 + }, + { + "epoch": 0.8309473684210527, + "grad_norm": 0.4921875, + "learning_rate": 3.7567251461988305e-05, + "loss": 3.3946, + "step": 15788 + }, + { + "epoch": 0.831, + "grad_norm": 0.48828125, + "learning_rate": 3.7555555555555554e-05, + "loss": 3.2796, + "step": 15789 + }, + { + "epoch": 0.8310526315789474, + "grad_norm": 0.52734375, + "learning_rate": 3.754385964912281e-05, + "loss": 3.6309, + "step": 15790 + }, + { + "epoch": 0.8311052631578948, + "grad_norm": 0.53515625, + "learning_rate": 3.753216374269006e-05, + "loss": 3.3969, + "step": 15791 + }, + { + "epoch": 0.8311578947368421, + "grad_norm": 0.4921875, + "learning_rate": 3.752046783625731e-05, + "loss": 2.8378, + "step": 15792 + }, + { + "epoch": 0.8312105263157895, + "grad_norm": 0.490234375, + "learning_rate": 3.7508771929824566e-05, + "loss": 3.6444, + "step": 15793 + }, + { + "epoch": 0.8312631578947368, + "grad_norm": 0.50390625, + "learning_rate": 3.7497076023391816e-05, + "loss": 2.8102, + "step": 15794 + }, + { + "epoch": 0.8313157894736842, + "grad_norm": 0.478515625, + "learning_rate": 3.7485380116959065e-05, + "loss": 3.0705, + "step": 15795 + }, + { + "epoch": 0.8313684210526315, + "grad_norm": 0.48828125, + "learning_rate": 3.7473684210526315e-05, + "loss": 3.666, + "step": 15796 + }, + { + "epoch": 0.831421052631579, + "grad_norm": 0.498046875, + "learning_rate": 3.746198830409357e-05, + "loss": 3.5571, + "step": 15797 + }, + { + "epoch": 0.8314736842105264, + "grad_norm": 0.490234375, + "learning_rate": 3.745029239766082e-05, + "loss": 3.2318, + "step": 15798 + }, + { + "epoch": 0.8315263157894737, + "grad_norm": 0.5, + "learning_rate": 3.743859649122807e-05, + "loss": 3.1962, + "step": 15799 + }, + { + "epoch": 0.8315789473684211, + "grad_norm": 0.48046875, + "learning_rate": 3.742690058479532e-05, + "loss": 3.0565, + "step": 15800 + }, + { + "epoch": 0.8316315789473684, + "grad_norm": 0.5078125, + "learning_rate": 3.7415204678362577e-05, + "loss": 3.4736, + "step": 15801 + }, + { + "epoch": 0.8316842105263158, + "grad_norm": 0.51171875, + "learning_rate": 3.7403508771929826e-05, + "loss": 3.7677, + "step": 15802 + }, + { + "epoch": 0.8317368421052631, + "grad_norm": 0.51171875, + "learning_rate": 3.7391812865497076e-05, + "loss": 3.3464, + "step": 15803 + }, + { + "epoch": 0.8317894736842105, + "grad_norm": 0.5, + "learning_rate": 3.738011695906433e-05, + "loss": 3.3575, + "step": 15804 + }, + { + "epoch": 0.8318421052631579, + "grad_norm": 0.5, + "learning_rate": 3.736842105263158e-05, + "loss": 3.6578, + "step": 15805 + }, + { + "epoch": 0.8318947368421052, + "grad_norm": 0.51171875, + "learning_rate": 3.735672514619883e-05, + "loss": 3.3162, + "step": 15806 + }, + { + "epoch": 0.8319473684210527, + "grad_norm": 0.5, + "learning_rate": 3.734502923976608e-05, + "loss": 3.3098, + "step": 15807 + }, + { + "epoch": 0.832, + "grad_norm": 0.51953125, + "learning_rate": 3.733333333333334e-05, + "loss": 3.1795, + "step": 15808 + }, + { + "epoch": 0.8320526315789474, + "grad_norm": 0.51171875, + "learning_rate": 3.732163742690059e-05, + "loss": 3.3831, + "step": 15809 + }, + { + "epoch": 0.8321052631578948, + "grad_norm": 0.52734375, + "learning_rate": 3.7309941520467836e-05, + "loss": 2.9341, + "step": 15810 + }, + { + "epoch": 0.8321578947368421, + "grad_norm": 0.51171875, + "learning_rate": 3.7298245614035086e-05, + "loss": 3.529, + "step": 15811 + }, + { + "epoch": 0.8322105263157895, + "grad_norm": 0.50390625, + "learning_rate": 3.728654970760234e-05, + "loss": 3.0318, + "step": 15812 + }, + { + "epoch": 0.8322631578947368, + "grad_norm": 0.51953125, + "learning_rate": 3.727485380116959e-05, + "loss": 3.164, + "step": 15813 + }, + { + "epoch": 0.8323157894736842, + "grad_norm": 0.51953125, + "learning_rate": 3.726315789473684e-05, + "loss": 3.384, + "step": 15814 + }, + { + "epoch": 0.8323684210526315, + "grad_norm": 0.50390625, + "learning_rate": 3.725146198830409e-05, + "loss": 3.4455, + "step": 15815 + }, + { + "epoch": 0.832421052631579, + "grad_norm": 0.51171875, + "learning_rate": 3.723976608187135e-05, + "loss": 2.976, + "step": 15816 + }, + { + "epoch": 0.8324736842105264, + "grad_norm": 0.46875, + "learning_rate": 3.7228070175438604e-05, + "loss": 2.8126, + "step": 15817 + }, + { + "epoch": 0.8325263157894737, + "grad_norm": 0.51953125, + "learning_rate": 3.721637426900585e-05, + "loss": 3.1021, + "step": 15818 + }, + { + "epoch": 0.8325789473684211, + "grad_norm": 0.50390625, + "learning_rate": 3.72046783625731e-05, + "loss": 3.3329, + "step": 15819 + }, + { + "epoch": 0.8326315789473684, + "grad_norm": 0.53515625, + "learning_rate": 3.719298245614035e-05, + "loss": 2.7331, + "step": 15820 + }, + { + "epoch": 0.8326842105263158, + "grad_norm": 0.50390625, + "learning_rate": 3.718128654970761e-05, + "loss": 3.0744, + "step": 15821 + }, + { + "epoch": 0.8327368421052631, + "grad_norm": 0.474609375, + "learning_rate": 3.716959064327485e-05, + "loss": 2.9419, + "step": 15822 + }, + { + "epoch": 0.8327894736842105, + "grad_norm": 0.478515625, + "learning_rate": 3.715789473684211e-05, + "loss": 3.2847, + "step": 15823 + }, + { + "epoch": 0.8328421052631579, + "grad_norm": 0.482421875, + "learning_rate": 3.714619883040936e-05, + "loss": 3.4144, + "step": 15824 + }, + { + "epoch": 0.8328947368421052, + "grad_norm": 0.50390625, + "learning_rate": 3.713450292397661e-05, + "loss": 3.3506, + "step": 15825 + }, + { + "epoch": 0.8329473684210527, + "grad_norm": 0.50390625, + "learning_rate": 3.712280701754386e-05, + "loss": 3.6138, + "step": 15826 + }, + { + "epoch": 0.833, + "grad_norm": 0.490234375, + "learning_rate": 3.7111111111111113e-05, + "loss": 3.1302, + "step": 15827 + }, + { + "epoch": 0.8330526315789474, + "grad_norm": 0.5625, + "learning_rate": 3.709941520467836e-05, + "loss": 3.3994, + "step": 15828 + }, + { + "epoch": 0.8331052631578947, + "grad_norm": 0.50390625, + "learning_rate": 3.708771929824561e-05, + "loss": 2.7485, + "step": 15829 + }, + { + "epoch": 0.8331578947368421, + "grad_norm": 0.4921875, + "learning_rate": 3.707602339181287e-05, + "loss": 3.0041, + "step": 15830 + }, + { + "epoch": 0.8332105263157895, + "grad_norm": 0.5078125, + "learning_rate": 3.706432748538012e-05, + "loss": 3.4569, + "step": 15831 + }, + { + "epoch": 0.8332631578947368, + "grad_norm": 0.4921875, + "learning_rate": 3.7052631578947375e-05, + "loss": 3.6206, + "step": 15832 + }, + { + "epoch": 0.8333157894736842, + "grad_norm": 0.50390625, + "learning_rate": 3.704093567251462e-05, + "loss": 3.3488, + "step": 15833 + }, + { + "epoch": 0.8333684210526315, + "grad_norm": 0.50390625, + "learning_rate": 3.7029239766081874e-05, + "loss": 3.1795, + "step": 15834 + }, + { + "epoch": 0.833421052631579, + "grad_norm": 0.5234375, + "learning_rate": 3.7017543859649124e-05, + "loss": 3.064, + "step": 15835 + }, + { + "epoch": 0.8334736842105264, + "grad_norm": 0.494140625, + "learning_rate": 3.700584795321638e-05, + "loss": 3.3996, + "step": 15836 + }, + { + "epoch": 0.8335263157894737, + "grad_norm": 0.5, + "learning_rate": 3.699415204678362e-05, + "loss": 3.0939, + "step": 15837 + }, + { + "epoch": 0.8335789473684211, + "grad_norm": 0.490234375, + "learning_rate": 3.698245614035088e-05, + "loss": 3.3909, + "step": 15838 + }, + { + "epoch": 0.8336315789473684, + "grad_norm": 0.482421875, + "learning_rate": 3.697076023391813e-05, + "loss": 2.9832, + "step": 15839 + }, + { + "epoch": 0.8336842105263158, + "grad_norm": 0.498046875, + "learning_rate": 3.6959064327485385e-05, + "loss": 3.1938, + "step": 15840 + }, + { + "epoch": 0.8337368421052631, + "grad_norm": 0.50390625, + "learning_rate": 3.6947368421052635e-05, + "loss": 3.2383, + "step": 15841 + }, + { + "epoch": 0.8337894736842105, + "grad_norm": 0.4765625, + "learning_rate": 3.6935672514619884e-05, + "loss": 2.7729, + "step": 15842 + }, + { + "epoch": 0.8338421052631579, + "grad_norm": 0.515625, + "learning_rate": 3.692397660818714e-05, + "loss": 3.6187, + "step": 15843 + }, + { + "epoch": 0.8338947368421052, + "grad_norm": 0.49609375, + "learning_rate": 3.691228070175439e-05, + "loss": 3.3678, + "step": 15844 + }, + { + "epoch": 0.8339473684210527, + "grad_norm": 0.5078125, + "learning_rate": 3.690058479532164e-05, + "loss": 3.3769, + "step": 15845 + }, + { + "epoch": 0.834, + "grad_norm": 0.48828125, + "learning_rate": 3.688888888888889e-05, + "loss": 3.4905, + "step": 15846 + }, + { + "epoch": 0.8340526315789474, + "grad_norm": 0.482421875, + "learning_rate": 3.6877192982456146e-05, + "loss": 3.7071, + "step": 15847 + }, + { + "epoch": 0.8341052631578947, + "grad_norm": 0.4921875, + "learning_rate": 3.686549707602339e-05, + "loss": 3.1904, + "step": 15848 + }, + { + "epoch": 0.8341578947368421, + "grad_norm": 0.47265625, + "learning_rate": 3.6853801169590645e-05, + "loss": 2.8759, + "step": 15849 + }, + { + "epoch": 0.8342105263157895, + "grad_norm": 0.4921875, + "learning_rate": 3.6842105263157895e-05, + "loss": 3.405, + "step": 15850 + }, + { + "epoch": 0.8342631578947368, + "grad_norm": 0.490234375, + "learning_rate": 3.683040935672515e-05, + "loss": 3.2465, + "step": 15851 + }, + { + "epoch": 0.8343157894736842, + "grad_norm": 0.50390625, + "learning_rate": 3.6818713450292394e-05, + "loss": 3.0229, + "step": 15852 + }, + { + "epoch": 0.8343684210526315, + "grad_norm": 0.51953125, + "learning_rate": 3.680701754385965e-05, + "loss": 2.8644, + "step": 15853 + }, + { + "epoch": 0.834421052631579, + "grad_norm": 0.48046875, + "learning_rate": 3.679532163742691e-05, + "loss": 3.3034, + "step": 15854 + }, + { + "epoch": 0.8344736842105264, + "grad_norm": 0.51953125, + "learning_rate": 3.6783625730994156e-05, + "loss": 3.4259, + "step": 15855 + }, + { + "epoch": 0.8345263157894737, + "grad_norm": 0.51171875, + "learning_rate": 3.6771929824561406e-05, + "loss": 3.187, + "step": 15856 + }, + { + "epoch": 0.8345789473684211, + "grad_norm": 0.50390625, + "learning_rate": 3.6760233918128655e-05, + "loss": 2.9228, + "step": 15857 + }, + { + "epoch": 0.8346315789473684, + "grad_norm": 0.5, + "learning_rate": 3.674853801169591e-05, + "loss": 3.3636, + "step": 15858 + }, + { + "epoch": 0.8346842105263158, + "grad_norm": 0.515625, + "learning_rate": 3.673684210526316e-05, + "loss": 3.3356, + "step": 15859 + }, + { + "epoch": 0.8347368421052631, + "grad_norm": 0.455078125, + "learning_rate": 3.672514619883041e-05, + "loss": 2.8695, + "step": 15860 + }, + { + "epoch": 0.8347894736842105, + "grad_norm": 0.46484375, + "learning_rate": 3.671345029239766e-05, + "loss": 2.9133, + "step": 15861 + }, + { + "epoch": 0.8348421052631579, + "grad_norm": 0.482421875, + "learning_rate": 3.670175438596492e-05, + "loss": 3.6559, + "step": 15862 + }, + { + "epoch": 0.8348947368421052, + "grad_norm": 0.48828125, + "learning_rate": 3.6690058479532167e-05, + "loss": 3.2108, + "step": 15863 + }, + { + "epoch": 0.8349473684210527, + "grad_norm": 0.494140625, + "learning_rate": 3.6678362573099416e-05, + "loss": 3.312, + "step": 15864 + }, + { + "epoch": 0.835, + "grad_norm": 0.50390625, + "learning_rate": 3.6666666666666666e-05, + "loss": 3.2262, + "step": 15865 + }, + { + "epoch": 0.8350526315789474, + "grad_norm": 0.498046875, + "learning_rate": 3.665497076023392e-05, + "loss": 3.1933, + "step": 15866 + }, + { + "epoch": 0.8351052631578947, + "grad_norm": 0.48828125, + "learning_rate": 3.664327485380117e-05, + "loss": 3.3008, + "step": 15867 + }, + { + "epoch": 0.8351578947368421, + "grad_norm": 0.49609375, + "learning_rate": 3.663157894736842e-05, + "loss": 2.923, + "step": 15868 + }, + { + "epoch": 0.8352105263157895, + "grad_norm": 0.50390625, + "learning_rate": 3.661988304093568e-05, + "loss": 3.453, + "step": 15869 + }, + { + "epoch": 0.8352631578947368, + "grad_norm": 0.51171875, + "learning_rate": 3.660818713450293e-05, + "loss": 3.6341, + "step": 15870 + }, + { + "epoch": 0.8353157894736842, + "grad_norm": 0.5, + "learning_rate": 3.659649122807018e-05, + "loss": 3.1667, + "step": 15871 + }, + { + "epoch": 0.8353684210526315, + "grad_norm": 0.490234375, + "learning_rate": 3.6584795321637426e-05, + "loss": 3.2124, + "step": 15872 + }, + { + "epoch": 0.835421052631579, + "grad_norm": 0.5078125, + "learning_rate": 3.657309941520468e-05, + "loss": 3.4159, + "step": 15873 + }, + { + "epoch": 0.8354736842105264, + "grad_norm": 0.474609375, + "learning_rate": 3.656140350877193e-05, + "loss": 3.1574, + "step": 15874 + }, + { + "epoch": 0.8355263157894737, + "grad_norm": 0.50390625, + "learning_rate": 3.654970760233918e-05, + "loss": 3.4364, + "step": 15875 + }, + { + "epoch": 0.8355789473684211, + "grad_norm": 0.494140625, + "learning_rate": 3.653801169590643e-05, + "loss": 2.8465, + "step": 15876 + }, + { + "epoch": 0.8356315789473684, + "grad_norm": 0.51953125, + "learning_rate": 3.652631578947369e-05, + "loss": 3.5421, + "step": 15877 + }, + { + "epoch": 0.8356842105263158, + "grad_norm": 0.482421875, + "learning_rate": 3.651461988304094e-05, + "loss": 3.074, + "step": 15878 + }, + { + "epoch": 0.8357368421052631, + "grad_norm": 0.515625, + "learning_rate": 3.650292397660819e-05, + "loss": 3.3413, + "step": 15879 + }, + { + "epoch": 0.8357894736842105, + "grad_norm": 0.4765625, + "learning_rate": 3.6491228070175443e-05, + "loss": 3.5464, + "step": 15880 + }, + { + "epoch": 0.8358421052631579, + "grad_norm": 0.546875, + "learning_rate": 3.647953216374269e-05, + "loss": 3.1975, + "step": 15881 + }, + { + "epoch": 0.8358947368421052, + "grad_norm": 0.546875, + "learning_rate": 3.646783625730994e-05, + "loss": 2.6448, + "step": 15882 + }, + { + "epoch": 0.8359473684210527, + "grad_norm": 0.490234375, + "learning_rate": 3.645614035087719e-05, + "loss": 3.1912, + "step": 15883 + }, + { + "epoch": 0.836, + "grad_norm": 0.50390625, + "learning_rate": 3.644444444444445e-05, + "loss": 3.0267, + "step": 15884 + }, + { + "epoch": 0.8360526315789474, + "grad_norm": 0.490234375, + "learning_rate": 3.64327485380117e-05, + "loss": 2.9939, + "step": 15885 + }, + { + "epoch": 0.8361052631578947, + "grad_norm": 0.53125, + "learning_rate": 3.642105263157895e-05, + "loss": 3.4505, + "step": 15886 + }, + { + "epoch": 0.8361578947368421, + "grad_norm": 0.48046875, + "learning_rate": 3.64093567251462e-05, + "loss": 2.9825, + "step": 15887 + }, + { + "epoch": 0.8362105263157895, + "grad_norm": 0.4921875, + "learning_rate": 3.6397660818713454e-05, + "loss": 3.6131, + "step": 15888 + }, + { + "epoch": 0.8362631578947368, + "grad_norm": 0.49609375, + "learning_rate": 3.63859649122807e-05, + "loss": 3.231, + "step": 15889 + }, + { + "epoch": 0.8363157894736842, + "grad_norm": 0.50390625, + "learning_rate": 3.637426900584795e-05, + "loss": 3.225, + "step": 15890 + }, + { + "epoch": 0.8363684210526315, + "grad_norm": 0.4921875, + "learning_rate": 3.636257309941521e-05, + "loss": 3.4569, + "step": 15891 + }, + { + "epoch": 0.836421052631579, + "grad_norm": 0.54296875, + "learning_rate": 3.635087719298246e-05, + "loss": 3.144, + "step": 15892 + }, + { + "epoch": 0.8364736842105264, + "grad_norm": 0.51953125, + "learning_rate": 3.633918128654971e-05, + "loss": 3.7068, + "step": 15893 + }, + { + "epoch": 0.8365263157894737, + "grad_norm": 0.4921875, + "learning_rate": 3.632748538011696e-05, + "loss": 3.4943, + "step": 15894 + }, + { + "epoch": 0.8365789473684211, + "grad_norm": 0.5078125, + "learning_rate": 3.6315789473684214e-05, + "loss": 3.1216, + "step": 15895 + }, + { + "epoch": 0.8366315789473684, + "grad_norm": 0.490234375, + "learning_rate": 3.6304093567251464e-05, + "loss": 3.5671, + "step": 15896 + }, + { + "epoch": 0.8366842105263158, + "grad_norm": 0.50390625, + "learning_rate": 3.6292397660818714e-05, + "loss": 2.7807, + "step": 15897 + }, + { + "epoch": 0.8367368421052631, + "grad_norm": 0.484375, + "learning_rate": 3.628070175438596e-05, + "loss": 2.8178, + "step": 15898 + }, + { + "epoch": 0.8367894736842105, + "grad_norm": 0.5078125, + "learning_rate": 3.626900584795322e-05, + "loss": 3.1285, + "step": 15899 + }, + { + "epoch": 0.8368421052631579, + "grad_norm": 0.4921875, + "learning_rate": 3.625730994152047e-05, + "loss": 3.5499, + "step": 15900 + }, + { + "epoch": 0.8368947368421052, + "grad_norm": 0.59765625, + "learning_rate": 3.624561403508772e-05, + "loss": 2.853, + "step": 15901 + }, + { + "epoch": 0.8369473684210527, + "grad_norm": 0.53125, + "learning_rate": 3.623391812865497e-05, + "loss": 3.1134, + "step": 15902 + }, + { + "epoch": 0.837, + "grad_norm": 0.48828125, + "learning_rate": 3.6222222222222225e-05, + "loss": 3.1749, + "step": 15903 + }, + { + "epoch": 0.8370526315789474, + "grad_norm": 0.486328125, + "learning_rate": 3.621052631578948e-05, + "loss": 2.9598, + "step": 15904 + }, + { + "epoch": 0.8371052631578947, + "grad_norm": 0.486328125, + "learning_rate": 3.6198830409356724e-05, + "loss": 3.0117, + "step": 15905 + }, + { + "epoch": 0.8371578947368421, + "grad_norm": 0.5, + "learning_rate": 3.618713450292398e-05, + "loss": 3.156, + "step": 15906 + }, + { + "epoch": 0.8372105263157895, + "grad_norm": 0.5, + "learning_rate": 3.617543859649123e-05, + "loss": 3.3801, + "step": 15907 + }, + { + "epoch": 0.8372631578947368, + "grad_norm": 0.49609375, + "learning_rate": 3.6163742690058486e-05, + "loss": 2.8542, + "step": 15908 + }, + { + "epoch": 0.8373157894736842, + "grad_norm": 0.51171875, + "learning_rate": 3.615204678362573e-05, + "loss": 3.3097, + "step": 15909 + }, + { + "epoch": 0.8373684210526315, + "grad_norm": 0.498046875, + "learning_rate": 3.6140350877192985e-05, + "loss": 3.4951, + "step": 15910 + }, + { + "epoch": 0.837421052631579, + "grad_norm": 0.48828125, + "learning_rate": 3.6128654970760235e-05, + "loss": 3.5592, + "step": 15911 + }, + { + "epoch": 0.8374736842105264, + "grad_norm": 0.494140625, + "learning_rate": 3.6116959064327485e-05, + "loss": 3.2192, + "step": 15912 + }, + { + "epoch": 0.8375263157894737, + "grad_norm": 0.50390625, + "learning_rate": 3.6105263157894734e-05, + "loss": 3.1175, + "step": 15913 + }, + { + "epoch": 0.8375789473684211, + "grad_norm": 0.51171875, + "learning_rate": 3.609356725146199e-05, + "loss": 3.5284, + "step": 15914 + }, + { + "epoch": 0.8376315789473684, + "grad_norm": 0.50390625, + "learning_rate": 3.608187134502924e-05, + "loss": 3.031, + "step": 15915 + }, + { + "epoch": 0.8376842105263158, + "grad_norm": 0.4921875, + "learning_rate": 3.607017543859649e-05, + "loss": 2.9444, + "step": 15916 + }, + { + "epoch": 0.8377368421052631, + "grad_norm": 0.51171875, + "learning_rate": 3.6058479532163746e-05, + "loss": 3.1051, + "step": 15917 + }, + { + "epoch": 0.8377894736842105, + "grad_norm": 0.62109375, + "learning_rate": 3.6046783625730996e-05, + "loss": 3.1136, + "step": 15918 + }, + { + "epoch": 0.8378421052631579, + "grad_norm": 0.5234375, + "learning_rate": 3.603508771929825e-05, + "loss": 3.1314, + "step": 15919 + }, + { + "epoch": 0.8378947368421052, + "grad_norm": 0.486328125, + "learning_rate": 3.6023391812865495e-05, + "loss": 3.5038, + "step": 15920 + }, + { + "epoch": 0.8379473684210527, + "grad_norm": 0.48046875, + "learning_rate": 3.601169590643275e-05, + "loss": 3.1972, + "step": 15921 + }, + { + "epoch": 0.838, + "grad_norm": 0.482421875, + "learning_rate": 3.6e-05, + "loss": 3.2057, + "step": 15922 + }, + { + "epoch": 0.8380526315789474, + "grad_norm": 0.478515625, + "learning_rate": 3.598830409356726e-05, + "loss": 3.5946, + "step": 15923 + }, + { + "epoch": 0.8381052631578947, + "grad_norm": 0.4921875, + "learning_rate": 3.59766081871345e-05, + "loss": 3.556, + "step": 15924 + }, + { + "epoch": 0.8381578947368421, + "grad_norm": 0.51953125, + "learning_rate": 3.5964912280701756e-05, + "loss": 2.8931, + "step": 15925 + }, + { + "epoch": 0.8382105263157895, + "grad_norm": 0.49609375, + "learning_rate": 3.5953216374269006e-05, + "loss": 3.7881, + "step": 15926 + }, + { + "epoch": 0.8382631578947368, + "grad_norm": 0.515625, + "learning_rate": 3.594152046783626e-05, + "loss": 3.1284, + "step": 15927 + }, + { + "epoch": 0.8383157894736842, + "grad_norm": 0.4921875, + "learning_rate": 3.5929824561403505e-05, + "loss": 3.296, + "step": 15928 + }, + { + "epoch": 0.8383684210526315, + "grad_norm": 0.48046875, + "learning_rate": 3.591812865497076e-05, + "loss": 3.3818, + "step": 15929 + }, + { + "epoch": 0.838421052631579, + "grad_norm": 0.494140625, + "learning_rate": 3.590643274853802e-05, + "loss": 3.0946, + "step": 15930 + }, + { + "epoch": 0.8384736842105264, + "grad_norm": 0.546875, + "learning_rate": 3.589473684210527e-05, + "loss": 3.3859, + "step": 15931 + }, + { + "epoch": 0.8385263157894737, + "grad_norm": 0.50390625, + "learning_rate": 3.588304093567252e-05, + "loss": 3.2341, + "step": 15932 + }, + { + "epoch": 0.8385789473684211, + "grad_norm": 0.51953125, + "learning_rate": 3.587134502923977e-05, + "loss": 3.6497, + "step": 15933 + }, + { + "epoch": 0.8386315789473684, + "grad_norm": 0.515625, + "learning_rate": 3.585964912280702e-05, + "loss": 2.9873, + "step": 15934 + }, + { + "epoch": 0.8386842105263158, + "grad_norm": 0.5234375, + "learning_rate": 3.5847953216374266e-05, + "loss": 3.0507, + "step": 15935 + }, + { + "epoch": 0.8387368421052631, + "grad_norm": 0.482421875, + "learning_rate": 3.583625730994152e-05, + "loss": 3.2581, + "step": 15936 + }, + { + "epoch": 0.8387894736842105, + "grad_norm": 0.486328125, + "learning_rate": 3.582456140350877e-05, + "loss": 3.0186, + "step": 15937 + }, + { + "epoch": 0.8388421052631579, + "grad_norm": 0.490234375, + "learning_rate": 3.581286549707603e-05, + "loss": 3.6919, + "step": 15938 + }, + { + "epoch": 0.8388947368421052, + "grad_norm": 0.5078125, + "learning_rate": 3.580116959064327e-05, + "loss": 3.4752, + "step": 15939 + }, + { + "epoch": 0.8389473684210527, + "grad_norm": 0.53515625, + "learning_rate": 3.578947368421053e-05, + "loss": 3.0883, + "step": 15940 + }, + { + "epoch": 0.839, + "grad_norm": 0.474609375, + "learning_rate": 3.577777777777778e-05, + "loss": 3.4289, + "step": 15941 + }, + { + "epoch": 0.8390526315789474, + "grad_norm": 0.53125, + "learning_rate": 3.5766081871345033e-05, + "loss": 3.1197, + "step": 15942 + }, + { + "epoch": 0.8391052631578947, + "grad_norm": 0.4921875, + "learning_rate": 3.575438596491228e-05, + "loss": 2.7553, + "step": 15943 + }, + { + "epoch": 0.8391578947368421, + "grad_norm": 0.51953125, + "learning_rate": 3.574269005847953e-05, + "loss": 3.0309, + "step": 15944 + }, + { + "epoch": 0.8392105263157895, + "grad_norm": 0.48828125, + "learning_rate": 3.573099415204679e-05, + "loss": 3.4033, + "step": 15945 + }, + { + "epoch": 0.8392631578947368, + "grad_norm": 0.5390625, + "learning_rate": 3.571929824561404e-05, + "loss": 2.6145, + "step": 15946 + }, + { + "epoch": 0.8393157894736842, + "grad_norm": 0.51171875, + "learning_rate": 3.570760233918129e-05, + "loss": 3.4196, + "step": 15947 + }, + { + "epoch": 0.8393684210526315, + "grad_norm": 0.515625, + "learning_rate": 3.569590643274854e-05, + "loss": 3.0889, + "step": 15948 + }, + { + "epoch": 0.839421052631579, + "grad_norm": 0.50390625, + "learning_rate": 3.5684210526315794e-05, + "loss": 3.3931, + "step": 15949 + }, + { + "epoch": 0.8394736842105263, + "grad_norm": 0.494140625, + "learning_rate": 3.5672514619883044e-05, + "loss": 3.0066, + "step": 15950 + }, + { + "epoch": 0.8395263157894737, + "grad_norm": 0.498046875, + "learning_rate": 3.566081871345029e-05, + "loss": 3.1268, + "step": 15951 + }, + { + "epoch": 0.8395789473684211, + "grad_norm": 0.482421875, + "learning_rate": 3.564912280701754e-05, + "loss": 3.0633, + "step": 15952 + }, + { + "epoch": 0.8396315789473684, + "grad_norm": 0.490234375, + "learning_rate": 3.56374269005848e-05, + "loss": 3.7048, + "step": 15953 + }, + { + "epoch": 0.8396842105263158, + "grad_norm": 0.5, + "learning_rate": 3.562573099415205e-05, + "loss": 3.1581, + "step": 15954 + }, + { + "epoch": 0.8397368421052631, + "grad_norm": 0.5546875, + "learning_rate": 3.56140350877193e-05, + "loss": 3.802, + "step": 15955 + }, + { + "epoch": 0.8397894736842105, + "grad_norm": 0.51171875, + "learning_rate": 3.5602339181286555e-05, + "loss": 3.5708, + "step": 15956 + }, + { + "epoch": 0.839842105263158, + "grad_norm": 0.53125, + "learning_rate": 3.5590643274853804e-05, + "loss": 2.9369, + "step": 15957 + }, + { + "epoch": 0.8398947368421052, + "grad_norm": 0.51171875, + "learning_rate": 3.5578947368421054e-05, + "loss": 3.66, + "step": 15958 + }, + { + "epoch": 0.8399473684210527, + "grad_norm": 0.4921875, + "learning_rate": 3.5567251461988304e-05, + "loss": 3.1109, + "step": 15959 + }, + { + "epoch": 0.84, + "grad_norm": 0.54296875, + "learning_rate": 3.555555555555556e-05, + "loss": 3.1085, + "step": 15960 + }, + { + "epoch": 0.8400526315789474, + "grad_norm": 0.515625, + "learning_rate": 3.554385964912281e-05, + "loss": 3.2042, + "step": 15961 + }, + { + "epoch": 0.8401052631578947, + "grad_norm": 0.494140625, + "learning_rate": 3.553216374269006e-05, + "loss": 3.2834, + "step": 15962 + }, + { + "epoch": 0.8401578947368421, + "grad_norm": 0.51171875, + "learning_rate": 3.552046783625731e-05, + "loss": 3.1826, + "step": 15963 + }, + { + "epoch": 0.8402105263157895, + "grad_norm": 0.50390625, + "learning_rate": 3.5508771929824565e-05, + "loss": 3.5384, + "step": 15964 + }, + { + "epoch": 0.8402631578947368, + "grad_norm": 0.47265625, + "learning_rate": 3.5497076023391815e-05, + "loss": 3.1772, + "step": 15965 + }, + { + "epoch": 0.8403157894736842, + "grad_norm": 0.515625, + "learning_rate": 3.5485380116959064e-05, + "loss": 3.0085, + "step": 15966 + }, + { + "epoch": 0.8403684210526315, + "grad_norm": 0.51171875, + "learning_rate": 3.547368421052632e-05, + "loss": 3.2732, + "step": 15967 + }, + { + "epoch": 0.840421052631579, + "grad_norm": 0.48046875, + "learning_rate": 3.546198830409357e-05, + "loss": 3.3877, + "step": 15968 + }, + { + "epoch": 0.8404736842105263, + "grad_norm": 0.49609375, + "learning_rate": 3.545029239766082e-05, + "loss": 2.9851, + "step": 15969 + }, + { + "epoch": 0.8405263157894737, + "grad_norm": 0.5078125, + "learning_rate": 3.543859649122807e-05, + "loss": 3.8405, + "step": 15970 + }, + { + "epoch": 0.8405789473684211, + "grad_norm": 0.474609375, + "learning_rate": 3.5426900584795326e-05, + "loss": 2.9272, + "step": 15971 + }, + { + "epoch": 0.8406315789473684, + "grad_norm": 0.46875, + "learning_rate": 3.5415204678362575e-05, + "loss": 3.6455, + "step": 15972 + }, + { + "epoch": 0.8406842105263158, + "grad_norm": 0.52734375, + "learning_rate": 3.5403508771929825e-05, + "loss": 3.2978, + "step": 15973 + }, + { + "epoch": 0.8407368421052631, + "grad_norm": 0.5703125, + "learning_rate": 3.5391812865497075e-05, + "loss": 3.4948, + "step": 15974 + }, + { + "epoch": 0.8407894736842105, + "grad_norm": 0.50390625, + "learning_rate": 3.538011695906433e-05, + "loss": 3.5177, + "step": 15975 + }, + { + "epoch": 0.840842105263158, + "grad_norm": 0.5546875, + "learning_rate": 3.536842105263158e-05, + "loss": 3.2993, + "step": 15976 + }, + { + "epoch": 0.8408947368421053, + "grad_norm": 0.5234375, + "learning_rate": 3.535672514619883e-05, + "loss": 3.2147, + "step": 15977 + }, + { + "epoch": 0.8409473684210527, + "grad_norm": 0.5078125, + "learning_rate": 3.534502923976608e-05, + "loss": 3.6459, + "step": 15978 + }, + { + "epoch": 0.841, + "grad_norm": 0.53515625, + "learning_rate": 3.5333333333333336e-05, + "loss": 2.8027, + "step": 15979 + }, + { + "epoch": 0.8410526315789474, + "grad_norm": 0.49609375, + "learning_rate": 3.5321637426900586e-05, + "loss": 2.8598, + "step": 15980 + }, + { + "epoch": 0.8411052631578947, + "grad_norm": 0.51171875, + "learning_rate": 3.5309941520467835e-05, + "loss": 3.6817, + "step": 15981 + }, + { + "epoch": 0.8411578947368421, + "grad_norm": 0.5078125, + "learning_rate": 3.529824561403509e-05, + "loss": 3.2576, + "step": 15982 + }, + { + "epoch": 0.8412105263157895, + "grad_norm": 0.50390625, + "learning_rate": 3.528654970760234e-05, + "loss": 3.6508, + "step": 15983 + }, + { + "epoch": 0.8412631578947368, + "grad_norm": 0.50390625, + "learning_rate": 3.527485380116959e-05, + "loss": 3.4374, + "step": 15984 + }, + { + "epoch": 0.8413157894736842, + "grad_norm": 0.50390625, + "learning_rate": 3.526315789473684e-05, + "loss": 2.9762, + "step": 15985 + }, + { + "epoch": 0.8413684210526315, + "grad_norm": 0.486328125, + "learning_rate": 3.52514619883041e-05, + "loss": 3.2236, + "step": 15986 + }, + { + "epoch": 0.841421052631579, + "grad_norm": 0.498046875, + "learning_rate": 3.5239766081871346e-05, + "loss": 3.1005, + "step": 15987 + }, + { + "epoch": 0.8414736842105263, + "grad_norm": 0.498046875, + "learning_rate": 3.5228070175438596e-05, + "loss": 3.6378, + "step": 15988 + }, + { + "epoch": 0.8415263157894737, + "grad_norm": 0.4921875, + "learning_rate": 3.5216374269005846e-05, + "loss": 3.2978, + "step": 15989 + }, + { + "epoch": 0.8415789473684211, + "grad_norm": 0.486328125, + "learning_rate": 3.52046783625731e-05, + "loss": 3.5627, + "step": 15990 + }, + { + "epoch": 0.8416315789473684, + "grad_norm": 0.4921875, + "learning_rate": 3.519298245614035e-05, + "loss": 3.0722, + "step": 15991 + }, + { + "epoch": 0.8416842105263158, + "grad_norm": 0.54296875, + "learning_rate": 3.51812865497076e-05, + "loss": 3.1928, + "step": 15992 + }, + { + "epoch": 0.8417368421052631, + "grad_norm": 0.515625, + "learning_rate": 3.516959064327486e-05, + "loss": 3.2839, + "step": 15993 + }, + { + "epoch": 0.8417894736842105, + "grad_norm": 0.50390625, + "learning_rate": 3.515789473684211e-05, + "loss": 3.67, + "step": 15994 + }, + { + "epoch": 0.841842105263158, + "grad_norm": 0.53125, + "learning_rate": 3.5146198830409363e-05, + "loss": 3.2063, + "step": 15995 + }, + { + "epoch": 0.8418947368421053, + "grad_norm": 0.52734375, + "learning_rate": 3.5134502923976606e-05, + "loss": 2.9966, + "step": 15996 + }, + { + "epoch": 0.8419473684210527, + "grad_norm": 0.54296875, + "learning_rate": 3.512280701754386e-05, + "loss": 3.6409, + "step": 15997 + }, + { + "epoch": 0.842, + "grad_norm": 0.494140625, + "learning_rate": 3.511111111111111e-05, + "loss": 3.4022, + "step": 15998 + }, + { + "epoch": 0.8420526315789474, + "grad_norm": 0.478515625, + "learning_rate": 3.509941520467837e-05, + "loss": 3.2007, + "step": 15999 + }, + { + "epoch": 0.8421052631578947, + "grad_norm": 0.5, + "learning_rate": 3.508771929824561e-05, + "loss": 3.1864, + "step": 16000 + }, + { + "epoch": 0.8421578947368421, + "grad_norm": 0.52734375, + "learning_rate": 3.507602339181287e-05, + "loss": 3.1222, + "step": 16001 + }, + { + "epoch": 0.8422105263157895, + "grad_norm": 0.484375, + "learning_rate": 3.506432748538012e-05, + "loss": 3.2016, + "step": 16002 + }, + { + "epoch": 0.8422631578947368, + "grad_norm": 0.458984375, + "learning_rate": 3.505263157894737e-05, + "loss": 3.0141, + "step": 16003 + }, + { + "epoch": 0.8423157894736842, + "grad_norm": 0.51953125, + "learning_rate": 3.504093567251462e-05, + "loss": 3.0305, + "step": 16004 + }, + { + "epoch": 0.8423684210526315, + "grad_norm": 0.48828125, + "learning_rate": 3.502923976608187e-05, + "loss": 3.5575, + "step": 16005 + }, + { + "epoch": 0.842421052631579, + "grad_norm": 0.50390625, + "learning_rate": 3.501754385964913e-05, + "loss": 3.2627, + "step": 16006 + }, + { + "epoch": 0.8424736842105263, + "grad_norm": 0.51171875, + "learning_rate": 3.500584795321637e-05, + "loss": 3.6907, + "step": 16007 + }, + { + "epoch": 0.8425263157894737, + "grad_norm": 0.5078125, + "learning_rate": 3.499415204678363e-05, + "loss": 2.9136, + "step": 16008 + }, + { + "epoch": 0.8425789473684211, + "grad_norm": 0.498046875, + "learning_rate": 3.498245614035088e-05, + "loss": 3.1508, + "step": 16009 + }, + { + "epoch": 0.8426315789473684, + "grad_norm": 0.482421875, + "learning_rate": 3.4970760233918134e-05, + "loss": 3.0747, + "step": 16010 + }, + { + "epoch": 0.8426842105263158, + "grad_norm": 0.482421875, + "learning_rate": 3.495906432748538e-05, + "loss": 3.3582, + "step": 16011 + }, + { + "epoch": 0.8427368421052631, + "grad_norm": 0.484375, + "learning_rate": 3.4947368421052634e-05, + "loss": 3.1569, + "step": 16012 + }, + { + "epoch": 0.8427894736842105, + "grad_norm": 0.4765625, + "learning_rate": 3.493567251461988e-05, + "loss": 3.4697, + "step": 16013 + }, + { + "epoch": 0.842842105263158, + "grad_norm": 0.490234375, + "learning_rate": 3.492397660818714e-05, + "loss": 3.3818, + "step": 16014 + }, + { + "epoch": 0.8428947368421053, + "grad_norm": 0.44921875, + "learning_rate": 3.491228070175438e-05, + "loss": 2.809, + "step": 16015 + }, + { + "epoch": 0.8429473684210527, + "grad_norm": 0.51171875, + "learning_rate": 3.490058479532164e-05, + "loss": 3.0897, + "step": 16016 + }, + { + "epoch": 0.843, + "grad_norm": 0.478515625, + "learning_rate": 3.4888888888888895e-05, + "loss": 2.9305, + "step": 16017 + }, + { + "epoch": 0.8430526315789474, + "grad_norm": 0.55078125, + "learning_rate": 3.4877192982456145e-05, + "loss": 3.1241, + "step": 16018 + }, + { + "epoch": 0.8431052631578947, + "grad_norm": 0.54296875, + "learning_rate": 3.4865497076023394e-05, + "loss": 3.0017, + "step": 16019 + }, + { + "epoch": 0.8431578947368421, + "grad_norm": 0.486328125, + "learning_rate": 3.4853801169590644e-05, + "loss": 3.4044, + "step": 16020 + }, + { + "epoch": 0.8432105263157895, + "grad_norm": 0.5234375, + "learning_rate": 3.48421052631579e-05, + "loss": 3.2844, + "step": 16021 + }, + { + "epoch": 0.8432631578947368, + "grad_norm": 0.5078125, + "learning_rate": 3.483040935672515e-05, + "loss": 3.0362, + "step": 16022 + }, + { + "epoch": 0.8433157894736842, + "grad_norm": 0.478515625, + "learning_rate": 3.48187134502924e-05, + "loss": 3.2715, + "step": 16023 + }, + { + "epoch": 0.8433684210526315, + "grad_norm": 0.486328125, + "learning_rate": 3.480701754385965e-05, + "loss": 2.977, + "step": 16024 + }, + { + "epoch": 0.843421052631579, + "grad_norm": 0.482421875, + "learning_rate": 3.4795321637426905e-05, + "loss": 3.3665, + "step": 16025 + }, + { + "epoch": 0.8434736842105263, + "grad_norm": 0.494140625, + "learning_rate": 3.478362573099415e-05, + "loss": 3.1866, + "step": 16026 + }, + { + "epoch": 0.8435263157894737, + "grad_norm": 0.5078125, + "learning_rate": 3.4771929824561405e-05, + "loss": 3.5795, + "step": 16027 + }, + { + "epoch": 0.8435789473684211, + "grad_norm": 0.49609375, + "learning_rate": 3.4760233918128654e-05, + "loss": 3.125, + "step": 16028 + }, + { + "epoch": 0.8436315789473684, + "grad_norm": 0.515625, + "learning_rate": 3.474853801169591e-05, + "loss": 2.8124, + "step": 16029 + }, + { + "epoch": 0.8436842105263158, + "grad_norm": 0.515625, + "learning_rate": 3.473684210526316e-05, + "loss": 3.3742, + "step": 16030 + }, + { + "epoch": 0.8437368421052631, + "grad_norm": 0.482421875, + "learning_rate": 3.472514619883041e-05, + "loss": 3.1777, + "step": 16031 + }, + { + "epoch": 0.8437894736842105, + "grad_norm": 0.5546875, + "learning_rate": 3.4713450292397666e-05, + "loss": 3.2429, + "step": 16032 + }, + { + "epoch": 0.843842105263158, + "grad_norm": 0.51171875, + "learning_rate": 3.4701754385964916e-05, + "loss": 3.5365, + "step": 16033 + }, + { + "epoch": 0.8438947368421053, + "grad_norm": 0.490234375, + "learning_rate": 3.4690058479532165e-05, + "loss": 3.3845, + "step": 16034 + }, + { + "epoch": 0.8439473684210527, + "grad_norm": 0.4921875, + "learning_rate": 3.4678362573099415e-05, + "loss": 3.4756, + "step": 16035 + }, + { + "epoch": 0.844, + "grad_norm": 0.494140625, + "learning_rate": 3.466666666666667e-05, + "loss": 3.3968, + "step": 16036 + }, + { + "epoch": 0.8440526315789474, + "grad_norm": 0.53125, + "learning_rate": 3.465497076023392e-05, + "loss": 2.986, + "step": 16037 + }, + { + "epoch": 0.8441052631578947, + "grad_norm": 0.5, + "learning_rate": 3.464327485380117e-05, + "loss": 3.4886, + "step": 16038 + }, + { + "epoch": 0.8441578947368421, + "grad_norm": 0.48046875, + "learning_rate": 3.463157894736842e-05, + "loss": 3.4683, + "step": 16039 + }, + { + "epoch": 0.8442105263157895, + "grad_norm": 0.484375, + "learning_rate": 3.4619883040935676e-05, + "loss": 3.3178, + "step": 16040 + }, + { + "epoch": 0.8442631578947368, + "grad_norm": 0.5, + "learning_rate": 3.4608187134502926e-05, + "loss": 2.8481, + "step": 16041 + }, + { + "epoch": 0.8443157894736842, + "grad_norm": 0.51171875, + "learning_rate": 3.4596491228070176e-05, + "loss": 3.2473, + "step": 16042 + }, + { + "epoch": 0.8443684210526315, + "grad_norm": 0.5078125, + "learning_rate": 3.458479532163743e-05, + "loss": 3.5413, + "step": 16043 + }, + { + "epoch": 0.844421052631579, + "grad_norm": 0.48828125, + "learning_rate": 3.457309941520468e-05, + "loss": 3.3273, + "step": 16044 + }, + { + "epoch": 0.8444736842105263, + "grad_norm": 0.48046875, + "learning_rate": 3.456140350877193e-05, + "loss": 2.9311, + "step": 16045 + }, + { + "epoch": 0.8445263157894737, + "grad_norm": 0.546875, + "learning_rate": 3.454970760233918e-05, + "loss": 2.9358, + "step": 16046 + }, + { + "epoch": 0.8445789473684211, + "grad_norm": 0.48828125, + "learning_rate": 3.453801169590644e-05, + "loss": 3.2705, + "step": 16047 + }, + { + "epoch": 0.8446315789473684, + "grad_norm": 0.5390625, + "learning_rate": 3.452631578947369e-05, + "loss": 3.0617, + "step": 16048 + }, + { + "epoch": 0.8446842105263158, + "grad_norm": 0.50390625, + "learning_rate": 3.4514619883040936e-05, + "loss": 3.4158, + "step": 16049 + }, + { + "epoch": 0.8447368421052631, + "grad_norm": 0.5078125, + "learning_rate": 3.4502923976608186e-05, + "loss": 3.2795, + "step": 16050 + }, + { + "epoch": 0.8447894736842105, + "grad_norm": 0.51953125, + "learning_rate": 3.449122807017544e-05, + "loss": 3.4, + "step": 16051 + }, + { + "epoch": 0.844842105263158, + "grad_norm": 0.486328125, + "learning_rate": 3.447953216374269e-05, + "loss": 3.3696, + "step": 16052 + }, + { + "epoch": 0.8448947368421053, + "grad_norm": 0.51171875, + "learning_rate": 3.446783625730994e-05, + "loss": 3.452, + "step": 16053 + }, + { + "epoch": 0.8449473684210527, + "grad_norm": 0.53125, + "learning_rate": 3.44561403508772e-05, + "loss": 2.9537, + "step": 16054 + }, + { + "epoch": 0.845, + "grad_norm": 0.4921875, + "learning_rate": 3.444444444444445e-05, + "loss": 3.4299, + "step": 16055 + }, + { + "epoch": 0.8450526315789474, + "grad_norm": 0.55859375, + "learning_rate": 3.44327485380117e-05, + "loss": 3.1392, + "step": 16056 + }, + { + "epoch": 0.8451052631578947, + "grad_norm": 0.4921875, + "learning_rate": 3.442105263157895e-05, + "loss": 3.5728, + "step": 16057 + }, + { + "epoch": 0.8451578947368421, + "grad_norm": 0.51171875, + "learning_rate": 3.44093567251462e-05, + "loss": 3.2054, + "step": 16058 + }, + { + "epoch": 0.8452105263157895, + "grad_norm": 0.484375, + "learning_rate": 3.439766081871345e-05, + "loss": 3.5177, + "step": 16059 + }, + { + "epoch": 0.8452631578947368, + "grad_norm": 0.515625, + "learning_rate": 3.43859649122807e-05, + "loss": 3.308, + "step": 16060 + }, + { + "epoch": 0.8453157894736842, + "grad_norm": 0.515625, + "learning_rate": 3.437426900584795e-05, + "loss": 3.5426, + "step": 16061 + }, + { + "epoch": 0.8453684210526315, + "grad_norm": 0.52734375, + "learning_rate": 3.436257309941521e-05, + "loss": 3.2797, + "step": 16062 + }, + { + "epoch": 0.845421052631579, + "grad_norm": 0.494140625, + "learning_rate": 3.435087719298246e-05, + "loss": 3.0439, + "step": 16063 + }, + { + "epoch": 0.8454736842105263, + "grad_norm": 0.48046875, + "learning_rate": 3.433918128654971e-05, + "loss": 3.3438, + "step": 16064 + }, + { + "epoch": 0.8455263157894737, + "grad_norm": 0.4921875, + "learning_rate": 3.432748538011696e-05, + "loss": 3.1235, + "step": 16065 + }, + { + "epoch": 0.8455789473684211, + "grad_norm": 0.490234375, + "learning_rate": 3.431578947368421e-05, + "loss": 3.2084, + "step": 16066 + }, + { + "epoch": 0.8456315789473684, + "grad_norm": 0.5078125, + "learning_rate": 3.430409356725147e-05, + "loss": 3.3608, + "step": 16067 + }, + { + "epoch": 0.8456842105263158, + "grad_norm": 0.482421875, + "learning_rate": 3.429239766081871e-05, + "loss": 3.2046, + "step": 16068 + }, + { + "epoch": 0.8457368421052631, + "grad_norm": 0.478515625, + "learning_rate": 3.428070175438597e-05, + "loss": 2.9659, + "step": 16069 + }, + { + "epoch": 0.8457894736842105, + "grad_norm": 0.49609375, + "learning_rate": 3.426900584795322e-05, + "loss": 3.2758, + "step": 16070 + }, + { + "epoch": 0.8458421052631578, + "grad_norm": 0.53125, + "learning_rate": 3.425730994152047e-05, + "loss": 3.2749, + "step": 16071 + }, + { + "epoch": 0.8458947368421053, + "grad_norm": 0.484375, + "learning_rate": 3.424561403508772e-05, + "loss": 3.5783, + "step": 16072 + }, + { + "epoch": 0.8459473684210527, + "grad_norm": 0.490234375, + "learning_rate": 3.4233918128654974e-05, + "loss": 3.2037, + "step": 16073 + }, + { + "epoch": 0.846, + "grad_norm": 0.51171875, + "learning_rate": 3.4222222222222224e-05, + "loss": 2.946, + "step": 16074 + }, + { + "epoch": 0.8460526315789474, + "grad_norm": 0.5, + "learning_rate": 3.421052631578947e-05, + "loss": 3.172, + "step": 16075 + }, + { + "epoch": 0.8461052631578947, + "grad_norm": 0.490234375, + "learning_rate": 3.419883040935672e-05, + "loss": 3.2785, + "step": 16076 + }, + { + "epoch": 0.8461578947368421, + "grad_norm": 0.484375, + "learning_rate": 3.418713450292398e-05, + "loss": 3.0388, + "step": 16077 + }, + { + "epoch": 0.8462105263157895, + "grad_norm": 0.50390625, + "learning_rate": 3.417543859649123e-05, + "loss": 3.0852, + "step": 16078 + }, + { + "epoch": 0.8462631578947368, + "grad_norm": 0.5078125, + "learning_rate": 3.416374269005848e-05, + "loss": 3.3779, + "step": 16079 + }, + { + "epoch": 0.8463157894736842, + "grad_norm": 0.498046875, + "learning_rate": 3.4152046783625735e-05, + "loss": 3.1636, + "step": 16080 + }, + { + "epoch": 0.8463684210526315, + "grad_norm": 0.52734375, + "learning_rate": 3.4140350877192984e-05, + "loss": 3.4887, + "step": 16081 + }, + { + "epoch": 0.846421052631579, + "grad_norm": 0.474609375, + "learning_rate": 3.412865497076024e-05, + "loss": 3.1774, + "step": 16082 + }, + { + "epoch": 0.8464736842105263, + "grad_norm": 0.494140625, + "learning_rate": 3.4116959064327483e-05, + "loss": 3.1914, + "step": 16083 + }, + { + "epoch": 0.8465263157894737, + "grad_norm": 0.53515625, + "learning_rate": 3.410526315789474e-05, + "loss": 3.2212, + "step": 16084 + }, + { + "epoch": 0.8465789473684211, + "grad_norm": 0.51171875, + "learning_rate": 3.409356725146199e-05, + "loss": 3.3158, + "step": 16085 + }, + { + "epoch": 0.8466315789473684, + "grad_norm": 0.51953125, + "learning_rate": 3.4081871345029246e-05, + "loss": 3.298, + "step": 16086 + }, + { + "epoch": 0.8466842105263158, + "grad_norm": 0.51953125, + "learning_rate": 3.407017543859649e-05, + "loss": 3.2901, + "step": 16087 + }, + { + "epoch": 0.8467368421052631, + "grad_norm": 0.515625, + "learning_rate": 3.4058479532163745e-05, + "loss": 3.045, + "step": 16088 + }, + { + "epoch": 0.8467894736842105, + "grad_norm": 0.55859375, + "learning_rate": 3.4046783625730995e-05, + "loss": 2.926, + "step": 16089 + }, + { + "epoch": 0.8468421052631578, + "grad_norm": 0.52734375, + "learning_rate": 3.403508771929825e-05, + "loss": 2.9299, + "step": 16090 + }, + { + "epoch": 0.8468947368421053, + "grad_norm": 0.470703125, + "learning_rate": 3.4023391812865494e-05, + "loss": 3.1534, + "step": 16091 + }, + { + "epoch": 0.8469473684210527, + "grad_norm": 0.5078125, + "learning_rate": 3.401169590643275e-05, + "loss": 3.5204, + "step": 16092 + }, + { + "epoch": 0.847, + "grad_norm": 0.50390625, + "learning_rate": 3.4000000000000007e-05, + "loss": 3.4864, + "step": 16093 + }, + { + "epoch": 0.8470526315789474, + "grad_norm": 0.47265625, + "learning_rate": 3.398830409356725e-05, + "loss": 3.1706, + "step": 16094 + }, + { + "epoch": 0.8471052631578947, + "grad_norm": 0.54296875, + "learning_rate": 3.3976608187134506e-05, + "loss": 3.2993, + "step": 16095 + }, + { + "epoch": 0.8471578947368421, + "grad_norm": 0.470703125, + "learning_rate": 3.3964912280701755e-05, + "loss": 2.8488, + "step": 16096 + }, + { + "epoch": 0.8472105263157895, + "grad_norm": 0.5234375, + "learning_rate": 3.395321637426901e-05, + "loss": 3.0207, + "step": 16097 + }, + { + "epoch": 0.8472631578947368, + "grad_norm": 0.47265625, + "learning_rate": 3.3941520467836254e-05, + "loss": 2.9014, + "step": 16098 + }, + { + "epoch": 0.8473157894736842, + "grad_norm": 0.51953125, + "learning_rate": 3.392982456140351e-05, + "loss": 3.3679, + "step": 16099 + }, + { + "epoch": 0.8473684210526315, + "grad_norm": 0.494140625, + "learning_rate": 3.391812865497076e-05, + "loss": 3.2201, + "step": 16100 + }, + { + "epoch": 0.847421052631579, + "grad_norm": 0.50390625, + "learning_rate": 3.390643274853802e-05, + "loss": 3.6344, + "step": 16101 + }, + { + "epoch": 0.8474736842105263, + "grad_norm": 0.48828125, + "learning_rate": 3.389473684210526e-05, + "loss": 3.0707, + "step": 16102 + }, + { + "epoch": 0.8475263157894737, + "grad_norm": 0.484375, + "learning_rate": 3.3883040935672516e-05, + "loss": 3.1065, + "step": 16103 + }, + { + "epoch": 0.8475789473684211, + "grad_norm": 0.48046875, + "learning_rate": 3.3871345029239766e-05, + "loss": 2.9947, + "step": 16104 + }, + { + "epoch": 0.8476315789473684, + "grad_norm": 0.484375, + "learning_rate": 3.385964912280702e-05, + "loss": 2.7745, + "step": 16105 + }, + { + "epoch": 0.8476842105263158, + "grad_norm": 0.48828125, + "learning_rate": 3.384795321637427e-05, + "loss": 3.4998, + "step": 16106 + }, + { + "epoch": 0.8477368421052631, + "grad_norm": 0.55078125, + "learning_rate": 3.383625730994152e-05, + "loss": 2.9544, + "step": 16107 + }, + { + "epoch": 0.8477894736842105, + "grad_norm": 0.54296875, + "learning_rate": 3.382456140350878e-05, + "loss": 3.142, + "step": 16108 + }, + { + "epoch": 0.8478421052631578, + "grad_norm": 0.5, + "learning_rate": 3.381286549707603e-05, + "loss": 3.5428, + "step": 16109 + }, + { + "epoch": 0.8478947368421053, + "grad_norm": 0.51953125, + "learning_rate": 3.380116959064328e-05, + "loss": 3.2874, + "step": 16110 + }, + { + "epoch": 0.8479473684210527, + "grad_norm": 0.50390625, + "learning_rate": 3.3789473684210526e-05, + "loss": 3.6435, + "step": 16111 + }, + { + "epoch": 0.848, + "grad_norm": 0.5078125, + "learning_rate": 3.377777777777778e-05, + "loss": 2.8968, + "step": 16112 + }, + { + "epoch": 0.8480526315789474, + "grad_norm": 0.486328125, + "learning_rate": 3.376608187134503e-05, + "loss": 3.447, + "step": 16113 + }, + { + "epoch": 0.8481052631578947, + "grad_norm": 0.494140625, + "learning_rate": 3.375438596491228e-05, + "loss": 3.2729, + "step": 16114 + }, + { + "epoch": 0.8481578947368421, + "grad_norm": 0.470703125, + "learning_rate": 3.374269005847953e-05, + "loss": 2.8924, + "step": 16115 + }, + { + "epoch": 0.8482105263157895, + "grad_norm": 0.515625, + "learning_rate": 3.373099415204679e-05, + "loss": 3.2258, + "step": 16116 + }, + { + "epoch": 0.8482631578947368, + "grad_norm": 0.515625, + "learning_rate": 3.371929824561404e-05, + "loss": 3.2064, + "step": 16117 + }, + { + "epoch": 0.8483157894736842, + "grad_norm": 0.5234375, + "learning_rate": 3.370760233918129e-05, + "loss": 3.1267, + "step": 16118 + }, + { + "epoch": 0.8483684210526315, + "grad_norm": 0.498046875, + "learning_rate": 3.369590643274854e-05, + "loss": 2.685, + "step": 16119 + }, + { + "epoch": 0.848421052631579, + "grad_norm": 0.5078125, + "learning_rate": 3.368421052631579e-05, + "loss": 3.4323, + "step": 16120 + }, + { + "epoch": 0.8484736842105263, + "grad_norm": 0.515625, + "learning_rate": 3.367251461988304e-05, + "loss": 3.2404, + "step": 16121 + }, + { + "epoch": 0.8485263157894737, + "grad_norm": 0.51171875, + "learning_rate": 3.366081871345029e-05, + "loss": 3.2441, + "step": 16122 + }, + { + "epoch": 0.8485789473684211, + "grad_norm": 0.5, + "learning_rate": 3.364912280701755e-05, + "loss": 3.3614, + "step": 16123 + }, + { + "epoch": 0.8486315789473684, + "grad_norm": 0.51953125, + "learning_rate": 3.36374269005848e-05, + "loss": 3.5418, + "step": 16124 + }, + { + "epoch": 0.8486842105263158, + "grad_norm": 0.5078125, + "learning_rate": 3.362573099415205e-05, + "loss": 3.1637, + "step": 16125 + }, + { + "epoch": 0.8487368421052631, + "grad_norm": 0.50390625, + "learning_rate": 3.36140350877193e-05, + "loss": 3.2327, + "step": 16126 + }, + { + "epoch": 0.8487894736842105, + "grad_norm": 0.50390625, + "learning_rate": 3.3602339181286554e-05, + "loss": 3.4233, + "step": 16127 + }, + { + "epoch": 0.8488421052631578, + "grad_norm": 0.49609375, + "learning_rate": 3.35906432748538e-05, + "loss": 3.6742, + "step": 16128 + }, + { + "epoch": 0.8488947368421053, + "grad_norm": 0.53125, + "learning_rate": 3.357894736842105e-05, + "loss": 2.9666, + "step": 16129 + }, + { + "epoch": 0.8489473684210527, + "grad_norm": 0.51171875, + "learning_rate": 3.356725146198831e-05, + "loss": 3.7085, + "step": 16130 + }, + { + "epoch": 0.849, + "grad_norm": 0.5, + "learning_rate": 3.355555555555556e-05, + "loss": 3.1522, + "step": 16131 + }, + { + "epoch": 0.8490526315789474, + "grad_norm": 0.486328125, + "learning_rate": 3.354385964912281e-05, + "loss": 3.4296, + "step": 16132 + }, + { + "epoch": 0.8491052631578947, + "grad_norm": 0.498046875, + "learning_rate": 3.353216374269006e-05, + "loss": 3.2432, + "step": 16133 + }, + { + "epoch": 0.8491578947368421, + "grad_norm": 0.48828125, + "learning_rate": 3.3520467836257314e-05, + "loss": 3.2252, + "step": 16134 + }, + { + "epoch": 0.8492105263157895, + "grad_norm": 0.55078125, + "learning_rate": 3.3508771929824564e-05, + "loss": 3.1425, + "step": 16135 + }, + { + "epoch": 0.8492631578947368, + "grad_norm": 0.48828125, + "learning_rate": 3.3497076023391814e-05, + "loss": 3.3502, + "step": 16136 + }, + { + "epoch": 0.8493157894736842, + "grad_norm": 0.50390625, + "learning_rate": 3.348538011695906e-05, + "loss": 3.1192, + "step": 16137 + }, + { + "epoch": 0.8493684210526316, + "grad_norm": 0.50390625, + "learning_rate": 3.347368421052632e-05, + "loss": 3.0464, + "step": 16138 + }, + { + "epoch": 0.849421052631579, + "grad_norm": 0.490234375, + "learning_rate": 3.346198830409357e-05, + "loss": 3.2097, + "step": 16139 + }, + { + "epoch": 0.8494736842105263, + "grad_norm": 0.486328125, + "learning_rate": 3.345029239766082e-05, + "loss": 3.115, + "step": 16140 + }, + { + "epoch": 0.8495263157894737, + "grad_norm": 0.51171875, + "learning_rate": 3.343859649122807e-05, + "loss": 3.4482, + "step": 16141 + }, + { + "epoch": 0.8495789473684211, + "grad_norm": 0.494140625, + "learning_rate": 3.3426900584795325e-05, + "loss": 2.8732, + "step": 16142 + }, + { + "epoch": 0.8496315789473684, + "grad_norm": 0.49609375, + "learning_rate": 3.3415204678362574e-05, + "loss": 3.6904, + "step": 16143 + }, + { + "epoch": 0.8496842105263158, + "grad_norm": 0.51171875, + "learning_rate": 3.3403508771929824e-05, + "loss": 3.3504, + "step": 16144 + }, + { + "epoch": 0.8497368421052631, + "grad_norm": 0.494140625, + "learning_rate": 3.339181286549708e-05, + "loss": 2.8573, + "step": 16145 + }, + { + "epoch": 0.8497894736842105, + "grad_norm": 0.47265625, + "learning_rate": 3.338011695906433e-05, + "loss": 2.9339, + "step": 16146 + }, + { + "epoch": 0.8498421052631578, + "grad_norm": 0.451171875, + "learning_rate": 3.336842105263158e-05, + "loss": 3.0351, + "step": 16147 + }, + { + "epoch": 0.8498947368421053, + "grad_norm": 0.478515625, + "learning_rate": 3.335672514619883e-05, + "loss": 3.3131, + "step": 16148 + }, + { + "epoch": 0.8499473684210527, + "grad_norm": 0.51953125, + "learning_rate": 3.3345029239766085e-05, + "loss": 2.9007, + "step": 16149 + }, + { + "epoch": 0.85, + "grad_norm": 0.478515625, + "learning_rate": 3.3333333333333335e-05, + "loss": 2.4504, + "step": 16150 + }, + { + "epoch": 0.8500526315789474, + "grad_norm": 0.51171875, + "learning_rate": 3.3321637426900585e-05, + "loss": 3.0107, + "step": 16151 + }, + { + "epoch": 0.8501052631578947, + "grad_norm": 0.52734375, + "learning_rate": 3.3309941520467834e-05, + "loss": 3.0124, + "step": 16152 + }, + { + "epoch": 0.8501578947368421, + "grad_norm": 0.51171875, + "learning_rate": 3.329824561403509e-05, + "loss": 2.8811, + "step": 16153 + }, + { + "epoch": 0.8502105263157895, + "grad_norm": 0.494140625, + "learning_rate": 3.328654970760234e-05, + "loss": 3.7997, + "step": 16154 + }, + { + "epoch": 0.8502631578947368, + "grad_norm": 0.56640625, + "learning_rate": 3.327485380116959e-05, + "loss": 3.0173, + "step": 16155 + }, + { + "epoch": 0.8503157894736842, + "grad_norm": 0.56640625, + "learning_rate": 3.3263157894736846e-05, + "loss": 3.0108, + "step": 16156 + }, + { + "epoch": 0.8503684210526316, + "grad_norm": 0.5, + "learning_rate": 3.3251461988304096e-05, + "loss": 3.2553, + "step": 16157 + }, + { + "epoch": 0.850421052631579, + "grad_norm": 0.5, + "learning_rate": 3.3239766081871345e-05, + "loss": 3.235, + "step": 16158 + }, + { + "epoch": 0.8504736842105263, + "grad_norm": 0.52734375, + "learning_rate": 3.3228070175438595e-05, + "loss": 2.7834, + "step": 16159 + }, + { + "epoch": 0.8505263157894737, + "grad_norm": 0.50390625, + "learning_rate": 3.321637426900585e-05, + "loss": 3.5732, + "step": 16160 + }, + { + "epoch": 0.8505789473684211, + "grad_norm": 0.51953125, + "learning_rate": 3.32046783625731e-05, + "loss": 3.2226, + "step": 16161 + }, + { + "epoch": 0.8506315789473684, + "grad_norm": 0.50390625, + "learning_rate": 3.319298245614035e-05, + "loss": 3.3053, + "step": 16162 + }, + { + "epoch": 0.8506842105263158, + "grad_norm": 0.484375, + "learning_rate": 3.31812865497076e-05, + "loss": 3.6412, + "step": 16163 + }, + { + "epoch": 0.8507368421052631, + "grad_norm": 0.52734375, + "learning_rate": 3.3169590643274856e-05, + "loss": 3.5027, + "step": 16164 + }, + { + "epoch": 0.8507894736842105, + "grad_norm": 0.59765625, + "learning_rate": 3.3157894736842106e-05, + "loss": 3.6656, + "step": 16165 + }, + { + "epoch": 0.8508421052631578, + "grad_norm": 0.515625, + "learning_rate": 3.3146198830409356e-05, + "loss": 3.4029, + "step": 16166 + }, + { + "epoch": 0.8508947368421053, + "grad_norm": 0.478515625, + "learning_rate": 3.313450292397661e-05, + "loss": 3.3732, + "step": 16167 + }, + { + "epoch": 0.8509473684210527, + "grad_norm": 0.470703125, + "learning_rate": 3.312280701754386e-05, + "loss": 2.9329, + "step": 16168 + }, + { + "epoch": 0.851, + "grad_norm": 1.34375, + "learning_rate": 3.311111111111112e-05, + "loss": 3.007, + "step": 16169 + }, + { + "epoch": 0.8510526315789474, + "grad_norm": 0.51953125, + "learning_rate": 3.309941520467836e-05, + "loss": 3.562, + "step": 16170 + }, + { + "epoch": 0.8511052631578947, + "grad_norm": 0.546875, + "learning_rate": 3.308771929824562e-05, + "loss": 2.9597, + "step": 16171 + }, + { + "epoch": 0.8511578947368421, + "grad_norm": 0.498046875, + "learning_rate": 3.307602339181287e-05, + "loss": 3.2552, + "step": 16172 + }, + { + "epoch": 0.8512105263157895, + "grad_norm": 0.474609375, + "learning_rate": 3.306432748538012e-05, + "loss": 2.7988, + "step": 16173 + }, + { + "epoch": 0.8512631578947368, + "grad_norm": 0.546875, + "learning_rate": 3.3052631578947366e-05, + "loss": 2.9755, + "step": 16174 + }, + { + "epoch": 0.8513157894736842, + "grad_norm": 0.494140625, + "learning_rate": 3.304093567251462e-05, + "loss": 3.1563, + "step": 16175 + }, + { + "epoch": 0.8513684210526316, + "grad_norm": 0.50390625, + "learning_rate": 3.302923976608187e-05, + "loss": 3.0077, + "step": 16176 + }, + { + "epoch": 0.851421052631579, + "grad_norm": 0.50390625, + "learning_rate": 3.301754385964913e-05, + "loss": 3.4169, + "step": 16177 + }, + { + "epoch": 0.8514736842105263, + "grad_norm": 0.486328125, + "learning_rate": 3.300584795321637e-05, + "loss": 2.8282, + "step": 16178 + }, + { + "epoch": 0.8515263157894737, + "grad_norm": 0.5234375, + "learning_rate": 3.299415204678363e-05, + "loss": 3.2405, + "step": 16179 + }, + { + "epoch": 0.8515789473684211, + "grad_norm": 0.515625, + "learning_rate": 3.2982456140350884e-05, + "loss": 2.9168, + "step": 16180 + }, + { + "epoch": 0.8516315789473684, + "grad_norm": 0.490234375, + "learning_rate": 3.2970760233918127e-05, + "loss": 3.1598, + "step": 16181 + }, + { + "epoch": 0.8516842105263158, + "grad_norm": 0.484375, + "learning_rate": 3.295906432748538e-05, + "loss": 3.4154, + "step": 16182 + }, + { + "epoch": 0.8517368421052631, + "grad_norm": 0.498046875, + "learning_rate": 3.294736842105263e-05, + "loss": 2.9787, + "step": 16183 + }, + { + "epoch": 0.8517894736842105, + "grad_norm": 0.52734375, + "learning_rate": 3.293567251461989e-05, + "loss": 3.181, + "step": 16184 + }, + { + "epoch": 0.8518421052631578, + "grad_norm": 0.48828125, + "learning_rate": 3.292397660818713e-05, + "loss": 3.0837, + "step": 16185 + }, + { + "epoch": 0.8518947368421053, + "grad_norm": 0.478515625, + "learning_rate": 3.291228070175439e-05, + "loss": 3.091, + "step": 16186 + }, + { + "epoch": 0.8519473684210527, + "grad_norm": 0.5078125, + "learning_rate": 3.290058479532164e-05, + "loss": 3.4448, + "step": 16187 + }, + { + "epoch": 0.852, + "grad_norm": 0.52734375, + "learning_rate": 3.2888888888888894e-05, + "loss": 3.5341, + "step": 16188 + }, + { + "epoch": 0.8520526315789474, + "grad_norm": 0.50390625, + "learning_rate": 3.287719298245614e-05, + "loss": 2.993, + "step": 16189 + }, + { + "epoch": 0.8521052631578947, + "grad_norm": 0.51171875, + "learning_rate": 3.286549707602339e-05, + "loss": 3.4067, + "step": 16190 + }, + { + "epoch": 0.8521578947368421, + "grad_norm": 0.47265625, + "learning_rate": 3.285380116959064e-05, + "loss": 3.0787, + "step": 16191 + }, + { + "epoch": 0.8522105263157894, + "grad_norm": 0.5078125, + "learning_rate": 3.28421052631579e-05, + "loss": 3.1902, + "step": 16192 + }, + { + "epoch": 0.8522631578947368, + "grad_norm": 0.470703125, + "learning_rate": 3.283040935672515e-05, + "loss": 3.1664, + "step": 16193 + }, + { + "epoch": 0.8523157894736842, + "grad_norm": 0.50390625, + "learning_rate": 3.28187134502924e-05, + "loss": 3.3529, + "step": 16194 + }, + { + "epoch": 0.8523684210526316, + "grad_norm": 0.490234375, + "learning_rate": 3.2807017543859655e-05, + "loss": 3.2659, + "step": 16195 + }, + { + "epoch": 0.852421052631579, + "grad_norm": 0.59765625, + "learning_rate": 3.2795321637426904e-05, + "loss": 3.1501, + "step": 16196 + }, + { + "epoch": 0.8524736842105263, + "grad_norm": 0.5078125, + "learning_rate": 3.2783625730994154e-05, + "loss": 3.0801, + "step": 16197 + }, + { + "epoch": 0.8525263157894737, + "grad_norm": 0.5546875, + "learning_rate": 3.2771929824561403e-05, + "loss": 2.8626, + "step": 16198 + }, + { + "epoch": 0.8525789473684211, + "grad_norm": 0.490234375, + "learning_rate": 3.276023391812866e-05, + "loss": 3.0896, + "step": 16199 + }, + { + "epoch": 0.8526315789473684, + "grad_norm": 0.4921875, + "learning_rate": 3.274853801169591e-05, + "loss": 3.2783, + "step": 16200 + }, + { + "epoch": 0.8526842105263158, + "grad_norm": 0.486328125, + "learning_rate": 3.273684210526316e-05, + "loss": 3.092, + "step": 16201 + }, + { + "epoch": 0.8527368421052631, + "grad_norm": 0.494140625, + "learning_rate": 3.272514619883041e-05, + "loss": 3.4882, + "step": 16202 + }, + { + "epoch": 0.8527894736842105, + "grad_norm": 0.462890625, + "learning_rate": 3.2713450292397665e-05, + "loss": 3.1498, + "step": 16203 + }, + { + "epoch": 0.8528421052631578, + "grad_norm": 0.5078125, + "learning_rate": 3.270175438596491e-05, + "loss": 3.471, + "step": 16204 + }, + { + "epoch": 0.8528947368421053, + "grad_norm": 0.609375, + "learning_rate": 3.2690058479532164e-05, + "loss": 3.1503, + "step": 16205 + }, + { + "epoch": 0.8529473684210527, + "grad_norm": 0.51171875, + "learning_rate": 3.267836257309942e-05, + "loss": 2.8603, + "step": 16206 + }, + { + "epoch": 0.853, + "grad_norm": 0.4921875, + "learning_rate": 3.266666666666667e-05, + "loss": 3.1893, + "step": 16207 + }, + { + "epoch": 0.8530526315789474, + "grad_norm": 0.494140625, + "learning_rate": 3.265497076023392e-05, + "loss": 3.3012, + "step": 16208 + }, + { + "epoch": 0.8531052631578947, + "grad_norm": 0.478515625, + "learning_rate": 3.264327485380117e-05, + "loss": 2.984, + "step": 16209 + }, + { + "epoch": 0.8531578947368421, + "grad_norm": 0.486328125, + "learning_rate": 3.2631578947368426e-05, + "loss": 3.5726, + "step": 16210 + }, + { + "epoch": 0.8532105263157894, + "grad_norm": 0.5390625, + "learning_rate": 3.2619883040935675e-05, + "loss": 3.304, + "step": 16211 + }, + { + "epoch": 0.8532631578947368, + "grad_norm": 0.515625, + "learning_rate": 3.2608187134502925e-05, + "loss": 2.9166, + "step": 16212 + }, + { + "epoch": 0.8533157894736842, + "grad_norm": 0.51953125, + "learning_rate": 3.2596491228070174e-05, + "loss": 2.7702, + "step": 16213 + }, + { + "epoch": 0.8533684210526316, + "grad_norm": 0.490234375, + "learning_rate": 3.258479532163743e-05, + "loss": 3.5392, + "step": 16214 + }, + { + "epoch": 0.853421052631579, + "grad_norm": 0.498046875, + "learning_rate": 3.257309941520468e-05, + "loss": 2.9803, + "step": 16215 + }, + { + "epoch": 0.8534736842105263, + "grad_norm": 0.515625, + "learning_rate": 3.256140350877193e-05, + "loss": 3.0742, + "step": 16216 + }, + { + "epoch": 0.8535263157894737, + "grad_norm": 0.49609375, + "learning_rate": 3.2549707602339186e-05, + "loss": 3.1333, + "step": 16217 + }, + { + "epoch": 0.8535789473684211, + "grad_norm": 0.48828125, + "learning_rate": 3.2538011695906436e-05, + "loss": 3.1401, + "step": 16218 + }, + { + "epoch": 0.8536315789473684, + "grad_norm": 0.498046875, + "learning_rate": 3.2526315789473686e-05, + "loss": 3.6367, + "step": 16219 + }, + { + "epoch": 0.8536842105263158, + "grad_norm": 0.5, + "learning_rate": 3.2514619883040935e-05, + "loss": 3.3791, + "step": 16220 + }, + { + "epoch": 0.8537368421052631, + "grad_norm": 0.51171875, + "learning_rate": 3.250292397660819e-05, + "loss": 3.5133, + "step": 16221 + }, + { + "epoch": 0.8537894736842105, + "grad_norm": 0.50390625, + "learning_rate": 3.249122807017544e-05, + "loss": 3.1371, + "step": 16222 + }, + { + "epoch": 0.8538421052631578, + "grad_norm": 0.4921875, + "learning_rate": 3.247953216374269e-05, + "loss": 3.2637, + "step": 16223 + }, + { + "epoch": 0.8538947368421053, + "grad_norm": 0.5078125, + "learning_rate": 3.246783625730994e-05, + "loss": 3.1687, + "step": 16224 + }, + { + "epoch": 0.8539473684210527, + "grad_norm": 0.50390625, + "learning_rate": 3.24561403508772e-05, + "loss": 3.3458, + "step": 16225 + }, + { + "epoch": 0.854, + "grad_norm": 0.5546875, + "learning_rate": 3.2444444444444446e-05, + "loss": 2.9924, + "step": 16226 + }, + { + "epoch": 0.8540526315789474, + "grad_norm": 0.455078125, + "learning_rate": 3.2432748538011696e-05, + "loss": 2.8607, + "step": 16227 + }, + { + "epoch": 0.8541052631578947, + "grad_norm": 0.53125, + "learning_rate": 3.2421052631578945e-05, + "loss": 3.3139, + "step": 16228 + }, + { + "epoch": 0.8541578947368421, + "grad_norm": 0.5, + "learning_rate": 3.24093567251462e-05, + "loss": 3.7788, + "step": 16229 + }, + { + "epoch": 0.8542105263157894, + "grad_norm": 0.5078125, + "learning_rate": 3.239766081871345e-05, + "loss": 2.9581, + "step": 16230 + }, + { + "epoch": 0.8542631578947368, + "grad_norm": 0.50390625, + "learning_rate": 3.23859649122807e-05, + "loss": 3.5226, + "step": 16231 + }, + { + "epoch": 0.8543157894736843, + "grad_norm": 0.5078125, + "learning_rate": 3.237426900584796e-05, + "loss": 3.1807, + "step": 16232 + }, + { + "epoch": 0.8543684210526316, + "grad_norm": 0.498046875, + "learning_rate": 3.236257309941521e-05, + "loss": 3.2423, + "step": 16233 + }, + { + "epoch": 0.854421052631579, + "grad_norm": 0.48828125, + "learning_rate": 3.2350877192982457e-05, + "loss": 3.2879, + "step": 16234 + }, + { + "epoch": 0.8544736842105263, + "grad_norm": 0.5, + "learning_rate": 3.2339181286549706e-05, + "loss": 2.9098, + "step": 16235 + }, + { + "epoch": 0.8545263157894737, + "grad_norm": 0.484375, + "learning_rate": 3.232748538011696e-05, + "loss": 3.0358, + "step": 16236 + }, + { + "epoch": 0.8545789473684211, + "grad_norm": 0.5234375, + "learning_rate": 3.231578947368421e-05, + "loss": 3.9981, + "step": 16237 + }, + { + "epoch": 0.8546315789473684, + "grad_norm": 0.482421875, + "learning_rate": 3.230409356725146e-05, + "loss": 3.5293, + "step": 16238 + }, + { + "epoch": 0.8546842105263158, + "grad_norm": 0.515625, + "learning_rate": 3.229239766081871e-05, + "loss": 3.4421, + "step": 16239 + }, + { + "epoch": 0.8547368421052631, + "grad_norm": 0.5, + "learning_rate": 3.228070175438597e-05, + "loss": 3.1508, + "step": 16240 + }, + { + "epoch": 0.8547894736842105, + "grad_norm": 0.498046875, + "learning_rate": 3.226900584795322e-05, + "loss": 3.5089, + "step": 16241 + }, + { + "epoch": 0.8548421052631578, + "grad_norm": 0.482421875, + "learning_rate": 3.225730994152047e-05, + "loss": 3.2383, + "step": 16242 + }, + { + "epoch": 0.8548947368421053, + "grad_norm": 0.50390625, + "learning_rate": 3.224561403508772e-05, + "loss": 3.0773, + "step": 16243 + }, + { + "epoch": 0.8549473684210527, + "grad_norm": 0.515625, + "learning_rate": 3.223391812865497e-05, + "loss": 2.9989, + "step": 16244 + }, + { + "epoch": 0.855, + "grad_norm": 0.50390625, + "learning_rate": 3.222222222222223e-05, + "loss": 3.3222, + "step": 16245 + }, + { + "epoch": 0.8550526315789474, + "grad_norm": 0.50390625, + "learning_rate": 3.221052631578947e-05, + "loss": 3.1329, + "step": 16246 + }, + { + "epoch": 0.8551052631578947, + "grad_norm": 0.53125, + "learning_rate": 3.219883040935673e-05, + "loss": 2.8129, + "step": 16247 + }, + { + "epoch": 0.8551578947368421, + "grad_norm": 0.482421875, + "learning_rate": 3.218713450292398e-05, + "loss": 3.3759, + "step": 16248 + }, + { + "epoch": 0.8552105263157894, + "grad_norm": 0.49609375, + "learning_rate": 3.217543859649123e-05, + "loss": 3.3562, + "step": 16249 + }, + { + "epoch": 0.8552631578947368, + "grad_norm": 0.5703125, + "learning_rate": 3.216374269005848e-05, + "loss": 3.5181, + "step": 16250 + }, + { + "epoch": 0.8553157894736843, + "grad_norm": 0.546875, + "learning_rate": 3.2152046783625734e-05, + "loss": 3.2445, + "step": 16251 + }, + { + "epoch": 0.8553684210526316, + "grad_norm": 0.482421875, + "learning_rate": 3.214035087719298e-05, + "loss": 3.4765, + "step": 16252 + }, + { + "epoch": 0.855421052631579, + "grad_norm": 0.53515625, + "learning_rate": 3.212865497076023e-05, + "loss": 2.9554, + "step": 16253 + }, + { + "epoch": 0.8554736842105263, + "grad_norm": 0.486328125, + "learning_rate": 3.211695906432748e-05, + "loss": 3.299, + "step": 16254 + }, + { + "epoch": 0.8555263157894737, + "grad_norm": 0.515625, + "learning_rate": 3.210526315789474e-05, + "loss": 3.166, + "step": 16255 + }, + { + "epoch": 0.8555789473684211, + "grad_norm": 0.48828125, + "learning_rate": 3.2093567251461995e-05, + "loss": 3.389, + "step": 16256 + }, + { + "epoch": 0.8556315789473684, + "grad_norm": 0.51171875, + "learning_rate": 3.208187134502924e-05, + "loss": 2.9496, + "step": 16257 + }, + { + "epoch": 0.8556842105263158, + "grad_norm": 0.498046875, + "learning_rate": 3.2070175438596494e-05, + "loss": 3.4295, + "step": 16258 + }, + { + "epoch": 0.8557368421052631, + "grad_norm": 0.5078125, + "learning_rate": 3.2058479532163744e-05, + "loss": 3.2458, + "step": 16259 + }, + { + "epoch": 0.8557894736842105, + "grad_norm": 0.4765625, + "learning_rate": 3.2046783625731e-05, + "loss": 3.2663, + "step": 16260 + }, + { + "epoch": 0.8558421052631578, + "grad_norm": 0.4921875, + "learning_rate": 3.203508771929824e-05, + "loss": 3.1545, + "step": 16261 + }, + { + "epoch": 0.8558947368421053, + "grad_norm": 0.5, + "learning_rate": 3.20233918128655e-05, + "loss": 3.0869, + "step": 16262 + }, + { + "epoch": 0.8559473684210527, + "grad_norm": 0.490234375, + "learning_rate": 3.201169590643275e-05, + "loss": 3.3677, + "step": 16263 + }, + { + "epoch": 0.856, + "grad_norm": 0.51171875, + "learning_rate": 3.2000000000000005e-05, + "loss": 3.0253, + "step": 16264 + }, + { + "epoch": 0.8560526315789474, + "grad_norm": 0.50390625, + "learning_rate": 3.198830409356725e-05, + "loss": 3.7018, + "step": 16265 + }, + { + "epoch": 0.8561052631578947, + "grad_norm": 0.4921875, + "learning_rate": 3.1976608187134505e-05, + "loss": 3.1898, + "step": 16266 + }, + { + "epoch": 0.8561578947368421, + "grad_norm": 0.48046875, + "learning_rate": 3.1964912280701754e-05, + "loss": 3.099, + "step": 16267 + }, + { + "epoch": 0.8562105263157894, + "grad_norm": 0.46875, + "learning_rate": 3.195321637426901e-05, + "loss": 3.3333, + "step": 16268 + }, + { + "epoch": 0.8562631578947368, + "grad_norm": 0.51171875, + "learning_rate": 3.194152046783626e-05, + "loss": 3.3493, + "step": 16269 + }, + { + "epoch": 0.8563157894736843, + "grad_norm": 0.49609375, + "learning_rate": 3.192982456140351e-05, + "loss": 2.9882, + "step": 16270 + }, + { + "epoch": 0.8563684210526316, + "grad_norm": 0.53515625, + "learning_rate": 3.1918128654970766e-05, + "loss": 3.4675, + "step": 16271 + }, + { + "epoch": 0.856421052631579, + "grad_norm": 0.50390625, + "learning_rate": 3.190643274853801e-05, + "loss": 3.446, + "step": 16272 + }, + { + "epoch": 0.8564736842105263, + "grad_norm": 0.515625, + "learning_rate": 3.1894736842105265e-05, + "loss": 2.9815, + "step": 16273 + }, + { + "epoch": 0.8565263157894737, + "grad_norm": 0.5, + "learning_rate": 3.1883040935672515e-05, + "loss": 3.4822, + "step": 16274 + }, + { + "epoch": 0.8565789473684211, + "grad_norm": 0.498046875, + "learning_rate": 3.187134502923977e-05, + "loss": 3.1123, + "step": 16275 + }, + { + "epoch": 0.8566315789473684, + "grad_norm": 0.51171875, + "learning_rate": 3.1859649122807014e-05, + "loss": 3.0161, + "step": 16276 + }, + { + "epoch": 0.8566842105263158, + "grad_norm": 0.55859375, + "learning_rate": 3.184795321637427e-05, + "loss": 3.2847, + "step": 16277 + }, + { + "epoch": 0.8567368421052631, + "grad_norm": 0.490234375, + "learning_rate": 3.183625730994152e-05, + "loss": 3.1266, + "step": 16278 + }, + { + "epoch": 0.8567894736842105, + "grad_norm": 0.52734375, + "learning_rate": 3.1824561403508776e-05, + "loss": 3.123, + "step": 16279 + }, + { + "epoch": 0.8568421052631578, + "grad_norm": 0.51953125, + "learning_rate": 3.1812865497076026e-05, + "loss": 3.1617, + "step": 16280 + }, + { + "epoch": 0.8568947368421053, + "grad_norm": 0.5078125, + "learning_rate": 3.1801169590643276e-05, + "loss": 3.6334, + "step": 16281 + }, + { + "epoch": 0.8569473684210527, + "grad_norm": 0.578125, + "learning_rate": 3.178947368421053e-05, + "loss": 3.3824, + "step": 16282 + }, + { + "epoch": 0.857, + "grad_norm": 0.498046875, + "learning_rate": 3.177777777777778e-05, + "loss": 3.8074, + "step": 16283 + }, + { + "epoch": 0.8570526315789474, + "grad_norm": 0.54296875, + "learning_rate": 3.176608187134503e-05, + "loss": 3.1782, + "step": 16284 + }, + { + "epoch": 0.8571052631578947, + "grad_norm": 0.4921875, + "learning_rate": 3.175438596491228e-05, + "loss": 3.4348, + "step": 16285 + }, + { + "epoch": 0.8571578947368421, + "grad_norm": 0.515625, + "learning_rate": 3.174269005847954e-05, + "loss": 3.1343, + "step": 16286 + }, + { + "epoch": 0.8572105263157894, + "grad_norm": 0.49609375, + "learning_rate": 3.173099415204679e-05, + "loss": 3.4552, + "step": 16287 + }, + { + "epoch": 0.8572631578947368, + "grad_norm": 0.474609375, + "learning_rate": 3.1719298245614036e-05, + "loss": 3.0891, + "step": 16288 + }, + { + "epoch": 0.8573157894736843, + "grad_norm": 0.51953125, + "learning_rate": 3.1707602339181286e-05, + "loss": 3.0699, + "step": 16289 + }, + { + "epoch": 0.8573684210526316, + "grad_norm": 0.5, + "learning_rate": 3.169590643274854e-05, + "loss": 3.3721, + "step": 16290 + }, + { + "epoch": 0.857421052631579, + "grad_norm": 0.4921875, + "learning_rate": 3.168421052631579e-05, + "loss": 3.1767, + "step": 16291 + }, + { + "epoch": 0.8574736842105263, + "grad_norm": 0.515625, + "learning_rate": 3.167251461988304e-05, + "loss": 3.5888, + "step": 16292 + }, + { + "epoch": 0.8575263157894737, + "grad_norm": 0.5546875, + "learning_rate": 3.16608187134503e-05, + "loss": 2.8968, + "step": 16293 + }, + { + "epoch": 0.8575789473684211, + "grad_norm": 0.53515625, + "learning_rate": 3.164912280701755e-05, + "loss": 2.9634, + "step": 16294 + }, + { + "epoch": 0.8576315789473684, + "grad_norm": 0.49609375, + "learning_rate": 3.16374269005848e-05, + "loss": 3.1616, + "step": 16295 + }, + { + "epoch": 0.8576842105263158, + "grad_norm": 0.486328125, + "learning_rate": 3.1625730994152047e-05, + "loss": 3.2083, + "step": 16296 + }, + { + "epoch": 0.8577368421052631, + "grad_norm": 0.5078125, + "learning_rate": 3.16140350877193e-05, + "loss": 2.6529, + "step": 16297 + }, + { + "epoch": 0.8577894736842105, + "grad_norm": 0.486328125, + "learning_rate": 3.160233918128655e-05, + "loss": 3.5068, + "step": 16298 + }, + { + "epoch": 0.8578421052631579, + "grad_norm": 0.5078125, + "learning_rate": 3.15906432748538e-05, + "loss": 3.5936, + "step": 16299 + }, + { + "epoch": 0.8578947368421053, + "grad_norm": 0.52734375, + "learning_rate": 3.157894736842105e-05, + "loss": 3.2488, + "step": 16300 + }, + { + "epoch": 0.8579473684210527, + "grad_norm": 0.49609375, + "learning_rate": 3.156725146198831e-05, + "loss": 3.2309, + "step": 16301 + }, + { + "epoch": 0.858, + "grad_norm": 0.51953125, + "learning_rate": 3.155555555555556e-05, + "loss": 3.7138, + "step": 16302 + }, + { + "epoch": 0.8580526315789474, + "grad_norm": 0.5234375, + "learning_rate": 3.154385964912281e-05, + "loss": 3.1845, + "step": 16303 + }, + { + "epoch": 0.8581052631578947, + "grad_norm": 0.53125, + "learning_rate": 3.153216374269006e-05, + "loss": 3.3205, + "step": 16304 + }, + { + "epoch": 0.8581578947368421, + "grad_norm": 0.51171875, + "learning_rate": 3.152046783625731e-05, + "loss": 3.4741, + "step": 16305 + }, + { + "epoch": 0.8582105263157894, + "grad_norm": 0.490234375, + "learning_rate": 3.150877192982456e-05, + "loss": 3.0931, + "step": 16306 + }, + { + "epoch": 0.8582631578947368, + "grad_norm": 0.5625, + "learning_rate": 3.149707602339181e-05, + "loss": 3.0689, + "step": 16307 + }, + { + "epoch": 0.8583157894736843, + "grad_norm": 0.53125, + "learning_rate": 3.148538011695907e-05, + "loss": 3.2861, + "step": 16308 + }, + { + "epoch": 0.8583684210526316, + "grad_norm": 0.482421875, + "learning_rate": 3.147368421052632e-05, + "loss": 3.3451, + "step": 16309 + }, + { + "epoch": 0.858421052631579, + "grad_norm": 0.625, + "learning_rate": 3.146198830409357e-05, + "loss": 3.3003, + "step": 16310 + }, + { + "epoch": 0.8584736842105263, + "grad_norm": 0.51953125, + "learning_rate": 3.145029239766082e-05, + "loss": 2.845, + "step": 16311 + }, + { + "epoch": 0.8585263157894737, + "grad_norm": 0.51171875, + "learning_rate": 3.1438596491228074e-05, + "loss": 3.3131, + "step": 16312 + }, + { + "epoch": 0.858578947368421, + "grad_norm": 0.4921875, + "learning_rate": 3.1426900584795323e-05, + "loss": 3.2414, + "step": 16313 + }, + { + "epoch": 0.8586315789473684, + "grad_norm": 0.5, + "learning_rate": 3.141520467836257e-05, + "loss": 2.924, + "step": 16314 + }, + { + "epoch": 0.8586842105263158, + "grad_norm": 0.482421875, + "learning_rate": 3.140350877192982e-05, + "loss": 3.0958, + "step": 16315 + }, + { + "epoch": 0.8587368421052631, + "grad_norm": 0.5078125, + "learning_rate": 3.139181286549708e-05, + "loss": 3.1506, + "step": 16316 + }, + { + "epoch": 0.8587894736842105, + "grad_norm": 0.484375, + "learning_rate": 3.138011695906433e-05, + "loss": 3.0679, + "step": 16317 + }, + { + "epoch": 0.8588421052631579, + "grad_norm": 0.51171875, + "learning_rate": 3.136842105263158e-05, + "loss": 3.2457, + "step": 16318 + }, + { + "epoch": 0.8588947368421053, + "grad_norm": 0.48828125, + "learning_rate": 3.1356725146198835e-05, + "loss": 3.4178, + "step": 16319 + }, + { + "epoch": 0.8589473684210527, + "grad_norm": 0.51171875, + "learning_rate": 3.1345029239766084e-05, + "loss": 3.3508, + "step": 16320 + }, + { + "epoch": 0.859, + "grad_norm": 0.49609375, + "learning_rate": 3.1333333333333334e-05, + "loss": 3.1646, + "step": 16321 + }, + { + "epoch": 0.8590526315789474, + "grad_norm": 0.50390625, + "learning_rate": 3.132163742690058e-05, + "loss": 3.53, + "step": 16322 + }, + { + "epoch": 0.8591052631578947, + "grad_norm": 0.478515625, + "learning_rate": 3.130994152046784e-05, + "loss": 2.9513, + "step": 16323 + }, + { + "epoch": 0.8591578947368421, + "grad_norm": 0.48828125, + "learning_rate": 3.129824561403509e-05, + "loss": 3.2394, + "step": 16324 + }, + { + "epoch": 0.8592105263157894, + "grad_norm": 0.515625, + "learning_rate": 3.128654970760234e-05, + "loss": 3.0538, + "step": 16325 + }, + { + "epoch": 0.8592631578947368, + "grad_norm": 0.48046875, + "learning_rate": 3.127485380116959e-05, + "loss": 3.5667, + "step": 16326 + }, + { + "epoch": 0.8593157894736843, + "grad_norm": 0.498046875, + "learning_rate": 3.1263157894736845e-05, + "loss": 3.1899, + "step": 16327 + }, + { + "epoch": 0.8593684210526316, + "grad_norm": 0.51953125, + "learning_rate": 3.1251461988304094e-05, + "loss": 3.17, + "step": 16328 + }, + { + "epoch": 0.859421052631579, + "grad_norm": 0.6484375, + "learning_rate": 3.1239766081871344e-05, + "loss": 2.9721, + "step": 16329 + }, + { + "epoch": 0.8594736842105263, + "grad_norm": 0.484375, + "learning_rate": 3.12280701754386e-05, + "loss": 3.534, + "step": 16330 + }, + { + "epoch": 0.8595263157894737, + "grad_norm": 0.515625, + "learning_rate": 3.121637426900585e-05, + "loss": 2.6672, + "step": 16331 + }, + { + "epoch": 0.859578947368421, + "grad_norm": 0.498046875, + "learning_rate": 3.1204678362573106e-05, + "loss": 3.0488, + "step": 16332 + }, + { + "epoch": 0.8596315789473684, + "grad_norm": 0.51953125, + "learning_rate": 3.119298245614035e-05, + "loss": 3.4008, + "step": 16333 + }, + { + "epoch": 0.8596842105263158, + "grad_norm": 0.4921875, + "learning_rate": 3.1181286549707606e-05, + "loss": 2.915, + "step": 16334 + }, + { + "epoch": 0.8597368421052631, + "grad_norm": 0.490234375, + "learning_rate": 3.1169590643274855e-05, + "loss": 3.064, + "step": 16335 + }, + { + "epoch": 0.8597894736842105, + "grad_norm": 0.54296875, + "learning_rate": 3.115789473684211e-05, + "loss": 3.5478, + "step": 16336 + }, + { + "epoch": 0.8598421052631579, + "grad_norm": 0.57421875, + "learning_rate": 3.1146198830409354e-05, + "loss": 3.3587, + "step": 16337 + }, + { + "epoch": 0.8598947368421053, + "grad_norm": 0.50390625, + "learning_rate": 3.113450292397661e-05, + "loss": 3.5468, + "step": 16338 + }, + { + "epoch": 0.8599473684210527, + "grad_norm": 0.490234375, + "learning_rate": 3.112280701754386e-05, + "loss": 3.4188, + "step": 16339 + }, + { + "epoch": 0.86, + "grad_norm": 0.47265625, + "learning_rate": 3.111111111111111e-05, + "loss": 3.2205, + "step": 16340 + }, + { + "epoch": 0.8600526315789474, + "grad_norm": 0.53125, + "learning_rate": 3.109941520467836e-05, + "loss": 3.2347, + "step": 16341 + }, + { + "epoch": 0.8601052631578947, + "grad_norm": 0.51953125, + "learning_rate": 3.1087719298245616e-05, + "loss": 3.7481, + "step": 16342 + }, + { + "epoch": 0.8601578947368421, + "grad_norm": 0.494140625, + "learning_rate": 3.107602339181287e-05, + "loss": 3.4516, + "step": 16343 + }, + { + "epoch": 0.8602105263157894, + "grad_norm": 0.5, + "learning_rate": 3.1064327485380115e-05, + "loss": 3.4245, + "step": 16344 + }, + { + "epoch": 0.8602631578947368, + "grad_norm": 0.52734375, + "learning_rate": 3.105263157894737e-05, + "loss": 3.379, + "step": 16345 + }, + { + "epoch": 0.8603157894736843, + "grad_norm": 0.5, + "learning_rate": 3.104093567251462e-05, + "loss": 3.2224, + "step": 16346 + }, + { + "epoch": 0.8603684210526316, + "grad_norm": 0.54296875, + "learning_rate": 3.102923976608188e-05, + "loss": 3.5279, + "step": 16347 + }, + { + "epoch": 0.860421052631579, + "grad_norm": 0.484375, + "learning_rate": 3.101754385964912e-05, + "loss": 3.3737, + "step": 16348 + }, + { + "epoch": 0.8604736842105263, + "grad_norm": 0.5078125, + "learning_rate": 3.1005847953216377e-05, + "loss": 3.2109, + "step": 16349 + }, + { + "epoch": 0.8605263157894737, + "grad_norm": 0.48828125, + "learning_rate": 3.0994152046783626e-05, + "loss": 3.5463, + "step": 16350 + }, + { + "epoch": 0.860578947368421, + "grad_norm": 0.50390625, + "learning_rate": 3.098245614035088e-05, + "loss": 3.2539, + "step": 16351 + }, + { + "epoch": 0.8606315789473684, + "grad_norm": 0.494140625, + "learning_rate": 3.0970760233918125e-05, + "loss": 3.3285, + "step": 16352 + }, + { + "epoch": 0.8606842105263158, + "grad_norm": 0.482421875, + "learning_rate": 3.095906432748538e-05, + "loss": 3.468, + "step": 16353 + }, + { + "epoch": 0.8607368421052631, + "grad_norm": 0.490234375, + "learning_rate": 3.094736842105263e-05, + "loss": 3.3584, + "step": 16354 + }, + { + "epoch": 0.8607894736842105, + "grad_norm": 0.58984375, + "learning_rate": 3.093567251461989e-05, + "loss": 2.9162, + "step": 16355 + }, + { + "epoch": 0.8608421052631579, + "grad_norm": 0.5, + "learning_rate": 3.092397660818714e-05, + "loss": 3.1284, + "step": 16356 + }, + { + "epoch": 0.8608947368421053, + "grad_norm": 0.53515625, + "learning_rate": 3.091228070175439e-05, + "loss": 3.2398, + "step": 16357 + }, + { + "epoch": 0.8609473684210527, + "grad_norm": 0.52734375, + "learning_rate": 3.090058479532164e-05, + "loss": 3.4741, + "step": 16358 + }, + { + "epoch": 0.861, + "grad_norm": 0.478515625, + "learning_rate": 3.088888888888889e-05, + "loss": 3.0992, + "step": 16359 + }, + { + "epoch": 0.8610526315789474, + "grad_norm": 0.482421875, + "learning_rate": 3.087719298245614e-05, + "loss": 2.7707, + "step": 16360 + }, + { + "epoch": 0.8611052631578947, + "grad_norm": 0.470703125, + "learning_rate": 3.086549707602339e-05, + "loss": 3.0537, + "step": 16361 + }, + { + "epoch": 0.8611578947368421, + "grad_norm": 0.478515625, + "learning_rate": 3.085380116959065e-05, + "loss": 3.0556, + "step": 16362 + }, + { + "epoch": 0.8612105263157894, + "grad_norm": 0.5234375, + "learning_rate": 3.084210526315789e-05, + "loss": 3.3564, + "step": 16363 + }, + { + "epoch": 0.8612631578947368, + "grad_norm": 0.46875, + "learning_rate": 3.083040935672515e-05, + "loss": 3.0361, + "step": 16364 + }, + { + "epoch": 0.8613157894736843, + "grad_norm": 0.53125, + "learning_rate": 3.08187134502924e-05, + "loss": 3.6513, + "step": 16365 + }, + { + "epoch": 0.8613684210526316, + "grad_norm": 0.515625, + "learning_rate": 3.0807017543859654e-05, + "loss": 3.2424, + "step": 16366 + }, + { + "epoch": 0.861421052631579, + "grad_norm": 0.470703125, + "learning_rate": 3.0795321637426896e-05, + "loss": 3.2998, + "step": 16367 + }, + { + "epoch": 0.8614736842105263, + "grad_norm": 0.50390625, + "learning_rate": 3.078362573099415e-05, + "loss": 3.29, + "step": 16368 + }, + { + "epoch": 0.8615263157894737, + "grad_norm": 0.515625, + "learning_rate": 3.077192982456141e-05, + "loss": 3.5296, + "step": 16369 + }, + { + "epoch": 0.861578947368421, + "grad_norm": 0.515625, + "learning_rate": 3.076023391812866e-05, + "loss": 3.4554, + "step": 16370 + }, + { + "epoch": 0.8616315789473684, + "grad_norm": 0.48828125, + "learning_rate": 3.074853801169591e-05, + "loss": 3.1909, + "step": 16371 + }, + { + "epoch": 0.8616842105263158, + "grad_norm": 0.478515625, + "learning_rate": 3.073684210526316e-05, + "loss": 2.7863, + "step": 16372 + }, + { + "epoch": 0.8617368421052631, + "grad_norm": 0.48828125, + "learning_rate": 3.0725146198830414e-05, + "loss": 3.1094, + "step": 16373 + }, + { + "epoch": 0.8617894736842105, + "grad_norm": 0.51953125, + "learning_rate": 3.0713450292397664e-05, + "loss": 3.1202, + "step": 16374 + }, + { + "epoch": 0.8618421052631579, + "grad_norm": 0.515625, + "learning_rate": 3.0701754385964913e-05, + "loss": 3.0825, + "step": 16375 + }, + { + "epoch": 0.8618947368421053, + "grad_norm": 0.474609375, + "learning_rate": 3.069005847953216e-05, + "loss": 2.8818, + "step": 16376 + }, + { + "epoch": 0.8619473684210527, + "grad_norm": 0.5, + "learning_rate": 3.067836257309942e-05, + "loss": 3.4168, + "step": 16377 + }, + { + "epoch": 0.862, + "grad_norm": 0.462890625, + "learning_rate": 3.066666666666667e-05, + "loss": 2.8078, + "step": 16378 + }, + { + "epoch": 0.8620526315789474, + "grad_norm": 0.51953125, + "learning_rate": 3.065497076023392e-05, + "loss": 3.3831, + "step": 16379 + }, + { + "epoch": 0.8621052631578947, + "grad_norm": 0.51171875, + "learning_rate": 3.0643274853801175e-05, + "loss": 2.7197, + "step": 16380 + }, + { + "epoch": 0.8621578947368421, + "grad_norm": 0.5859375, + "learning_rate": 3.0631578947368425e-05, + "loss": 3.4058, + "step": 16381 + }, + { + "epoch": 0.8622105263157894, + "grad_norm": 0.50390625, + "learning_rate": 3.0619883040935674e-05, + "loss": 3.1288, + "step": 16382 + }, + { + "epoch": 0.8622631578947368, + "grad_norm": 0.53515625, + "learning_rate": 3.0608187134502924e-05, + "loss": 3.1224, + "step": 16383 + }, + { + "epoch": 0.8623157894736843, + "grad_norm": 0.5625, + "learning_rate": 3.059649122807018e-05, + "loss": 3.3046, + "step": 16384 + }, + { + "epoch": 0.8623684210526316, + "grad_norm": 0.5, + "learning_rate": 3.058479532163743e-05, + "loss": 3.1168, + "step": 16385 + }, + { + "epoch": 0.862421052631579, + "grad_norm": 0.51953125, + "learning_rate": 3.057309941520468e-05, + "loss": 3.2675, + "step": 16386 + }, + { + "epoch": 0.8624736842105263, + "grad_norm": 0.5390625, + "learning_rate": 3.056140350877193e-05, + "loss": 3.497, + "step": 16387 + }, + { + "epoch": 0.8625263157894737, + "grad_norm": 0.490234375, + "learning_rate": 3.0549707602339185e-05, + "loss": 3.301, + "step": 16388 + }, + { + "epoch": 0.862578947368421, + "grad_norm": 0.51953125, + "learning_rate": 3.0538011695906435e-05, + "loss": 3.4277, + "step": 16389 + }, + { + "epoch": 0.8626315789473684, + "grad_norm": 0.5078125, + "learning_rate": 3.0526315789473684e-05, + "loss": 3.5384, + "step": 16390 + }, + { + "epoch": 0.8626842105263158, + "grad_norm": 0.490234375, + "learning_rate": 3.0514619883040934e-05, + "loss": 3.1985, + "step": 16391 + }, + { + "epoch": 0.8627368421052631, + "grad_norm": 0.5078125, + "learning_rate": 3.050292397660819e-05, + "loss": 3.2409, + "step": 16392 + }, + { + "epoch": 0.8627894736842106, + "grad_norm": 0.482421875, + "learning_rate": 3.0491228070175443e-05, + "loss": 3.4646, + "step": 16393 + }, + { + "epoch": 0.8628421052631579, + "grad_norm": 0.4921875, + "learning_rate": 3.047953216374269e-05, + "loss": 3.1938, + "step": 16394 + }, + { + "epoch": 0.8628947368421053, + "grad_norm": 0.5, + "learning_rate": 3.0467836257309946e-05, + "loss": 3.4815, + "step": 16395 + }, + { + "epoch": 0.8629473684210527, + "grad_norm": 0.458984375, + "learning_rate": 3.0456140350877192e-05, + "loss": 2.9986, + "step": 16396 + }, + { + "epoch": 0.863, + "grad_norm": 0.515625, + "learning_rate": 3.044444444444445e-05, + "loss": 3.6215, + "step": 16397 + }, + { + "epoch": 0.8630526315789474, + "grad_norm": 0.5078125, + "learning_rate": 3.0432748538011695e-05, + "loss": 3.5335, + "step": 16398 + }, + { + "epoch": 0.8631052631578947, + "grad_norm": 0.478515625, + "learning_rate": 3.042105263157895e-05, + "loss": 2.7134, + "step": 16399 + }, + { + "epoch": 0.8631578947368421, + "grad_norm": 0.49609375, + "learning_rate": 3.0409356725146197e-05, + "loss": 3.0878, + "step": 16400 + }, + { + "epoch": 0.8632105263157894, + "grad_norm": 0.50390625, + "learning_rate": 3.0397660818713454e-05, + "loss": 3.5192, + "step": 16401 + }, + { + "epoch": 0.8632631578947368, + "grad_norm": 0.486328125, + "learning_rate": 3.03859649122807e-05, + "loss": 3.3827, + "step": 16402 + }, + { + "epoch": 0.8633157894736843, + "grad_norm": 0.50390625, + "learning_rate": 3.0374269005847956e-05, + "loss": 3.4157, + "step": 16403 + }, + { + "epoch": 0.8633684210526316, + "grad_norm": 0.50390625, + "learning_rate": 3.0362573099415202e-05, + "loss": 3.195, + "step": 16404 + }, + { + "epoch": 0.863421052631579, + "grad_norm": 0.515625, + "learning_rate": 3.035087719298246e-05, + "loss": 3.0324, + "step": 16405 + }, + { + "epoch": 0.8634736842105263, + "grad_norm": 0.48828125, + "learning_rate": 3.0339181286549712e-05, + "loss": 2.9971, + "step": 16406 + }, + { + "epoch": 0.8635263157894737, + "grad_norm": 0.50390625, + "learning_rate": 3.032748538011696e-05, + "loss": 3.3258, + "step": 16407 + }, + { + "epoch": 0.863578947368421, + "grad_norm": 0.5, + "learning_rate": 3.0315789473684214e-05, + "loss": 3.3321, + "step": 16408 + }, + { + "epoch": 0.8636315789473684, + "grad_norm": 0.48828125, + "learning_rate": 3.0304093567251464e-05, + "loss": 3.3106, + "step": 16409 + }, + { + "epoch": 0.8636842105263158, + "grad_norm": 0.470703125, + "learning_rate": 3.0292397660818717e-05, + "loss": 2.81, + "step": 16410 + }, + { + "epoch": 0.8637368421052631, + "grad_norm": 0.51953125, + "learning_rate": 3.0280701754385966e-05, + "loss": 3.2937, + "step": 16411 + }, + { + "epoch": 0.8637894736842106, + "grad_norm": 0.490234375, + "learning_rate": 3.026900584795322e-05, + "loss": 3.2366, + "step": 16412 + }, + { + "epoch": 0.8638421052631579, + "grad_norm": 0.5078125, + "learning_rate": 3.025730994152047e-05, + "loss": 3.3065, + "step": 16413 + }, + { + "epoch": 0.8638947368421053, + "grad_norm": 0.49609375, + "learning_rate": 3.0245614035087722e-05, + "loss": 3.3974, + "step": 16414 + }, + { + "epoch": 0.8639473684210527, + "grad_norm": 0.58203125, + "learning_rate": 3.023391812865497e-05, + "loss": 3.058, + "step": 16415 + }, + { + "epoch": 0.864, + "grad_norm": 0.50390625, + "learning_rate": 3.0222222222222225e-05, + "loss": 3.7437, + "step": 16416 + }, + { + "epoch": 0.8640526315789474, + "grad_norm": 0.52734375, + "learning_rate": 3.021052631578947e-05, + "loss": 3.29, + "step": 16417 + }, + { + "epoch": 0.8641052631578947, + "grad_norm": 0.498046875, + "learning_rate": 3.0198830409356727e-05, + "loss": 3.4089, + "step": 16418 + }, + { + "epoch": 0.8641578947368421, + "grad_norm": 0.484375, + "learning_rate": 3.018713450292398e-05, + "loss": 3.7152, + "step": 16419 + }, + { + "epoch": 0.8642105263157894, + "grad_norm": 0.50390625, + "learning_rate": 3.017543859649123e-05, + "loss": 3.3756, + "step": 16420 + }, + { + "epoch": 0.8642631578947368, + "grad_norm": 0.486328125, + "learning_rate": 3.0163742690058483e-05, + "loss": 2.8936, + "step": 16421 + }, + { + "epoch": 0.8643157894736843, + "grad_norm": 0.5390625, + "learning_rate": 3.0152046783625732e-05, + "loss": 3.1958, + "step": 16422 + }, + { + "epoch": 0.8643684210526316, + "grad_norm": 0.5546875, + "learning_rate": 3.0140350877192985e-05, + "loss": 3.5931, + "step": 16423 + }, + { + "epoch": 0.864421052631579, + "grad_norm": 0.5234375, + "learning_rate": 3.0128654970760235e-05, + "loss": 3.6253, + "step": 16424 + }, + { + "epoch": 0.8644736842105263, + "grad_norm": 0.490234375, + "learning_rate": 3.0116959064327488e-05, + "loss": 3.4103, + "step": 16425 + }, + { + "epoch": 0.8645263157894737, + "grad_norm": 0.52734375, + "learning_rate": 3.0105263157894737e-05, + "loss": 2.9323, + "step": 16426 + }, + { + "epoch": 0.864578947368421, + "grad_norm": 0.494140625, + "learning_rate": 3.009356725146199e-05, + "loss": 3.0186, + "step": 16427 + }, + { + "epoch": 0.8646315789473684, + "grad_norm": 0.47265625, + "learning_rate": 3.008187134502924e-05, + "loss": 3.2887, + "step": 16428 + }, + { + "epoch": 0.8646842105263158, + "grad_norm": 0.4765625, + "learning_rate": 3.0070175438596493e-05, + "loss": 2.7921, + "step": 16429 + }, + { + "epoch": 0.8647368421052631, + "grad_norm": 0.48828125, + "learning_rate": 3.0058479532163746e-05, + "loss": 3.5923, + "step": 16430 + }, + { + "epoch": 0.8647894736842106, + "grad_norm": 0.47265625, + "learning_rate": 3.0046783625730996e-05, + "loss": 3.1735, + "step": 16431 + }, + { + "epoch": 0.8648421052631579, + "grad_norm": 0.478515625, + "learning_rate": 3.003508771929825e-05, + "loss": 2.3938, + "step": 16432 + }, + { + "epoch": 0.8648947368421053, + "grad_norm": 0.5078125, + "learning_rate": 3.0023391812865498e-05, + "loss": 3.1468, + "step": 16433 + }, + { + "epoch": 0.8649473684210527, + "grad_norm": 0.5625, + "learning_rate": 3.001169590643275e-05, + "loss": 3.5333, + "step": 16434 + }, + { + "epoch": 0.865, + "grad_norm": 0.51953125, + "learning_rate": 3e-05, + "loss": 2.9915, + "step": 16435 + }, + { + "epoch": 0.8650526315789474, + "grad_norm": 0.51171875, + "learning_rate": 2.9988304093567254e-05, + "loss": 3.7091, + "step": 16436 + }, + { + "epoch": 0.8651052631578947, + "grad_norm": 0.4921875, + "learning_rate": 2.9976608187134503e-05, + "loss": 2.1931, + "step": 16437 + }, + { + "epoch": 0.8651578947368421, + "grad_norm": 0.48828125, + "learning_rate": 2.9964912280701756e-05, + "loss": 3.6491, + "step": 16438 + }, + { + "epoch": 0.8652105263157894, + "grad_norm": 0.48046875, + "learning_rate": 2.9953216374269006e-05, + "loss": 3.558, + "step": 16439 + }, + { + "epoch": 0.8652631578947368, + "grad_norm": 0.5078125, + "learning_rate": 2.994152046783626e-05, + "loss": 3.4199, + "step": 16440 + }, + { + "epoch": 0.8653157894736843, + "grad_norm": 0.5, + "learning_rate": 2.992982456140351e-05, + "loss": 3.0337, + "step": 16441 + }, + { + "epoch": 0.8653684210526316, + "grad_norm": 0.46875, + "learning_rate": 2.991812865497076e-05, + "loss": 3.3466, + "step": 16442 + }, + { + "epoch": 0.865421052631579, + "grad_norm": 0.484375, + "learning_rate": 2.9906432748538014e-05, + "loss": 3.3404, + "step": 16443 + }, + { + "epoch": 0.8654736842105263, + "grad_norm": 0.52734375, + "learning_rate": 2.9894736842105264e-05, + "loss": 2.9376, + "step": 16444 + }, + { + "epoch": 0.8655263157894737, + "grad_norm": 0.4921875, + "learning_rate": 2.9883040935672517e-05, + "loss": 3.4009, + "step": 16445 + }, + { + "epoch": 0.865578947368421, + "grad_norm": 0.49609375, + "learning_rate": 2.9871345029239767e-05, + "loss": 3.3216, + "step": 16446 + }, + { + "epoch": 0.8656315789473684, + "grad_norm": 0.486328125, + "learning_rate": 2.985964912280702e-05, + "loss": 3.5334, + "step": 16447 + }, + { + "epoch": 0.8656842105263158, + "grad_norm": 0.48828125, + "learning_rate": 2.984795321637427e-05, + "loss": 3.0672, + "step": 16448 + }, + { + "epoch": 0.8657368421052631, + "grad_norm": 0.5078125, + "learning_rate": 2.9836257309941522e-05, + "loss": 3.1194, + "step": 16449 + }, + { + "epoch": 0.8657894736842106, + "grad_norm": 0.5, + "learning_rate": 2.9824561403508772e-05, + "loss": 3.5325, + "step": 16450 + }, + { + "epoch": 0.8658421052631579, + "grad_norm": 0.51171875, + "learning_rate": 2.9812865497076025e-05, + "loss": 3.5666, + "step": 16451 + }, + { + "epoch": 0.8658947368421053, + "grad_norm": 0.482421875, + "learning_rate": 2.9801169590643274e-05, + "loss": 3.0018, + "step": 16452 + }, + { + "epoch": 0.8659473684210526, + "grad_norm": 0.50390625, + "learning_rate": 2.9789473684210527e-05, + "loss": 3.1005, + "step": 16453 + }, + { + "epoch": 0.866, + "grad_norm": 0.484375, + "learning_rate": 2.9777777777777777e-05, + "loss": 3.1619, + "step": 16454 + }, + { + "epoch": 0.8660526315789474, + "grad_norm": 0.50390625, + "learning_rate": 2.976608187134503e-05, + "loss": 3.3135, + "step": 16455 + }, + { + "epoch": 0.8661052631578947, + "grad_norm": 0.494140625, + "learning_rate": 2.9754385964912286e-05, + "loss": 2.9066, + "step": 16456 + }, + { + "epoch": 0.8661578947368421, + "grad_norm": 0.462890625, + "learning_rate": 2.9742690058479532e-05, + "loss": 3.4006, + "step": 16457 + }, + { + "epoch": 0.8662105263157894, + "grad_norm": 0.490234375, + "learning_rate": 2.973099415204679e-05, + "loss": 3.5161, + "step": 16458 + }, + { + "epoch": 0.8662631578947368, + "grad_norm": 0.4921875, + "learning_rate": 2.9719298245614035e-05, + "loss": 3.1349, + "step": 16459 + }, + { + "epoch": 0.8663157894736843, + "grad_norm": 0.50390625, + "learning_rate": 2.9707602339181288e-05, + "loss": 3.3277, + "step": 16460 + }, + { + "epoch": 0.8663684210526316, + "grad_norm": 0.50390625, + "learning_rate": 2.9695906432748538e-05, + "loss": 3.3548, + "step": 16461 + }, + { + "epoch": 0.866421052631579, + "grad_norm": 0.50390625, + "learning_rate": 2.968421052631579e-05, + "loss": 2.6389, + "step": 16462 + }, + { + "epoch": 0.8664736842105263, + "grad_norm": 0.474609375, + "learning_rate": 2.967251461988304e-05, + "loss": 3.5351, + "step": 16463 + }, + { + "epoch": 0.8665263157894737, + "grad_norm": 0.50390625, + "learning_rate": 2.9660818713450293e-05, + "loss": 3.5927, + "step": 16464 + }, + { + "epoch": 0.866578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.9649122807017543e-05, + "loss": 3.368, + "step": 16465 + }, + { + "epoch": 0.8666315789473684, + "grad_norm": 0.50390625, + "learning_rate": 2.9637426900584796e-05, + "loss": 3.6024, + "step": 16466 + }, + { + "epoch": 0.8666842105263158, + "grad_norm": 0.49609375, + "learning_rate": 2.9625730994152045e-05, + "loss": 2.9998, + "step": 16467 + }, + { + "epoch": 0.8667368421052631, + "grad_norm": 0.51953125, + "learning_rate": 2.9614035087719298e-05, + "loss": 3.3778, + "step": 16468 + }, + { + "epoch": 0.8667894736842106, + "grad_norm": 0.5078125, + "learning_rate": 2.9602339181286555e-05, + "loss": 3.6526, + "step": 16469 + }, + { + "epoch": 0.8668421052631579, + "grad_norm": 0.490234375, + "learning_rate": 2.95906432748538e-05, + "loss": 3.3744, + "step": 16470 + }, + { + "epoch": 0.8668947368421053, + "grad_norm": 0.4921875, + "learning_rate": 2.9578947368421057e-05, + "loss": 3.0169, + "step": 16471 + }, + { + "epoch": 0.8669473684210526, + "grad_norm": 0.54296875, + "learning_rate": 2.9567251461988303e-05, + "loss": 3.356, + "step": 16472 + }, + { + "epoch": 0.867, + "grad_norm": 0.4765625, + "learning_rate": 2.955555555555556e-05, + "loss": 3.265, + "step": 16473 + }, + { + "epoch": 0.8670526315789474, + "grad_norm": 0.50390625, + "learning_rate": 2.9543859649122806e-05, + "loss": 3.2125, + "step": 16474 + }, + { + "epoch": 0.8671052631578947, + "grad_norm": 0.484375, + "learning_rate": 2.9532163742690062e-05, + "loss": 3.3195, + "step": 16475 + }, + { + "epoch": 0.8671578947368421, + "grad_norm": 0.5390625, + "learning_rate": 2.952046783625731e-05, + "loss": 3.3495, + "step": 16476 + }, + { + "epoch": 0.8672105263157894, + "grad_norm": 0.51953125, + "learning_rate": 2.9508771929824565e-05, + "loss": 3.4593, + "step": 16477 + }, + { + "epoch": 0.8672631578947368, + "grad_norm": 0.5078125, + "learning_rate": 2.949707602339181e-05, + "loss": 3.3254, + "step": 16478 + }, + { + "epoch": 0.8673157894736843, + "grad_norm": 0.51171875, + "learning_rate": 2.9485380116959068e-05, + "loss": 3.288, + "step": 16479 + }, + { + "epoch": 0.8673684210526316, + "grad_norm": 0.5078125, + "learning_rate": 2.9473684210526314e-05, + "loss": 3.7807, + "step": 16480 + }, + { + "epoch": 0.867421052631579, + "grad_norm": 0.51171875, + "learning_rate": 2.946198830409357e-05, + "loss": 3.6176, + "step": 16481 + }, + { + "epoch": 0.8674736842105263, + "grad_norm": 0.48046875, + "learning_rate": 2.9450292397660823e-05, + "loss": 3.2965, + "step": 16482 + }, + { + "epoch": 0.8675263157894737, + "grad_norm": 0.470703125, + "learning_rate": 2.943859649122807e-05, + "loss": 2.9637, + "step": 16483 + }, + { + "epoch": 0.867578947368421, + "grad_norm": 0.498046875, + "learning_rate": 2.9426900584795326e-05, + "loss": 3.1347, + "step": 16484 + }, + { + "epoch": 0.8676315789473684, + "grad_norm": 0.53125, + "learning_rate": 2.9415204678362572e-05, + "loss": 2.9721, + "step": 16485 + }, + { + "epoch": 0.8676842105263158, + "grad_norm": 0.470703125, + "learning_rate": 2.9403508771929828e-05, + "loss": 2.8989, + "step": 16486 + }, + { + "epoch": 0.8677368421052631, + "grad_norm": 0.5234375, + "learning_rate": 2.9391812865497074e-05, + "loss": 3.4452, + "step": 16487 + }, + { + "epoch": 0.8677894736842106, + "grad_norm": 0.484375, + "learning_rate": 2.938011695906433e-05, + "loss": 3.5385, + "step": 16488 + }, + { + "epoch": 0.8678421052631579, + "grad_norm": 0.5390625, + "learning_rate": 2.9368421052631577e-05, + "loss": 3.2483, + "step": 16489 + }, + { + "epoch": 0.8678947368421053, + "grad_norm": 0.51953125, + "learning_rate": 2.9356725146198833e-05, + "loss": 3.3082, + "step": 16490 + }, + { + "epoch": 0.8679473684210526, + "grad_norm": 0.4921875, + "learning_rate": 2.934502923976608e-05, + "loss": 3.3369, + "step": 16491 + }, + { + "epoch": 0.868, + "grad_norm": 0.5390625, + "learning_rate": 2.9333333333333336e-05, + "loss": 3.4078, + "step": 16492 + }, + { + "epoch": 0.8680526315789474, + "grad_norm": 0.49609375, + "learning_rate": 2.932163742690059e-05, + "loss": 3.294, + "step": 16493 + }, + { + "epoch": 0.8681052631578947, + "grad_norm": 0.482421875, + "learning_rate": 2.930994152046784e-05, + "loss": 3.6401, + "step": 16494 + }, + { + "epoch": 0.8681578947368421, + "grad_norm": 0.4921875, + "learning_rate": 2.929824561403509e-05, + "loss": 3.0758, + "step": 16495 + }, + { + "epoch": 0.8682105263157894, + "grad_norm": 0.51953125, + "learning_rate": 2.928654970760234e-05, + "loss": 3.0174, + "step": 16496 + }, + { + "epoch": 0.8682631578947368, + "grad_norm": 0.5078125, + "learning_rate": 2.9274853801169594e-05, + "loss": 3.4509, + "step": 16497 + }, + { + "epoch": 0.8683157894736843, + "grad_norm": 0.494140625, + "learning_rate": 2.9263157894736844e-05, + "loss": 3.2692, + "step": 16498 + }, + { + "epoch": 0.8683684210526316, + "grad_norm": 0.484375, + "learning_rate": 2.9251461988304097e-05, + "loss": 3.3452, + "step": 16499 + }, + { + "epoch": 0.868421052631579, + "grad_norm": 0.486328125, + "learning_rate": 2.9239766081871346e-05, + "loss": 3.3217, + "step": 16500 + }, + { + "epoch": 0.8684736842105263, + "grad_norm": 0.4609375, + "learning_rate": 2.92280701754386e-05, + "loss": 3.4631, + "step": 16501 + }, + { + "epoch": 0.8685263157894737, + "grad_norm": 0.5078125, + "learning_rate": 2.921637426900585e-05, + "loss": 3.4492, + "step": 16502 + }, + { + "epoch": 0.868578947368421, + "grad_norm": 0.8125, + "learning_rate": 2.9204678362573102e-05, + "loss": 3.0906, + "step": 16503 + }, + { + "epoch": 0.8686315789473684, + "grad_norm": 0.50390625, + "learning_rate": 2.919298245614035e-05, + "loss": 2.8687, + "step": 16504 + }, + { + "epoch": 0.8686842105263158, + "grad_norm": 0.54296875, + "learning_rate": 2.9181286549707604e-05, + "loss": 3.4444, + "step": 16505 + }, + { + "epoch": 0.8687368421052631, + "grad_norm": 0.56640625, + "learning_rate": 2.9169590643274857e-05, + "loss": 3.1989, + "step": 16506 + }, + { + "epoch": 0.8687894736842106, + "grad_norm": 0.52734375, + "learning_rate": 2.9157894736842107e-05, + "loss": 2.92, + "step": 16507 + }, + { + "epoch": 0.8688421052631579, + "grad_norm": 0.53515625, + "learning_rate": 2.914619883040936e-05, + "loss": 3.4994, + "step": 16508 + }, + { + "epoch": 0.8688947368421053, + "grad_norm": 0.486328125, + "learning_rate": 2.913450292397661e-05, + "loss": 3.1498, + "step": 16509 + }, + { + "epoch": 0.8689473684210526, + "grad_norm": 0.51171875, + "learning_rate": 2.9122807017543863e-05, + "loss": 3.4646, + "step": 16510 + }, + { + "epoch": 0.869, + "grad_norm": 0.4921875, + "learning_rate": 2.9111111111111112e-05, + "loss": 3.2559, + "step": 16511 + }, + { + "epoch": 0.8690526315789474, + "grad_norm": 0.515625, + "learning_rate": 2.9099415204678365e-05, + "loss": 3.064, + "step": 16512 + }, + { + "epoch": 0.8691052631578947, + "grad_norm": 0.5078125, + "learning_rate": 2.9087719298245615e-05, + "loss": 3.1118, + "step": 16513 + }, + { + "epoch": 0.8691578947368421, + "grad_norm": 0.478515625, + "learning_rate": 2.9076023391812868e-05, + "loss": 2.6453, + "step": 16514 + }, + { + "epoch": 0.8692105263157894, + "grad_norm": 0.486328125, + "learning_rate": 2.9064327485380117e-05, + "loss": 3.4614, + "step": 16515 + }, + { + "epoch": 0.8692631578947368, + "grad_norm": 0.478515625, + "learning_rate": 2.905263157894737e-05, + "loss": 3.3807, + "step": 16516 + }, + { + "epoch": 0.8693157894736843, + "grad_norm": 0.5, + "learning_rate": 2.904093567251462e-05, + "loss": 3.2163, + "step": 16517 + }, + { + "epoch": 0.8693684210526316, + "grad_norm": 0.5078125, + "learning_rate": 2.9029239766081873e-05, + "loss": 3.5411, + "step": 16518 + }, + { + "epoch": 0.869421052631579, + "grad_norm": 0.4921875, + "learning_rate": 2.9017543859649126e-05, + "loss": 3.0732, + "step": 16519 + }, + { + "epoch": 0.8694736842105263, + "grad_norm": 0.51171875, + "learning_rate": 2.9005847953216375e-05, + "loss": 3.1607, + "step": 16520 + }, + { + "epoch": 0.8695263157894737, + "grad_norm": 0.50390625, + "learning_rate": 2.899415204678363e-05, + "loss": 3.1038, + "step": 16521 + }, + { + "epoch": 0.869578947368421, + "grad_norm": 0.4921875, + "learning_rate": 2.8982456140350878e-05, + "loss": 3.4038, + "step": 16522 + }, + { + "epoch": 0.8696315789473684, + "grad_norm": 0.4765625, + "learning_rate": 2.897076023391813e-05, + "loss": 3.6708, + "step": 16523 + }, + { + "epoch": 0.8696842105263158, + "grad_norm": 0.478515625, + "learning_rate": 2.895906432748538e-05, + "loss": 3.142, + "step": 16524 + }, + { + "epoch": 0.8697368421052631, + "grad_norm": 0.486328125, + "learning_rate": 2.8947368421052634e-05, + "loss": 3.0483, + "step": 16525 + }, + { + "epoch": 0.8697894736842106, + "grad_norm": 0.486328125, + "learning_rate": 2.8935672514619883e-05, + "loss": 3.0556, + "step": 16526 + }, + { + "epoch": 0.8698421052631579, + "grad_norm": 0.51953125, + "learning_rate": 2.8923976608187136e-05, + "loss": 3.1999, + "step": 16527 + }, + { + "epoch": 0.8698947368421053, + "grad_norm": 0.4609375, + "learning_rate": 2.8912280701754386e-05, + "loss": 2.8097, + "step": 16528 + }, + { + "epoch": 0.8699473684210526, + "grad_norm": 0.515625, + "learning_rate": 2.890058479532164e-05, + "loss": 3.4438, + "step": 16529 + }, + { + "epoch": 0.87, + "grad_norm": 0.5078125, + "learning_rate": 2.8888888888888888e-05, + "loss": 2.9909, + "step": 16530 + }, + { + "epoch": 0.8700526315789474, + "grad_norm": 0.50390625, + "learning_rate": 2.887719298245614e-05, + "loss": 3.6399, + "step": 16531 + }, + { + "epoch": 0.8701052631578947, + "grad_norm": 0.50390625, + "learning_rate": 2.8865497076023394e-05, + "loss": 3.5465, + "step": 16532 + }, + { + "epoch": 0.8701578947368421, + "grad_norm": 0.5234375, + "learning_rate": 2.8853801169590644e-05, + "loss": 3.2871, + "step": 16533 + }, + { + "epoch": 0.8702105263157894, + "grad_norm": 0.494140625, + "learning_rate": 2.8842105263157897e-05, + "loss": 3.4616, + "step": 16534 + }, + { + "epoch": 0.8702631578947368, + "grad_norm": 0.4765625, + "learning_rate": 2.8830409356725146e-05, + "loss": 3.4288, + "step": 16535 + }, + { + "epoch": 0.8703157894736843, + "grad_norm": 0.54296875, + "learning_rate": 2.88187134502924e-05, + "loss": 2.8363, + "step": 16536 + }, + { + "epoch": 0.8703684210526316, + "grad_norm": 0.4921875, + "learning_rate": 2.880701754385965e-05, + "loss": 3.0822, + "step": 16537 + }, + { + "epoch": 0.870421052631579, + "grad_norm": 0.498046875, + "learning_rate": 2.8795321637426902e-05, + "loss": 2.9958, + "step": 16538 + }, + { + "epoch": 0.8704736842105263, + "grad_norm": 0.484375, + "learning_rate": 2.878362573099415e-05, + "loss": 3.2542, + "step": 16539 + }, + { + "epoch": 0.8705263157894737, + "grad_norm": 0.474609375, + "learning_rate": 2.8771929824561404e-05, + "loss": 3.1464, + "step": 16540 + }, + { + "epoch": 0.870578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.8760233918128654e-05, + "loss": 2.973, + "step": 16541 + }, + { + "epoch": 0.8706315789473684, + "grad_norm": 0.51953125, + "learning_rate": 2.8748538011695907e-05, + "loss": 3.5299, + "step": 16542 + }, + { + "epoch": 0.8706842105263158, + "grad_norm": 0.51953125, + "learning_rate": 2.8736842105263163e-05, + "loss": 3.0222, + "step": 16543 + }, + { + "epoch": 0.8707368421052631, + "grad_norm": 0.4765625, + "learning_rate": 2.872514619883041e-05, + "loss": 3.246, + "step": 16544 + }, + { + "epoch": 0.8707894736842106, + "grad_norm": 0.5078125, + "learning_rate": 2.8713450292397666e-05, + "loss": 3.5559, + "step": 16545 + }, + { + "epoch": 0.8708421052631579, + "grad_norm": 0.5234375, + "learning_rate": 2.8701754385964912e-05, + "loss": 3.0039, + "step": 16546 + }, + { + "epoch": 0.8708947368421053, + "grad_norm": 0.4765625, + "learning_rate": 2.869005847953217e-05, + "loss": 2.8039, + "step": 16547 + }, + { + "epoch": 0.8709473684210526, + "grad_norm": 0.494140625, + "learning_rate": 2.8678362573099415e-05, + "loss": 3.3564, + "step": 16548 + }, + { + "epoch": 0.871, + "grad_norm": 0.5078125, + "learning_rate": 2.8666666666666668e-05, + "loss": 3.5042, + "step": 16549 + }, + { + "epoch": 0.8710526315789474, + "grad_norm": 0.4765625, + "learning_rate": 2.8654970760233917e-05, + "loss": 2.7727, + "step": 16550 + }, + { + "epoch": 0.8711052631578947, + "grad_norm": 0.50390625, + "learning_rate": 2.864327485380117e-05, + "loss": 3.1897, + "step": 16551 + }, + { + "epoch": 0.8711578947368421, + "grad_norm": 0.52734375, + "learning_rate": 2.863157894736842e-05, + "loss": 3.3622, + "step": 16552 + }, + { + "epoch": 0.8712105263157894, + "grad_norm": 0.546875, + "learning_rate": 2.8619883040935673e-05, + "loss": 3.6042, + "step": 16553 + }, + { + "epoch": 0.8712631578947369, + "grad_norm": 0.4921875, + "learning_rate": 2.8608187134502923e-05, + "loss": 3.5723, + "step": 16554 + }, + { + "epoch": 0.8713157894736843, + "grad_norm": 0.49609375, + "learning_rate": 2.8596491228070175e-05, + "loss": 3.3, + "step": 16555 + }, + { + "epoch": 0.8713684210526316, + "grad_norm": 0.5234375, + "learning_rate": 2.8584795321637432e-05, + "loss": 3.2978, + "step": 16556 + }, + { + "epoch": 0.871421052631579, + "grad_norm": 0.49609375, + "learning_rate": 2.8573099415204678e-05, + "loss": 3.3402, + "step": 16557 + }, + { + "epoch": 0.8714736842105263, + "grad_norm": 0.51953125, + "learning_rate": 2.8561403508771934e-05, + "loss": 2.9474, + "step": 16558 + }, + { + "epoch": 0.8715263157894737, + "grad_norm": 0.4609375, + "learning_rate": 2.854970760233918e-05, + "loss": 3.0086, + "step": 16559 + }, + { + "epoch": 0.871578947368421, + "grad_norm": 0.5, + "learning_rate": 2.8538011695906437e-05, + "loss": 3.071, + "step": 16560 + }, + { + "epoch": 0.8716315789473684, + "grad_norm": 0.5, + "learning_rate": 2.8526315789473683e-05, + "loss": 3.1808, + "step": 16561 + }, + { + "epoch": 0.8716842105263158, + "grad_norm": 0.49609375, + "learning_rate": 2.851461988304094e-05, + "loss": 3.4862, + "step": 16562 + }, + { + "epoch": 0.8717368421052631, + "grad_norm": 0.498046875, + "learning_rate": 2.8502923976608186e-05, + "loss": 3.1909, + "step": 16563 + }, + { + "epoch": 0.8717894736842106, + "grad_norm": 0.51171875, + "learning_rate": 2.8491228070175442e-05, + "loss": 3.1853, + "step": 16564 + }, + { + "epoch": 0.8718421052631579, + "grad_norm": 0.490234375, + "learning_rate": 2.847953216374269e-05, + "loss": 3.1664, + "step": 16565 + }, + { + "epoch": 0.8718947368421053, + "grad_norm": 0.515625, + "learning_rate": 2.8467836257309945e-05, + "loss": 3.4397, + "step": 16566 + }, + { + "epoch": 0.8719473684210526, + "grad_norm": 0.498046875, + "learning_rate": 2.845614035087719e-05, + "loss": 3.2063, + "step": 16567 + }, + { + "epoch": 0.872, + "grad_norm": 0.54296875, + "learning_rate": 2.8444444444444447e-05, + "loss": 3.1491, + "step": 16568 + }, + { + "epoch": 0.8720526315789474, + "grad_norm": 0.48828125, + "learning_rate": 2.84327485380117e-05, + "loss": 3.0589, + "step": 16569 + }, + { + "epoch": 0.8721052631578947, + "grad_norm": 0.5078125, + "learning_rate": 2.842105263157895e-05, + "loss": 2.7431, + "step": 16570 + }, + { + "epoch": 0.8721578947368421, + "grad_norm": 0.4921875, + "learning_rate": 2.8409356725146203e-05, + "loss": 3.5982, + "step": 16571 + }, + { + "epoch": 0.8722105263157894, + "grad_norm": 0.51171875, + "learning_rate": 2.839766081871345e-05, + "loss": 3.6412, + "step": 16572 + }, + { + "epoch": 0.8722631578947369, + "grad_norm": 0.50390625, + "learning_rate": 2.8385964912280705e-05, + "loss": 3.6451, + "step": 16573 + }, + { + "epoch": 0.8723157894736842, + "grad_norm": 0.470703125, + "learning_rate": 2.837426900584795e-05, + "loss": 3.2579, + "step": 16574 + }, + { + "epoch": 0.8723684210526316, + "grad_norm": 0.498046875, + "learning_rate": 2.8362573099415208e-05, + "loss": 3.3297, + "step": 16575 + }, + { + "epoch": 0.872421052631579, + "grad_norm": 0.53125, + "learning_rate": 2.8350877192982454e-05, + "loss": 3.2269, + "step": 16576 + }, + { + "epoch": 0.8724736842105263, + "grad_norm": 0.4921875, + "learning_rate": 2.833918128654971e-05, + "loss": 3.3478, + "step": 16577 + }, + { + "epoch": 0.8725263157894737, + "grad_norm": 0.5, + "learning_rate": 2.8327485380116957e-05, + "loss": 3.5035, + "step": 16578 + }, + { + "epoch": 0.872578947368421, + "grad_norm": 0.47265625, + "learning_rate": 2.8315789473684213e-05, + "loss": 3.0706, + "step": 16579 + }, + { + "epoch": 0.8726315789473684, + "grad_norm": 0.50390625, + "learning_rate": 2.830409356725146e-05, + "loss": 2.944, + "step": 16580 + }, + { + "epoch": 0.8726842105263158, + "grad_norm": 0.50390625, + "learning_rate": 2.8292397660818716e-05, + "loss": 3.2094, + "step": 16581 + }, + { + "epoch": 0.8727368421052631, + "grad_norm": 0.515625, + "learning_rate": 2.828070175438597e-05, + "loss": 3.3139, + "step": 16582 + }, + { + "epoch": 0.8727894736842106, + "grad_norm": 0.5, + "learning_rate": 2.8269005847953218e-05, + "loss": 3.6033, + "step": 16583 + }, + { + "epoch": 0.8728421052631579, + "grad_norm": 0.4921875, + "learning_rate": 2.825730994152047e-05, + "loss": 3.3061, + "step": 16584 + }, + { + "epoch": 0.8728947368421053, + "grad_norm": 0.482421875, + "learning_rate": 2.824561403508772e-05, + "loss": 3.5548, + "step": 16585 + }, + { + "epoch": 0.8729473684210526, + "grad_norm": 0.55078125, + "learning_rate": 2.8233918128654974e-05, + "loss": 3.5664, + "step": 16586 + }, + { + "epoch": 0.873, + "grad_norm": 0.5, + "learning_rate": 2.8222222222222223e-05, + "loss": 3.3432, + "step": 16587 + }, + { + "epoch": 0.8730526315789474, + "grad_norm": 0.490234375, + "learning_rate": 2.8210526315789476e-05, + "loss": 3.1606, + "step": 16588 + }, + { + "epoch": 0.8731052631578947, + "grad_norm": 0.48046875, + "learning_rate": 2.8198830409356726e-05, + "loss": 3.4736, + "step": 16589 + }, + { + "epoch": 0.8731578947368421, + "grad_norm": 0.484375, + "learning_rate": 2.818713450292398e-05, + "loss": 3.1581, + "step": 16590 + }, + { + "epoch": 0.8732105263157894, + "grad_norm": 0.51171875, + "learning_rate": 2.817543859649123e-05, + "loss": 3.3495, + "step": 16591 + }, + { + "epoch": 0.8732631578947369, + "grad_norm": 0.46875, + "learning_rate": 2.816374269005848e-05, + "loss": 3.4143, + "step": 16592 + }, + { + "epoch": 0.8733157894736842, + "grad_norm": 0.486328125, + "learning_rate": 2.8152046783625735e-05, + "loss": 2.9859, + "step": 16593 + }, + { + "epoch": 0.8733684210526316, + "grad_norm": 0.49609375, + "learning_rate": 2.8140350877192984e-05, + "loss": 2.9448, + "step": 16594 + }, + { + "epoch": 0.873421052631579, + "grad_norm": 0.57421875, + "learning_rate": 2.8128654970760237e-05, + "loss": 3.2185, + "step": 16595 + }, + { + "epoch": 0.8734736842105263, + "grad_norm": 0.49609375, + "learning_rate": 2.8116959064327487e-05, + "loss": 3.2013, + "step": 16596 + }, + { + "epoch": 0.8735263157894737, + "grad_norm": 0.490234375, + "learning_rate": 2.810526315789474e-05, + "loss": 3.2108, + "step": 16597 + }, + { + "epoch": 0.873578947368421, + "grad_norm": 0.5234375, + "learning_rate": 2.809356725146199e-05, + "loss": 3.1601, + "step": 16598 + }, + { + "epoch": 0.8736315789473684, + "grad_norm": 0.5, + "learning_rate": 2.8081871345029242e-05, + "loss": 3.2056, + "step": 16599 + }, + { + "epoch": 0.8736842105263158, + "grad_norm": 0.49609375, + "learning_rate": 2.8070175438596492e-05, + "loss": 2.8818, + "step": 16600 + }, + { + "epoch": 0.8737368421052631, + "grad_norm": 0.5, + "learning_rate": 2.8058479532163745e-05, + "loss": 2.9401, + "step": 16601 + }, + { + "epoch": 0.8737894736842106, + "grad_norm": 0.484375, + "learning_rate": 2.8046783625730994e-05, + "loss": 3.3262, + "step": 16602 + }, + { + "epoch": 0.8738421052631579, + "grad_norm": 0.482421875, + "learning_rate": 2.8035087719298247e-05, + "loss": 3.32, + "step": 16603 + }, + { + "epoch": 0.8738947368421053, + "grad_norm": 0.498046875, + "learning_rate": 2.8023391812865497e-05, + "loss": 2.947, + "step": 16604 + }, + { + "epoch": 0.8739473684210526, + "grad_norm": 0.486328125, + "learning_rate": 2.801169590643275e-05, + "loss": 3.3418, + "step": 16605 + }, + { + "epoch": 0.874, + "grad_norm": 0.47265625, + "learning_rate": 2.8000000000000003e-05, + "loss": 3.0125, + "step": 16606 + }, + { + "epoch": 0.8740526315789474, + "grad_norm": 0.515625, + "learning_rate": 2.7988304093567253e-05, + "loss": 3.526, + "step": 16607 + }, + { + "epoch": 0.8741052631578947, + "grad_norm": 0.51953125, + "learning_rate": 2.7976608187134506e-05, + "loss": 3.0243, + "step": 16608 + }, + { + "epoch": 0.8741578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.7964912280701755e-05, + "loss": 3.6239, + "step": 16609 + }, + { + "epoch": 0.8742105263157894, + "grad_norm": 0.4921875, + "learning_rate": 2.7953216374269008e-05, + "loss": 3.049, + "step": 16610 + }, + { + "epoch": 0.8742631578947369, + "grad_norm": 0.50390625, + "learning_rate": 2.7941520467836258e-05, + "loss": 2.8591, + "step": 16611 + }, + { + "epoch": 0.8743157894736842, + "grad_norm": 0.498046875, + "learning_rate": 2.792982456140351e-05, + "loss": 2.7899, + "step": 16612 + }, + { + "epoch": 0.8743684210526316, + "grad_norm": 0.54296875, + "learning_rate": 2.791812865497076e-05, + "loss": 2.8088, + "step": 16613 + }, + { + "epoch": 0.874421052631579, + "grad_norm": 0.515625, + "learning_rate": 2.7906432748538013e-05, + "loss": 3.3083, + "step": 16614 + }, + { + "epoch": 0.8744736842105263, + "grad_norm": 0.50390625, + "learning_rate": 2.7894736842105263e-05, + "loss": 3.0108, + "step": 16615 + }, + { + "epoch": 0.8745263157894737, + "grad_norm": 0.49609375, + "learning_rate": 2.7883040935672516e-05, + "loss": 3.5866, + "step": 16616 + }, + { + "epoch": 0.874578947368421, + "grad_norm": 0.4921875, + "learning_rate": 2.7871345029239765e-05, + "loss": 2.9545, + "step": 16617 + }, + { + "epoch": 0.8746315789473684, + "grad_norm": 0.470703125, + "learning_rate": 2.785964912280702e-05, + "loss": 2.7401, + "step": 16618 + }, + { + "epoch": 0.8746842105263158, + "grad_norm": 0.5390625, + "learning_rate": 2.784795321637427e-05, + "loss": 3.2958, + "step": 16619 + }, + { + "epoch": 0.8747368421052631, + "grad_norm": 0.51953125, + "learning_rate": 2.783625730994152e-05, + "loss": 2.8956, + "step": 16620 + }, + { + "epoch": 0.8747894736842106, + "grad_norm": 0.482421875, + "learning_rate": 2.7824561403508774e-05, + "loss": 3.2771, + "step": 16621 + }, + { + "epoch": 0.8748421052631579, + "grad_norm": 0.50390625, + "learning_rate": 2.7812865497076024e-05, + "loss": 3.3044, + "step": 16622 + }, + { + "epoch": 0.8748947368421053, + "grad_norm": 0.48046875, + "learning_rate": 2.7801169590643277e-05, + "loss": 2.7496, + "step": 16623 + }, + { + "epoch": 0.8749473684210526, + "grad_norm": 0.515625, + "learning_rate": 2.7789473684210526e-05, + "loss": 3.1775, + "step": 16624 + }, + { + "epoch": 0.875, + "grad_norm": 0.494140625, + "learning_rate": 2.777777777777778e-05, + "loss": 3.3022, + "step": 16625 + }, + { + "epoch": 0.8750526315789474, + "grad_norm": 0.52734375, + "learning_rate": 2.776608187134503e-05, + "loss": 3.5976, + "step": 16626 + }, + { + "epoch": 0.8751052631578947, + "grad_norm": 0.54296875, + "learning_rate": 2.775438596491228e-05, + "loss": 3.3498, + "step": 16627 + }, + { + "epoch": 0.8751578947368421, + "grad_norm": 0.53125, + "learning_rate": 2.774269005847953e-05, + "loss": 3.6119, + "step": 16628 + }, + { + "epoch": 0.8752105263157894, + "grad_norm": 0.51171875, + "learning_rate": 2.7730994152046784e-05, + "loss": 3.6318, + "step": 16629 + }, + { + "epoch": 0.8752631578947369, + "grad_norm": 0.55078125, + "learning_rate": 2.7719298245614034e-05, + "loss": 3.3795, + "step": 16630 + }, + { + "epoch": 0.8753157894736842, + "grad_norm": 0.5078125, + "learning_rate": 2.7707602339181287e-05, + "loss": 3.6475, + "step": 16631 + }, + { + "epoch": 0.8753684210526316, + "grad_norm": 0.5078125, + "learning_rate": 2.7695906432748543e-05, + "loss": 3.4097, + "step": 16632 + }, + { + "epoch": 0.875421052631579, + "grad_norm": 0.515625, + "learning_rate": 2.768421052631579e-05, + "loss": 3.5136, + "step": 16633 + }, + { + "epoch": 0.8754736842105263, + "grad_norm": 0.50390625, + "learning_rate": 2.7672514619883046e-05, + "loss": 3.4532, + "step": 16634 + }, + { + "epoch": 0.8755263157894737, + "grad_norm": 0.51171875, + "learning_rate": 2.7660818713450292e-05, + "loss": 3.2467, + "step": 16635 + }, + { + "epoch": 0.875578947368421, + "grad_norm": 0.65234375, + "learning_rate": 2.764912280701755e-05, + "loss": 3.5449, + "step": 16636 + }, + { + "epoch": 0.8756315789473684, + "grad_norm": 0.50390625, + "learning_rate": 2.7637426900584795e-05, + "loss": 3.3628, + "step": 16637 + }, + { + "epoch": 0.8756842105263158, + "grad_norm": 0.490234375, + "learning_rate": 2.762573099415205e-05, + "loss": 3.313, + "step": 16638 + }, + { + "epoch": 0.8757368421052631, + "grad_norm": 0.51171875, + "learning_rate": 2.7614035087719297e-05, + "loss": 3.0715, + "step": 16639 + }, + { + "epoch": 0.8757894736842106, + "grad_norm": 0.486328125, + "learning_rate": 2.760233918128655e-05, + "loss": 3.5433, + "step": 16640 + }, + { + "epoch": 0.8758421052631579, + "grad_norm": 0.5078125, + "learning_rate": 2.75906432748538e-05, + "loss": 3.6539, + "step": 16641 + }, + { + "epoch": 0.8758947368421053, + "grad_norm": 0.51953125, + "learning_rate": 2.7578947368421053e-05, + "loss": 3.3206, + "step": 16642 + }, + { + "epoch": 0.8759473684210526, + "grad_norm": 0.51171875, + "learning_rate": 2.7567251461988302e-05, + "loss": 3.4604, + "step": 16643 + }, + { + "epoch": 0.876, + "grad_norm": 0.4765625, + "learning_rate": 2.7555555555555555e-05, + "loss": 3.3098, + "step": 16644 + }, + { + "epoch": 0.8760526315789474, + "grad_norm": 0.50390625, + "learning_rate": 2.754385964912281e-05, + "loss": 3.1294, + "step": 16645 + }, + { + "epoch": 0.8761052631578947, + "grad_norm": 0.486328125, + "learning_rate": 2.7532163742690058e-05, + "loss": 3.3397, + "step": 16646 + }, + { + "epoch": 0.8761578947368421, + "grad_norm": 0.498046875, + "learning_rate": 2.7520467836257314e-05, + "loss": 3.1904, + "step": 16647 + }, + { + "epoch": 0.8762105263157894, + "grad_norm": 0.4765625, + "learning_rate": 2.750877192982456e-05, + "loss": 2.9608, + "step": 16648 + }, + { + "epoch": 0.8762631578947369, + "grad_norm": 0.51171875, + "learning_rate": 2.7497076023391817e-05, + "loss": 3.5726, + "step": 16649 + }, + { + "epoch": 0.8763157894736842, + "grad_norm": 0.51953125, + "learning_rate": 2.7485380116959063e-05, + "loss": 3.3426, + "step": 16650 + }, + { + "epoch": 0.8763684210526316, + "grad_norm": 0.4765625, + "learning_rate": 2.747368421052632e-05, + "loss": 3.344, + "step": 16651 + }, + { + "epoch": 0.876421052631579, + "grad_norm": 0.5703125, + "learning_rate": 2.7461988304093566e-05, + "loss": 3.1026, + "step": 16652 + }, + { + "epoch": 0.8764736842105263, + "grad_norm": 0.49609375, + "learning_rate": 2.7450292397660822e-05, + "loss": 3.3772, + "step": 16653 + }, + { + "epoch": 0.8765263157894737, + "grad_norm": 0.486328125, + "learning_rate": 2.7438596491228068e-05, + "loss": 3.1073, + "step": 16654 + }, + { + "epoch": 0.876578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.7426900584795324e-05, + "loss": 3.3045, + "step": 16655 + }, + { + "epoch": 0.8766315789473684, + "grad_norm": 0.498046875, + "learning_rate": 2.7415204678362577e-05, + "loss": 2.9484, + "step": 16656 + }, + { + "epoch": 0.8766842105263158, + "grad_norm": 0.4921875, + "learning_rate": 2.7403508771929827e-05, + "loss": 3.498, + "step": 16657 + }, + { + "epoch": 0.8767368421052631, + "grad_norm": 0.51171875, + "learning_rate": 2.739181286549708e-05, + "loss": 3.4372, + "step": 16658 + }, + { + "epoch": 0.8767894736842106, + "grad_norm": 0.478515625, + "learning_rate": 2.738011695906433e-05, + "loss": 3.0595, + "step": 16659 + }, + { + "epoch": 0.8768421052631579, + "grad_norm": 0.5, + "learning_rate": 2.7368421052631583e-05, + "loss": 3.5139, + "step": 16660 + }, + { + "epoch": 0.8768947368421053, + "grad_norm": 0.494140625, + "learning_rate": 2.7356725146198832e-05, + "loss": 3.2357, + "step": 16661 + }, + { + "epoch": 0.8769473684210526, + "grad_norm": 0.50390625, + "learning_rate": 2.7345029239766085e-05, + "loss": 3.7196, + "step": 16662 + }, + { + "epoch": 0.877, + "grad_norm": 0.51953125, + "learning_rate": 2.733333333333333e-05, + "loss": 3.4039, + "step": 16663 + }, + { + "epoch": 0.8770526315789474, + "grad_norm": 0.59765625, + "learning_rate": 2.7321637426900588e-05, + "loss": 3.2073, + "step": 16664 + }, + { + "epoch": 0.8771052631578947, + "grad_norm": 0.53125, + "learning_rate": 2.7309941520467834e-05, + "loss": 3.7238, + "step": 16665 + }, + { + "epoch": 0.8771578947368421, + "grad_norm": 0.494140625, + "learning_rate": 2.729824561403509e-05, + "loss": 3.2025, + "step": 16666 + }, + { + "epoch": 0.8772105263157894, + "grad_norm": 0.482421875, + "learning_rate": 2.7286549707602337e-05, + "loss": 3.5507, + "step": 16667 + }, + { + "epoch": 0.8772631578947369, + "grad_norm": 0.51171875, + "learning_rate": 2.7274853801169593e-05, + "loss": 3.3444, + "step": 16668 + }, + { + "epoch": 0.8773157894736842, + "grad_norm": 0.58203125, + "learning_rate": 2.7263157894736846e-05, + "loss": 2.7866, + "step": 16669 + }, + { + "epoch": 0.8773684210526316, + "grad_norm": 0.498046875, + "learning_rate": 2.7251461988304095e-05, + "loss": 3.4108, + "step": 16670 + }, + { + "epoch": 0.877421052631579, + "grad_norm": 0.490234375, + "learning_rate": 2.723976608187135e-05, + "loss": 3.4854, + "step": 16671 + }, + { + "epoch": 0.8774736842105263, + "grad_norm": 0.52734375, + "learning_rate": 2.7228070175438598e-05, + "loss": 2.7543, + "step": 16672 + }, + { + "epoch": 0.8775263157894737, + "grad_norm": 0.5, + "learning_rate": 2.721637426900585e-05, + "loss": 3.1165, + "step": 16673 + }, + { + "epoch": 0.877578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.72046783625731e-05, + "loss": 3.2045, + "step": 16674 + }, + { + "epoch": 0.8776315789473684, + "grad_norm": 0.484375, + "learning_rate": 2.7192982456140354e-05, + "loss": 3.2625, + "step": 16675 + }, + { + "epoch": 0.8776842105263158, + "grad_norm": 0.58984375, + "learning_rate": 2.7181286549707603e-05, + "loss": 3.0781, + "step": 16676 + }, + { + "epoch": 0.8777368421052631, + "grad_norm": 0.57421875, + "learning_rate": 2.7169590643274856e-05, + "loss": 2.862, + "step": 16677 + }, + { + "epoch": 0.8777894736842106, + "grad_norm": 0.51953125, + "learning_rate": 2.7157894736842106e-05, + "loss": 3.0335, + "step": 16678 + }, + { + "epoch": 0.8778421052631579, + "grad_norm": 0.54296875, + "learning_rate": 2.714619883040936e-05, + "loss": 3.2517, + "step": 16679 + }, + { + "epoch": 0.8778947368421053, + "grad_norm": 0.54296875, + "learning_rate": 2.713450292397661e-05, + "loss": 3.1659, + "step": 16680 + }, + { + "epoch": 0.8779473684210526, + "grad_norm": 0.51953125, + "learning_rate": 2.712280701754386e-05, + "loss": 3.1644, + "step": 16681 + }, + { + "epoch": 0.878, + "grad_norm": 0.51953125, + "learning_rate": 2.7111111111111114e-05, + "loss": 3.1486, + "step": 16682 + }, + { + "epoch": 0.8780526315789474, + "grad_norm": 0.494140625, + "learning_rate": 2.7099415204678364e-05, + "loss": 2.9472, + "step": 16683 + }, + { + "epoch": 0.8781052631578947, + "grad_norm": 0.515625, + "learning_rate": 2.7087719298245617e-05, + "loss": 3.121, + "step": 16684 + }, + { + "epoch": 0.8781578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.7076023391812866e-05, + "loss": 3.1431, + "step": 16685 + }, + { + "epoch": 0.8782105263157894, + "grad_norm": 0.494140625, + "learning_rate": 2.706432748538012e-05, + "loss": 3.1307, + "step": 16686 + }, + { + "epoch": 0.8782631578947369, + "grad_norm": 0.494140625, + "learning_rate": 2.705263157894737e-05, + "loss": 3.1033, + "step": 16687 + }, + { + "epoch": 0.8783157894736842, + "grad_norm": 0.50390625, + "learning_rate": 2.7040935672514622e-05, + "loss": 2.8851, + "step": 16688 + }, + { + "epoch": 0.8783684210526316, + "grad_norm": 0.5390625, + "learning_rate": 2.702923976608187e-05, + "loss": 3.362, + "step": 16689 + }, + { + "epoch": 0.878421052631579, + "grad_norm": 0.5078125, + "learning_rate": 2.7017543859649125e-05, + "loss": 3.3086, + "step": 16690 + }, + { + "epoch": 0.8784736842105263, + "grad_norm": 0.498046875, + "learning_rate": 2.7005847953216374e-05, + "loss": 3.1248, + "step": 16691 + }, + { + "epoch": 0.8785263157894737, + "grad_norm": 0.5234375, + "learning_rate": 2.6994152046783627e-05, + "loss": 3.304, + "step": 16692 + }, + { + "epoch": 0.878578947368421, + "grad_norm": 0.4765625, + "learning_rate": 2.6982456140350877e-05, + "loss": 3.2172, + "step": 16693 + }, + { + "epoch": 0.8786315789473684, + "grad_norm": 0.50390625, + "learning_rate": 2.697076023391813e-05, + "loss": 3.3818, + "step": 16694 + }, + { + "epoch": 0.8786842105263157, + "grad_norm": 0.5, + "learning_rate": 2.6959064327485383e-05, + "loss": 2.964, + "step": 16695 + }, + { + "epoch": 0.8787368421052631, + "grad_norm": 0.515625, + "learning_rate": 2.6947368421052632e-05, + "loss": 3.2782, + "step": 16696 + }, + { + "epoch": 0.8787894736842106, + "grad_norm": 0.51171875, + "learning_rate": 2.6935672514619885e-05, + "loss": 3.0277, + "step": 16697 + }, + { + "epoch": 0.8788421052631579, + "grad_norm": 0.51953125, + "learning_rate": 2.6923976608187135e-05, + "loss": 3.6688, + "step": 16698 + }, + { + "epoch": 0.8788947368421053, + "grad_norm": 0.5, + "learning_rate": 2.6912280701754388e-05, + "loss": 3.4706, + "step": 16699 + }, + { + "epoch": 0.8789473684210526, + "grad_norm": 0.4921875, + "learning_rate": 2.6900584795321637e-05, + "loss": 3.3661, + "step": 16700 + }, + { + "epoch": 0.879, + "grad_norm": 0.50390625, + "learning_rate": 2.688888888888889e-05, + "loss": 2.8617, + "step": 16701 + }, + { + "epoch": 0.8790526315789474, + "grad_norm": 0.50390625, + "learning_rate": 2.687719298245614e-05, + "loss": 3.2528, + "step": 16702 + }, + { + "epoch": 0.8791052631578947, + "grad_norm": 0.50390625, + "learning_rate": 2.6865497076023393e-05, + "loss": 3.2625, + "step": 16703 + }, + { + "epoch": 0.8791578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.6853801169590643e-05, + "loss": 3.0257, + "step": 16704 + }, + { + "epoch": 0.8792105263157894, + "grad_norm": 0.5234375, + "learning_rate": 2.6842105263157896e-05, + "loss": 2.8481, + "step": 16705 + }, + { + "epoch": 0.8792631578947369, + "grad_norm": 0.5, + "learning_rate": 2.683040935672515e-05, + "loss": 3.3075, + "step": 16706 + }, + { + "epoch": 0.8793157894736842, + "grad_norm": 0.49609375, + "learning_rate": 2.6818713450292398e-05, + "loss": 3.2538, + "step": 16707 + }, + { + "epoch": 0.8793684210526316, + "grad_norm": 0.490234375, + "learning_rate": 2.680701754385965e-05, + "loss": 3.0276, + "step": 16708 + }, + { + "epoch": 0.879421052631579, + "grad_norm": 0.466796875, + "learning_rate": 2.67953216374269e-05, + "loss": 2.9458, + "step": 16709 + }, + { + "epoch": 0.8794736842105263, + "grad_norm": 0.490234375, + "learning_rate": 2.6783625730994154e-05, + "loss": 3.1836, + "step": 16710 + }, + { + "epoch": 0.8795263157894737, + "grad_norm": 0.494140625, + "learning_rate": 2.6771929824561403e-05, + "loss": 3.3042, + "step": 16711 + }, + { + "epoch": 0.879578947368421, + "grad_norm": 0.5078125, + "learning_rate": 2.6760233918128656e-05, + "loss": 3.3614, + "step": 16712 + }, + { + "epoch": 0.8796315789473684, + "grad_norm": 0.5078125, + "learning_rate": 2.6748538011695906e-05, + "loss": 3.5379, + "step": 16713 + }, + { + "epoch": 0.8796842105263157, + "grad_norm": 0.51953125, + "learning_rate": 2.673684210526316e-05, + "loss": 3.2183, + "step": 16714 + }, + { + "epoch": 0.8797368421052632, + "grad_norm": 0.51953125, + "learning_rate": 2.672514619883041e-05, + "loss": 3.847, + "step": 16715 + }, + { + "epoch": 0.8797894736842106, + "grad_norm": 0.48828125, + "learning_rate": 2.671345029239766e-05, + "loss": 3.1264, + "step": 16716 + }, + { + "epoch": 0.8798421052631579, + "grad_norm": 0.486328125, + "learning_rate": 2.670175438596491e-05, + "loss": 3.0906, + "step": 16717 + }, + { + "epoch": 0.8798947368421053, + "grad_norm": 0.49609375, + "learning_rate": 2.6690058479532164e-05, + "loss": 2.8372, + "step": 16718 + }, + { + "epoch": 0.8799473684210526, + "grad_norm": 0.498046875, + "learning_rate": 2.667836257309942e-05, + "loss": 2.9356, + "step": 16719 + }, + { + "epoch": 0.88, + "grad_norm": 0.5078125, + "learning_rate": 2.6666666666666667e-05, + "loss": 3.2266, + "step": 16720 + }, + { + "epoch": 0.8800526315789474, + "grad_norm": 0.5078125, + "learning_rate": 2.6654970760233923e-05, + "loss": 2.9354, + "step": 16721 + }, + { + "epoch": 0.8801052631578947, + "grad_norm": 0.53515625, + "learning_rate": 2.664327485380117e-05, + "loss": 3.3525, + "step": 16722 + }, + { + "epoch": 0.8801578947368421, + "grad_norm": 0.515625, + "learning_rate": 2.6631578947368426e-05, + "loss": 3.3998, + "step": 16723 + }, + { + "epoch": 0.8802105263157894, + "grad_norm": 0.5234375, + "learning_rate": 2.6619883040935672e-05, + "loss": 3.1099, + "step": 16724 + }, + { + "epoch": 0.8802631578947369, + "grad_norm": 0.52734375, + "learning_rate": 2.6608187134502928e-05, + "loss": 2.7028, + "step": 16725 + }, + { + "epoch": 0.8803157894736842, + "grad_norm": 0.515625, + "learning_rate": 2.6596491228070174e-05, + "loss": 3.3422, + "step": 16726 + }, + { + "epoch": 0.8803684210526316, + "grad_norm": 0.474609375, + "learning_rate": 2.658479532163743e-05, + "loss": 3.4729, + "step": 16727 + }, + { + "epoch": 0.880421052631579, + "grad_norm": 0.50390625, + "learning_rate": 2.6573099415204677e-05, + "loss": 3.3582, + "step": 16728 + }, + { + "epoch": 0.8804736842105263, + "grad_norm": 0.484375, + "learning_rate": 2.656140350877193e-05, + "loss": 3.1082, + "step": 16729 + }, + { + "epoch": 0.8805263157894737, + "grad_norm": 0.4453125, + "learning_rate": 2.654970760233918e-05, + "loss": 3.0319, + "step": 16730 + }, + { + "epoch": 0.880578947368421, + "grad_norm": 0.48828125, + "learning_rate": 2.6538011695906432e-05, + "loss": 3.4722, + "step": 16731 + }, + { + "epoch": 0.8806315789473684, + "grad_norm": 0.490234375, + "learning_rate": 2.652631578947369e-05, + "loss": 3.3556, + "step": 16732 + }, + { + "epoch": 0.8806842105263157, + "grad_norm": 0.482421875, + "learning_rate": 2.6514619883040935e-05, + "loss": 3.3031, + "step": 16733 + }, + { + "epoch": 0.8807368421052632, + "grad_norm": 0.5078125, + "learning_rate": 2.650292397660819e-05, + "loss": 3.2281, + "step": 16734 + }, + { + "epoch": 0.8807894736842106, + "grad_norm": 0.51171875, + "learning_rate": 2.6491228070175438e-05, + "loss": 3.2277, + "step": 16735 + }, + { + "epoch": 0.8808421052631579, + "grad_norm": 0.49609375, + "learning_rate": 2.6479532163742694e-05, + "loss": 3.4756, + "step": 16736 + }, + { + "epoch": 0.8808947368421053, + "grad_norm": 0.51953125, + "learning_rate": 2.646783625730994e-05, + "loss": 3.455, + "step": 16737 + }, + { + "epoch": 0.8809473684210526, + "grad_norm": 0.50390625, + "learning_rate": 2.6456140350877197e-05, + "loss": 3.1316, + "step": 16738 + }, + { + "epoch": 0.881, + "grad_norm": 0.50390625, + "learning_rate": 2.6444444444444443e-05, + "loss": 3.1252, + "step": 16739 + }, + { + "epoch": 0.8810526315789474, + "grad_norm": 0.5, + "learning_rate": 2.64327485380117e-05, + "loss": 3.5043, + "step": 16740 + }, + { + "epoch": 0.8811052631578947, + "grad_norm": 0.46875, + "learning_rate": 2.6421052631578945e-05, + "loss": 3.4687, + "step": 16741 + }, + { + "epoch": 0.8811578947368421, + "grad_norm": 0.49609375, + "learning_rate": 2.64093567251462e-05, + "loss": 3.5628, + "step": 16742 + }, + { + "epoch": 0.8812105263157894, + "grad_norm": 0.474609375, + "learning_rate": 2.6397660818713448e-05, + "loss": 2.7791, + "step": 16743 + }, + { + "epoch": 0.8812631578947369, + "grad_norm": 0.5078125, + "learning_rate": 2.6385964912280704e-05, + "loss": 3.3824, + "step": 16744 + }, + { + "epoch": 0.8813157894736842, + "grad_norm": 0.48828125, + "learning_rate": 2.6374269005847957e-05, + "loss": 3.0527, + "step": 16745 + }, + { + "epoch": 0.8813684210526316, + "grad_norm": 0.484375, + "learning_rate": 2.6362573099415207e-05, + "loss": 3.4713, + "step": 16746 + }, + { + "epoch": 0.881421052631579, + "grad_norm": 0.486328125, + "learning_rate": 2.635087719298246e-05, + "loss": 2.6403, + "step": 16747 + }, + { + "epoch": 0.8814736842105263, + "grad_norm": 0.53125, + "learning_rate": 2.633918128654971e-05, + "loss": 2.8589, + "step": 16748 + }, + { + "epoch": 0.8815263157894737, + "grad_norm": 0.494140625, + "learning_rate": 2.6327485380116962e-05, + "loss": 3.0268, + "step": 16749 + }, + { + "epoch": 0.881578947368421, + "grad_norm": 0.47265625, + "learning_rate": 2.6315789473684212e-05, + "loss": 3.1348, + "step": 16750 + }, + { + "epoch": 0.8816315789473684, + "grad_norm": 0.50390625, + "learning_rate": 2.6304093567251465e-05, + "loss": 3.3018, + "step": 16751 + }, + { + "epoch": 0.8816842105263157, + "grad_norm": 0.5234375, + "learning_rate": 2.629239766081871e-05, + "loss": 3.0747, + "step": 16752 + }, + { + "epoch": 0.8817368421052632, + "grad_norm": 0.50390625, + "learning_rate": 2.6280701754385968e-05, + "loss": 3.063, + "step": 16753 + }, + { + "epoch": 0.8817894736842106, + "grad_norm": 0.515625, + "learning_rate": 2.6269005847953214e-05, + "loss": 3.2861, + "step": 16754 + }, + { + "epoch": 0.8818421052631579, + "grad_norm": 0.49609375, + "learning_rate": 2.625730994152047e-05, + "loss": 3.5501, + "step": 16755 + }, + { + "epoch": 0.8818947368421053, + "grad_norm": 0.498046875, + "learning_rate": 2.6245614035087723e-05, + "loss": 3.3601, + "step": 16756 + }, + { + "epoch": 0.8819473684210526, + "grad_norm": 0.494140625, + "learning_rate": 2.6233918128654973e-05, + "loss": 3.2787, + "step": 16757 + }, + { + "epoch": 0.882, + "grad_norm": 0.486328125, + "learning_rate": 2.6222222222222226e-05, + "loss": 2.936, + "step": 16758 + }, + { + "epoch": 0.8820526315789474, + "grad_norm": 0.5078125, + "learning_rate": 2.6210526315789475e-05, + "loss": 3.2731, + "step": 16759 + }, + { + "epoch": 0.8821052631578947, + "grad_norm": 0.498046875, + "learning_rate": 2.6198830409356728e-05, + "loss": 3.2648, + "step": 16760 + }, + { + "epoch": 0.8821578947368421, + "grad_norm": 0.5, + "learning_rate": 2.6187134502923978e-05, + "loss": 3.1823, + "step": 16761 + }, + { + "epoch": 0.8822105263157894, + "grad_norm": 0.5, + "learning_rate": 2.617543859649123e-05, + "loss": 3.3974, + "step": 16762 + }, + { + "epoch": 0.8822631578947369, + "grad_norm": 0.5078125, + "learning_rate": 2.616374269005848e-05, + "loss": 3.6553, + "step": 16763 + }, + { + "epoch": 0.8823157894736842, + "grad_norm": 0.474609375, + "learning_rate": 2.6152046783625733e-05, + "loss": 2.9347, + "step": 16764 + }, + { + "epoch": 0.8823684210526316, + "grad_norm": 0.486328125, + "learning_rate": 2.6140350877192983e-05, + "loss": 3.54, + "step": 16765 + }, + { + "epoch": 0.882421052631579, + "grad_norm": 0.490234375, + "learning_rate": 2.6128654970760236e-05, + "loss": 2.8222, + "step": 16766 + }, + { + "epoch": 0.8824736842105263, + "grad_norm": 0.5, + "learning_rate": 2.6116959064327486e-05, + "loss": 3.3029, + "step": 16767 + }, + { + "epoch": 0.8825263157894737, + "grad_norm": 0.482421875, + "learning_rate": 2.610526315789474e-05, + "loss": 3.1553, + "step": 16768 + }, + { + "epoch": 0.882578947368421, + "grad_norm": 0.4921875, + "learning_rate": 2.609356725146199e-05, + "loss": 3.2051, + "step": 16769 + }, + { + "epoch": 0.8826315789473684, + "grad_norm": 0.478515625, + "learning_rate": 2.608187134502924e-05, + "loss": 3.1287, + "step": 16770 + }, + { + "epoch": 0.8826842105263157, + "grad_norm": 0.48828125, + "learning_rate": 2.6070175438596494e-05, + "loss": 3.1754, + "step": 16771 + }, + { + "epoch": 0.8827368421052632, + "grad_norm": 0.51171875, + "learning_rate": 2.6058479532163744e-05, + "loss": 3.3752, + "step": 16772 + }, + { + "epoch": 0.8827894736842106, + "grad_norm": 0.462890625, + "learning_rate": 2.6046783625730997e-05, + "loss": 3.0493, + "step": 16773 + }, + { + "epoch": 0.8828421052631579, + "grad_norm": 0.5078125, + "learning_rate": 2.6035087719298246e-05, + "loss": 3.453, + "step": 16774 + }, + { + "epoch": 0.8828947368421053, + "grad_norm": 0.4921875, + "learning_rate": 2.60233918128655e-05, + "loss": 3.8474, + "step": 16775 + }, + { + "epoch": 0.8829473684210526, + "grad_norm": 0.53125, + "learning_rate": 2.601169590643275e-05, + "loss": 3.4357, + "step": 16776 + }, + { + "epoch": 0.883, + "grad_norm": 0.478515625, + "learning_rate": 2.6000000000000002e-05, + "loss": 3.2797, + "step": 16777 + }, + { + "epoch": 0.8830526315789474, + "grad_norm": 0.5078125, + "learning_rate": 2.598830409356725e-05, + "loss": 3.033, + "step": 16778 + }, + { + "epoch": 0.8831052631578947, + "grad_norm": 0.5, + "learning_rate": 2.5976608187134504e-05, + "loss": 3.1131, + "step": 16779 + }, + { + "epoch": 0.8831578947368421, + "grad_norm": 0.470703125, + "learning_rate": 2.5964912280701754e-05, + "loss": 2.8476, + "step": 16780 + }, + { + "epoch": 0.8832105263157894, + "grad_norm": 0.52734375, + "learning_rate": 2.5953216374269007e-05, + "loss": 3.4285, + "step": 16781 + }, + { + "epoch": 0.8832631578947369, + "grad_norm": 0.47265625, + "learning_rate": 2.594152046783626e-05, + "loss": 3.0001, + "step": 16782 + }, + { + "epoch": 0.8833157894736842, + "grad_norm": 0.482421875, + "learning_rate": 2.592982456140351e-05, + "loss": 3.3339, + "step": 16783 + }, + { + "epoch": 0.8833684210526316, + "grad_norm": 0.51953125, + "learning_rate": 2.5918128654970762e-05, + "loss": 3.0834, + "step": 16784 + }, + { + "epoch": 0.883421052631579, + "grad_norm": 0.6796875, + "learning_rate": 2.5906432748538012e-05, + "loss": 3.5901, + "step": 16785 + }, + { + "epoch": 0.8834736842105263, + "grad_norm": 0.478515625, + "learning_rate": 2.5894736842105265e-05, + "loss": 3.2156, + "step": 16786 + }, + { + "epoch": 0.8835263157894737, + "grad_norm": 0.484375, + "learning_rate": 2.5883040935672515e-05, + "loss": 3.3196, + "step": 16787 + }, + { + "epoch": 0.883578947368421, + "grad_norm": 0.51953125, + "learning_rate": 2.5871345029239768e-05, + "loss": 3.0712, + "step": 16788 + }, + { + "epoch": 0.8836315789473684, + "grad_norm": 0.486328125, + "learning_rate": 2.5859649122807017e-05, + "loss": 3.1148, + "step": 16789 + }, + { + "epoch": 0.8836842105263157, + "grad_norm": 0.466796875, + "learning_rate": 2.584795321637427e-05, + "loss": 2.7971, + "step": 16790 + }, + { + "epoch": 0.8837368421052632, + "grad_norm": 0.50390625, + "learning_rate": 2.583625730994152e-05, + "loss": 3.2888, + "step": 16791 + }, + { + "epoch": 0.8837894736842106, + "grad_norm": 0.49609375, + "learning_rate": 2.5824561403508773e-05, + "loss": 3.2863, + "step": 16792 + }, + { + "epoch": 0.8838421052631579, + "grad_norm": 0.48046875, + "learning_rate": 2.5812865497076022e-05, + "loss": 3.0251, + "step": 16793 + }, + { + "epoch": 0.8838947368421053, + "grad_norm": 0.54296875, + "learning_rate": 2.5801169590643275e-05, + "loss": 3.3446, + "step": 16794 + }, + { + "epoch": 0.8839473684210526, + "grad_norm": 0.5078125, + "learning_rate": 2.578947368421053e-05, + "loss": 3.2401, + "step": 16795 + }, + { + "epoch": 0.884, + "grad_norm": 0.52734375, + "learning_rate": 2.5777777777777778e-05, + "loss": 3.1109, + "step": 16796 + }, + { + "epoch": 0.8840526315789474, + "grad_norm": 0.51171875, + "learning_rate": 2.576608187134503e-05, + "loss": 3.5745, + "step": 16797 + }, + { + "epoch": 0.8841052631578947, + "grad_norm": 0.50390625, + "learning_rate": 2.575438596491228e-05, + "loss": 3.1348, + "step": 16798 + }, + { + "epoch": 0.8841578947368421, + "grad_norm": 0.48046875, + "learning_rate": 2.5742690058479533e-05, + "loss": 3.6106, + "step": 16799 + }, + { + "epoch": 0.8842105263157894, + "grad_norm": 0.490234375, + "learning_rate": 2.5730994152046783e-05, + "loss": 3.0706, + "step": 16800 + }, + { + "epoch": 0.8842631578947369, + "grad_norm": 0.48828125, + "learning_rate": 2.5719298245614036e-05, + "loss": 3.0709, + "step": 16801 + }, + { + "epoch": 0.8843157894736842, + "grad_norm": 0.49609375, + "learning_rate": 2.5707602339181286e-05, + "loss": 3.302, + "step": 16802 + }, + { + "epoch": 0.8843684210526316, + "grad_norm": 0.498046875, + "learning_rate": 2.569590643274854e-05, + "loss": 3.504, + "step": 16803 + }, + { + "epoch": 0.884421052631579, + "grad_norm": 0.474609375, + "learning_rate": 2.5684210526315788e-05, + "loss": 3.1389, + "step": 16804 + }, + { + "epoch": 0.8844736842105263, + "grad_norm": 0.474609375, + "learning_rate": 2.567251461988304e-05, + "loss": 3.0757, + "step": 16805 + }, + { + "epoch": 0.8845263157894737, + "grad_norm": 0.515625, + "learning_rate": 2.566081871345029e-05, + "loss": 3.0318, + "step": 16806 + }, + { + "epoch": 0.884578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.5649122807017544e-05, + "loss": 3.4944, + "step": 16807 + }, + { + "epoch": 0.8846315789473684, + "grad_norm": 0.5, + "learning_rate": 2.56374269005848e-05, + "loss": 3.517, + "step": 16808 + }, + { + "epoch": 0.8846842105263157, + "grad_norm": 0.462890625, + "learning_rate": 2.5625730994152046e-05, + "loss": 2.7326, + "step": 16809 + }, + { + "epoch": 0.8847368421052632, + "grad_norm": 0.50390625, + "learning_rate": 2.5614035087719303e-05, + "loss": 3.248, + "step": 16810 + }, + { + "epoch": 0.8847894736842106, + "grad_norm": 0.49609375, + "learning_rate": 2.560233918128655e-05, + "loss": 3.0324, + "step": 16811 + }, + { + "epoch": 0.8848421052631579, + "grad_norm": 0.54296875, + "learning_rate": 2.5590643274853805e-05, + "loss": 3.1256, + "step": 16812 + }, + { + "epoch": 0.8848947368421053, + "grad_norm": 0.51171875, + "learning_rate": 2.557894736842105e-05, + "loss": 3.641, + "step": 16813 + }, + { + "epoch": 0.8849473684210526, + "grad_norm": 0.4921875, + "learning_rate": 2.5567251461988308e-05, + "loss": 3.5055, + "step": 16814 + }, + { + "epoch": 0.885, + "grad_norm": 0.48046875, + "learning_rate": 2.5555555555555554e-05, + "loss": 3.2861, + "step": 16815 + }, + { + "epoch": 0.8850526315789473, + "grad_norm": 0.49609375, + "learning_rate": 2.554385964912281e-05, + "loss": 3.582, + "step": 16816 + }, + { + "epoch": 0.8851052631578947, + "grad_norm": 0.48828125, + "learning_rate": 2.5532163742690057e-05, + "loss": 3.375, + "step": 16817 + }, + { + "epoch": 0.8851578947368421, + "grad_norm": 0.482421875, + "learning_rate": 2.552046783625731e-05, + "loss": 3.0545, + "step": 16818 + }, + { + "epoch": 0.8852105263157894, + "grad_norm": 0.4921875, + "learning_rate": 2.5508771929824566e-05, + "loss": 3.5366, + "step": 16819 + }, + { + "epoch": 0.8852631578947369, + "grad_norm": 0.48828125, + "learning_rate": 2.5497076023391812e-05, + "loss": 3.5549, + "step": 16820 + }, + { + "epoch": 0.8853157894736842, + "grad_norm": 0.578125, + "learning_rate": 2.548538011695907e-05, + "loss": 3.2113, + "step": 16821 + }, + { + "epoch": 0.8853684210526316, + "grad_norm": 0.515625, + "learning_rate": 2.5473684210526315e-05, + "loss": 3.2405, + "step": 16822 + }, + { + "epoch": 0.885421052631579, + "grad_norm": 0.462890625, + "learning_rate": 2.546198830409357e-05, + "loss": 2.7126, + "step": 16823 + }, + { + "epoch": 0.8854736842105263, + "grad_norm": 0.515625, + "learning_rate": 2.5450292397660817e-05, + "loss": 3.1629, + "step": 16824 + }, + { + "epoch": 0.8855263157894737, + "grad_norm": 0.48828125, + "learning_rate": 2.5438596491228074e-05, + "loss": 3.2583, + "step": 16825 + }, + { + "epoch": 0.885578947368421, + "grad_norm": 0.515625, + "learning_rate": 2.542690058479532e-05, + "loss": 3.4626, + "step": 16826 + }, + { + "epoch": 0.8856315789473684, + "grad_norm": 0.4921875, + "learning_rate": 2.5415204678362576e-05, + "loss": 3.1921, + "step": 16827 + }, + { + "epoch": 0.8856842105263157, + "grad_norm": 0.515625, + "learning_rate": 2.5403508771929822e-05, + "loss": 3.2557, + "step": 16828 + }, + { + "epoch": 0.8857368421052632, + "grad_norm": 0.5078125, + "learning_rate": 2.539181286549708e-05, + "loss": 3.4181, + "step": 16829 + }, + { + "epoch": 0.8857894736842106, + "grad_norm": 0.5078125, + "learning_rate": 2.5380116959064325e-05, + "loss": 3.2874, + "step": 16830 + }, + { + "epoch": 0.8858421052631579, + "grad_norm": 0.5, + "learning_rate": 2.536842105263158e-05, + "loss": 3.345, + "step": 16831 + }, + { + "epoch": 0.8858947368421053, + "grad_norm": 0.54296875, + "learning_rate": 2.5356725146198834e-05, + "loss": 3.2672, + "step": 16832 + }, + { + "epoch": 0.8859473684210526, + "grad_norm": 0.48046875, + "learning_rate": 2.5345029239766084e-05, + "loss": 3.0724, + "step": 16833 + }, + { + "epoch": 0.886, + "grad_norm": 0.5234375, + "learning_rate": 2.5333333333333337e-05, + "loss": 3.0217, + "step": 16834 + }, + { + "epoch": 0.8860526315789473, + "grad_norm": 0.578125, + "learning_rate": 2.5321637426900587e-05, + "loss": 3.1403, + "step": 16835 + }, + { + "epoch": 0.8861052631578947, + "grad_norm": 0.515625, + "learning_rate": 2.530994152046784e-05, + "loss": 3.2369, + "step": 16836 + }, + { + "epoch": 0.8861578947368421, + "grad_norm": 0.49609375, + "learning_rate": 2.529824561403509e-05, + "loss": 2.97, + "step": 16837 + }, + { + "epoch": 0.8862105263157894, + "grad_norm": 0.50390625, + "learning_rate": 2.5286549707602342e-05, + "loss": 2.9705, + "step": 16838 + }, + { + "epoch": 0.8862631578947369, + "grad_norm": 0.515625, + "learning_rate": 2.5274853801169592e-05, + "loss": 3.3583, + "step": 16839 + }, + { + "epoch": 0.8863157894736842, + "grad_norm": 0.48828125, + "learning_rate": 2.5263157894736845e-05, + "loss": 3.3181, + "step": 16840 + }, + { + "epoch": 0.8863684210526316, + "grad_norm": 0.546875, + "learning_rate": 2.525146198830409e-05, + "loss": 3.5341, + "step": 16841 + }, + { + "epoch": 0.886421052631579, + "grad_norm": 0.51953125, + "learning_rate": 2.5239766081871347e-05, + "loss": 2.7647, + "step": 16842 + }, + { + "epoch": 0.8864736842105263, + "grad_norm": 0.515625, + "learning_rate": 2.5228070175438593e-05, + "loss": 3.2897, + "step": 16843 + }, + { + "epoch": 0.8865263157894737, + "grad_norm": 0.48046875, + "learning_rate": 2.521637426900585e-05, + "loss": 3.3029, + "step": 16844 + }, + { + "epoch": 0.886578947368421, + "grad_norm": 0.48828125, + "learning_rate": 2.5204678362573103e-05, + "loss": 3.208, + "step": 16845 + }, + { + "epoch": 0.8866315789473684, + "grad_norm": 0.5390625, + "learning_rate": 2.5192982456140352e-05, + "loss": 3.1148, + "step": 16846 + }, + { + "epoch": 0.8866842105263157, + "grad_norm": 0.51953125, + "learning_rate": 2.5181286549707605e-05, + "loss": 3.5342, + "step": 16847 + }, + { + "epoch": 0.8867368421052632, + "grad_norm": 0.498046875, + "learning_rate": 2.5169590643274855e-05, + "loss": 3.245, + "step": 16848 + }, + { + "epoch": 0.8867894736842106, + "grad_norm": 0.49609375, + "learning_rate": 2.5157894736842108e-05, + "loss": 3.1661, + "step": 16849 + }, + { + "epoch": 0.8868421052631579, + "grad_norm": 0.490234375, + "learning_rate": 2.5146198830409358e-05, + "loss": 3.6188, + "step": 16850 + }, + { + "epoch": 0.8868947368421053, + "grad_norm": 0.50390625, + "learning_rate": 2.513450292397661e-05, + "loss": 3.2179, + "step": 16851 + }, + { + "epoch": 0.8869473684210526, + "grad_norm": 0.490234375, + "learning_rate": 2.512280701754386e-05, + "loss": 3.2703, + "step": 16852 + }, + { + "epoch": 0.887, + "grad_norm": 0.46875, + "learning_rate": 2.5111111111111113e-05, + "loss": 3.2616, + "step": 16853 + }, + { + "epoch": 0.8870526315789473, + "grad_norm": 0.49609375, + "learning_rate": 2.5099415204678363e-05, + "loss": 3.4044, + "step": 16854 + }, + { + "epoch": 0.8871052631578947, + "grad_norm": 0.484375, + "learning_rate": 2.5087719298245616e-05, + "loss": 3.1707, + "step": 16855 + }, + { + "epoch": 0.8871578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.5076023391812865e-05, + "loss": 2.9584, + "step": 16856 + }, + { + "epoch": 0.8872105263157894, + "grad_norm": 0.5, + "learning_rate": 2.5064327485380118e-05, + "loss": 2.9935, + "step": 16857 + }, + { + "epoch": 0.8872631578947369, + "grad_norm": 0.51171875, + "learning_rate": 2.505263157894737e-05, + "loss": 3.2264, + "step": 16858 + }, + { + "epoch": 0.8873157894736842, + "grad_norm": 0.494140625, + "learning_rate": 2.504093567251462e-05, + "loss": 3.1989, + "step": 16859 + }, + { + "epoch": 0.8873684210526316, + "grad_norm": 0.49609375, + "learning_rate": 2.5029239766081874e-05, + "loss": 3.0735, + "step": 16860 + }, + { + "epoch": 0.887421052631579, + "grad_norm": 0.48828125, + "learning_rate": 2.5017543859649123e-05, + "loss": 2.8871, + "step": 16861 + }, + { + "epoch": 0.8874736842105263, + "grad_norm": 0.5078125, + "learning_rate": 2.5005847953216376e-05, + "loss": 3.0639, + "step": 16862 + }, + { + "epoch": 0.8875263157894737, + "grad_norm": 0.51171875, + "learning_rate": 2.499415204678363e-05, + "loss": 3.0848, + "step": 16863 + }, + { + "epoch": 0.887578947368421, + "grad_norm": 0.486328125, + "learning_rate": 2.498245614035088e-05, + "loss": 3.2939, + "step": 16864 + }, + { + "epoch": 0.8876315789473684, + "grad_norm": 0.54296875, + "learning_rate": 2.4970760233918132e-05, + "loss": 3.39, + "step": 16865 + }, + { + "epoch": 0.8876842105263157, + "grad_norm": 0.482421875, + "learning_rate": 2.495906432748538e-05, + "loss": 2.966, + "step": 16866 + }, + { + "epoch": 0.8877368421052632, + "grad_norm": 0.5, + "learning_rate": 2.4947368421052635e-05, + "loss": 3.497, + "step": 16867 + }, + { + "epoch": 0.8877894736842106, + "grad_norm": 0.462890625, + "learning_rate": 2.4935672514619884e-05, + "loss": 2.7376, + "step": 16868 + }, + { + "epoch": 0.8878421052631579, + "grad_norm": 0.4921875, + "learning_rate": 2.4923976608187137e-05, + "loss": 2.9141, + "step": 16869 + }, + { + "epoch": 0.8878947368421053, + "grad_norm": 0.5, + "learning_rate": 2.4912280701754387e-05, + "loss": 2.6461, + "step": 16870 + }, + { + "epoch": 0.8879473684210526, + "grad_norm": 0.51953125, + "learning_rate": 2.490058479532164e-05, + "loss": 3.0573, + "step": 16871 + }, + { + "epoch": 0.888, + "grad_norm": 0.5078125, + "learning_rate": 2.488888888888889e-05, + "loss": 2.8785, + "step": 16872 + }, + { + "epoch": 0.8880526315789473, + "grad_norm": 0.48828125, + "learning_rate": 2.4877192982456142e-05, + "loss": 3.1294, + "step": 16873 + }, + { + "epoch": 0.8881052631578947, + "grad_norm": 0.5, + "learning_rate": 2.4865497076023392e-05, + "loss": 3.5817, + "step": 16874 + }, + { + "epoch": 0.8881578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.485380116959064e-05, + "loss": 3.8896, + "step": 16875 + }, + { + "epoch": 0.8882105263157895, + "grad_norm": 0.49609375, + "learning_rate": 2.4842105263157898e-05, + "loss": 3.192, + "step": 16876 + }, + { + "epoch": 0.8882631578947369, + "grad_norm": 0.498046875, + "learning_rate": 2.4830409356725147e-05, + "loss": 3.3532, + "step": 16877 + }, + { + "epoch": 0.8883157894736842, + "grad_norm": 0.498046875, + "learning_rate": 2.48187134502924e-05, + "loss": 3.4118, + "step": 16878 + }, + { + "epoch": 0.8883684210526316, + "grad_norm": 0.50390625, + "learning_rate": 2.480701754385965e-05, + "loss": 3.275, + "step": 16879 + }, + { + "epoch": 0.888421052631579, + "grad_norm": 0.5, + "learning_rate": 2.4795321637426903e-05, + "loss": 3.2575, + "step": 16880 + }, + { + "epoch": 0.8884736842105263, + "grad_norm": 0.478515625, + "learning_rate": 2.4783625730994153e-05, + "loss": 3.3379, + "step": 16881 + }, + { + "epoch": 0.8885263157894737, + "grad_norm": 0.50390625, + "learning_rate": 2.4771929824561406e-05, + "loss": 3.3006, + "step": 16882 + }, + { + "epoch": 0.888578947368421, + "grad_norm": 0.498046875, + "learning_rate": 2.4760233918128655e-05, + "loss": 3.1134, + "step": 16883 + }, + { + "epoch": 0.8886315789473684, + "grad_norm": 0.46484375, + "learning_rate": 2.4748538011695908e-05, + "loss": 3.1767, + "step": 16884 + }, + { + "epoch": 0.8886842105263157, + "grad_norm": 0.5, + "learning_rate": 2.4736842105263158e-05, + "loss": 3.3568, + "step": 16885 + }, + { + "epoch": 0.8887368421052632, + "grad_norm": 0.48828125, + "learning_rate": 2.472514619883041e-05, + "loss": 3.7097, + "step": 16886 + }, + { + "epoch": 0.8887894736842106, + "grad_norm": 0.494140625, + "learning_rate": 2.471345029239766e-05, + "loss": 3.5537, + "step": 16887 + }, + { + "epoch": 0.8888421052631579, + "grad_norm": 0.5, + "learning_rate": 2.4701754385964913e-05, + "loss": 2.869, + "step": 16888 + }, + { + "epoch": 0.8888947368421053, + "grad_norm": 0.5234375, + "learning_rate": 2.4690058479532166e-05, + "loss": 3.2702, + "step": 16889 + }, + { + "epoch": 0.8889473684210526, + "grad_norm": 0.5234375, + "learning_rate": 2.4678362573099416e-05, + "loss": 3.3162, + "step": 16890 + }, + { + "epoch": 0.889, + "grad_norm": 0.51953125, + "learning_rate": 2.466666666666667e-05, + "loss": 3.2549, + "step": 16891 + }, + { + "epoch": 0.8890526315789473, + "grad_norm": 0.52734375, + "learning_rate": 2.465497076023392e-05, + "loss": 3.0559, + "step": 16892 + }, + { + "epoch": 0.8891052631578947, + "grad_norm": 0.490234375, + "learning_rate": 2.464327485380117e-05, + "loss": 3.5083, + "step": 16893 + }, + { + "epoch": 0.8891578947368421, + "grad_norm": 0.494140625, + "learning_rate": 2.463157894736842e-05, + "loss": 3.5119, + "step": 16894 + }, + { + "epoch": 0.8892105263157895, + "grad_norm": 0.60546875, + "learning_rate": 2.4619883040935674e-05, + "loss": 2.9232, + "step": 16895 + }, + { + "epoch": 0.8892631578947369, + "grad_norm": 0.52734375, + "learning_rate": 2.4608187134502924e-05, + "loss": 3.6337, + "step": 16896 + }, + { + "epoch": 0.8893157894736842, + "grad_norm": 0.498046875, + "learning_rate": 2.4596491228070177e-05, + "loss": 3.2901, + "step": 16897 + }, + { + "epoch": 0.8893684210526316, + "grad_norm": 0.474609375, + "learning_rate": 2.4584795321637426e-05, + "loss": 3.0994, + "step": 16898 + }, + { + "epoch": 0.889421052631579, + "grad_norm": 0.482421875, + "learning_rate": 2.457309941520468e-05, + "loss": 2.9117, + "step": 16899 + }, + { + "epoch": 0.8894736842105263, + "grad_norm": 0.51953125, + "learning_rate": 2.456140350877193e-05, + "loss": 3.4578, + "step": 16900 + }, + { + "epoch": 0.8895263157894737, + "grad_norm": 0.49609375, + "learning_rate": 2.4549707602339185e-05, + "loss": 3.1434, + "step": 16901 + }, + { + "epoch": 0.889578947368421, + "grad_norm": 0.5078125, + "learning_rate": 2.4538011695906435e-05, + "loss": 3.0933, + "step": 16902 + }, + { + "epoch": 0.8896315789473684, + "grad_norm": 0.50390625, + "learning_rate": 2.4526315789473688e-05, + "loss": 3.47, + "step": 16903 + }, + { + "epoch": 0.8896842105263157, + "grad_norm": 0.5234375, + "learning_rate": 2.4514619883040937e-05, + "loss": 3.1461, + "step": 16904 + }, + { + "epoch": 0.8897368421052632, + "grad_norm": 0.498046875, + "learning_rate": 2.450292397660819e-05, + "loss": 3.493, + "step": 16905 + }, + { + "epoch": 0.8897894736842106, + "grad_norm": 0.53125, + "learning_rate": 2.449122807017544e-05, + "loss": 3.4218, + "step": 16906 + }, + { + "epoch": 0.8898421052631579, + "grad_norm": 0.51171875, + "learning_rate": 2.4479532163742693e-05, + "loss": 3.5638, + "step": 16907 + }, + { + "epoch": 0.8898947368421053, + "grad_norm": 0.494140625, + "learning_rate": 2.4467836257309942e-05, + "loss": 3.6919, + "step": 16908 + }, + { + "epoch": 0.8899473684210526, + "grad_norm": 0.55859375, + "learning_rate": 2.4456140350877192e-05, + "loss": 3.1808, + "step": 16909 + }, + { + "epoch": 0.89, + "grad_norm": 0.48046875, + "learning_rate": 2.4444444444444445e-05, + "loss": 3.1452, + "step": 16910 + }, + { + "epoch": 0.8900526315789473, + "grad_norm": 0.515625, + "learning_rate": 2.4432748538011695e-05, + "loss": 3.3169, + "step": 16911 + }, + { + "epoch": 0.8901052631578947, + "grad_norm": 0.5, + "learning_rate": 2.4421052631578948e-05, + "loss": 2.8776, + "step": 16912 + }, + { + "epoch": 0.8901578947368421, + "grad_norm": 0.5, + "learning_rate": 2.44093567251462e-05, + "loss": 3.0116, + "step": 16913 + }, + { + "epoch": 0.8902105263157895, + "grad_norm": 0.482421875, + "learning_rate": 2.4397660818713453e-05, + "loss": 3.1729, + "step": 16914 + }, + { + "epoch": 0.8902631578947369, + "grad_norm": 0.515625, + "learning_rate": 2.4385964912280703e-05, + "loss": 3.4584, + "step": 16915 + }, + { + "epoch": 0.8903157894736842, + "grad_norm": 0.5, + "learning_rate": 2.4374269005847956e-05, + "loss": 3.5243, + "step": 16916 + }, + { + "epoch": 0.8903684210526316, + "grad_norm": 0.4921875, + "learning_rate": 2.4362573099415206e-05, + "loss": 3.178, + "step": 16917 + }, + { + "epoch": 0.890421052631579, + "grad_norm": 0.494140625, + "learning_rate": 2.435087719298246e-05, + "loss": 3.516, + "step": 16918 + }, + { + "epoch": 0.8904736842105263, + "grad_norm": 0.50390625, + "learning_rate": 2.4339181286549708e-05, + "loss": 3.1141, + "step": 16919 + }, + { + "epoch": 0.8905263157894737, + "grad_norm": 0.5, + "learning_rate": 2.432748538011696e-05, + "loss": 3.4189, + "step": 16920 + }, + { + "epoch": 0.890578947368421, + "grad_norm": 0.5, + "learning_rate": 2.431578947368421e-05, + "loss": 3.0502, + "step": 16921 + }, + { + "epoch": 0.8906315789473684, + "grad_norm": 0.51171875, + "learning_rate": 2.4304093567251464e-05, + "loss": 3.2007, + "step": 16922 + }, + { + "epoch": 0.8906842105263157, + "grad_norm": 0.51953125, + "learning_rate": 2.4292397660818713e-05, + "loss": 3.6688, + "step": 16923 + }, + { + "epoch": 0.8907368421052632, + "grad_norm": 0.51171875, + "learning_rate": 2.4280701754385966e-05, + "loss": 3.5153, + "step": 16924 + }, + { + "epoch": 0.8907894736842106, + "grad_norm": 0.482421875, + "learning_rate": 2.4269005847953216e-05, + "loss": 3.0214, + "step": 16925 + }, + { + "epoch": 0.8908421052631579, + "grad_norm": 0.498046875, + "learning_rate": 2.425730994152047e-05, + "loss": 3.7421, + "step": 16926 + }, + { + "epoch": 0.8908947368421053, + "grad_norm": 0.490234375, + "learning_rate": 2.4245614035087722e-05, + "loss": 3.2026, + "step": 16927 + }, + { + "epoch": 0.8909473684210526, + "grad_norm": 0.51171875, + "learning_rate": 2.423391812865497e-05, + "loss": 3.4668, + "step": 16928 + }, + { + "epoch": 0.891, + "grad_norm": 0.48828125, + "learning_rate": 2.4222222222222224e-05, + "loss": 3.527, + "step": 16929 + }, + { + "epoch": 0.8910526315789473, + "grad_norm": 0.50390625, + "learning_rate": 2.4210526315789474e-05, + "loss": 3.2554, + "step": 16930 + }, + { + "epoch": 0.8911052631578947, + "grad_norm": 0.466796875, + "learning_rate": 2.4198830409356727e-05, + "loss": 2.9315, + "step": 16931 + }, + { + "epoch": 0.8911578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.4187134502923977e-05, + "loss": 2.7632, + "step": 16932 + }, + { + "epoch": 0.8912105263157895, + "grad_norm": 0.5234375, + "learning_rate": 2.417543859649123e-05, + "loss": 3.4848, + "step": 16933 + }, + { + "epoch": 0.8912631578947369, + "grad_norm": 0.482421875, + "learning_rate": 2.416374269005848e-05, + "loss": 3.2327, + "step": 16934 + }, + { + "epoch": 0.8913157894736842, + "grad_norm": 0.5390625, + "learning_rate": 2.4152046783625732e-05, + "loss": 3.4134, + "step": 16935 + }, + { + "epoch": 0.8913684210526316, + "grad_norm": 0.478515625, + "learning_rate": 2.4140350877192982e-05, + "loss": 2.947, + "step": 16936 + }, + { + "epoch": 0.891421052631579, + "grad_norm": 0.51171875, + "learning_rate": 2.4128654970760235e-05, + "loss": 3.4694, + "step": 16937 + }, + { + "epoch": 0.8914736842105263, + "grad_norm": 0.4921875, + "learning_rate": 2.4116959064327488e-05, + "loss": 3.3165, + "step": 16938 + }, + { + "epoch": 0.8915263157894737, + "grad_norm": 0.482421875, + "learning_rate": 2.410526315789474e-05, + "loss": 2.9508, + "step": 16939 + }, + { + "epoch": 0.891578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.409356725146199e-05, + "loss": 3.2304, + "step": 16940 + }, + { + "epoch": 0.8916315789473684, + "grad_norm": 0.5, + "learning_rate": 2.408187134502924e-05, + "loss": 3.4052, + "step": 16941 + }, + { + "epoch": 0.8916842105263157, + "grad_norm": 0.51171875, + "learning_rate": 2.4070175438596493e-05, + "loss": 3.4543, + "step": 16942 + }, + { + "epoch": 0.8917368421052632, + "grad_norm": 0.48828125, + "learning_rate": 2.4058479532163742e-05, + "loss": 3.023, + "step": 16943 + }, + { + "epoch": 0.8917894736842106, + "grad_norm": 0.451171875, + "learning_rate": 2.4046783625730995e-05, + "loss": 3.1484, + "step": 16944 + }, + { + "epoch": 0.8918421052631579, + "grad_norm": 0.5078125, + "learning_rate": 2.4035087719298245e-05, + "loss": 3.4138, + "step": 16945 + }, + { + "epoch": 0.8918947368421053, + "grad_norm": 0.5625, + "learning_rate": 2.4023391812865498e-05, + "loss": 3.2913, + "step": 16946 + }, + { + "epoch": 0.8919473684210526, + "grad_norm": 0.50390625, + "learning_rate": 2.4011695906432748e-05, + "loss": 3.5169, + "step": 16947 + }, + { + "epoch": 0.892, + "grad_norm": 0.45703125, + "learning_rate": 2.4e-05, + "loss": 2.8548, + "step": 16948 + }, + { + "epoch": 0.8920526315789473, + "grad_norm": 0.50390625, + "learning_rate": 2.398830409356725e-05, + "loss": 3.2983, + "step": 16949 + }, + { + "epoch": 0.8921052631578947, + "grad_norm": 0.498046875, + "learning_rate": 2.3976608187134503e-05, + "loss": 3.1579, + "step": 16950 + }, + { + "epoch": 0.8921578947368421, + "grad_norm": 0.53515625, + "learning_rate": 2.3964912280701756e-05, + "loss": 3.3399, + "step": 16951 + }, + { + "epoch": 0.8922105263157895, + "grad_norm": 0.51171875, + "learning_rate": 2.395321637426901e-05, + "loss": 3.526, + "step": 16952 + }, + { + "epoch": 0.8922631578947369, + "grad_norm": 0.49609375, + "learning_rate": 2.394152046783626e-05, + "loss": 2.9267, + "step": 16953 + }, + { + "epoch": 0.8923157894736842, + "grad_norm": 0.5234375, + "learning_rate": 2.3929824561403512e-05, + "loss": 3.2507, + "step": 16954 + }, + { + "epoch": 0.8923684210526316, + "grad_norm": 0.49609375, + "learning_rate": 2.391812865497076e-05, + "loss": 3.703, + "step": 16955 + }, + { + "epoch": 0.8924210526315789, + "grad_norm": 0.5, + "learning_rate": 2.3906432748538014e-05, + "loss": 3.4288, + "step": 16956 + }, + { + "epoch": 0.8924736842105263, + "grad_norm": 0.474609375, + "learning_rate": 2.3894736842105264e-05, + "loss": 3.1303, + "step": 16957 + }, + { + "epoch": 0.8925263157894737, + "grad_norm": 0.5234375, + "learning_rate": 2.3883040935672517e-05, + "loss": 2.9434, + "step": 16958 + }, + { + "epoch": 0.892578947368421, + "grad_norm": 0.48828125, + "learning_rate": 2.3871345029239766e-05, + "loss": 3.6075, + "step": 16959 + }, + { + "epoch": 0.8926315789473684, + "grad_norm": 0.5, + "learning_rate": 2.385964912280702e-05, + "loss": 3.1306, + "step": 16960 + }, + { + "epoch": 0.8926842105263157, + "grad_norm": 0.515625, + "learning_rate": 2.384795321637427e-05, + "loss": 3.2209, + "step": 16961 + }, + { + "epoch": 0.8927368421052632, + "grad_norm": 0.48046875, + "learning_rate": 2.3836257309941522e-05, + "loss": 3.1304, + "step": 16962 + }, + { + "epoch": 0.8927894736842106, + "grad_norm": 0.5546875, + "learning_rate": 2.3824561403508775e-05, + "loss": 3.2395, + "step": 16963 + }, + { + "epoch": 0.8928421052631579, + "grad_norm": 0.490234375, + "learning_rate": 2.3812865497076025e-05, + "loss": 3.5629, + "step": 16964 + }, + { + "epoch": 0.8928947368421053, + "grad_norm": 0.50390625, + "learning_rate": 2.3801169590643278e-05, + "loss": 3.7149, + "step": 16965 + }, + { + "epoch": 0.8929473684210526, + "grad_norm": 0.5234375, + "learning_rate": 2.3789473684210527e-05, + "loss": 3.1968, + "step": 16966 + }, + { + "epoch": 0.893, + "grad_norm": 0.49609375, + "learning_rate": 2.377777777777778e-05, + "loss": 2.9874, + "step": 16967 + }, + { + "epoch": 0.8930526315789473, + "grad_norm": 0.48046875, + "learning_rate": 2.376608187134503e-05, + "loss": 3.0727, + "step": 16968 + }, + { + "epoch": 0.8931052631578947, + "grad_norm": 0.51171875, + "learning_rate": 2.3754385964912283e-05, + "loss": 3.1747, + "step": 16969 + }, + { + "epoch": 0.8931578947368422, + "grad_norm": 0.53125, + "learning_rate": 2.3742690058479532e-05, + "loss": 3.2231, + "step": 16970 + }, + { + "epoch": 0.8932105263157895, + "grad_norm": 0.5, + "learning_rate": 2.3730994152046785e-05, + "loss": 3.5845, + "step": 16971 + }, + { + "epoch": 0.8932631578947369, + "grad_norm": 0.5, + "learning_rate": 2.3719298245614035e-05, + "loss": 2.8976, + "step": 16972 + }, + { + "epoch": 0.8933157894736842, + "grad_norm": 0.52734375, + "learning_rate": 2.3707602339181288e-05, + "loss": 3.4654, + "step": 16973 + }, + { + "epoch": 0.8933684210526316, + "grad_norm": 0.5078125, + "learning_rate": 2.3695906432748537e-05, + "loss": 2.8758, + "step": 16974 + }, + { + "epoch": 0.8934210526315789, + "grad_norm": 0.5, + "learning_rate": 2.368421052631579e-05, + "loss": 3.2165, + "step": 16975 + }, + { + "epoch": 0.8934736842105263, + "grad_norm": 0.5234375, + "learning_rate": 2.3672514619883043e-05, + "loss": 3.5257, + "step": 16976 + }, + { + "epoch": 0.8935263157894737, + "grad_norm": 0.51953125, + "learning_rate": 2.3660818713450293e-05, + "loss": 2.6813, + "step": 16977 + }, + { + "epoch": 0.893578947368421, + "grad_norm": 0.482421875, + "learning_rate": 2.3649122807017546e-05, + "loss": 3.6372, + "step": 16978 + }, + { + "epoch": 0.8936315789473684, + "grad_norm": 0.5078125, + "learning_rate": 2.3637426900584796e-05, + "loss": 3.3585, + "step": 16979 + }, + { + "epoch": 0.8936842105263157, + "grad_norm": 0.53125, + "learning_rate": 2.362573099415205e-05, + "loss": 3.2047, + "step": 16980 + }, + { + "epoch": 0.8937368421052632, + "grad_norm": 0.52734375, + "learning_rate": 2.3614035087719298e-05, + "loss": 3.0416, + "step": 16981 + }, + { + "epoch": 0.8937894736842106, + "grad_norm": 0.54296875, + "learning_rate": 2.360233918128655e-05, + "loss": 3.098, + "step": 16982 + }, + { + "epoch": 0.8938421052631579, + "grad_norm": 0.486328125, + "learning_rate": 2.35906432748538e-05, + "loss": 3.2223, + "step": 16983 + }, + { + "epoch": 0.8938947368421053, + "grad_norm": 0.50390625, + "learning_rate": 2.3578947368421054e-05, + "loss": 3.1558, + "step": 16984 + }, + { + "epoch": 0.8939473684210526, + "grad_norm": 0.490234375, + "learning_rate": 2.3567251461988303e-05, + "loss": 3.1617, + "step": 16985 + }, + { + "epoch": 0.894, + "grad_norm": 0.51171875, + "learning_rate": 2.3555555555555556e-05, + "loss": 3.7852, + "step": 16986 + }, + { + "epoch": 0.8940526315789473, + "grad_norm": 0.51953125, + "learning_rate": 2.3543859649122806e-05, + "loss": 3.1809, + "step": 16987 + }, + { + "epoch": 0.8941052631578947, + "grad_norm": 0.498046875, + "learning_rate": 2.3532163742690062e-05, + "loss": 3.2107, + "step": 16988 + }, + { + "epoch": 0.8941578947368422, + "grad_norm": 0.50390625, + "learning_rate": 2.3520467836257312e-05, + "loss": 3.2674, + "step": 16989 + }, + { + "epoch": 0.8942105263157895, + "grad_norm": 0.54296875, + "learning_rate": 2.3508771929824565e-05, + "loss": 3.6237, + "step": 16990 + }, + { + "epoch": 0.8942631578947369, + "grad_norm": 0.5078125, + "learning_rate": 2.3497076023391814e-05, + "loss": 3.3372, + "step": 16991 + }, + { + "epoch": 0.8943157894736842, + "grad_norm": 0.5078125, + "learning_rate": 2.3485380116959067e-05, + "loss": 3.2417, + "step": 16992 + }, + { + "epoch": 0.8943684210526316, + "grad_norm": 0.486328125, + "learning_rate": 2.3473684210526317e-05, + "loss": 3.0992, + "step": 16993 + }, + { + "epoch": 0.8944210526315789, + "grad_norm": 0.53125, + "learning_rate": 2.346198830409357e-05, + "loss": 3.2996, + "step": 16994 + }, + { + "epoch": 0.8944736842105263, + "grad_norm": 0.5, + "learning_rate": 2.345029239766082e-05, + "loss": 3.3976, + "step": 16995 + }, + { + "epoch": 0.8945263157894737, + "grad_norm": 0.4921875, + "learning_rate": 2.3438596491228073e-05, + "loss": 3.2101, + "step": 16996 + }, + { + "epoch": 0.894578947368421, + "grad_norm": 0.478515625, + "learning_rate": 2.3426900584795322e-05, + "loss": 3.4452, + "step": 16997 + }, + { + "epoch": 0.8946315789473684, + "grad_norm": 0.5, + "learning_rate": 2.3415204678362572e-05, + "loss": 3.4387, + "step": 16998 + }, + { + "epoch": 0.8946842105263157, + "grad_norm": 0.494140625, + "learning_rate": 2.3403508771929825e-05, + "loss": 2.8641, + "step": 16999 + }, + { + "epoch": 0.8947368421052632, + "grad_norm": 0.5078125, + "learning_rate": 2.3391812865497074e-05, + "loss": 3.1956, + "step": 17000 + }, + { + "epoch": 0.8947894736842106, + "grad_norm": 0.494140625, + "learning_rate": 2.338011695906433e-05, + "loss": 3.2727, + "step": 17001 + }, + { + "epoch": 0.8948421052631579, + "grad_norm": 0.482421875, + "learning_rate": 2.336842105263158e-05, + "loss": 3.1536, + "step": 17002 + }, + { + "epoch": 0.8948947368421053, + "grad_norm": 0.48046875, + "learning_rate": 2.3356725146198833e-05, + "loss": 3.3168, + "step": 17003 + }, + { + "epoch": 0.8949473684210526, + "grad_norm": 0.486328125, + "learning_rate": 2.3345029239766083e-05, + "loss": 2.8208, + "step": 17004 + }, + { + "epoch": 0.895, + "grad_norm": 0.51171875, + "learning_rate": 2.3333333333333336e-05, + "loss": 3.3213, + "step": 17005 + }, + { + "epoch": 0.8950526315789473, + "grad_norm": 0.50390625, + "learning_rate": 2.3321637426900585e-05, + "loss": 3.3506, + "step": 17006 + }, + { + "epoch": 0.8951052631578947, + "grad_norm": 0.5078125, + "learning_rate": 2.330994152046784e-05, + "loss": 3.5748, + "step": 17007 + }, + { + "epoch": 0.8951578947368422, + "grad_norm": 0.515625, + "learning_rate": 2.3298245614035088e-05, + "loss": 2.9808, + "step": 17008 + }, + { + "epoch": 0.8952105263157895, + "grad_norm": 0.470703125, + "learning_rate": 2.328654970760234e-05, + "loss": 3.1442, + "step": 17009 + }, + { + "epoch": 0.8952631578947369, + "grad_norm": 0.494140625, + "learning_rate": 2.327485380116959e-05, + "loss": 3.1651, + "step": 17010 + }, + { + "epoch": 0.8953157894736842, + "grad_norm": 0.53125, + "learning_rate": 2.3263157894736844e-05, + "loss": 2.6576, + "step": 17011 + }, + { + "epoch": 0.8953684210526316, + "grad_norm": 0.494140625, + "learning_rate": 2.3251461988304093e-05, + "loss": 3.0631, + "step": 17012 + }, + { + "epoch": 0.8954210526315789, + "grad_norm": 0.5, + "learning_rate": 2.3239766081871346e-05, + "loss": 3.3058, + "step": 17013 + }, + { + "epoch": 0.8954736842105263, + "grad_norm": 0.58203125, + "learning_rate": 2.32280701754386e-05, + "loss": 2.7984, + "step": 17014 + }, + { + "epoch": 0.8955263157894737, + "grad_norm": 0.494140625, + "learning_rate": 2.321637426900585e-05, + "loss": 3.2062, + "step": 17015 + }, + { + "epoch": 0.895578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.32046783625731e-05, + "loss": 3.0548, + "step": 17016 + }, + { + "epoch": 0.8956315789473684, + "grad_norm": 0.51171875, + "learning_rate": 2.319298245614035e-05, + "loss": 2.9493, + "step": 17017 + }, + { + "epoch": 0.8956842105263157, + "grad_norm": 0.51171875, + "learning_rate": 2.3181286549707604e-05, + "loss": 3.0372, + "step": 17018 + }, + { + "epoch": 0.8957368421052632, + "grad_norm": 0.478515625, + "learning_rate": 2.3169590643274854e-05, + "loss": 3.5136, + "step": 17019 + }, + { + "epoch": 0.8957894736842106, + "grad_norm": 0.54296875, + "learning_rate": 2.3157894736842107e-05, + "loss": 3.4718, + "step": 17020 + }, + { + "epoch": 0.8958421052631579, + "grad_norm": 0.48046875, + "learning_rate": 2.3146198830409356e-05, + "loss": 2.7477, + "step": 17021 + }, + { + "epoch": 0.8958947368421053, + "grad_norm": 0.5078125, + "learning_rate": 2.313450292397661e-05, + "loss": 3.3095, + "step": 17022 + }, + { + "epoch": 0.8959473684210526, + "grad_norm": 0.48046875, + "learning_rate": 2.312280701754386e-05, + "loss": 2.8037, + "step": 17023 + }, + { + "epoch": 0.896, + "grad_norm": 0.5078125, + "learning_rate": 2.3111111111111112e-05, + "loss": 3.1855, + "step": 17024 + }, + { + "epoch": 0.8960526315789473, + "grad_norm": 0.515625, + "learning_rate": 2.309941520467836e-05, + "loss": 3.3448, + "step": 17025 + }, + { + "epoch": 0.8961052631578947, + "grad_norm": 0.50390625, + "learning_rate": 2.3087719298245618e-05, + "loss": 3.5269, + "step": 17026 + }, + { + "epoch": 0.8961578947368422, + "grad_norm": 0.61328125, + "learning_rate": 2.3076023391812868e-05, + "loss": 3.0128, + "step": 17027 + }, + { + "epoch": 0.8962105263157895, + "grad_norm": 0.51171875, + "learning_rate": 2.306432748538012e-05, + "loss": 3.6642, + "step": 17028 + }, + { + "epoch": 0.8962631578947369, + "grad_norm": 0.498046875, + "learning_rate": 2.305263157894737e-05, + "loss": 2.7373, + "step": 17029 + }, + { + "epoch": 0.8963157894736842, + "grad_norm": 0.515625, + "learning_rate": 2.3040935672514623e-05, + "loss": 3.1828, + "step": 17030 + }, + { + "epoch": 0.8963684210526316, + "grad_norm": 0.498046875, + "learning_rate": 2.3029239766081873e-05, + "loss": 3.1906, + "step": 17031 + }, + { + "epoch": 0.8964210526315789, + "grad_norm": 0.50390625, + "learning_rate": 2.3017543859649122e-05, + "loss": 3.3559, + "step": 17032 + }, + { + "epoch": 0.8964736842105263, + "grad_norm": 0.5234375, + "learning_rate": 2.3005847953216375e-05, + "loss": 3.1979, + "step": 17033 + }, + { + "epoch": 0.8965263157894737, + "grad_norm": 0.5078125, + "learning_rate": 2.2994152046783625e-05, + "loss": 3.6128, + "step": 17034 + }, + { + "epoch": 0.896578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.2982456140350878e-05, + "loss": 3.3614, + "step": 17035 + }, + { + "epoch": 0.8966315789473684, + "grad_norm": 0.5078125, + "learning_rate": 2.2970760233918127e-05, + "loss": 3.2025, + "step": 17036 + }, + { + "epoch": 0.8966842105263158, + "grad_norm": 0.47265625, + "learning_rate": 2.295906432748538e-05, + "loss": 2.9444, + "step": 17037 + }, + { + "epoch": 0.8967368421052632, + "grad_norm": 0.515625, + "learning_rate": 2.294736842105263e-05, + "loss": 3.6156, + "step": 17038 + }, + { + "epoch": 0.8967894736842106, + "grad_norm": 0.5078125, + "learning_rate": 2.2935672514619886e-05, + "loss": 2.7666, + "step": 17039 + }, + { + "epoch": 0.8968421052631579, + "grad_norm": 0.494140625, + "learning_rate": 2.2923976608187136e-05, + "loss": 3.151, + "step": 17040 + }, + { + "epoch": 0.8968947368421053, + "grad_norm": 0.498046875, + "learning_rate": 2.291228070175439e-05, + "loss": 3.6008, + "step": 17041 + }, + { + "epoch": 0.8969473684210526, + "grad_norm": 0.5234375, + "learning_rate": 2.290058479532164e-05, + "loss": 2.7188, + "step": 17042 + }, + { + "epoch": 0.897, + "grad_norm": 0.56640625, + "learning_rate": 2.288888888888889e-05, + "loss": 3.208, + "step": 17043 + }, + { + "epoch": 0.8970526315789473, + "grad_norm": 0.51953125, + "learning_rate": 2.287719298245614e-05, + "loss": 3.3815, + "step": 17044 + }, + { + "epoch": 0.8971052631578947, + "grad_norm": 0.6171875, + "learning_rate": 2.2865497076023394e-05, + "loss": 3.1336, + "step": 17045 + }, + { + "epoch": 0.8971578947368422, + "grad_norm": 0.5078125, + "learning_rate": 2.2853801169590644e-05, + "loss": 3.6254, + "step": 17046 + }, + { + "epoch": 0.8972105263157895, + "grad_norm": 0.478515625, + "learning_rate": 2.2842105263157897e-05, + "loss": 3.2384, + "step": 17047 + }, + { + "epoch": 0.8972631578947369, + "grad_norm": 0.51953125, + "learning_rate": 2.2830409356725146e-05, + "loss": 2.8527, + "step": 17048 + }, + { + "epoch": 0.8973157894736842, + "grad_norm": 0.48828125, + "learning_rate": 2.28187134502924e-05, + "loss": 3.2609, + "step": 17049 + }, + { + "epoch": 0.8973684210526316, + "grad_norm": 0.490234375, + "learning_rate": 2.280701754385965e-05, + "loss": 3.3062, + "step": 17050 + }, + { + "epoch": 0.8974210526315789, + "grad_norm": 0.48828125, + "learning_rate": 2.2795321637426902e-05, + "loss": 3.3082, + "step": 17051 + }, + { + "epoch": 0.8974736842105263, + "grad_norm": 0.4921875, + "learning_rate": 2.2783625730994155e-05, + "loss": 3.4343, + "step": 17052 + }, + { + "epoch": 0.8975263157894737, + "grad_norm": 0.48828125, + "learning_rate": 2.2771929824561404e-05, + "loss": 3.2428, + "step": 17053 + }, + { + "epoch": 0.897578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.2760233918128657e-05, + "loss": 2.9786, + "step": 17054 + }, + { + "epoch": 0.8976315789473684, + "grad_norm": 0.490234375, + "learning_rate": 2.2748538011695907e-05, + "loss": 3.3697, + "step": 17055 + }, + { + "epoch": 0.8976842105263158, + "grad_norm": 0.54296875, + "learning_rate": 2.273684210526316e-05, + "loss": 2.9315, + "step": 17056 + }, + { + "epoch": 0.8977368421052632, + "grad_norm": 0.5234375, + "learning_rate": 2.272514619883041e-05, + "loss": 3.0761, + "step": 17057 + }, + { + "epoch": 0.8977894736842106, + "grad_norm": 0.5078125, + "learning_rate": 2.2713450292397662e-05, + "loss": 3.2992, + "step": 17058 + }, + { + "epoch": 0.8978421052631579, + "grad_norm": 0.5234375, + "learning_rate": 2.2701754385964912e-05, + "loss": 3.3962, + "step": 17059 + }, + { + "epoch": 0.8978947368421053, + "grad_norm": 0.5546875, + "learning_rate": 2.2690058479532165e-05, + "loss": 2.9121, + "step": 17060 + }, + { + "epoch": 0.8979473684210526, + "grad_norm": 0.5, + "learning_rate": 2.2678362573099415e-05, + "loss": 3.1541, + "step": 17061 + }, + { + "epoch": 0.898, + "grad_norm": 0.48828125, + "learning_rate": 2.2666666666666668e-05, + "loss": 3.2689, + "step": 17062 + }, + { + "epoch": 0.8980526315789473, + "grad_norm": 0.53125, + "learning_rate": 2.2654970760233917e-05, + "loss": 3.2903, + "step": 17063 + }, + { + "epoch": 0.8981052631578947, + "grad_norm": 0.515625, + "learning_rate": 2.264327485380117e-05, + "loss": 3.3197, + "step": 17064 + }, + { + "epoch": 0.8981578947368422, + "grad_norm": 0.498046875, + "learning_rate": 2.2631578947368423e-05, + "loss": 3.4004, + "step": 17065 + }, + { + "epoch": 0.8982105263157895, + "grad_norm": 0.49609375, + "learning_rate": 2.2619883040935673e-05, + "loss": 3.0054, + "step": 17066 + }, + { + "epoch": 0.8982631578947369, + "grad_norm": 0.49609375, + "learning_rate": 2.2608187134502926e-05, + "loss": 3.1674, + "step": 17067 + }, + { + "epoch": 0.8983157894736842, + "grad_norm": 0.5, + "learning_rate": 2.2596491228070175e-05, + "loss": 3.4829, + "step": 17068 + }, + { + "epoch": 0.8983684210526316, + "grad_norm": 0.50390625, + "learning_rate": 2.258479532163743e-05, + "loss": 3.0536, + "step": 17069 + }, + { + "epoch": 0.8984210526315789, + "grad_norm": 0.48828125, + "learning_rate": 2.2573099415204678e-05, + "loss": 3.1053, + "step": 17070 + }, + { + "epoch": 0.8984736842105263, + "grad_norm": 0.482421875, + "learning_rate": 2.256140350877193e-05, + "loss": 2.9999, + "step": 17071 + }, + { + "epoch": 0.8985263157894737, + "grad_norm": 0.490234375, + "learning_rate": 2.254970760233918e-05, + "loss": 3.5938, + "step": 17072 + }, + { + "epoch": 0.898578947368421, + "grad_norm": 0.4921875, + "learning_rate": 2.2538011695906433e-05, + "loss": 3.4056, + "step": 17073 + }, + { + "epoch": 0.8986315789473684, + "grad_norm": 0.578125, + "learning_rate": 2.2526315789473683e-05, + "loss": 2.8091, + "step": 17074 + }, + { + "epoch": 0.8986842105263158, + "grad_norm": 0.515625, + "learning_rate": 2.2514619883040936e-05, + "loss": 2.9465, + "step": 17075 + }, + { + "epoch": 0.8987368421052632, + "grad_norm": 0.49609375, + "learning_rate": 2.250292397660819e-05, + "loss": 3.2142, + "step": 17076 + }, + { + "epoch": 0.8987894736842105, + "grad_norm": 0.490234375, + "learning_rate": 2.2491228070175442e-05, + "loss": 3.2742, + "step": 17077 + }, + { + "epoch": 0.8988421052631579, + "grad_norm": 0.478515625, + "learning_rate": 2.247953216374269e-05, + "loss": 2.967, + "step": 17078 + }, + { + "epoch": 0.8988947368421053, + "grad_norm": 0.484375, + "learning_rate": 2.2467836257309945e-05, + "loss": 3.3779, + "step": 17079 + }, + { + "epoch": 0.8989473684210526, + "grad_norm": 0.5390625, + "learning_rate": 2.2456140350877194e-05, + "loss": 3.1006, + "step": 17080 + }, + { + "epoch": 0.899, + "grad_norm": 0.50390625, + "learning_rate": 2.2444444444444447e-05, + "loss": 3.2021, + "step": 17081 + }, + { + "epoch": 0.8990526315789473, + "grad_norm": 0.498046875, + "learning_rate": 2.2432748538011697e-05, + "loss": 3.1443, + "step": 17082 + }, + { + "epoch": 0.8991052631578947, + "grad_norm": 0.5, + "learning_rate": 2.242105263157895e-05, + "loss": 3.1021, + "step": 17083 + }, + { + "epoch": 0.8991578947368422, + "grad_norm": 0.482421875, + "learning_rate": 2.24093567251462e-05, + "loss": 2.6177, + "step": 17084 + }, + { + "epoch": 0.8992105263157895, + "grad_norm": 0.4765625, + "learning_rate": 2.2397660818713452e-05, + "loss": 3.3316, + "step": 17085 + }, + { + "epoch": 0.8992631578947369, + "grad_norm": 0.5078125, + "learning_rate": 2.2385964912280702e-05, + "loss": 2.901, + "step": 17086 + }, + { + "epoch": 0.8993157894736842, + "grad_norm": 0.48828125, + "learning_rate": 2.237426900584795e-05, + "loss": 3.3628, + "step": 17087 + }, + { + "epoch": 0.8993684210526316, + "grad_norm": 0.48828125, + "learning_rate": 2.2362573099415204e-05, + "loss": 3.2409, + "step": 17088 + }, + { + "epoch": 0.8994210526315789, + "grad_norm": 0.498046875, + "learning_rate": 2.2350877192982457e-05, + "loss": 3.1187, + "step": 17089 + }, + { + "epoch": 0.8994736842105263, + "grad_norm": 0.49609375, + "learning_rate": 2.233918128654971e-05, + "loss": 3.0363, + "step": 17090 + }, + { + "epoch": 0.8995263157894737, + "grad_norm": 0.462890625, + "learning_rate": 2.232748538011696e-05, + "loss": 2.6412, + "step": 17091 + }, + { + "epoch": 0.899578947368421, + "grad_norm": 0.4921875, + "learning_rate": 2.2315789473684213e-05, + "loss": 2.9379, + "step": 17092 + }, + { + "epoch": 0.8996315789473684, + "grad_norm": 0.54296875, + "learning_rate": 2.2304093567251463e-05, + "loss": 3.1336, + "step": 17093 + }, + { + "epoch": 0.8996842105263158, + "grad_norm": 0.5, + "learning_rate": 2.2292397660818716e-05, + "loss": 3.515, + "step": 17094 + }, + { + "epoch": 0.8997368421052632, + "grad_norm": 0.498046875, + "learning_rate": 2.2280701754385965e-05, + "loss": 2.9847, + "step": 17095 + }, + { + "epoch": 0.8997894736842105, + "grad_norm": 0.5, + "learning_rate": 2.2269005847953218e-05, + "loss": 3.4355, + "step": 17096 + }, + { + "epoch": 0.8998421052631579, + "grad_norm": 0.4765625, + "learning_rate": 2.2257309941520468e-05, + "loss": 3.3016, + "step": 17097 + }, + { + "epoch": 0.8998947368421053, + "grad_norm": 0.4921875, + "learning_rate": 2.224561403508772e-05, + "loss": 3.4046, + "step": 17098 + }, + { + "epoch": 0.8999473684210526, + "grad_norm": 0.482421875, + "learning_rate": 2.223391812865497e-05, + "loss": 2.9481, + "step": 17099 + }, + { + "epoch": 0.9, + "grad_norm": 0.515625, + "learning_rate": 2.2222222222222223e-05, + "loss": 2.9475, + "step": 17100 + }, + { + "epoch": 0.9000526315789473, + "grad_norm": 0.515625, + "learning_rate": 2.2210526315789476e-05, + "loss": 3.236, + "step": 17101 + }, + { + "epoch": 0.9001052631578947, + "grad_norm": 0.494140625, + "learning_rate": 2.2198830409356726e-05, + "loss": 3.4441, + "step": 17102 + }, + { + "epoch": 0.9001578947368422, + "grad_norm": 0.5, + "learning_rate": 2.218713450292398e-05, + "loss": 3.2257, + "step": 17103 + }, + { + "epoch": 0.9002105263157895, + "grad_norm": 0.484375, + "learning_rate": 2.217543859649123e-05, + "loss": 3.0543, + "step": 17104 + }, + { + "epoch": 0.9002631578947369, + "grad_norm": 0.51953125, + "learning_rate": 2.216374269005848e-05, + "loss": 3.2076, + "step": 17105 + }, + { + "epoch": 0.9003157894736842, + "grad_norm": 0.486328125, + "learning_rate": 2.215204678362573e-05, + "loss": 3.4317, + "step": 17106 + }, + { + "epoch": 0.9003684210526316, + "grad_norm": 0.50390625, + "learning_rate": 2.2140350877192984e-05, + "loss": 3.3698, + "step": 17107 + }, + { + "epoch": 0.9004210526315789, + "grad_norm": 0.5, + "learning_rate": 2.2128654970760234e-05, + "loss": 3.1506, + "step": 17108 + }, + { + "epoch": 0.9004736842105263, + "grad_norm": 0.50390625, + "learning_rate": 2.2116959064327487e-05, + "loss": 2.9799, + "step": 17109 + }, + { + "epoch": 0.9005263157894737, + "grad_norm": 0.50390625, + "learning_rate": 2.2105263157894736e-05, + "loss": 3.3137, + "step": 17110 + }, + { + "epoch": 0.900578947368421, + "grad_norm": 0.484375, + "learning_rate": 2.209356725146199e-05, + "loss": 3.2922, + "step": 17111 + }, + { + "epoch": 0.9006315789473684, + "grad_norm": 0.5078125, + "learning_rate": 2.208187134502924e-05, + "loss": 3.0219, + "step": 17112 + }, + { + "epoch": 0.9006842105263158, + "grad_norm": 0.50390625, + "learning_rate": 2.2070175438596492e-05, + "loss": 3.0319, + "step": 17113 + }, + { + "epoch": 0.9007368421052632, + "grad_norm": 0.51171875, + "learning_rate": 2.2058479532163745e-05, + "loss": 3.2631, + "step": 17114 + }, + { + "epoch": 0.9007894736842105, + "grad_norm": 0.494140625, + "learning_rate": 2.2046783625730998e-05, + "loss": 3.5486, + "step": 17115 + }, + { + "epoch": 0.9008421052631579, + "grad_norm": 0.50390625, + "learning_rate": 2.2035087719298247e-05, + "loss": 2.8576, + "step": 17116 + }, + { + "epoch": 0.9008947368421053, + "grad_norm": 0.546875, + "learning_rate": 2.20233918128655e-05, + "loss": 2.8577, + "step": 17117 + }, + { + "epoch": 0.9009473684210526, + "grad_norm": 0.50390625, + "learning_rate": 2.201169590643275e-05, + "loss": 3.2752, + "step": 17118 + }, + { + "epoch": 0.901, + "grad_norm": 0.52734375, + "learning_rate": 2.2000000000000003e-05, + "loss": 3.4693, + "step": 17119 + }, + { + "epoch": 0.9010526315789473, + "grad_norm": 0.5078125, + "learning_rate": 2.1988304093567252e-05, + "loss": 3.4138, + "step": 17120 + }, + { + "epoch": 0.9011052631578947, + "grad_norm": 0.494140625, + "learning_rate": 2.1976608187134502e-05, + "loss": 3.1298, + "step": 17121 + }, + { + "epoch": 0.9011578947368422, + "grad_norm": 0.486328125, + "learning_rate": 2.1964912280701755e-05, + "loss": 2.8001, + "step": 17122 + }, + { + "epoch": 0.9012105263157895, + "grad_norm": 0.5234375, + "learning_rate": 2.1953216374269005e-05, + "loss": 3.5982, + "step": 17123 + }, + { + "epoch": 0.9012631578947369, + "grad_norm": 0.490234375, + "learning_rate": 2.1941520467836258e-05, + "loss": 3.5472, + "step": 17124 + }, + { + "epoch": 0.9013157894736842, + "grad_norm": 0.51953125, + "learning_rate": 2.1929824561403507e-05, + "loss": 3.0257, + "step": 17125 + }, + { + "epoch": 0.9013684210526316, + "grad_norm": 0.53125, + "learning_rate": 2.1918128654970764e-05, + "loss": 3.7809, + "step": 17126 + }, + { + "epoch": 0.9014210526315789, + "grad_norm": 0.490234375, + "learning_rate": 2.1906432748538013e-05, + "loss": 3.4225, + "step": 17127 + }, + { + "epoch": 0.9014736842105263, + "grad_norm": 0.51171875, + "learning_rate": 2.1894736842105266e-05, + "loss": 3.3598, + "step": 17128 + }, + { + "epoch": 0.9015263157894737, + "grad_norm": 0.51171875, + "learning_rate": 2.1883040935672516e-05, + "loss": 3.1191, + "step": 17129 + }, + { + "epoch": 0.901578947368421, + "grad_norm": 0.51953125, + "learning_rate": 2.187134502923977e-05, + "loss": 3.0988, + "step": 17130 + }, + { + "epoch": 0.9016315789473685, + "grad_norm": 0.50390625, + "learning_rate": 2.1859649122807018e-05, + "loss": 3.093, + "step": 17131 + }, + { + "epoch": 0.9016842105263158, + "grad_norm": 0.53125, + "learning_rate": 2.184795321637427e-05, + "loss": 3.0012, + "step": 17132 + }, + { + "epoch": 0.9017368421052632, + "grad_norm": 0.490234375, + "learning_rate": 2.183625730994152e-05, + "loss": 3.1626, + "step": 17133 + }, + { + "epoch": 0.9017894736842105, + "grad_norm": 0.61328125, + "learning_rate": 2.1824561403508774e-05, + "loss": 3.4663, + "step": 17134 + }, + { + "epoch": 0.9018421052631579, + "grad_norm": 0.53515625, + "learning_rate": 2.1812865497076023e-05, + "loss": 3.1534, + "step": 17135 + }, + { + "epoch": 0.9018947368421053, + "grad_norm": 0.52734375, + "learning_rate": 2.1801169590643276e-05, + "loss": 3.3093, + "step": 17136 + }, + { + "epoch": 0.9019473684210526, + "grad_norm": 0.5390625, + "learning_rate": 2.1789473684210526e-05, + "loss": 3.3557, + "step": 17137 + }, + { + "epoch": 0.902, + "grad_norm": 0.515625, + "learning_rate": 2.177777777777778e-05, + "loss": 3.064, + "step": 17138 + }, + { + "epoch": 0.9020526315789473, + "grad_norm": 0.5078125, + "learning_rate": 2.1766081871345032e-05, + "loss": 3.5155, + "step": 17139 + }, + { + "epoch": 0.9021052631578947, + "grad_norm": 0.50390625, + "learning_rate": 2.175438596491228e-05, + "loss": 3.092, + "step": 17140 + }, + { + "epoch": 0.9021578947368422, + "grad_norm": 0.474609375, + "learning_rate": 2.1742690058479535e-05, + "loss": 3.3432, + "step": 17141 + }, + { + "epoch": 0.9022105263157895, + "grad_norm": 0.48828125, + "learning_rate": 2.1730994152046784e-05, + "loss": 3.1802, + "step": 17142 + }, + { + "epoch": 0.9022631578947369, + "grad_norm": 0.478515625, + "learning_rate": 2.1719298245614037e-05, + "loss": 3.7073, + "step": 17143 + }, + { + "epoch": 0.9023157894736842, + "grad_norm": 0.49609375, + "learning_rate": 2.1707602339181287e-05, + "loss": 3.477, + "step": 17144 + }, + { + "epoch": 0.9023684210526316, + "grad_norm": 0.53125, + "learning_rate": 2.169590643274854e-05, + "loss": 3.4982, + "step": 17145 + }, + { + "epoch": 0.9024210526315789, + "grad_norm": 0.51953125, + "learning_rate": 2.168421052631579e-05, + "loss": 3.2292, + "step": 17146 + }, + { + "epoch": 0.9024736842105263, + "grad_norm": 0.54296875, + "learning_rate": 2.1672514619883042e-05, + "loss": 3.2341, + "step": 17147 + }, + { + "epoch": 0.9025263157894737, + "grad_norm": 0.5078125, + "learning_rate": 2.1660818713450292e-05, + "loss": 3.5914, + "step": 17148 + }, + { + "epoch": 0.902578947368421, + "grad_norm": 0.50390625, + "learning_rate": 2.1649122807017545e-05, + "loss": 3.5043, + "step": 17149 + }, + { + "epoch": 0.9026315789473685, + "grad_norm": 0.515625, + "learning_rate": 2.1637426900584794e-05, + "loss": 2.8952, + "step": 17150 + }, + { + "epoch": 0.9026842105263158, + "grad_norm": 0.5, + "learning_rate": 2.162573099415205e-05, + "loss": 3.2692, + "step": 17151 + }, + { + "epoch": 0.9027368421052632, + "grad_norm": 0.51171875, + "learning_rate": 2.16140350877193e-05, + "loss": 3.182, + "step": 17152 + }, + { + "epoch": 0.9027894736842105, + "grad_norm": 0.5, + "learning_rate": 2.1602339181286553e-05, + "loss": 3.2873, + "step": 17153 + }, + { + "epoch": 0.9028421052631579, + "grad_norm": 0.51171875, + "learning_rate": 2.1590643274853803e-05, + "loss": 2.8506, + "step": 17154 + }, + { + "epoch": 0.9028947368421053, + "grad_norm": 0.50390625, + "learning_rate": 2.1578947368421053e-05, + "loss": 3.0758, + "step": 17155 + }, + { + "epoch": 0.9029473684210526, + "grad_norm": 0.50390625, + "learning_rate": 2.1567251461988306e-05, + "loss": 3.532, + "step": 17156 + }, + { + "epoch": 0.903, + "grad_norm": 0.515625, + "learning_rate": 2.1555555555555555e-05, + "loss": 2.8374, + "step": 17157 + }, + { + "epoch": 0.9030526315789473, + "grad_norm": 0.51171875, + "learning_rate": 2.1543859649122808e-05, + "loss": 3.1341, + "step": 17158 + }, + { + "epoch": 0.9031052631578947, + "grad_norm": 0.515625, + "learning_rate": 2.1532163742690058e-05, + "loss": 3.7507, + "step": 17159 + }, + { + "epoch": 0.9031578947368422, + "grad_norm": 0.50390625, + "learning_rate": 2.152046783625731e-05, + "loss": 2.9135, + "step": 17160 + }, + { + "epoch": 0.9032105263157895, + "grad_norm": 0.5, + "learning_rate": 2.150877192982456e-05, + "loss": 3.6592, + "step": 17161 + }, + { + "epoch": 0.9032631578947369, + "grad_norm": 0.515625, + "learning_rate": 2.1497076023391813e-05, + "loss": 3.3507, + "step": 17162 + }, + { + "epoch": 0.9033157894736842, + "grad_norm": 0.49609375, + "learning_rate": 2.1485380116959063e-05, + "loss": 3.4312, + "step": 17163 + }, + { + "epoch": 0.9033684210526316, + "grad_norm": 0.478515625, + "learning_rate": 2.147368421052632e-05, + "loss": 2.8475, + "step": 17164 + }, + { + "epoch": 0.9034210526315789, + "grad_norm": 0.484375, + "learning_rate": 2.146198830409357e-05, + "loss": 3.3122, + "step": 17165 + }, + { + "epoch": 0.9034736842105263, + "grad_norm": 0.54296875, + "learning_rate": 2.1450292397660822e-05, + "loss": 3.2965, + "step": 17166 + }, + { + "epoch": 0.9035263157894737, + "grad_norm": 0.498046875, + "learning_rate": 2.143859649122807e-05, + "loss": 3.464, + "step": 17167 + }, + { + "epoch": 0.903578947368421, + "grad_norm": 0.57421875, + "learning_rate": 2.1426900584795324e-05, + "loss": 3.4347, + "step": 17168 + }, + { + "epoch": 0.9036315789473685, + "grad_norm": 0.4921875, + "learning_rate": 2.1415204678362574e-05, + "loss": 3.1157, + "step": 17169 + }, + { + "epoch": 0.9036842105263158, + "grad_norm": 0.4921875, + "learning_rate": 2.1403508771929827e-05, + "loss": 3.0563, + "step": 17170 + }, + { + "epoch": 0.9037368421052632, + "grad_norm": 0.458984375, + "learning_rate": 2.1391812865497076e-05, + "loss": 3.1074, + "step": 17171 + }, + { + "epoch": 0.9037894736842105, + "grad_norm": 0.515625, + "learning_rate": 2.138011695906433e-05, + "loss": 3.3558, + "step": 17172 + }, + { + "epoch": 0.9038421052631579, + "grad_norm": 0.5390625, + "learning_rate": 2.136842105263158e-05, + "loss": 3.1691, + "step": 17173 + }, + { + "epoch": 0.9038947368421053, + "grad_norm": 0.51171875, + "learning_rate": 2.1356725146198832e-05, + "loss": 3.5907, + "step": 17174 + }, + { + "epoch": 0.9039473684210526, + "grad_norm": 0.498046875, + "learning_rate": 2.134502923976608e-05, + "loss": 3.2294, + "step": 17175 + }, + { + "epoch": 0.904, + "grad_norm": 0.515625, + "learning_rate": 2.1333333333333335e-05, + "loss": 3.5299, + "step": 17176 + }, + { + "epoch": 0.9040526315789473, + "grad_norm": 0.51171875, + "learning_rate": 2.1321637426900588e-05, + "loss": 3.8385, + "step": 17177 + }, + { + "epoch": 0.9041052631578947, + "grad_norm": 0.478515625, + "learning_rate": 2.1309941520467837e-05, + "loss": 3.419, + "step": 17178 + }, + { + "epoch": 0.9041578947368422, + "grad_norm": 0.51171875, + "learning_rate": 2.129824561403509e-05, + "loss": 3.6274, + "step": 17179 + }, + { + "epoch": 0.9042105263157895, + "grad_norm": 0.53125, + "learning_rate": 2.128654970760234e-05, + "loss": 3.027, + "step": 17180 + }, + { + "epoch": 0.9042631578947369, + "grad_norm": 0.490234375, + "learning_rate": 2.1274853801169593e-05, + "loss": 3.3062, + "step": 17181 + }, + { + "epoch": 0.9043157894736842, + "grad_norm": 0.546875, + "learning_rate": 2.1263157894736842e-05, + "loss": 3.0835, + "step": 17182 + }, + { + "epoch": 0.9043684210526316, + "grad_norm": 0.486328125, + "learning_rate": 2.1251461988304095e-05, + "loss": 2.8786, + "step": 17183 + }, + { + "epoch": 0.9044210526315789, + "grad_norm": 0.5234375, + "learning_rate": 2.1239766081871345e-05, + "loss": 3.3309, + "step": 17184 + }, + { + "epoch": 0.9044736842105263, + "grad_norm": 0.5390625, + "learning_rate": 2.1228070175438598e-05, + "loss": 3.4049, + "step": 17185 + }, + { + "epoch": 0.9045263157894737, + "grad_norm": 0.5234375, + "learning_rate": 2.1216374269005847e-05, + "loss": 2.8849, + "step": 17186 + }, + { + "epoch": 0.904578947368421, + "grad_norm": 0.5234375, + "learning_rate": 2.12046783625731e-05, + "loss": 3.1284, + "step": 17187 + }, + { + "epoch": 0.9046315789473685, + "grad_norm": 0.5, + "learning_rate": 2.119298245614035e-05, + "loss": 2.6808, + "step": 17188 + }, + { + "epoch": 0.9046842105263158, + "grad_norm": 0.515625, + "learning_rate": 2.1181286549707603e-05, + "loss": 3.4624, + "step": 17189 + }, + { + "epoch": 0.9047368421052632, + "grad_norm": 0.4921875, + "learning_rate": 2.1169590643274856e-05, + "loss": 2.9541, + "step": 17190 + }, + { + "epoch": 0.9047894736842105, + "grad_norm": 0.5078125, + "learning_rate": 2.1157894736842106e-05, + "loss": 3.071, + "step": 17191 + }, + { + "epoch": 0.9048421052631579, + "grad_norm": 0.5, + "learning_rate": 2.114619883040936e-05, + "loss": 3.2292, + "step": 17192 + }, + { + "epoch": 0.9048947368421053, + "grad_norm": 0.498046875, + "learning_rate": 2.1134502923976608e-05, + "loss": 3.2965, + "step": 17193 + }, + { + "epoch": 0.9049473684210526, + "grad_norm": 0.5078125, + "learning_rate": 2.112280701754386e-05, + "loss": 2.9906, + "step": 17194 + }, + { + "epoch": 0.905, + "grad_norm": 0.498046875, + "learning_rate": 2.111111111111111e-05, + "loss": 3.1734, + "step": 17195 + }, + { + "epoch": 0.9050526315789473, + "grad_norm": 0.5390625, + "learning_rate": 2.1099415204678364e-05, + "loss": 2.9533, + "step": 17196 + }, + { + "epoch": 0.9051052631578947, + "grad_norm": 0.5, + "learning_rate": 2.1087719298245613e-05, + "loss": 3.0102, + "step": 17197 + }, + { + "epoch": 0.905157894736842, + "grad_norm": 0.5703125, + "learning_rate": 2.1076023391812866e-05, + "loss": 2.8895, + "step": 17198 + }, + { + "epoch": 0.9052105263157895, + "grad_norm": 0.484375, + "learning_rate": 2.1064327485380116e-05, + "loss": 3.304, + "step": 17199 + }, + { + "epoch": 0.9052631578947369, + "grad_norm": 0.54296875, + "learning_rate": 2.105263157894737e-05, + "loss": 2.9593, + "step": 17200 + }, + { + "epoch": 0.9053157894736842, + "grad_norm": 0.5, + "learning_rate": 2.1040935672514622e-05, + "loss": 3.309, + "step": 17201 + }, + { + "epoch": 0.9053684210526316, + "grad_norm": 0.5, + "learning_rate": 2.1029239766081875e-05, + "loss": 3.1013, + "step": 17202 + }, + { + "epoch": 0.9054210526315789, + "grad_norm": 0.48046875, + "learning_rate": 2.1017543859649124e-05, + "loss": 3.0142, + "step": 17203 + }, + { + "epoch": 0.9054736842105263, + "grad_norm": 0.51171875, + "learning_rate": 2.1005847953216377e-05, + "loss": 3.2759, + "step": 17204 + }, + { + "epoch": 0.9055263157894737, + "grad_norm": 0.47265625, + "learning_rate": 2.0994152046783627e-05, + "loss": 3.1946, + "step": 17205 + }, + { + "epoch": 0.905578947368421, + "grad_norm": 0.51953125, + "learning_rate": 2.098245614035088e-05, + "loss": 3.4085, + "step": 17206 + }, + { + "epoch": 0.9056315789473685, + "grad_norm": 0.54296875, + "learning_rate": 2.097076023391813e-05, + "loss": 3.6399, + "step": 17207 + }, + { + "epoch": 0.9056842105263158, + "grad_norm": 0.51171875, + "learning_rate": 2.0959064327485383e-05, + "loss": 3.1222, + "step": 17208 + }, + { + "epoch": 0.9057368421052632, + "grad_norm": 0.55078125, + "learning_rate": 2.0947368421052632e-05, + "loss": 3.6595, + "step": 17209 + }, + { + "epoch": 0.9057894736842105, + "grad_norm": 0.48046875, + "learning_rate": 2.0935672514619882e-05, + "loss": 3.4008, + "step": 17210 + }, + { + "epoch": 0.9058421052631579, + "grad_norm": 0.515625, + "learning_rate": 2.0923976608187135e-05, + "loss": 3.0966, + "step": 17211 + }, + { + "epoch": 0.9058947368421053, + "grad_norm": 0.51171875, + "learning_rate": 2.0912280701754384e-05, + "loss": 2.7211, + "step": 17212 + }, + { + "epoch": 0.9059473684210526, + "grad_norm": 0.49609375, + "learning_rate": 2.0900584795321637e-05, + "loss": 3.3232, + "step": 17213 + }, + { + "epoch": 0.906, + "grad_norm": 0.498046875, + "learning_rate": 2.088888888888889e-05, + "loss": 3.3369, + "step": 17214 + }, + { + "epoch": 0.9060526315789473, + "grad_norm": 0.5078125, + "learning_rate": 2.0877192982456143e-05, + "loss": 2.7934, + "step": 17215 + }, + { + "epoch": 0.9061052631578947, + "grad_norm": 0.52734375, + "learning_rate": 2.0865497076023393e-05, + "loss": 3.5283, + "step": 17216 + }, + { + "epoch": 0.906157894736842, + "grad_norm": 0.50390625, + "learning_rate": 2.0853801169590646e-05, + "loss": 3.2553, + "step": 17217 + }, + { + "epoch": 0.9062105263157895, + "grad_norm": 0.498046875, + "learning_rate": 2.0842105263157895e-05, + "loss": 2.9929, + "step": 17218 + }, + { + "epoch": 0.9062631578947369, + "grad_norm": 0.54296875, + "learning_rate": 2.083040935672515e-05, + "loss": 2.9406, + "step": 17219 + }, + { + "epoch": 0.9063157894736842, + "grad_norm": 0.51171875, + "learning_rate": 2.0818713450292398e-05, + "loss": 3.2413, + "step": 17220 + }, + { + "epoch": 0.9063684210526316, + "grad_norm": 0.4921875, + "learning_rate": 2.080701754385965e-05, + "loss": 2.9585, + "step": 17221 + }, + { + "epoch": 0.9064210526315789, + "grad_norm": 0.5078125, + "learning_rate": 2.07953216374269e-05, + "loss": 3.3807, + "step": 17222 + }, + { + "epoch": 0.9064736842105263, + "grad_norm": 0.482421875, + "learning_rate": 2.0783625730994154e-05, + "loss": 3.0944, + "step": 17223 + }, + { + "epoch": 0.9065263157894737, + "grad_norm": 0.52734375, + "learning_rate": 2.0771929824561403e-05, + "loss": 3.4051, + "step": 17224 + }, + { + "epoch": 0.906578947368421, + "grad_norm": 0.5078125, + "learning_rate": 2.0760233918128656e-05, + "loss": 3.6099, + "step": 17225 + }, + { + "epoch": 0.9066315789473685, + "grad_norm": 0.484375, + "learning_rate": 2.0748538011695906e-05, + "loss": 3.0311, + "step": 17226 + }, + { + "epoch": 0.9066842105263158, + "grad_norm": 0.498046875, + "learning_rate": 2.073684210526316e-05, + "loss": 3.474, + "step": 17227 + }, + { + "epoch": 0.9067368421052632, + "grad_norm": 0.494140625, + "learning_rate": 2.0725146198830412e-05, + "loss": 2.9689, + "step": 17228 + }, + { + "epoch": 0.9067894736842105, + "grad_norm": 0.51171875, + "learning_rate": 2.071345029239766e-05, + "loss": 3.0933, + "step": 17229 + }, + { + "epoch": 0.9068421052631579, + "grad_norm": 0.498046875, + "learning_rate": 2.0701754385964914e-05, + "loss": 3.3205, + "step": 17230 + }, + { + "epoch": 0.9068947368421053, + "grad_norm": 0.50390625, + "learning_rate": 2.0690058479532164e-05, + "loss": 3.5004, + "step": 17231 + }, + { + "epoch": 0.9069473684210526, + "grad_norm": 0.50390625, + "learning_rate": 2.0678362573099417e-05, + "loss": 3.2745, + "step": 17232 + }, + { + "epoch": 0.907, + "grad_norm": 0.498046875, + "learning_rate": 2.0666666666666666e-05, + "loss": 3.471, + "step": 17233 + }, + { + "epoch": 0.9070526315789473, + "grad_norm": 0.5078125, + "learning_rate": 2.065497076023392e-05, + "loss": 3.2087, + "step": 17234 + }, + { + "epoch": 0.9071052631578947, + "grad_norm": 0.482421875, + "learning_rate": 2.064327485380117e-05, + "loss": 3.5049, + "step": 17235 + }, + { + "epoch": 0.907157894736842, + "grad_norm": 0.53125, + "learning_rate": 2.0631578947368422e-05, + "loss": 3.6055, + "step": 17236 + }, + { + "epoch": 0.9072105263157895, + "grad_norm": 0.470703125, + "learning_rate": 2.061988304093567e-05, + "loss": 3.2235, + "step": 17237 + }, + { + "epoch": 0.9072631578947369, + "grad_norm": 0.5078125, + "learning_rate": 2.0608187134502925e-05, + "loss": 3.1636, + "step": 17238 + }, + { + "epoch": 0.9073157894736842, + "grad_norm": 0.5078125, + "learning_rate": 2.0596491228070178e-05, + "loss": 3.3847, + "step": 17239 + }, + { + "epoch": 0.9073684210526316, + "grad_norm": 0.498046875, + "learning_rate": 2.058479532163743e-05, + "loss": 3.4044, + "step": 17240 + }, + { + "epoch": 0.9074210526315789, + "grad_norm": 0.486328125, + "learning_rate": 2.057309941520468e-05, + "loss": 3.4439, + "step": 17241 + }, + { + "epoch": 0.9074736842105263, + "grad_norm": 0.50390625, + "learning_rate": 2.0561403508771933e-05, + "loss": 3.7154, + "step": 17242 + }, + { + "epoch": 0.9075263157894737, + "grad_norm": 0.4765625, + "learning_rate": 2.0549707602339183e-05, + "loss": 3.4507, + "step": 17243 + }, + { + "epoch": 0.907578947368421, + "grad_norm": 0.4921875, + "learning_rate": 2.0538011695906432e-05, + "loss": 3.0942, + "step": 17244 + }, + { + "epoch": 0.9076315789473685, + "grad_norm": 0.5078125, + "learning_rate": 2.0526315789473685e-05, + "loss": 3.6592, + "step": 17245 + }, + { + "epoch": 0.9076842105263158, + "grad_norm": 0.498046875, + "learning_rate": 2.0514619883040935e-05, + "loss": 3.3757, + "step": 17246 + }, + { + "epoch": 0.9077368421052632, + "grad_norm": 0.48828125, + "learning_rate": 2.0502923976608188e-05, + "loss": 3.239, + "step": 17247 + }, + { + "epoch": 0.9077894736842105, + "grad_norm": 0.490234375, + "learning_rate": 2.0491228070175437e-05, + "loss": 3.3787, + "step": 17248 + }, + { + "epoch": 0.9078421052631579, + "grad_norm": 0.466796875, + "learning_rate": 2.047953216374269e-05, + "loss": 2.7247, + "step": 17249 + }, + { + "epoch": 0.9078947368421053, + "grad_norm": 0.515625, + "learning_rate": 2.046783625730994e-05, + "loss": 3.2106, + "step": 17250 + }, + { + "epoch": 0.9079473684210526, + "grad_norm": 0.4765625, + "learning_rate": 2.0456140350877193e-05, + "loss": 2.9569, + "step": 17251 + }, + { + "epoch": 0.908, + "grad_norm": 0.50390625, + "learning_rate": 2.0444444444444446e-05, + "loss": 3.6169, + "step": 17252 + }, + { + "epoch": 0.9080526315789473, + "grad_norm": 0.50390625, + "learning_rate": 2.04327485380117e-05, + "loss": 3.0655, + "step": 17253 + }, + { + "epoch": 0.9081052631578947, + "grad_norm": 0.486328125, + "learning_rate": 2.042105263157895e-05, + "loss": 3.0942, + "step": 17254 + }, + { + "epoch": 0.908157894736842, + "grad_norm": 0.4921875, + "learning_rate": 2.04093567251462e-05, + "loss": 2.9576, + "step": 17255 + }, + { + "epoch": 0.9082105263157895, + "grad_norm": 0.5234375, + "learning_rate": 2.039766081871345e-05, + "loss": 3.5296, + "step": 17256 + }, + { + "epoch": 0.9082631578947369, + "grad_norm": 0.50390625, + "learning_rate": 2.0385964912280704e-05, + "loss": 2.9764, + "step": 17257 + }, + { + "epoch": 0.9083157894736842, + "grad_norm": 0.482421875, + "learning_rate": 2.0374269005847954e-05, + "loss": 3.2601, + "step": 17258 + }, + { + "epoch": 0.9083684210526316, + "grad_norm": 0.490234375, + "learning_rate": 2.0362573099415207e-05, + "loss": 2.8326, + "step": 17259 + }, + { + "epoch": 0.9084210526315789, + "grad_norm": 0.50390625, + "learning_rate": 2.0350877192982456e-05, + "loss": 3.3413, + "step": 17260 + }, + { + "epoch": 0.9084736842105263, + "grad_norm": 0.515625, + "learning_rate": 2.033918128654971e-05, + "loss": 3.1505, + "step": 17261 + }, + { + "epoch": 0.9085263157894737, + "grad_norm": 0.5234375, + "learning_rate": 2.032748538011696e-05, + "loss": 3.2551, + "step": 17262 + }, + { + "epoch": 0.908578947368421, + "grad_norm": 0.515625, + "learning_rate": 2.0315789473684212e-05, + "loss": 3.3061, + "step": 17263 + }, + { + "epoch": 0.9086315789473685, + "grad_norm": 0.54296875, + "learning_rate": 2.0304093567251465e-05, + "loss": 3.104, + "step": 17264 + }, + { + "epoch": 0.9086842105263158, + "grad_norm": 0.51171875, + "learning_rate": 2.0292397660818714e-05, + "loss": 3.3231, + "step": 17265 + }, + { + "epoch": 0.9087368421052632, + "grad_norm": 0.478515625, + "learning_rate": 2.0280701754385967e-05, + "loss": 2.7504, + "step": 17266 + }, + { + "epoch": 0.9087894736842105, + "grad_norm": 0.5078125, + "learning_rate": 2.0269005847953217e-05, + "loss": 3.3381, + "step": 17267 + }, + { + "epoch": 0.9088421052631579, + "grad_norm": 0.51953125, + "learning_rate": 2.025730994152047e-05, + "loss": 3.4813, + "step": 17268 + }, + { + "epoch": 0.9088947368421053, + "grad_norm": 0.50390625, + "learning_rate": 2.024561403508772e-05, + "loss": 3.2899, + "step": 17269 + }, + { + "epoch": 0.9089473684210526, + "grad_norm": 0.5, + "learning_rate": 2.0233918128654973e-05, + "loss": 3.4209, + "step": 17270 + }, + { + "epoch": 0.909, + "grad_norm": 0.51171875, + "learning_rate": 2.0222222222222222e-05, + "loss": 3.4993, + "step": 17271 + }, + { + "epoch": 0.9090526315789473, + "grad_norm": 0.486328125, + "learning_rate": 2.0210526315789475e-05, + "loss": 3.1126, + "step": 17272 + }, + { + "epoch": 0.9091052631578947, + "grad_norm": 0.5078125, + "learning_rate": 2.0198830409356725e-05, + "loss": 3.1875, + "step": 17273 + }, + { + "epoch": 0.909157894736842, + "grad_norm": 0.5, + "learning_rate": 2.0187134502923978e-05, + "loss": 3.2557, + "step": 17274 + }, + { + "epoch": 0.9092105263157895, + "grad_norm": 0.50390625, + "learning_rate": 2.0175438596491227e-05, + "loss": 3.2763, + "step": 17275 + }, + { + "epoch": 0.9092631578947369, + "grad_norm": 0.4921875, + "learning_rate": 2.016374269005848e-05, + "loss": 3.0448, + "step": 17276 + }, + { + "epoch": 0.9093157894736842, + "grad_norm": 0.482421875, + "learning_rate": 2.0152046783625733e-05, + "loss": 3.1232, + "step": 17277 + }, + { + "epoch": 0.9093684210526316, + "grad_norm": 0.51171875, + "learning_rate": 2.0140350877192983e-05, + "loss": 3.2821, + "step": 17278 + }, + { + "epoch": 0.9094210526315789, + "grad_norm": 0.515625, + "learning_rate": 2.0128654970760236e-05, + "loss": 3.4254, + "step": 17279 + }, + { + "epoch": 0.9094736842105263, + "grad_norm": 0.484375, + "learning_rate": 2.0116959064327485e-05, + "loss": 2.7702, + "step": 17280 + }, + { + "epoch": 0.9095263157894737, + "grad_norm": 0.5, + "learning_rate": 2.010526315789474e-05, + "loss": 3.4558, + "step": 17281 + }, + { + "epoch": 0.909578947368421, + "grad_norm": 0.52734375, + "learning_rate": 2.0093567251461988e-05, + "loss": 3.1813, + "step": 17282 + }, + { + "epoch": 0.9096315789473685, + "grad_norm": 0.482421875, + "learning_rate": 2.008187134502924e-05, + "loss": 3.3281, + "step": 17283 + }, + { + "epoch": 0.9096842105263158, + "grad_norm": 0.5, + "learning_rate": 2.007017543859649e-05, + "loss": 3.5701, + "step": 17284 + }, + { + "epoch": 0.9097368421052632, + "grad_norm": 0.49609375, + "learning_rate": 2.0058479532163744e-05, + "loss": 3.3437, + "step": 17285 + }, + { + "epoch": 0.9097894736842105, + "grad_norm": 0.49609375, + "learning_rate": 2.0046783625730993e-05, + "loss": 3.6254, + "step": 17286 + }, + { + "epoch": 0.9098421052631579, + "grad_norm": 0.515625, + "learning_rate": 2.0035087719298246e-05, + "loss": 3.3682, + "step": 17287 + }, + { + "epoch": 0.9098947368421053, + "grad_norm": 0.5, + "learning_rate": 2.0023391812865496e-05, + "loss": 3.0794, + "step": 17288 + }, + { + "epoch": 0.9099473684210526, + "grad_norm": 0.51171875, + "learning_rate": 2.0011695906432752e-05, + "loss": 3.5461, + "step": 17289 + }, + { + "epoch": 0.91, + "grad_norm": 0.4609375, + "learning_rate": 2e-05, + "loss": 3.2123, + "step": 17290 + }, + { + "epoch": 0.9100526315789473, + "grad_norm": 0.48828125, + "learning_rate": 1.9988304093567255e-05, + "loss": 2.9516, + "step": 17291 + }, + { + "epoch": 0.9101052631578948, + "grad_norm": 0.50390625, + "learning_rate": 1.9976608187134504e-05, + "loss": 2.9637, + "step": 17292 + }, + { + "epoch": 0.910157894736842, + "grad_norm": 0.515625, + "learning_rate": 1.9964912280701757e-05, + "loss": 3.2019, + "step": 17293 + }, + { + "epoch": 0.9102105263157895, + "grad_norm": 0.50390625, + "learning_rate": 1.9953216374269007e-05, + "loss": 3.2137, + "step": 17294 + }, + { + "epoch": 0.9102631578947369, + "grad_norm": 0.484375, + "learning_rate": 1.994152046783626e-05, + "loss": 3.3727, + "step": 17295 + }, + { + "epoch": 0.9103157894736842, + "grad_norm": 0.5078125, + "learning_rate": 1.992982456140351e-05, + "loss": 3.0627, + "step": 17296 + }, + { + "epoch": 0.9103684210526316, + "grad_norm": 0.51953125, + "learning_rate": 1.9918128654970762e-05, + "loss": 3.3133, + "step": 17297 + }, + { + "epoch": 0.9104210526315789, + "grad_norm": 0.490234375, + "learning_rate": 1.9906432748538012e-05, + "loss": 2.816, + "step": 17298 + }, + { + "epoch": 0.9104736842105263, + "grad_norm": 0.53125, + "learning_rate": 1.9894736842105265e-05, + "loss": 2.9905, + "step": 17299 + }, + { + "epoch": 0.9105263157894737, + "grad_norm": 0.5, + "learning_rate": 1.9883040935672515e-05, + "loss": 3.2197, + "step": 17300 + }, + { + "epoch": 0.910578947368421, + "grad_norm": 0.52734375, + "learning_rate": 1.9871345029239764e-05, + "loss": 3.3793, + "step": 17301 + }, + { + "epoch": 0.9106315789473685, + "grad_norm": 0.5703125, + "learning_rate": 1.985964912280702e-05, + "loss": 3.1525, + "step": 17302 + }, + { + "epoch": 0.9106842105263158, + "grad_norm": 0.45703125, + "learning_rate": 1.984795321637427e-05, + "loss": 3.0405, + "step": 17303 + }, + { + "epoch": 0.9107368421052632, + "grad_norm": 0.49609375, + "learning_rate": 1.9836257309941523e-05, + "loss": 3.4977, + "step": 17304 + }, + { + "epoch": 0.9107894736842105, + "grad_norm": 0.5546875, + "learning_rate": 1.9824561403508773e-05, + "loss": 3.0848, + "step": 17305 + }, + { + "epoch": 0.9108421052631579, + "grad_norm": 0.50390625, + "learning_rate": 1.9812865497076026e-05, + "loss": 3.4493, + "step": 17306 + }, + { + "epoch": 0.9108947368421053, + "grad_norm": 0.498046875, + "learning_rate": 1.9801169590643275e-05, + "loss": 3.2288, + "step": 17307 + }, + { + "epoch": 0.9109473684210526, + "grad_norm": 0.515625, + "learning_rate": 1.9789473684210528e-05, + "loss": 3.2825, + "step": 17308 + }, + { + "epoch": 0.911, + "grad_norm": 0.498046875, + "learning_rate": 1.9777777777777778e-05, + "loss": 2.9664, + "step": 17309 + }, + { + "epoch": 0.9110526315789473, + "grad_norm": 0.48046875, + "learning_rate": 1.976608187134503e-05, + "loss": 3.1803, + "step": 17310 + }, + { + "epoch": 0.9111052631578948, + "grad_norm": 0.494140625, + "learning_rate": 1.975438596491228e-05, + "loss": 3.2285, + "step": 17311 + }, + { + "epoch": 0.9111578947368421, + "grad_norm": 0.470703125, + "learning_rate": 1.9742690058479533e-05, + "loss": 3.2782, + "step": 17312 + }, + { + "epoch": 0.9112105263157895, + "grad_norm": 0.484375, + "learning_rate": 1.9730994152046783e-05, + "loss": 3.5262, + "step": 17313 + }, + { + "epoch": 0.9112631578947369, + "grad_norm": 0.484375, + "learning_rate": 1.9719298245614036e-05, + "loss": 3.0676, + "step": 17314 + }, + { + "epoch": 0.9113157894736842, + "grad_norm": 0.50390625, + "learning_rate": 1.970760233918129e-05, + "loss": 3.1035, + "step": 17315 + }, + { + "epoch": 0.9113684210526316, + "grad_norm": 0.5078125, + "learning_rate": 1.969590643274854e-05, + "loss": 3.6052, + "step": 17316 + }, + { + "epoch": 0.9114210526315789, + "grad_norm": 0.4921875, + "learning_rate": 1.968421052631579e-05, + "loss": 3.5782, + "step": 17317 + }, + { + "epoch": 0.9114736842105263, + "grad_norm": 0.474609375, + "learning_rate": 1.967251461988304e-05, + "loss": 3.5077, + "step": 17318 + }, + { + "epoch": 0.9115263157894736, + "grad_norm": 0.52734375, + "learning_rate": 1.9660818713450294e-05, + "loss": 3.3617, + "step": 17319 + }, + { + "epoch": 0.911578947368421, + "grad_norm": 0.515625, + "learning_rate": 1.9649122807017544e-05, + "loss": 3.4703, + "step": 17320 + }, + { + "epoch": 0.9116315789473685, + "grad_norm": 0.5078125, + "learning_rate": 1.9637426900584797e-05, + "loss": 2.9331, + "step": 17321 + }, + { + "epoch": 0.9116842105263158, + "grad_norm": 0.52734375, + "learning_rate": 1.9625730994152046e-05, + "loss": 3.1847, + "step": 17322 + }, + { + "epoch": 0.9117368421052632, + "grad_norm": 0.515625, + "learning_rate": 1.96140350877193e-05, + "loss": 3.3881, + "step": 17323 + }, + { + "epoch": 0.9117894736842105, + "grad_norm": 0.5078125, + "learning_rate": 1.960233918128655e-05, + "loss": 3.5865, + "step": 17324 + }, + { + "epoch": 0.9118421052631579, + "grad_norm": 0.498046875, + "learning_rate": 1.9590643274853802e-05, + "loss": 3.1188, + "step": 17325 + }, + { + "epoch": 0.9118947368421053, + "grad_norm": 0.5078125, + "learning_rate": 1.957894736842105e-05, + "loss": 3.2616, + "step": 17326 + }, + { + "epoch": 0.9119473684210526, + "grad_norm": 0.478515625, + "learning_rate": 1.9567251461988308e-05, + "loss": 3.1627, + "step": 17327 + }, + { + "epoch": 0.912, + "grad_norm": 0.4921875, + "learning_rate": 1.9555555555555557e-05, + "loss": 3.063, + "step": 17328 + }, + { + "epoch": 0.9120526315789473, + "grad_norm": 0.51953125, + "learning_rate": 1.954385964912281e-05, + "loss": 3.1231, + "step": 17329 + }, + { + "epoch": 0.9121052631578948, + "grad_norm": 0.482421875, + "learning_rate": 1.953216374269006e-05, + "loss": 3.1484, + "step": 17330 + }, + { + "epoch": 0.9121578947368421, + "grad_norm": 0.5234375, + "learning_rate": 1.9520467836257313e-05, + "loss": 3.0444, + "step": 17331 + }, + { + "epoch": 0.9122105263157895, + "grad_norm": 0.51171875, + "learning_rate": 1.9508771929824562e-05, + "loss": 2.8556, + "step": 17332 + }, + { + "epoch": 0.9122631578947369, + "grad_norm": 0.51171875, + "learning_rate": 1.9497076023391812e-05, + "loss": 3.4055, + "step": 17333 + }, + { + "epoch": 0.9123157894736842, + "grad_norm": 0.486328125, + "learning_rate": 1.9485380116959065e-05, + "loss": 2.925, + "step": 17334 + }, + { + "epoch": 0.9123684210526316, + "grad_norm": 0.490234375, + "learning_rate": 1.9473684210526315e-05, + "loss": 3.4607, + "step": 17335 + }, + { + "epoch": 0.9124210526315789, + "grad_norm": 0.48046875, + "learning_rate": 1.9461988304093568e-05, + "loss": 3.4492, + "step": 17336 + }, + { + "epoch": 0.9124736842105263, + "grad_norm": 0.484375, + "learning_rate": 1.9450292397660817e-05, + "loss": 3.5131, + "step": 17337 + }, + { + "epoch": 0.9125263157894736, + "grad_norm": 0.498046875, + "learning_rate": 1.943859649122807e-05, + "loss": 3.0988, + "step": 17338 + }, + { + "epoch": 0.912578947368421, + "grad_norm": 0.482421875, + "learning_rate": 1.9426900584795323e-05, + "loss": 3.3968, + "step": 17339 + }, + { + "epoch": 0.9126315789473685, + "grad_norm": 0.494140625, + "learning_rate": 1.9415204678362576e-05, + "loss": 2.5564, + "step": 17340 + }, + { + "epoch": 0.9126842105263158, + "grad_norm": 0.490234375, + "learning_rate": 1.9403508771929826e-05, + "loss": 3.0763, + "step": 17341 + }, + { + "epoch": 0.9127368421052632, + "grad_norm": 0.5234375, + "learning_rate": 1.939181286549708e-05, + "loss": 3.06, + "step": 17342 + }, + { + "epoch": 0.9127894736842105, + "grad_norm": 0.515625, + "learning_rate": 1.9380116959064328e-05, + "loss": 3.2611, + "step": 17343 + }, + { + "epoch": 0.9128421052631579, + "grad_norm": 0.49609375, + "learning_rate": 1.936842105263158e-05, + "loss": 3.7999, + "step": 17344 + }, + { + "epoch": 0.9128947368421053, + "grad_norm": 0.5234375, + "learning_rate": 1.935672514619883e-05, + "loss": 3.5493, + "step": 17345 + }, + { + "epoch": 0.9129473684210526, + "grad_norm": 0.515625, + "learning_rate": 1.9345029239766084e-05, + "loss": 3.5604, + "step": 17346 + }, + { + "epoch": 0.913, + "grad_norm": 0.51171875, + "learning_rate": 1.9333333333333333e-05, + "loss": 3.2136, + "step": 17347 + }, + { + "epoch": 0.9130526315789473, + "grad_norm": 0.478515625, + "learning_rate": 1.9321637426900586e-05, + "loss": 3.3954, + "step": 17348 + }, + { + "epoch": 0.9131052631578948, + "grad_norm": 0.49609375, + "learning_rate": 1.9309941520467836e-05, + "loss": 3.1757, + "step": 17349 + }, + { + "epoch": 0.9131578947368421, + "grad_norm": 0.51171875, + "learning_rate": 1.929824561403509e-05, + "loss": 3.4398, + "step": 17350 + }, + { + "epoch": 0.9132105263157895, + "grad_norm": 0.53125, + "learning_rate": 1.928654970760234e-05, + "loss": 3.4554, + "step": 17351 + }, + { + "epoch": 0.9132631578947369, + "grad_norm": 0.490234375, + "learning_rate": 1.927485380116959e-05, + "loss": 3.4732, + "step": 17352 + }, + { + "epoch": 0.9133157894736842, + "grad_norm": 0.5078125, + "learning_rate": 1.9263157894736845e-05, + "loss": 3.421, + "step": 17353 + }, + { + "epoch": 0.9133684210526316, + "grad_norm": 0.484375, + "learning_rate": 1.9251461988304094e-05, + "loss": 3.1374, + "step": 17354 + }, + { + "epoch": 0.9134210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.9239766081871347e-05, + "loss": 3.3118, + "step": 17355 + }, + { + "epoch": 0.9134736842105263, + "grad_norm": 0.515625, + "learning_rate": 1.9228070175438597e-05, + "loss": 3.0891, + "step": 17356 + }, + { + "epoch": 0.9135263157894736, + "grad_norm": 0.5078125, + "learning_rate": 1.921637426900585e-05, + "loss": 3.2546, + "step": 17357 + }, + { + "epoch": 0.913578947368421, + "grad_norm": 0.51953125, + "learning_rate": 1.92046783625731e-05, + "loss": 3.5536, + "step": 17358 + }, + { + "epoch": 0.9136315789473685, + "grad_norm": 0.5078125, + "learning_rate": 1.9192982456140352e-05, + "loss": 3.3566, + "step": 17359 + }, + { + "epoch": 0.9136842105263158, + "grad_norm": 0.49609375, + "learning_rate": 1.9181286549707602e-05, + "loss": 3.2069, + "step": 17360 + }, + { + "epoch": 0.9137368421052632, + "grad_norm": 0.5078125, + "learning_rate": 1.9169590643274855e-05, + "loss": 3.4829, + "step": 17361 + }, + { + "epoch": 0.9137894736842105, + "grad_norm": 0.52734375, + "learning_rate": 1.9157894736842104e-05, + "loss": 3.6108, + "step": 17362 + }, + { + "epoch": 0.9138421052631579, + "grad_norm": 0.73046875, + "learning_rate": 1.9146198830409357e-05, + "loss": 3.0637, + "step": 17363 + }, + { + "epoch": 0.9138947368421053, + "grad_norm": 0.51171875, + "learning_rate": 1.913450292397661e-05, + "loss": 3.4356, + "step": 17364 + }, + { + "epoch": 0.9139473684210526, + "grad_norm": 0.58203125, + "learning_rate": 1.9122807017543863e-05, + "loss": 3.322, + "step": 17365 + }, + { + "epoch": 0.914, + "grad_norm": 0.53125, + "learning_rate": 1.9111111111111113e-05, + "loss": 2.8955, + "step": 17366 + }, + { + "epoch": 0.9140526315789473, + "grad_norm": 0.5234375, + "learning_rate": 1.9099415204678363e-05, + "loss": 3.4544, + "step": 17367 + }, + { + "epoch": 0.9141052631578948, + "grad_norm": 0.50390625, + "learning_rate": 1.9087719298245616e-05, + "loss": 3.4024, + "step": 17368 + }, + { + "epoch": 0.9141578947368421, + "grad_norm": 0.49609375, + "learning_rate": 1.9076023391812865e-05, + "loss": 3.4951, + "step": 17369 + }, + { + "epoch": 0.9142105263157895, + "grad_norm": 0.478515625, + "learning_rate": 1.9064327485380118e-05, + "loss": 2.9471, + "step": 17370 + }, + { + "epoch": 0.9142631578947369, + "grad_norm": 0.51171875, + "learning_rate": 1.9052631578947368e-05, + "loss": 2.9642, + "step": 17371 + }, + { + "epoch": 0.9143157894736842, + "grad_norm": 0.498046875, + "learning_rate": 1.904093567251462e-05, + "loss": 2.83, + "step": 17372 + }, + { + "epoch": 0.9143684210526316, + "grad_norm": 0.5078125, + "learning_rate": 1.902923976608187e-05, + "loss": 3.5092, + "step": 17373 + }, + { + "epoch": 0.9144210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.9017543859649123e-05, + "loss": 3.6537, + "step": 17374 + }, + { + "epoch": 0.9144736842105263, + "grad_norm": 0.498046875, + "learning_rate": 1.9005847953216373e-05, + "loss": 3.4147, + "step": 17375 + }, + { + "epoch": 0.9145263157894736, + "grad_norm": 0.5546875, + "learning_rate": 1.8994152046783626e-05, + "loss": 3.435, + "step": 17376 + }, + { + "epoch": 0.914578947368421, + "grad_norm": 0.48828125, + "learning_rate": 1.898245614035088e-05, + "loss": 3.6505, + "step": 17377 + }, + { + "epoch": 0.9146315789473685, + "grad_norm": 0.51953125, + "learning_rate": 1.8970760233918132e-05, + "loss": 3.5453, + "step": 17378 + }, + { + "epoch": 0.9146842105263158, + "grad_norm": 0.5234375, + "learning_rate": 1.895906432748538e-05, + "loss": 3.0712, + "step": 17379 + }, + { + "epoch": 0.9147368421052632, + "grad_norm": 0.51953125, + "learning_rate": 1.8947368421052634e-05, + "loss": 2.6559, + "step": 17380 + }, + { + "epoch": 0.9147894736842105, + "grad_norm": 0.51171875, + "learning_rate": 1.8935672514619884e-05, + "loss": 3.3528, + "step": 17381 + }, + { + "epoch": 0.9148421052631579, + "grad_norm": 0.54296875, + "learning_rate": 1.8923976608187137e-05, + "loss": 3.5738, + "step": 17382 + }, + { + "epoch": 0.9148947368421053, + "grad_norm": 0.52734375, + "learning_rate": 1.8912280701754387e-05, + "loss": 3.4099, + "step": 17383 + }, + { + "epoch": 0.9149473684210526, + "grad_norm": 0.5390625, + "learning_rate": 1.890058479532164e-05, + "loss": 3.145, + "step": 17384 + }, + { + "epoch": 0.915, + "grad_norm": 0.51953125, + "learning_rate": 1.888888888888889e-05, + "loss": 2.9412, + "step": 17385 + }, + { + "epoch": 0.9150526315789473, + "grad_norm": 0.48828125, + "learning_rate": 1.8877192982456142e-05, + "loss": 3.0757, + "step": 17386 + }, + { + "epoch": 0.9151052631578948, + "grad_norm": 0.515625, + "learning_rate": 1.886549707602339e-05, + "loss": 3.2821, + "step": 17387 + }, + { + "epoch": 0.9151578947368421, + "grad_norm": 0.5, + "learning_rate": 1.8853801169590645e-05, + "loss": 2.8606, + "step": 17388 + }, + { + "epoch": 0.9152105263157895, + "grad_norm": 0.5234375, + "learning_rate": 1.8842105263157894e-05, + "loss": 3.3978, + "step": 17389 + }, + { + "epoch": 0.9152631578947369, + "grad_norm": 0.578125, + "learning_rate": 1.8830409356725147e-05, + "loss": 3.3549, + "step": 17390 + }, + { + "epoch": 0.9153157894736842, + "grad_norm": 0.51171875, + "learning_rate": 1.88187134502924e-05, + "loss": 3.2728, + "step": 17391 + }, + { + "epoch": 0.9153684210526316, + "grad_norm": 0.49609375, + "learning_rate": 1.880701754385965e-05, + "loss": 3.339, + "step": 17392 + }, + { + "epoch": 0.9154210526315789, + "grad_norm": 0.46875, + "learning_rate": 1.8795321637426903e-05, + "loss": 2.8997, + "step": 17393 + }, + { + "epoch": 0.9154736842105263, + "grad_norm": 0.5, + "learning_rate": 1.8783625730994152e-05, + "loss": 3.3336, + "step": 17394 + }, + { + "epoch": 0.9155263157894736, + "grad_norm": 0.486328125, + "learning_rate": 1.8771929824561405e-05, + "loss": 2.9677, + "step": 17395 + }, + { + "epoch": 0.915578947368421, + "grad_norm": 0.56640625, + "learning_rate": 1.8760233918128655e-05, + "loss": 3.0119, + "step": 17396 + }, + { + "epoch": 0.9156315789473685, + "grad_norm": 0.494140625, + "learning_rate": 1.8748538011695908e-05, + "loss": 3.2025, + "step": 17397 + }, + { + "epoch": 0.9156842105263158, + "grad_norm": 0.5546875, + "learning_rate": 1.8736842105263158e-05, + "loss": 3.0745, + "step": 17398 + }, + { + "epoch": 0.9157368421052632, + "grad_norm": 0.482421875, + "learning_rate": 1.872514619883041e-05, + "loss": 3.1197, + "step": 17399 + }, + { + "epoch": 0.9157894736842105, + "grad_norm": 0.5078125, + "learning_rate": 1.871345029239766e-05, + "loss": 3.7139, + "step": 17400 + }, + { + "epoch": 0.9158421052631579, + "grad_norm": 0.478515625, + "learning_rate": 1.8701754385964913e-05, + "loss": 2.7775, + "step": 17401 + }, + { + "epoch": 0.9158947368421053, + "grad_norm": 0.5234375, + "learning_rate": 1.8690058479532166e-05, + "loss": 3.6689, + "step": 17402 + }, + { + "epoch": 0.9159473684210526, + "grad_norm": 0.51171875, + "learning_rate": 1.8678362573099416e-05, + "loss": 3.1492, + "step": 17403 + }, + { + "epoch": 0.916, + "grad_norm": 0.51171875, + "learning_rate": 1.866666666666667e-05, + "loss": 3.354, + "step": 17404 + }, + { + "epoch": 0.9160526315789473, + "grad_norm": 0.5234375, + "learning_rate": 1.8654970760233918e-05, + "loss": 3.6461, + "step": 17405 + }, + { + "epoch": 0.9161052631578948, + "grad_norm": 0.48046875, + "learning_rate": 1.864327485380117e-05, + "loss": 2.9427, + "step": 17406 + }, + { + "epoch": 0.9161578947368421, + "grad_norm": 0.49609375, + "learning_rate": 1.863157894736842e-05, + "loss": 3.0802, + "step": 17407 + }, + { + "epoch": 0.9162105263157895, + "grad_norm": 0.490234375, + "learning_rate": 1.8619883040935674e-05, + "loss": 3.4369, + "step": 17408 + }, + { + "epoch": 0.9162631578947369, + "grad_norm": 0.51171875, + "learning_rate": 1.8608187134502923e-05, + "loss": 3.4962, + "step": 17409 + }, + { + "epoch": 0.9163157894736842, + "grad_norm": 0.5234375, + "learning_rate": 1.8596491228070176e-05, + "loss": 3.5732, + "step": 17410 + }, + { + "epoch": 0.9163684210526316, + "grad_norm": 0.55078125, + "learning_rate": 1.8584795321637426e-05, + "loss": 3.1555, + "step": 17411 + }, + { + "epoch": 0.9164210526315789, + "grad_norm": 0.478515625, + "learning_rate": 1.857309941520468e-05, + "loss": 3.7227, + "step": 17412 + }, + { + "epoch": 0.9164736842105263, + "grad_norm": 0.482421875, + "learning_rate": 1.856140350877193e-05, + "loss": 3.1609, + "step": 17413 + }, + { + "epoch": 0.9165263157894736, + "grad_norm": 0.4921875, + "learning_rate": 1.854970760233918e-05, + "loss": 3.0743, + "step": 17414 + }, + { + "epoch": 0.916578947368421, + "grad_norm": 0.484375, + "learning_rate": 1.8538011695906434e-05, + "loss": 3.0995, + "step": 17415 + }, + { + "epoch": 0.9166315789473685, + "grad_norm": 0.51171875, + "learning_rate": 1.8526315789473687e-05, + "loss": 3.5241, + "step": 17416 + }, + { + "epoch": 0.9166842105263158, + "grad_norm": 0.54296875, + "learning_rate": 1.8514619883040937e-05, + "loss": 3.5963, + "step": 17417 + }, + { + "epoch": 0.9167368421052632, + "grad_norm": 0.482421875, + "learning_rate": 1.850292397660819e-05, + "loss": 3.2212, + "step": 17418 + }, + { + "epoch": 0.9167894736842105, + "grad_norm": 0.51953125, + "learning_rate": 1.849122807017544e-05, + "loss": 3.2962, + "step": 17419 + }, + { + "epoch": 0.9168421052631579, + "grad_norm": 0.484375, + "learning_rate": 1.8479532163742693e-05, + "loss": 3.3471, + "step": 17420 + }, + { + "epoch": 0.9168947368421053, + "grad_norm": 0.51171875, + "learning_rate": 1.8467836257309942e-05, + "loss": 3.377, + "step": 17421 + }, + { + "epoch": 0.9169473684210526, + "grad_norm": 0.5, + "learning_rate": 1.8456140350877195e-05, + "loss": 3.5775, + "step": 17422 + }, + { + "epoch": 0.917, + "grad_norm": 0.53125, + "learning_rate": 1.8444444444444445e-05, + "loss": 3.1492, + "step": 17423 + }, + { + "epoch": 0.9170526315789473, + "grad_norm": 0.494140625, + "learning_rate": 1.8432748538011694e-05, + "loss": 3.5464, + "step": 17424 + }, + { + "epoch": 0.9171052631578948, + "grad_norm": 0.5, + "learning_rate": 1.8421052631578947e-05, + "loss": 3.121, + "step": 17425 + }, + { + "epoch": 0.9171578947368421, + "grad_norm": 0.50390625, + "learning_rate": 1.8409356725146197e-05, + "loss": 3.0043, + "step": 17426 + }, + { + "epoch": 0.9172105263157895, + "grad_norm": 0.51171875, + "learning_rate": 1.8397660818713453e-05, + "loss": 2.9687, + "step": 17427 + }, + { + "epoch": 0.9172631578947369, + "grad_norm": 0.5, + "learning_rate": 1.8385964912280703e-05, + "loss": 3.382, + "step": 17428 + }, + { + "epoch": 0.9173157894736842, + "grad_norm": 0.494140625, + "learning_rate": 1.8374269005847956e-05, + "loss": 3.6084, + "step": 17429 + }, + { + "epoch": 0.9173684210526316, + "grad_norm": 0.52734375, + "learning_rate": 1.8362573099415205e-05, + "loss": 3.4596, + "step": 17430 + }, + { + "epoch": 0.9174210526315789, + "grad_norm": 0.4921875, + "learning_rate": 1.835087719298246e-05, + "loss": 3.2467, + "step": 17431 + }, + { + "epoch": 0.9174736842105263, + "grad_norm": 0.5234375, + "learning_rate": 1.8339181286549708e-05, + "loss": 3.3702, + "step": 17432 + }, + { + "epoch": 0.9175263157894736, + "grad_norm": 0.482421875, + "learning_rate": 1.832748538011696e-05, + "loss": 3.0411, + "step": 17433 + }, + { + "epoch": 0.917578947368421, + "grad_norm": 0.4921875, + "learning_rate": 1.831578947368421e-05, + "loss": 3.3608, + "step": 17434 + }, + { + "epoch": 0.9176315789473685, + "grad_norm": 0.515625, + "learning_rate": 1.8304093567251464e-05, + "loss": 3.1292, + "step": 17435 + }, + { + "epoch": 0.9176842105263158, + "grad_norm": 0.50390625, + "learning_rate": 1.8292397660818713e-05, + "loss": 3.1488, + "step": 17436 + }, + { + "epoch": 0.9177368421052632, + "grad_norm": 0.5234375, + "learning_rate": 1.8280701754385966e-05, + "loss": 3.2375, + "step": 17437 + }, + { + "epoch": 0.9177894736842105, + "grad_norm": 0.5078125, + "learning_rate": 1.8269005847953216e-05, + "loss": 3.2437, + "step": 17438 + }, + { + "epoch": 0.9178421052631579, + "grad_norm": 0.5234375, + "learning_rate": 1.825730994152047e-05, + "loss": 3.0773, + "step": 17439 + }, + { + "epoch": 0.9178947368421052, + "grad_norm": 0.49609375, + "learning_rate": 1.8245614035087722e-05, + "loss": 3.4403, + "step": 17440 + }, + { + "epoch": 0.9179473684210526, + "grad_norm": 0.53125, + "learning_rate": 1.823391812865497e-05, + "loss": 3.1157, + "step": 17441 + }, + { + "epoch": 0.918, + "grad_norm": 0.51171875, + "learning_rate": 1.8222222222222224e-05, + "loss": 3.286, + "step": 17442 + }, + { + "epoch": 0.9180526315789473, + "grad_norm": 0.52734375, + "learning_rate": 1.8210526315789474e-05, + "loss": 3.1218, + "step": 17443 + }, + { + "epoch": 0.9181052631578948, + "grad_norm": 0.49609375, + "learning_rate": 1.8198830409356727e-05, + "loss": 3.2315, + "step": 17444 + }, + { + "epoch": 0.9181578947368421, + "grad_norm": 0.50390625, + "learning_rate": 1.8187134502923976e-05, + "loss": 3.2709, + "step": 17445 + }, + { + "epoch": 0.9182105263157895, + "grad_norm": 0.50390625, + "learning_rate": 1.817543859649123e-05, + "loss": 3.7179, + "step": 17446 + }, + { + "epoch": 0.9182631578947369, + "grad_norm": 0.5234375, + "learning_rate": 1.816374269005848e-05, + "loss": 3.2224, + "step": 17447 + }, + { + "epoch": 0.9183157894736842, + "grad_norm": 0.50390625, + "learning_rate": 1.8152046783625732e-05, + "loss": 3.5238, + "step": 17448 + }, + { + "epoch": 0.9183684210526316, + "grad_norm": 0.5546875, + "learning_rate": 1.814035087719298e-05, + "loss": 3.4274, + "step": 17449 + }, + { + "epoch": 0.9184210526315789, + "grad_norm": 0.51953125, + "learning_rate": 1.8128654970760235e-05, + "loss": 2.9817, + "step": 17450 + }, + { + "epoch": 0.9184736842105263, + "grad_norm": 0.48828125, + "learning_rate": 1.8116959064327484e-05, + "loss": 2.7442, + "step": 17451 + }, + { + "epoch": 0.9185263157894736, + "grad_norm": 0.490234375, + "learning_rate": 1.810526315789474e-05, + "loss": 3.3568, + "step": 17452 + }, + { + "epoch": 0.918578947368421, + "grad_norm": 0.498046875, + "learning_rate": 1.809356725146199e-05, + "loss": 3.1104, + "step": 17453 + }, + { + "epoch": 0.9186315789473685, + "grad_norm": 0.498046875, + "learning_rate": 1.8081871345029243e-05, + "loss": 3.3874, + "step": 17454 + }, + { + "epoch": 0.9186842105263158, + "grad_norm": 0.490234375, + "learning_rate": 1.8070175438596493e-05, + "loss": 3.3124, + "step": 17455 + }, + { + "epoch": 0.9187368421052632, + "grad_norm": 0.490234375, + "learning_rate": 1.8058479532163742e-05, + "loss": 3.4779, + "step": 17456 + }, + { + "epoch": 0.9187894736842105, + "grad_norm": 0.515625, + "learning_rate": 1.8046783625730995e-05, + "loss": 3.4284, + "step": 17457 + }, + { + "epoch": 0.9188421052631579, + "grad_norm": 0.5078125, + "learning_rate": 1.8035087719298245e-05, + "loss": 3.3666, + "step": 17458 + }, + { + "epoch": 0.9188947368421052, + "grad_norm": 0.51953125, + "learning_rate": 1.8023391812865498e-05, + "loss": 3.2393, + "step": 17459 + }, + { + "epoch": 0.9189473684210526, + "grad_norm": 0.4921875, + "learning_rate": 1.8011695906432747e-05, + "loss": 3.7248, + "step": 17460 + }, + { + "epoch": 0.919, + "grad_norm": 0.51171875, + "learning_rate": 1.8e-05, + "loss": 3.0557, + "step": 17461 + }, + { + "epoch": 0.9190526315789473, + "grad_norm": 0.53125, + "learning_rate": 1.798830409356725e-05, + "loss": 3.5882, + "step": 17462 + }, + { + "epoch": 0.9191052631578948, + "grad_norm": 0.5, + "learning_rate": 1.7976608187134503e-05, + "loss": 3.4648, + "step": 17463 + }, + { + "epoch": 0.9191578947368421, + "grad_norm": 0.48828125, + "learning_rate": 1.7964912280701753e-05, + "loss": 2.9915, + "step": 17464 + }, + { + "epoch": 0.9192105263157895, + "grad_norm": 0.466796875, + "learning_rate": 1.795321637426901e-05, + "loss": 3.287, + "step": 17465 + }, + { + "epoch": 0.9192631578947369, + "grad_norm": 0.490234375, + "learning_rate": 1.794152046783626e-05, + "loss": 3.4706, + "step": 17466 + }, + { + "epoch": 0.9193157894736842, + "grad_norm": 0.51953125, + "learning_rate": 1.792982456140351e-05, + "loss": 3.3555, + "step": 17467 + }, + { + "epoch": 0.9193684210526316, + "grad_norm": 0.50390625, + "learning_rate": 1.791812865497076e-05, + "loss": 3.441, + "step": 17468 + }, + { + "epoch": 0.9194210526315789, + "grad_norm": 0.546875, + "learning_rate": 1.7906432748538014e-05, + "loss": 2.9818, + "step": 17469 + }, + { + "epoch": 0.9194736842105263, + "grad_norm": 0.5859375, + "learning_rate": 1.7894736842105264e-05, + "loss": 3.4544, + "step": 17470 + }, + { + "epoch": 0.9195263157894736, + "grad_norm": 0.4921875, + "learning_rate": 1.7883040935672517e-05, + "loss": 3.5033, + "step": 17471 + }, + { + "epoch": 0.919578947368421, + "grad_norm": 0.484375, + "learning_rate": 1.7871345029239766e-05, + "loss": 3.1659, + "step": 17472 + }, + { + "epoch": 0.9196315789473685, + "grad_norm": 0.4921875, + "learning_rate": 1.785964912280702e-05, + "loss": 3.2598, + "step": 17473 + }, + { + "epoch": 0.9196842105263158, + "grad_norm": 0.515625, + "learning_rate": 1.784795321637427e-05, + "loss": 2.9839, + "step": 17474 + }, + { + "epoch": 0.9197368421052632, + "grad_norm": 0.494140625, + "learning_rate": 1.7836257309941522e-05, + "loss": 3.319, + "step": 17475 + }, + { + "epoch": 0.9197894736842105, + "grad_norm": 0.474609375, + "learning_rate": 1.782456140350877e-05, + "loss": 3.1814, + "step": 17476 + }, + { + "epoch": 0.9198421052631579, + "grad_norm": 0.6640625, + "learning_rate": 1.7812865497076024e-05, + "loss": 2.9456, + "step": 17477 + }, + { + "epoch": 0.9198947368421052, + "grad_norm": 0.46875, + "learning_rate": 1.7801169590643277e-05, + "loss": 2.822, + "step": 17478 + }, + { + "epoch": 0.9199473684210526, + "grad_norm": 0.474609375, + "learning_rate": 1.7789473684210527e-05, + "loss": 3.116, + "step": 17479 + }, + { + "epoch": 0.92, + "grad_norm": 0.52734375, + "learning_rate": 1.777777777777778e-05, + "loss": 3.5197, + "step": 17480 + }, + { + "epoch": 0.9200526315789473, + "grad_norm": 0.498046875, + "learning_rate": 1.776608187134503e-05, + "loss": 3.2693, + "step": 17481 + }, + { + "epoch": 0.9201052631578948, + "grad_norm": 0.51171875, + "learning_rate": 1.7754385964912283e-05, + "loss": 3.2115, + "step": 17482 + }, + { + "epoch": 0.9201578947368421, + "grad_norm": 0.494140625, + "learning_rate": 1.7742690058479532e-05, + "loss": 3.4364, + "step": 17483 + }, + { + "epoch": 0.9202105263157895, + "grad_norm": 0.51171875, + "learning_rate": 1.7730994152046785e-05, + "loss": 3.1031, + "step": 17484 + }, + { + "epoch": 0.9202631578947369, + "grad_norm": 0.498046875, + "learning_rate": 1.7719298245614035e-05, + "loss": 3.3098, + "step": 17485 + }, + { + "epoch": 0.9203157894736842, + "grad_norm": 0.4921875, + "learning_rate": 1.7707602339181288e-05, + "loss": 3.3529, + "step": 17486 + }, + { + "epoch": 0.9203684210526316, + "grad_norm": 0.5078125, + "learning_rate": 1.7695906432748537e-05, + "loss": 3.0192, + "step": 17487 + }, + { + "epoch": 0.9204210526315789, + "grad_norm": 0.5078125, + "learning_rate": 1.768421052631579e-05, + "loss": 3.0636, + "step": 17488 + }, + { + "epoch": 0.9204736842105263, + "grad_norm": 0.5078125, + "learning_rate": 1.767251461988304e-05, + "loss": 3.4084, + "step": 17489 + }, + { + "epoch": 0.9205263157894736, + "grad_norm": 0.49609375, + "learning_rate": 1.7660818713450293e-05, + "loss": 3.1332, + "step": 17490 + }, + { + "epoch": 0.920578947368421, + "grad_norm": 0.48046875, + "learning_rate": 1.7649122807017546e-05, + "loss": 3.3877, + "step": 17491 + }, + { + "epoch": 0.9206315789473685, + "grad_norm": 0.49609375, + "learning_rate": 1.7637426900584795e-05, + "loss": 3.5309, + "step": 17492 + }, + { + "epoch": 0.9206842105263158, + "grad_norm": 0.546875, + "learning_rate": 1.762573099415205e-05, + "loss": 3.4109, + "step": 17493 + }, + { + "epoch": 0.9207368421052632, + "grad_norm": 0.515625, + "learning_rate": 1.7614035087719298e-05, + "loss": 3.1111, + "step": 17494 + }, + { + "epoch": 0.9207894736842105, + "grad_norm": 0.5234375, + "learning_rate": 1.760233918128655e-05, + "loss": 3.4889, + "step": 17495 + }, + { + "epoch": 0.9208421052631579, + "grad_norm": 0.51171875, + "learning_rate": 1.75906432748538e-05, + "loss": 3.4728, + "step": 17496 + }, + { + "epoch": 0.9208947368421052, + "grad_norm": 0.482421875, + "learning_rate": 1.7578947368421054e-05, + "loss": 3.098, + "step": 17497 + }, + { + "epoch": 0.9209473684210526, + "grad_norm": 0.490234375, + "learning_rate": 1.7567251461988303e-05, + "loss": 2.8391, + "step": 17498 + }, + { + "epoch": 0.921, + "grad_norm": 0.47265625, + "learning_rate": 1.7555555555555556e-05, + "loss": 3.1034, + "step": 17499 + }, + { + "epoch": 0.9210526315789473, + "grad_norm": 0.51171875, + "learning_rate": 1.7543859649122806e-05, + "loss": 3.3298, + "step": 17500 + }, + { + "epoch": 0.9211052631578948, + "grad_norm": 0.478515625, + "learning_rate": 1.753216374269006e-05, + "loss": 3.3827, + "step": 17501 + }, + { + "epoch": 0.9211578947368421, + "grad_norm": 0.515625, + "learning_rate": 1.752046783625731e-05, + "loss": 3.2799, + "step": 17502 + }, + { + "epoch": 0.9212105263157895, + "grad_norm": 0.49609375, + "learning_rate": 1.7508771929824565e-05, + "loss": 3.6106, + "step": 17503 + }, + { + "epoch": 0.9212631578947369, + "grad_norm": 0.494140625, + "learning_rate": 1.7497076023391814e-05, + "loss": 3.3864, + "step": 17504 + }, + { + "epoch": 0.9213157894736842, + "grad_norm": 0.52734375, + "learning_rate": 1.7485380116959067e-05, + "loss": 2.969, + "step": 17505 + }, + { + "epoch": 0.9213684210526316, + "grad_norm": 0.515625, + "learning_rate": 1.7473684210526317e-05, + "loss": 2.7808, + "step": 17506 + }, + { + "epoch": 0.9214210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.746198830409357e-05, + "loss": 3.1853, + "step": 17507 + }, + { + "epoch": 0.9214736842105263, + "grad_norm": 0.484375, + "learning_rate": 1.745029239766082e-05, + "loss": 3.034, + "step": 17508 + }, + { + "epoch": 0.9215263157894736, + "grad_norm": 0.515625, + "learning_rate": 1.7438596491228072e-05, + "loss": 3.2442, + "step": 17509 + }, + { + "epoch": 0.921578947368421, + "grad_norm": 0.484375, + "learning_rate": 1.7426900584795322e-05, + "loss": 3.3326, + "step": 17510 + }, + { + "epoch": 0.9216315789473685, + "grad_norm": 0.490234375, + "learning_rate": 1.7415204678362575e-05, + "loss": 2.6569, + "step": 17511 + }, + { + "epoch": 0.9216842105263158, + "grad_norm": 0.494140625, + "learning_rate": 1.7403508771929825e-05, + "loss": 3.4036, + "step": 17512 + }, + { + "epoch": 0.9217368421052632, + "grad_norm": 0.48828125, + "learning_rate": 1.7391812865497074e-05, + "loss": 3.2285, + "step": 17513 + }, + { + "epoch": 0.9217894736842105, + "grad_norm": 0.5078125, + "learning_rate": 1.7380116959064327e-05, + "loss": 3.5582, + "step": 17514 + }, + { + "epoch": 0.9218421052631579, + "grad_norm": 0.478515625, + "learning_rate": 1.736842105263158e-05, + "loss": 3.202, + "step": 17515 + }, + { + "epoch": 0.9218947368421052, + "grad_norm": 0.515625, + "learning_rate": 1.7356725146198833e-05, + "loss": 3.4879, + "step": 17516 + }, + { + "epoch": 0.9219473684210526, + "grad_norm": 0.50390625, + "learning_rate": 1.7345029239766083e-05, + "loss": 3.4668, + "step": 17517 + }, + { + "epoch": 0.922, + "grad_norm": 0.498046875, + "learning_rate": 1.7333333333333336e-05, + "loss": 3.1957, + "step": 17518 + }, + { + "epoch": 0.9220526315789473, + "grad_norm": 0.54296875, + "learning_rate": 1.7321637426900585e-05, + "loss": 3.2462, + "step": 17519 + }, + { + "epoch": 0.9221052631578948, + "grad_norm": 0.498046875, + "learning_rate": 1.7309941520467838e-05, + "loss": 3.3554, + "step": 17520 + }, + { + "epoch": 0.9221578947368421, + "grad_norm": 0.484375, + "learning_rate": 1.7298245614035088e-05, + "loss": 3.0985, + "step": 17521 + }, + { + "epoch": 0.9222105263157895, + "grad_norm": 0.515625, + "learning_rate": 1.728654970760234e-05, + "loss": 3.5204, + "step": 17522 + }, + { + "epoch": 0.9222631578947369, + "grad_norm": 0.50390625, + "learning_rate": 1.727485380116959e-05, + "loss": 3.6013, + "step": 17523 + }, + { + "epoch": 0.9223157894736842, + "grad_norm": 0.5078125, + "learning_rate": 1.7263157894736843e-05, + "loss": 3.3642, + "step": 17524 + }, + { + "epoch": 0.9223684210526316, + "grad_norm": 0.51171875, + "learning_rate": 1.7251461988304093e-05, + "loss": 3.3816, + "step": 17525 + }, + { + "epoch": 0.9224210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.7239766081871346e-05, + "loss": 3.3881, + "step": 17526 + }, + { + "epoch": 0.9224736842105263, + "grad_norm": 0.546875, + "learning_rate": 1.72280701754386e-05, + "loss": 3.6431, + "step": 17527 + }, + { + "epoch": 0.9225263157894736, + "grad_norm": 0.52734375, + "learning_rate": 1.721637426900585e-05, + "loss": 3.3042, + "step": 17528 + }, + { + "epoch": 0.922578947368421, + "grad_norm": 0.49609375, + "learning_rate": 1.72046783625731e-05, + "loss": 3.1203, + "step": 17529 + }, + { + "epoch": 0.9226315789473685, + "grad_norm": 0.52734375, + "learning_rate": 1.719298245614035e-05, + "loss": 3.5756, + "step": 17530 + }, + { + "epoch": 0.9226842105263158, + "grad_norm": 0.48828125, + "learning_rate": 1.7181286549707604e-05, + "loss": 3.5187, + "step": 17531 + }, + { + "epoch": 0.9227368421052632, + "grad_norm": 0.484375, + "learning_rate": 1.7169590643274854e-05, + "loss": 3.3357, + "step": 17532 + }, + { + "epoch": 0.9227894736842105, + "grad_norm": 0.515625, + "learning_rate": 1.7157894736842107e-05, + "loss": 3.2308, + "step": 17533 + }, + { + "epoch": 0.9228421052631579, + "grad_norm": 0.5234375, + "learning_rate": 1.7146198830409356e-05, + "loss": 3.4517, + "step": 17534 + }, + { + "epoch": 0.9228947368421052, + "grad_norm": 0.49609375, + "learning_rate": 1.713450292397661e-05, + "loss": 3.568, + "step": 17535 + }, + { + "epoch": 0.9229473684210526, + "grad_norm": 0.515625, + "learning_rate": 1.712280701754386e-05, + "loss": 3.5255, + "step": 17536 + }, + { + "epoch": 0.923, + "grad_norm": 0.486328125, + "learning_rate": 1.7111111111111112e-05, + "loss": 3.079, + "step": 17537 + }, + { + "epoch": 0.9230526315789473, + "grad_norm": 0.490234375, + "learning_rate": 1.709941520467836e-05, + "loss": 3.1671, + "step": 17538 + }, + { + "epoch": 0.9231052631578948, + "grad_norm": 0.50390625, + "learning_rate": 1.7087719298245614e-05, + "loss": 3.418, + "step": 17539 + }, + { + "epoch": 0.9231578947368421, + "grad_norm": 0.46484375, + "learning_rate": 1.7076023391812867e-05, + "loss": 2.6441, + "step": 17540 + }, + { + "epoch": 0.9232105263157895, + "grad_norm": 0.49609375, + "learning_rate": 1.706432748538012e-05, + "loss": 3.3691, + "step": 17541 + }, + { + "epoch": 0.9232631578947369, + "grad_norm": 0.4921875, + "learning_rate": 1.705263157894737e-05, + "loss": 3.2049, + "step": 17542 + }, + { + "epoch": 0.9233157894736842, + "grad_norm": 0.50390625, + "learning_rate": 1.7040935672514623e-05, + "loss": 3.2377, + "step": 17543 + }, + { + "epoch": 0.9233684210526316, + "grad_norm": 0.51171875, + "learning_rate": 1.7029239766081872e-05, + "loss": 3.6058, + "step": 17544 + }, + { + "epoch": 0.9234210526315789, + "grad_norm": 0.515625, + "learning_rate": 1.7017543859649125e-05, + "loss": 2.8958, + "step": 17545 + }, + { + "epoch": 0.9234736842105263, + "grad_norm": 0.48828125, + "learning_rate": 1.7005847953216375e-05, + "loss": 2.8592, + "step": 17546 + }, + { + "epoch": 0.9235263157894736, + "grad_norm": 0.5078125, + "learning_rate": 1.6994152046783625e-05, + "loss": 3.1412, + "step": 17547 + }, + { + "epoch": 0.923578947368421, + "grad_norm": 0.478515625, + "learning_rate": 1.6982456140350878e-05, + "loss": 2.7447, + "step": 17548 + }, + { + "epoch": 0.9236315789473685, + "grad_norm": 0.5, + "learning_rate": 1.6970760233918127e-05, + "loss": 3.2951, + "step": 17549 + }, + { + "epoch": 0.9236842105263158, + "grad_norm": 0.470703125, + "learning_rate": 1.695906432748538e-05, + "loss": 3.2031, + "step": 17550 + }, + { + "epoch": 0.9237368421052632, + "grad_norm": 0.5078125, + "learning_rate": 1.694736842105263e-05, + "loss": 3.706, + "step": 17551 + }, + { + "epoch": 0.9237894736842105, + "grad_norm": 0.515625, + "learning_rate": 1.6935672514619883e-05, + "loss": 3.0037, + "step": 17552 + }, + { + "epoch": 0.9238421052631579, + "grad_norm": 0.50390625, + "learning_rate": 1.6923976608187136e-05, + "loss": 3.184, + "step": 17553 + }, + { + "epoch": 0.9238947368421052, + "grad_norm": 0.54296875, + "learning_rate": 1.691228070175439e-05, + "loss": 3.396, + "step": 17554 + }, + { + "epoch": 0.9239473684210526, + "grad_norm": 0.51953125, + "learning_rate": 1.690058479532164e-05, + "loss": 3.1, + "step": 17555 + }, + { + "epoch": 0.924, + "grad_norm": 0.5234375, + "learning_rate": 1.688888888888889e-05, + "loss": 3.4253, + "step": 17556 + }, + { + "epoch": 0.9240526315789473, + "grad_norm": 0.546875, + "learning_rate": 1.687719298245614e-05, + "loss": 3.1873, + "step": 17557 + }, + { + "epoch": 0.9241052631578948, + "grad_norm": 0.50390625, + "learning_rate": 1.6865497076023394e-05, + "loss": 3.0648, + "step": 17558 + }, + { + "epoch": 0.9241578947368421, + "grad_norm": 0.498046875, + "learning_rate": 1.6853801169590643e-05, + "loss": 3.186, + "step": 17559 + }, + { + "epoch": 0.9242105263157895, + "grad_norm": 0.484375, + "learning_rate": 1.6842105263157896e-05, + "loss": 3.115, + "step": 17560 + }, + { + "epoch": 0.9242631578947369, + "grad_norm": 0.482421875, + "learning_rate": 1.6830409356725146e-05, + "loss": 3.044, + "step": 17561 + }, + { + "epoch": 0.9243157894736842, + "grad_norm": 0.478515625, + "learning_rate": 1.68187134502924e-05, + "loss": 3.3802, + "step": 17562 + }, + { + "epoch": 0.9243684210526316, + "grad_norm": 0.494140625, + "learning_rate": 1.680701754385965e-05, + "loss": 3.1655, + "step": 17563 + }, + { + "epoch": 0.9244210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.67953216374269e-05, + "loss": 3.0891, + "step": 17564 + }, + { + "epoch": 0.9244736842105263, + "grad_norm": 0.46875, + "learning_rate": 1.6783625730994155e-05, + "loss": 3.0019, + "step": 17565 + }, + { + "epoch": 0.9245263157894736, + "grad_norm": 0.50390625, + "learning_rate": 1.6771929824561404e-05, + "loss": 2.9559, + "step": 17566 + }, + { + "epoch": 0.9245789473684211, + "grad_norm": 0.478515625, + "learning_rate": 1.6760233918128657e-05, + "loss": 3.304, + "step": 17567 + }, + { + "epoch": 0.9246315789473685, + "grad_norm": 0.51953125, + "learning_rate": 1.6748538011695907e-05, + "loss": 3.0572, + "step": 17568 + }, + { + "epoch": 0.9246842105263158, + "grad_norm": 0.494140625, + "learning_rate": 1.673684210526316e-05, + "loss": 3.1343, + "step": 17569 + }, + { + "epoch": 0.9247368421052632, + "grad_norm": 0.48046875, + "learning_rate": 1.672514619883041e-05, + "loss": 3.3105, + "step": 17570 + }, + { + "epoch": 0.9247894736842105, + "grad_norm": 0.50390625, + "learning_rate": 1.6713450292397662e-05, + "loss": 3.1774, + "step": 17571 + }, + { + "epoch": 0.9248421052631579, + "grad_norm": 0.51953125, + "learning_rate": 1.6701754385964912e-05, + "loss": 3.0584, + "step": 17572 + }, + { + "epoch": 0.9248947368421052, + "grad_norm": 0.494140625, + "learning_rate": 1.6690058479532165e-05, + "loss": 3.2908, + "step": 17573 + }, + { + "epoch": 0.9249473684210526, + "grad_norm": 0.498046875, + "learning_rate": 1.6678362573099414e-05, + "loss": 3.0996, + "step": 17574 + }, + { + "epoch": 0.925, + "grad_norm": 0.5, + "learning_rate": 1.6666666666666667e-05, + "loss": 3.4701, + "step": 17575 + }, + { + "epoch": 0.9250526315789473, + "grad_norm": 0.490234375, + "learning_rate": 1.6654970760233917e-05, + "loss": 3.4042, + "step": 17576 + }, + { + "epoch": 0.9251052631578948, + "grad_norm": 0.484375, + "learning_rate": 1.664327485380117e-05, + "loss": 3.4034, + "step": 17577 + }, + { + "epoch": 0.9251578947368421, + "grad_norm": 0.5, + "learning_rate": 1.6631578947368423e-05, + "loss": 3.4744, + "step": 17578 + }, + { + "epoch": 0.9252105263157895, + "grad_norm": 0.52734375, + "learning_rate": 1.6619883040935673e-05, + "loss": 3.3019, + "step": 17579 + }, + { + "epoch": 0.9252631578947368, + "grad_norm": 0.5078125, + "learning_rate": 1.6608187134502926e-05, + "loss": 3.4171, + "step": 17580 + }, + { + "epoch": 0.9253157894736842, + "grad_norm": 0.47265625, + "learning_rate": 1.6596491228070175e-05, + "loss": 2.7362, + "step": 17581 + }, + { + "epoch": 0.9253684210526316, + "grad_norm": 0.5078125, + "learning_rate": 1.6584795321637428e-05, + "loss": 3.4826, + "step": 17582 + }, + { + "epoch": 0.9254210526315789, + "grad_norm": 0.52734375, + "learning_rate": 1.6573099415204678e-05, + "loss": 3.0557, + "step": 17583 + }, + { + "epoch": 0.9254736842105263, + "grad_norm": 0.494140625, + "learning_rate": 1.656140350877193e-05, + "loss": 3.4202, + "step": 17584 + }, + { + "epoch": 0.9255263157894736, + "grad_norm": 0.5, + "learning_rate": 1.654970760233918e-05, + "loss": 3.066, + "step": 17585 + }, + { + "epoch": 0.9255789473684211, + "grad_norm": 0.49609375, + "learning_rate": 1.6538011695906433e-05, + "loss": 2.8823, + "step": 17586 + }, + { + "epoch": 0.9256315789473685, + "grad_norm": 0.53515625, + "learning_rate": 1.6526315789473683e-05, + "loss": 3.3182, + "step": 17587 + }, + { + "epoch": 0.9256842105263158, + "grad_norm": 0.50390625, + "learning_rate": 1.6514619883040936e-05, + "loss": 2.5149, + "step": 17588 + }, + { + "epoch": 0.9257368421052632, + "grad_norm": 0.5, + "learning_rate": 1.6502923976608185e-05, + "loss": 3.3093, + "step": 17589 + }, + { + "epoch": 0.9257894736842105, + "grad_norm": 0.4921875, + "learning_rate": 1.6491228070175442e-05, + "loss": 3.392, + "step": 17590 + }, + { + "epoch": 0.9258421052631579, + "grad_norm": 0.474609375, + "learning_rate": 1.647953216374269e-05, + "loss": 3.2853, + "step": 17591 + }, + { + "epoch": 0.9258947368421052, + "grad_norm": 0.50390625, + "learning_rate": 1.6467836257309944e-05, + "loss": 3.2142, + "step": 17592 + }, + { + "epoch": 0.9259473684210526, + "grad_norm": 0.48046875, + "learning_rate": 1.6456140350877194e-05, + "loss": 3.1786, + "step": 17593 + }, + { + "epoch": 0.926, + "grad_norm": 0.50390625, + "learning_rate": 1.6444444444444447e-05, + "loss": 3.6957, + "step": 17594 + }, + { + "epoch": 0.9260526315789473, + "grad_norm": 0.52734375, + "learning_rate": 1.6432748538011697e-05, + "loss": 3.4367, + "step": 17595 + }, + { + "epoch": 0.9261052631578948, + "grad_norm": 0.515625, + "learning_rate": 1.642105263157895e-05, + "loss": 3.7313, + "step": 17596 + }, + { + "epoch": 0.9261578947368421, + "grad_norm": 0.51171875, + "learning_rate": 1.64093567251462e-05, + "loss": 3.2481, + "step": 17597 + }, + { + "epoch": 0.9262105263157895, + "grad_norm": 0.50390625, + "learning_rate": 1.6397660818713452e-05, + "loss": 3.5081, + "step": 17598 + }, + { + "epoch": 0.9262631578947368, + "grad_norm": 0.5, + "learning_rate": 1.6385964912280702e-05, + "loss": 3.0592, + "step": 17599 + }, + { + "epoch": 0.9263157894736842, + "grad_norm": 0.51953125, + "learning_rate": 1.6374269005847955e-05, + "loss": 3.2021, + "step": 17600 + }, + { + "epoch": 0.9263684210526316, + "grad_norm": 0.486328125, + "learning_rate": 1.6362573099415204e-05, + "loss": 3.4918, + "step": 17601 + }, + { + "epoch": 0.9264210526315789, + "grad_norm": 0.482421875, + "learning_rate": 1.6350877192982454e-05, + "loss": 3.4201, + "step": 17602 + }, + { + "epoch": 0.9264736842105263, + "grad_norm": 0.494140625, + "learning_rate": 1.633918128654971e-05, + "loss": 2.8549, + "step": 17603 + }, + { + "epoch": 0.9265263157894736, + "grad_norm": 0.49609375, + "learning_rate": 1.632748538011696e-05, + "loss": 3.1092, + "step": 17604 + }, + { + "epoch": 0.9265789473684211, + "grad_norm": 0.54296875, + "learning_rate": 1.6315789473684213e-05, + "loss": 2.8128, + "step": 17605 + }, + { + "epoch": 0.9266315789473685, + "grad_norm": 0.49609375, + "learning_rate": 1.6304093567251462e-05, + "loss": 3.2978, + "step": 17606 + }, + { + "epoch": 0.9266842105263158, + "grad_norm": 0.5, + "learning_rate": 1.6292397660818715e-05, + "loss": 3.0027, + "step": 17607 + }, + { + "epoch": 0.9267368421052632, + "grad_norm": 0.5234375, + "learning_rate": 1.6280701754385965e-05, + "loss": 3.6739, + "step": 17608 + }, + { + "epoch": 0.9267894736842105, + "grad_norm": 0.486328125, + "learning_rate": 1.6269005847953218e-05, + "loss": 3.5844, + "step": 17609 + }, + { + "epoch": 0.9268421052631579, + "grad_norm": 0.48046875, + "learning_rate": 1.6257309941520468e-05, + "loss": 3.3229, + "step": 17610 + }, + { + "epoch": 0.9268947368421052, + "grad_norm": 0.56640625, + "learning_rate": 1.624561403508772e-05, + "loss": 3.3119, + "step": 17611 + }, + { + "epoch": 0.9269473684210526, + "grad_norm": 0.498046875, + "learning_rate": 1.623391812865497e-05, + "loss": 3.4388, + "step": 17612 + }, + { + "epoch": 0.927, + "grad_norm": 0.48828125, + "learning_rate": 1.6222222222222223e-05, + "loss": 3.2424, + "step": 17613 + }, + { + "epoch": 0.9270526315789474, + "grad_norm": 0.48046875, + "learning_rate": 1.6210526315789473e-05, + "loss": 3.001, + "step": 17614 + }, + { + "epoch": 0.9271052631578948, + "grad_norm": 0.486328125, + "learning_rate": 1.6198830409356726e-05, + "loss": 3.3869, + "step": 17615 + }, + { + "epoch": 0.9271578947368421, + "grad_norm": 0.51171875, + "learning_rate": 1.618713450292398e-05, + "loss": 3.0972, + "step": 17616 + }, + { + "epoch": 0.9272105263157895, + "grad_norm": 0.53125, + "learning_rate": 1.6175438596491228e-05, + "loss": 3.5656, + "step": 17617 + }, + { + "epoch": 0.9272631578947368, + "grad_norm": 0.52734375, + "learning_rate": 1.616374269005848e-05, + "loss": 3.7212, + "step": 17618 + }, + { + "epoch": 0.9273157894736842, + "grad_norm": 0.4921875, + "learning_rate": 1.615204678362573e-05, + "loss": 3.4885, + "step": 17619 + }, + { + "epoch": 0.9273684210526316, + "grad_norm": 0.515625, + "learning_rate": 1.6140350877192984e-05, + "loss": 3.4346, + "step": 17620 + }, + { + "epoch": 0.9274210526315789, + "grad_norm": 0.515625, + "learning_rate": 1.6128654970760233e-05, + "loss": 3.2752, + "step": 17621 + }, + { + "epoch": 0.9274736842105263, + "grad_norm": 0.5, + "learning_rate": 1.6116959064327486e-05, + "loss": 3.5586, + "step": 17622 + }, + { + "epoch": 0.9275263157894736, + "grad_norm": 0.51171875, + "learning_rate": 1.6105263157894736e-05, + "loss": 3.4044, + "step": 17623 + }, + { + "epoch": 0.9275789473684211, + "grad_norm": 0.49609375, + "learning_rate": 1.609356725146199e-05, + "loss": 3.1885, + "step": 17624 + }, + { + "epoch": 0.9276315789473685, + "grad_norm": 0.5078125, + "learning_rate": 1.608187134502924e-05, + "loss": 2.7688, + "step": 17625 + }, + { + "epoch": 0.9276842105263158, + "grad_norm": 0.51171875, + "learning_rate": 1.607017543859649e-05, + "loss": 3.3317, + "step": 17626 + }, + { + "epoch": 0.9277368421052632, + "grad_norm": 0.4765625, + "learning_rate": 1.605847953216374e-05, + "loss": 3.1873, + "step": 17627 + }, + { + "epoch": 0.9277894736842105, + "grad_norm": 0.5234375, + "learning_rate": 1.6046783625730998e-05, + "loss": 3.2957, + "step": 17628 + }, + { + "epoch": 0.9278421052631579, + "grad_norm": 0.50390625, + "learning_rate": 1.6035087719298247e-05, + "loss": 3.3279, + "step": 17629 + }, + { + "epoch": 0.9278947368421052, + "grad_norm": 0.498046875, + "learning_rate": 1.60233918128655e-05, + "loss": 3.1262, + "step": 17630 + }, + { + "epoch": 0.9279473684210526, + "grad_norm": 0.49609375, + "learning_rate": 1.601169590643275e-05, + "loss": 3.3186, + "step": 17631 + }, + { + "epoch": 0.928, + "grad_norm": 0.50390625, + "learning_rate": 1.6000000000000003e-05, + "loss": 3.0457, + "step": 17632 + }, + { + "epoch": 0.9280526315789474, + "grad_norm": 0.51953125, + "learning_rate": 1.5988304093567252e-05, + "loss": 3.436, + "step": 17633 + }, + { + "epoch": 0.9281052631578948, + "grad_norm": 0.54296875, + "learning_rate": 1.5976608187134505e-05, + "loss": 3.5695, + "step": 17634 + }, + { + "epoch": 0.9281578947368421, + "grad_norm": 0.4765625, + "learning_rate": 1.5964912280701755e-05, + "loss": 3.006, + "step": 17635 + }, + { + "epoch": 0.9282105263157895, + "grad_norm": 0.51171875, + "learning_rate": 1.5953216374269004e-05, + "loss": 3.4756, + "step": 17636 + }, + { + "epoch": 0.9282631578947368, + "grad_norm": 0.494140625, + "learning_rate": 1.5941520467836257e-05, + "loss": 3.3924, + "step": 17637 + }, + { + "epoch": 0.9283157894736842, + "grad_norm": 0.50390625, + "learning_rate": 1.5929824561403507e-05, + "loss": 3.1844, + "step": 17638 + }, + { + "epoch": 0.9283684210526316, + "grad_norm": 0.50390625, + "learning_rate": 1.591812865497076e-05, + "loss": 3.3468, + "step": 17639 + }, + { + "epoch": 0.9284210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.5906432748538013e-05, + "loss": 3.8488, + "step": 17640 + }, + { + "epoch": 0.9284736842105263, + "grad_norm": 0.53125, + "learning_rate": 1.5894736842105266e-05, + "loss": 3.3566, + "step": 17641 + }, + { + "epoch": 0.9285263157894736, + "grad_norm": 0.498046875, + "learning_rate": 1.5883040935672516e-05, + "loss": 3.2778, + "step": 17642 + }, + { + "epoch": 0.9285789473684211, + "grad_norm": 0.49609375, + "learning_rate": 1.587134502923977e-05, + "loss": 3.1165, + "step": 17643 + }, + { + "epoch": 0.9286315789473685, + "grad_norm": 0.474609375, + "learning_rate": 1.5859649122807018e-05, + "loss": 2.9731, + "step": 17644 + }, + { + "epoch": 0.9286842105263158, + "grad_norm": 0.51171875, + "learning_rate": 1.584795321637427e-05, + "loss": 3.3209, + "step": 17645 + }, + { + "epoch": 0.9287368421052632, + "grad_norm": 0.484375, + "learning_rate": 1.583625730994152e-05, + "loss": 3.5765, + "step": 17646 + }, + { + "epoch": 0.9287894736842105, + "grad_norm": 0.515625, + "learning_rate": 1.5824561403508774e-05, + "loss": 3.3089, + "step": 17647 + }, + { + "epoch": 0.9288421052631579, + "grad_norm": 0.50390625, + "learning_rate": 1.5812865497076023e-05, + "loss": 3.1355, + "step": 17648 + }, + { + "epoch": 0.9288947368421052, + "grad_norm": 0.5078125, + "learning_rate": 1.5801169590643276e-05, + "loss": 3.4829, + "step": 17649 + }, + { + "epoch": 0.9289473684210526, + "grad_norm": 0.515625, + "learning_rate": 1.5789473684210526e-05, + "loss": 3.2883, + "step": 17650 + }, + { + "epoch": 0.929, + "grad_norm": 0.5390625, + "learning_rate": 1.577777777777778e-05, + "loss": 3.4308, + "step": 17651 + }, + { + "epoch": 0.9290526315789474, + "grad_norm": 0.51953125, + "learning_rate": 1.576608187134503e-05, + "loss": 3.5467, + "step": 17652 + }, + { + "epoch": 0.9291052631578948, + "grad_norm": 0.51171875, + "learning_rate": 1.575438596491228e-05, + "loss": 3.2084, + "step": 17653 + }, + { + "epoch": 0.9291578947368421, + "grad_norm": 0.48046875, + "learning_rate": 1.5742690058479534e-05, + "loss": 2.9981, + "step": 17654 + }, + { + "epoch": 0.9292105263157895, + "grad_norm": 0.490234375, + "learning_rate": 1.5730994152046784e-05, + "loss": 2.4897, + "step": 17655 + }, + { + "epoch": 0.9292631578947368, + "grad_norm": 0.498046875, + "learning_rate": 1.5719298245614037e-05, + "loss": 3.036, + "step": 17656 + }, + { + "epoch": 0.9293157894736842, + "grad_norm": 0.490234375, + "learning_rate": 1.5707602339181287e-05, + "loss": 2.951, + "step": 17657 + }, + { + "epoch": 0.9293684210526316, + "grad_norm": 0.51953125, + "learning_rate": 1.569590643274854e-05, + "loss": 2.8881, + "step": 17658 + }, + { + "epoch": 0.9294210526315789, + "grad_norm": 0.51953125, + "learning_rate": 1.568421052631579e-05, + "loss": 3.3302, + "step": 17659 + }, + { + "epoch": 0.9294736842105263, + "grad_norm": 0.4921875, + "learning_rate": 1.5672514619883042e-05, + "loss": 3.3824, + "step": 17660 + }, + { + "epoch": 0.9295263157894736, + "grad_norm": 0.5078125, + "learning_rate": 1.566081871345029e-05, + "loss": 3.3476, + "step": 17661 + }, + { + "epoch": 0.9295789473684211, + "grad_norm": 0.515625, + "learning_rate": 1.5649122807017545e-05, + "loss": 3.1947, + "step": 17662 + }, + { + "epoch": 0.9296315789473685, + "grad_norm": 0.578125, + "learning_rate": 1.5637426900584794e-05, + "loss": 3.5667, + "step": 17663 + }, + { + "epoch": 0.9296842105263158, + "grad_norm": 0.51171875, + "learning_rate": 1.5625730994152047e-05, + "loss": 3.1259, + "step": 17664 + }, + { + "epoch": 0.9297368421052632, + "grad_norm": 0.5234375, + "learning_rate": 1.56140350877193e-05, + "loss": 3.103, + "step": 17665 + }, + { + "epoch": 0.9297894736842105, + "grad_norm": 0.52734375, + "learning_rate": 1.5602339181286553e-05, + "loss": 3.5197, + "step": 17666 + }, + { + "epoch": 0.9298421052631579, + "grad_norm": 0.494140625, + "learning_rate": 1.5590643274853803e-05, + "loss": 3.0885, + "step": 17667 + }, + { + "epoch": 0.9298947368421052, + "grad_norm": 0.5390625, + "learning_rate": 1.5578947368421056e-05, + "loss": 3.2303, + "step": 17668 + }, + { + "epoch": 0.9299473684210526, + "grad_norm": 0.490234375, + "learning_rate": 1.5567251461988305e-05, + "loss": 3.0969, + "step": 17669 + }, + { + "epoch": 0.93, + "grad_norm": 0.5078125, + "learning_rate": 1.5555555555555555e-05, + "loss": 3.1643, + "step": 17670 + }, + { + "epoch": 0.9300526315789474, + "grad_norm": 0.494140625, + "learning_rate": 1.5543859649122808e-05, + "loss": 3.5122, + "step": 17671 + }, + { + "epoch": 0.9301052631578948, + "grad_norm": 0.5546875, + "learning_rate": 1.5532163742690058e-05, + "loss": 2.6825, + "step": 17672 + }, + { + "epoch": 0.9301578947368421, + "grad_norm": 0.486328125, + "learning_rate": 1.552046783625731e-05, + "loss": 3.3269, + "step": 17673 + }, + { + "epoch": 0.9302105263157895, + "grad_norm": 0.51953125, + "learning_rate": 1.550877192982456e-05, + "loss": 3.4901, + "step": 17674 + }, + { + "epoch": 0.9302631578947368, + "grad_norm": 0.49609375, + "learning_rate": 1.5497076023391813e-05, + "loss": 3.3007, + "step": 17675 + }, + { + "epoch": 0.9303157894736842, + "grad_norm": 0.482421875, + "learning_rate": 1.5485380116959063e-05, + "loss": 3.2462, + "step": 17676 + }, + { + "epoch": 0.9303684210526316, + "grad_norm": 0.5234375, + "learning_rate": 1.5473684210526316e-05, + "loss": 3.2542, + "step": 17677 + }, + { + "epoch": 0.9304210526315789, + "grad_norm": 0.54296875, + "learning_rate": 1.546198830409357e-05, + "loss": 3.227, + "step": 17678 + }, + { + "epoch": 0.9304736842105263, + "grad_norm": 0.55078125, + "learning_rate": 1.545029239766082e-05, + "loss": 3.593, + "step": 17679 + }, + { + "epoch": 0.9305263157894736, + "grad_norm": 0.515625, + "learning_rate": 1.543859649122807e-05, + "loss": 3.1705, + "step": 17680 + }, + { + "epoch": 0.9305789473684211, + "grad_norm": 0.5, + "learning_rate": 1.5426900584795324e-05, + "loss": 3.0447, + "step": 17681 + }, + { + "epoch": 0.9306315789473685, + "grad_norm": 0.5234375, + "learning_rate": 1.5415204678362574e-05, + "loss": 3.3236, + "step": 17682 + }, + { + "epoch": 0.9306842105263158, + "grad_norm": 0.494140625, + "learning_rate": 1.5403508771929827e-05, + "loss": 3.3099, + "step": 17683 + }, + { + "epoch": 0.9307368421052632, + "grad_norm": 0.50390625, + "learning_rate": 1.5391812865497076e-05, + "loss": 3.2574, + "step": 17684 + }, + { + "epoch": 0.9307894736842105, + "grad_norm": 0.515625, + "learning_rate": 1.538011695906433e-05, + "loss": 3.8601, + "step": 17685 + }, + { + "epoch": 0.9308421052631579, + "grad_norm": 0.515625, + "learning_rate": 1.536842105263158e-05, + "loss": 2.9754, + "step": 17686 + }, + { + "epoch": 0.9308947368421052, + "grad_norm": 0.515625, + "learning_rate": 1.5356725146198832e-05, + "loss": 3.749, + "step": 17687 + }, + { + "epoch": 0.9309473684210526, + "grad_norm": 0.5234375, + "learning_rate": 1.534502923976608e-05, + "loss": 3.5363, + "step": 17688 + }, + { + "epoch": 0.931, + "grad_norm": 0.5078125, + "learning_rate": 1.5333333333333334e-05, + "loss": 3.0976, + "step": 17689 + }, + { + "epoch": 0.9310526315789474, + "grad_norm": 0.55859375, + "learning_rate": 1.5321637426900587e-05, + "loss": 3.0911, + "step": 17690 + }, + { + "epoch": 0.9311052631578948, + "grad_norm": 0.498046875, + "learning_rate": 1.5309941520467837e-05, + "loss": 3.2218, + "step": 17691 + }, + { + "epoch": 0.9311578947368421, + "grad_norm": 0.5078125, + "learning_rate": 1.529824561403509e-05, + "loss": 3.0043, + "step": 17692 + }, + { + "epoch": 0.9312105263157895, + "grad_norm": 0.51171875, + "learning_rate": 1.528654970760234e-05, + "loss": 3.1264, + "step": 17693 + }, + { + "epoch": 0.9312631578947368, + "grad_norm": 0.494140625, + "learning_rate": 1.5274853801169593e-05, + "loss": 3.5617, + "step": 17694 + }, + { + "epoch": 0.9313157894736842, + "grad_norm": 0.498046875, + "learning_rate": 1.5263157894736842e-05, + "loss": 3.0772, + "step": 17695 + }, + { + "epoch": 0.9313684210526316, + "grad_norm": 0.51953125, + "learning_rate": 1.5251461988304095e-05, + "loss": 3.1716, + "step": 17696 + }, + { + "epoch": 0.9314210526315789, + "grad_norm": 0.5, + "learning_rate": 1.5239766081871345e-05, + "loss": 3.3039, + "step": 17697 + }, + { + "epoch": 0.9314736842105263, + "grad_norm": 0.5, + "learning_rate": 1.5228070175438596e-05, + "loss": 3.1853, + "step": 17698 + }, + { + "epoch": 0.9315263157894736, + "grad_norm": 0.46875, + "learning_rate": 1.5216374269005847e-05, + "loss": 3.2833, + "step": 17699 + }, + { + "epoch": 0.9315789473684211, + "grad_norm": 0.53125, + "learning_rate": 1.5204678362573099e-05, + "loss": 3.1221, + "step": 17700 + }, + { + "epoch": 0.9316315789473684, + "grad_norm": 0.51953125, + "learning_rate": 1.519298245614035e-05, + "loss": 3.2463, + "step": 17701 + }, + { + "epoch": 0.9316842105263158, + "grad_norm": 0.486328125, + "learning_rate": 1.5181286549707601e-05, + "loss": 3.3121, + "step": 17702 + }, + { + "epoch": 0.9317368421052632, + "grad_norm": 0.51953125, + "learning_rate": 1.5169590643274856e-05, + "loss": 3.4514, + "step": 17703 + }, + { + "epoch": 0.9317894736842105, + "grad_norm": 0.5, + "learning_rate": 1.5157894736842107e-05, + "loss": 3.4277, + "step": 17704 + }, + { + "epoch": 0.9318421052631579, + "grad_norm": 0.51171875, + "learning_rate": 1.5146198830409358e-05, + "loss": 3.254, + "step": 17705 + }, + { + "epoch": 0.9318947368421052, + "grad_norm": 0.546875, + "learning_rate": 1.513450292397661e-05, + "loss": 3.0509, + "step": 17706 + }, + { + "epoch": 0.9319473684210526, + "grad_norm": 0.5078125, + "learning_rate": 1.5122807017543861e-05, + "loss": 3.3485, + "step": 17707 + }, + { + "epoch": 0.932, + "grad_norm": 0.486328125, + "learning_rate": 1.5111111111111112e-05, + "loss": 3.2301, + "step": 17708 + }, + { + "epoch": 0.9320526315789474, + "grad_norm": 0.51171875, + "learning_rate": 1.5099415204678364e-05, + "loss": 3.44, + "step": 17709 + }, + { + "epoch": 0.9321052631578948, + "grad_norm": 0.515625, + "learning_rate": 1.5087719298245615e-05, + "loss": 3.1272, + "step": 17710 + }, + { + "epoch": 0.9321578947368421, + "grad_norm": 0.5234375, + "learning_rate": 1.5076023391812866e-05, + "loss": 3.6398, + "step": 17711 + }, + { + "epoch": 0.9322105263157895, + "grad_norm": 0.49609375, + "learning_rate": 1.5064327485380117e-05, + "loss": 3.3226, + "step": 17712 + }, + { + "epoch": 0.9322631578947368, + "grad_norm": 0.498046875, + "learning_rate": 1.5052631578947369e-05, + "loss": 3.591, + "step": 17713 + }, + { + "epoch": 0.9323157894736842, + "grad_norm": 0.51953125, + "learning_rate": 1.504093567251462e-05, + "loss": 3.518, + "step": 17714 + }, + { + "epoch": 0.9323684210526316, + "grad_norm": 0.62890625, + "learning_rate": 1.5029239766081873e-05, + "loss": 3.1201, + "step": 17715 + }, + { + "epoch": 0.9324210526315789, + "grad_norm": 0.51171875, + "learning_rate": 1.5017543859649124e-05, + "loss": 3.4327, + "step": 17716 + }, + { + "epoch": 0.9324736842105263, + "grad_norm": 0.48828125, + "learning_rate": 1.5005847953216376e-05, + "loss": 3.4156, + "step": 17717 + }, + { + "epoch": 0.9325263157894736, + "grad_norm": 0.486328125, + "learning_rate": 1.4994152046783627e-05, + "loss": 3.006, + "step": 17718 + }, + { + "epoch": 0.9325789473684211, + "grad_norm": 0.50390625, + "learning_rate": 1.4982456140350878e-05, + "loss": 3.1642, + "step": 17719 + }, + { + "epoch": 0.9326315789473684, + "grad_norm": 0.54296875, + "learning_rate": 1.497076023391813e-05, + "loss": 3.2177, + "step": 17720 + }, + { + "epoch": 0.9326842105263158, + "grad_norm": 0.51953125, + "learning_rate": 1.495906432748538e-05, + "loss": 3.589, + "step": 17721 + }, + { + "epoch": 0.9327368421052632, + "grad_norm": 0.5234375, + "learning_rate": 1.4947368421052632e-05, + "loss": 3.6249, + "step": 17722 + }, + { + "epoch": 0.9327894736842105, + "grad_norm": 0.494140625, + "learning_rate": 1.4935672514619883e-05, + "loss": 3.3704, + "step": 17723 + }, + { + "epoch": 0.9328421052631579, + "grad_norm": 0.486328125, + "learning_rate": 1.4923976608187135e-05, + "loss": 3.4959, + "step": 17724 + }, + { + "epoch": 0.9328947368421052, + "grad_norm": 0.4921875, + "learning_rate": 1.4912280701754386e-05, + "loss": 3.4216, + "step": 17725 + }, + { + "epoch": 0.9329473684210526, + "grad_norm": 0.498046875, + "learning_rate": 1.4900584795321637e-05, + "loss": 2.9945, + "step": 17726 + }, + { + "epoch": 0.933, + "grad_norm": 0.51953125, + "learning_rate": 1.4888888888888888e-05, + "loss": 2.9886, + "step": 17727 + }, + { + "epoch": 0.9330526315789474, + "grad_norm": 0.484375, + "learning_rate": 1.4877192982456143e-05, + "loss": 3.3387, + "step": 17728 + }, + { + "epoch": 0.9331052631578948, + "grad_norm": 0.515625, + "learning_rate": 1.4865497076023394e-05, + "loss": 3.3855, + "step": 17729 + }, + { + "epoch": 0.9331578947368421, + "grad_norm": 0.5, + "learning_rate": 1.4853801169590644e-05, + "loss": 3.483, + "step": 17730 + }, + { + "epoch": 0.9332105263157895, + "grad_norm": 0.51953125, + "learning_rate": 1.4842105263157895e-05, + "loss": 3.3815, + "step": 17731 + }, + { + "epoch": 0.9332631578947368, + "grad_norm": 0.515625, + "learning_rate": 1.4830409356725147e-05, + "loss": 2.4679, + "step": 17732 + }, + { + "epoch": 0.9333157894736842, + "grad_norm": 0.51171875, + "learning_rate": 1.4818713450292398e-05, + "loss": 3.5061, + "step": 17733 + }, + { + "epoch": 0.9333684210526316, + "grad_norm": 0.52734375, + "learning_rate": 1.4807017543859649e-05, + "loss": 3.1903, + "step": 17734 + }, + { + "epoch": 0.9334210526315789, + "grad_norm": 0.625, + "learning_rate": 1.47953216374269e-05, + "loss": 3.1513, + "step": 17735 + }, + { + "epoch": 0.9334736842105263, + "grad_norm": 0.51953125, + "learning_rate": 1.4783625730994152e-05, + "loss": 3.2521, + "step": 17736 + }, + { + "epoch": 0.9335263157894736, + "grad_norm": 0.4921875, + "learning_rate": 1.4771929824561403e-05, + "loss": 3.3504, + "step": 17737 + }, + { + "epoch": 0.9335789473684211, + "grad_norm": 0.478515625, + "learning_rate": 1.4760233918128654e-05, + "loss": 3.2534, + "step": 17738 + }, + { + "epoch": 0.9336315789473684, + "grad_norm": 0.5078125, + "learning_rate": 1.4748538011695906e-05, + "loss": 3.3436, + "step": 17739 + }, + { + "epoch": 0.9336842105263158, + "grad_norm": 0.5, + "learning_rate": 1.4736842105263157e-05, + "loss": 2.636, + "step": 17740 + }, + { + "epoch": 0.9337368421052632, + "grad_norm": 0.515625, + "learning_rate": 1.4725146198830412e-05, + "loss": 3.5255, + "step": 17741 + }, + { + "epoch": 0.9337894736842105, + "grad_norm": 0.50390625, + "learning_rate": 1.4713450292397663e-05, + "loss": 3.6261, + "step": 17742 + }, + { + "epoch": 0.9338421052631579, + "grad_norm": 0.48046875, + "learning_rate": 1.4701754385964914e-05, + "loss": 2.779, + "step": 17743 + }, + { + "epoch": 0.9338947368421052, + "grad_norm": 0.515625, + "learning_rate": 1.4690058479532165e-05, + "loss": 3.4334, + "step": 17744 + }, + { + "epoch": 0.9339473684210526, + "grad_norm": 0.51953125, + "learning_rate": 1.4678362573099417e-05, + "loss": 3.0248, + "step": 17745 + }, + { + "epoch": 0.934, + "grad_norm": 0.48046875, + "learning_rate": 1.4666666666666668e-05, + "loss": 3.1295, + "step": 17746 + }, + { + "epoch": 0.9340526315789474, + "grad_norm": 0.52734375, + "learning_rate": 1.465497076023392e-05, + "loss": 3.4784, + "step": 17747 + }, + { + "epoch": 0.9341052631578948, + "grad_norm": 0.49609375, + "learning_rate": 1.464327485380117e-05, + "loss": 3.1383, + "step": 17748 + }, + { + "epoch": 0.9341578947368421, + "grad_norm": 0.50390625, + "learning_rate": 1.4631578947368422e-05, + "loss": 3.475, + "step": 17749 + }, + { + "epoch": 0.9342105263157895, + "grad_norm": 0.470703125, + "learning_rate": 1.4619883040935673e-05, + "loss": 2.8721, + "step": 17750 + }, + { + "epoch": 0.9342631578947368, + "grad_norm": 0.625, + "learning_rate": 1.4608187134502924e-05, + "loss": 2.9904, + "step": 17751 + }, + { + "epoch": 0.9343157894736842, + "grad_norm": 0.486328125, + "learning_rate": 1.4596491228070176e-05, + "loss": 2.7353, + "step": 17752 + }, + { + "epoch": 0.9343684210526316, + "grad_norm": 0.4921875, + "learning_rate": 1.4584795321637429e-05, + "loss": 3.3495, + "step": 17753 + }, + { + "epoch": 0.9344210526315789, + "grad_norm": 0.498046875, + "learning_rate": 1.457309941520468e-05, + "loss": 3.2362, + "step": 17754 + }, + { + "epoch": 0.9344736842105263, + "grad_norm": 0.484375, + "learning_rate": 1.4561403508771931e-05, + "loss": 2.9263, + "step": 17755 + }, + { + "epoch": 0.9345263157894736, + "grad_norm": 0.5, + "learning_rate": 1.4549707602339183e-05, + "loss": 3.0736, + "step": 17756 + }, + { + "epoch": 0.9345789473684211, + "grad_norm": 0.490234375, + "learning_rate": 1.4538011695906434e-05, + "loss": 3.6524, + "step": 17757 + }, + { + "epoch": 0.9346315789473684, + "grad_norm": 0.515625, + "learning_rate": 1.4526315789473685e-05, + "loss": 3.1013, + "step": 17758 + }, + { + "epoch": 0.9346842105263158, + "grad_norm": 0.53515625, + "learning_rate": 1.4514619883040936e-05, + "loss": 3.6115, + "step": 17759 + }, + { + "epoch": 0.9347368421052632, + "grad_norm": 0.50390625, + "learning_rate": 1.4502923976608188e-05, + "loss": 3.5694, + "step": 17760 + }, + { + "epoch": 0.9347894736842105, + "grad_norm": 0.52734375, + "learning_rate": 1.4491228070175439e-05, + "loss": 3.4978, + "step": 17761 + }, + { + "epoch": 0.9348421052631579, + "grad_norm": 0.515625, + "learning_rate": 1.447953216374269e-05, + "loss": 3.435, + "step": 17762 + }, + { + "epoch": 0.9348947368421052, + "grad_norm": 0.486328125, + "learning_rate": 1.4467836257309942e-05, + "loss": 3.0026, + "step": 17763 + }, + { + "epoch": 0.9349473684210526, + "grad_norm": 0.5703125, + "learning_rate": 1.4456140350877193e-05, + "loss": 3.0703, + "step": 17764 + }, + { + "epoch": 0.935, + "grad_norm": 0.5078125, + "learning_rate": 1.4444444444444444e-05, + "loss": 3.3248, + "step": 17765 + }, + { + "epoch": 0.9350526315789474, + "grad_norm": 0.4921875, + "learning_rate": 1.4432748538011697e-05, + "loss": 3.0567, + "step": 17766 + }, + { + "epoch": 0.9351052631578948, + "grad_norm": 0.5234375, + "learning_rate": 1.4421052631578948e-05, + "loss": 3.3513, + "step": 17767 + }, + { + "epoch": 0.9351578947368421, + "grad_norm": 0.50390625, + "learning_rate": 1.44093567251462e-05, + "loss": 3.0082, + "step": 17768 + }, + { + "epoch": 0.9352105263157895, + "grad_norm": 0.486328125, + "learning_rate": 1.4397660818713451e-05, + "loss": 3.4792, + "step": 17769 + }, + { + "epoch": 0.9352631578947368, + "grad_norm": 0.490234375, + "learning_rate": 1.4385964912280702e-05, + "loss": 3.1514, + "step": 17770 + }, + { + "epoch": 0.9353157894736842, + "grad_norm": 0.5, + "learning_rate": 1.4374269005847954e-05, + "loss": 3.2766, + "step": 17771 + }, + { + "epoch": 0.9353684210526316, + "grad_norm": 0.51171875, + "learning_rate": 1.4362573099415205e-05, + "loss": 3.4319, + "step": 17772 + }, + { + "epoch": 0.9354210526315789, + "grad_norm": 0.494140625, + "learning_rate": 1.4350877192982456e-05, + "loss": 2.8002, + "step": 17773 + }, + { + "epoch": 0.9354736842105263, + "grad_norm": 0.5234375, + "learning_rate": 1.4339181286549707e-05, + "loss": 3.2923, + "step": 17774 + }, + { + "epoch": 0.9355263157894737, + "grad_norm": 0.48046875, + "learning_rate": 1.4327485380116959e-05, + "loss": 3.0479, + "step": 17775 + }, + { + "epoch": 0.9355789473684211, + "grad_norm": 0.484375, + "learning_rate": 1.431578947368421e-05, + "loss": 3.1759, + "step": 17776 + }, + { + "epoch": 0.9356315789473684, + "grad_norm": 0.51171875, + "learning_rate": 1.4304093567251461e-05, + "loss": 3.3097, + "step": 17777 + }, + { + "epoch": 0.9356842105263158, + "grad_norm": 0.498046875, + "learning_rate": 1.4292397660818716e-05, + "loss": 3.3685, + "step": 17778 + }, + { + "epoch": 0.9357368421052632, + "grad_norm": 0.51953125, + "learning_rate": 1.4280701754385967e-05, + "loss": 3.3269, + "step": 17779 + }, + { + "epoch": 0.9357894736842105, + "grad_norm": 0.50390625, + "learning_rate": 1.4269005847953219e-05, + "loss": 3.4368, + "step": 17780 + }, + { + "epoch": 0.9358421052631579, + "grad_norm": 0.484375, + "learning_rate": 1.425730994152047e-05, + "loss": 3.4329, + "step": 17781 + }, + { + "epoch": 0.9358947368421052, + "grad_norm": 0.5234375, + "learning_rate": 1.4245614035087721e-05, + "loss": 3.1309, + "step": 17782 + }, + { + "epoch": 0.9359473684210526, + "grad_norm": 0.498046875, + "learning_rate": 1.4233918128654972e-05, + "loss": 3.2907, + "step": 17783 + }, + { + "epoch": 0.936, + "grad_norm": 0.482421875, + "learning_rate": 1.4222222222222224e-05, + "loss": 3.2056, + "step": 17784 + }, + { + "epoch": 0.9360526315789474, + "grad_norm": 0.48046875, + "learning_rate": 1.4210526315789475e-05, + "loss": 3.2255, + "step": 17785 + }, + { + "epoch": 0.9361052631578948, + "grad_norm": 0.484375, + "learning_rate": 1.4198830409356725e-05, + "loss": 3.2211, + "step": 17786 + }, + { + "epoch": 0.9361578947368421, + "grad_norm": 0.478515625, + "learning_rate": 1.4187134502923976e-05, + "loss": 3.5287, + "step": 17787 + }, + { + "epoch": 0.9362105263157895, + "grad_norm": 0.5, + "learning_rate": 1.4175438596491227e-05, + "loss": 3.0199, + "step": 17788 + }, + { + "epoch": 0.9362631578947368, + "grad_norm": 0.53125, + "learning_rate": 1.4163742690058478e-05, + "loss": 3.4769, + "step": 17789 + }, + { + "epoch": 0.9363157894736842, + "grad_norm": 0.5078125, + "learning_rate": 1.415204678362573e-05, + "loss": 3.0338, + "step": 17790 + }, + { + "epoch": 0.9363684210526316, + "grad_norm": 0.498046875, + "learning_rate": 1.4140350877192984e-05, + "loss": 3.4938, + "step": 17791 + }, + { + "epoch": 0.9364210526315789, + "grad_norm": 0.5, + "learning_rate": 1.4128654970760236e-05, + "loss": 3.1425, + "step": 17792 + }, + { + "epoch": 0.9364736842105263, + "grad_norm": 0.4765625, + "learning_rate": 1.4116959064327487e-05, + "loss": 2.9595, + "step": 17793 + }, + { + "epoch": 0.9365263157894737, + "grad_norm": 0.5078125, + "learning_rate": 1.4105263157894738e-05, + "loss": 3.5504, + "step": 17794 + }, + { + "epoch": 0.9365789473684211, + "grad_norm": 0.498046875, + "learning_rate": 1.409356725146199e-05, + "loss": 3.1448, + "step": 17795 + }, + { + "epoch": 0.9366315789473684, + "grad_norm": 0.55859375, + "learning_rate": 1.408187134502924e-05, + "loss": 3.1729, + "step": 17796 + }, + { + "epoch": 0.9366842105263158, + "grad_norm": 0.48828125, + "learning_rate": 1.4070175438596492e-05, + "loss": 3.2111, + "step": 17797 + }, + { + "epoch": 0.9367368421052632, + "grad_norm": 0.5390625, + "learning_rate": 1.4058479532163743e-05, + "loss": 3.4332, + "step": 17798 + }, + { + "epoch": 0.9367894736842105, + "grad_norm": 0.5078125, + "learning_rate": 1.4046783625730995e-05, + "loss": 3.1417, + "step": 17799 + }, + { + "epoch": 0.9368421052631579, + "grad_norm": 0.515625, + "learning_rate": 1.4035087719298246e-05, + "loss": 3.3151, + "step": 17800 + }, + { + "epoch": 0.9368947368421052, + "grad_norm": 0.50390625, + "learning_rate": 1.4023391812865497e-05, + "loss": 3.3271, + "step": 17801 + }, + { + "epoch": 0.9369473684210526, + "grad_norm": 0.5234375, + "learning_rate": 1.4011695906432749e-05, + "loss": 3.0275, + "step": 17802 + }, + { + "epoch": 0.937, + "grad_norm": 0.50390625, + "learning_rate": 1.4000000000000001e-05, + "loss": 3.635, + "step": 17803 + }, + { + "epoch": 0.9370526315789474, + "grad_norm": 0.5, + "learning_rate": 1.3988304093567253e-05, + "loss": 3.1386, + "step": 17804 + }, + { + "epoch": 0.9371052631578948, + "grad_norm": 0.498046875, + "learning_rate": 1.3976608187134504e-05, + "loss": 3.2267, + "step": 17805 + }, + { + "epoch": 0.9371578947368421, + "grad_norm": 0.515625, + "learning_rate": 1.3964912280701755e-05, + "loss": 3.1148, + "step": 17806 + }, + { + "epoch": 0.9372105263157895, + "grad_norm": 0.482421875, + "learning_rate": 1.3953216374269007e-05, + "loss": 3.3404, + "step": 17807 + }, + { + "epoch": 0.9372631578947368, + "grad_norm": 0.49609375, + "learning_rate": 1.3941520467836258e-05, + "loss": 3.2689, + "step": 17808 + }, + { + "epoch": 0.9373157894736842, + "grad_norm": 0.52734375, + "learning_rate": 1.392982456140351e-05, + "loss": 2.9293, + "step": 17809 + }, + { + "epoch": 0.9373684210526316, + "grad_norm": 0.52734375, + "learning_rate": 1.391812865497076e-05, + "loss": 3.1241, + "step": 17810 + }, + { + "epoch": 0.9374210526315789, + "grad_norm": 0.482421875, + "learning_rate": 1.3906432748538012e-05, + "loss": 2.9599, + "step": 17811 + }, + { + "epoch": 0.9374736842105263, + "grad_norm": 0.56640625, + "learning_rate": 1.3894736842105263e-05, + "loss": 2.7314, + "step": 17812 + }, + { + "epoch": 0.9375263157894737, + "grad_norm": 0.51953125, + "learning_rate": 1.3883040935672514e-05, + "loss": 3.4446, + "step": 17813 + }, + { + "epoch": 0.9375789473684211, + "grad_norm": 0.53515625, + "learning_rate": 1.3871345029239766e-05, + "loss": 3.6052, + "step": 17814 + }, + { + "epoch": 0.9376315789473684, + "grad_norm": 0.51953125, + "learning_rate": 1.3859649122807017e-05, + "loss": 3.0535, + "step": 17815 + }, + { + "epoch": 0.9376842105263158, + "grad_norm": 0.4921875, + "learning_rate": 1.3847953216374272e-05, + "loss": 3.041, + "step": 17816 + }, + { + "epoch": 0.9377368421052632, + "grad_norm": 0.486328125, + "learning_rate": 1.3836257309941523e-05, + "loss": 3.0257, + "step": 17817 + }, + { + "epoch": 0.9377894736842105, + "grad_norm": 0.490234375, + "learning_rate": 1.3824561403508774e-05, + "loss": 3.4428, + "step": 17818 + }, + { + "epoch": 0.9378421052631579, + "grad_norm": 0.50390625, + "learning_rate": 1.3812865497076025e-05, + "loss": 3.5755, + "step": 17819 + }, + { + "epoch": 0.9378947368421052, + "grad_norm": 0.53515625, + "learning_rate": 1.3801169590643275e-05, + "loss": 3.3178, + "step": 17820 + }, + { + "epoch": 0.9379473684210526, + "grad_norm": 0.490234375, + "learning_rate": 1.3789473684210526e-05, + "loss": 3.2972, + "step": 17821 + }, + { + "epoch": 0.938, + "grad_norm": 0.51171875, + "learning_rate": 1.3777777777777778e-05, + "loss": 3.401, + "step": 17822 + }, + { + "epoch": 0.9380526315789474, + "grad_norm": 0.470703125, + "learning_rate": 1.3766081871345029e-05, + "loss": 2.9526, + "step": 17823 + }, + { + "epoch": 0.9381052631578948, + "grad_norm": 0.51953125, + "learning_rate": 1.375438596491228e-05, + "loss": 3.608, + "step": 17824 + }, + { + "epoch": 0.9381578947368421, + "grad_norm": 0.5234375, + "learning_rate": 1.3742690058479531e-05, + "loss": 3.5877, + "step": 17825 + }, + { + "epoch": 0.9382105263157895, + "grad_norm": 0.484375, + "learning_rate": 1.3730994152046783e-05, + "loss": 3.3879, + "step": 17826 + }, + { + "epoch": 0.9382631578947368, + "grad_norm": 0.5078125, + "learning_rate": 1.3719298245614034e-05, + "loss": 3.6423, + "step": 17827 + }, + { + "epoch": 0.9383157894736842, + "grad_norm": 0.58203125, + "learning_rate": 1.3707602339181289e-05, + "loss": 2.997, + "step": 17828 + }, + { + "epoch": 0.9383684210526316, + "grad_norm": 0.52734375, + "learning_rate": 1.369590643274854e-05, + "loss": 3.681, + "step": 17829 + }, + { + "epoch": 0.9384210526315789, + "grad_norm": 0.494140625, + "learning_rate": 1.3684210526315791e-05, + "loss": 3.2348, + "step": 17830 + }, + { + "epoch": 0.9384736842105263, + "grad_norm": 0.52734375, + "learning_rate": 1.3672514619883043e-05, + "loss": 2.8994, + "step": 17831 + }, + { + "epoch": 0.9385263157894737, + "grad_norm": 0.51953125, + "learning_rate": 1.3660818713450294e-05, + "loss": 3.5269, + "step": 17832 + }, + { + "epoch": 0.9385789473684211, + "grad_norm": 0.482421875, + "learning_rate": 1.3649122807017545e-05, + "loss": 2.9438, + "step": 17833 + }, + { + "epoch": 0.9386315789473684, + "grad_norm": 0.49609375, + "learning_rate": 1.3637426900584796e-05, + "loss": 3.4214, + "step": 17834 + }, + { + "epoch": 0.9386842105263158, + "grad_norm": 0.53125, + "learning_rate": 1.3625730994152048e-05, + "loss": 3.1403, + "step": 17835 + }, + { + "epoch": 0.9387368421052632, + "grad_norm": 0.52734375, + "learning_rate": 1.3614035087719299e-05, + "loss": 2.9543, + "step": 17836 + }, + { + "epoch": 0.9387894736842105, + "grad_norm": 0.52734375, + "learning_rate": 1.360233918128655e-05, + "loss": 3.3655, + "step": 17837 + }, + { + "epoch": 0.9388421052631579, + "grad_norm": 0.486328125, + "learning_rate": 1.3590643274853802e-05, + "loss": 3.017, + "step": 17838 + }, + { + "epoch": 0.9388947368421052, + "grad_norm": 0.4765625, + "learning_rate": 1.3578947368421053e-05, + "loss": 3.1246, + "step": 17839 + }, + { + "epoch": 0.9389473684210526, + "grad_norm": 0.498046875, + "learning_rate": 1.3567251461988304e-05, + "loss": 3.1397, + "step": 17840 + }, + { + "epoch": 0.939, + "grad_norm": 0.51953125, + "learning_rate": 1.3555555555555557e-05, + "loss": 3.1486, + "step": 17841 + }, + { + "epoch": 0.9390526315789474, + "grad_norm": 0.482421875, + "learning_rate": 1.3543859649122808e-05, + "loss": 3.2375, + "step": 17842 + }, + { + "epoch": 0.9391052631578948, + "grad_norm": 0.51953125, + "learning_rate": 1.353216374269006e-05, + "loss": 3.3735, + "step": 17843 + }, + { + "epoch": 0.9391578947368421, + "grad_norm": 0.494140625, + "learning_rate": 1.3520467836257311e-05, + "loss": 3.3892, + "step": 17844 + }, + { + "epoch": 0.9392105263157895, + "grad_norm": 0.54296875, + "learning_rate": 1.3508771929824562e-05, + "loss": 3.1163, + "step": 17845 + }, + { + "epoch": 0.9392631578947368, + "grad_norm": 0.474609375, + "learning_rate": 1.3497076023391814e-05, + "loss": 3.2437, + "step": 17846 + }, + { + "epoch": 0.9393157894736842, + "grad_norm": 0.5, + "learning_rate": 1.3485380116959065e-05, + "loss": 3.2184, + "step": 17847 + }, + { + "epoch": 0.9393684210526316, + "grad_norm": 0.4921875, + "learning_rate": 1.3473684210526316e-05, + "loss": 3.3297, + "step": 17848 + }, + { + "epoch": 0.9394210526315789, + "grad_norm": 0.478515625, + "learning_rate": 1.3461988304093567e-05, + "loss": 3.2353, + "step": 17849 + }, + { + "epoch": 0.9394736842105263, + "grad_norm": 0.49609375, + "learning_rate": 1.3450292397660819e-05, + "loss": 2.8783, + "step": 17850 + }, + { + "epoch": 0.9395263157894737, + "grad_norm": 0.5078125, + "learning_rate": 1.343859649122807e-05, + "loss": 3.4254, + "step": 17851 + }, + { + "epoch": 0.9395789473684211, + "grad_norm": 0.5078125, + "learning_rate": 1.3426900584795321e-05, + "loss": 3.071, + "step": 17852 + }, + { + "epoch": 0.9396315789473684, + "grad_norm": 0.494140625, + "learning_rate": 1.3415204678362574e-05, + "loss": 3.5042, + "step": 17853 + }, + { + "epoch": 0.9396842105263158, + "grad_norm": 0.51171875, + "learning_rate": 1.3403508771929826e-05, + "loss": 3.7472, + "step": 17854 + }, + { + "epoch": 0.9397368421052632, + "grad_norm": 0.5390625, + "learning_rate": 1.3391812865497077e-05, + "loss": 2.8914, + "step": 17855 + }, + { + "epoch": 0.9397894736842105, + "grad_norm": 0.5, + "learning_rate": 1.3380116959064328e-05, + "loss": 3.0247, + "step": 17856 + }, + { + "epoch": 0.9398421052631579, + "grad_norm": 0.5234375, + "learning_rate": 1.336842105263158e-05, + "loss": 2.9854, + "step": 17857 + }, + { + "epoch": 0.9398947368421052, + "grad_norm": 0.5234375, + "learning_rate": 1.335672514619883e-05, + "loss": 3.254, + "step": 17858 + }, + { + "epoch": 0.9399473684210526, + "grad_norm": 0.49609375, + "learning_rate": 1.3345029239766082e-05, + "loss": 3.706, + "step": 17859 + }, + { + "epoch": 0.94, + "grad_norm": 0.50390625, + "learning_rate": 1.3333333333333333e-05, + "loss": 3.2256, + "step": 17860 + }, + { + "epoch": 0.9400526315789474, + "grad_norm": 0.58203125, + "learning_rate": 1.3321637426900585e-05, + "loss": 3.2094, + "step": 17861 + }, + { + "epoch": 0.9401052631578948, + "grad_norm": 0.5234375, + "learning_rate": 1.3309941520467836e-05, + "loss": 3.3385, + "step": 17862 + }, + { + "epoch": 0.9401578947368421, + "grad_norm": 0.474609375, + "learning_rate": 1.3298245614035087e-05, + "loss": 2.7369, + "step": 17863 + }, + { + "epoch": 0.9402105263157895, + "grad_norm": 0.59765625, + "learning_rate": 1.3286549707602338e-05, + "loss": 3.5002, + "step": 17864 + }, + { + "epoch": 0.9402631578947368, + "grad_norm": 0.4765625, + "learning_rate": 1.327485380116959e-05, + "loss": 3.2865, + "step": 17865 + }, + { + "epoch": 0.9403157894736842, + "grad_norm": 0.490234375, + "learning_rate": 1.3263157894736844e-05, + "loss": 3.4154, + "step": 17866 + }, + { + "epoch": 0.9403684210526316, + "grad_norm": 0.46875, + "learning_rate": 1.3251461988304096e-05, + "loss": 2.8955, + "step": 17867 + }, + { + "epoch": 0.9404210526315789, + "grad_norm": 0.5078125, + "learning_rate": 1.3239766081871347e-05, + "loss": 2.9246, + "step": 17868 + }, + { + "epoch": 0.9404736842105264, + "grad_norm": 0.51953125, + "learning_rate": 1.3228070175438598e-05, + "loss": 3.3488, + "step": 17869 + }, + { + "epoch": 0.9405263157894737, + "grad_norm": 0.50390625, + "learning_rate": 1.321637426900585e-05, + "loss": 3.3079, + "step": 17870 + }, + { + "epoch": 0.9405789473684211, + "grad_norm": 0.51171875, + "learning_rate": 1.32046783625731e-05, + "loss": 3.4924, + "step": 17871 + }, + { + "epoch": 0.9406315789473684, + "grad_norm": 0.54296875, + "learning_rate": 1.3192982456140352e-05, + "loss": 3.3058, + "step": 17872 + }, + { + "epoch": 0.9406842105263158, + "grad_norm": 0.490234375, + "learning_rate": 1.3181286549707603e-05, + "loss": 3.0304, + "step": 17873 + }, + { + "epoch": 0.9407368421052632, + "grad_norm": 0.51171875, + "learning_rate": 1.3169590643274855e-05, + "loss": 3.3624, + "step": 17874 + }, + { + "epoch": 0.9407894736842105, + "grad_norm": 0.53125, + "learning_rate": 1.3157894736842106e-05, + "loss": 3.6493, + "step": 17875 + }, + { + "epoch": 0.9408421052631579, + "grad_norm": 0.53515625, + "learning_rate": 1.3146198830409356e-05, + "loss": 3.6263, + "step": 17876 + }, + { + "epoch": 0.9408947368421052, + "grad_norm": 0.49609375, + "learning_rate": 1.3134502923976607e-05, + "loss": 3.3005, + "step": 17877 + }, + { + "epoch": 0.9409473684210526, + "grad_norm": 0.47265625, + "learning_rate": 1.3122807017543862e-05, + "loss": 3.2235, + "step": 17878 + }, + { + "epoch": 0.941, + "grad_norm": 0.478515625, + "learning_rate": 1.3111111111111113e-05, + "loss": 3.5047, + "step": 17879 + }, + { + "epoch": 0.9410526315789474, + "grad_norm": 0.50390625, + "learning_rate": 1.3099415204678364e-05, + "loss": 3.34, + "step": 17880 + }, + { + "epoch": 0.9411052631578948, + "grad_norm": 0.515625, + "learning_rate": 1.3087719298245615e-05, + "loss": 3.4458, + "step": 17881 + }, + { + "epoch": 0.9411578947368421, + "grad_norm": 0.49609375, + "learning_rate": 1.3076023391812867e-05, + "loss": 3.4403, + "step": 17882 + }, + { + "epoch": 0.9412105263157895, + "grad_norm": 0.482421875, + "learning_rate": 1.3064327485380118e-05, + "loss": 3.1623, + "step": 17883 + }, + { + "epoch": 0.9412631578947368, + "grad_norm": 0.52734375, + "learning_rate": 1.305263157894737e-05, + "loss": 3.2396, + "step": 17884 + }, + { + "epoch": 0.9413157894736842, + "grad_norm": 0.5, + "learning_rate": 1.304093567251462e-05, + "loss": 3.2531, + "step": 17885 + }, + { + "epoch": 0.9413684210526316, + "grad_norm": 0.51171875, + "learning_rate": 1.3029239766081872e-05, + "loss": 2.5047, + "step": 17886 + }, + { + "epoch": 0.9414210526315789, + "grad_norm": 0.48828125, + "learning_rate": 1.3017543859649123e-05, + "loss": 3.2929, + "step": 17887 + }, + { + "epoch": 0.9414736842105264, + "grad_norm": 0.515625, + "learning_rate": 1.3005847953216374e-05, + "loss": 3.3398, + "step": 17888 + }, + { + "epoch": 0.9415263157894737, + "grad_norm": 0.54296875, + "learning_rate": 1.2994152046783626e-05, + "loss": 3.1823, + "step": 17889 + }, + { + "epoch": 0.9415789473684211, + "grad_norm": 0.51953125, + "learning_rate": 1.2982456140350877e-05, + "loss": 3.225, + "step": 17890 + }, + { + "epoch": 0.9416315789473684, + "grad_norm": 0.5, + "learning_rate": 1.297076023391813e-05, + "loss": 3.7073, + "step": 17891 + }, + { + "epoch": 0.9416842105263158, + "grad_norm": 0.51171875, + "learning_rate": 1.2959064327485381e-05, + "loss": 2.8827, + "step": 17892 + }, + { + "epoch": 0.9417368421052632, + "grad_norm": 0.5234375, + "learning_rate": 1.2947368421052633e-05, + "loss": 3.0418, + "step": 17893 + }, + { + "epoch": 0.9417894736842105, + "grad_norm": 0.5234375, + "learning_rate": 1.2935672514619884e-05, + "loss": 3.0229, + "step": 17894 + }, + { + "epoch": 0.9418421052631579, + "grad_norm": 0.5, + "learning_rate": 1.2923976608187135e-05, + "loss": 3.2014, + "step": 17895 + }, + { + "epoch": 0.9418947368421052, + "grad_norm": 0.5, + "learning_rate": 1.2912280701754386e-05, + "loss": 3.4143, + "step": 17896 + }, + { + "epoch": 0.9419473684210526, + "grad_norm": 0.51953125, + "learning_rate": 1.2900584795321638e-05, + "loss": 3.4396, + "step": 17897 + }, + { + "epoch": 0.942, + "grad_norm": 0.4921875, + "learning_rate": 1.2888888888888889e-05, + "loss": 2.9499, + "step": 17898 + }, + { + "epoch": 0.9420526315789474, + "grad_norm": 0.48828125, + "learning_rate": 1.287719298245614e-05, + "loss": 3.2397, + "step": 17899 + }, + { + "epoch": 0.9421052631578948, + "grad_norm": 0.494140625, + "learning_rate": 1.2865497076023392e-05, + "loss": 3.2071, + "step": 17900 + }, + { + "epoch": 0.9421578947368421, + "grad_norm": 0.515625, + "learning_rate": 1.2853801169590643e-05, + "loss": 3.6369, + "step": 17901 + }, + { + "epoch": 0.9422105263157895, + "grad_norm": 0.5, + "learning_rate": 1.2842105263157894e-05, + "loss": 2.951, + "step": 17902 + }, + { + "epoch": 0.9422631578947368, + "grad_norm": 0.498046875, + "learning_rate": 1.2830409356725145e-05, + "loss": 3.0247, + "step": 17903 + }, + { + "epoch": 0.9423157894736842, + "grad_norm": 0.5078125, + "learning_rate": 1.28187134502924e-05, + "loss": 3.4643, + "step": 17904 + }, + { + "epoch": 0.9423684210526316, + "grad_norm": 0.49609375, + "learning_rate": 1.2807017543859651e-05, + "loss": 3.3626, + "step": 17905 + }, + { + "epoch": 0.9424210526315789, + "grad_norm": 0.486328125, + "learning_rate": 1.2795321637426903e-05, + "loss": 3.5071, + "step": 17906 + }, + { + "epoch": 0.9424736842105264, + "grad_norm": 0.515625, + "learning_rate": 1.2783625730994154e-05, + "loss": 3.3367, + "step": 17907 + }, + { + "epoch": 0.9425263157894737, + "grad_norm": 0.49609375, + "learning_rate": 1.2771929824561405e-05, + "loss": 3.5321, + "step": 17908 + }, + { + "epoch": 0.9425789473684211, + "grad_norm": 0.50390625, + "learning_rate": 1.2760233918128655e-05, + "loss": 3.0368, + "step": 17909 + }, + { + "epoch": 0.9426315789473684, + "grad_norm": 0.5078125, + "learning_rate": 1.2748538011695906e-05, + "loss": 3.2359, + "step": 17910 + }, + { + "epoch": 0.9426842105263158, + "grad_norm": 0.50390625, + "learning_rate": 1.2736842105263157e-05, + "loss": 2.944, + "step": 17911 + }, + { + "epoch": 0.9427368421052632, + "grad_norm": 0.5078125, + "learning_rate": 1.2725146198830409e-05, + "loss": 3.1515, + "step": 17912 + }, + { + "epoch": 0.9427894736842105, + "grad_norm": 0.5078125, + "learning_rate": 1.271345029239766e-05, + "loss": 3.0964, + "step": 17913 + }, + { + "epoch": 0.9428421052631579, + "grad_norm": 0.51171875, + "learning_rate": 1.2701754385964911e-05, + "loss": 2.986, + "step": 17914 + }, + { + "epoch": 0.9428947368421052, + "grad_norm": 0.494140625, + "learning_rate": 1.2690058479532163e-05, + "loss": 3.167, + "step": 17915 + }, + { + "epoch": 0.9429473684210526, + "grad_norm": 0.50390625, + "learning_rate": 1.2678362573099417e-05, + "loss": 3.3382, + "step": 17916 + }, + { + "epoch": 0.943, + "grad_norm": 0.498046875, + "learning_rate": 1.2666666666666668e-05, + "loss": 3.2973, + "step": 17917 + }, + { + "epoch": 0.9430526315789474, + "grad_norm": 0.5078125, + "learning_rate": 1.265497076023392e-05, + "loss": 3.9007, + "step": 17918 + }, + { + "epoch": 0.9431052631578948, + "grad_norm": 0.494140625, + "learning_rate": 1.2643274853801171e-05, + "loss": 3.1383, + "step": 17919 + }, + { + "epoch": 0.9431578947368421, + "grad_norm": 0.4921875, + "learning_rate": 1.2631578947368422e-05, + "loss": 3.2845, + "step": 17920 + }, + { + "epoch": 0.9432105263157895, + "grad_norm": 0.50390625, + "learning_rate": 1.2619883040935674e-05, + "loss": 3.1499, + "step": 17921 + }, + { + "epoch": 0.9432631578947368, + "grad_norm": 0.498046875, + "learning_rate": 1.2608187134502925e-05, + "loss": 3.2122, + "step": 17922 + }, + { + "epoch": 0.9433157894736842, + "grad_norm": 0.53125, + "learning_rate": 1.2596491228070176e-05, + "loss": 3.1327, + "step": 17923 + }, + { + "epoch": 0.9433684210526316, + "grad_norm": 0.490234375, + "learning_rate": 1.2584795321637428e-05, + "loss": 3.3358, + "step": 17924 + }, + { + "epoch": 0.9434210526315789, + "grad_norm": 0.5, + "learning_rate": 1.2573099415204679e-05, + "loss": 3.5552, + "step": 17925 + }, + { + "epoch": 0.9434736842105264, + "grad_norm": 0.51171875, + "learning_rate": 1.256140350877193e-05, + "loss": 3.0278, + "step": 17926 + }, + { + "epoch": 0.9435263157894737, + "grad_norm": 0.490234375, + "learning_rate": 1.2549707602339181e-05, + "loss": 3.1659, + "step": 17927 + }, + { + "epoch": 0.9435789473684211, + "grad_norm": 0.5546875, + "learning_rate": 1.2538011695906433e-05, + "loss": 3.3078, + "step": 17928 + }, + { + "epoch": 0.9436315789473684, + "grad_norm": 0.51953125, + "learning_rate": 1.2526315789473686e-05, + "loss": 2.9819, + "step": 17929 + }, + { + "epoch": 0.9436842105263158, + "grad_norm": 0.51171875, + "learning_rate": 1.2514619883040937e-05, + "loss": 3.3592, + "step": 17930 + }, + { + "epoch": 0.9437368421052632, + "grad_norm": 0.490234375, + "learning_rate": 1.2502923976608188e-05, + "loss": 3.2624, + "step": 17931 + }, + { + "epoch": 0.9437894736842105, + "grad_norm": 0.498046875, + "learning_rate": 1.249122807017544e-05, + "loss": 3.2696, + "step": 17932 + }, + { + "epoch": 0.9438421052631579, + "grad_norm": 0.51171875, + "learning_rate": 1.247953216374269e-05, + "loss": 3.3016, + "step": 17933 + }, + { + "epoch": 0.9438947368421052, + "grad_norm": 0.50390625, + "learning_rate": 1.2467836257309942e-05, + "loss": 3.5102, + "step": 17934 + }, + { + "epoch": 0.9439473684210526, + "grad_norm": 0.486328125, + "learning_rate": 1.2456140350877193e-05, + "loss": 3.1114, + "step": 17935 + }, + { + "epoch": 0.944, + "grad_norm": 0.51953125, + "learning_rate": 1.2444444444444445e-05, + "loss": 3.2082, + "step": 17936 + }, + { + "epoch": 0.9440526315789474, + "grad_norm": 0.51171875, + "learning_rate": 1.2432748538011696e-05, + "loss": 3.1497, + "step": 17937 + }, + { + "epoch": 0.9441052631578948, + "grad_norm": 0.4921875, + "learning_rate": 1.2421052631578949e-05, + "loss": 3.0453, + "step": 17938 + }, + { + "epoch": 0.9441578947368421, + "grad_norm": 0.52734375, + "learning_rate": 1.24093567251462e-05, + "loss": 3.0321, + "step": 17939 + }, + { + "epoch": 0.9442105263157895, + "grad_norm": 0.478515625, + "learning_rate": 1.2397660818713451e-05, + "loss": 3.3806, + "step": 17940 + }, + { + "epoch": 0.9442631578947368, + "grad_norm": 0.52734375, + "learning_rate": 1.2385964912280703e-05, + "loss": 3.1353, + "step": 17941 + }, + { + "epoch": 0.9443157894736842, + "grad_norm": 0.52734375, + "learning_rate": 1.2374269005847954e-05, + "loss": 3.648, + "step": 17942 + }, + { + "epoch": 0.9443684210526315, + "grad_norm": 0.50390625, + "learning_rate": 1.2362573099415205e-05, + "loss": 3.5517, + "step": 17943 + }, + { + "epoch": 0.9444210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.2350877192982457e-05, + "loss": 3.0632, + "step": 17944 + }, + { + "epoch": 0.9444736842105264, + "grad_norm": 0.494140625, + "learning_rate": 1.2339181286549708e-05, + "loss": 3.147, + "step": 17945 + }, + { + "epoch": 0.9445263157894737, + "grad_norm": 0.5, + "learning_rate": 1.232748538011696e-05, + "loss": 2.8817, + "step": 17946 + }, + { + "epoch": 0.9445789473684211, + "grad_norm": 0.50390625, + "learning_rate": 1.231578947368421e-05, + "loss": 2.834, + "step": 17947 + }, + { + "epoch": 0.9446315789473684, + "grad_norm": 0.49609375, + "learning_rate": 1.2304093567251462e-05, + "loss": 3.6661, + "step": 17948 + }, + { + "epoch": 0.9446842105263158, + "grad_norm": 0.4609375, + "learning_rate": 1.2292397660818713e-05, + "loss": 2.8902, + "step": 17949 + }, + { + "epoch": 0.9447368421052632, + "grad_norm": 0.4921875, + "learning_rate": 1.2280701754385964e-05, + "loss": 3.5941, + "step": 17950 + }, + { + "epoch": 0.9447894736842105, + "grad_norm": 0.51171875, + "learning_rate": 1.2269005847953217e-05, + "loss": 2.9475, + "step": 17951 + }, + { + "epoch": 0.9448421052631579, + "grad_norm": 0.51171875, + "learning_rate": 1.2257309941520469e-05, + "loss": 3.2469, + "step": 17952 + }, + { + "epoch": 0.9448947368421052, + "grad_norm": 0.52734375, + "learning_rate": 1.224561403508772e-05, + "loss": 3.2431, + "step": 17953 + }, + { + "epoch": 0.9449473684210526, + "grad_norm": 0.47265625, + "learning_rate": 1.2233918128654971e-05, + "loss": 3.1094, + "step": 17954 + }, + { + "epoch": 0.945, + "grad_norm": 0.50390625, + "learning_rate": 1.2222222222222222e-05, + "loss": 3.481, + "step": 17955 + }, + { + "epoch": 0.9450526315789474, + "grad_norm": 0.515625, + "learning_rate": 1.2210526315789474e-05, + "loss": 3.3963, + "step": 17956 + }, + { + "epoch": 0.9451052631578948, + "grad_norm": 0.515625, + "learning_rate": 1.2198830409356727e-05, + "loss": 3.2616, + "step": 17957 + }, + { + "epoch": 0.9451578947368421, + "grad_norm": 0.4921875, + "learning_rate": 1.2187134502923978e-05, + "loss": 2.9887, + "step": 17958 + }, + { + "epoch": 0.9452105263157895, + "grad_norm": 0.490234375, + "learning_rate": 1.217543859649123e-05, + "loss": 3.1712, + "step": 17959 + }, + { + "epoch": 0.9452631578947368, + "grad_norm": 0.5078125, + "learning_rate": 1.216374269005848e-05, + "loss": 2.8836, + "step": 17960 + }, + { + "epoch": 0.9453157894736842, + "grad_norm": 0.55078125, + "learning_rate": 1.2152046783625732e-05, + "loss": 3.0489, + "step": 17961 + }, + { + "epoch": 0.9453684210526315, + "grad_norm": 0.5078125, + "learning_rate": 1.2140350877192983e-05, + "loss": 3.2447, + "step": 17962 + }, + { + "epoch": 0.9454210526315789, + "grad_norm": 0.5, + "learning_rate": 1.2128654970760234e-05, + "loss": 3.0889, + "step": 17963 + }, + { + "epoch": 0.9454736842105264, + "grad_norm": 0.50390625, + "learning_rate": 1.2116959064327486e-05, + "loss": 2.7766, + "step": 17964 + }, + { + "epoch": 0.9455263157894737, + "grad_norm": 0.52734375, + "learning_rate": 1.2105263157894737e-05, + "loss": 3.4238, + "step": 17965 + }, + { + "epoch": 0.9455789473684211, + "grad_norm": 0.51171875, + "learning_rate": 1.2093567251461988e-05, + "loss": 3.4615, + "step": 17966 + }, + { + "epoch": 0.9456315789473684, + "grad_norm": 0.494140625, + "learning_rate": 1.208187134502924e-05, + "loss": 3.5796, + "step": 17967 + }, + { + "epoch": 0.9456842105263158, + "grad_norm": 0.498046875, + "learning_rate": 1.2070175438596491e-05, + "loss": 2.8808, + "step": 17968 + }, + { + "epoch": 0.9457368421052632, + "grad_norm": 0.49609375, + "learning_rate": 1.2058479532163744e-05, + "loss": 2.7626, + "step": 17969 + }, + { + "epoch": 0.9457894736842105, + "grad_norm": 0.48828125, + "learning_rate": 1.2046783625730995e-05, + "loss": 3.4154, + "step": 17970 + }, + { + "epoch": 0.9458421052631579, + "grad_norm": 0.5, + "learning_rate": 1.2035087719298246e-05, + "loss": 3.2282, + "step": 17971 + }, + { + "epoch": 0.9458947368421052, + "grad_norm": 0.5078125, + "learning_rate": 1.2023391812865498e-05, + "loss": 2.7435, + "step": 17972 + }, + { + "epoch": 0.9459473684210526, + "grad_norm": 0.51171875, + "learning_rate": 1.2011695906432749e-05, + "loss": 3.091, + "step": 17973 + }, + { + "epoch": 0.946, + "grad_norm": 0.49609375, + "learning_rate": 1.2e-05, + "loss": 2.7192, + "step": 17974 + }, + { + "epoch": 0.9460526315789474, + "grad_norm": 0.490234375, + "learning_rate": 1.1988304093567252e-05, + "loss": 3.5627, + "step": 17975 + }, + { + "epoch": 0.9461052631578948, + "grad_norm": 0.5234375, + "learning_rate": 1.1976608187134505e-05, + "loss": 3.2294, + "step": 17976 + }, + { + "epoch": 0.9461578947368421, + "grad_norm": 0.49609375, + "learning_rate": 1.1964912280701756e-05, + "loss": 3.2643, + "step": 17977 + }, + { + "epoch": 0.9462105263157895, + "grad_norm": 0.486328125, + "learning_rate": 1.1953216374269007e-05, + "loss": 3.4582, + "step": 17978 + }, + { + "epoch": 0.9462631578947368, + "grad_norm": 0.48828125, + "learning_rate": 1.1941520467836258e-05, + "loss": 3.3472, + "step": 17979 + }, + { + "epoch": 0.9463157894736842, + "grad_norm": 0.6328125, + "learning_rate": 1.192982456140351e-05, + "loss": 3.4679, + "step": 17980 + }, + { + "epoch": 0.9463684210526315, + "grad_norm": 0.5078125, + "learning_rate": 1.1918128654970761e-05, + "loss": 3.1137, + "step": 17981 + }, + { + "epoch": 0.9464210526315789, + "grad_norm": 0.462890625, + "learning_rate": 1.1906432748538012e-05, + "loss": 3.007, + "step": 17982 + }, + { + "epoch": 0.9464736842105264, + "grad_norm": 0.478515625, + "learning_rate": 1.1894736842105264e-05, + "loss": 3.1414, + "step": 17983 + }, + { + "epoch": 0.9465263157894737, + "grad_norm": 0.498046875, + "learning_rate": 1.1883040935672515e-05, + "loss": 3.4767, + "step": 17984 + }, + { + "epoch": 0.9465789473684211, + "grad_norm": 0.4921875, + "learning_rate": 1.1871345029239766e-05, + "loss": 3.3324, + "step": 17985 + }, + { + "epoch": 0.9466315789473684, + "grad_norm": 0.51171875, + "learning_rate": 1.1859649122807017e-05, + "loss": 3.1503, + "step": 17986 + }, + { + "epoch": 0.9466842105263158, + "grad_norm": 0.53125, + "learning_rate": 1.1847953216374269e-05, + "loss": 3.0761, + "step": 17987 + }, + { + "epoch": 0.9467368421052632, + "grad_norm": 0.51171875, + "learning_rate": 1.1836257309941522e-05, + "loss": 3.2061, + "step": 17988 + }, + { + "epoch": 0.9467894736842105, + "grad_norm": 0.515625, + "learning_rate": 1.1824561403508773e-05, + "loss": 3.0889, + "step": 17989 + }, + { + "epoch": 0.9468421052631579, + "grad_norm": 0.48046875, + "learning_rate": 1.1812865497076024e-05, + "loss": 3.2379, + "step": 17990 + }, + { + "epoch": 0.9468947368421052, + "grad_norm": 0.50390625, + "learning_rate": 1.1801169590643276e-05, + "loss": 3.4331, + "step": 17991 + }, + { + "epoch": 0.9469473684210526, + "grad_norm": 0.494140625, + "learning_rate": 1.1789473684210527e-05, + "loss": 3.4565, + "step": 17992 + }, + { + "epoch": 0.947, + "grad_norm": 0.546875, + "learning_rate": 1.1777777777777778e-05, + "loss": 3.431, + "step": 17993 + }, + { + "epoch": 0.9470526315789474, + "grad_norm": 0.50390625, + "learning_rate": 1.1766081871345031e-05, + "loss": 3.5203, + "step": 17994 + }, + { + "epoch": 0.9471052631578948, + "grad_norm": 0.5, + "learning_rate": 1.1754385964912282e-05, + "loss": 3.1866, + "step": 17995 + }, + { + "epoch": 0.9471578947368421, + "grad_norm": 0.48828125, + "learning_rate": 1.1742690058479534e-05, + "loss": 3.1028, + "step": 17996 + }, + { + "epoch": 0.9472105263157895, + "grad_norm": 0.490234375, + "learning_rate": 1.1730994152046785e-05, + "loss": 2.7231, + "step": 17997 + }, + { + "epoch": 0.9472631578947368, + "grad_norm": 0.51953125, + "learning_rate": 1.1719298245614036e-05, + "loss": 3.1236, + "step": 17998 + }, + { + "epoch": 0.9473157894736842, + "grad_norm": 0.5234375, + "learning_rate": 1.1707602339181286e-05, + "loss": 3.2321, + "step": 17999 + }, + { + "epoch": 0.9473684210526315, + "grad_norm": 0.53515625, + "learning_rate": 1.1695906432748537e-05, + "loss": 3.2355, + "step": 18000 + }, + { + "epoch": 0.9474210526315789, + "grad_norm": 0.515625, + "learning_rate": 1.168421052631579e-05, + "loss": 3.5173, + "step": 18001 + }, + { + "epoch": 0.9474736842105264, + "grad_norm": 0.5078125, + "learning_rate": 1.1672514619883041e-05, + "loss": 3.0477, + "step": 18002 + }, + { + "epoch": 0.9475263157894737, + "grad_norm": 0.5, + "learning_rate": 1.1660818713450293e-05, + "loss": 3.5164, + "step": 18003 + }, + { + "epoch": 0.9475789473684211, + "grad_norm": 0.52734375, + "learning_rate": 1.1649122807017544e-05, + "loss": 3.0023, + "step": 18004 + }, + { + "epoch": 0.9476315789473684, + "grad_norm": 0.51171875, + "learning_rate": 1.1637426900584795e-05, + "loss": 3.397, + "step": 18005 + }, + { + "epoch": 0.9476842105263158, + "grad_norm": 0.494140625, + "learning_rate": 1.1625730994152047e-05, + "loss": 3.1799, + "step": 18006 + }, + { + "epoch": 0.9477368421052632, + "grad_norm": 0.49609375, + "learning_rate": 1.16140350877193e-05, + "loss": 3.0373, + "step": 18007 + }, + { + "epoch": 0.9477894736842105, + "grad_norm": 0.5, + "learning_rate": 1.160233918128655e-05, + "loss": 3.3714, + "step": 18008 + }, + { + "epoch": 0.9478421052631579, + "grad_norm": 0.5078125, + "learning_rate": 1.1590643274853802e-05, + "loss": 2.7774, + "step": 18009 + }, + { + "epoch": 0.9478947368421052, + "grad_norm": 0.470703125, + "learning_rate": 1.1578947368421053e-05, + "loss": 2.9164, + "step": 18010 + }, + { + "epoch": 0.9479473684210526, + "grad_norm": 0.515625, + "learning_rate": 1.1567251461988305e-05, + "loss": 3.243, + "step": 18011 + }, + { + "epoch": 0.948, + "grad_norm": 0.51953125, + "learning_rate": 1.1555555555555556e-05, + "loss": 3.0879, + "step": 18012 + }, + { + "epoch": 0.9480526315789474, + "grad_norm": 0.50390625, + "learning_rate": 1.1543859649122809e-05, + "loss": 3.1919, + "step": 18013 + }, + { + "epoch": 0.9481052631578948, + "grad_norm": 0.53515625, + "learning_rate": 1.153216374269006e-05, + "loss": 2.9789, + "step": 18014 + }, + { + "epoch": 0.9481578947368421, + "grad_norm": 0.5078125, + "learning_rate": 1.1520467836257312e-05, + "loss": 3.2657, + "step": 18015 + }, + { + "epoch": 0.9482105263157895, + "grad_norm": 0.5078125, + "learning_rate": 1.1508771929824561e-05, + "loss": 3.3391, + "step": 18016 + }, + { + "epoch": 0.9482631578947368, + "grad_norm": 0.5078125, + "learning_rate": 1.1497076023391812e-05, + "loss": 3.2196, + "step": 18017 + }, + { + "epoch": 0.9483157894736842, + "grad_norm": 0.482421875, + "learning_rate": 1.1485380116959064e-05, + "loss": 3.1705, + "step": 18018 + }, + { + "epoch": 0.9483684210526315, + "grad_norm": 0.48828125, + "learning_rate": 1.1473684210526315e-05, + "loss": 3.2995, + "step": 18019 + }, + { + "epoch": 0.9484210526315789, + "grad_norm": 0.490234375, + "learning_rate": 1.1461988304093568e-05, + "loss": 3.3594, + "step": 18020 + }, + { + "epoch": 0.9484736842105264, + "grad_norm": 0.47265625, + "learning_rate": 1.145029239766082e-05, + "loss": 2.8986, + "step": 18021 + }, + { + "epoch": 0.9485263157894737, + "grad_norm": 0.490234375, + "learning_rate": 1.143859649122807e-05, + "loss": 3.2522, + "step": 18022 + }, + { + "epoch": 0.9485789473684211, + "grad_norm": 0.484375, + "learning_rate": 1.1426900584795322e-05, + "loss": 3.3958, + "step": 18023 + }, + { + "epoch": 0.9486315789473684, + "grad_norm": 0.515625, + "learning_rate": 1.1415204678362573e-05, + "loss": 3.2882, + "step": 18024 + }, + { + "epoch": 0.9486842105263158, + "grad_norm": 0.48828125, + "learning_rate": 1.1403508771929824e-05, + "loss": 3.3951, + "step": 18025 + }, + { + "epoch": 0.9487368421052632, + "grad_norm": 0.5, + "learning_rate": 1.1391812865497077e-05, + "loss": 3.1004, + "step": 18026 + }, + { + "epoch": 0.9487894736842105, + "grad_norm": 0.5, + "learning_rate": 1.1380116959064329e-05, + "loss": 3.2555, + "step": 18027 + }, + { + "epoch": 0.9488421052631579, + "grad_norm": 0.494140625, + "learning_rate": 1.136842105263158e-05, + "loss": 3.3642, + "step": 18028 + }, + { + "epoch": 0.9488947368421052, + "grad_norm": 0.48828125, + "learning_rate": 1.1356725146198831e-05, + "loss": 3.1116, + "step": 18029 + }, + { + "epoch": 0.9489473684210527, + "grad_norm": 0.51171875, + "learning_rate": 1.1345029239766083e-05, + "loss": 3.1918, + "step": 18030 + }, + { + "epoch": 0.949, + "grad_norm": 0.494140625, + "learning_rate": 1.1333333333333334e-05, + "loss": 3.1937, + "step": 18031 + }, + { + "epoch": 0.9490526315789474, + "grad_norm": 0.5234375, + "learning_rate": 1.1321637426900585e-05, + "loss": 3.4845, + "step": 18032 + }, + { + "epoch": 0.9491052631578948, + "grad_norm": 0.51171875, + "learning_rate": 1.1309941520467836e-05, + "loss": 3.4815, + "step": 18033 + }, + { + "epoch": 0.9491578947368421, + "grad_norm": 0.486328125, + "learning_rate": 1.1298245614035088e-05, + "loss": 3.0765, + "step": 18034 + }, + { + "epoch": 0.9492105263157895, + "grad_norm": 0.4921875, + "learning_rate": 1.1286549707602339e-05, + "loss": 2.863, + "step": 18035 + }, + { + "epoch": 0.9492631578947368, + "grad_norm": 0.546875, + "learning_rate": 1.127485380116959e-05, + "loss": 3.215, + "step": 18036 + }, + { + "epoch": 0.9493157894736842, + "grad_norm": 0.51171875, + "learning_rate": 1.1263157894736842e-05, + "loss": 3.4708, + "step": 18037 + }, + { + "epoch": 0.9493684210526315, + "grad_norm": 0.48046875, + "learning_rate": 1.1251461988304095e-05, + "loss": 3.0808, + "step": 18038 + }, + { + "epoch": 0.9494210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.1239766081871346e-05, + "loss": 2.9475, + "step": 18039 + }, + { + "epoch": 0.9494736842105264, + "grad_norm": 0.49609375, + "learning_rate": 1.1228070175438597e-05, + "loss": 2.9817, + "step": 18040 + }, + { + "epoch": 0.9495263157894737, + "grad_norm": 0.4921875, + "learning_rate": 1.1216374269005848e-05, + "loss": 3.3119, + "step": 18041 + }, + { + "epoch": 0.9495789473684211, + "grad_norm": 0.5, + "learning_rate": 1.12046783625731e-05, + "loss": 3.0489, + "step": 18042 + }, + { + "epoch": 0.9496315789473684, + "grad_norm": 0.49609375, + "learning_rate": 1.1192982456140351e-05, + "loss": 3.3928, + "step": 18043 + }, + { + "epoch": 0.9496842105263158, + "grad_norm": 0.5078125, + "learning_rate": 1.1181286549707602e-05, + "loss": 3.3488, + "step": 18044 + }, + { + "epoch": 0.9497368421052632, + "grad_norm": 0.50390625, + "learning_rate": 1.1169590643274855e-05, + "loss": 3.1803, + "step": 18045 + }, + { + "epoch": 0.9497894736842105, + "grad_norm": 0.5390625, + "learning_rate": 1.1157894736842106e-05, + "loss": 2.8362, + "step": 18046 + }, + { + "epoch": 0.9498421052631579, + "grad_norm": 0.486328125, + "learning_rate": 1.1146198830409358e-05, + "loss": 3.3422, + "step": 18047 + }, + { + "epoch": 0.9498947368421052, + "grad_norm": 0.484375, + "learning_rate": 1.1134502923976609e-05, + "loss": 3.4074, + "step": 18048 + }, + { + "epoch": 0.9499473684210527, + "grad_norm": 0.515625, + "learning_rate": 1.112280701754386e-05, + "loss": 3.5765, + "step": 18049 + }, + { + "epoch": 0.95, + "grad_norm": 0.52734375, + "learning_rate": 1.1111111111111112e-05, + "loss": 3.5225, + "step": 18050 + }, + { + "epoch": 0.9500526315789474, + "grad_norm": 0.5078125, + "learning_rate": 1.1099415204678363e-05, + "loss": 3.4492, + "step": 18051 + }, + { + "epoch": 0.9501052631578948, + "grad_norm": 0.5, + "learning_rate": 1.1087719298245614e-05, + "loss": 3.5544, + "step": 18052 + }, + { + "epoch": 0.9501578947368421, + "grad_norm": 0.515625, + "learning_rate": 1.1076023391812866e-05, + "loss": 3.2365, + "step": 18053 + }, + { + "epoch": 0.9502105263157895, + "grad_norm": 0.484375, + "learning_rate": 1.1064327485380117e-05, + "loss": 3.0666, + "step": 18054 + }, + { + "epoch": 0.9502631578947368, + "grad_norm": 0.5078125, + "learning_rate": 1.1052631578947368e-05, + "loss": 3.5122, + "step": 18055 + }, + { + "epoch": 0.9503157894736842, + "grad_norm": 0.478515625, + "learning_rate": 1.104093567251462e-05, + "loss": 2.6121, + "step": 18056 + }, + { + "epoch": 0.9503684210526315, + "grad_norm": 0.51171875, + "learning_rate": 1.1029239766081872e-05, + "loss": 3.5162, + "step": 18057 + }, + { + "epoch": 0.9504210526315789, + "grad_norm": 0.50390625, + "learning_rate": 1.1017543859649124e-05, + "loss": 3.441, + "step": 18058 + }, + { + "epoch": 0.9504736842105264, + "grad_norm": 0.494140625, + "learning_rate": 1.1005847953216375e-05, + "loss": 2.898, + "step": 18059 + }, + { + "epoch": 0.9505263157894737, + "grad_norm": 1.171875, + "learning_rate": 1.0994152046783626e-05, + "loss": 3.3761, + "step": 18060 + }, + { + "epoch": 0.9505789473684211, + "grad_norm": 0.51171875, + "learning_rate": 1.0982456140350877e-05, + "loss": 3.4148, + "step": 18061 + }, + { + "epoch": 0.9506315789473684, + "grad_norm": 0.5078125, + "learning_rate": 1.0970760233918129e-05, + "loss": 3.0159, + "step": 18062 + }, + { + "epoch": 0.9506842105263158, + "grad_norm": 0.515625, + "learning_rate": 1.0959064327485382e-05, + "loss": 3.3286, + "step": 18063 + }, + { + "epoch": 0.9507368421052631, + "grad_norm": 0.49609375, + "learning_rate": 1.0947368421052633e-05, + "loss": 3.9067, + "step": 18064 + }, + { + "epoch": 0.9507894736842105, + "grad_norm": 0.5, + "learning_rate": 1.0935672514619884e-05, + "loss": 3.2295, + "step": 18065 + }, + { + "epoch": 0.9508421052631579, + "grad_norm": 0.49609375, + "learning_rate": 1.0923976608187136e-05, + "loss": 3.27, + "step": 18066 + }, + { + "epoch": 0.9508947368421052, + "grad_norm": 0.5, + "learning_rate": 1.0912280701754387e-05, + "loss": 3.4827, + "step": 18067 + }, + { + "epoch": 0.9509473684210527, + "grad_norm": 0.4765625, + "learning_rate": 1.0900584795321638e-05, + "loss": 2.8108, + "step": 18068 + }, + { + "epoch": 0.951, + "grad_norm": 0.5078125, + "learning_rate": 1.088888888888889e-05, + "loss": 2.8815, + "step": 18069 + }, + { + "epoch": 0.9510526315789474, + "grad_norm": 0.5, + "learning_rate": 1.087719298245614e-05, + "loss": 3.1266, + "step": 18070 + }, + { + "epoch": 0.9511052631578948, + "grad_norm": 0.515625, + "learning_rate": 1.0865497076023392e-05, + "loss": 3.0565, + "step": 18071 + }, + { + "epoch": 0.9511578947368421, + "grad_norm": 0.50390625, + "learning_rate": 1.0853801169590643e-05, + "loss": 3.0865, + "step": 18072 + }, + { + "epoch": 0.9512105263157895, + "grad_norm": 0.53515625, + "learning_rate": 1.0842105263157895e-05, + "loss": 2.9079, + "step": 18073 + }, + { + "epoch": 0.9512631578947368, + "grad_norm": 0.50390625, + "learning_rate": 1.0830409356725146e-05, + "loss": 3.337, + "step": 18074 + }, + { + "epoch": 0.9513157894736842, + "grad_norm": 0.52734375, + "learning_rate": 1.0818713450292397e-05, + "loss": 3.6795, + "step": 18075 + }, + { + "epoch": 0.9513684210526315, + "grad_norm": 0.53125, + "learning_rate": 1.080701754385965e-05, + "loss": 3.5303, + "step": 18076 + }, + { + "epoch": 0.951421052631579, + "grad_norm": 0.5078125, + "learning_rate": 1.0795321637426901e-05, + "loss": 3.4105, + "step": 18077 + }, + { + "epoch": 0.9514736842105264, + "grad_norm": 0.51171875, + "learning_rate": 1.0783625730994153e-05, + "loss": 3.0958, + "step": 18078 + }, + { + "epoch": 0.9515263157894737, + "grad_norm": 0.4921875, + "learning_rate": 1.0771929824561404e-05, + "loss": 3.2926, + "step": 18079 + }, + { + "epoch": 0.9515789473684211, + "grad_norm": 0.494140625, + "learning_rate": 1.0760233918128655e-05, + "loss": 3.3227, + "step": 18080 + }, + { + "epoch": 0.9516315789473684, + "grad_norm": 0.51171875, + "learning_rate": 1.0748538011695907e-05, + "loss": 3.2973, + "step": 18081 + }, + { + "epoch": 0.9516842105263158, + "grad_norm": 0.498046875, + "learning_rate": 1.073684210526316e-05, + "loss": 3.157, + "step": 18082 + }, + { + "epoch": 0.9517368421052631, + "grad_norm": 0.48828125, + "learning_rate": 1.0725146198830411e-05, + "loss": 3.1303, + "step": 18083 + }, + { + "epoch": 0.9517894736842105, + "grad_norm": 0.462890625, + "learning_rate": 1.0713450292397662e-05, + "loss": 3.0698, + "step": 18084 + }, + { + "epoch": 0.9518421052631579, + "grad_norm": 0.53515625, + "learning_rate": 1.0701754385964913e-05, + "loss": 3.0789, + "step": 18085 + }, + { + "epoch": 0.9518947368421052, + "grad_norm": 0.49609375, + "learning_rate": 1.0690058479532165e-05, + "loss": 3.0277, + "step": 18086 + }, + { + "epoch": 0.9519473684210527, + "grad_norm": 0.51953125, + "learning_rate": 1.0678362573099416e-05, + "loss": 3.4491, + "step": 18087 + }, + { + "epoch": 0.952, + "grad_norm": 0.490234375, + "learning_rate": 1.0666666666666667e-05, + "loss": 2.798, + "step": 18088 + }, + { + "epoch": 0.9520526315789474, + "grad_norm": 0.5390625, + "learning_rate": 1.0654970760233919e-05, + "loss": 2.9056, + "step": 18089 + }, + { + "epoch": 0.9521052631578948, + "grad_norm": 0.46875, + "learning_rate": 1.064327485380117e-05, + "loss": 3.0878, + "step": 18090 + }, + { + "epoch": 0.9521578947368421, + "grad_norm": 0.515625, + "learning_rate": 1.0631578947368421e-05, + "loss": 3.6541, + "step": 18091 + }, + { + "epoch": 0.9522105263157895, + "grad_norm": 0.53515625, + "learning_rate": 1.0619883040935672e-05, + "loss": 3.5103, + "step": 18092 + }, + { + "epoch": 0.9522631578947368, + "grad_norm": 0.53515625, + "learning_rate": 1.0608187134502924e-05, + "loss": 2.8437, + "step": 18093 + }, + { + "epoch": 0.9523157894736842, + "grad_norm": 0.474609375, + "learning_rate": 1.0596491228070175e-05, + "loss": 3.1042, + "step": 18094 + }, + { + "epoch": 0.9523684210526315, + "grad_norm": 0.515625, + "learning_rate": 1.0584795321637428e-05, + "loss": 3.273, + "step": 18095 + }, + { + "epoch": 0.952421052631579, + "grad_norm": 0.494140625, + "learning_rate": 1.057309941520468e-05, + "loss": 3.2388, + "step": 18096 + }, + { + "epoch": 0.9524736842105264, + "grad_norm": 0.50390625, + "learning_rate": 1.056140350877193e-05, + "loss": 3.0192, + "step": 18097 + }, + { + "epoch": 0.9525263157894737, + "grad_norm": 0.490234375, + "learning_rate": 1.0549707602339182e-05, + "loss": 3.3895, + "step": 18098 + }, + { + "epoch": 0.9525789473684211, + "grad_norm": 0.51171875, + "learning_rate": 1.0538011695906433e-05, + "loss": 3.2427, + "step": 18099 + }, + { + "epoch": 0.9526315789473684, + "grad_norm": 0.48828125, + "learning_rate": 1.0526315789473684e-05, + "loss": 3.3376, + "step": 18100 + }, + { + "epoch": 0.9526842105263158, + "grad_norm": 0.498046875, + "learning_rate": 1.0514619883040937e-05, + "loss": 2.5994, + "step": 18101 + }, + { + "epoch": 0.9527368421052631, + "grad_norm": 0.490234375, + "learning_rate": 1.0502923976608189e-05, + "loss": 3.0593, + "step": 18102 + }, + { + "epoch": 0.9527894736842105, + "grad_norm": 0.50390625, + "learning_rate": 1.049122807017544e-05, + "loss": 3.5233, + "step": 18103 + }, + { + "epoch": 0.9528421052631579, + "grad_norm": 0.486328125, + "learning_rate": 1.0479532163742691e-05, + "loss": 3.0435, + "step": 18104 + }, + { + "epoch": 0.9528947368421052, + "grad_norm": 0.48046875, + "learning_rate": 1.0467836257309941e-05, + "loss": 3.364, + "step": 18105 + }, + { + "epoch": 0.9529473684210527, + "grad_norm": 0.50390625, + "learning_rate": 1.0456140350877192e-05, + "loss": 3.4211, + "step": 18106 + }, + { + "epoch": 0.953, + "grad_norm": 0.50390625, + "learning_rate": 1.0444444444444445e-05, + "loss": 3.2398, + "step": 18107 + }, + { + "epoch": 0.9530526315789474, + "grad_norm": 0.52734375, + "learning_rate": 1.0432748538011696e-05, + "loss": 2.9333, + "step": 18108 + }, + { + "epoch": 0.9531052631578948, + "grad_norm": 0.5234375, + "learning_rate": 1.0421052631578948e-05, + "loss": 3.4026, + "step": 18109 + }, + { + "epoch": 0.9531578947368421, + "grad_norm": 0.5, + "learning_rate": 1.0409356725146199e-05, + "loss": 3.2539, + "step": 18110 + }, + { + "epoch": 0.9532105263157895, + "grad_norm": 0.50390625, + "learning_rate": 1.039766081871345e-05, + "loss": 3.2356, + "step": 18111 + }, + { + "epoch": 0.9532631578947368, + "grad_norm": 0.5, + "learning_rate": 1.0385964912280702e-05, + "loss": 3.5023, + "step": 18112 + }, + { + "epoch": 0.9533157894736842, + "grad_norm": 0.5, + "learning_rate": 1.0374269005847953e-05, + "loss": 3.5098, + "step": 18113 + }, + { + "epoch": 0.9533684210526315, + "grad_norm": 0.53125, + "learning_rate": 1.0362573099415206e-05, + "loss": 3.2061, + "step": 18114 + }, + { + "epoch": 0.953421052631579, + "grad_norm": 0.5078125, + "learning_rate": 1.0350877192982457e-05, + "loss": 3.6146, + "step": 18115 + }, + { + "epoch": 0.9534736842105264, + "grad_norm": 0.51171875, + "learning_rate": 1.0339181286549708e-05, + "loss": 3.2196, + "step": 18116 + }, + { + "epoch": 0.9535263157894737, + "grad_norm": 0.5078125, + "learning_rate": 1.032748538011696e-05, + "loss": 2.9117, + "step": 18117 + }, + { + "epoch": 0.9535789473684211, + "grad_norm": 0.494140625, + "learning_rate": 1.0315789473684211e-05, + "loss": 3.3993, + "step": 18118 + }, + { + "epoch": 0.9536315789473684, + "grad_norm": 0.5078125, + "learning_rate": 1.0304093567251462e-05, + "loss": 3.4543, + "step": 18119 + }, + { + "epoch": 0.9536842105263158, + "grad_norm": 0.4921875, + "learning_rate": 1.0292397660818715e-05, + "loss": 3.0659, + "step": 18120 + }, + { + "epoch": 0.9537368421052631, + "grad_norm": 0.5, + "learning_rate": 1.0280701754385967e-05, + "loss": 3.3334, + "step": 18121 + }, + { + "epoch": 0.9537894736842105, + "grad_norm": 0.5078125, + "learning_rate": 1.0269005847953216e-05, + "loss": 3.2347, + "step": 18122 + }, + { + "epoch": 0.9538421052631579, + "grad_norm": 0.6171875, + "learning_rate": 1.0257309941520467e-05, + "loss": 3.4646, + "step": 18123 + }, + { + "epoch": 0.9538947368421052, + "grad_norm": 0.46484375, + "learning_rate": 1.0245614035087719e-05, + "loss": 3.544, + "step": 18124 + }, + { + "epoch": 0.9539473684210527, + "grad_norm": 0.52734375, + "learning_rate": 1.023391812865497e-05, + "loss": 3.0231, + "step": 18125 + }, + { + "epoch": 0.954, + "grad_norm": 0.5, + "learning_rate": 1.0222222222222223e-05, + "loss": 3.3371, + "step": 18126 + }, + { + "epoch": 0.9540526315789474, + "grad_norm": 0.5, + "learning_rate": 1.0210526315789474e-05, + "loss": 3.6096, + "step": 18127 + }, + { + "epoch": 0.9541052631578948, + "grad_norm": 0.515625, + "learning_rate": 1.0198830409356726e-05, + "loss": 3.3472, + "step": 18128 + }, + { + "epoch": 0.9541578947368421, + "grad_norm": 0.51171875, + "learning_rate": 1.0187134502923977e-05, + "loss": 3.1051, + "step": 18129 + }, + { + "epoch": 0.9542105263157895, + "grad_norm": 0.5078125, + "learning_rate": 1.0175438596491228e-05, + "loss": 3.4277, + "step": 18130 + }, + { + "epoch": 0.9542631578947368, + "grad_norm": 0.51171875, + "learning_rate": 1.016374269005848e-05, + "loss": 3.3671, + "step": 18131 + }, + { + "epoch": 0.9543157894736842, + "grad_norm": 0.51171875, + "learning_rate": 1.0152046783625732e-05, + "loss": 3.2751, + "step": 18132 + }, + { + "epoch": 0.9543684210526315, + "grad_norm": 0.5234375, + "learning_rate": 1.0140350877192984e-05, + "loss": 2.9838, + "step": 18133 + }, + { + "epoch": 0.954421052631579, + "grad_norm": 0.5234375, + "learning_rate": 1.0128654970760235e-05, + "loss": 3.4655, + "step": 18134 + }, + { + "epoch": 0.9544736842105264, + "grad_norm": 0.50390625, + "learning_rate": 1.0116959064327486e-05, + "loss": 3.163, + "step": 18135 + }, + { + "epoch": 0.9545263157894737, + "grad_norm": 0.5390625, + "learning_rate": 1.0105263157894738e-05, + "loss": 2.8582, + "step": 18136 + }, + { + "epoch": 0.9545789473684211, + "grad_norm": 0.5, + "learning_rate": 1.0093567251461989e-05, + "loss": 3.0938, + "step": 18137 + }, + { + "epoch": 0.9546315789473684, + "grad_norm": 0.48046875, + "learning_rate": 1.008187134502924e-05, + "loss": 2.9576, + "step": 18138 + }, + { + "epoch": 0.9546842105263158, + "grad_norm": 0.50390625, + "learning_rate": 1.0070175438596491e-05, + "loss": 3.1557, + "step": 18139 + }, + { + "epoch": 0.9547368421052631, + "grad_norm": 0.50390625, + "learning_rate": 1.0058479532163743e-05, + "loss": 3.2886, + "step": 18140 + }, + { + "epoch": 0.9547894736842105, + "grad_norm": 0.53515625, + "learning_rate": 1.0046783625730994e-05, + "loss": 3.1237, + "step": 18141 + }, + { + "epoch": 0.9548421052631579, + "grad_norm": 0.55078125, + "learning_rate": 1.0035087719298245e-05, + "loss": 3.4001, + "step": 18142 + }, + { + "epoch": 0.9548947368421052, + "grad_norm": 0.51953125, + "learning_rate": 1.0023391812865497e-05, + "loss": 2.8707, + "step": 18143 + }, + { + "epoch": 0.9549473684210527, + "grad_norm": 0.5078125, + "learning_rate": 1.0011695906432748e-05, + "loss": 3.5526, + "step": 18144 + }, + { + "epoch": 0.955, + "grad_norm": 0.50390625, + "learning_rate": 1e-05, + "loss": 3.2795, + "step": 18145 + }, + { + "epoch": 0.9550526315789474, + "grad_norm": 0.51171875, + "learning_rate": 9.988304093567252e-06, + "loss": 3.1319, + "step": 18146 + }, + { + "epoch": 0.9551052631578948, + "grad_norm": 0.51171875, + "learning_rate": 9.976608187134503e-06, + "loss": 3.3342, + "step": 18147 + }, + { + "epoch": 0.9551578947368421, + "grad_norm": 0.5625, + "learning_rate": 9.964912280701755e-06, + "loss": 3.2828, + "step": 18148 + }, + { + "epoch": 0.9552105263157895, + "grad_norm": 0.51953125, + "learning_rate": 9.953216374269006e-06, + "loss": 3.5313, + "step": 18149 + }, + { + "epoch": 0.9552631578947368, + "grad_norm": 0.5, + "learning_rate": 9.941520467836257e-06, + "loss": 3.013, + "step": 18150 + }, + { + "epoch": 0.9553157894736842, + "grad_norm": 0.4921875, + "learning_rate": 9.92982456140351e-06, + "loss": 3.1794, + "step": 18151 + }, + { + "epoch": 0.9553684210526315, + "grad_norm": 0.4921875, + "learning_rate": 9.918128654970762e-06, + "loss": 3.4225, + "step": 18152 + }, + { + "epoch": 0.955421052631579, + "grad_norm": 0.48046875, + "learning_rate": 9.906432748538013e-06, + "loss": 3.4647, + "step": 18153 + }, + { + "epoch": 0.9554736842105264, + "grad_norm": 0.5, + "learning_rate": 9.894736842105264e-06, + "loss": 3.2377, + "step": 18154 + }, + { + "epoch": 0.9555263157894737, + "grad_norm": 0.490234375, + "learning_rate": 9.883040935672515e-06, + "loss": 3.4617, + "step": 18155 + }, + { + "epoch": 0.9555789473684211, + "grad_norm": 0.50390625, + "learning_rate": 9.871345029239767e-06, + "loss": 3.6327, + "step": 18156 + }, + { + "epoch": 0.9556315789473684, + "grad_norm": 0.47265625, + "learning_rate": 9.859649122807018e-06, + "loss": 3.2194, + "step": 18157 + }, + { + "epoch": 0.9556842105263158, + "grad_norm": 0.48046875, + "learning_rate": 9.84795321637427e-06, + "loss": 3.3591, + "step": 18158 + }, + { + "epoch": 0.9557368421052631, + "grad_norm": 0.5234375, + "learning_rate": 9.83625730994152e-06, + "loss": 3.035, + "step": 18159 + }, + { + "epoch": 0.9557894736842105, + "grad_norm": 0.51953125, + "learning_rate": 9.824561403508772e-06, + "loss": 3.1932, + "step": 18160 + }, + { + "epoch": 0.9558421052631579, + "grad_norm": 0.49609375, + "learning_rate": 9.812865497076023e-06, + "loss": 3.2946, + "step": 18161 + }, + { + "epoch": 0.9558947368421052, + "grad_norm": 0.5, + "learning_rate": 9.801169590643274e-06, + "loss": 3.3466, + "step": 18162 + }, + { + "epoch": 0.9559473684210527, + "grad_norm": 0.51953125, + "learning_rate": 9.789473684210526e-06, + "loss": 3.1265, + "step": 18163 + }, + { + "epoch": 0.956, + "grad_norm": 0.478515625, + "learning_rate": 9.777777777777779e-06, + "loss": 3.1335, + "step": 18164 + }, + { + "epoch": 0.9560526315789474, + "grad_norm": 0.546875, + "learning_rate": 9.76608187134503e-06, + "loss": 3.075, + "step": 18165 + }, + { + "epoch": 0.9561052631578948, + "grad_norm": 0.50390625, + "learning_rate": 9.754385964912281e-06, + "loss": 3.4884, + "step": 18166 + }, + { + "epoch": 0.9561578947368421, + "grad_norm": 0.498046875, + "learning_rate": 9.742690058479533e-06, + "loss": 3.1616, + "step": 18167 + }, + { + "epoch": 0.9562105263157895, + "grad_norm": 0.48046875, + "learning_rate": 9.730994152046784e-06, + "loss": 3.3642, + "step": 18168 + }, + { + "epoch": 0.9562631578947368, + "grad_norm": 0.53515625, + "learning_rate": 9.719298245614035e-06, + "loss": 3.1696, + "step": 18169 + }, + { + "epoch": 0.9563157894736842, + "grad_norm": 0.50390625, + "learning_rate": 9.707602339181288e-06, + "loss": 3.0403, + "step": 18170 + }, + { + "epoch": 0.9563684210526315, + "grad_norm": 0.5234375, + "learning_rate": 9.69590643274854e-06, + "loss": 3.3183, + "step": 18171 + }, + { + "epoch": 0.956421052631579, + "grad_norm": 0.494140625, + "learning_rate": 9.68421052631579e-06, + "loss": 3.0367, + "step": 18172 + }, + { + "epoch": 0.9564736842105264, + "grad_norm": 0.51953125, + "learning_rate": 9.672514619883042e-06, + "loss": 3.3059, + "step": 18173 + }, + { + "epoch": 0.9565263157894737, + "grad_norm": 0.51171875, + "learning_rate": 9.660818713450293e-06, + "loss": 3.8652, + "step": 18174 + }, + { + "epoch": 0.9565789473684211, + "grad_norm": 0.4765625, + "learning_rate": 9.649122807017545e-06, + "loss": 3.1261, + "step": 18175 + }, + { + "epoch": 0.9566315789473684, + "grad_norm": 0.494140625, + "learning_rate": 9.637426900584796e-06, + "loss": 3.1594, + "step": 18176 + }, + { + "epoch": 0.9566842105263158, + "grad_norm": 0.51171875, + "learning_rate": 9.625730994152047e-06, + "loss": 3.0052, + "step": 18177 + }, + { + "epoch": 0.9567368421052631, + "grad_norm": 0.51953125, + "learning_rate": 9.614035087719298e-06, + "loss": 3.2573, + "step": 18178 + }, + { + "epoch": 0.9567894736842105, + "grad_norm": 0.5, + "learning_rate": 9.60233918128655e-06, + "loss": 3.3302, + "step": 18179 + }, + { + "epoch": 0.9568421052631579, + "grad_norm": 0.5, + "learning_rate": 9.590643274853801e-06, + "loss": 3.1378, + "step": 18180 + }, + { + "epoch": 0.9568947368421052, + "grad_norm": 0.50390625, + "learning_rate": 9.578947368421052e-06, + "loss": 3.3289, + "step": 18181 + }, + { + "epoch": 0.9569473684210527, + "grad_norm": 0.46875, + "learning_rate": 9.567251461988305e-06, + "loss": 2.7202, + "step": 18182 + }, + { + "epoch": 0.957, + "grad_norm": 0.482421875, + "learning_rate": 9.555555555555556e-06, + "loss": 3.5353, + "step": 18183 + }, + { + "epoch": 0.9570526315789474, + "grad_norm": 0.51171875, + "learning_rate": 9.543859649122808e-06, + "loss": 3.1353, + "step": 18184 + }, + { + "epoch": 0.9571052631578948, + "grad_norm": 0.515625, + "learning_rate": 9.532163742690059e-06, + "loss": 3.189, + "step": 18185 + }, + { + "epoch": 0.9571578947368421, + "grad_norm": 0.5078125, + "learning_rate": 9.52046783625731e-06, + "loss": 3.6863, + "step": 18186 + }, + { + "epoch": 0.9572105263157895, + "grad_norm": 0.498046875, + "learning_rate": 9.508771929824562e-06, + "loss": 3.1895, + "step": 18187 + }, + { + "epoch": 0.9572631578947368, + "grad_norm": 0.470703125, + "learning_rate": 9.497076023391813e-06, + "loss": 3.7566, + "step": 18188 + }, + { + "epoch": 0.9573157894736842, + "grad_norm": 0.498046875, + "learning_rate": 9.485380116959066e-06, + "loss": 3.2796, + "step": 18189 + }, + { + "epoch": 0.9573684210526315, + "grad_norm": 0.486328125, + "learning_rate": 9.473684210526317e-06, + "loss": 3.0543, + "step": 18190 + }, + { + "epoch": 0.957421052631579, + "grad_norm": 0.5, + "learning_rate": 9.461988304093568e-06, + "loss": 3.1391, + "step": 18191 + }, + { + "epoch": 0.9574736842105264, + "grad_norm": 0.53515625, + "learning_rate": 9.45029239766082e-06, + "loss": 3.001, + "step": 18192 + }, + { + "epoch": 0.9575263157894737, + "grad_norm": 0.5, + "learning_rate": 9.438596491228071e-06, + "loss": 3.3047, + "step": 18193 + }, + { + "epoch": 0.9575789473684211, + "grad_norm": 0.51171875, + "learning_rate": 9.426900584795322e-06, + "loss": 3.3919, + "step": 18194 + }, + { + "epoch": 0.9576315789473684, + "grad_norm": 0.515625, + "learning_rate": 9.415204678362574e-06, + "loss": 3.2902, + "step": 18195 + }, + { + "epoch": 0.9576842105263158, + "grad_norm": 0.50390625, + "learning_rate": 9.403508771929825e-06, + "loss": 3.163, + "step": 18196 + }, + { + "epoch": 0.9577368421052631, + "grad_norm": 0.48828125, + "learning_rate": 9.391812865497076e-06, + "loss": 3.1236, + "step": 18197 + }, + { + "epoch": 0.9577894736842105, + "grad_norm": 0.515625, + "learning_rate": 9.380116959064327e-06, + "loss": 3.7249, + "step": 18198 + }, + { + "epoch": 0.9578421052631579, + "grad_norm": 0.4609375, + "learning_rate": 9.368421052631579e-06, + "loss": 2.5572, + "step": 18199 + }, + { + "epoch": 0.9578947368421052, + "grad_norm": 0.48046875, + "learning_rate": 9.35672514619883e-06, + "loss": 3.4839, + "step": 18200 + }, + { + "epoch": 0.9579473684210527, + "grad_norm": 0.48828125, + "learning_rate": 9.345029239766083e-06, + "loss": 3.4192, + "step": 18201 + }, + { + "epoch": 0.958, + "grad_norm": 0.5, + "learning_rate": 9.333333333333334e-06, + "loss": 3.4115, + "step": 18202 + }, + { + "epoch": 0.9580526315789474, + "grad_norm": 0.51171875, + "learning_rate": 9.321637426900586e-06, + "loss": 3.3908, + "step": 18203 + }, + { + "epoch": 0.9581052631578947, + "grad_norm": 0.4765625, + "learning_rate": 9.309941520467837e-06, + "loss": 3.1828, + "step": 18204 + }, + { + "epoch": 0.9581578947368421, + "grad_norm": 0.494140625, + "learning_rate": 9.298245614035088e-06, + "loss": 3.1719, + "step": 18205 + }, + { + "epoch": 0.9582105263157895, + "grad_norm": 0.51953125, + "learning_rate": 9.28654970760234e-06, + "loss": 3.2052, + "step": 18206 + }, + { + "epoch": 0.9582631578947368, + "grad_norm": 0.498046875, + "learning_rate": 9.27485380116959e-06, + "loss": 3.1725, + "step": 18207 + }, + { + "epoch": 0.9583157894736842, + "grad_norm": 0.50390625, + "learning_rate": 9.263157894736844e-06, + "loss": 3.2535, + "step": 18208 + }, + { + "epoch": 0.9583684210526315, + "grad_norm": 0.51953125, + "learning_rate": 9.251461988304095e-06, + "loss": 3.1352, + "step": 18209 + }, + { + "epoch": 0.958421052631579, + "grad_norm": 0.53125, + "learning_rate": 9.239766081871346e-06, + "loss": 3.0753, + "step": 18210 + }, + { + "epoch": 0.9584736842105264, + "grad_norm": 0.51171875, + "learning_rate": 9.228070175438598e-06, + "loss": 3.0818, + "step": 18211 + }, + { + "epoch": 0.9585263157894737, + "grad_norm": 0.50390625, + "learning_rate": 9.216374269005847e-06, + "loss": 2.9402, + "step": 18212 + }, + { + "epoch": 0.9585789473684211, + "grad_norm": 0.50390625, + "learning_rate": 9.204678362573098e-06, + "loss": 3.5742, + "step": 18213 + }, + { + "epoch": 0.9586315789473684, + "grad_norm": 0.52734375, + "learning_rate": 9.192982456140351e-06, + "loss": 3.3952, + "step": 18214 + }, + { + "epoch": 0.9586842105263158, + "grad_norm": 0.5234375, + "learning_rate": 9.181286549707603e-06, + "loss": 2.9627, + "step": 18215 + }, + { + "epoch": 0.9587368421052631, + "grad_norm": 0.5234375, + "learning_rate": 9.169590643274854e-06, + "loss": 3.3012, + "step": 18216 + }, + { + "epoch": 0.9587894736842105, + "grad_norm": 0.51171875, + "learning_rate": 9.157894736842105e-06, + "loss": 3.2668, + "step": 18217 + }, + { + "epoch": 0.9588421052631579, + "grad_norm": 0.515625, + "learning_rate": 9.146198830409357e-06, + "loss": 3.0829, + "step": 18218 + }, + { + "epoch": 0.9588947368421052, + "grad_norm": 0.5, + "learning_rate": 9.134502923976608e-06, + "loss": 2.9891, + "step": 18219 + }, + { + "epoch": 0.9589473684210527, + "grad_norm": 0.494140625, + "learning_rate": 9.122807017543861e-06, + "loss": 3.1726, + "step": 18220 + }, + { + "epoch": 0.959, + "grad_norm": 0.498046875, + "learning_rate": 9.111111111111112e-06, + "loss": 3.0695, + "step": 18221 + }, + { + "epoch": 0.9590526315789474, + "grad_norm": 0.50390625, + "learning_rate": 9.099415204678363e-06, + "loss": 3.2382, + "step": 18222 + }, + { + "epoch": 0.9591052631578947, + "grad_norm": 0.51171875, + "learning_rate": 9.087719298245615e-06, + "loss": 3.2234, + "step": 18223 + }, + { + "epoch": 0.9591578947368421, + "grad_norm": 0.5078125, + "learning_rate": 9.076023391812866e-06, + "loss": 3.4116, + "step": 18224 + }, + { + "epoch": 0.9592105263157895, + "grad_norm": 0.484375, + "learning_rate": 9.064327485380117e-06, + "loss": 3.1685, + "step": 18225 + }, + { + "epoch": 0.9592631578947368, + "grad_norm": 0.51171875, + "learning_rate": 9.05263157894737e-06, + "loss": 3.2516, + "step": 18226 + }, + { + "epoch": 0.9593157894736842, + "grad_norm": 0.52734375, + "learning_rate": 9.040935672514622e-06, + "loss": 3.4306, + "step": 18227 + }, + { + "epoch": 0.9593684210526315, + "grad_norm": 0.52734375, + "learning_rate": 9.029239766081871e-06, + "loss": 3.173, + "step": 18228 + }, + { + "epoch": 0.959421052631579, + "grad_norm": 0.46875, + "learning_rate": 9.017543859649122e-06, + "loss": 2.9876, + "step": 18229 + }, + { + "epoch": 0.9594736842105264, + "grad_norm": 0.51171875, + "learning_rate": 9.005847953216374e-06, + "loss": 3.2797, + "step": 18230 + }, + { + "epoch": 0.9595263157894737, + "grad_norm": 0.5234375, + "learning_rate": 8.994152046783625e-06, + "loss": 3.16, + "step": 18231 + }, + { + "epoch": 0.9595789473684211, + "grad_norm": 0.515625, + "learning_rate": 8.982456140350876e-06, + "loss": 3.2735, + "step": 18232 + }, + { + "epoch": 0.9596315789473684, + "grad_norm": 0.46875, + "learning_rate": 8.97076023391813e-06, + "loss": 3.1922, + "step": 18233 + }, + { + "epoch": 0.9596842105263158, + "grad_norm": 0.490234375, + "learning_rate": 8.95906432748538e-06, + "loss": 3.3701, + "step": 18234 + }, + { + "epoch": 0.9597368421052631, + "grad_norm": 0.51953125, + "learning_rate": 8.947368421052632e-06, + "loss": 3.4321, + "step": 18235 + }, + { + "epoch": 0.9597894736842105, + "grad_norm": 0.515625, + "learning_rate": 8.935672514619883e-06, + "loss": 3.0832, + "step": 18236 + }, + { + "epoch": 0.9598421052631579, + "grad_norm": 0.484375, + "learning_rate": 8.923976608187134e-06, + "loss": 2.957, + "step": 18237 + }, + { + "epoch": 0.9598947368421052, + "grad_norm": 0.4921875, + "learning_rate": 8.912280701754386e-06, + "loss": 3.2935, + "step": 18238 + }, + { + "epoch": 0.9599473684210527, + "grad_norm": 0.5078125, + "learning_rate": 8.900584795321639e-06, + "loss": 2.917, + "step": 18239 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 8.88888888888889e-06, + "loss": 3.1272, + "step": 18240 + }, + { + "epoch": 0.9600526315789474, + "grad_norm": 0.49609375, + "learning_rate": 8.877192982456141e-06, + "loss": 2.9557, + "step": 18241 + }, + { + "epoch": 0.9601052631578947, + "grad_norm": 0.50390625, + "learning_rate": 8.865497076023393e-06, + "loss": 3.2256, + "step": 18242 + }, + { + "epoch": 0.9601578947368421, + "grad_norm": 0.4921875, + "learning_rate": 8.853801169590644e-06, + "loss": 3.3048, + "step": 18243 + }, + { + "epoch": 0.9602105263157895, + "grad_norm": 0.48828125, + "learning_rate": 8.842105263157895e-06, + "loss": 3.1149, + "step": 18244 + }, + { + "epoch": 0.9602631578947368, + "grad_norm": 0.5, + "learning_rate": 8.830409356725146e-06, + "loss": 3.6848, + "step": 18245 + }, + { + "epoch": 0.9603157894736842, + "grad_norm": 0.5, + "learning_rate": 8.818713450292398e-06, + "loss": 3.3552, + "step": 18246 + }, + { + "epoch": 0.9603684210526315, + "grad_norm": 0.59375, + "learning_rate": 8.807017543859649e-06, + "loss": 3.3907, + "step": 18247 + }, + { + "epoch": 0.960421052631579, + "grad_norm": 0.47265625, + "learning_rate": 8.7953216374269e-06, + "loss": 2.8529, + "step": 18248 + }, + { + "epoch": 0.9604736842105264, + "grad_norm": 0.515625, + "learning_rate": 8.783625730994152e-06, + "loss": 3.229, + "step": 18249 + }, + { + "epoch": 0.9605263157894737, + "grad_norm": 0.5234375, + "learning_rate": 8.771929824561403e-06, + "loss": 3.5548, + "step": 18250 + }, + { + "epoch": 0.9605789473684211, + "grad_norm": 0.498046875, + "learning_rate": 8.760233918128656e-06, + "loss": 3.2097, + "step": 18251 + }, + { + "epoch": 0.9606315789473684, + "grad_norm": 0.5078125, + "learning_rate": 8.748538011695907e-06, + "loss": 3.1595, + "step": 18252 + }, + { + "epoch": 0.9606842105263158, + "grad_norm": 0.5390625, + "learning_rate": 8.736842105263158e-06, + "loss": 2.8569, + "step": 18253 + }, + { + "epoch": 0.9607368421052631, + "grad_norm": 0.50390625, + "learning_rate": 8.72514619883041e-06, + "loss": 3.4628, + "step": 18254 + }, + { + "epoch": 0.9607894736842105, + "grad_norm": 0.51953125, + "learning_rate": 8.713450292397661e-06, + "loss": 3.5237, + "step": 18255 + }, + { + "epoch": 0.9608421052631579, + "grad_norm": 0.498046875, + "learning_rate": 8.701754385964912e-06, + "loss": 3.6631, + "step": 18256 + }, + { + "epoch": 0.9608947368421052, + "grad_norm": 0.462890625, + "learning_rate": 8.690058479532164e-06, + "loss": 3.2077, + "step": 18257 + }, + { + "epoch": 0.9609473684210527, + "grad_norm": 0.494140625, + "learning_rate": 8.678362573099417e-06, + "loss": 3.2969, + "step": 18258 + }, + { + "epoch": 0.961, + "grad_norm": 0.53125, + "learning_rate": 8.666666666666668e-06, + "loss": 3.3444, + "step": 18259 + }, + { + "epoch": 0.9610526315789474, + "grad_norm": 0.515625, + "learning_rate": 8.654970760233919e-06, + "loss": 3.6367, + "step": 18260 + }, + { + "epoch": 0.9611052631578947, + "grad_norm": 0.53125, + "learning_rate": 8.64327485380117e-06, + "loss": 3.0031, + "step": 18261 + }, + { + "epoch": 0.9611578947368421, + "grad_norm": 0.49609375, + "learning_rate": 8.631578947368422e-06, + "loss": 3.1579, + "step": 18262 + }, + { + "epoch": 0.9612105263157895, + "grad_norm": 0.494140625, + "learning_rate": 8.619883040935673e-06, + "loss": 3.3006, + "step": 18263 + }, + { + "epoch": 0.9612631578947368, + "grad_norm": 0.53125, + "learning_rate": 8.608187134502924e-06, + "loss": 3.1823, + "step": 18264 + }, + { + "epoch": 0.9613157894736842, + "grad_norm": 0.498046875, + "learning_rate": 8.596491228070176e-06, + "loss": 3.1618, + "step": 18265 + }, + { + "epoch": 0.9613684210526315, + "grad_norm": 0.48046875, + "learning_rate": 8.584795321637427e-06, + "loss": 3.3006, + "step": 18266 + }, + { + "epoch": 0.961421052631579, + "grad_norm": 0.494140625, + "learning_rate": 8.573099415204678e-06, + "loss": 3.6037, + "step": 18267 + }, + { + "epoch": 0.9614736842105264, + "grad_norm": 0.52734375, + "learning_rate": 8.56140350877193e-06, + "loss": 3.2796, + "step": 18268 + }, + { + "epoch": 0.9615263157894737, + "grad_norm": 0.5078125, + "learning_rate": 8.54970760233918e-06, + "loss": 3.1661, + "step": 18269 + }, + { + "epoch": 0.9615789473684211, + "grad_norm": 0.51953125, + "learning_rate": 8.538011695906434e-06, + "loss": 3.4126, + "step": 18270 + }, + { + "epoch": 0.9616315789473684, + "grad_norm": 0.484375, + "learning_rate": 8.526315789473685e-06, + "loss": 2.9967, + "step": 18271 + }, + { + "epoch": 0.9616842105263158, + "grad_norm": 0.484375, + "learning_rate": 8.514619883040936e-06, + "loss": 3.2314, + "step": 18272 + }, + { + "epoch": 0.9617368421052631, + "grad_norm": 0.5078125, + "learning_rate": 8.502923976608188e-06, + "loss": 2.8953, + "step": 18273 + }, + { + "epoch": 0.9617894736842105, + "grad_norm": 0.51953125, + "learning_rate": 8.491228070175439e-06, + "loss": 3.3692, + "step": 18274 + }, + { + "epoch": 0.9618421052631579, + "grad_norm": 0.490234375, + "learning_rate": 8.47953216374269e-06, + "loss": 3.2739, + "step": 18275 + }, + { + "epoch": 0.9618947368421052, + "grad_norm": 0.5, + "learning_rate": 8.467836257309941e-06, + "loss": 3.205, + "step": 18276 + }, + { + "epoch": 0.9619473684210527, + "grad_norm": 0.48046875, + "learning_rate": 8.456140350877194e-06, + "loss": 3.3011, + "step": 18277 + }, + { + "epoch": 0.962, + "grad_norm": 0.515625, + "learning_rate": 8.444444444444446e-06, + "loss": 3.2765, + "step": 18278 + }, + { + "epoch": 0.9620526315789474, + "grad_norm": 0.51171875, + "learning_rate": 8.432748538011697e-06, + "loss": 3.1167, + "step": 18279 + }, + { + "epoch": 0.9621052631578947, + "grad_norm": 0.4921875, + "learning_rate": 8.421052631578948e-06, + "loss": 2.9926, + "step": 18280 + }, + { + "epoch": 0.9621578947368421, + "grad_norm": 0.48828125, + "learning_rate": 8.4093567251462e-06, + "loss": 3.0902, + "step": 18281 + }, + { + "epoch": 0.9622105263157895, + "grad_norm": 0.50390625, + "learning_rate": 8.39766081871345e-06, + "loss": 3.4965, + "step": 18282 + }, + { + "epoch": 0.9622631578947368, + "grad_norm": 0.546875, + "learning_rate": 8.385964912280702e-06, + "loss": 3.0138, + "step": 18283 + }, + { + "epoch": 0.9623157894736842, + "grad_norm": 0.51953125, + "learning_rate": 8.374269005847953e-06, + "loss": 3.2002, + "step": 18284 + }, + { + "epoch": 0.9623684210526315, + "grad_norm": 0.51171875, + "learning_rate": 8.362573099415205e-06, + "loss": 3.2891, + "step": 18285 + }, + { + "epoch": 0.962421052631579, + "grad_norm": 0.5078125, + "learning_rate": 8.350877192982456e-06, + "loss": 3.3829, + "step": 18286 + }, + { + "epoch": 0.9624736842105264, + "grad_norm": 0.5078125, + "learning_rate": 8.339181286549707e-06, + "loss": 3.5092, + "step": 18287 + }, + { + "epoch": 0.9625263157894737, + "grad_norm": 0.4921875, + "learning_rate": 8.327485380116959e-06, + "loss": 3.3674, + "step": 18288 + }, + { + "epoch": 0.9625789473684211, + "grad_norm": 0.5234375, + "learning_rate": 8.315789473684212e-06, + "loss": 3.2738, + "step": 18289 + }, + { + "epoch": 0.9626315789473684, + "grad_norm": 0.50390625, + "learning_rate": 8.304093567251463e-06, + "loss": 3.3748, + "step": 18290 + }, + { + "epoch": 0.9626842105263158, + "grad_norm": 0.48046875, + "learning_rate": 8.292397660818714e-06, + "loss": 2.6104, + "step": 18291 + }, + { + "epoch": 0.9627368421052631, + "grad_norm": 0.48828125, + "learning_rate": 8.280701754385965e-06, + "loss": 3.4287, + "step": 18292 + }, + { + "epoch": 0.9627894736842105, + "grad_norm": 0.49609375, + "learning_rate": 8.269005847953217e-06, + "loss": 3.6937, + "step": 18293 + }, + { + "epoch": 0.9628421052631579, + "grad_norm": 0.49609375, + "learning_rate": 8.257309941520468e-06, + "loss": 3.4498, + "step": 18294 + }, + { + "epoch": 0.9628947368421052, + "grad_norm": 0.482421875, + "learning_rate": 8.245614035087721e-06, + "loss": 2.7136, + "step": 18295 + }, + { + "epoch": 0.9629473684210527, + "grad_norm": 0.4765625, + "learning_rate": 8.233918128654972e-06, + "loss": 3.4435, + "step": 18296 + }, + { + "epoch": 0.963, + "grad_norm": 0.474609375, + "learning_rate": 8.222222222222223e-06, + "loss": 3.0169, + "step": 18297 + }, + { + "epoch": 0.9630526315789474, + "grad_norm": 0.515625, + "learning_rate": 8.210526315789475e-06, + "loss": 3.6705, + "step": 18298 + }, + { + "epoch": 0.9631052631578947, + "grad_norm": 0.49609375, + "learning_rate": 8.198830409356726e-06, + "loss": 3.1836, + "step": 18299 + }, + { + "epoch": 0.9631578947368421, + "grad_norm": 0.5078125, + "learning_rate": 8.187134502923977e-06, + "loss": 3.4778, + "step": 18300 + }, + { + "epoch": 0.9632105263157895, + "grad_norm": 0.478515625, + "learning_rate": 8.175438596491227e-06, + "loss": 3.1093, + "step": 18301 + }, + { + "epoch": 0.9632631578947368, + "grad_norm": 0.4921875, + "learning_rate": 8.16374269005848e-06, + "loss": 3.431, + "step": 18302 + }, + { + "epoch": 0.9633157894736842, + "grad_norm": 0.53125, + "learning_rate": 8.152046783625731e-06, + "loss": 3.439, + "step": 18303 + }, + { + "epoch": 0.9633684210526315, + "grad_norm": 0.5078125, + "learning_rate": 8.140350877192983e-06, + "loss": 3.4793, + "step": 18304 + }, + { + "epoch": 0.963421052631579, + "grad_norm": 0.49609375, + "learning_rate": 8.128654970760234e-06, + "loss": 2.7665, + "step": 18305 + }, + { + "epoch": 0.9634736842105264, + "grad_norm": 0.490234375, + "learning_rate": 8.116959064327485e-06, + "loss": 3.0038, + "step": 18306 + }, + { + "epoch": 0.9635263157894737, + "grad_norm": 0.494140625, + "learning_rate": 8.105263157894736e-06, + "loss": 3.1862, + "step": 18307 + }, + { + "epoch": 0.9635789473684211, + "grad_norm": 0.478515625, + "learning_rate": 8.09356725146199e-06, + "loss": 3.013, + "step": 18308 + }, + { + "epoch": 0.9636315789473684, + "grad_norm": 0.54296875, + "learning_rate": 8.08187134502924e-06, + "loss": 3.1833, + "step": 18309 + }, + { + "epoch": 0.9636842105263158, + "grad_norm": 0.515625, + "learning_rate": 8.070175438596492e-06, + "loss": 3.2204, + "step": 18310 + }, + { + "epoch": 0.9637368421052631, + "grad_norm": 0.53515625, + "learning_rate": 8.058479532163743e-06, + "loss": 3.2436, + "step": 18311 + }, + { + "epoch": 0.9637894736842105, + "grad_norm": 0.5078125, + "learning_rate": 8.046783625730994e-06, + "loss": 3.3119, + "step": 18312 + }, + { + "epoch": 0.9638421052631579, + "grad_norm": 0.58203125, + "learning_rate": 8.035087719298246e-06, + "loss": 3.7375, + "step": 18313 + }, + { + "epoch": 0.9638947368421052, + "grad_norm": 0.5078125, + "learning_rate": 8.023391812865499e-06, + "loss": 3.2466, + "step": 18314 + }, + { + "epoch": 0.9639473684210527, + "grad_norm": 0.51171875, + "learning_rate": 8.01169590643275e-06, + "loss": 3.2733, + "step": 18315 + }, + { + "epoch": 0.964, + "grad_norm": 0.4921875, + "learning_rate": 8.000000000000001e-06, + "loss": 3.1919, + "step": 18316 + }, + { + "epoch": 0.9640526315789474, + "grad_norm": 0.5390625, + "learning_rate": 7.988304093567253e-06, + "loss": 2.9032, + "step": 18317 + }, + { + "epoch": 0.9641052631578947, + "grad_norm": 0.53125, + "learning_rate": 7.976608187134502e-06, + "loss": 3.2176, + "step": 18318 + }, + { + "epoch": 0.9641578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.964912280701753e-06, + "loss": 3.4216, + "step": 18319 + }, + { + "epoch": 0.9642105263157895, + "grad_norm": 0.5078125, + "learning_rate": 7.953216374269006e-06, + "loss": 3.4755, + "step": 18320 + }, + { + "epoch": 0.9642631578947368, + "grad_norm": 0.490234375, + "learning_rate": 7.941520467836258e-06, + "loss": 2.9652, + "step": 18321 + }, + { + "epoch": 0.9643157894736842, + "grad_norm": 0.5, + "learning_rate": 7.929824561403509e-06, + "loss": 3.0211, + "step": 18322 + }, + { + "epoch": 0.9643684210526315, + "grad_norm": 0.515625, + "learning_rate": 7.91812865497076e-06, + "loss": 3.3251, + "step": 18323 + }, + { + "epoch": 0.964421052631579, + "grad_norm": 0.57421875, + "learning_rate": 7.906432748538012e-06, + "loss": 3.423, + "step": 18324 + }, + { + "epoch": 0.9644736842105263, + "grad_norm": 0.515625, + "learning_rate": 7.894736842105263e-06, + "loss": 3.5557, + "step": 18325 + }, + { + "epoch": 0.9645263157894737, + "grad_norm": 0.515625, + "learning_rate": 7.883040935672514e-06, + "loss": 2.9997, + "step": 18326 + }, + { + "epoch": 0.9645789473684211, + "grad_norm": 0.5078125, + "learning_rate": 7.871345029239767e-06, + "loss": 3.0759, + "step": 18327 + }, + { + "epoch": 0.9646315789473684, + "grad_norm": 0.51171875, + "learning_rate": 7.859649122807018e-06, + "loss": 3.2997, + "step": 18328 + }, + { + "epoch": 0.9646842105263158, + "grad_norm": 0.486328125, + "learning_rate": 7.84795321637427e-06, + "loss": 3.1381, + "step": 18329 + }, + { + "epoch": 0.9647368421052631, + "grad_norm": 0.49609375, + "learning_rate": 7.836257309941521e-06, + "loss": 3.3444, + "step": 18330 + }, + { + "epoch": 0.9647894736842105, + "grad_norm": 0.5, + "learning_rate": 7.824561403508772e-06, + "loss": 3.5013, + "step": 18331 + }, + { + "epoch": 0.964842105263158, + "grad_norm": 0.55078125, + "learning_rate": 7.812865497076024e-06, + "loss": 3.5557, + "step": 18332 + }, + { + "epoch": 0.9648947368421052, + "grad_norm": 0.5078125, + "learning_rate": 7.801169590643277e-06, + "loss": 3.2382, + "step": 18333 + }, + { + "epoch": 0.9649473684210527, + "grad_norm": 0.515625, + "learning_rate": 7.789473684210528e-06, + "loss": 3.4284, + "step": 18334 + }, + { + "epoch": 0.965, + "grad_norm": 0.50390625, + "learning_rate": 7.777777777777777e-06, + "loss": 3.5544, + "step": 18335 + }, + { + "epoch": 0.9650526315789474, + "grad_norm": 0.5078125, + "learning_rate": 7.766081871345029e-06, + "loss": 3.2415, + "step": 18336 + }, + { + "epoch": 0.9651052631578947, + "grad_norm": 0.515625, + "learning_rate": 7.75438596491228e-06, + "loss": 2.8773, + "step": 18337 + }, + { + "epoch": 0.9651578947368421, + "grad_norm": 0.54296875, + "learning_rate": 7.742690058479531e-06, + "loss": 3.3404, + "step": 18338 + }, + { + "epoch": 0.9652105263157895, + "grad_norm": 0.546875, + "learning_rate": 7.730994152046784e-06, + "loss": 3.1636, + "step": 18339 + }, + { + "epoch": 0.9652631578947368, + "grad_norm": 0.5234375, + "learning_rate": 7.719298245614036e-06, + "loss": 3.1779, + "step": 18340 + }, + { + "epoch": 0.9653157894736842, + "grad_norm": 0.5390625, + "learning_rate": 7.707602339181287e-06, + "loss": 3.5597, + "step": 18341 + }, + { + "epoch": 0.9653684210526315, + "grad_norm": 0.47265625, + "learning_rate": 7.695906432748538e-06, + "loss": 3.3766, + "step": 18342 + }, + { + "epoch": 0.965421052631579, + "grad_norm": 0.52734375, + "learning_rate": 7.68421052631579e-06, + "loss": 3.1325, + "step": 18343 + }, + { + "epoch": 0.9654736842105263, + "grad_norm": 0.5078125, + "learning_rate": 7.67251461988304e-06, + "loss": 2.8076, + "step": 18344 + }, + { + "epoch": 0.9655263157894737, + "grad_norm": 0.5234375, + "learning_rate": 7.660818713450294e-06, + "loss": 3.4886, + "step": 18345 + }, + { + "epoch": 0.9655789473684211, + "grad_norm": 0.484375, + "learning_rate": 7.649122807017545e-06, + "loss": 3.1006, + "step": 18346 + }, + { + "epoch": 0.9656315789473684, + "grad_norm": 0.609375, + "learning_rate": 7.637426900584796e-06, + "loss": 3.4286, + "step": 18347 + }, + { + "epoch": 0.9656842105263158, + "grad_norm": 0.5234375, + "learning_rate": 7.625730994152048e-06, + "loss": 3.4829, + "step": 18348 + }, + { + "epoch": 0.9657368421052631, + "grad_norm": 0.5390625, + "learning_rate": 7.614035087719298e-06, + "loss": 3.1955, + "step": 18349 + }, + { + "epoch": 0.9657894736842105, + "grad_norm": 0.51171875, + "learning_rate": 7.602339181286549e-06, + "loss": 3.2271, + "step": 18350 + }, + { + "epoch": 0.965842105263158, + "grad_norm": 0.490234375, + "learning_rate": 7.590643274853801e-06, + "loss": 3.1306, + "step": 18351 + }, + { + "epoch": 0.9658947368421053, + "grad_norm": 0.498046875, + "learning_rate": 7.578947368421054e-06, + "loss": 3.4277, + "step": 18352 + }, + { + "epoch": 0.9659473684210527, + "grad_norm": 0.51171875, + "learning_rate": 7.567251461988305e-06, + "loss": 3.1152, + "step": 18353 + }, + { + "epoch": 0.966, + "grad_norm": 0.474609375, + "learning_rate": 7.555555555555556e-06, + "loss": 3.0682, + "step": 18354 + }, + { + "epoch": 0.9660526315789474, + "grad_norm": 0.48046875, + "learning_rate": 7.5438596491228074e-06, + "loss": 3.1534, + "step": 18355 + }, + { + "epoch": 0.9661052631578947, + "grad_norm": 0.51171875, + "learning_rate": 7.532163742690059e-06, + "loss": 3.3496, + "step": 18356 + }, + { + "epoch": 0.9661578947368421, + "grad_norm": 0.484375, + "learning_rate": 7.52046783625731e-06, + "loss": 3.3294, + "step": 18357 + }, + { + "epoch": 0.9662105263157895, + "grad_norm": 0.490234375, + "learning_rate": 7.508771929824562e-06, + "loss": 3.2387, + "step": 18358 + }, + { + "epoch": 0.9662631578947368, + "grad_norm": 0.52734375, + "learning_rate": 7.4970760233918134e-06, + "loss": 3.1054, + "step": 18359 + }, + { + "epoch": 0.9663157894736842, + "grad_norm": 0.5234375, + "learning_rate": 7.485380116959065e-06, + "loss": 2.8852, + "step": 18360 + }, + { + "epoch": 0.9663684210526315, + "grad_norm": 0.470703125, + "learning_rate": 7.473684210526316e-06, + "loss": 3.2008, + "step": 18361 + }, + { + "epoch": 0.966421052631579, + "grad_norm": 0.5078125, + "learning_rate": 7.461988304093567e-06, + "loss": 3.1148, + "step": 18362 + }, + { + "epoch": 0.9664736842105263, + "grad_norm": 0.50390625, + "learning_rate": 7.450292397660819e-06, + "loss": 3.6014, + "step": 18363 + }, + { + "epoch": 0.9665263157894737, + "grad_norm": 0.55078125, + "learning_rate": 7.4385964912280716e-06, + "loss": 3.2221, + "step": 18364 + }, + { + "epoch": 0.9665789473684211, + "grad_norm": 0.5078125, + "learning_rate": 7.426900584795322e-06, + "loss": 3.3137, + "step": 18365 + }, + { + "epoch": 0.9666315789473684, + "grad_norm": 0.49609375, + "learning_rate": 7.415204678362573e-06, + "loss": 3.3998, + "step": 18366 + }, + { + "epoch": 0.9666842105263158, + "grad_norm": 0.51953125, + "learning_rate": 7.4035087719298246e-06, + "loss": 3.3458, + "step": 18367 + }, + { + "epoch": 0.9667368421052631, + "grad_norm": 0.478515625, + "learning_rate": 7.391812865497076e-06, + "loss": 3.1421, + "step": 18368 + }, + { + "epoch": 0.9667894736842105, + "grad_norm": 0.51171875, + "learning_rate": 7.380116959064327e-06, + "loss": 3.5311, + "step": 18369 + }, + { + "epoch": 0.966842105263158, + "grad_norm": 0.5078125, + "learning_rate": 7.3684210526315784e-06, + "loss": 3.5275, + "step": 18370 + }, + { + "epoch": 0.9668947368421053, + "grad_norm": 0.4765625, + "learning_rate": 7.356725146198831e-06, + "loss": 2.8448, + "step": 18371 + }, + { + "epoch": 0.9669473684210527, + "grad_norm": 0.494140625, + "learning_rate": 7.345029239766083e-06, + "loss": 2.9942, + "step": 18372 + }, + { + "epoch": 0.967, + "grad_norm": 0.50390625, + "learning_rate": 7.333333333333334e-06, + "loss": 2.9331, + "step": 18373 + }, + { + "epoch": 0.9670526315789474, + "grad_norm": 0.4921875, + "learning_rate": 7.321637426900585e-06, + "loss": 3.2537, + "step": 18374 + }, + { + "epoch": 0.9671052631578947, + "grad_norm": 0.53125, + "learning_rate": 7.3099415204678366e-06, + "loss": 3.4555, + "step": 18375 + }, + { + "epoch": 0.9671578947368421, + "grad_norm": 0.5234375, + "learning_rate": 7.298245614035088e-06, + "loss": 3.4569, + "step": 18376 + }, + { + "epoch": 0.9672105263157895, + "grad_norm": 0.51953125, + "learning_rate": 7.28654970760234e-06, + "loss": 3.2879, + "step": 18377 + }, + { + "epoch": 0.9672631578947368, + "grad_norm": 0.53125, + "learning_rate": 7.274853801169591e-06, + "loss": 3.3185, + "step": 18378 + }, + { + "epoch": 0.9673157894736842, + "grad_norm": 0.5078125, + "learning_rate": 7.2631578947368426e-06, + "loss": 3.4534, + "step": 18379 + }, + { + "epoch": 0.9673684210526315, + "grad_norm": 0.53515625, + "learning_rate": 7.251461988304094e-06, + "loss": 3.3179, + "step": 18380 + }, + { + "epoch": 0.967421052631579, + "grad_norm": 0.50390625, + "learning_rate": 7.239766081871345e-06, + "loss": 3.1057, + "step": 18381 + }, + { + "epoch": 0.9674736842105263, + "grad_norm": 0.486328125, + "learning_rate": 7.228070175438596e-06, + "loss": 3.1495, + "step": 18382 + }, + { + "epoch": 0.9675263157894737, + "grad_norm": 0.5078125, + "learning_rate": 7.2163742690058486e-06, + "loss": 3.3723, + "step": 18383 + }, + { + "epoch": 0.9675789473684211, + "grad_norm": 0.470703125, + "learning_rate": 7.2046783625731e-06, + "loss": 2.9009, + "step": 18384 + }, + { + "epoch": 0.9676315789473684, + "grad_norm": 0.52734375, + "learning_rate": 7.192982456140351e-06, + "loss": 3.4478, + "step": 18385 + }, + { + "epoch": 0.9676842105263158, + "grad_norm": 0.4921875, + "learning_rate": 7.181286549707602e-06, + "loss": 3.2578, + "step": 18386 + }, + { + "epoch": 0.9677368421052631, + "grad_norm": 0.490234375, + "learning_rate": 7.169590643274854e-06, + "loss": 3.143, + "step": 18387 + }, + { + "epoch": 0.9677894736842105, + "grad_norm": 0.48828125, + "learning_rate": 7.157894736842105e-06, + "loss": 3.2001, + "step": 18388 + }, + { + "epoch": 0.967842105263158, + "grad_norm": 0.515625, + "learning_rate": 7.146198830409358e-06, + "loss": 3.2413, + "step": 18389 + }, + { + "epoch": 0.9678947368421053, + "grad_norm": 0.498046875, + "learning_rate": 7.134502923976609e-06, + "loss": 3.3174, + "step": 18390 + }, + { + "epoch": 0.9679473684210527, + "grad_norm": 0.490234375, + "learning_rate": 7.1228070175438605e-06, + "loss": 2.8417, + "step": 18391 + }, + { + "epoch": 0.968, + "grad_norm": 0.51171875, + "learning_rate": 7.111111111111112e-06, + "loss": 3.5443, + "step": 18392 + }, + { + "epoch": 0.9680526315789474, + "grad_norm": 0.515625, + "learning_rate": 7.099415204678362e-06, + "loss": 3.3294, + "step": 18393 + }, + { + "epoch": 0.9681052631578947, + "grad_norm": 0.494140625, + "learning_rate": 7.0877192982456136e-06, + "loss": 3.4182, + "step": 18394 + }, + { + "epoch": 0.9681578947368421, + "grad_norm": 0.50390625, + "learning_rate": 7.076023391812865e-06, + "loss": 3.0991, + "step": 18395 + }, + { + "epoch": 0.9682105263157895, + "grad_norm": 0.486328125, + "learning_rate": 7.064327485380118e-06, + "loss": 3.5877, + "step": 18396 + }, + { + "epoch": 0.9682631578947368, + "grad_norm": 0.474609375, + "learning_rate": 7.052631578947369e-06, + "loss": 3.2863, + "step": 18397 + }, + { + "epoch": 0.9683157894736842, + "grad_norm": 0.50390625, + "learning_rate": 7.04093567251462e-06, + "loss": 3.4038, + "step": 18398 + }, + { + "epoch": 0.9683684210526315, + "grad_norm": 0.484375, + "learning_rate": 7.029239766081872e-06, + "loss": 3.3315, + "step": 18399 + }, + { + "epoch": 0.968421052631579, + "grad_norm": 0.494140625, + "learning_rate": 7.017543859649123e-06, + "loss": 3.3297, + "step": 18400 + }, + { + "epoch": 0.9684736842105263, + "grad_norm": 0.515625, + "learning_rate": 7.005847953216374e-06, + "loss": 3.3652, + "step": 18401 + }, + { + "epoch": 0.9685263157894737, + "grad_norm": 0.53515625, + "learning_rate": 6.994152046783626e-06, + "loss": 2.9471, + "step": 18402 + }, + { + "epoch": 0.9685789473684211, + "grad_norm": 0.5078125, + "learning_rate": 6.982456140350878e-06, + "loss": 3.5255, + "step": 18403 + }, + { + "epoch": 0.9686315789473684, + "grad_norm": 0.52734375, + "learning_rate": 6.970760233918129e-06, + "loss": 3.3599, + "step": 18404 + }, + { + "epoch": 0.9686842105263158, + "grad_norm": 0.484375, + "learning_rate": 6.95906432748538e-06, + "loss": 2.988, + "step": 18405 + }, + { + "epoch": 0.9687368421052631, + "grad_norm": 0.484375, + "learning_rate": 6.9473684210526315e-06, + "loss": 3.1425, + "step": 18406 + }, + { + "epoch": 0.9687894736842105, + "grad_norm": 0.55078125, + "learning_rate": 6.935672514619883e-06, + "loss": 3.1571, + "step": 18407 + }, + { + "epoch": 0.968842105263158, + "grad_norm": 0.51953125, + "learning_rate": 6.923976608187136e-06, + "loss": 3.5928, + "step": 18408 + }, + { + "epoch": 0.9688947368421053, + "grad_norm": 0.5078125, + "learning_rate": 6.912280701754387e-06, + "loss": 3.3409, + "step": 18409 + }, + { + "epoch": 0.9689473684210527, + "grad_norm": 0.486328125, + "learning_rate": 6.9005847953216375e-06, + "loss": 3.1638, + "step": 18410 + }, + { + "epoch": 0.969, + "grad_norm": 0.515625, + "learning_rate": 6.888888888888889e-06, + "loss": 3.6402, + "step": 18411 + }, + { + "epoch": 0.9690526315789474, + "grad_norm": 0.5078125, + "learning_rate": 6.87719298245614e-06, + "loss": 3.3228, + "step": 18412 + }, + { + "epoch": 0.9691052631578947, + "grad_norm": 0.490234375, + "learning_rate": 6.865497076023391e-06, + "loss": 3.1781, + "step": 18413 + }, + { + "epoch": 0.9691578947368421, + "grad_norm": 0.466796875, + "learning_rate": 6.853801169590644e-06, + "loss": 3.3589, + "step": 18414 + }, + { + "epoch": 0.9692105263157895, + "grad_norm": 0.5078125, + "learning_rate": 6.842105263157896e-06, + "loss": 3.5343, + "step": 18415 + }, + { + "epoch": 0.9692631578947368, + "grad_norm": 0.5234375, + "learning_rate": 6.830409356725147e-06, + "loss": 3.2128, + "step": 18416 + }, + { + "epoch": 0.9693157894736842, + "grad_norm": 0.494140625, + "learning_rate": 6.818713450292398e-06, + "loss": 3.1668, + "step": 18417 + }, + { + "epoch": 0.9693684210526315, + "grad_norm": 0.5078125, + "learning_rate": 6.8070175438596495e-06, + "loss": 2.5318, + "step": 18418 + }, + { + "epoch": 0.969421052631579, + "grad_norm": 0.498046875, + "learning_rate": 6.795321637426901e-06, + "loss": 3.105, + "step": 18419 + }, + { + "epoch": 0.9694736842105263, + "grad_norm": 0.5, + "learning_rate": 6.783625730994152e-06, + "loss": 2.7599, + "step": 18420 + }, + { + "epoch": 0.9695263157894737, + "grad_norm": 0.5078125, + "learning_rate": 6.771929824561404e-06, + "loss": 3.2891, + "step": 18421 + }, + { + "epoch": 0.9695789473684211, + "grad_norm": 0.53515625, + "learning_rate": 6.7602339181286555e-06, + "loss": 3.1107, + "step": 18422 + }, + { + "epoch": 0.9696315789473684, + "grad_norm": 0.5, + "learning_rate": 6.748538011695907e-06, + "loss": 3.4185, + "step": 18423 + }, + { + "epoch": 0.9696842105263158, + "grad_norm": 0.5234375, + "learning_rate": 6.736842105263158e-06, + "loss": 3.5209, + "step": 18424 + }, + { + "epoch": 0.9697368421052631, + "grad_norm": 0.5234375, + "learning_rate": 6.725146198830409e-06, + "loss": 3.446, + "step": 18425 + }, + { + "epoch": 0.9697894736842105, + "grad_norm": 0.48828125, + "learning_rate": 6.713450292397661e-06, + "loss": 3.4639, + "step": 18426 + }, + { + "epoch": 0.969842105263158, + "grad_norm": 0.50390625, + "learning_rate": 6.701754385964913e-06, + "loss": 2.8488, + "step": 18427 + }, + { + "epoch": 0.9698947368421053, + "grad_norm": 0.50390625, + "learning_rate": 6.690058479532164e-06, + "loss": 3.4317, + "step": 18428 + }, + { + "epoch": 0.9699473684210527, + "grad_norm": 0.4921875, + "learning_rate": 6.678362573099415e-06, + "loss": 2.8859, + "step": 18429 + }, + { + "epoch": 0.97, + "grad_norm": 0.6015625, + "learning_rate": 6.666666666666667e-06, + "loss": 2.7775, + "step": 18430 + }, + { + "epoch": 0.9700526315789474, + "grad_norm": 0.515625, + "learning_rate": 6.654970760233918e-06, + "loss": 2.8102, + "step": 18431 + }, + { + "epoch": 0.9701052631578947, + "grad_norm": 0.458984375, + "learning_rate": 6.643274853801169e-06, + "loss": 3.1299, + "step": 18432 + }, + { + "epoch": 0.9701578947368421, + "grad_norm": 0.53125, + "learning_rate": 6.631578947368422e-06, + "loss": 3.2157, + "step": 18433 + }, + { + "epoch": 0.9702105263157895, + "grad_norm": 0.51171875, + "learning_rate": 6.6198830409356735e-06, + "loss": 3.0933, + "step": 18434 + }, + { + "epoch": 0.9702631578947368, + "grad_norm": 0.55859375, + "learning_rate": 6.608187134502925e-06, + "loss": 3.342, + "step": 18435 + }, + { + "epoch": 0.9703157894736842, + "grad_norm": 0.51953125, + "learning_rate": 6.596491228070176e-06, + "loss": 3.3873, + "step": 18436 + }, + { + "epoch": 0.9703684210526315, + "grad_norm": 0.5078125, + "learning_rate": 6.584795321637427e-06, + "loss": 3.1696, + "step": 18437 + }, + { + "epoch": 0.970421052631579, + "grad_norm": 0.478515625, + "learning_rate": 6.573099415204678e-06, + "loss": 3.587, + "step": 18438 + }, + { + "epoch": 0.9704736842105263, + "grad_norm": 0.490234375, + "learning_rate": 6.561403508771931e-06, + "loss": 2.9587, + "step": 18439 + }, + { + "epoch": 0.9705263157894737, + "grad_norm": 0.5, + "learning_rate": 6.549707602339182e-06, + "loss": 3.3314, + "step": 18440 + }, + { + "epoch": 0.9705789473684211, + "grad_norm": 0.5, + "learning_rate": 6.538011695906433e-06, + "loss": 3.396, + "step": 18441 + }, + { + "epoch": 0.9706315789473684, + "grad_norm": 0.498046875, + "learning_rate": 6.526315789473685e-06, + "loss": 3.1761, + "step": 18442 + }, + { + "epoch": 0.9706842105263158, + "grad_norm": 0.482421875, + "learning_rate": 6.514619883040936e-06, + "loss": 2.6345, + "step": 18443 + }, + { + "epoch": 0.9707368421052631, + "grad_norm": 0.51171875, + "learning_rate": 6.502923976608187e-06, + "loss": 3.2003, + "step": 18444 + }, + { + "epoch": 0.9707894736842105, + "grad_norm": 0.494140625, + "learning_rate": 6.4912280701754385e-06, + "loss": 3.0201, + "step": 18445 + }, + { + "epoch": 0.9708421052631578, + "grad_norm": 0.51953125, + "learning_rate": 6.479532163742691e-06, + "loss": 3.247, + "step": 18446 + }, + { + "epoch": 0.9708947368421053, + "grad_norm": 0.50390625, + "learning_rate": 6.467836257309942e-06, + "loss": 3.2829, + "step": 18447 + }, + { + "epoch": 0.9709473684210527, + "grad_norm": 0.515625, + "learning_rate": 6.456140350877193e-06, + "loss": 3.3466, + "step": 18448 + }, + { + "epoch": 0.971, + "grad_norm": 0.54296875, + "learning_rate": 6.4444444444444445e-06, + "loss": 3.1495, + "step": 18449 + }, + { + "epoch": 0.9710526315789474, + "grad_norm": 0.56640625, + "learning_rate": 6.432748538011696e-06, + "loss": 3.4306, + "step": 18450 + }, + { + "epoch": 0.9711052631578947, + "grad_norm": 0.53125, + "learning_rate": 6.421052631578947e-06, + "loss": 3.2321, + "step": 18451 + }, + { + "epoch": 0.9711578947368421, + "grad_norm": 0.48828125, + "learning_rate": 6.4093567251462e-06, + "loss": 2.9739, + "step": 18452 + }, + { + "epoch": 0.9712105263157895, + "grad_norm": 0.5, + "learning_rate": 6.397660818713451e-06, + "loss": 3.2709, + "step": 18453 + }, + { + "epoch": 0.9712631578947368, + "grad_norm": 0.4765625, + "learning_rate": 6.385964912280703e-06, + "loss": 3.1036, + "step": 18454 + }, + { + "epoch": 0.9713157894736842, + "grad_norm": 0.54296875, + "learning_rate": 6.374269005847953e-06, + "loss": 3.6631, + "step": 18455 + }, + { + "epoch": 0.9713684210526315, + "grad_norm": 0.5078125, + "learning_rate": 6.362573099415204e-06, + "loss": 3.2566, + "step": 18456 + }, + { + "epoch": 0.971421052631579, + "grad_norm": 0.515625, + "learning_rate": 6.350877192982456e-06, + "loss": 3.1791, + "step": 18457 + }, + { + "epoch": 0.9714736842105263, + "grad_norm": 0.5078125, + "learning_rate": 6.339181286549709e-06, + "loss": 3.0102, + "step": 18458 + }, + { + "epoch": 0.9715263157894737, + "grad_norm": 0.486328125, + "learning_rate": 6.32748538011696e-06, + "loss": 3.3457, + "step": 18459 + }, + { + "epoch": 0.9715789473684211, + "grad_norm": 0.48828125, + "learning_rate": 6.315789473684211e-06, + "loss": 3.2721, + "step": 18460 + }, + { + "epoch": 0.9716315789473684, + "grad_norm": 0.51171875, + "learning_rate": 6.3040935672514625e-06, + "loss": 3.4196, + "step": 18461 + }, + { + "epoch": 0.9716842105263158, + "grad_norm": 0.51953125, + "learning_rate": 6.292397660818714e-06, + "loss": 3.1379, + "step": 18462 + }, + { + "epoch": 0.9717368421052631, + "grad_norm": 0.50390625, + "learning_rate": 6.280701754385965e-06, + "loss": 3.5859, + "step": 18463 + }, + { + "epoch": 0.9717894736842105, + "grad_norm": 0.55859375, + "learning_rate": 6.269005847953216e-06, + "loss": 2.9944, + "step": 18464 + }, + { + "epoch": 0.9718421052631578, + "grad_norm": 0.55078125, + "learning_rate": 6.2573099415204685e-06, + "loss": 3.369, + "step": 18465 + }, + { + "epoch": 0.9718947368421053, + "grad_norm": 0.5, + "learning_rate": 6.24561403508772e-06, + "loss": 3.2986, + "step": 18466 + }, + { + "epoch": 0.9719473684210527, + "grad_norm": 0.48046875, + "learning_rate": 6.233918128654971e-06, + "loss": 3.4706, + "step": 18467 + }, + { + "epoch": 0.972, + "grad_norm": 0.49609375, + "learning_rate": 6.222222222222222e-06, + "loss": 3.4256, + "step": 18468 + }, + { + "epoch": 0.9720526315789474, + "grad_norm": 0.5, + "learning_rate": 6.2105263157894745e-06, + "loss": 2.9385, + "step": 18469 + }, + { + "epoch": 0.9721052631578947, + "grad_norm": 0.48828125, + "learning_rate": 6.198830409356726e-06, + "loss": 3.4337, + "step": 18470 + }, + { + "epoch": 0.9721578947368421, + "grad_norm": 0.515625, + "learning_rate": 6.187134502923977e-06, + "loss": 3.5714, + "step": 18471 + }, + { + "epoch": 0.9722105263157895, + "grad_norm": 0.48046875, + "learning_rate": 6.175438596491228e-06, + "loss": 3.2777, + "step": 18472 + }, + { + "epoch": 0.9722631578947368, + "grad_norm": 0.490234375, + "learning_rate": 6.16374269005848e-06, + "loss": 3.1763, + "step": 18473 + }, + { + "epoch": 0.9723157894736842, + "grad_norm": 0.5, + "learning_rate": 6.152046783625731e-06, + "loss": 3.4822, + "step": 18474 + }, + { + "epoch": 0.9723684210526315, + "grad_norm": 0.54296875, + "learning_rate": 6.140350877192982e-06, + "loss": 3.4801, + "step": 18475 + }, + { + "epoch": 0.972421052631579, + "grad_norm": 0.5, + "learning_rate": 6.128654970760234e-06, + "loss": 3.5144, + "step": 18476 + }, + { + "epoch": 0.9724736842105263, + "grad_norm": 0.51171875, + "learning_rate": 6.116959064327486e-06, + "loss": 3.0251, + "step": 18477 + }, + { + "epoch": 0.9725263157894737, + "grad_norm": 0.4921875, + "learning_rate": 6.105263157894737e-06, + "loss": 3.3572, + "step": 18478 + }, + { + "epoch": 0.9725789473684211, + "grad_norm": 0.486328125, + "learning_rate": 6.093567251461989e-06, + "loss": 3.3666, + "step": 18479 + }, + { + "epoch": 0.9726315789473684, + "grad_norm": 0.51171875, + "learning_rate": 6.08187134502924e-06, + "loss": 3.0917, + "step": 18480 + }, + { + "epoch": 0.9726842105263158, + "grad_norm": 0.51171875, + "learning_rate": 6.070175438596492e-06, + "loss": 3.3764, + "step": 18481 + }, + { + "epoch": 0.9727368421052631, + "grad_norm": 0.51171875, + "learning_rate": 6.058479532163743e-06, + "loss": 3.514, + "step": 18482 + }, + { + "epoch": 0.9727894736842105, + "grad_norm": 0.51953125, + "learning_rate": 6.046783625730994e-06, + "loss": 3.3664, + "step": 18483 + }, + { + "epoch": 0.9728421052631578, + "grad_norm": 0.486328125, + "learning_rate": 6.0350877192982454e-06, + "loss": 3.0874, + "step": 18484 + }, + { + "epoch": 0.9728947368421053, + "grad_norm": 0.478515625, + "learning_rate": 6.023391812865498e-06, + "loss": 2.9608, + "step": 18485 + }, + { + "epoch": 0.9729473684210527, + "grad_norm": 0.5078125, + "learning_rate": 6.011695906432749e-06, + "loss": 3.1638, + "step": 18486 + }, + { + "epoch": 0.973, + "grad_norm": 0.52734375, + "learning_rate": 6e-06, + "loss": 3.3039, + "step": 18487 + }, + { + "epoch": 0.9730526315789474, + "grad_norm": 0.486328125, + "learning_rate": 5.988304093567252e-06, + "loss": 3.0671, + "step": 18488 + }, + { + "epoch": 0.9731052631578947, + "grad_norm": 0.51953125, + "learning_rate": 5.9766081871345036e-06, + "loss": 3.2895, + "step": 18489 + }, + { + "epoch": 0.9731578947368421, + "grad_norm": 0.515625, + "learning_rate": 5.964912280701755e-06, + "loss": 3.258, + "step": 18490 + }, + { + "epoch": 0.9732105263157895, + "grad_norm": 0.47265625, + "learning_rate": 5.953216374269006e-06, + "loss": 3.0884, + "step": 18491 + }, + { + "epoch": 0.9732631578947368, + "grad_norm": 0.515625, + "learning_rate": 5.9415204678362574e-06, + "loss": 3.5855, + "step": 18492 + }, + { + "epoch": 0.9733157894736842, + "grad_norm": 0.53125, + "learning_rate": 5.929824561403509e-06, + "loss": 3.2691, + "step": 18493 + }, + { + "epoch": 0.9733684210526315, + "grad_norm": 0.5, + "learning_rate": 5.918128654970761e-06, + "loss": 2.6908, + "step": 18494 + }, + { + "epoch": 0.973421052631579, + "grad_norm": 0.50390625, + "learning_rate": 5.906432748538012e-06, + "loss": 3.1951, + "step": 18495 + }, + { + "epoch": 0.9734736842105263, + "grad_norm": 0.4921875, + "learning_rate": 5.8947368421052634e-06, + "loss": 3.2222, + "step": 18496 + }, + { + "epoch": 0.9735263157894737, + "grad_norm": 0.494140625, + "learning_rate": 5.8830409356725156e-06, + "loss": 3.3267, + "step": 18497 + }, + { + "epoch": 0.9735789473684211, + "grad_norm": 0.51171875, + "learning_rate": 5.871345029239767e-06, + "loss": 2.6777, + "step": 18498 + }, + { + "epoch": 0.9736315789473684, + "grad_norm": 0.5078125, + "learning_rate": 5.859649122807018e-06, + "loss": 3.5772, + "step": 18499 + }, + { + "epoch": 0.9736842105263158, + "grad_norm": 0.462890625, + "learning_rate": 5.8479532163742686e-06, + "loss": 2.8547, + "step": 18500 + }, + { + "epoch": 0.9737368421052631, + "grad_norm": 0.51953125, + "learning_rate": 5.836257309941521e-06, + "loss": 3.0491, + "step": 18501 + }, + { + "epoch": 0.9737894736842105, + "grad_norm": 0.482421875, + "learning_rate": 5.824561403508772e-06, + "loss": 3.0771, + "step": 18502 + }, + { + "epoch": 0.9738421052631578, + "grad_norm": 0.50390625, + "learning_rate": 5.812865497076023e-06, + "loss": 3.4357, + "step": 18503 + }, + { + "epoch": 0.9738947368421053, + "grad_norm": 0.494140625, + "learning_rate": 5.801169590643275e-06, + "loss": 3.5406, + "step": 18504 + }, + { + "epoch": 0.9739473684210527, + "grad_norm": 0.55078125, + "learning_rate": 5.789473684210527e-06, + "loss": 3.1151, + "step": 18505 + }, + { + "epoch": 0.974, + "grad_norm": 0.48046875, + "learning_rate": 5.777777777777778e-06, + "loss": 2.9744, + "step": 18506 + }, + { + "epoch": 0.9740526315789474, + "grad_norm": 0.53125, + "learning_rate": 5.76608187134503e-06, + "loss": 3.1364, + "step": 18507 + }, + { + "epoch": 0.9741052631578947, + "grad_norm": 0.515625, + "learning_rate": 5.7543859649122806e-06, + "loss": 3.5951, + "step": 18508 + }, + { + "epoch": 0.9741578947368421, + "grad_norm": 0.498046875, + "learning_rate": 5.742690058479532e-06, + "loss": 3.3623, + "step": 18509 + }, + { + "epoch": 0.9742105263157895, + "grad_norm": 0.55078125, + "learning_rate": 5.730994152046784e-06, + "loss": 3.0249, + "step": 18510 + }, + { + "epoch": 0.9742631578947368, + "grad_norm": 0.5234375, + "learning_rate": 5.719298245614035e-06, + "loss": 3.6043, + "step": 18511 + }, + { + "epoch": 0.9743157894736842, + "grad_norm": 0.50390625, + "learning_rate": 5.7076023391812866e-06, + "loss": 3.2449, + "step": 18512 + }, + { + "epoch": 0.9743684210526316, + "grad_norm": 0.48828125, + "learning_rate": 5.695906432748539e-06, + "loss": 3.4829, + "step": 18513 + }, + { + "epoch": 0.974421052631579, + "grad_norm": 0.48046875, + "learning_rate": 5.68421052631579e-06, + "loss": 3.3864, + "step": 18514 + }, + { + "epoch": 0.9744736842105263, + "grad_norm": 0.51953125, + "learning_rate": 5.672514619883041e-06, + "loss": 3.4594, + "step": 18515 + }, + { + "epoch": 0.9745263157894737, + "grad_norm": 0.51953125, + "learning_rate": 5.6608187134502925e-06, + "loss": 3.4262, + "step": 18516 + }, + { + "epoch": 0.9745789473684211, + "grad_norm": 0.53125, + "learning_rate": 5.649122807017544e-06, + "loss": 3.3831, + "step": 18517 + }, + { + "epoch": 0.9746315789473684, + "grad_norm": 0.52734375, + "learning_rate": 5.637426900584795e-06, + "loss": 3.079, + "step": 18518 + }, + { + "epoch": 0.9746842105263158, + "grad_norm": 0.49609375, + "learning_rate": 5.625730994152047e-06, + "loss": 3.1941, + "step": 18519 + }, + { + "epoch": 0.9747368421052631, + "grad_norm": 0.53125, + "learning_rate": 5.6140350877192985e-06, + "loss": 3.1428, + "step": 18520 + }, + { + "epoch": 0.9747894736842105, + "grad_norm": 0.58984375, + "learning_rate": 5.60233918128655e-06, + "loss": 3.0724, + "step": 18521 + }, + { + "epoch": 0.9748421052631578, + "grad_norm": 0.5546875, + "learning_rate": 5.590643274853801e-06, + "loss": 3.2295, + "step": 18522 + }, + { + "epoch": 0.9748947368421053, + "grad_norm": 0.50390625, + "learning_rate": 5.578947368421053e-06, + "loss": 3.1792, + "step": 18523 + }, + { + "epoch": 0.9749473684210527, + "grad_norm": 0.53125, + "learning_rate": 5.5672514619883045e-06, + "loss": 2.968, + "step": 18524 + }, + { + "epoch": 0.975, + "grad_norm": 0.4609375, + "learning_rate": 5.555555555555556e-06, + "loss": 2.9199, + "step": 18525 + }, + { + "epoch": 0.9750526315789474, + "grad_norm": 0.48046875, + "learning_rate": 5.543859649122807e-06, + "loss": 3.4923, + "step": 18526 + }, + { + "epoch": 0.9751052631578947, + "grad_norm": 0.498046875, + "learning_rate": 5.532163742690058e-06, + "loss": 3.3526, + "step": 18527 + }, + { + "epoch": 0.9751578947368421, + "grad_norm": 0.53125, + "learning_rate": 5.52046783625731e-06, + "loss": 2.9702, + "step": 18528 + }, + { + "epoch": 0.9752105263157895, + "grad_norm": 0.5, + "learning_rate": 5.508771929824562e-06, + "loss": 2.9938, + "step": 18529 + }, + { + "epoch": 0.9752631578947368, + "grad_norm": 0.48828125, + "learning_rate": 5.497076023391813e-06, + "loss": 3.1243, + "step": 18530 + }, + { + "epoch": 0.9753157894736842, + "grad_norm": 0.5078125, + "learning_rate": 5.485380116959064e-06, + "loss": 3.2832, + "step": 18531 + }, + { + "epoch": 0.9753684210526316, + "grad_norm": 0.53125, + "learning_rate": 5.4736842105263165e-06, + "loss": 3.6336, + "step": 18532 + }, + { + "epoch": 0.975421052631579, + "grad_norm": 0.49609375, + "learning_rate": 5.461988304093568e-06, + "loss": 3.2304, + "step": 18533 + }, + { + "epoch": 0.9754736842105263, + "grad_norm": 0.498046875, + "learning_rate": 5.450292397660819e-06, + "loss": 3.1037, + "step": 18534 + }, + { + "epoch": 0.9755263157894737, + "grad_norm": 0.484375, + "learning_rate": 5.43859649122807e-06, + "loss": 2.8645, + "step": 18535 + }, + { + "epoch": 0.9755789473684211, + "grad_norm": 0.5, + "learning_rate": 5.426900584795322e-06, + "loss": 3.083, + "step": 18536 + }, + { + "epoch": 0.9756315789473684, + "grad_norm": 0.5234375, + "learning_rate": 5.415204678362573e-06, + "loss": 3.5826, + "step": 18537 + }, + { + "epoch": 0.9756842105263158, + "grad_norm": 0.50390625, + "learning_rate": 5.403508771929825e-06, + "loss": 3.5114, + "step": 18538 + }, + { + "epoch": 0.9757368421052631, + "grad_norm": 0.5, + "learning_rate": 5.391812865497076e-06, + "loss": 3.6236, + "step": 18539 + }, + { + "epoch": 0.9757894736842105, + "grad_norm": 0.498046875, + "learning_rate": 5.380116959064328e-06, + "loss": 3.5085, + "step": 18540 + }, + { + "epoch": 0.9758421052631578, + "grad_norm": 0.51171875, + "learning_rate": 5.36842105263158e-06, + "loss": 2.9527, + "step": 18541 + }, + { + "epoch": 0.9758947368421053, + "grad_norm": 0.55078125, + "learning_rate": 5.356725146198831e-06, + "loss": 2.9188, + "step": 18542 + }, + { + "epoch": 0.9759473684210527, + "grad_norm": 0.54296875, + "learning_rate": 5.345029239766082e-06, + "loss": 3.1134, + "step": 18543 + }, + { + "epoch": 0.976, + "grad_norm": 0.5, + "learning_rate": 5.333333333333334e-06, + "loss": 3.3252, + "step": 18544 + }, + { + "epoch": 0.9760526315789474, + "grad_norm": 0.515625, + "learning_rate": 5.321637426900585e-06, + "loss": 3.4682, + "step": 18545 + }, + { + "epoch": 0.9761052631578947, + "grad_norm": 0.490234375, + "learning_rate": 5.309941520467836e-06, + "loss": 2.7716, + "step": 18546 + }, + { + "epoch": 0.9761578947368421, + "grad_norm": 0.51171875, + "learning_rate": 5.2982456140350875e-06, + "loss": 3.3262, + "step": 18547 + }, + { + "epoch": 0.9762105263157895, + "grad_norm": 0.5234375, + "learning_rate": 5.28654970760234e-06, + "loss": 3.1433, + "step": 18548 + }, + { + "epoch": 0.9762631578947368, + "grad_norm": 0.5078125, + "learning_rate": 5.274853801169591e-06, + "loss": 3.4105, + "step": 18549 + }, + { + "epoch": 0.9763157894736842, + "grad_norm": 0.5390625, + "learning_rate": 5.263157894736842e-06, + "loss": 3.5096, + "step": 18550 + }, + { + "epoch": 0.9763684210526316, + "grad_norm": 0.494140625, + "learning_rate": 5.251461988304094e-06, + "loss": 3.074, + "step": 18551 + }, + { + "epoch": 0.976421052631579, + "grad_norm": 0.50390625, + "learning_rate": 5.239766081871346e-06, + "loss": 3.1877, + "step": 18552 + }, + { + "epoch": 0.9764736842105263, + "grad_norm": 0.47265625, + "learning_rate": 5.228070175438596e-06, + "loss": 2.8421, + "step": 18553 + }, + { + "epoch": 0.9765263157894737, + "grad_norm": 0.494140625, + "learning_rate": 5.216374269005848e-06, + "loss": 3.4776, + "step": 18554 + }, + { + "epoch": 0.9765789473684211, + "grad_norm": 0.482421875, + "learning_rate": 5.2046783625730995e-06, + "loss": 3.3914, + "step": 18555 + }, + { + "epoch": 0.9766315789473684, + "grad_norm": 0.494140625, + "learning_rate": 5.192982456140351e-06, + "loss": 3.5067, + "step": 18556 + }, + { + "epoch": 0.9766842105263158, + "grad_norm": 0.478515625, + "learning_rate": 5.181286549707603e-06, + "loss": 2.9714, + "step": 18557 + }, + { + "epoch": 0.9767368421052631, + "grad_norm": 0.494140625, + "learning_rate": 5.169590643274854e-06, + "loss": 3.6422, + "step": 18558 + }, + { + "epoch": 0.9767894736842105, + "grad_norm": 0.5, + "learning_rate": 5.1578947368421055e-06, + "loss": 3.4299, + "step": 18559 + }, + { + "epoch": 0.9768421052631578, + "grad_norm": 0.48828125, + "learning_rate": 5.146198830409358e-06, + "loss": 3.2173, + "step": 18560 + }, + { + "epoch": 0.9768947368421053, + "grad_norm": 0.515625, + "learning_rate": 5.134502923976608e-06, + "loss": 3.5723, + "step": 18561 + }, + { + "epoch": 0.9769473684210527, + "grad_norm": 0.51171875, + "learning_rate": 5.122807017543859e-06, + "loss": 3.5141, + "step": 18562 + }, + { + "epoch": 0.977, + "grad_norm": 0.484375, + "learning_rate": 5.1111111111111115e-06, + "loss": 2.6583, + "step": 18563 + }, + { + "epoch": 0.9770526315789474, + "grad_norm": 0.5078125, + "learning_rate": 5.099415204678363e-06, + "loss": 3.1996, + "step": 18564 + }, + { + "epoch": 0.9771052631578947, + "grad_norm": 0.51953125, + "learning_rate": 5.087719298245614e-06, + "loss": 2.9843, + "step": 18565 + }, + { + "epoch": 0.9771578947368421, + "grad_norm": 0.470703125, + "learning_rate": 5.076023391812866e-06, + "loss": 2.8505, + "step": 18566 + }, + { + "epoch": 0.9772105263157894, + "grad_norm": 0.50390625, + "learning_rate": 5.0643274853801175e-06, + "loss": 3.4494, + "step": 18567 + }, + { + "epoch": 0.9772631578947368, + "grad_norm": 0.47265625, + "learning_rate": 5.052631578947369e-06, + "loss": 2.9415, + "step": 18568 + }, + { + "epoch": 0.9773157894736842, + "grad_norm": 0.51171875, + "learning_rate": 5.04093567251462e-06, + "loss": 3.4166, + "step": 18569 + }, + { + "epoch": 0.9773684210526316, + "grad_norm": 0.49609375, + "learning_rate": 5.029239766081871e-06, + "loss": 3.137, + "step": 18570 + }, + { + "epoch": 0.977421052631579, + "grad_norm": 0.490234375, + "learning_rate": 5.017543859649123e-06, + "loss": 3.0047, + "step": 18571 + }, + { + "epoch": 0.9774736842105263, + "grad_norm": 0.51171875, + "learning_rate": 5.005847953216374e-06, + "loss": 3.1525, + "step": 18572 + }, + { + "epoch": 0.9775263157894737, + "grad_norm": 0.5, + "learning_rate": 4.994152046783626e-06, + "loss": 3.5419, + "step": 18573 + }, + { + "epoch": 0.9775789473684211, + "grad_norm": 0.51171875, + "learning_rate": 4.982456140350877e-06, + "loss": 3.5029, + "step": 18574 + }, + { + "epoch": 0.9776315789473684, + "grad_norm": 0.494140625, + "learning_rate": 4.970760233918129e-06, + "loss": 3.3598, + "step": 18575 + }, + { + "epoch": 0.9776842105263158, + "grad_norm": 0.5390625, + "learning_rate": 4.959064327485381e-06, + "loss": 3.3393, + "step": 18576 + }, + { + "epoch": 0.9777368421052631, + "grad_norm": 0.53125, + "learning_rate": 4.947368421052632e-06, + "loss": 3.053, + "step": 18577 + }, + { + "epoch": 0.9777894736842105, + "grad_norm": 0.51171875, + "learning_rate": 4.935672514619883e-06, + "loss": 3.3789, + "step": 18578 + }, + { + "epoch": 0.9778421052631578, + "grad_norm": 0.5234375, + "learning_rate": 4.923976608187135e-06, + "loss": 3.5588, + "step": 18579 + }, + { + "epoch": 0.9778947368421053, + "grad_norm": 0.515625, + "learning_rate": 4.912280701754386e-06, + "loss": 3.4027, + "step": 18580 + }, + { + "epoch": 0.9779473684210527, + "grad_norm": 0.515625, + "learning_rate": 4.900584795321637e-06, + "loss": 2.9954, + "step": 18581 + }, + { + "epoch": 0.978, + "grad_norm": 0.515625, + "learning_rate": 4.888888888888889e-06, + "loss": 3.0485, + "step": 18582 + }, + { + "epoch": 0.9780526315789474, + "grad_norm": 0.4765625, + "learning_rate": 4.877192982456141e-06, + "loss": 3.356, + "step": 18583 + }, + { + "epoch": 0.9781052631578947, + "grad_norm": 0.5, + "learning_rate": 4.865497076023392e-06, + "loss": 3.449, + "step": 18584 + }, + { + "epoch": 0.9781578947368421, + "grad_norm": 0.484375, + "learning_rate": 4.853801169590644e-06, + "loss": 2.9239, + "step": 18585 + }, + { + "epoch": 0.9782105263157894, + "grad_norm": 0.50390625, + "learning_rate": 4.842105263157895e-06, + "loss": 3.2858, + "step": 18586 + }, + { + "epoch": 0.9782631578947368, + "grad_norm": 0.5078125, + "learning_rate": 4.830409356725147e-06, + "loss": 3.0166, + "step": 18587 + }, + { + "epoch": 0.9783157894736842, + "grad_norm": 0.498046875, + "learning_rate": 4.818713450292398e-06, + "loss": 3.3418, + "step": 18588 + }, + { + "epoch": 0.9783684210526316, + "grad_norm": 0.515625, + "learning_rate": 4.807017543859649e-06, + "loss": 3.1956, + "step": 18589 + }, + { + "epoch": 0.978421052631579, + "grad_norm": 0.515625, + "learning_rate": 4.7953216374269005e-06, + "loss": 3.1895, + "step": 18590 + }, + { + "epoch": 0.9784736842105263, + "grad_norm": 0.455078125, + "learning_rate": 4.783625730994153e-06, + "loss": 3.02, + "step": 18591 + }, + { + "epoch": 0.9785263157894737, + "grad_norm": 0.53515625, + "learning_rate": 4.771929824561404e-06, + "loss": 3.1288, + "step": 18592 + }, + { + "epoch": 0.9785789473684211, + "grad_norm": 0.515625, + "learning_rate": 4.760233918128655e-06, + "loss": 3.6766, + "step": 18593 + }, + { + "epoch": 0.9786315789473684, + "grad_norm": 0.4921875, + "learning_rate": 4.7485380116959065e-06, + "loss": 2.9137, + "step": 18594 + }, + { + "epoch": 0.9786842105263158, + "grad_norm": 0.50390625, + "learning_rate": 4.736842105263159e-06, + "loss": 3.1271, + "step": 18595 + }, + { + "epoch": 0.9787368421052631, + "grad_norm": 0.49609375, + "learning_rate": 4.72514619883041e-06, + "loss": 3.0877, + "step": 18596 + }, + { + "epoch": 0.9787894736842105, + "grad_norm": 0.515625, + "learning_rate": 4.713450292397661e-06, + "loss": 3.1772, + "step": 18597 + }, + { + "epoch": 0.9788421052631578, + "grad_norm": 0.494140625, + "learning_rate": 4.7017543859649125e-06, + "loss": 3.2896, + "step": 18598 + }, + { + "epoch": 0.9788947368421053, + "grad_norm": 0.49609375, + "learning_rate": 4.690058479532164e-06, + "loss": 3.4476, + "step": 18599 + }, + { + "epoch": 0.9789473684210527, + "grad_norm": 0.515625, + "learning_rate": 4.678362573099415e-06, + "loss": 3.1279, + "step": 18600 + }, + { + "epoch": 0.979, + "grad_norm": 0.55859375, + "learning_rate": 4.666666666666667e-06, + "loss": 2.6952, + "step": 18601 + }, + { + "epoch": 0.9790526315789474, + "grad_norm": 0.51171875, + "learning_rate": 4.6549707602339184e-06, + "loss": 3.5048, + "step": 18602 + }, + { + "epoch": 0.9791052631578947, + "grad_norm": 0.515625, + "learning_rate": 4.64327485380117e-06, + "loss": 2.9428, + "step": 18603 + }, + { + "epoch": 0.9791578947368421, + "grad_norm": 0.494140625, + "learning_rate": 4.631578947368422e-06, + "loss": 3.2126, + "step": 18604 + }, + { + "epoch": 0.9792105263157894, + "grad_norm": 0.51953125, + "learning_rate": 4.619883040935673e-06, + "loss": 3.3164, + "step": 18605 + }, + { + "epoch": 0.9792631578947368, + "grad_norm": 0.5078125, + "learning_rate": 4.608187134502924e-06, + "loss": 3.4563, + "step": 18606 + }, + { + "epoch": 0.9793157894736843, + "grad_norm": 0.5234375, + "learning_rate": 4.596491228070176e-06, + "loss": 3.1642, + "step": 18607 + }, + { + "epoch": 0.9793684210526316, + "grad_norm": 0.5078125, + "learning_rate": 4.584795321637427e-06, + "loss": 3.1783, + "step": 18608 + }, + { + "epoch": 0.979421052631579, + "grad_norm": 0.54296875, + "learning_rate": 4.573099415204678e-06, + "loss": 2.724, + "step": 18609 + }, + { + "epoch": 0.9794736842105263, + "grad_norm": 0.48828125, + "learning_rate": 4.5614035087719304e-06, + "loss": 3.2241, + "step": 18610 + }, + { + "epoch": 0.9795263157894737, + "grad_norm": 0.490234375, + "learning_rate": 4.549707602339182e-06, + "loss": 3.1449, + "step": 18611 + }, + { + "epoch": 0.9795789473684211, + "grad_norm": 0.5, + "learning_rate": 4.538011695906433e-06, + "loss": 3.2896, + "step": 18612 + }, + { + "epoch": 0.9796315789473684, + "grad_norm": 0.498046875, + "learning_rate": 4.526315789473685e-06, + "loss": 3.5015, + "step": 18613 + }, + { + "epoch": 0.9796842105263158, + "grad_norm": 0.50390625, + "learning_rate": 4.514619883040936e-06, + "loss": 2.8911, + "step": 18614 + }, + { + "epoch": 0.9797368421052631, + "grad_norm": 0.53125, + "learning_rate": 4.502923976608187e-06, + "loss": 3.015, + "step": 18615 + }, + { + "epoch": 0.9797894736842105, + "grad_norm": 0.49609375, + "learning_rate": 4.491228070175438e-06, + "loss": 3.1748, + "step": 18616 + }, + { + "epoch": 0.9798421052631578, + "grad_norm": 0.4765625, + "learning_rate": 4.47953216374269e-06, + "loss": 3.2972, + "step": 18617 + }, + { + "epoch": 0.9798947368421053, + "grad_norm": 0.4765625, + "learning_rate": 4.4678362573099416e-06, + "loss": 2.9987, + "step": 18618 + }, + { + "epoch": 0.9799473684210527, + "grad_norm": 0.50390625, + "learning_rate": 4.456140350877193e-06, + "loss": 3.3208, + "step": 18619 + }, + { + "epoch": 0.98, + "grad_norm": 0.5, + "learning_rate": 4.444444444444445e-06, + "loss": 2.8686, + "step": 18620 + }, + { + "epoch": 0.9800526315789474, + "grad_norm": 0.4609375, + "learning_rate": 4.432748538011696e-06, + "loss": 3.1694, + "step": 18621 + }, + { + "epoch": 0.9801052631578947, + "grad_norm": 0.51953125, + "learning_rate": 4.4210526315789476e-06, + "loss": 3.602, + "step": 18622 + }, + { + "epoch": 0.9801578947368421, + "grad_norm": 0.53515625, + "learning_rate": 4.409356725146199e-06, + "loss": 2.9198, + "step": 18623 + }, + { + "epoch": 0.9802105263157894, + "grad_norm": 0.5234375, + "learning_rate": 4.39766081871345e-06, + "loss": 3.1751, + "step": 18624 + }, + { + "epoch": 0.9802631578947368, + "grad_norm": 0.51953125, + "learning_rate": 4.3859649122807014e-06, + "loss": 3.6165, + "step": 18625 + }, + { + "epoch": 0.9803157894736843, + "grad_norm": 0.490234375, + "learning_rate": 4.3742690058479536e-06, + "loss": 3.3217, + "step": 18626 + }, + { + "epoch": 0.9803684210526316, + "grad_norm": 0.515625, + "learning_rate": 4.362573099415205e-06, + "loss": 3.155, + "step": 18627 + }, + { + "epoch": 0.980421052631579, + "grad_norm": 0.51953125, + "learning_rate": 4.350877192982456e-06, + "loss": 3.0658, + "step": 18628 + }, + { + "epoch": 0.9804736842105263, + "grad_norm": 0.52734375, + "learning_rate": 4.339181286549708e-06, + "loss": 3.1993, + "step": 18629 + }, + { + "epoch": 0.9805263157894737, + "grad_norm": 0.5078125, + "learning_rate": 4.3274853801169596e-06, + "loss": 3.2104, + "step": 18630 + }, + { + "epoch": 0.9805789473684211, + "grad_norm": 0.50390625, + "learning_rate": 4.315789473684211e-06, + "loss": 2.808, + "step": 18631 + }, + { + "epoch": 0.9806315789473684, + "grad_norm": 0.51171875, + "learning_rate": 4.304093567251462e-06, + "loss": 3.3425, + "step": 18632 + }, + { + "epoch": 0.9806842105263158, + "grad_norm": 0.5234375, + "learning_rate": 4.292397660818713e-06, + "loss": 3.3965, + "step": 18633 + }, + { + "epoch": 0.9807368421052631, + "grad_norm": 0.51953125, + "learning_rate": 4.280701754385965e-06, + "loss": 3.1639, + "step": 18634 + }, + { + "epoch": 0.9807894736842105, + "grad_norm": 0.51171875, + "learning_rate": 4.269005847953217e-06, + "loss": 3.1797, + "step": 18635 + }, + { + "epoch": 0.9808421052631578, + "grad_norm": 0.498046875, + "learning_rate": 4.257309941520468e-06, + "loss": 3.1371, + "step": 18636 + }, + { + "epoch": 0.9808947368421053, + "grad_norm": 0.50390625, + "learning_rate": 4.245614035087719e-06, + "loss": 2.8004, + "step": 18637 + }, + { + "epoch": 0.9809473684210527, + "grad_norm": 0.54296875, + "learning_rate": 4.233918128654971e-06, + "loss": 3.3528, + "step": 18638 + }, + { + "epoch": 0.981, + "grad_norm": 0.5625, + "learning_rate": 4.222222222222223e-06, + "loss": 3.5502, + "step": 18639 + }, + { + "epoch": 0.9810526315789474, + "grad_norm": 0.50390625, + "learning_rate": 4.210526315789474e-06, + "loss": 3.0873, + "step": 18640 + }, + { + "epoch": 0.9811052631578947, + "grad_norm": 0.515625, + "learning_rate": 4.198830409356725e-06, + "loss": 3.2429, + "step": 18641 + }, + { + "epoch": 0.9811578947368421, + "grad_norm": 0.51171875, + "learning_rate": 4.187134502923977e-06, + "loss": 3.5396, + "step": 18642 + }, + { + "epoch": 0.9812105263157894, + "grad_norm": 0.5, + "learning_rate": 4.175438596491228e-06, + "loss": 3.6001, + "step": 18643 + }, + { + "epoch": 0.9812631578947368, + "grad_norm": 0.48828125, + "learning_rate": 4.163742690058479e-06, + "loss": 3.4805, + "step": 18644 + }, + { + "epoch": 0.9813157894736843, + "grad_norm": 0.49609375, + "learning_rate": 4.152046783625731e-06, + "loss": 3.015, + "step": 18645 + }, + { + "epoch": 0.9813684210526316, + "grad_norm": 0.494140625, + "learning_rate": 4.140350877192983e-06, + "loss": 2.909, + "step": 18646 + }, + { + "epoch": 0.981421052631579, + "grad_norm": 0.51953125, + "learning_rate": 4.128654970760234e-06, + "loss": 3.2564, + "step": 18647 + }, + { + "epoch": 0.9814736842105263, + "grad_norm": 0.515625, + "learning_rate": 4.116959064327486e-06, + "loss": 3.2609, + "step": 18648 + }, + { + "epoch": 0.9815263157894737, + "grad_norm": 0.5390625, + "learning_rate": 4.105263157894737e-06, + "loss": 3.283, + "step": 18649 + }, + { + "epoch": 0.9815789473684211, + "grad_norm": 0.57421875, + "learning_rate": 4.093567251461989e-06, + "loss": 3.4628, + "step": 18650 + }, + { + "epoch": 0.9816315789473684, + "grad_norm": 0.51953125, + "learning_rate": 4.08187134502924e-06, + "loss": 3.0572, + "step": 18651 + }, + { + "epoch": 0.9816842105263158, + "grad_norm": 0.486328125, + "learning_rate": 4.070175438596491e-06, + "loss": 3.5456, + "step": 18652 + }, + { + "epoch": 0.9817368421052631, + "grad_norm": 0.50390625, + "learning_rate": 4.0584795321637425e-06, + "loss": 3.0308, + "step": 18653 + }, + { + "epoch": 0.9817894736842105, + "grad_norm": 0.498046875, + "learning_rate": 4.046783625730995e-06, + "loss": 3.2587, + "step": 18654 + }, + { + "epoch": 0.9818421052631578, + "grad_norm": 0.5234375, + "learning_rate": 4.035087719298246e-06, + "loss": 3.3019, + "step": 18655 + }, + { + "epoch": 0.9818947368421053, + "grad_norm": 0.53125, + "learning_rate": 4.023391812865497e-06, + "loss": 3.4547, + "step": 18656 + }, + { + "epoch": 0.9819473684210527, + "grad_norm": 0.5078125, + "learning_rate": 4.011695906432749e-06, + "loss": 2.9218, + "step": 18657 + }, + { + "epoch": 0.982, + "grad_norm": 0.51171875, + "learning_rate": 4.000000000000001e-06, + "loss": 3.4638, + "step": 18658 + }, + { + "epoch": 0.9820526315789474, + "grad_norm": 0.484375, + "learning_rate": 3.988304093567251e-06, + "loss": 3.182, + "step": 18659 + }, + { + "epoch": 0.9821052631578947, + "grad_norm": 0.490234375, + "learning_rate": 3.976608187134503e-06, + "loss": 3.4709, + "step": 18660 + }, + { + "epoch": 0.9821578947368421, + "grad_norm": 0.482421875, + "learning_rate": 3.9649122807017545e-06, + "loss": 3.1368, + "step": 18661 + }, + { + "epoch": 0.9822105263157894, + "grad_norm": 0.59375, + "learning_rate": 3.953216374269006e-06, + "loss": 3.1669, + "step": 18662 + }, + { + "epoch": 0.9822631578947368, + "grad_norm": 0.546875, + "learning_rate": 3.941520467836257e-06, + "loss": 3.5169, + "step": 18663 + }, + { + "epoch": 0.9823157894736843, + "grad_norm": 0.5390625, + "learning_rate": 3.929824561403509e-06, + "loss": 3.3445, + "step": 18664 + }, + { + "epoch": 0.9823684210526316, + "grad_norm": 0.5234375, + "learning_rate": 3.9181286549707605e-06, + "loss": 3.3731, + "step": 18665 + }, + { + "epoch": 0.982421052631579, + "grad_norm": 0.515625, + "learning_rate": 3.906432748538012e-06, + "loss": 3.4181, + "step": 18666 + }, + { + "epoch": 0.9824736842105263, + "grad_norm": 0.486328125, + "learning_rate": 3.894736842105264e-06, + "loss": 3.4331, + "step": 18667 + }, + { + "epoch": 0.9825263157894737, + "grad_norm": 0.494140625, + "learning_rate": 3.883040935672514e-06, + "loss": 3.2117, + "step": 18668 + }, + { + "epoch": 0.9825789473684211, + "grad_norm": 0.51171875, + "learning_rate": 3.871345029239766e-06, + "loss": 3.3036, + "step": 18669 + }, + { + "epoch": 0.9826315789473684, + "grad_norm": 0.498046875, + "learning_rate": 3.859649122807018e-06, + "loss": 3.237, + "step": 18670 + }, + { + "epoch": 0.9826842105263158, + "grad_norm": 0.5234375, + "learning_rate": 3.847953216374269e-06, + "loss": 3.4946, + "step": 18671 + }, + { + "epoch": 0.9827368421052631, + "grad_norm": 0.50390625, + "learning_rate": 3.83625730994152e-06, + "loss": 3.126, + "step": 18672 + }, + { + "epoch": 0.9827894736842105, + "grad_norm": 0.61328125, + "learning_rate": 3.8245614035087725e-06, + "loss": 2.9856, + "step": 18673 + }, + { + "epoch": 0.9828421052631579, + "grad_norm": 0.5234375, + "learning_rate": 3.812865497076024e-06, + "loss": 3.0061, + "step": 18674 + }, + { + "epoch": 0.9828947368421053, + "grad_norm": 0.5078125, + "learning_rate": 3.8011695906432747e-06, + "loss": 3.1473, + "step": 18675 + }, + { + "epoch": 0.9829473684210527, + "grad_norm": 0.48828125, + "learning_rate": 3.789473684210527e-06, + "loss": 3.3914, + "step": 18676 + }, + { + "epoch": 0.983, + "grad_norm": 0.53125, + "learning_rate": 3.777777777777778e-06, + "loss": 3.4541, + "step": 18677 + }, + { + "epoch": 0.9830526315789474, + "grad_norm": 0.5078125, + "learning_rate": 3.7660818713450294e-06, + "loss": 3.2498, + "step": 18678 + }, + { + "epoch": 0.9831052631578947, + "grad_norm": 0.51171875, + "learning_rate": 3.754385964912281e-06, + "loss": 2.8758, + "step": 18679 + }, + { + "epoch": 0.9831578947368421, + "grad_norm": 0.50390625, + "learning_rate": 3.7426900584795324e-06, + "loss": 3.577, + "step": 18680 + }, + { + "epoch": 0.9832105263157894, + "grad_norm": 0.5078125, + "learning_rate": 3.7309941520467836e-06, + "loss": 2.9447, + "step": 18681 + }, + { + "epoch": 0.9832631578947368, + "grad_norm": 0.51171875, + "learning_rate": 3.7192982456140358e-06, + "loss": 2.9756, + "step": 18682 + }, + { + "epoch": 0.9833157894736843, + "grad_norm": 0.5078125, + "learning_rate": 3.7076023391812866e-06, + "loss": 3.5441, + "step": 18683 + }, + { + "epoch": 0.9833684210526316, + "grad_norm": 0.490234375, + "learning_rate": 3.695906432748538e-06, + "loss": 3.4546, + "step": 18684 + }, + { + "epoch": 0.983421052631579, + "grad_norm": 0.5, + "learning_rate": 3.6842105263157892e-06, + "loss": 3.4029, + "step": 18685 + }, + { + "epoch": 0.9834736842105263, + "grad_norm": 0.51953125, + "learning_rate": 3.6725146198830414e-06, + "loss": 3.4506, + "step": 18686 + }, + { + "epoch": 0.9835263157894737, + "grad_norm": 0.4765625, + "learning_rate": 3.6608187134502926e-06, + "loss": 2.9811, + "step": 18687 + }, + { + "epoch": 0.983578947368421, + "grad_norm": 0.5078125, + "learning_rate": 3.649122807017544e-06, + "loss": 2.8264, + "step": 18688 + }, + { + "epoch": 0.9836315789473684, + "grad_norm": 0.51171875, + "learning_rate": 3.6374269005847956e-06, + "loss": 3.0798, + "step": 18689 + }, + { + "epoch": 0.9836842105263158, + "grad_norm": 0.50390625, + "learning_rate": 3.625730994152047e-06, + "loss": 3.0682, + "step": 18690 + }, + { + "epoch": 0.9837368421052631, + "grad_norm": 0.5078125, + "learning_rate": 3.614035087719298e-06, + "loss": 2.7075, + "step": 18691 + }, + { + "epoch": 0.9837894736842105, + "grad_norm": 0.482421875, + "learning_rate": 3.60233918128655e-06, + "loss": 3.1701, + "step": 18692 + }, + { + "epoch": 0.9838421052631579, + "grad_norm": 0.49609375, + "learning_rate": 3.590643274853801e-06, + "loss": 3.1112, + "step": 18693 + }, + { + "epoch": 0.9838947368421053, + "grad_norm": 0.498046875, + "learning_rate": 3.5789473684210525e-06, + "loss": 3.0291, + "step": 18694 + }, + { + "epoch": 0.9839473684210527, + "grad_norm": 0.5, + "learning_rate": 3.5672514619883046e-06, + "loss": 3.041, + "step": 18695 + }, + { + "epoch": 0.984, + "grad_norm": 0.4921875, + "learning_rate": 3.555555555555556e-06, + "loss": 3.2056, + "step": 18696 + }, + { + "epoch": 0.9840526315789474, + "grad_norm": 0.5234375, + "learning_rate": 3.5438596491228068e-06, + "loss": 3.1675, + "step": 18697 + }, + { + "epoch": 0.9841052631578947, + "grad_norm": 0.49609375, + "learning_rate": 3.532163742690059e-06, + "loss": 3.5394, + "step": 18698 + }, + { + "epoch": 0.9841578947368421, + "grad_norm": 0.5078125, + "learning_rate": 3.52046783625731e-06, + "loss": 2.8955, + "step": 18699 + }, + { + "epoch": 0.9842105263157894, + "grad_norm": 0.54296875, + "learning_rate": 3.5087719298245615e-06, + "loss": 2.9542, + "step": 18700 + }, + { + "epoch": 0.9842631578947368, + "grad_norm": 0.498046875, + "learning_rate": 3.497076023391813e-06, + "loss": 3.3045, + "step": 18701 + }, + { + "epoch": 0.9843157894736843, + "grad_norm": 0.5, + "learning_rate": 3.4853801169590645e-06, + "loss": 3.2772, + "step": 18702 + }, + { + "epoch": 0.9843684210526316, + "grad_norm": 0.5078125, + "learning_rate": 3.4736842105263158e-06, + "loss": 3.3957, + "step": 18703 + }, + { + "epoch": 0.984421052631579, + "grad_norm": 0.48828125, + "learning_rate": 3.461988304093568e-06, + "loss": 3.4511, + "step": 18704 + }, + { + "epoch": 0.9844736842105263, + "grad_norm": 0.55078125, + "learning_rate": 3.4502923976608188e-06, + "loss": 3.8154, + "step": 18705 + }, + { + "epoch": 0.9845263157894737, + "grad_norm": 0.51171875, + "learning_rate": 3.43859649122807e-06, + "loss": 3.2788, + "step": 18706 + }, + { + "epoch": 0.984578947368421, + "grad_norm": 0.5078125, + "learning_rate": 3.426900584795322e-06, + "loss": 2.723, + "step": 18707 + }, + { + "epoch": 0.9846315789473684, + "grad_norm": 0.5234375, + "learning_rate": 3.4152046783625735e-06, + "loss": 3.4905, + "step": 18708 + }, + { + "epoch": 0.9846842105263158, + "grad_norm": 0.50390625, + "learning_rate": 3.4035087719298248e-06, + "loss": 3.4506, + "step": 18709 + }, + { + "epoch": 0.9847368421052631, + "grad_norm": 0.5078125, + "learning_rate": 3.391812865497076e-06, + "loss": 3.1864, + "step": 18710 + }, + { + "epoch": 0.9847894736842105, + "grad_norm": 0.515625, + "learning_rate": 3.3801169590643278e-06, + "loss": 2.9272, + "step": 18711 + }, + { + "epoch": 0.9848421052631579, + "grad_norm": 0.5234375, + "learning_rate": 3.368421052631579e-06, + "loss": 3.2609, + "step": 18712 + }, + { + "epoch": 0.9848947368421053, + "grad_norm": 0.51171875, + "learning_rate": 3.3567251461988303e-06, + "loss": 3.5599, + "step": 18713 + }, + { + "epoch": 0.9849473684210527, + "grad_norm": 0.486328125, + "learning_rate": 3.345029239766082e-06, + "loss": 3.2152, + "step": 18714 + }, + { + "epoch": 0.985, + "grad_norm": 0.50390625, + "learning_rate": 3.3333333333333333e-06, + "loss": 3.1377, + "step": 18715 + }, + { + "epoch": 0.9850526315789474, + "grad_norm": 0.494140625, + "learning_rate": 3.3216374269005846e-06, + "loss": 3.5291, + "step": 18716 + }, + { + "epoch": 0.9851052631578947, + "grad_norm": 0.515625, + "learning_rate": 3.3099415204678367e-06, + "loss": 3.5827, + "step": 18717 + }, + { + "epoch": 0.9851578947368421, + "grad_norm": 0.5234375, + "learning_rate": 3.298245614035088e-06, + "loss": 3.298, + "step": 18718 + }, + { + "epoch": 0.9852105263157894, + "grad_norm": 0.51171875, + "learning_rate": 3.286549707602339e-06, + "loss": 3.3887, + "step": 18719 + }, + { + "epoch": 0.9852631578947368, + "grad_norm": 0.50390625, + "learning_rate": 3.274853801169591e-06, + "loss": 3.305, + "step": 18720 + }, + { + "epoch": 0.9853157894736843, + "grad_norm": 0.51171875, + "learning_rate": 3.2631578947368423e-06, + "loss": 3.3961, + "step": 18721 + }, + { + "epoch": 0.9853684210526316, + "grad_norm": 0.5, + "learning_rate": 3.2514619883040936e-06, + "loss": 3.1362, + "step": 18722 + }, + { + "epoch": 0.985421052631579, + "grad_norm": 0.48828125, + "learning_rate": 3.2397660818713453e-06, + "loss": 2.9914, + "step": 18723 + }, + { + "epoch": 0.9854736842105263, + "grad_norm": 0.51953125, + "learning_rate": 3.2280701754385966e-06, + "loss": 3.5119, + "step": 18724 + }, + { + "epoch": 0.9855263157894737, + "grad_norm": 0.490234375, + "learning_rate": 3.216374269005848e-06, + "loss": 3.1714, + "step": 18725 + }, + { + "epoch": 0.985578947368421, + "grad_norm": 0.482421875, + "learning_rate": 3.2046783625731e-06, + "loss": 3.2746, + "step": 18726 + }, + { + "epoch": 0.9856315789473684, + "grad_norm": 0.5078125, + "learning_rate": 3.1929824561403513e-06, + "loss": 3.3753, + "step": 18727 + }, + { + "epoch": 0.9856842105263158, + "grad_norm": 0.49609375, + "learning_rate": 3.181286549707602e-06, + "loss": 2.9356, + "step": 18728 + }, + { + "epoch": 0.9857368421052631, + "grad_norm": 0.498046875, + "learning_rate": 3.1695906432748543e-06, + "loss": 3.2086, + "step": 18729 + }, + { + "epoch": 0.9857894736842105, + "grad_norm": 0.51171875, + "learning_rate": 3.1578947368421056e-06, + "loss": 3.5823, + "step": 18730 + }, + { + "epoch": 0.9858421052631579, + "grad_norm": 0.5, + "learning_rate": 3.146198830409357e-06, + "loss": 3.1112, + "step": 18731 + }, + { + "epoch": 0.9858947368421053, + "grad_norm": 0.484375, + "learning_rate": 3.134502923976608e-06, + "loss": 3.2937, + "step": 18732 + }, + { + "epoch": 0.9859473684210527, + "grad_norm": 0.5078125, + "learning_rate": 3.12280701754386e-06, + "loss": 3.0149, + "step": 18733 + }, + { + "epoch": 0.986, + "grad_norm": 0.56640625, + "learning_rate": 3.111111111111111e-06, + "loss": 2.7137, + "step": 18734 + }, + { + "epoch": 0.9860526315789474, + "grad_norm": 0.486328125, + "learning_rate": 3.099415204678363e-06, + "loss": 3.4008, + "step": 18735 + }, + { + "epoch": 0.9861052631578947, + "grad_norm": 0.51171875, + "learning_rate": 3.087719298245614e-06, + "loss": 3.0668, + "step": 18736 + }, + { + "epoch": 0.9861578947368421, + "grad_norm": 0.5, + "learning_rate": 3.0760233918128654e-06, + "loss": 3.1435, + "step": 18737 + }, + { + "epoch": 0.9862105263157894, + "grad_norm": 0.51953125, + "learning_rate": 3.064327485380117e-06, + "loss": 3.3558, + "step": 18738 + }, + { + "epoch": 0.9862631578947368, + "grad_norm": 0.5, + "learning_rate": 3.0526315789473684e-06, + "loss": 3.4792, + "step": 18739 + }, + { + "epoch": 0.9863157894736843, + "grad_norm": 0.50390625, + "learning_rate": 3.04093567251462e-06, + "loss": 3.3165, + "step": 18740 + }, + { + "epoch": 0.9863684210526316, + "grad_norm": 0.53515625, + "learning_rate": 3.0292397660818714e-06, + "loss": 3.3735, + "step": 18741 + }, + { + "epoch": 0.986421052631579, + "grad_norm": 0.50390625, + "learning_rate": 3.0175438596491227e-06, + "loss": 3.3317, + "step": 18742 + }, + { + "epoch": 0.9864736842105263, + "grad_norm": 0.5078125, + "learning_rate": 3.0058479532163744e-06, + "loss": 3.2173, + "step": 18743 + }, + { + "epoch": 0.9865263157894737, + "grad_norm": 0.515625, + "learning_rate": 2.994152046783626e-06, + "loss": 3.1278, + "step": 18744 + }, + { + "epoch": 0.986578947368421, + "grad_norm": 0.486328125, + "learning_rate": 2.9824561403508774e-06, + "loss": 2.6816, + "step": 18745 + }, + { + "epoch": 0.9866315789473684, + "grad_norm": 0.51171875, + "learning_rate": 2.9707602339181287e-06, + "loss": 3.235, + "step": 18746 + }, + { + "epoch": 0.9866842105263158, + "grad_norm": 0.48046875, + "learning_rate": 2.9590643274853804e-06, + "loss": 3.0653, + "step": 18747 + }, + { + "epoch": 0.9867368421052631, + "grad_norm": 0.494140625, + "learning_rate": 2.9473684210526317e-06, + "loss": 3.3987, + "step": 18748 + }, + { + "epoch": 0.9867894736842105, + "grad_norm": 0.515625, + "learning_rate": 2.9356725146198834e-06, + "loss": 3.5845, + "step": 18749 + }, + { + "epoch": 0.9868421052631579, + "grad_norm": 0.4921875, + "learning_rate": 2.9239766081871343e-06, + "loss": 3.5983, + "step": 18750 + }, + { + "epoch": 0.9868947368421053, + "grad_norm": 0.52734375, + "learning_rate": 2.912280701754386e-06, + "loss": 3.3335, + "step": 18751 + }, + { + "epoch": 0.9869473684210527, + "grad_norm": 0.470703125, + "learning_rate": 2.9005847953216377e-06, + "loss": 3.122, + "step": 18752 + }, + { + "epoch": 0.987, + "grad_norm": 0.5078125, + "learning_rate": 2.888888888888889e-06, + "loss": 3.233, + "step": 18753 + }, + { + "epoch": 0.9870526315789474, + "grad_norm": 0.48046875, + "learning_rate": 2.8771929824561403e-06, + "loss": 3.1232, + "step": 18754 + }, + { + "epoch": 0.9871052631578947, + "grad_norm": 0.498046875, + "learning_rate": 2.865497076023392e-06, + "loss": 3.5284, + "step": 18755 + }, + { + "epoch": 0.9871578947368421, + "grad_norm": 0.49609375, + "learning_rate": 2.8538011695906433e-06, + "loss": 2.7713, + "step": 18756 + }, + { + "epoch": 0.9872105263157894, + "grad_norm": 0.54296875, + "learning_rate": 2.842105263157895e-06, + "loss": 3.4491, + "step": 18757 + }, + { + "epoch": 0.9872631578947368, + "grad_norm": 0.478515625, + "learning_rate": 2.8304093567251463e-06, + "loss": 3.3599, + "step": 18758 + }, + { + "epoch": 0.9873157894736843, + "grad_norm": 0.494140625, + "learning_rate": 2.8187134502923976e-06, + "loss": 2.729, + "step": 18759 + }, + { + "epoch": 0.9873684210526316, + "grad_norm": 0.482421875, + "learning_rate": 2.8070175438596493e-06, + "loss": 3.3597, + "step": 18760 + }, + { + "epoch": 0.987421052631579, + "grad_norm": 0.494140625, + "learning_rate": 2.7953216374269006e-06, + "loss": 3.2045, + "step": 18761 + }, + { + "epoch": 0.9874736842105263, + "grad_norm": 0.5390625, + "learning_rate": 2.7836257309941523e-06, + "loss": 3.1623, + "step": 18762 + }, + { + "epoch": 0.9875263157894737, + "grad_norm": 0.5, + "learning_rate": 2.7719298245614036e-06, + "loss": 3.332, + "step": 18763 + }, + { + "epoch": 0.987578947368421, + "grad_norm": 0.54296875, + "learning_rate": 2.760233918128655e-06, + "loss": 3.2053, + "step": 18764 + }, + { + "epoch": 0.9876315789473684, + "grad_norm": 0.498046875, + "learning_rate": 2.7485380116959066e-06, + "loss": 3.4354, + "step": 18765 + }, + { + "epoch": 0.9876842105263158, + "grad_norm": 0.5078125, + "learning_rate": 2.7368421052631583e-06, + "loss": 2.9203, + "step": 18766 + }, + { + "epoch": 0.9877368421052631, + "grad_norm": 0.5078125, + "learning_rate": 2.7251461988304095e-06, + "loss": 3.1633, + "step": 18767 + }, + { + "epoch": 0.9877894736842106, + "grad_norm": 0.50390625, + "learning_rate": 2.713450292397661e-06, + "loss": 3.3563, + "step": 18768 + }, + { + "epoch": 0.9878421052631579, + "grad_norm": 0.5, + "learning_rate": 2.7017543859649125e-06, + "loss": 3.2921, + "step": 18769 + }, + { + "epoch": 0.9878947368421053, + "grad_norm": 0.50390625, + "learning_rate": 2.690058479532164e-06, + "loss": 3.291, + "step": 18770 + }, + { + "epoch": 0.9879473684210527, + "grad_norm": 0.546875, + "learning_rate": 2.6783625730994155e-06, + "loss": 3.0927, + "step": 18771 + }, + { + "epoch": 0.988, + "grad_norm": 0.51171875, + "learning_rate": 2.666666666666667e-06, + "loss": 3.2518, + "step": 18772 + }, + { + "epoch": 0.9880526315789474, + "grad_norm": 0.52734375, + "learning_rate": 2.654970760233918e-06, + "loss": 3.416, + "step": 18773 + }, + { + "epoch": 0.9881052631578947, + "grad_norm": 0.50390625, + "learning_rate": 2.64327485380117e-06, + "loss": 3.238, + "step": 18774 + }, + { + "epoch": 0.9881578947368421, + "grad_norm": 0.5078125, + "learning_rate": 2.631578947368421e-06, + "loss": 3.3812, + "step": 18775 + }, + { + "epoch": 0.9882105263157894, + "grad_norm": 0.51171875, + "learning_rate": 2.619883040935673e-06, + "loss": 3.1976, + "step": 18776 + }, + { + "epoch": 0.9882631578947368, + "grad_norm": 0.50390625, + "learning_rate": 2.608187134502924e-06, + "loss": 3.4071, + "step": 18777 + }, + { + "epoch": 0.9883157894736843, + "grad_norm": 0.48828125, + "learning_rate": 2.5964912280701754e-06, + "loss": 2.8228, + "step": 18778 + }, + { + "epoch": 0.9883684210526316, + "grad_norm": 0.51953125, + "learning_rate": 2.584795321637427e-06, + "loss": 3.5784, + "step": 18779 + }, + { + "epoch": 0.988421052631579, + "grad_norm": 0.57421875, + "learning_rate": 2.573099415204679e-06, + "loss": 3.2636, + "step": 18780 + }, + { + "epoch": 0.9884736842105263, + "grad_norm": 0.51171875, + "learning_rate": 2.5614035087719297e-06, + "loss": 3.2385, + "step": 18781 + }, + { + "epoch": 0.9885263157894737, + "grad_norm": 0.498046875, + "learning_rate": 2.5497076023391814e-06, + "loss": 3.5404, + "step": 18782 + }, + { + "epoch": 0.988578947368421, + "grad_norm": 0.5234375, + "learning_rate": 2.538011695906433e-06, + "loss": 3.2029, + "step": 18783 + }, + { + "epoch": 0.9886315789473684, + "grad_norm": 0.51953125, + "learning_rate": 2.5263157894736844e-06, + "loss": 3.1988, + "step": 18784 + }, + { + "epoch": 0.9886842105263158, + "grad_norm": 0.5234375, + "learning_rate": 2.5146198830409357e-06, + "loss": 3.5278, + "step": 18785 + }, + { + "epoch": 0.9887368421052631, + "grad_norm": 0.51953125, + "learning_rate": 2.502923976608187e-06, + "loss": 3.0902, + "step": 18786 + }, + { + "epoch": 0.9887894736842106, + "grad_norm": 0.484375, + "learning_rate": 2.4912280701754387e-06, + "loss": 3.2814, + "step": 18787 + }, + { + "epoch": 0.9888421052631579, + "grad_norm": 0.53515625, + "learning_rate": 2.4795321637426904e-06, + "loss": 2.9612, + "step": 18788 + }, + { + "epoch": 0.9888947368421053, + "grad_norm": 0.474609375, + "learning_rate": 2.4678362573099417e-06, + "loss": 2.8338, + "step": 18789 + }, + { + "epoch": 0.9889473684210527, + "grad_norm": 0.5390625, + "learning_rate": 2.456140350877193e-06, + "loss": 2.9451, + "step": 18790 + }, + { + "epoch": 0.989, + "grad_norm": 0.48828125, + "learning_rate": 2.4444444444444447e-06, + "loss": 3.2045, + "step": 18791 + }, + { + "epoch": 0.9890526315789474, + "grad_norm": 0.52734375, + "learning_rate": 2.432748538011696e-06, + "loss": 2.9935, + "step": 18792 + }, + { + "epoch": 0.9891052631578947, + "grad_norm": 0.5, + "learning_rate": 2.4210526315789477e-06, + "loss": 3.0098, + "step": 18793 + }, + { + "epoch": 0.9891578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.409356725146199e-06, + "loss": 3.2596, + "step": 18794 + }, + { + "epoch": 0.9892105263157894, + "grad_norm": 0.51171875, + "learning_rate": 2.3976608187134502e-06, + "loss": 3.4437, + "step": 18795 + }, + { + "epoch": 0.9892631578947368, + "grad_norm": 0.5, + "learning_rate": 2.385964912280702e-06, + "loss": 3.2721, + "step": 18796 + }, + { + "epoch": 0.9893157894736843, + "grad_norm": 0.5078125, + "learning_rate": 2.3742690058479532e-06, + "loss": 3.2916, + "step": 18797 + }, + { + "epoch": 0.9893684210526316, + "grad_norm": 0.546875, + "learning_rate": 2.362573099415205e-06, + "loss": 3.0661, + "step": 18798 + }, + { + "epoch": 0.989421052631579, + "grad_norm": 0.4765625, + "learning_rate": 2.3508771929824562e-06, + "loss": 2.8894, + "step": 18799 + }, + { + "epoch": 0.9894736842105263, + "grad_norm": 0.494140625, + "learning_rate": 2.3391812865497075e-06, + "loss": 2.9744, + "step": 18800 + }, + { + "epoch": 0.9895263157894737, + "grad_norm": 0.51953125, + "learning_rate": 2.3274853801169592e-06, + "loss": 3.5872, + "step": 18801 + }, + { + "epoch": 0.989578947368421, + "grad_norm": 0.51171875, + "learning_rate": 2.315789473684211e-06, + "loss": 3.1068, + "step": 18802 + }, + { + "epoch": 0.9896315789473684, + "grad_norm": 0.4921875, + "learning_rate": 2.304093567251462e-06, + "loss": 3.5558, + "step": 18803 + }, + { + "epoch": 0.9896842105263158, + "grad_norm": 0.51953125, + "learning_rate": 2.2923976608187135e-06, + "loss": 3.4752, + "step": 18804 + }, + { + "epoch": 0.9897368421052631, + "grad_norm": 0.50390625, + "learning_rate": 2.2807017543859652e-06, + "loss": 3.1419, + "step": 18805 + }, + { + "epoch": 0.9897894736842106, + "grad_norm": 0.5078125, + "learning_rate": 2.2690058479532165e-06, + "loss": 3.3014, + "step": 18806 + }, + { + "epoch": 0.9898421052631579, + "grad_norm": 0.478515625, + "learning_rate": 2.257309941520468e-06, + "loss": 2.9816, + "step": 18807 + }, + { + "epoch": 0.9898947368421053, + "grad_norm": 0.498046875, + "learning_rate": 2.245614035087719e-06, + "loss": 3.5526, + "step": 18808 + }, + { + "epoch": 0.9899473684210527, + "grad_norm": 0.5078125, + "learning_rate": 2.2339181286549708e-06, + "loss": 3.7486, + "step": 18809 + }, + { + "epoch": 0.99, + "grad_norm": 0.5, + "learning_rate": 2.2222222222222225e-06, + "loss": 3.3175, + "step": 18810 + }, + { + "epoch": 0.9900526315789474, + "grad_norm": 0.515625, + "learning_rate": 2.2105263157894738e-06, + "loss": 3.3, + "step": 18811 + }, + { + "epoch": 0.9901052631578947, + "grad_norm": 0.515625, + "learning_rate": 2.198830409356725e-06, + "loss": 3.4114, + "step": 18812 + }, + { + "epoch": 0.9901578947368421, + "grad_norm": 0.515625, + "learning_rate": 2.1871345029239768e-06, + "loss": 3.6816, + "step": 18813 + }, + { + "epoch": 0.9902105263157894, + "grad_norm": 0.5234375, + "learning_rate": 2.175438596491228e-06, + "loss": 2.7702, + "step": 18814 + }, + { + "epoch": 0.9902631578947368, + "grad_norm": 0.51953125, + "learning_rate": 2.1637426900584798e-06, + "loss": 3.2962, + "step": 18815 + }, + { + "epoch": 0.9903157894736843, + "grad_norm": 0.48828125, + "learning_rate": 2.152046783625731e-06, + "loss": 3.4388, + "step": 18816 + }, + { + "epoch": 0.9903684210526316, + "grad_norm": 0.51953125, + "learning_rate": 2.1403508771929824e-06, + "loss": 3.3254, + "step": 18817 + }, + { + "epoch": 0.990421052631579, + "grad_norm": 0.50390625, + "learning_rate": 2.128654970760234e-06, + "loss": 3.6737, + "step": 18818 + }, + { + "epoch": 0.9904736842105263, + "grad_norm": 0.498046875, + "learning_rate": 2.1169590643274853e-06, + "loss": 3.2158, + "step": 18819 + }, + { + "epoch": 0.9905263157894737, + "grad_norm": 0.52734375, + "learning_rate": 2.105263157894737e-06, + "loss": 3.1744, + "step": 18820 + }, + { + "epoch": 0.990578947368421, + "grad_norm": 0.5, + "learning_rate": 2.0935672514619883e-06, + "loss": 2.9356, + "step": 18821 + }, + { + "epoch": 0.9906315789473684, + "grad_norm": 0.5078125, + "learning_rate": 2.0818713450292396e-06, + "loss": 3.7341, + "step": 18822 + }, + { + "epoch": 0.9906842105263158, + "grad_norm": 0.51171875, + "learning_rate": 2.0701754385964913e-06, + "loss": 3.1537, + "step": 18823 + }, + { + "epoch": 0.9907368421052631, + "grad_norm": 0.54296875, + "learning_rate": 2.058479532163743e-06, + "loss": 3.1395, + "step": 18824 + }, + { + "epoch": 0.9907894736842106, + "grad_norm": 0.5234375, + "learning_rate": 2.0467836257309943e-06, + "loss": 3.6919, + "step": 18825 + }, + { + "epoch": 0.9908421052631579, + "grad_norm": 0.48828125, + "learning_rate": 2.0350877192982456e-06, + "loss": 2.8981, + "step": 18826 + }, + { + "epoch": 0.9908947368421053, + "grad_norm": 0.5234375, + "learning_rate": 2.0233918128654973e-06, + "loss": 3.5284, + "step": 18827 + }, + { + "epoch": 0.9909473684210526, + "grad_norm": 0.625, + "learning_rate": 2.0116959064327486e-06, + "loss": 3.2105, + "step": 18828 + }, + { + "epoch": 0.991, + "grad_norm": 0.5234375, + "learning_rate": 2.0000000000000003e-06, + "loss": 3.3516, + "step": 18829 + }, + { + "epoch": 0.9910526315789474, + "grad_norm": 0.54296875, + "learning_rate": 1.9883040935672516e-06, + "loss": 3.4505, + "step": 18830 + }, + { + "epoch": 0.9911052631578947, + "grad_norm": 0.52734375, + "learning_rate": 1.976608187134503e-06, + "loss": 2.9886, + "step": 18831 + }, + { + "epoch": 0.9911578947368421, + "grad_norm": 0.51953125, + "learning_rate": 1.9649122807017546e-06, + "loss": 3.2064, + "step": 18832 + }, + { + "epoch": 0.9912105263157894, + "grad_norm": 0.498046875, + "learning_rate": 1.953216374269006e-06, + "loss": 3.0844, + "step": 18833 + }, + { + "epoch": 0.9912631578947368, + "grad_norm": 0.474609375, + "learning_rate": 1.941520467836257e-06, + "loss": 2.8974, + "step": 18834 + }, + { + "epoch": 0.9913157894736843, + "grad_norm": 0.494140625, + "learning_rate": 1.929824561403509e-06, + "loss": 3.1259, + "step": 18835 + }, + { + "epoch": 0.9913684210526316, + "grad_norm": 0.498046875, + "learning_rate": 1.91812865497076e-06, + "loss": 3.3656, + "step": 18836 + }, + { + "epoch": 0.991421052631579, + "grad_norm": 0.48828125, + "learning_rate": 1.906432748538012e-06, + "loss": 3.1082, + "step": 18837 + }, + { + "epoch": 0.9914736842105263, + "grad_norm": 0.490234375, + "learning_rate": 1.8947368421052634e-06, + "loss": 3.1432, + "step": 18838 + }, + { + "epoch": 0.9915263157894737, + "grad_norm": 0.48828125, + "learning_rate": 1.8830409356725147e-06, + "loss": 3.3463, + "step": 18839 + }, + { + "epoch": 0.991578947368421, + "grad_norm": 0.5, + "learning_rate": 1.8713450292397662e-06, + "loss": 3.3505, + "step": 18840 + }, + { + "epoch": 0.9916315789473684, + "grad_norm": 0.5234375, + "learning_rate": 1.8596491228070179e-06, + "loss": 3.4081, + "step": 18841 + }, + { + "epoch": 0.9916842105263158, + "grad_norm": 0.486328125, + "learning_rate": 1.847953216374269e-06, + "loss": 2.7309, + "step": 18842 + }, + { + "epoch": 0.9917368421052631, + "grad_norm": 0.48828125, + "learning_rate": 1.8362573099415207e-06, + "loss": 3.2229, + "step": 18843 + }, + { + "epoch": 0.9917894736842106, + "grad_norm": 0.5078125, + "learning_rate": 1.824561403508772e-06, + "loss": 3.1046, + "step": 18844 + }, + { + "epoch": 0.9918421052631579, + "grad_norm": 0.5, + "learning_rate": 1.8128654970760235e-06, + "loss": 3.1694, + "step": 18845 + }, + { + "epoch": 0.9918947368421053, + "grad_norm": 0.5390625, + "learning_rate": 1.801169590643275e-06, + "loss": 3.45, + "step": 18846 + }, + { + "epoch": 0.9919473684210526, + "grad_norm": 0.5, + "learning_rate": 1.7894736842105262e-06, + "loss": 2.845, + "step": 18847 + }, + { + "epoch": 0.992, + "grad_norm": 0.490234375, + "learning_rate": 1.777777777777778e-06, + "loss": 3.1005, + "step": 18848 + }, + { + "epoch": 0.9920526315789474, + "grad_norm": 0.5, + "learning_rate": 1.7660818713450295e-06, + "loss": 2.6833, + "step": 18849 + }, + { + "epoch": 0.9921052631578947, + "grad_norm": 0.482421875, + "learning_rate": 1.7543859649122807e-06, + "loss": 2.9939, + "step": 18850 + }, + { + "epoch": 0.9921578947368421, + "grad_norm": 0.482421875, + "learning_rate": 1.7426900584795322e-06, + "loss": 3.0616, + "step": 18851 + }, + { + "epoch": 0.9922105263157894, + "grad_norm": 0.5078125, + "learning_rate": 1.730994152046784e-06, + "loss": 2.6906, + "step": 18852 + }, + { + "epoch": 0.9922631578947368, + "grad_norm": 0.52734375, + "learning_rate": 1.719298245614035e-06, + "loss": 2.8473, + "step": 18853 + }, + { + "epoch": 0.9923157894736843, + "grad_norm": 0.5, + "learning_rate": 1.7076023391812867e-06, + "loss": 3.2996, + "step": 18854 + }, + { + "epoch": 0.9923684210526316, + "grad_norm": 0.515625, + "learning_rate": 1.695906432748538e-06, + "loss": 3.3279, + "step": 18855 + }, + { + "epoch": 0.992421052631579, + "grad_norm": 0.484375, + "learning_rate": 1.6842105263157895e-06, + "loss": 3.0572, + "step": 18856 + }, + { + "epoch": 0.9924736842105263, + "grad_norm": 0.49609375, + "learning_rate": 1.672514619883041e-06, + "loss": 3.3911, + "step": 18857 + }, + { + "epoch": 0.9925263157894737, + "grad_norm": 0.52734375, + "learning_rate": 1.6608187134502923e-06, + "loss": 3.4038, + "step": 18858 + }, + { + "epoch": 0.992578947368421, + "grad_norm": 0.4921875, + "learning_rate": 1.649122807017544e-06, + "loss": 3.1493, + "step": 18859 + }, + { + "epoch": 0.9926315789473684, + "grad_norm": 0.546875, + "learning_rate": 1.6374269005847955e-06, + "loss": 3.3896, + "step": 18860 + }, + { + "epoch": 0.9926842105263158, + "grad_norm": 0.5078125, + "learning_rate": 1.6257309941520468e-06, + "loss": 3.5922, + "step": 18861 + }, + { + "epoch": 0.9927368421052631, + "grad_norm": 0.486328125, + "learning_rate": 1.6140350877192983e-06, + "loss": 3.1119, + "step": 18862 + }, + { + "epoch": 0.9927894736842106, + "grad_norm": 0.5234375, + "learning_rate": 1.60233918128655e-06, + "loss": 3.2996, + "step": 18863 + }, + { + "epoch": 0.9928421052631579, + "grad_norm": 0.5546875, + "learning_rate": 1.590643274853801e-06, + "loss": 3.335, + "step": 18864 + }, + { + "epoch": 0.9928947368421053, + "grad_norm": 0.515625, + "learning_rate": 1.5789473684210528e-06, + "loss": 3.2948, + "step": 18865 + }, + { + "epoch": 0.9929473684210526, + "grad_norm": 0.51953125, + "learning_rate": 1.567251461988304e-06, + "loss": 3.1747, + "step": 18866 + }, + { + "epoch": 0.993, + "grad_norm": 0.478515625, + "learning_rate": 1.5555555555555556e-06, + "loss": 2.9716, + "step": 18867 + }, + { + "epoch": 0.9930526315789474, + "grad_norm": 0.4765625, + "learning_rate": 1.543859649122807e-06, + "loss": 3.1717, + "step": 18868 + }, + { + "epoch": 0.9931052631578947, + "grad_norm": 0.4921875, + "learning_rate": 1.5321637426900586e-06, + "loss": 3.2804, + "step": 18869 + }, + { + "epoch": 0.9931578947368421, + "grad_norm": 0.478515625, + "learning_rate": 1.52046783625731e-06, + "loss": 3.6641, + "step": 18870 + }, + { + "epoch": 0.9932105263157894, + "grad_norm": 0.5078125, + "learning_rate": 1.5087719298245614e-06, + "loss": 3.2458, + "step": 18871 + }, + { + "epoch": 0.9932631578947368, + "grad_norm": 0.53125, + "learning_rate": 1.497076023391813e-06, + "loss": 3.1913, + "step": 18872 + }, + { + "epoch": 0.9933157894736843, + "grad_norm": 0.5234375, + "learning_rate": 1.4853801169590644e-06, + "loss": 3.1694, + "step": 18873 + }, + { + "epoch": 0.9933684210526316, + "grad_norm": 0.53125, + "learning_rate": 1.4736842105263159e-06, + "loss": 3.2463, + "step": 18874 + }, + { + "epoch": 0.993421052631579, + "grad_norm": 0.49609375, + "learning_rate": 1.4619883040935671e-06, + "loss": 3.6317, + "step": 18875 + }, + { + "epoch": 0.9934736842105263, + "grad_norm": 0.49609375, + "learning_rate": 1.4502923976608189e-06, + "loss": 3.3634, + "step": 18876 + }, + { + "epoch": 0.9935263157894737, + "grad_norm": 0.5, + "learning_rate": 1.4385964912280701e-06, + "loss": 3.496, + "step": 18877 + }, + { + "epoch": 0.993578947368421, + "grad_norm": 0.462890625, + "learning_rate": 1.4269005847953216e-06, + "loss": 2.8813, + "step": 18878 + }, + { + "epoch": 0.9936315789473684, + "grad_norm": 0.484375, + "learning_rate": 1.4152046783625731e-06, + "loss": 3.5346, + "step": 18879 + }, + { + "epoch": 0.9936842105263158, + "grad_norm": 0.51953125, + "learning_rate": 1.4035087719298246e-06, + "loss": 3.4752, + "step": 18880 + }, + { + "epoch": 0.9937368421052631, + "grad_norm": 0.48046875, + "learning_rate": 1.3918128654970761e-06, + "loss": 3.1208, + "step": 18881 + }, + { + "epoch": 0.9937894736842106, + "grad_norm": 0.5, + "learning_rate": 1.3801169590643274e-06, + "loss": 2.9876, + "step": 18882 + }, + { + "epoch": 0.9938421052631579, + "grad_norm": 0.51953125, + "learning_rate": 1.3684210526315791e-06, + "loss": 3.2145, + "step": 18883 + }, + { + "epoch": 0.9938947368421053, + "grad_norm": 0.494140625, + "learning_rate": 1.3567251461988304e-06, + "loss": 3.3048, + "step": 18884 + }, + { + "epoch": 0.9939473684210526, + "grad_norm": 0.50390625, + "learning_rate": 1.345029239766082e-06, + "loss": 3.2225, + "step": 18885 + }, + { + "epoch": 0.994, + "grad_norm": 0.50390625, + "learning_rate": 1.3333333333333334e-06, + "loss": 3.5696, + "step": 18886 + }, + { + "epoch": 0.9940526315789474, + "grad_norm": 0.53515625, + "learning_rate": 1.321637426900585e-06, + "loss": 3.0488, + "step": 18887 + }, + { + "epoch": 0.9941052631578947, + "grad_norm": 0.50390625, + "learning_rate": 1.3099415204678364e-06, + "loss": 3.2205, + "step": 18888 + }, + { + "epoch": 0.9941578947368421, + "grad_norm": 0.5390625, + "learning_rate": 1.2982456140350877e-06, + "loss": 3.8393, + "step": 18889 + }, + { + "epoch": 0.9942105263157894, + "grad_norm": 0.494140625, + "learning_rate": 1.2865497076023394e-06, + "loss": 3.3705, + "step": 18890 + }, + { + "epoch": 0.9942631578947368, + "grad_norm": 0.486328125, + "learning_rate": 1.2748538011695907e-06, + "loss": 3.4211, + "step": 18891 + }, + { + "epoch": 0.9943157894736843, + "grad_norm": 0.55859375, + "learning_rate": 1.2631578947368422e-06, + "loss": 3.1323, + "step": 18892 + }, + { + "epoch": 0.9943684210526316, + "grad_norm": 0.474609375, + "learning_rate": 1.2514619883040935e-06, + "loss": 3.3817, + "step": 18893 + }, + { + "epoch": 0.994421052631579, + "grad_norm": 0.50390625, + "learning_rate": 1.2397660818713452e-06, + "loss": 3.579, + "step": 18894 + }, + { + "epoch": 0.9944736842105263, + "grad_norm": 0.51171875, + "learning_rate": 1.2280701754385965e-06, + "loss": 3.4429, + "step": 18895 + }, + { + "epoch": 0.9945263157894737, + "grad_norm": 0.478515625, + "learning_rate": 1.216374269005848e-06, + "loss": 3.2187, + "step": 18896 + }, + { + "epoch": 0.994578947368421, + "grad_norm": 0.53125, + "learning_rate": 1.2046783625730995e-06, + "loss": 3.4218, + "step": 18897 + }, + { + "epoch": 0.9946315789473684, + "grad_norm": 0.490234375, + "learning_rate": 1.192982456140351e-06, + "loss": 3.3024, + "step": 18898 + }, + { + "epoch": 0.9946842105263158, + "grad_norm": 0.5, + "learning_rate": 1.1812865497076025e-06, + "loss": 2.9397, + "step": 18899 + }, + { + "epoch": 0.9947368421052631, + "grad_norm": 0.51171875, + "learning_rate": 1.1695906432748538e-06, + "loss": 2.9523, + "step": 18900 + }, + { + "epoch": 0.9947894736842106, + "grad_norm": 0.5, + "learning_rate": 1.1578947368421055e-06, + "loss": 3.4995, + "step": 18901 + }, + { + "epoch": 0.9948421052631579, + "grad_norm": 0.51171875, + "learning_rate": 1.1461988304093568e-06, + "loss": 3.1044, + "step": 18902 + }, + { + "epoch": 0.9948947368421053, + "grad_norm": 0.5078125, + "learning_rate": 1.1345029239766083e-06, + "loss": 2.9601, + "step": 18903 + }, + { + "epoch": 0.9949473684210526, + "grad_norm": 0.5078125, + "learning_rate": 1.1228070175438595e-06, + "loss": 2.9416, + "step": 18904 + }, + { + "epoch": 0.995, + "grad_norm": 0.51953125, + "learning_rate": 1.1111111111111112e-06, + "loss": 3.4881, + "step": 18905 + }, + { + "epoch": 0.9950526315789474, + "grad_norm": 0.486328125, + "learning_rate": 1.0994152046783625e-06, + "loss": 3.4006, + "step": 18906 + }, + { + "epoch": 0.9951052631578947, + "grad_norm": 0.48828125, + "learning_rate": 1.087719298245614e-06, + "loss": 3.4513, + "step": 18907 + }, + { + "epoch": 0.9951578947368421, + "grad_norm": 0.51171875, + "learning_rate": 1.0760233918128655e-06, + "loss": 3.238, + "step": 18908 + }, + { + "epoch": 0.9952105263157894, + "grad_norm": 0.51171875, + "learning_rate": 1.064327485380117e-06, + "loss": 3.6381, + "step": 18909 + }, + { + "epoch": 0.9952631578947368, + "grad_norm": 0.49609375, + "learning_rate": 1.0526315789473685e-06, + "loss": 3.0103, + "step": 18910 + }, + { + "epoch": 0.9953157894736843, + "grad_norm": 0.486328125, + "learning_rate": 1.0409356725146198e-06, + "loss": 3.3831, + "step": 18911 + }, + { + "epoch": 0.9953684210526316, + "grad_norm": 0.5078125, + "learning_rate": 1.0292397660818715e-06, + "loss": 2.7026, + "step": 18912 + }, + { + "epoch": 0.995421052631579, + "grad_norm": 0.48828125, + "learning_rate": 1.0175438596491228e-06, + "loss": 3.3572, + "step": 18913 + }, + { + "epoch": 0.9954736842105263, + "grad_norm": 0.53515625, + "learning_rate": 1.0058479532163743e-06, + "loss": 3.0732, + "step": 18914 + }, + { + "epoch": 0.9955263157894737, + "grad_norm": 0.5, + "learning_rate": 9.941520467836258e-07, + "loss": 3.1648, + "step": 18915 + }, + { + "epoch": 0.995578947368421, + "grad_norm": 0.478515625, + "learning_rate": 9.824561403508773e-07, + "loss": 2.8752, + "step": 18916 + }, + { + "epoch": 0.9956315789473684, + "grad_norm": 0.4921875, + "learning_rate": 9.707602339181286e-07, + "loss": 3.0952, + "step": 18917 + }, + { + "epoch": 0.9956842105263158, + "grad_norm": 0.5625, + "learning_rate": 9.5906432748538e-07, + "loss": 3.5349, + "step": 18918 + }, + { + "epoch": 0.9957368421052631, + "grad_norm": 0.490234375, + "learning_rate": 9.473684210526317e-07, + "loss": 3.2299, + "step": 18919 + }, + { + "epoch": 0.9957894736842106, + "grad_norm": 0.53515625, + "learning_rate": 9.356725146198831e-07, + "loss": 3.2992, + "step": 18920 + }, + { + "epoch": 0.9958421052631579, + "grad_norm": 0.5078125, + "learning_rate": 9.239766081871345e-07, + "loss": 2.6764, + "step": 18921 + }, + { + "epoch": 0.9958947368421053, + "grad_norm": 0.51171875, + "learning_rate": 9.12280701754386e-07, + "loss": 3.3256, + "step": 18922 + }, + { + "epoch": 0.9959473684210526, + "grad_norm": 0.474609375, + "learning_rate": 9.005847953216375e-07, + "loss": 3.0031, + "step": 18923 + }, + { + "epoch": 0.996, + "grad_norm": 0.5078125, + "learning_rate": 8.88888888888889e-07, + "loss": 3.5007, + "step": 18924 + }, + { + "epoch": 0.9960526315789474, + "grad_norm": 0.5078125, + "learning_rate": 8.771929824561404e-07, + "loss": 3.8728, + "step": 18925 + }, + { + "epoch": 0.9961052631578947, + "grad_norm": 0.498046875, + "learning_rate": 8.65497076023392e-07, + "loss": 3.1699, + "step": 18926 + }, + { + "epoch": 0.9961578947368421, + "grad_norm": 0.50390625, + "learning_rate": 8.538011695906434e-07, + "loss": 2.8043, + "step": 18927 + }, + { + "epoch": 0.9962105263157894, + "grad_norm": 0.51171875, + "learning_rate": 8.421052631578948e-07, + "loss": 3.2404, + "step": 18928 + }, + { + "epoch": 0.9962631578947369, + "grad_norm": 0.50390625, + "learning_rate": 8.304093567251462e-07, + "loss": 3.2155, + "step": 18929 + }, + { + "epoch": 0.9963157894736843, + "grad_norm": 0.494140625, + "learning_rate": 8.187134502923978e-07, + "loss": 3.2682, + "step": 18930 + }, + { + "epoch": 0.9963684210526316, + "grad_norm": 0.490234375, + "learning_rate": 8.070175438596491e-07, + "loss": 3.3304, + "step": 18931 + }, + { + "epoch": 0.996421052631579, + "grad_norm": 0.466796875, + "learning_rate": 7.953216374269005e-07, + "loss": 2.7649, + "step": 18932 + }, + { + "epoch": 0.9964736842105263, + "grad_norm": 0.447265625, + "learning_rate": 7.83625730994152e-07, + "loss": 2.8866, + "step": 18933 + }, + { + "epoch": 0.9965263157894737, + "grad_norm": 0.51171875, + "learning_rate": 7.719298245614035e-07, + "loss": 3.2201, + "step": 18934 + }, + { + "epoch": 0.996578947368421, + "grad_norm": 0.5, + "learning_rate": 7.60233918128655e-07, + "loss": 3.5144, + "step": 18935 + }, + { + "epoch": 0.9966315789473684, + "grad_norm": 0.494140625, + "learning_rate": 7.485380116959065e-07, + "loss": 3.5233, + "step": 18936 + }, + { + "epoch": 0.9966842105263158, + "grad_norm": 0.48828125, + "learning_rate": 7.368421052631579e-07, + "loss": 2.844, + "step": 18937 + }, + { + "epoch": 0.9967368421052631, + "grad_norm": 0.50390625, + "learning_rate": 7.251461988304094e-07, + "loss": 2.9476, + "step": 18938 + }, + { + "epoch": 0.9967894736842106, + "grad_norm": 0.48828125, + "learning_rate": 7.134502923976608e-07, + "loss": 3.2841, + "step": 18939 + }, + { + "epoch": 0.9968421052631579, + "grad_norm": 0.49609375, + "learning_rate": 7.017543859649123e-07, + "loss": 3.4248, + "step": 18940 + }, + { + "epoch": 0.9968947368421053, + "grad_norm": 0.6328125, + "learning_rate": 6.900584795321637e-07, + "loss": 2.972, + "step": 18941 + }, + { + "epoch": 0.9969473684210526, + "grad_norm": 0.5, + "learning_rate": 6.783625730994152e-07, + "loss": 3.0625, + "step": 18942 + }, + { + "epoch": 0.997, + "grad_norm": 0.54296875, + "learning_rate": 6.666666666666667e-07, + "loss": 3.6488, + "step": 18943 + }, + { + "epoch": 0.9970526315789474, + "grad_norm": 0.53125, + "learning_rate": 6.549707602339182e-07, + "loss": 3.011, + "step": 18944 + }, + { + "epoch": 0.9971052631578947, + "grad_norm": 0.4921875, + "learning_rate": 6.432748538011697e-07, + "loss": 3.1072, + "step": 18945 + }, + { + "epoch": 0.9971578947368421, + "grad_norm": 0.50390625, + "learning_rate": 6.315789473684211e-07, + "loss": 3.1897, + "step": 18946 + }, + { + "epoch": 0.9972105263157894, + "grad_norm": 0.4921875, + "learning_rate": 6.198830409356726e-07, + "loss": 3.427, + "step": 18947 + }, + { + "epoch": 0.9972631578947369, + "grad_norm": 0.5, + "learning_rate": 6.08187134502924e-07, + "loss": 3.3063, + "step": 18948 + }, + { + "epoch": 0.9973157894736842, + "grad_norm": 0.51171875, + "learning_rate": 5.964912280701755e-07, + "loss": 3.469, + "step": 18949 + }, + { + "epoch": 0.9973684210526316, + "grad_norm": 0.49609375, + "learning_rate": 5.847953216374269e-07, + "loss": 3.1516, + "step": 18950 + }, + { + "epoch": 0.997421052631579, + "grad_norm": 0.498046875, + "learning_rate": 5.730994152046784e-07, + "loss": 3.4487, + "step": 18951 + }, + { + "epoch": 0.9974736842105263, + "grad_norm": 0.490234375, + "learning_rate": 5.614035087719298e-07, + "loss": 3.2412, + "step": 18952 + }, + { + "epoch": 0.9975263157894737, + "grad_norm": 0.51953125, + "learning_rate": 5.497076023391813e-07, + "loss": 2.8502, + "step": 18953 + }, + { + "epoch": 0.997578947368421, + "grad_norm": 0.53515625, + "learning_rate": 5.380116959064328e-07, + "loss": 3.4207, + "step": 18954 + }, + { + "epoch": 0.9976315789473684, + "grad_norm": 0.50390625, + "learning_rate": 5.263157894736843e-07, + "loss": 3.479, + "step": 18955 + }, + { + "epoch": 0.9976842105263158, + "grad_norm": 0.474609375, + "learning_rate": 5.146198830409358e-07, + "loss": 3.1995, + "step": 18956 + }, + { + "epoch": 0.9977368421052631, + "grad_norm": 0.4921875, + "learning_rate": 5.029239766081872e-07, + "loss": 2.8603, + "step": 18957 + }, + { + "epoch": 0.9977894736842106, + "grad_norm": 0.48828125, + "learning_rate": 4.912280701754387e-07, + "loss": 2.8229, + "step": 18958 + }, + { + "epoch": 0.9978421052631579, + "grad_norm": 0.50390625, + "learning_rate": 4.7953216374269e-07, + "loss": 3.0461, + "step": 18959 + }, + { + "epoch": 0.9978947368421053, + "grad_norm": 0.5234375, + "learning_rate": 4.6783625730994155e-07, + "loss": 3.11, + "step": 18960 + }, + { + "epoch": 0.9979473684210526, + "grad_norm": 0.546875, + "learning_rate": 4.56140350877193e-07, + "loss": 3.364, + "step": 18961 + }, + { + "epoch": 0.998, + "grad_norm": 0.5, + "learning_rate": 4.444444444444445e-07, + "loss": 2.6014, + "step": 18962 + }, + { + "epoch": 0.9980526315789474, + "grad_norm": 0.515625, + "learning_rate": 4.32748538011696e-07, + "loss": 3.2184, + "step": 18963 + }, + { + "epoch": 0.9981052631578947, + "grad_norm": 0.5, + "learning_rate": 4.210526315789474e-07, + "loss": 3.4861, + "step": 18964 + }, + { + "epoch": 0.9981578947368421, + "grad_norm": 0.51171875, + "learning_rate": 4.093567251461989e-07, + "loss": 3.3056, + "step": 18965 + }, + { + "epoch": 0.9982105263157894, + "grad_norm": 0.51171875, + "learning_rate": 3.9766081871345027e-07, + "loss": 3.361, + "step": 18966 + }, + { + "epoch": 0.9982631578947369, + "grad_norm": 0.54296875, + "learning_rate": 3.8596491228070177e-07, + "loss": 3.1787, + "step": 18967 + }, + { + "epoch": 0.9983157894736842, + "grad_norm": 0.478515625, + "learning_rate": 3.7426900584795327e-07, + "loss": 3.201, + "step": 18968 + }, + { + "epoch": 0.9983684210526316, + "grad_norm": 0.50390625, + "learning_rate": 3.625730994152047e-07, + "loss": 3.1442, + "step": 18969 + }, + { + "epoch": 0.998421052631579, + "grad_norm": 0.5078125, + "learning_rate": 3.5087719298245616e-07, + "loss": 2.9533, + "step": 18970 + }, + { + "epoch": 0.9984736842105263, + "grad_norm": 0.51171875, + "learning_rate": 3.391812865497076e-07, + "loss": 3.4701, + "step": 18971 + }, + { + "epoch": 0.9985263157894737, + "grad_norm": 0.5078125, + "learning_rate": 3.274853801169591e-07, + "loss": 3.0543, + "step": 18972 + }, + { + "epoch": 0.998578947368421, + "grad_norm": 0.4765625, + "learning_rate": 3.1578947368421055e-07, + "loss": 3.1304, + "step": 18973 + }, + { + "epoch": 0.9986315789473684, + "grad_norm": 0.49609375, + "learning_rate": 3.04093567251462e-07, + "loss": 3.4778, + "step": 18974 + }, + { + "epoch": 0.9986842105263158, + "grad_norm": 0.52734375, + "learning_rate": 2.9239766081871344e-07, + "loss": 3.0237, + "step": 18975 + }, + { + "epoch": 0.9987368421052631, + "grad_norm": 0.51171875, + "learning_rate": 2.807017543859649e-07, + "loss": 2.5957, + "step": 18976 + }, + { + "epoch": 0.9987894736842106, + "grad_norm": 0.49609375, + "learning_rate": 2.690058479532164e-07, + "loss": 3.3196, + "step": 18977 + }, + { + "epoch": 0.9988421052631579, + "grad_norm": 0.4921875, + "learning_rate": 2.573099415204679e-07, + "loss": 3.2859, + "step": 18978 + }, + { + "epoch": 0.9988947368421053, + "grad_norm": 0.50390625, + "learning_rate": 2.4561403508771933e-07, + "loss": 3.394, + "step": 18979 + }, + { + "epoch": 0.9989473684210526, + "grad_norm": 0.4921875, + "learning_rate": 2.3391812865497077e-07, + "loss": 3.1491, + "step": 18980 + }, + { + "epoch": 0.999, + "grad_norm": 0.4921875, + "learning_rate": 2.2222222222222224e-07, + "loss": 3.0977, + "step": 18981 + }, + { + "epoch": 0.9990526315789474, + "grad_norm": 0.48828125, + "learning_rate": 2.105263157894737e-07, + "loss": 3.6204, + "step": 18982 + }, + { + "epoch": 0.9991052631578947, + "grad_norm": 0.5234375, + "learning_rate": 1.9883040935672514e-07, + "loss": 3.4118, + "step": 18983 + }, + { + "epoch": 0.9991578947368421, + "grad_norm": 0.49609375, + "learning_rate": 1.8713450292397663e-07, + "loss": 3.0766, + "step": 18984 + }, + { + "epoch": 0.9992105263157894, + "grad_norm": 0.52734375, + "learning_rate": 1.7543859649122808e-07, + "loss": 3.5315, + "step": 18985 + }, + { + "epoch": 0.9992631578947369, + "grad_norm": 0.5078125, + "learning_rate": 1.6374269005847955e-07, + "loss": 3.1416, + "step": 18986 + }, + { + "epoch": 0.9993157894736842, + "grad_norm": 0.515625, + "learning_rate": 1.52046783625731e-07, + "loss": 3.2438, + "step": 18987 + }, + { + "epoch": 0.9993684210526316, + "grad_norm": 0.5234375, + "learning_rate": 1.4035087719298244e-07, + "loss": 3.6965, + "step": 18988 + }, + { + "epoch": 0.999421052631579, + "grad_norm": 0.51953125, + "learning_rate": 1.2865497076023394e-07, + "loss": 2.9424, + "step": 18989 + }, + { + "epoch": 0.9994736842105263, + "grad_norm": 0.46484375, + "learning_rate": 1.1695906432748539e-07, + "loss": 3.282, + "step": 18990 + }, + { + "epoch": 0.9995263157894737, + "grad_norm": 0.51171875, + "learning_rate": 1.0526315789473685e-07, + "loss": 2.8389, + "step": 18991 + }, + { + "epoch": 0.999578947368421, + "grad_norm": 0.546875, + "learning_rate": 9.356725146198832e-08, + "loss": 3.3525, + "step": 18992 + }, + { + "epoch": 0.9996315789473684, + "grad_norm": 0.49609375, + "learning_rate": 8.187134502923978e-08, + "loss": 3.4968, + "step": 18993 + }, + { + "epoch": 0.9996842105263158, + "grad_norm": 0.50390625, + "learning_rate": 7.017543859649122e-08, + "loss": 3.1322, + "step": 18994 + }, + { + "epoch": 0.9997368421052631, + "grad_norm": 0.51171875, + "learning_rate": 5.847953216374269e-08, + "loss": 3.3716, + "step": 18995 + }, + { + "epoch": 0.9997894736842106, + "grad_norm": 0.482421875, + "learning_rate": 4.678362573099416e-08, + "loss": 3.3022, + "step": 18996 + }, + { + "epoch": 0.9998421052631579, + "grad_norm": 0.51171875, + "learning_rate": 3.508771929824561e-08, + "loss": 2.6884, + "step": 18997 + }, + { + "epoch": 0.9998947368421053, + "grad_norm": 0.498046875, + "learning_rate": 2.339181286549708e-08, + "loss": 3.2359, + "step": 18998 + }, + { + "epoch": 0.9999473684210526, + "grad_norm": 0.51171875, + "learning_rate": 1.169590643274854e-08, + "loss": 3.4151, + "step": 18999 + }, + { + "epoch": 1.0, + "grad_norm": 0.5, + "learning_rate": 0.0, + "loss": 3.4362, + "step": 19000 + }, + { + "epoch": 1.0, + "eval_loss": 3.2381837368011475, + "eval_runtime": 134.1592, + "eval_samples_per_second": 111.807, + "eval_steps_per_second": 13.976, + "step": 19000 + } + ], + "logging_steps": 1, + "max_steps": 19000, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.3317042610176e+17, + "train_batch_size": 15, + "trial_name": null, + "trial_params": null +}