| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9992422328870927, | |
| "eval_steps": 495, | |
| "global_step": 989, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 53.622260681945534, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 1.1058, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "eval_loss": 1.1559669971466064, | |
| "eval_runtime": 134.2853, | |
| "eval_samples_per_second": 9.733, | |
| "eval_steps_per_second": 0.276, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 40.68249059631914, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.1361, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 43.27766914416912, | |
| "learning_rate": 1.5e-06, | |
| "loss": 1.1577, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 64.72729420725338, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.1292, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 56.372636612148945, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.1337, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 50.14861511180959, | |
| "learning_rate": 3e-06, | |
| "loss": 1.1778, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 62.95626490942665, | |
| "learning_rate": 3.5e-06, | |
| "loss": 1.1317, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 25.825764536536617, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.0504, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 30.983005109311666, | |
| "learning_rate": 4.5e-06, | |
| "loss": 1.0636, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 42.69870581147392, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1061, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 30.4102208779435, | |
| "learning_rate": 4.9999968146329895e-06, | |
| "loss": 1.0577, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 19.477219524924408, | |
| "learning_rate": 4.999987258540075e-06, | |
| "loss": 1.0341, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 16.065115351967986, | |
| "learning_rate": 4.9999713317456065e-06, | |
| "loss": 1.0212, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 20.305999270879685, | |
| "learning_rate": 4.999949034290173e-06, | |
| "loss": 0.9934, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 15.296134261870147, | |
| "learning_rate": 4.999920366230593e-06, | |
| "loss": 1.0452, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 21.688573186878063, | |
| "learning_rate": 4.9998853276399215e-06, | |
| "loss": 1.0699, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 14.335847546461085, | |
| "learning_rate": 4.999843918607447e-06, | |
| "loss": 1.0098, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 22.3952228510413, | |
| "learning_rate": 4.999796139238694e-06, | |
| "loss": 1.0408, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 9.166221867386932, | |
| "learning_rate": 4.999741989655415e-06, | |
| "loss": 1.0028, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 11.838379188476093, | |
| "learning_rate": 4.999681469995601e-06, | |
| "loss": 1.0009, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 10.419092295341123, | |
| "learning_rate": 4.9996145804134735e-06, | |
| "loss": 1.022, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 9.35336523909701, | |
| "learning_rate": 4.999541321079486e-06, | |
| "loss": 0.9748, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 9.198682807281516, | |
| "learning_rate": 4.999461692180326e-06, | |
| "loss": 1.0549, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 12.547373233238517, | |
| "learning_rate": 4.999375693918911e-06, | |
| "loss": 0.9437, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.595672331068808, | |
| "learning_rate": 4.99928332651439e-06, | |
| "loss": 1.0163, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 12.993405908580561, | |
| "learning_rate": 4.999184590202142e-06, | |
| "loss": 1.0146, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.549924685378986, | |
| "learning_rate": 4.999079485233775e-06, | |
| "loss": 0.9542, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 10.198320823624972, | |
| "learning_rate": 4.998968011877129e-06, | |
| "loss": 0.9641, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.2545266805555, | |
| "learning_rate": 4.99885017041627e-06, | |
| "loss": 0.9832, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 7.9313407875475646, | |
| "learning_rate": 4.998725961151493e-06, | |
| "loss": 0.9685, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 9.807588698454, | |
| "learning_rate": 4.998595384399319e-06, | |
| "loss": 0.9193, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 9.876327124281431, | |
| "learning_rate": 4.998458440492498e-06, | |
| "loss": 0.9616, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.98501521873747, | |
| "learning_rate": 4.998315129780001e-06, | |
| "loss": 1.0099, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 10.044564125219294, | |
| "learning_rate": 4.998165452627025e-06, | |
| "loss": 0.9615, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 11.210345590334603, | |
| "learning_rate": 4.9980094094149945e-06, | |
| "loss": 1.1204, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 7.886709354787903, | |
| "learning_rate": 4.997847000541551e-06, | |
| "loss": 0.9783, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 7.5254666516991335, | |
| "learning_rate": 4.997678226420561e-06, | |
| "loss": 0.9161, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 8.272181825775892, | |
| "learning_rate": 4.99750308748211e-06, | |
| "loss": 0.9425, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 6.553520865941246, | |
| "learning_rate": 4.997321584172504e-06, | |
| "loss": 0.9962, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 8.229506711694187, | |
| "learning_rate": 4.997133716954266e-06, | |
| "loss": 0.9347, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 8.567484614549636, | |
| "learning_rate": 4.996939486306138e-06, | |
| "loss": 0.9485, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 6.61521839575945, | |
| "learning_rate": 4.996738892723076e-06, | |
| "loss": 0.883, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 8.10212907196996, | |
| "learning_rate": 4.99653193671625e-06, | |
| "loss": 1.1048, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 8.80462692647593, | |
| "learning_rate": 4.996318618813046e-06, | |
| "loss": 0.9816, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.097343998505614, | |
| "learning_rate": 4.996098939557062e-06, | |
| "loss": 0.9561, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.282799226817679, | |
| "learning_rate": 4.995872899508103e-06, | |
| "loss": 0.9617, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 12.216537774733089, | |
| "learning_rate": 4.995640499242187e-06, | |
| "loss": 0.9243, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 6.359198206461496, | |
| "learning_rate": 4.995401739351537e-06, | |
| "loss": 0.9436, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 6.346073589036576, | |
| "learning_rate": 4.995156620444584e-06, | |
| "loss": 0.8964, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 5.688812989274403, | |
| "learning_rate": 4.9949051431459615e-06, | |
| "loss": 0.9005, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.002996068199546, | |
| "learning_rate": 4.994647308096509e-06, | |
| "loss": 0.9334, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.083307002817586, | |
| "learning_rate": 4.994383115953266e-06, | |
| "loss": 0.8988, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 6.468005918770268, | |
| "learning_rate": 4.9941125673894705e-06, | |
| "loss": 0.9222, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 6.444789411001062, | |
| "learning_rate": 4.993835663094562e-06, | |
| "loss": 0.9338, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 7.644539402695577, | |
| "learning_rate": 4.9935524037741705e-06, | |
| "loss": 1.0552, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.40049014930346, | |
| "learning_rate": 4.9932627901501265e-06, | |
| "loss": 0.9328, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 8.906097856477972, | |
| "learning_rate": 4.99296682296045e-06, | |
| "loss": 1.1228, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 7.079585029259878, | |
| "learning_rate": 4.992664502959351e-06, | |
| "loss": 1.0413, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 10.34845460315542, | |
| "learning_rate": 4.992355830917232e-06, | |
| "loss": 0.9309, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.2525622909780765, | |
| "learning_rate": 4.992040807620678e-06, | |
| "loss": 0.9742, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.750158577512716, | |
| "learning_rate": 4.991719433872461e-06, | |
| "loss": 0.9132, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.2784065217091145, | |
| "learning_rate": 4.9913917104915374e-06, | |
| "loss": 0.9586, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.747179104322235, | |
| "learning_rate": 4.991057638313042e-06, | |
| "loss": 0.8983, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 13.39546333182503, | |
| "learning_rate": 4.990717218188286e-06, | |
| "loss": 0.9282, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 6.225226219601699, | |
| "learning_rate": 4.990370450984764e-06, | |
| "loss": 0.8964, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 7.297944379719202, | |
| "learning_rate": 4.990017337586137e-06, | |
| "loss": 0.886, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 6.684537834909945, | |
| "learning_rate": 4.989657878892245e-06, | |
| "loss": 0.9474, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 6.800644526106279, | |
| "learning_rate": 4.989292075819091e-06, | |
| "loss": 0.9222, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 5.935752803831776, | |
| "learning_rate": 4.988919929298851e-06, | |
| "loss": 0.9675, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 6.0962518911658385, | |
| "learning_rate": 4.988541440279862e-06, | |
| "loss": 0.8945, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 8.739181681630852, | |
| "learning_rate": 4.988156609726628e-06, | |
| "loss": 0.9216, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 6.531531611955777, | |
| "learning_rate": 4.987765438619806e-06, | |
| "loss": 0.9153, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 11.36936049546851, | |
| "learning_rate": 4.987367927956218e-06, | |
| "loss": 1.2286, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 6.073878979737398, | |
| "learning_rate": 4.9869640787488375e-06, | |
| "loss": 0.9115, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.8272340999466445, | |
| "learning_rate": 4.98655389202679e-06, | |
| "loss": 0.9509, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.397354616805974, | |
| "learning_rate": 4.986137368835351e-06, | |
| "loss": 1.0375, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 7.966255876371102, | |
| "learning_rate": 4.985714510235945e-06, | |
| "loss": 1.0293, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.185986548775906, | |
| "learning_rate": 4.985285317306141e-06, | |
| "loss": 0.9109, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 7.024522011830165, | |
| "learning_rate": 4.984849791139647e-06, | |
| "loss": 0.9061, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.995863723081675, | |
| "learning_rate": 4.984407932846311e-06, | |
| "loss": 0.8925, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.590869203505634, | |
| "learning_rate": 4.983959743552118e-06, | |
| "loss": 0.9194, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.037924790619887, | |
| "learning_rate": 4.983505224399188e-06, | |
| "loss": 0.9598, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.517206233677988, | |
| "learning_rate": 4.983044376545767e-06, | |
| "loss": 0.8645, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.230061445243754, | |
| "learning_rate": 4.982577201166232e-06, | |
| "loss": 0.9021, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 6.911990663487462, | |
| "learning_rate": 4.9821036994510816e-06, | |
| "loss": 0.858, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.984021384400987, | |
| "learning_rate": 4.981623872606938e-06, | |
| "loss": 0.8723, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 6.703129385119859, | |
| "learning_rate": 4.981137721856541e-06, | |
| "loss": 0.9518, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 7.12479649165451, | |
| "learning_rate": 4.980645248438746e-06, | |
| "loss": 0.9136, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 11.122916834992573, | |
| "learning_rate": 4.9801464536085184e-06, | |
| "loss": 0.9066, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 7.841215482759505, | |
| "learning_rate": 4.9796413386369344e-06, | |
| "loss": 1.0641, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.958073787403037, | |
| "learning_rate": 4.979129904811177e-06, | |
| "loss": 0.9263, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 7.020080058824307, | |
| "learning_rate": 4.978612153434527e-06, | |
| "loss": 0.9603, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.574162423590218, | |
| "learning_rate": 4.9780880858263684e-06, | |
| "loss": 0.8855, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.834495325366524, | |
| "learning_rate": 4.977557703322178e-06, | |
| "loss": 0.8873, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.461734817525809, | |
| "learning_rate": 4.9770210072735286e-06, | |
| "loss": 0.8965, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.653047237885576, | |
| "learning_rate": 4.976477999048078e-06, | |
| "loss": 1.0051, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.092048158281413, | |
| "learning_rate": 4.975928680029571e-06, | |
| "loss": 0.8779, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.730921274763532, | |
| "learning_rate": 4.975373051617831e-06, | |
| "loss": 0.8886, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 7.999386874766411, | |
| "learning_rate": 4.974811115228767e-06, | |
| "loss": 0.8932, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.569281154720038, | |
| "learning_rate": 4.974242872294354e-06, | |
| "loss": 0.9093, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.243416936742017, | |
| "learning_rate": 4.973668324262645e-06, | |
| "loss": 0.8787, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.425532736177959, | |
| "learning_rate": 4.973087472597754e-06, | |
| "loss": 0.8689, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.834945735385297, | |
| "learning_rate": 4.972500318779864e-06, | |
| "loss": 1.0439, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 7.010749117953628, | |
| "learning_rate": 4.971906864305214e-06, | |
| "loss": 1.0111, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 5.399008383715632, | |
| "learning_rate": 4.9713071106861e-06, | |
| "loss": 0.9488, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 6.852009850723007, | |
| "learning_rate": 4.970701059450872e-06, | |
| "loss": 0.9165, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 5.490567287919795, | |
| "learning_rate": 4.970088712143924e-06, | |
| "loss": 0.93, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 5.466174338824931, | |
| "learning_rate": 4.969470070325699e-06, | |
| "loss": 0.8258, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 5.3052755303637325, | |
| "learning_rate": 4.968845135572678e-06, | |
| "loss": 0.8643, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 5.79116786116443, | |
| "learning_rate": 4.968213909477376e-06, | |
| "loss": 0.9092, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 6.243372776357591, | |
| "learning_rate": 4.967576393648344e-06, | |
| "loss": 0.8854, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 5.554172933410867, | |
| "learning_rate": 4.96693258971016e-06, | |
| "loss": 0.9373, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 14.37748982425115, | |
| "learning_rate": 4.9662824993034245e-06, | |
| "loss": 0.8937, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 5.41048524226914, | |
| "learning_rate": 4.965626124084759e-06, | |
| "loss": 0.9385, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 5.3566507966065355, | |
| "learning_rate": 4.9649634657268e-06, | |
| "loss": 0.9047, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 6.532323859056279, | |
| "learning_rate": 4.964294525918196e-06, | |
| "loss": 0.8841, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 8.816489663062468, | |
| "learning_rate": 4.963619306363602e-06, | |
| "loss": 0.8959, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 7.372965529585357, | |
| "learning_rate": 4.962937808783675e-06, | |
| "loss": 1.0141, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 5.68129098809545, | |
| "learning_rate": 4.962250034915072e-06, | |
| "loss": 0.8776, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 6.138429945353462, | |
| "learning_rate": 4.961555986510443e-06, | |
| "loss": 0.9064, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 6.2074388893831625, | |
| "learning_rate": 4.960855665338424e-06, | |
| "loss": 0.8984, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 6.423407890211476, | |
| "learning_rate": 4.960149073183643e-06, | |
| "loss": 0.8478, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 15.084543572578946, | |
| "learning_rate": 4.959436211846702e-06, | |
| "loss": 1.0345, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.093406275304441, | |
| "learning_rate": 4.958717083144182e-06, | |
| "loss": 0.8912, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.398558925299228, | |
| "learning_rate": 4.9579916889086336e-06, | |
| "loss": 0.9241, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.573650372167155, | |
| "learning_rate": 4.957260030988575e-06, | |
| "loss": 0.9476, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.430534040644023, | |
| "learning_rate": 4.956522111248483e-06, | |
| "loss": 0.9208, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.56067970590197, | |
| "learning_rate": 4.955777931568797e-06, | |
| "loss": 0.9274, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.601548738488328, | |
| "learning_rate": 4.955027493845903e-06, | |
| "loss": 0.8426, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 8.927604688533856, | |
| "learning_rate": 4.954270799992138e-06, | |
| "loss": 0.8618, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.6362861369729895, | |
| "learning_rate": 4.953507851935779e-06, | |
| "loss": 0.921, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 7.966230987922584, | |
| "learning_rate": 4.952738651621043e-06, | |
| "loss": 0.92, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.8283544064228145, | |
| "learning_rate": 4.9519632010080765e-06, | |
| "loss": 0.8719, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 6.733524485804168, | |
| "learning_rate": 4.951181502072957e-06, | |
| "loss": 0.8664, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 4.93365687921664, | |
| "learning_rate": 4.950393556807682e-06, | |
| "loss": 0.9109, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 5.720266452276143, | |
| "learning_rate": 4.9495993672201675e-06, | |
| "loss": 1.0141, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 6.577392904905377, | |
| "learning_rate": 4.948798935334242e-06, | |
| "loss": 0.8947, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 5.248018611453421, | |
| "learning_rate": 4.947992263189641e-06, | |
| "loss": 0.8519, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 4.964178402688073, | |
| "learning_rate": 4.947179352842001e-06, | |
| "loss": 0.8732, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 7.727649852322783, | |
| "learning_rate": 4.946360206362858e-06, | |
| "loss": 0.9003, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 5.678031094542374, | |
| "learning_rate": 4.945534825839637e-06, | |
| "loss": 0.931, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 4.8975198222340195, | |
| "learning_rate": 4.944703213375648e-06, | |
| "loss": 0.9022, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 5.2460458320933485, | |
| "learning_rate": 4.943865371090087e-06, | |
| "loss": 0.8835, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.080524254102179, | |
| "learning_rate": 4.94302130111802e-06, | |
| "loss": 0.8957, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.614189755845358, | |
| "learning_rate": 4.942171005610385e-06, | |
| "loss": 0.8961, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.2283236967224, | |
| "learning_rate": 4.941314486733986e-06, | |
| "loss": 0.82, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 4.903937090069375, | |
| "learning_rate": 4.940451746671484e-06, | |
| "loss": 0.8733, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.044592476065338, | |
| "learning_rate": 4.939582787621394e-06, | |
| "loss": 0.9423, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 10.766060880843327, | |
| "learning_rate": 4.938707611798078e-06, | |
| "loss": 0.8449, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 4.623203037666776, | |
| "learning_rate": 4.937826221431742e-06, | |
| "loss": 0.8623, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.0474657841525525, | |
| "learning_rate": 4.936938618768427e-06, | |
| "loss": 0.9026, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 31.531685747916384, | |
| "learning_rate": 4.936044806070005e-06, | |
| "loss": 0.8545, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.2640420937536, | |
| "learning_rate": 4.935144785614173e-06, | |
| "loss": 0.8793, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.645683842200697, | |
| "learning_rate": 4.934238559694448e-06, | |
| "loss": 0.8472, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 6.108168033210612, | |
| "learning_rate": 4.93332613062016e-06, | |
| "loss": 1.004, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 5.832130176498325, | |
| "learning_rate": 4.932407500716445e-06, | |
| "loss": 0.9278, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 6.2388094461996575, | |
| "learning_rate": 4.9314826723242425e-06, | |
| "loss": 1.0358, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 6.954370983448125, | |
| "learning_rate": 4.930551647800287e-06, | |
| "loss": 0.9919, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.997616854364736, | |
| "learning_rate": 4.9296144295171025e-06, | |
| "loss": 0.8445, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 7.115062150515696, | |
| "learning_rate": 4.928671019862995e-06, | |
| "loss": 0.8546, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 6.146468237609964, | |
| "learning_rate": 4.927721421242051e-06, | |
| "loss": 0.8879, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 5.269100479953049, | |
| "learning_rate": 4.926765636074124e-06, | |
| "loss": 0.8928, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.829206775414763, | |
| "learning_rate": 4.925803666794839e-06, | |
| "loss": 0.8912, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 13.500834231735407, | |
| "learning_rate": 4.924835515855573e-06, | |
| "loss": 0.89, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 5.876726093761002, | |
| "learning_rate": 4.923861185723461e-06, | |
| "loss": 1.0295, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 6.552028600594945, | |
| "learning_rate": 4.92288067888138e-06, | |
| "loss": 0.8523, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 9.604721277839626, | |
| "learning_rate": 4.921893997827951e-06, | |
| "loss": 0.8793, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 19.05047261104596, | |
| "learning_rate": 4.920901145077527e-06, | |
| "loss": 1.0183, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 8.670208792155977, | |
| "learning_rate": 4.919902123160187e-06, | |
| "loss": 0.916, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 5.963579222763698, | |
| "learning_rate": 4.918896934621734e-06, | |
| "loss": 0.8927, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 15.698308686475166, | |
| "learning_rate": 4.917885582023683e-06, | |
| "loss": 0.973, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 15.158791715235347, | |
| "learning_rate": 4.9168680679432565e-06, | |
| "loss": 0.8786, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 8.935267549779024, | |
| "learning_rate": 4.915844394973379e-06, | |
| "loss": 0.9105, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 7.85639856911323, | |
| "learning_rate": 4.914814565722671e-06, | |
| "loss": 0.8709, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 7.730019511721717, | |
| "learning_rate": 4.913778582815439e-06, | |
| "loss": 0.9203, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 30.04084618027315, | |
| "learning_rate": 4.912736448891672e-06, | |
| "loss": 0.8835, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 11.53316056998077, | |
| "learning_rate": 4.9116881666070325e-06, | |
| "loss": 0.8722, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 36.52217384345542, | |
| "learning_rate": 4.9106337386328524e-06, | |
| "loss": 0.8615, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 115.42434006092905, | |
| "learning_rate": 4.9095731676561246e-06, | |
| "loss": 0.8587, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 15.54123495731584, | |
| "learning_rate": 4.908506456379493e-06, | |
| "loss": 0.863, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 85.99384572854959, | |
| "learning_rate": 4.907433607521252e-06, | |
| "loss": 0.894, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 36.13462335675736, | |
| "learning_rate": 4.906354623815336e-06, | |
| "loss": 0.9055, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 27.02795167506559, | |
| "learning_rate": 4.905269508011312e-06, | |
| "loss": 0.9135, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 78.08453850338199, | |
| "learning_rate": 4.904178262874374e-06, | |
| "loss": 0.9165, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 36.25089616666813, | |
| "learning_rate": 4.903080891185335e-06, | |
| "loss": 0.9778, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 49.253048863400004, | |
| "learning_rate": 4.9019773957406194e-06, | |
| "loss": 0.8498, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 32.09689217815892, | |
| "learning_rate": 4.900867779352258e-06, | |
| "loss": 0.9056, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 26.828381083665327, | |
| "learning_rate": 4.899752044847881e-06, | |
| "loss": 0.8266, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 144.02426662483788, | |
| "learning_rate": 4.898630195070705e-06, | |
| "loss": 0.8645, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 47.52439763925523, | |
| "learning_rate": 4.897502232879533e-06, | |
| "loss": 0.8619, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 74.68661681875047, | |
| "learning_rate": 4.896368161148744e-06, | |
| "loss": 0.8985, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 75.68493630646316, | |
| "learning_rate": 4.895227982768287e-06, | |
| "loss": 0.8809, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 29.706946472395224, | |
| "learning_rate": 4.89408170064367e-06, | |
| "loss": 0.8396, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 161.48487196225275, | |
| "learning_rate": 4.892929317695957e-06, | |
| "loss": 0.8857, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 41.09638985280378, | |
| "learning_rate": 4.891770836861757e-06, | |
| "loss": 0.8633, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 69.69119197887811, | |
| "learning_rate": 4.890606261093221e-06, | |
| "loss": 1.0396, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 88.82274883602727, | |
| "learning_rate": 4.889435593358029e-06, | |
| "loss": 0.8651, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 81.79928830463821, | |
| "learning_rate": 4.888258836639386e-06, | |
| "loss": 0.8389, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 29.87578836447596, | |
| "learning_rate": 4.887075993936014e-06, | |
| "loss": 0.8775, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 108.15216101207292, | |
| "learning_rate": 4.885887068262143e-06, | |
| "loss": 0.8656, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 37.34405700410443, | |
| "learning_rate": 4.884692062647506e-06, | |
| "loss": 0.8682, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 18.005378230428164, | |
| "learning_rate": 4.883490980137327e-06, | |
| "loss": 0.9104, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 13.182657069815662, | |
| "learning_rate": 4.8822838237923164e-06, | |
| "loss": 0.8802, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 20.126124496397114, | |
| "learning_rate": 4.881070596688664e-06, | |
| "loss": 0.9229, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 8.025782615380805, | |
| "learning_rate": 4.8798513019180295e-06, | |
| "loss": 1.0137, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 11.303199949810612, | |
| "learning_rate": 4.878625942587531e-06, | |
| "loss": 0.8844, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 4.929950753578784, | |
| "learning_rate": 4.8773945218197475e-06, | |
| "loss": 0.8617, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 10.422846526317487, | |
| "learning_rate": 4.876157042752698e-06, | |
| "loss": 0.8628, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 5.406124330714249, | |
| "learning_rate": 4.874913508539844e-06, | |
| "loss": 0.8978, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 10.297329933831254, | |
| "learning_rate": 4.8736639223500734e-06, | |
| "loss": 0.9213, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 39.92063204593234, | |
| "learning_rate": 4.872408287367702e-06, | |
| "loss": 0.9151, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 12.707518091656361, | |
| "learning_rate": 4.871146606792455e-06, | |
| "loss": 0.8726, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 12.451656943060126, | |
| "learning_rate": 4.869878883839464e-06, | |
| "loss": 0.8831, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 11.149775907376688, | |
| "learning_rate": 4.868605121739261e-06, | |
| "loss": 0.7839, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 6.069136050487129, | |
| "learning_rate": 4.867325323737765e-06, | |
| "loss": 0.8095, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 33.341330536587115, | |
| "learning_rate": 4.866039493096276e-06, | |
| "loss": 0.8652, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 12.327283349644702, | |
| "learning_rate": 4.86474763309147e-06, | |
| "loss": 0.9332, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 14.886718679591452, | |
| "learning_rate": 4.863449747015384e-06, | |
| "loss": 0.9123, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 18.439054151579196, | |
| "learning_rate": 4.862145838175413e-06, | |
| "loss": 0.9952, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 6.450032781643884, | |
| "learning_rate": 4.8608359098943014e-06, | |
| "loss": 0.8619, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 5.76591848048908, | |
| "learning_rate": 4.859519965510129e-06, | |
| "loss": 0.8904, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 19.230556702347112, | |
| "learning_rate": 4.858198008376308e-06, | |
| "loss": 0.8454, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 12.812798200753768, | |
| "learning_rate": 4.856870041861576e-06, | |
| "loss": 0.8541, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 20.396035600104717, | |
| "learning_rate": 4.8555360693499786e-06, | |
| "loss": 0.8172, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 28.324370260593682, | |
| "learning_rate": 4.854196094240872e-06, | |
| "loss": 0.8665, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 24.090190708297367, | |
| "learning_rate": 4.8528501199489045e-06, | |
| "loss": 0.8964, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 51.69442396543579, | |
| "learning_rate": 4.8514981499040146e-06, | |
| "loss": 1.0263, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 13.682340026316318, | |
| "learning_rate": 4.850140187551417e-06, | |
| "loss": 0.8913, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 10.568651019154684, | |
| "learning_rate": 4.848776236351602e-06, | |
| "loss": 0.8711, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 17.95282322327484, | |
| "learning_rate": 4.847406299780316e-06, | |
| "loss": 0.8534, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 11.086531755411718, | |
| "learning_rate": 4.846030381328559e-06, | |
| "loss": 0.8577, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 20.43851904662153, | |
| "learning_rate": 4.8446484845025754e-06, | |
| "loss": 0.8711, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 9.019286512020804, | |
| "learning_rate": 4.843260612823844e-06, | |
| "loss": 0.8493, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 28.047694724485186, | |
| "learning_rate": 4.84186676982907e-06, | |
| "loss": 0.8913, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 12.170251493093959, | |
| "learning_rate": 4.840466959070174e-06, | |
| "loss": 0.8561, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 8.931963450098133, | |
| "learning_rate": 4.839061184114286e-06, | |
| "loss": 0.831, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 11.80446928845208, | |
| "learning_rate": 4.837649448543731e-06, | |
| "loss": 0.8314, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 6.844616516126078, | |
| "learning_rate": 4.836231755956028e-06, | |
| "loss": 0.8366, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 36.62950097854142, | |
| "learning_rate": 4.834808109963873e-06, | |
| "loss": 0.8674, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 7.123248446431571, | |
| "learning_rate": 4.8333785141951335e-06, | |
| "loss": 0.8705, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 53.32802137920739, | |
| "learning_rate": 4.83194297229284e-06, | |
| "loss": 0.8973, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 17.134182129435587, | |
| "learning_rate": 4.830501487915175e-06, | |
| "loss": 0.8552, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 18.056985775383417, | |
| "learning_rate": 4.829054064735463e-06, | |
| "loss": 0.8743, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 23.339383649662626, | |
| "learning_rate": 4.827600706442164e-06, | |
| "loss": 0.8401, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 44.31150837853898, | |
| "learning_rate": 4.826141416738861e-06, | |
| "loss": 0.8045, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 46.76596172244923, | |
| "learning_rate": 4.824676199344253e-06, | |
| "loss": 0.8392, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 19.466255218481894, | |
| "learning_rate": 4.823205057992145e-06, | |
| "loss": 0.8422, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 31.8224396652641, | |
| "learning_rate": 4.821727996431435e-06, | |
| "loss": 1.0119, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 28.30370717907097, | |
| "learning_rate": 4.8202450184261114e-06, | |
| "loss": 0.8858, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 14.967566855603655, | |
| "learning_rate": 4.8187561277552376e-06, | |
| "loss": 0.8458, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 40.45152173099994, | |
| "learning_rate": 4.817261328212942e-06, | |
| "loss": 0.8414, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 44.04696789627551, | |
| "learning_rate": 4.815760623608415e-06, | |
| "loss": 0.8429, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 22.669028244807055, | |
| "learning_rate": 4.8142540177658925e-06, | |
| "loss": 0.8377, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 36.679299334418545, | |
| "learning_rate": 4.812741514524647e-06, | |
| "loss": 0.8908, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 19.281368535984196, | |
| "learning_rate": 4.811223117738981e-06, | |
| "loss": 0.8768, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 44.69042639681561, | |
| "learning_rate": 4.809698831278217e-06, | |
| "loss": 0.8919, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 20.917629234011702, | |
| "learning_rate": 4.808168659026683e-06, | |
| "loss": 0.8443, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 20.01311483179014, | |
| "learning_rate": 4.806632604883709e-06, | |
| "loss": 0.8775, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 68.90691376267046, | |
| "learning_rate": 4.805090672763609e-06, | |
| "loss": 0.8913, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 41.87111801337849, | |
| "learning_rate": 4.803542866595681e-06, | |
| "loss": 0.823, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 25.071970545031796, | |
| "learning_rate": 4.801989190324187e-06, | |
| "loss": 0.8796, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 109.413625000989, | |
| "learning_rate": 4.800429647908354e-06, | |
| "loss": 0.858, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 43.38280987527835, | |
| "learning_rate": 4.7988642433223535e-06, | |
| "loss": 0.8521, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 32.13753250435139, | |
| "learning_rate": 4.797292980555293e-06, | |
| "loss": 0.8757, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 46.78789327684748, | |
| "learning_rate": 4.795715863611212e-06, | |
| "loss": 0.895, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 43.281917729979604, | |
| "learning_rate": 4.7941328965090705e-06, | |
| "loss": 0.8568, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 41.384686856921995, | |
| "learning_rate": 4.7925440832827305e-06, | |
| "loss": 0.9944, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 40.92527648821274, | |
| "learning_rate": 4.790949427980956e-06, | |
| "loss": 0.8338, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 82.6200683793308, | |
| "learning_rate": 4.789348934667397e-06, | |
| "loss": 0.844, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 46.48434653058032, | |
| "learning_rate": 4.787742607420579e-06, | |
| "loss": 0.8408, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 100.35892484768685, | |
| "learning_rate": 4.786130450333897e-06, | |
| "loss": 0.8429, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 104.75380555405746, | |
| "learning_rate": 4.784512467515599e-06, | |
| "loss": 0.8469, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 62.43051068810147, | |
| "learning_rate": 4.782888663088781e-06, | |
| "loss": 0.8564, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 93.44666911645837, | |
| "learning_rate": 4.7812590411913755e-06, | |
| "loss": 0.8957, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 19.862765945899945, | |
| "learning_rate": 4.779623605976135e-06, | |
| "loss": 1.0236, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 46.26361057838037, | |
| "learning_rate": 4.777982361610629e-06, | |
| "loss": 0.9026, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 52.97493760905037, | |
| "learning_rate": 4.77633531227723e-06, | |
| "loss": 0.871, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 34.041391067268314, | |
| "learning_rate": 4.774682462173105e-06, | |
| "loss": 0.8587, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 83.06466502493622, | |
| "learning_rate": 4.773023815510199e-06, | |
| "loss": 0.9954, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 32.778966883694586, | |
| "learning_rate": 4.771359376515231e-06, | |
| "loss": 0.8576, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 58.90680833070677, | |
| "learning_rate": 4.769689149429683e-06, | |
| "loss": 0.8974, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 18.71277300441414, | |
| "learning_rate": 4.768013138509781e-06, | |
| "loss": 0.804, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 107.32301554273414, | |
| "learning_rate": 4.766331348026493e-06, | |
| "loss": 0.7948, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 32.37059166820261, | |
| "learning_rate": 4.764643782265517e-06, | |
| "loss": 0.7972, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 60.12326441678363, | |
| "learning_rate": 4.762950445527264e-06, | |
| "loss": 0.8201, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 53.85163139056427, | |
| "learning_rate": 4.7612513421268546e-06, | |
| "loss": 0.8695, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 62.74196467038289, | |
| "learning_rate": 4.759546476394103e-06, | |
| "loss": 0.8713, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 39.83211475405053, | |
| "learning_rate": 4.757835852673506e-06, | |
| "loss": 0.7924, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 44.36619825791928, | |
| "learning_rate": 4.756119475324237e-06, | |
| "loss": 0.9933, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 37.10810986202249, | |
| "learning_rate": 4.754397348720129e-06, | |
| "loss": 0.8461, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 168.41339372943395, | |
| "learning_rate": 4.752669477249666e-06, | |
| "loss": 1.0045, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 38.375750184514935, | |
| "learning_rate": 4.750935865315972e-06, | |
| "loss": 0.8597, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 139.29153601089564, | |
| "learning_rate": 4.749196517336798e-06, | |
| "loss": 0.8526, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 84.96134212632315, | |
| "learning_rate": 4.7474514377445155e-06, | |
| "loss": 0.8705, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 79.39624824781504, | |
| "learning_rate": 4.745700630986097e-06, | |
| "loss": 0.8023, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 20.962475475808905, | |
| "learning_rate": 4.743944101523116e-06, | |
| "loss": 0.9936, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 18.755287063878363, | |
| "learning_rate": 4.742181853831721e-06, | |
| "loss": 0.822, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 14.452455674915326, | |
| "learning_rate": 4.740413892402639e-06, | |
| "loss": 0.8518, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 123.72945221996332, | |
| "learning_rate": 4.7386402217411555e-06, | |
| "loss": 0.846, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 114.85788665830623, | |
| "learning_rate": 4.7368608463671015e-06, | |
| "loss": 0.8264, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 47.28480009336326, | |
| "learning_rate": 4.73507577081485e-06, | |
| "loss": 0.8695, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 41.276139747783475, | |
| "learning_rate": 4.733284999633297e-06, | |
| "loss": 0.8486, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 17.925091199690936, | |
| "learning_rate": 4.731488537385853e-06, | |
| "loss": 0.8492, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 61.187483963098465, | |
| "learning_rate": 4.7296863886504315e-06, | |
| "loss": 0.8728, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 13.651808271590621, | |
| "learning_rate": 4.7278785580194365e-06, | |
| "loss": 0.8495, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 27.644885260501276, | |
| "learning_rate": 4.726065050099752e-06, | |
| "loss": 0.848, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 53.55797464707186, | |
| "learning_rate": 4.7242458695127275e-06, | |
| "loss": 0.894, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 48.26707421432808, | |
| "learning_rate": 4.72242102089417e-06, | |
| "loss": 0.8572, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 35.94106221579397, | |
| "learning_rate": 4.720590508894329e-06, | |
| "loss": 0.8329, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 118.69134781259584, | |
| "learning_rate": 4.718754338177887e-06, | |
| "loss": 0.8335, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 10.673509155086721, | |
| "learning_rate": 4.7169125134239456e-06, | |
| "loss": 0.8335, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 20.703878872323784, | |
| "learning_rate": 4.715065039326015e-06, | |
| "loss": 0.8741, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 42.51236987142145, | |
| "learning_rate": 4.713211920592003e-06, | |
| "loss": 0.8686, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 31.669223786445617, | |
| "learning_rate": 4.711353161944199e-06, | |
| "loss": 0.844, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 55.15593720360342, | |
| "learning_rate": 4.709488768119266e-06, | |
| "loss": 0.8519, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 95.39927875998099, | |
| "learning_rate": 4.707618743868226e-06, | |
| "loss": 0.8131, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 10.654454638290147, | |
| "learning_rate": 4.705743093956452e-06, | |
| "loss": 0.9639, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 38.23375131308745, | |
| "learning_rate": 4.703861823163649e-06, | |
| "loss": 0.8365, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 33.691960306044486, | |
| "learning_rate": 4.701974936283848e-06, | |
| "loss": 0.9664, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 10.611314812796477, | |
| "learning_rate": 4.700082438125391e-06, | |
| "loss": 0.8283, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 20.55619262108289, | |
| "learning_rate": 4.6981843335109176e-06, | |
| "loss": 0.8272, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 34.03935919948644, | |
| "learning_rate": 4.696280627277356e-06, | |
| "loss": 0.8481, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 20.169566617502596, | |
| "learning_rate": 4.69437132427591e-06, | |
| "loss": 0.8419, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 40.583282897440746, | |
| "learning_rate": 4.692456429372044e-06, | |
| "loss": 0.8535, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 8.056427978564315, | |
| "learning_rate": 4.690535947445471e-06, | |
| "loss": 0.851, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 24.00351867462924, | |
| "learning_rate": 4.688609883390144e-06, | |
| "loss": 0.8446, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 16.11523409967761, | |
| "learning_rate": 4.686678242114239e-06, | |
| "loss": 0.7832, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 8.214685204880942, | |
| "learning_rate": 4.6847410285401465e-06, | |
| "loss": 0.839, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 52.51672691072579, | |
| "learning_rate": 4.682798247604453e-06, | |
| "loss": 0.8327, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 8.016142819670371, | |
| "learning_rate": 4.680849904257938e-06, | |
| "loss": 0.8135, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 46.655885402782395, | |
| "learning_rate": 4.67889600346555e-06, | |
| "loss": 0.7983, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 16.138965226785817, | |
| "learning_rate": 4.676936550206402e-06, | |
| "loss": 0.8542, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 55.643105621600114, | |
| "learning_rate": 4.674971549473757e-06, | |
| "loss": 0.9038, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 16.31989957270255, | |
| "learning_rate": 4.673001006275013e-06, | |
| "loss": 0.8399, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 14.138316979026195, | |
| "learning_rate": 4.671024925631694e-06, | |
| "loss": 0.7812, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 50.2685770405374, | |
| "learning_rate": 4.669043312579433e-06, | |
| "loss": 0.8358, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 9.047670202154327, | |
| "learning_rate": 4.667056172167962e-06, | |
| "loss": 0.8361, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 15.329346838170585, | |
| "learning_rate": 4.665063509461098e-06, | |
| "loss": 0.8112, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 34.581957147308756, | |
| "learning_rate": 4.6630653295367286e-06, | |
| "loss": 0.8819, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 70.00837894867692, | |
| "learning_rate": 4.6610616374868066e-06, | |
| "loss": 0.8316, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 19.54707036059632, | |
| "learning_rate": 4.659052438417326e-06, | |
| "loss": 0.8171, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 122.51019371942445, | |
| "learning_rate": 4.6570377374483155e-06, | |
| "loss": 0.8908, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 24.88314319301423, | |
| "learning_rate": 4.655017539713826e-06, | |
| "loss": 0.8623, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 17.959493645537147, | |
| "learning_rate": 4.652991850361912e-06, | |
| "loss": 0.8354, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 13.1801020702163, | |
| "learning_rate": 4.650960674554627e-06, | |
| "loss": 0.8247, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 14.455355688503463, | |
| "learning_rate": 4.648924017468003e-06, | |
| "loss": 0.8286, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 8.731820345524651, | |
| "learning_rate": 4.64688188429204e-06, | |
| "loss": 0.8327, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 42.87517105426398, | |
| "learning_rate": 4.644834280230693e-06, | |
| "loss": 0.8393, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 50.868804589197836, | |
| "learning_rate": 4.642781210501858e-06, | |
| "loss": 0.8598, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 5.919995360552276, | |
| "learning_rate": 4.640722680337358e-06, | |
| "loss": 0.962, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 26.64669124982461, | |
| "learning_rate": 4.638658694982936e-06, | |
| "loss": 0.798, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 26.1192454778817, | |
| "learning_rate": 4.6365892596982295e-06, | |
| "loss": 0.8762, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 18.9062598091514, | |
| "learning_rate": 4.634514379756769e-06, | |
| "loss": 0.7937, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 11.482713649160976, | |
| "learning_rate": 4.632434060445956e-06, | |
| "loss": 0.8616, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 8.370599147559595, | |
| "learning_rate": 4.6303483070670574e-06, | |
| "loss": 0.7956, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 12.925585913640592, | |
| "learning_rate": 4.628257124935183e-06, | |
| "loss": 0.8246, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 13.662180432298918, | |
| "learning_rate": 4.626160519379279e-06, | |
| "loss": 0.8376, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 30.308232889166852, | |
| "learning_rate": 4.624058495742115e-06, | |
| "loss": 0.8591, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 64.26811094645683, | |
| "learning_rate": 4.621951059380259e-06, | |
| "loss": 0.8079, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 34.37337832128252, | |
| "learning_rate": 4.619838215664082e-06, | |
| "loss": 0.7918, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 15.079822107556895, | |
| "learning_rate": 4.617719969977729e-06, | |
| "loss": 0.8811, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 11.388476367649812, | |
| "learning_rate": 4.615596327719111e-06, | |
| "loss": 0.8445, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 8.881147580394376, | |
| "learning_rate": 4.613467294299893e-06, | |
| "loss": 0.8489, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 7.3102827631277885, | |
| "learning_rate": 4.611332875145476e-06, | |
| "loss": 0.8141, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 19.341931669078043, | |
| "learning_rate": 4.609193075694989e-06, | |
| "loss": 0.8301, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 16.293526833129757, | |
| "learning_rate": 4.607047901401267e-06, | |
| "loss": 0.8472, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 12.701861210256677, | |
| "learning_rate": 4.604897357730846e-06, | |
| "loss": 0.8258, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 12.022062250305899, | |
| "learning_rate": 4.60274145016394e-06, | |
| "loss": 0.7865, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 8.942347186149718, | |
| "learning_rate": 4.600580184194436e-06, | |
| "loss": 0.8538, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 7.084710601938015, | |
| "learning_rate": 4.598413565329876e-06, | |
| "loss": 0.8213, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 8.773416021429492, | |
| "learning_rate": 4.596241599091438e-06, | |
| "loss": 0.8281, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 11.412681654994278, | |
| "learning_rate": 4.59406429101393e-06, | |
| "loss": 0.8073, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 33.38456658772077, | |
| "learning_rate": 4.591881646645775e-06, | |
| "loss": 0.8078, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 20.116008377655895, | |
| "learning_rate": 4.589693671548989e-06, | |
| "loss": 0.8791, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 25.441358730169238, | |
| "learning_rate": 4.587500371299176e-06, | |
| "loss": 0.8096, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 10.632165430252845, | |
| "learning_rate": 4.585301751485508e-06, | |
| "loss": 0.8124, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 24.69121638703892, | |
| "learning_rate": 4.583097817710716e-06, | |
| "loss": 0.8292, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 42.373835243000144, | |
| "learning_rate": 4.580888575591068e-06, | |
| "loss": 0.8124, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 15.688601624623248, | |
| "learning_rate": 4.578674030756364e-06, | |
| "loss": 0.834, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 12.872910424438674, | |
| "learning_rate": 4.5764541888499116e-06, | |
| "loss": 0.8293, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 11.085141619803203, | |
| "learning_rate": 4.574229055528522e-06, | |
| "loss": 0.8419, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 20.128058817237306, | |
| "learning_rate": 4.571998636462487e-06, | |
| "loss": 0.8072, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 45.8693129268715, | |
| "learning_rate": 4.5697629373355694e-06, | |
| "loss": 0.8516, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 12.540832336612032, | |
| "learning_rate": 4.567521963844987e-06, | |
| "loss": 0.7809, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 7.272618009907343, | |
| "learning_rate": 4.5652757217014e-06, | |
| "loss": 0.8836, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 25.542789288767313, | |
| "learning_rate": 4.56302421662889e-06, | |
| "loss": 0.8266, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 21.60322316347993, | |
| "learning_rate": 4.560767454364955e-06, | |
| "loss": 0.8591, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 23.667006186382807, | |
| "learning_rate": 4.5585054406604865e-06, | |
| "loss": 0.8391, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 9.485503231699525, | |
| "learning_rate": 4.556238181279761e-06, | |
| "loss": 0.8561, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 24.57297216692732, | |
| "learning_rate": 4.55396568200042e-06, | |
| "loss": 0.8277, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 13.729895392458525, | |
| "learning_rate": 4.551687948613459e-06, | |
| "loss": 0.8032, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 18.865208120754964, | |
| "learning_rate": 4.549404986923213e-06, | |
| "loss": 0.863, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 22.190589070773914, | |
| "learning_rate": 4.547116802747335e-06, | |
| "loss": 0.8989, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 21.450617982266408, | |
| "learning_rate": 4.544823401916794e-06, | |
| "loss": 0.792, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 35.65422229133744, | |
| "learning_rate": 4.542524790275848e-06, | |
| "loss": 0.845, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 11.364569307665759, | |
| "learning_rate": 4.5402209736820325e-06, | |
| "loss": 0.829, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 21.94205192387575, | |
| "learning_rate": 4.537911958006149e-06, | |
| "loss": 0.8234, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 10.63181070102972, | |
| "learning_rate": 4.5355977491322485e-06, | |
| "loss": 0.9657, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 98.56328297192321, | |
| "learning_rate": 4.5332783529576145e-06, | |
| "loss": 0.851, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 15.486561420545545, | |
| "learning_rate": 4.530953775392749e-06, | |
| "loss": 1.0024, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 5.36867796146416, | |
| "learning_rate": 4.528624022361359e-06, | |
| "loss": 0.8203, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 23.16373737065163, | |
| "learning_rate": 4.5262890998003374e-06, | |
| "loss": 0.8405, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 24.557753611805698, | |
| "learning_rate": 4.523949013659754e-06, | |
| "loss": 0.8594, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 239.577821661523, | |
| "learning_rate": 4.521603769902835e-06, | |
| "loss": 0.8801, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 17.722681590802623, | |
| "learning_rate": 4.5192533745059494e-06, | |
| "loss": 0.7956, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 21.064904857136227, | |
| "learning_rate": 4.5168978334585955e-06, | |
| "loss": 0.849, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 29.204434577367632, | |
| "learning_rate": 4.514537152763384e-06, | |
| "loss": 0.8081, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 31.677494628738625, | |
| "learning_rate": 4.512171338436022e-06, | |
| "loss": 0.8123, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 28.259106443263146, | |
| "learning_rate": 4.509800396505298e-06, | |
| "loss": 0.7939, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 31.436330986692848, | |
| "learning_rate": 4.507424333013069e-06, | |
| "loss": 0.9276, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 62.15027720090623, | |
| "learning_rate": 4.505043154014243e-06, | |
| "loss": 0.8111, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 30.339486412023945, | |
| "learning_rate": 4.502656865576762e-06, | |
| "loss": 0.8102, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 29.14289188449159, | |
| "learning_rate": 4.500265473781591e-06, | |
| "loss": 0.8526, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 16.33958767689576, | |
| "learning_rate": 4.497868984722698e-06, | |
| "loss": 0.7931, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 6.946767762788769, | |
| "learning_rate": 4.495467404507039e-06, | |
| "loss": 0.7723, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 16.808686330416283, | |
| "learning_rate": 4.493060739254548e-06, | |
| "loss": 0.7941, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 11.193622425241314, | |
| "learning_rate": 4.4906489950981125e-06, | |
| "loss": 0.8121, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 7.172258662371558, | |
| "learning_rate": 4.4882321781835666e-06, | |
| "loss": 0.8399, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 7.172757454398357, | |
| "learning_rate": 4.485810294669668e-06, | |
| "loss": 0.7951, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 12.912002056626209, | |
| "learning_rate": 4.4833833507280884e-06, | |
| "loss": 0.8724, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 9.5299105957465, | |
| "learning_rate": 4.4809513525433925e-06, | |
| "loss": 0.8688, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 5.942401217715412, | |
| "learning_rate": 4.478514306313026e-06, | |
| "loss": 0.8349, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 14.70118213793697, | |
| "learning_rate": 4.476072218247297e-06, | |
| "loss": 0.8126, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 35.64062421380103, | |
| "learning_rate": 4.473625094569366e-06, | |
| "loss": 0.8502, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 9.515237433838145, | |
| "learning_rate": 4.471172941515219e-06, | |
| "loss": 0.8318, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 4.52525000568301, | |
| "learning_rate": 4.4687157653336645e-06, | |
| "loss": 0.836, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 14.93156121415628, | |
| "learning_rate": 4.466253572286308e-06, | |
| "loss": 0.8386, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 5.403603957950335, | |
| "learning_rate": 4.463786368647541e-06, | |
| "loss": 0.8026, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 5.101795351603165, | |
| "learning_rate": 4.461314160704521e-06, | |
| "loss": 0.8424, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 4.657146697628601, | |
| "learning_rate": 4.458836954757161e-06, | |
| "loss": 0.8853, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4.8266562887019315, | |
| "learning_rate": 4.456354757118109e-06, | |
| "loss": 0.8311, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4.734850079452724, | |
| "learning_rate": 4.453867574112733e-06, | |
| "loss": 0.8143, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 8.241211166354187, | |
| "learning_rate": 4.4513754120791065e-06, | |
| "loss": 0.8309, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4.984336730679723, | |
| "learning_rate": 4.448878277367988e-06, | |
| "loss": 0.834, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4.435146207156021, | |
| "learning_rate": 4.446376176342812e-06, | |
| "loss": 0.8344, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 5.354979176494093, | |
| "learning_rate": 4.443869115379667e-06, | |
| "loss": 0.8822, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4.740333651131116, | |
| "learning_rate": 4.441357100867278e-06, | |
| "loss": 0.7742, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 6.7966105580368605, | |
| "learning_rate": 4.438840139206998e-06, | |
| "loss": 0.8741, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 5.13633031473771, | |
| "learning_rate": 4.436318236812782e-06, | |
| "loss": 0.8654, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 5.791216706993845, | |
| "learning_rate": 4.433791400111179e-06, | |
| "loss": 0.8262, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 6.234562179308168, | |
| "learning_rate": 4.431259635541312e-06, | |
| "loss": 0.8392, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 5.150150040963309, | |
| "learning_rate": 4.428722949554858e-06, | |
| "loss": 0.8191, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 4.563262237056235, | |
| "learning_rate": 4.426181348616038e-06, | |
| "loss": 0.7961, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 4.409500438314574, | |
| "learning_rate": 4.423634839201601e-06, | |
| "loss": 0.8554, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 5.103854158767785, | |
| "learning_rate": 4.421083427800795e-06, | |
| "loss": 0.8073, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 4.956272764605345, | |
| "learning_rate": 4.41852712091537e-06, | |
| "loss": 0.822, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 47.36121108696217, | |
| "learning_rate": 4.415965925059544e-06, | |
| "loss": 0.8117, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 7.6663655272761515, | |
| "learning_rate": 4.413399846759998e-06, | |
| "loss": 0.8469, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 7.344767060016519, | |
| "learning_rate": 4.4108288925558505e-06, | |
| "loss": 0.8716, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 4.520414583397527, | |
| "learning_rate": 4.40825306899865e-06, | |
| "loss": 0.8154, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 5.337724239493111, | |
| "learning_rate": 4.405672382652349e-06, | |
| "loss": 0.8204, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 6.931095236367972, | |
| "learning_rate": 4.403086840093297e-06, | |
| "loss": 0.7721, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 4.283750585511646, | |
| "learning_rate": 4.400496447910212e-06, | |
| "loss": 0.8099, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 4.885379997216326, | |
| "learning_rate": 4.397901212704176e-06, | |
| "loss": 0.7897, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 6.083991051490237, | |
| "learning_rate": 4.3953011410886105e-06, | |
| "loss": 0.8497, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 4.86609034632746, | |
| "learning_rate": 4.392696239689261e-06, | |
| "loss": 0.8271, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 7.4379794775118615, | |
| "learning_rate": 4.390086515144179e-06, | |
| "loss": 0.8416, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 10.915187189394999, | |
| "learning_rate": 4.387471974103713e-06, | |
| "loss": 0.7873, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 5.839190871758675, | |
| "learning_rate": 4.384852623230478e-06, | |
| "loss": 0.7898, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 5.670517697152507, | |
| "learning_rate": 4.38222846919935e-06, | |
| "loss": 0.8431, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 6.4010524847241115, | |
| "learning_rate": 4.379599518697444e-06, | |
| "loss": 0.8489, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.505346933106831, | |
| "learning_rate": 4.3769657784240975e-06, | |
| "loss": 0.8056, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 4.519181163960888, | |
| "learning_rate": 4.3743272550908545e-06, | |
| "loss": 0.82, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.2720001800167475, | |
| "learning_rate": 4.3716839554214475e-06, | |
| "loss": 0.8261, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.48180004445423, | |
| "learning_rate": 4.369035886151778e-06, | |
| "loss": 0.7666, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 10.355727305381658, | |
| "learning_rate": 4.366383054029907e-06, | |
| "loss": 0.8203, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.185243645713917, | |
| "learning_rate": 4.363725465816028e-06, | |
| "loss": 0.8336, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 4.537447056649908, | |
| "learning_rate": 4.3610631282824556e-06, | |
| "loss": 0.8141, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 10.059335937073103, | |
| "learning_rate": 4.358396048213609e-06, | |
| "loss": 0.9323, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 7.558752810891884, | |
| "learning_rate": 4.355724232405989e-06, | |
| "loss": 0.8052, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 8.901083371299169, | |
| "learning_rate": 4.35304768766817e-06, | |
| "loss": 0.8681, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 5.749432223564655, | |
| "learning_rate": 4.350366420820771e-06, | |
| "loss": 0.7636, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 5.467983174988701, | |
| "learning_rate": 4.3476804386964486e-06, | |
| "loss": 0.798, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 14.340080127005345, | |
| "learning_rate": 4.3449897481398735e-06, | |
| "loss": 0.8325, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 10.900246873425738, | |
| "learning_rate": 4.342294356007715e-06, | |
| "loss": 0.8191, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 27.212678856649585, | |
| "learning_rate": 4.339594269168624e-06, | |
| "loss": 0.7983, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 12.291890124401506, | |
| "learning_rate": 4.336889494503215e-06, | |
| "loss": 0.8223, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 64.09126736889749, | |
| "learning_rate": 4.3341800389040465e-06, | |
| "loss": 0.7935, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 5.384829430193068, | |
| "learning_rate": 4.331465909275608e-06, | |
| "loss": 0.8063, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 5.980479175502184, | |
| "learning_rate": 4.3287471125342994e-06, | |
| "loss": 0.8264, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 14.891763292361196, | |
| "learning_rate": 4.326023655608412e-06, | |
| "loss": 0.8724, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 47.12998834101147, | |
| "learning_rate": 4.3232955454381126e-06, | |
| "loss": 0.9869, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 4.471610169915462, | |
| "learning_rate": 4.320562788975429e-06, | |
| "loss": 0.8504, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 5.687055238382911, | |
| "learning_rate": 4.317825393184226e-06, | |
| "loss": 0.8192, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 6.0149060191653625, | |
| "learning_rate": 4.315083365040193e-06, | |
| "loss": 0.8474, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 5.497573287322376, | |
| "learning_rate": 4.31233671153082e-06, | |
| "loss": 0.7821, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 5.877574667166528, | |
| "learning_rate": 4.30958543965539e-06, | |
| "loss": 0.7886, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 8.602706507929737, | |
| "learning_rate": 4.306829556424949e-06, | |
| "loss": 0.922, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 5.056672401512086, | |
| "learning_rate": 4.3040690688622965e-06, | |
| "loss": 0.824, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 19.382968104583764, | |
| "learning_rate": 4.3013039840019675e-06, | |
| "loss": 0.8193, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.880915789014604, | |
| "learning_rate": 4.2985343088902096e-06, | |
| "loss": 0.8291, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 12.940372419412574, | |
| "learning_rate": 4.295760050584966e-06, | |
| "loss": 0.8503, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 5.731004069006167, | |
| "learning_rate": 4.292981216155864e-06, | |
| "loss": 0.7975, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 14.301842614329003, | |
| "learning_rate": 4.2901978126841885e-06, | |
| "loss": 0.8179, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.304262507046851, | |
| "learning_rate": 4.287409847262868e-06, | |
| "loss": 0.8559, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.8057979345321655, | |
| "eval_runtime": 132.6219, | |
| "eval_samples_per_second": 9.855, | |
| "eval_steps_per_second": 0.279, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 12.900039338569156, | |
| "learning_rate": 4.284617326996458e-06, | |
| "loss": 0.8245, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 27.720291530372055, | |
| "learning_rate": 4.281820259001121e-06, | |
| "loss": 0.8123, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 9.875124909540267, | |
| "learning_rate": 4.2790186504046045e-06, | |
| "loss": 0.8077, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.959102865654325, | |
| "learning_rate": 4.276212508346232e-06, | |
| "loss": 0.8102, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 8.445468566132282, | |
| "learning_rate": 4.273401839976877e-06, | |
| "loss": 0.9495, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 7.168758336315809, | |
| "learning_rate": 4.2705866524589475e-06, | |
| "loss": 0.794, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 23.960316842270476, | |
| "learning_rate": 4.267766952966369e-06, | |
| "loss": 0.8234, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 18.37836287213558, | |
| "learning_rate": 4.264942748684563e-06, | |
| "loss": 0.8037, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 5.337874181371052, | |
| "learning_rate": 4.262114046810429e-06, | |
| "loss": 0.8341, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 23.756712822011433, | |
| "learning_rate": 4.2592808545523335e-06, | |
| "loss": 0.7835, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 6.5799458933423365, | |
| "learning_rate": 4.256443179130081e-06, | |
| "loss": 0.823, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 5.340830389502352, | |
| "learning_rate": 4.2536010277748995e-06, | |
| "loss": 0.7958, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 4.274138909771709, | |
| "learning_rate": 4.250754407729428e-06, | |
| "loss": 0.806, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 5.0306349490940985, | |
| "learning_rate": 4.2479033262476885e-06, | |
| "loss": 0.8728, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.644338434000799, | |
| "learning_rate": 4.2450477905950745e-06, | |
| "loss": 0.8304, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.447934903535319, | |
| "learning_rate": 4.242187808048329e-06, | |
| "loss": 0.8294, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.468749772112203, | |
| "learning_rate": 4.239323385895527e-06, | |
| "loss": 0.8134, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.595150327241665, | |
| "learning_rate": 4.236454531436058e-06, | |
| "loss": 0.9786, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 3.7601508175241807, | |
| "learning_rate": 4.233581251980604e-06, | |
| "loss": 0.8269, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.86366757480535, | |
| "learning_rate": 4.230703554851127e-06, | |
| "loss": 0.8258, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.352222075706524, | |
| "learning_rate": 4.227821447380842e-06, | |
| "loss": 0.9652, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 8.73857112989018, | |
| "learning_rate": 4.224934936914206e-06, | |
| "loss": 0.8383, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.044749221503628, | |
| "learning_rate": 4.222044030806894e-06, | |
| "loss": 0.8207, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.613463894710261, | |
| "learning_rate": 4.219148736425785e-06, | |
| "loss": 0.8116, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 5.8777142171863686, | |
| "learning_rate": 4.216249061148939e-06, | |
| "loss": 0.7927, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 4.297186700651947, | |
| "learning_rate": 4.2133450123655805e-06, | |
| "loss": 0.8784, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 4.9580457958209765, | |
| "learning_rate": 4.210436597476077e-06, | |
| "loss": 0.8232, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 4.156257841735334, | |
| "learning_rate": 4.207523823891924e-06, | |
| "loss": 0.8446, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 4.364319132512279, | |
| "learning_rate": 4.204606699035724e-06, | |
| "loss": 0.8001, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 4.626437632049346, | |
| "learning_rate": 4.201685230341168e-06, | |
| "loss": 0.8488, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 3.6979503756764536, | |
| "learning_rate": 4.198759425253015e-06, | |
| "loss": 0.8231, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 3.9375406125469117, | |
| "learning_rate": 4.195829291227076e-06, | |
| "loss": 0.8262, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 3.757592718600369, | |
| "learning_rate": 4.192894835730193e-06, | |
| "loss": 0.8325, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 5.166098354237156, | |
| "learning_rate": 4.1899560662402204e-06, | |
| "loss": 0.8207, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.5626403449402146, | |
| "learning_rate": 4.187012990246005e-06, | |
| "loss": 0.848, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 4.405523794955747, | |
| "learning_rate": 4.18406561524737e-06, | |
| "loss": 0.8153, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 4.461492151159491, | |
| "learning_rate": 4.18111394875509e-06, | |
| "loss": 0.7961, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.9579080031218945, | |
| "learning_rate": 4.178157998290879e-06, | |
| "loss": 0.7961, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.4157753139788114, | |
| "learning_rate": 4.175197771387368e-06, | |
| "loss": 0.77, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.6556329052007372, | |
| "learning_rate": 4.172233275588082e-06, | |
| "loss": 0.757, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.56739734349662, | |
| "learning_rate": 4.169264518447428e-06, | |
| "loss": 0.816, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.929236437983629, | |
| "learning_rate": 4.16629150753067e-06, | |
| "loss": 0.7774, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 4.162206048648253, | |
| "learning_rate": 4.163314250413913e-06, | |
| "loss": 0.8242, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.5348802562769683, | |
| "learning_rate": 4.160332754684085e-06, | |
| "loss": 0.8292, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 4.026476793505602, | |
| "learning_rate": 4.157347027938907e-06, | |
| "loss": 0.8438, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.6419095683671205, | |
| "learning_rate": 4.154357077786892e-06, | |
| "loss": 0.8358, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 4.410825648674678, | |
| "learning_rate": 4.1513629118473095e-06, | |
| "loss": 0.7966, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.7192835271993423, | |
| "learning_rate": 4.1483645377501726e-06, | |
| "loss": 0.7922, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 4.2093727504165575, | |
| "learning_rate": 4.145361963136219e-06, | |
| "loss": 0.8179, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.81543786003894, | |
| "learning_rate": 4.1423551956568916e-06, | |
| "loss": 0.8072, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 4.0092000330441335, | |
| "learning_rate": 4.139344242974317e-06, | |
| "loss": 0.811, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.927721495131119, | |
| "learning_rate": 4.136329112761285e-06, | |
| "loss": 0.8587, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.6480183964959076, | |
| "learning_rate": 4.133309812701233e-06, | |
| "loss": 0.7872, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.603038562895384, | |
| "learning_rate": 4.130286350488224e-06, | |
| "loss": 0.8211, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 4.502055880585956, | |
| "learning_rate": 4.127258733826929e-06, | |
| "loss": 0.7947, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 4.0596180713400845, | |
| "learning_rate": 4.124226970432602e-06, | |
| "loss": 0.8371, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 5.32641556527492, | |
| "learning_rate": 4.121191068031067e-06, | |
| "loss": 0.9996, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.657535832118382, | |
| "learning_rate": 4.118151034358696e-06, | |
| "loss": 0.7846, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 4.542970770312805, | |
| "learning_rate": 4.1151068771623864e-06, | |
| "loss": 0.9695, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.4165335352607418, | |
| "learning_rate": 4.112058604199544e-06, | |
| "loss": 0.7936, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 4.38297960645218, | |
| "learning_rate": 4.109006223238064e-06, | |
| "loss": 0.7862, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 4.597014556198219, | |
| "learning_rate": 4.105949742056309e-06, | |
| "loss": 0.9645, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.431709372470043, | |
| "learning_rate": 4.102889168443091e-06, | |
| "loss": 0.8077, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.787025575201097, | |
| "learning_rate": 4.0998245101976495e-06, | |
| "loss": 0.811, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 4.1176444935534535, | |
| "learning_rate": 4.096755775129634e-06, | |
| "loss": 0.824, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 3.874994826180069, | |
| "learning_rate": 4.093682971059081e-06, | |
| "loss": 0.8207, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 5.615505880427777, | |
| "learning_rate": 4.0906061058164e-06, | |
| "loss": 0.9337, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 4.159261391680892, | |
| "learning_rate": 4.087525187242345e-06, | |
| "loss": 0.7938, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 3.982328427595764, | |
| "learning_rate": 4.084440223188002e-06, | |
| "loss": 0.8353, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 3.501905560670443, | |
| "learning_rate": 4.081351221514765e-06, | |
| "loss": 0.8038, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 4.4798436725132795, | |
| "learning_rate": 4.078258190094318e-06, | |
| "loss": 0.8106, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 4.630604033125049, | |
| "learning_rate": 4.0751611368086115e-06, | |
| "loss": 0.8368, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 4.247722886052707, | |
| "learning_rate": 4.072060069549848e-06, | |
| "loss": 0.7804, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 3.976719242252589, | |
| "learning_rate": 4.068954996220457e-06, | |
| "loss": 0.8517, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 3.7378777926374647, | |
| "learning_rate": 4.065845924733077e-06, | |
| "loss": 0.7879, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 3.377152336223136, | |
| "learning_rate": 4.062732863010534e-06, | |
| "loss": 0.8151, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 3.958932705104774, | |
| "learning_rate": 4.059615818985826e-06, | |
| "loss": 0.8261, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 4.056870191904715, | |
| "learning_rate": 4.056494800602093e-06, | |
| "loss": 0.8036, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 4.276330614300562, | |
| "learning_rate": 4.053369815812608e-06, | |
| "loss": 0.7653, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 3.7171393121481557, | |
| "learning_rate": 4.050240872580749e-06, | |
| "loss": 0.811, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 3.8515114977006384, | |
| "learning_rate": 4.047107978879985e-06, | |
| "loss": 0.837, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 3.990125985702027, | |
| "learning_rate": 4.043971142693845e-06, | |
| "loss": 0.7478, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 5.038946309180022, | |
| "learning_rate": 4.040830372015909e-06, | |
| "loss": 0.8186, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 5.347765467237379, | |
| "learning_rate": 4.0376856748497865e-06, | |
| "loss": 0.8224, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 3.8056078948982957, | |
| "learning_rate": 4.034537059209085e-06, | |
| "loss": 0.7922, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.699749057539843, | |
| "learning_rate": 4.031384533117404e-06, | |
| "loss": 0.8004, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 3.985653883322575, | |
| "learning_rate": 4.0282281046083045e-06, | |
| "loss": 0.8352, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 3.6031461215401115, | |
| "learning_rate": 4.025067781725294e-06, | |
| "loss": 0.8557, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.374834805628173, | |
| "learning_rate": 4.021903572521802e-06, | |
| "loss": 0.8473, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.706489580366399, | |
| "learning_rate": 4.0187354850611634e-06, | |
| "loss": 0.8707, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 3.6196942519877844, | |
| "learning_rate": 4.015563527416596e-06, | |
| "loss": 0.7828, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.021419041448392, | |
| "learning_rate": 4.012387707671177e-06, | |
| "loss": 0.8212, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 7.5051269502991484, | |
| "learning_rate": 4.00920803391783e-06, | |
| "loss": 0.8125, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.816524818864674, | |
| "learning_rate": 4.006024514259295e-06, | |
| "loss": 0.8458, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.999047814247483, | |
| "learning_rate": 4.002837156808116e-06, | |
| "loss": 0.9794, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.053935034209712, | |
| "learning_rate": 3.999645969686616e-06, | |
| "loss": 0.8172, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.4988389719653408, | |
| "learning_rate": 3.996450961026876e-06, | |
| "loss": 0.8358, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.7828610862275447, | |
| "learning_rate": 3.993252138970716e-06, | |
| "loss": 0.8472, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.30179599391661, | |
| "learning_rate": 3.990049511669675e-06, | |
| "loss": 0.9382, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.7610184281848373, | |
| "learning_rate": 3.986843087284986e-06, | |
| "loss": 0.8122, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.7656350305584976, | |
| "learning_rate": 3.983632873987562e-06, | |
| "loss": 0.7727, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.760295447041278, | |
| "learning_rate": 3.980418879957967e-06, | |
| "loss": 0.7896, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.1049226948000386, | |
| "learning_rate": 3.977201113386402e-06, | |
| "loss": 0.7927, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 11.952856680061187, | |
| "learning_rate": 3.973979582472681e-06, | |
| "loss": 0.82, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.6775385257137136, | |
| "learning_rate": 3.970754295426211e-06, | |
| "loss": 0.7982, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 4.126563331457935, | |
| "learning_rate": 3.96752526046597e-06, | |
| "loss": 0.7973, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.9969733729191637, | |
| "learning_rate": 3.964292485820487e-06, | |
| "loss": 0.7848, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.843494951843216, | |
| "learning_rate": 3.961055979727822e-06, | |
| "loss": 0.7809, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.722737193377863, | |
| "learning_rate": 3.957815750435542e-06, | |
| "loss": 0.8272, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.505878864813977, | |
| "learning_rate": 3.954571806200702e-06, | |
| "loss": 0.7547, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.816576897584943, | |
| "learning_rate": 3.951324155289825e-06, | |
| "loss": 0.8178, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.6494681155339252, | |
| "learning_rate": 3.948072805978879e-06, | |
| "loss": 0.8163, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.31954721595174, | |
| "learning_rate": 3.9448177665532575e-06, | |
| "loss": 0.8036, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.99601790046458, | |
| "learning_rate": 3.941559045307756e-06, | |
| "loss": 0.8052, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.256204345344998, | |
| "learning_rate": 3.938296650546552e-06, | |
| "loss": 0.8062, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.6522897880772778, | |
| "learning_rate": 3.935030590583186e-06, | |
| "loss": 0.8461, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.5724478570414706, | |
| "learning_rate": 3.931760873740539e-06, | |
| "loss": 0.8173, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 4.369010546999307, | |
| "learning_rate": 3.928487508350808e-06, | |
| "loss": 0.8355, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.867196950400295, | |
| "learning_rate": 3.9252105027554885e-06, | |
| "loss": 0.7982, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.757894555280206, | |
| "learning_rate": 3.9219298653053544e-06, | |
| "loss": 0.744, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 4.483773308176931, | |
| "learning_rate": 3.918645604360434e-06, | |
| "loss": 0.8264, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.5936542631360995, | |
| "learning_rate": 3.915357728289985e-06, | |
| "loss": 0.773, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.7125514529042447, | |
| "learning_rate": 3.912066245472484e-06, | |
| "loss": 0.789, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.994678394733242, | |
| "learning_rate": 3.908771164295595e-06, | |
| "loss": 0.8054, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.440863065927586, | |
| "learning_rate": 3.9054724931561516e-06, | |
| "loss": 0.7667, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.6354877018750003, | |
| "learning_rate": 3.902170240460137e-06, | |
| "loss": 0.8434, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.6697310301169783, | |
| "learning_rate": 3.898864414622661e-06, | |
| "loss": 0.8277, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.4657019048153543, | |
| "learning_rate": 3.895555024067937e-06, | |
| "loss": 0.8058, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.569249201300816, | |
| "learning_rate": 3.8922420772292645e-06, | |
| "loss": 0.8646, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.6960654260718675, | |
| "learning_rate": 3.888925582549006e-06, | |
| "loss": 0.8439, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.206703534075406, | |
| "learning_rate": 3.8856055484785625e-06, | |
| "loss": 0.958, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.8024335105854963, | |
| "learning_rate": 3.8822819834783556e-06, | |
| "loss": 0.7933, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 3.311892134941326, | |
| "learning_rate": 3.878954896017804e-06, | |
| "loss": 0.7679, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 3.6892320743049383, | |
| "learning_rate": 3.875624294575306e-06, | |
| "loss": 0.8663, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.747838560381203, | |
| "learning_rate": 3.872290187638208e-06, | |
| "loss": 0.8028, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.130227898119578, | |
| "learning_rate": 3.868952583702798e-06, | |
| "loss": 0.7775, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 3.8622011098862004, | |
| "learning_rate": 3.865611491274267e-06, | |
| "loss": 0.8362, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.076018210248722, | |
| "learning_rate": 3.862266918866702e-06, | |
| "loss": 0.9772, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.361556277412654, | |
| "learning_rate": 3.858918875003053e-06, | |
| "loss": 0.7957, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.386167012643809, | |
| "learning_rate": 3.855567368215122e-06, | |
| "loss": 0.9516, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 3.9439193106997736, | |
| "learning_rate": 3.852212407043528e-06, | |
| "loss": 0.8231, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.465518470978496, | |
| "learning_rate": 3.848854000037702e-06, | |
| "loss": 0.8476, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.675424607415688, | |
| "learning_rate": 3.845492155755848e-06, | |
| "loss": 0.7827, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 6.031861648386004, | |
| "learning_rate": 3.8421268827649325e-06, | |
| "loss": 0.8044, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 5.884738538466192, | |
| "learning_rate": 3.83875818964066e-06, | |
| "loss": 0.9754, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 6.050577830000915, | |
| "learning_rate": 3.835386084967451e-06, | |
| "loss": 0.7774, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.992604354329953, | |
| "learning_rate": 3.832010577338414e-06, | |
| "loss": 0.8663, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.137703965401395, | |
| "learning_rate": 3.828631675355338e-06, | |
| "loss": 0.8247, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 3.9250818325319647, | |
| "learning_rate": 3.825249387628654e-06, | |
| "loss": 0.9427, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 5.876925468778531, | |
| "learning_rate": 3.821863722777427e-06, | |
| "loss": 0.7972, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.029506273549229, | |
| "learning_rate": 3.818474689429324e-06, | |
| "loss": 0.8085, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.095701024225992, | |
| "learning_rate": 3.815082296220596e-06, | |
| "loss": 0.8278, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.0688198508439255, | |
| "learning_rate": 3.8116865517960584e-06, | |
| "loss": 0.7869, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 4.679593702074773, | |
| "learning_rate": 3.808287464809063e-06, | |
| "loss": 0.8103, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 3.980055385102619, | |
| "learning_rate": 3.8048850439214844e-06, | |
| "loss": 0.9489, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 3.89441546745002, | |
| "learning_rate": 3.8014792978036874e-06, | |
| "loss": 0.8053, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 3.495335889379114, | |
| "learning_rate": 3.7980702351345146e-06, | |
| "loss": 0.796, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 3.69936931190018, | |
| "learning_rate": 3.7946578646012578e-06, | |
| "loss": 0.8388, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 3.711786581597773, | |
| "learning_rate": 3.7912421948996394e-06, | |
| "loss": 0.772, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 4.3115598706174225, | |
| "learning_rate": 3.787823234733788e-06, | |
| "loss": 0.8086, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 4.206524036573733, | |
| "learning_rate": 3.7844009928162195e-06, | |
| "loss": 0.8422, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 4.581806945598464, | |
| "learning_rate": 3.78097547786781e-06, | |
| "loss": 0.8299, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 3.649063963125477, | |
| "learning_rate": 3.7775466986177763e-06, | |
| "loss": 0.8233, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 3.9874069159693497, | |
| "learning_rate": 3.7741146638036567e-06, | |
| "loss": 0.836, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 5.101352434056481, | |
| "learning_rate": 3.770679382171283e-06, | |
| "loss": 0.8053, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.057518103549027, | |
| "learning_rate": 3.7672408624747598e-06, | |
| "loss": 0.8119, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.134292768219472, | |
| "learning_rate": 3.7637991134764475e-06, | |
| "loss": 0.8441, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.037293558864936, | |
| "learning_rate": 3.7603541439469314e-06, | |
| "loss": 0.8475, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.067939888641445, | |
| "learning_rate": 3.756905962665005e-06, | |
| "loss": 0.8176, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.741871018197297, | |
| "learning_rate": 3.7534545784176486e-06, | |
| "loss": 0.8252, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 3.7687469109452536, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.8222, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.0489323193480855, | |
| "learning_rate": 3.7465422362153416e-06, | |
| "loss": 0.8325, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.049191319739393, | |
| "learning_rate": 3.7430812958750695e-06, | |
| "loss": 0.8391, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 3.601920010559379, | |
| "learning_rate": 3.7396171877986766e-06, | |
| "loss": 0.7752, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.353923235687504, | |
| "learning_rate": 3.736149920813726e-06, | |
| "loss": 0.7507, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.022804558800863, | |
| "learning_rate": 3.7326795037558335e-06, | |
| "loss": 0.8216, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.056771582273963, | |
| "learning_rate": 3.72920594546864e-06, | |
| "loss": 0.7952, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 3.418690092445234, | |
| "learning_rate": 3.7257292548037917e-06, | |
| "loss": 0.7834, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.027419729266051, | |
| "learning_rate": 3.7222494406209174e-06, | |
| "loss": 0.8238, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.557745951925816, | |
| "learning_rate": 3.718766511787606e-06, | |
| "loss": 0.8367, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.16754573699968, | |
| "learning_rate": 3.715280477179382e-06, | |
| "loss": 0.8561, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.213296187055132, | |
| "learning_rate": 3.7117913456796855e-06, | |
| "loss": 0.8006, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.895541456046058, | |
| "learning_rate": 3.7082991261798473e-06, | |
| "loss": 0.8019, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 4.225646010065031, | |
| "learning_rate": 3.7048038275790695e-06, | |
| "loss": 0.7799, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.793719931690171, | |
| "learning_rate": 3.701305458784397e-06, | |
| "loss": 0.7877, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.6918538059697505, | |
| "learning_rate": 3.697804028710703e-06, | |
| "loss": 0.8472, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.7943218099807967, | |
| "learning_rate": 3.6942995462806574e-06, | |
| "loss": 0.84, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.603972411456462, | |
| "learning_rate": 3.6907920204247122e-06, | |
| "loss": 0.8205, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.755592793493984, | |
| "learning_rate": 3.6872814600810716e-06, | |
| "loss": 0.8354, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 4.658007950410123, | |
| "learning_rate": 3.6837678741956747e-06, | |
| "loss": 0.8767, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.4964527787354416, | |
| "learning_rate": 3.6802512717221694e-06, | |
| "loss": 0.7814, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.434255209389191, | |
| "learning_rate": 3.6767316616218927e-06, | |
| "loss": 0.8071, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.272321481007903, | |
| "learning_rate": 3.6732090528638432e-06, | |
| "loss": 0.7787, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.221528130430344, | |
| "learning_rate": 3.669683454424663e-06, | |
| "loss": 0.7729, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.9597068009542915, | |
| "learning_rate": 3.6661548752886113e-06, | |
| "loss": 0.7689, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.5568349195672253, | |
| "learning_rate": 3.6626233244475445e-06, | |
| "loss": 0.9255, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.719503943417636, | |
| "learning_rate": 3.6590888109008904e-06, | |
| "loss": 0.8281, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.6514574646812568, | |
| "learning_rate": 3.6555513436556285e-06, | |
| "loss": 0.8008, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.548664005213854, | |
| "learning_rate": 3.6520109317262624e-06, | |
| "loss": 0.7797, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.704773651343669, | |
| "learning_rate": 3.6484675841348024e-06, | |
| "loss": 0.813, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.6582715454571657, | |
| "learning_rate": 3.6449213099107376e-06, | |
| "loss": 0.8007, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.094141645704761, | |
| "learning_rate": 3.6413721180910165e-06, | |
| "loss": 0.837, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 4.618837723050664, | |
| "learning_rate": 3.6378200177200223e-06, | |
| "loss": 0.8316, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.38898352006268, | |
| "learning_rate": 3.6342650178495488e-06, | |
| "loss": 0.8103, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.7976368474709896, | |
| "learning_rate": 3.6307071275387807e-06, | |
| "loss": 0.8059, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.5189322963766205, | |
| "learning_rate": 3.6271463558542646e-06, | |
| "loss": 0.7959, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.3331229796313027, | |
| "learning_rate": 3.623582711869895e-06, | |
| "loss": 0.7882, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 5.530082944684263, | |
| "learning_rate": 3.6200162046668826e-06, | |
| "loss": 0.79, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.3051251617528568, | |
| "learning_rate": 3.616446843333733e-06, | |
| "loss": 0.7773, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.3812653747476427, | |
| "learning_rate": 3.6128746369662283e-06, | |
| "loss": 0.7819, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.5371020189256805, | |
| "learning_rate": 3.6092995946673996e-06, | |
| "loss": 0.8518, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 3.834778332537793, | |
| "learning_rate": 3.6057217255475034e-06, | |
| "loss": 0.8418, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 4.415666375807929, | |
| "learning_rate": 3.602141038724001e-06, | |
| "loss": 0.8146, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.641170161321054, | |
| "learning_rate": 3.5985575433215345e-06, | |
| "loss": 0.8347, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.7635739548814207, | |
| "learning_rate": 3.5949712484719013e-06, | |
| "loss": 0.8207, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 4.1180161320176225, | |
| "learning_rate": 3.591382163314034e-06, | |
| "loss": 0.9224, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 4.130810975608334, | |
| "learning_rate": 3.587790296993976e-06, | |
| "loss": 0.843, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 4.691833335291648, | |
| "learning_rate": 3.5841956586648553e-06, | |
| "loss": 0.9047, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 4.657933022135858, | |
| "learning_rate": 3.5805982574868673e-06, | |
| "loss": 1.1555, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.2260069416532207, | |
| "learning_rate": 3.5769981026272477e-06, | |
| "loss": 0.7869, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.4291174313472643, | |
| "learning_rate": 3.5733952032602454e-06, | |
| "loss": 0.7723, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.919942403046828, | |
| "learning_rate": 3.5697895685671076e-06, | |
| "loss": 0.8133, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 3.9420409702367167, | |
| "learning_rate": 3.5661812077360496e-06, | |
| "loss": 0.8135, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.062763811575889, | |
| "learning_rate": 3.562570129962234e-06, | |
| "loss": 0.8098, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 3.1692597156645306, | |
| "learning_rate": 3.5589563444477477e-06, | |
| "loss": 0.7864, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.529414675977005, | |
| "learning_rate": 3.5553398604015777e-06, | |
| "loss": 0.8194, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 3.6410323820649144, | |
| "learning_rate": 3.5517206870395852e-06, | |
| "loss": 0.8061, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 3.5825284359470673, | |
| "learning_rate": 3.548098833584489e-06, | |
| "loss": 0.8055, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 3.6994250667973314, | |
| "learning_rate": 3.544474309265834e-06, | |
| "loss": 0.8014, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.609457196655879, | |
| "learning_rate": 3.5408471233199713e-06, | |
| "loss": 0.8953, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.357338921046579, | |
| "learning_rate": 3.5372172849900377e-06, | |
| "loss": 0.815, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.534756505244773, | |
| "learning_rate": 3.5335848035259257e-06, | |
| "loss": 0.8281, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 4.46200388603068, | |
| "learning_rate": 3.5299496881842654e-06, | |
| "loss": 0.8979, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 3.8353237250725893, | |
| "learning_rate": 3.526311948228397e-06, | |
| "loss": 0.8392, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 3.513564550196794, | |
| "learning_rate": 3.5226715929283507e-06, | |
| "loss": 0.8276, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 4.840869909676071, | |
| "learning_rate": 3.5190286315608196e-06, | |
| "loss": 0.7808, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 3.6376115911500326, | |
| "learning_rate": 3.5153830734091404e-06, | |
| "loss": 0.7985, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 8.73646238891082, | |
| "learning_rate": 3.511734927763265e-06, | |
| "loss": 0.8393, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 4.215260596616087, | |
| "learning_rate": 3.508084203919739e-06, | |
| "loss": 0.8183, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 4.0886422935222395, | |
| "learning_rate": 3.50443091118168e-06, | |
| "loss": 0.7957, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 3.7885638513567383, | |
| "learning_rate": 3.5007750588587495e-06, | |
| "loss": 0.8112, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 3.9547192279690653, | |
| "learning_rate": 3.497116656267133e-06, | |
| "loss": 0.7731, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.86013814393367, | |
| "learning_rate": 3.493455712729514e-06, | |
| "loss": 0.8547, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.9264148109718997, | |
| "learning_rate": 3.4897922375750517e-06, | |
| "loss": 0.7858, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.5652346553585925, | |
| "learning_rate": 3.4861262401393563e-06, | |
| "loss": 0.785, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.1139632776522586, | |
| "learning_rate": 3.4824577297644663e-06, | |
| "loss": 0.9653, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 5.588885807787667, | |
| "learning_rate": 3.478786715798823e-06, | |
| "loss": 0.8312, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 6.383737000046943, | |
| "learning_rate": 3.4751132075972473e-06, | |
| "loss": 0.7787, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.038979136777571, | |
| "learning_rate": 3.471437214520917e-06, | |
| "loss": 0.8058, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.227693679919142, | |
| "learning_rate": 3.4677587459373417e-06, | |
| "loss": 0.9012, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.985607862312997, | |
| "learning_rate": 3.46407781122034e-06, | |
| "loss": 0.819, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.922668922254852, | |
| "learning_rate": 3.4603944197500126e-06, | |
| "loss": 0.7354, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.343492283466782, | |
| "learning_rate": 3.4567085809127247e-06, | |
| "loss": 0.786, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.484528470192613, | |
| "learning_rate": 3.4530203041010745e-06, | |
| "loss": 0.7849, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.367273259116789, | |
| "learning_rate": 3.449329598713874e-06, | |
| "loss": 0.7921, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 3.927650734068294, | |
| "learning_rate": 3.4456364741561256e-06, | |
| "loss": 0.7746, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.586041331240587, | |
| "learning_rate": 3.4419409398389937e-06, | |
| "loss": 0.8183, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 3.7164145526774224, | |
| "learning_rate": 3.4382430051797844e-06, | |
| "loss": 0.8202, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.394679226693479, | |
| "learning_rate": 3.434542679601922e-06, | |
| "loss": 0.7859, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.096618466435277, | |
| "learning_rate": 3.430839972534923e-06, | |
| "loss": 0.7583, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.013175572729404, | |
| "learning_rate": 3.4271348934143707e-06, | |
| "loss": 0.8071, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.292337945500214, | |
| "learning_rate": 3.423427451681895e-06, | |
| "loss": 0.8085, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 3.9544343369431463, | |
| "learning_rate": 3.4197176567851463e-06, | |
| "loss": 0.8177, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.433418332095438, | |
| "learning_rate": 3.4160055181777714e-06, | |
| "loss": 0.7884, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.956803986907128, | |
| "learning_rate": 3.4122910453193885e-06, | |
| "loss": 0.7959, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 5.450063143284857, | |
| "learning_rate": 3.4085742476755657e-06, | |
| "loss": 0.9269, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 5.196193261350649, | |
| "learning_rate": 3.404855134717795e-06, | |
| "loss": 1.0828, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 8.093995434969097, | |
| "learning_rate": 3.4011337159234674e-06, | |
| "loss": 0.7985, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.575208484444267, | |
| "learning_rate": 3.3974100007758514e-06, | |
| "loss": 0.7914, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.88561287010472, | |
| "learning_rate": 3.3936839987640665e-06, | |
| "loss": 0.8413, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.10258246501254, | |
| "learning_rate": 3.3899557193830585e-06, | |
| "loss": 0.7827, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.134612322280866, | |
| "learning_rate": 3.3862251721335793e-06, | |
| "loss": 0.7846, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.127521869620768, | |
| "learning_rate": 3.382492366522158e-06, | |
| "loss": 0.8187, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.694265995175311, | |
| "learning_rate": 3.3787573120610794e-06, | |
| "loss": 0.9095, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.832053974686459, | |
| "learning_rate": 3.375020018268359e-06, | |
| "loss": 0.8422, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 3.615649772558988, | |
| "learning_rate": 3.371280494667719e-06, | |
| "loss": 0.7797, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.377263570264409, | |
| "learning_rate": 3.367538750788563e-06, | |
| "loss": 0.9017, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 5.6300240178979815, | |
| "learning_rate": 3.3637947961659533e-06, | |
| "loss": 0.8064, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.977615821468784, | |
| "learning_rate": 3.360048640340585e-06, | |
| "loss": 0.7614, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 3.9739872782044285, | |
| "learning_rate": 3.356300292858763e-06, | |
| "loss": 0.8155, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.365429767411158, | |
| "learning_rate": 3.3525497632723786e-06, | |
| "loss": 0.7931, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.03510482219512, | |
| "learning_rate": 3.348797061138881e-06, | |
| "loss": 0.7533, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.423899720091732, | |
| "learning_rate": 3.345042196021257e-06, | |
| "loss": 0.849, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.3577483066057745, | |
| "learning_rate": 3.3412851774880066e-06, | |
| "loss": 0.914, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.8429976699063095, | |
| "learning_rate": 3.3375260151131156e-06, | |
| "loss": 0.8208, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 3.846433730762683, | |
| "learning_rate": 3.333764718476032e-06, | |
| "loss": 0.7667, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 7.424150572222642, | |
| "learning_rate": 3.3300012971616467e-06, | |
| "loss": 0.7906, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 3.78110316938379, | |
| "learning_rate": 3.3262357607602596e-06, | |
| "loss": 0.8568, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.484285048821905, | |
| "learning_rate": 3.3224681188675643e-06, | |
| "loss": 0.7874, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.710007453384313, | |
| "learning_rate": 3.318698381084619e-06, | |
| "loss": 0.812, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.303370184571771, | |
| "learning_rate": 3.3149265570178215e-06, | |
| "loss": 0.9035, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.26213714508879, | |
| "learning_rate": 3.3111526562788864e-06, | |
| "loss": 0.8047, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.9733132790695183, | |
| "learning_rate": 3.3073766884848235e-06, | |
| "loss": 0.7961, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 5.156615645995947, | |
| "learning_rate": 3.303598663257904e-06, | |
| "loss": 0.824, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.736017510667661, | |
| "learning_rate": 3.2998185902256475e-06, | |
| "loss": 0.8016, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 5.180043733793966, | |
| "learning_rate": 3.2960364790207897e-06, | |
| "loss": 0.7707, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.905861811752156, | |
| "learning_rate": 3.2922523392812605e-06, | |
| "loss": 0.7466, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.923996642696014, | |
| "learning_rate": 3.2884661806501576e-06, | |
| "loss": 0.8046, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.9631467827534426, | |
| "learning_rate": 3.284678012775727e-06, | |
| "loss": 0.9511, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.965325234874838, | |
| "learning_rate": 3.2808878453113317e-06, | |
| "loss": 0.8636, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.642633499379332, | |
| "learning_rate": 3.2770956879154305e-06, | |
| "loss": 0.7702, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.492042123923082, | |
| "learning_rate": 3.273301550251555e-06, | |
| "loss": 0.8537, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.8876988328768602, | |
| "learning_rate": 3.269505441988281e-06, | |
| "loss": 0.8605, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.766606063581812, | |
| "learning_rate": 3.2657073727992078e-06, | |
| "loss": 0.7985, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.267264555929007, | |
| "learning_rate": 3.2619073523629304e-06, | |
| "loss": 0.7971, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.919439020397232, | |
| "learning_rate": 3.258105390363016e-06, | |
| "loss": 0.8164, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.24583647910275, | |
| "learning_rate": 3.2543014964879814e-06, | |
| "loss": 0.8043, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.4364112443855985, | |
| "learning_rate": 3.250495680431264e-06, | |
| "loss": 0.769, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.40163318603266, | |
| "learning_rate": 3.246687951891201e-06, | |
| "loss": 0.7878, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.273441543771241, | |
| "learning_rate": 3.2428783205710023e-06, | |
| "loss": 0.8331, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.7031687463067633, | |
| "learning_rate": 3.2390667961787276e-06, | |
| "loss": 0.7873, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.5284627603373284, | |
| "learning_rate": 3.2352533884272596e-06, | |
| "loss": 0.7826, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.901059309743076, | |
| "learning_rate": 3.2314381070342815e-06, | |
| "loss": 0.9248, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 13.116776373765031, | |
| "learning_rate": 3.2276209617222497e-06, | |
| "loss": 0.8109, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.5825950028846907, | |
| "learning_rate": 3.223801962218372e-06, | |
| "loss": 0.7666, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.500505244324813, | |
| "learning_rate": 3.21998111825458e-06, | |
| "loss": 0.7398, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.496092976488147, | |
| "learning_rate": 3.216158439567506e-06, | |
| "loss": 0.7631, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 4.556135014522858, | |
| "learning_rate": 3.2123339358984573e-06, | |
| "loss": 0.8207, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.7277520517697136, | |
| "learning_rate": 3.208507616993393e-06, | |
| "loss": 0.8516, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 4.423036921270627, | |
| "learning_rate": 3.2046794926028966e-06, | |
| "loss": 0.791, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.235611807092007, | |
| "learning_rate": 3.200849572482153e-06, | |
| "loss": 0.7558, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 3.824187344341294, | |
| "learning_rate": 3.1970178663909233e-06, | |
| "loss": 0.7658, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.275625768452698, | |
| "learning_rate": 3.19318438409352e-06, | |
| "loss": 0.8375, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.566672221414914, | |
| "learning_rate": 3.189349135358781e-06, | |
| "loss": 0.762, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.020446057440514, | |
| "learning_rate": 3.1855121299600454e-06, | |
| "loss": 0.8817, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.2881734956366095, | |
| "learning_rate": 3.1816733776751313e-06, | |
| "loss": 0.8106, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.673485294928623, | |
| "learning_rate": 3.1778328882863058e-06, | |
| "loss": 0.9695, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 7.339180964057777, | |
| "learning_rate": 3.173990671580263e-06, | |
| "loss": 0.7922, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 3.7378322020107713, | |
| "learning_rate": 3.170146737348099e-06, | |
| "loss": 0.8213, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.440411381382036, | |
| "learning_rate": 3.166301095385288e-06, | |
| "loss": 0.8058, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 3.645734306429232, | |
| "learning_rate": 3.162453755491655e-06, | |
| "loss": 0.7686, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.646322329035178, | |
| "learning_rate": 3.1586047274713493e-06, | |
| "loss": 0.8386, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.205705304738026, | |
| "learning_rate": 3.154754021132827e-06, | |
| "loss": 0.857, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 49.63215769717914, | |
| "learning_rate": 3.1509016462888175e-06, | |
| "loss": 0.8218, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.780984494435262, | |
| "learning_rate": 3.147047612756302e-06, | |
| "loss": 0.8394, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.58164822302537, | |
| "learning_rate": 3.143191930356491e-06, | |
| "loss": 0.837, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.852494144118555, | |
| "learning_rate": 3.139334608914795e-06, | |
| "loss": 0.9018, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.145761545326323, | |
| "learning_rate": 3.135475658260801e-06, | |
| "loss": 0.8081, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 3.7323925690162945, | |
| "learning_rate": 3.1316150882282486e-06, | |
| "loss": 0.7904, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 3.828580217813585, | |
| "learning_rate": 3.1277529086550044e-06, | |
| "loss": 0.8209, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 7.169193160513471, | |
| "learning_rate": 3.1238891293830344e-06, | |
| "loss": 0.8518, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 3.7890433555958705, | |
| "learning_rate": 3.120023760258384e-06, | |
| "loss": 0.81, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 5.4100072300835915, | |
| "learning_rate": 3.1161568111311487e-06, | |
| "loss": 0.8143, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.745257896570691, | |
| "learning_rate": 3.112288291855449e-06, | |
| "loss": 0.7812, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 3.989422633795527, | |
| "learning_rate": 3.108418212289408e-06, | |
| "loss": 0.7939, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 3.69468243636268, | |
| "learning_rate": 3.1045465822951265e-06, | |
| "loss": 0.7719, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 3.734646448984784, | |
| "learning_rate": 3.1006734117386517e-06, | |
| "loss": 0.8206, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.406895307041761, | |
| "learning_rate": 3.0967987104899623e-06, | |
| "loss": 0.849, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 5.120926190738434, | |
| "learning_rate": 3.092922488422933e-06, | |
| "loss": 0.8067, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.395257481206765, | |
| "learning_rate": 3.0890447554153153e-06, | |
| "loss": 0.7899, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.381884034429759, | |
| "learning_rate": 3.0851655213487123e-06, | |
| "loss": 0.8286, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 3.4432938469880736, | |
| "learning_rate": 3.0812847961085527e-06, | |
| "loss": 0.7555, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 5.491662807975452, | |
| "learning_rate": 3.077402589584061e-06, | |
| "loss": 0.7953, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 5.109785316548567, | |
| "learning_rate": 3.073518911668241e-06, | |
| "loss": 0.8068, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.816255710860229, | |
| "learning_rate": 3.0696337722578444e-06, | |
| "loss": 0.8116, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.0550180873058945, | |
| "learning_rate": 3.0657471812533464e-06, | |
| "loss": 0.7445, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 6.428409369009171, | |
| "learning_rate": 3.061859148558922e-06, | |
| "loss": 0.7809, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 7.399300412933422, | |
| "learning_rate": 3.057969684082421e-06, | |
| "loss": 0.9132, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.438275269071275, | |
| "learning_rate": 3.05407879773534e-06, | |
| "loss": 0.7677, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.029683232488459, | |
| "learning_rate": 3.0501864994328002e-06, | |
| "loss": 0.8277, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 6.985086578254067, | |
| "learning_rate": 3.04629279909352e-06, | |
| "loss": 0.7709, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 3.989233419159114, | |
| "learning_rate": 3.0423977066397913e-06, | |
| "loss": 0.7942, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 5.59818214654168, | |
| "learning_rate": 3.038501231997454e-06, | |
| "loss": 0.7889, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.738840603418049, | |
| "learning_rate": 3.0346033850958685e-06, | |
| "loss": 0.8176, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.316647635809497, | |
| "learning_rate": 3.0307041758678933e-06, | |
| "loss": 0.771, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 5.364622380247118, | |
| "learning_rate": 3.0268036142498596e-06, | |
| "loss": 0.7959, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 9.785481265487443, | |
| "learning_rate": 3.0229017101815424e-06, | |
| "loss": 0.8661, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.766052528284133, | |
| "learning_rate": 3.018998473606139e-06, | |
| "loss": 0.7826, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.701099515941216, | |
| "learning_rate": 3.0150939144702425e-06, | |
| "loss": 0.8236, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 9.461997482098507, | |
| "learning_rate": 3.011188042723816e-06, | |
| "loss": 0.7458, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 5.380245649553912, | |
| "learning_rate": 3.007280868320167e-06, | |
| "loss": 0.9145, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 10.163712466924556, | |
| "learning_rate": 3.0033724012159244e-06, | |
| "loss": 0.8479, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.2825214697150455, | |
| "learning_rate": 2.9994626513710085e-06, | |
| "loss": 0.8418, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.133634973081057, | |
| "learning_rate": 2.9955516287486103e-06, | |
| "loss": 0.7698, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.448695379643913, | |
| "learning_rate": 2.9916393433151636e-06, | |
| "loss": 0.8066, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.466338581251602, | |
| "learning_rate": 2.9877258050403214e-06, | |
| "loss": 0.8007, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 10.810936084229239, | |
| "learning_rate": 2.9838110238969266e-06, | |
| "loss": 0.8186, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.803654636244916, | |
| "learning_rate": 2.9798950098609925e-06, | |
| "loss": 0.8928, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.603349966398751, | |
| "learning_rate": 2.975977772911671e-06, | |
| "loss": 0.82, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.1551088640505, | |
| "learning_rate": 2.9720593230312337e-06, | |
| "loss": 0.8218, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.340225183652387, | |
| "learning_rate": 2.968139670205041e-06, | |
| "loss": 0.8037, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.864128739357516, | |
| "learning_rate": 2.964218824421518e-06, | |
| "loss": 0.7639, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 3.9960489196439664, | |
| "learning_rate": 2.960296795672132e-06, | |
| "loss": 0.7699, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.852346612250402, | |
| "learning_rate": 2.9563735939513637e-06, | |
| "loss": 0.8014, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.3848451966395325, | |
| "learning_rate": 2.9524492292566824e-06, | |
| "loss": 0.8093, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 8.205235412527301, | |
| "learning_rate": 2.9485237115885223e-06, | |
| "loss": 0.8619, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 6.176360237119004, | |
| "learning_rate": 2.9445970509502547e-06, | |
| "loss": 0.7676, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.630007100247436, | |
| "learning_rate": 2.9406692573481634e-06, | |
| "loss": 0.8479, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.947598794426371, | |
| "learning_rate": 2.93674034079142e-06, | |
| "loss": 0.8159, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 5.208912980619626, | |
| "learning_rate": 2.932810311292058e-06, | |
| "loss": 0.7876, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.419301085882204, | |
| "learning_rate": 2.928879178864946e-06, | |
| "loss": 0.8171, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.453129993869502, | |
| "learning_rate": 2.9249469535277635e-06, | |
| "loss": 0.8192, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 3.978894479574167, | |
| "learning_rate": 2.9210136453009753e-06, | |
| "loss": 0.8177, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.085215406941065, | |
| "learning_rate": 2.9170792642078057e-06, | |
| "loss": 0.7986, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.6198554119473725, | |
| "learning_rate": 2.9131438202742123e-06, | |
| "loss": 0.7632, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.722723234078222, | |
| "learning_rate": 2.9092073235288633e-06, | |
| "loss": 0.8149, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.181515177263234, | |
| "learning_rate": 2.9052697840031065e-06, | |
| "loss": 0.8288, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.794572021110422, | |
| "learning_rate": 2.901331211730949e-06, | |
| "loss": 0.7967, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.424733629629019, | |
| "learning_rate": 2.897391616749031e-06, | |
| "loss": 0.7658, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.217478880240176, | |
| "learning_rate": 2.8934510090965943e-06, | |
| "loss": 0.7973, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 8.450381133458173, | |
| "learning_rate": 2.889509398815467e-06, | |
| "loss": 0.8968, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 6.678550127443945, | |
| "learning_rate": 2.885566795950028e-06, | |
| "loss": 0.7672, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.847140586125922, | |
| "learning_rate": 2.8816232105471864e-06, | |
| "loss": 0.7939, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.027451532927153, | |
| "learning_rate": 2.877678652656358e-06, | |
| "loss": 0.7753, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 5.483294011279615, | |
| "learning_rate": 2.8737331323294315e-06, | |
| "loss": 0.7848, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.358769311669604, | |
| "learning_rate": 2.8697866596207524e-06, | |
| "loss": 0.9225, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 3.9764053893031757, | |
| "learning_rate": 2.865839244587093e-06, | |
| "loss": 0.8153, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.817588176503937, | |
| "learning_rate": 2.861890897287625e-06, | |
| "loss": 0.8854, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 5.88032298358066, | |
| "learning_rate": 2.8579416277838952e-06, | |
| "loss": 0.8062, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 5.4231075227119225, | |
| "learning_rate": 2.8539914461398043e-06, | |
| "loss": 0.8082, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.255932637689089, | |
| "learning_rate": 2.8500403624215733e-06, | |
| "loss": 0.801, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.8397836719406153, | |
| "learning_rate": 2.846088386697723e-06, | |
| "loss": 0.8418, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.9932629786289753, | |
| "learning_rate": 2.8421355290390506e-06, | |
| "loss": 0.8162, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.487065147178443, | |
| "learning_rate": 2.838181799518595e-06, | |
| "loss": 0.8014, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.914529347080895, | |
| "learning_rate": 2.8342272082116214e-06, | |
| "loss": 0.8356, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.004022127968303, | |
| "learning_rate": 2.83027176519559e-06, | |
| "loss": 0.8389, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 8.828551661954858, | |
| "learning_rate": 2.82631548055013e-06, | |
| "loss": 0.7885, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.492552440308808, | |
| "learning_rate": 2.822358364357015e-06, | |
| "loss": 0.8, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.342294518537859, | |
| "learning_rate": 2.8184004267001427e-06, | |
| "loss": 0.7626, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.14513404654822, | |
| "learning_rate": 2.8144416776654964e-06, | |
| "loss": 0.7403, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.207420214999504, | |
| "learning_rate": 2.8104821273411333e-06, | |
| "loss": 0.7719, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 5.76849216288104, | |
| "learning_rate": 2.8065217858171495e-06, | |
| "loss": 0.772, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 5.087940784747258, | |
| "learning_rate": 2.802560663185658e-06, | |
| "loss": 0.755, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 5.804412743927739, | |
| "learning_rate": 2.7985987695407618e-06, | |
| "loss": 0.7937, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.3563337300652725, | |
| "learning_rate": 2.7946361149785304e-06, | |
| "loss": 0.7953, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 5.353344075424233, | |
| "learning_rate": 2.79067270959697e-06, | |
| "loss": 0.7835, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.606865353160219, | |
| "learning_rate": 2.786708563496002e-06, | |
| "loss": 0.8217, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 5.151168827461153, | |
| "learning_rate": 2.7827436867774334e-06, | |
| "loss": 0.7994, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.3077719748381815, | |
| "learning_rate": 2.7787780895449353e-06, | |
| "loss": 0.81, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 5.939054267141848, | |
| "learning_rate": 2.774811781904013e-06, | |
| "loss": 0.8043, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 3.6482393833284537, | |
| "learning_rate": 2.7708447739619833e-06, | |
| "loss": 0.7984, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.469643319922747, | |
| "learning_rate": 2.7668770758279473e-06, | |
| "loss": 0.7579, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.297430505606569, | |
| "learning_rate": 2.762908697612765e-06, | |
| "loss": 0.7396, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 5.299365762769891, | |
| "learning_rate": 2.7589396494290287e-06, | |
| "loss": 0.9516, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.820980097501582, | |
| "learning_rate": 2.7549699413910387e-06, | |
| "loss": 0.7753, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.070459691106533, | |
| "learning_rate": 2.750999583614777e-06, | |
| "loss": 0.7792, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 3.9507862249719974, | |
| "learning_rate": 2.7470285862178804e-06, | |
| "loss": 0.7307, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 3.6716263593517255, | |
| "learning_rate": 2.743056959319616e-06, | |
| "loss": 0.7952, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.098216751623958, | |
| "learning_rate": 2.739084713040856e-06, | |
| "loss": 0.7741, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 5.192569762900567, | |
| "learning_rate": 2.73511185750405e-06, | |
| "loss": 0.7985, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 4.261338885304019, | |
| "learning_rate": 2.7311384028332e-06, | |
| "loss": 0.7996, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 4.104647555320273, | |
| "learning_rate": 2.7271643591538355e-06, | |
| "loss": 0.7984, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 4.119079048596377, | |
| "learning_rate": 2.723189736592986e-06, | |
| "loss": 0.8057, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 15.046890388512228, | |
| "learning_rate": 2.719214545279158e-06, | |
| "loss": 0.805, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.22626992572264, | |
| "learning_rate": 2.7152387953423047e-06, | |
| "loss": 0.8173, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.450359321360147, | |
| "learning_rate": 2.711262496913805e-06, | |
| "loss": 0.7907, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 4.084842715465518, | |
| "learning_rate": 2.707285660126435e-06, | |
| "loss": 0.7497, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 4.0680641158312865, | |
| "learning_rate": 2.703308295114342e-06, | |
| "loss": 0.8914, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 3.9396881683951595, | |
| "learning_rate": 2.6993304120130197e-06, | |
| "loss": 0.7543, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 8.168996657657983, | |
| "learning_rate": 2.6953520209592827e-06, | |
| "loss": 0.832, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.615116753830786, | |
| "learning_rate": 2.69137313209124e-06, | |
| "loss": 0.7986, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.1733727377866865, | |
| "learning_rate": 2.6873937555482664e-06, | |
| "loss": 0.8034, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 13.47432961457841, | |
| "learning_rate": 2.683413901470984e-06, | |
| "loss": 0.8469, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.634549734894035, | |
| "learning_rate": 2.6794335800012294e-06, | |
| "loss": 0.7886, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 5.128741319695401, | |
| "learning_rate": 2.6754528012820288e-06, | |
| "loss": 0.7701, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.1198392744805545, | |
| "learning_rate": 2.6714715754575757e-06, | |
| "loss": 0.803, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.337325262757384, | |
| "learning_rate": 2.6674899126732045e-06, | |
| "loss": 0.8442, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.143567365031371, | |
| "learning_rate": 2.663507823075358e-06, | |
| "loss": 0.7851, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 5.745829541899124, | |
| "learning_rate": 2.6595253168115707e-06, | |
| "loss": 0.7785, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.933563947271949, | |
| "learning_rate": 2.65554240403044e-06, | |
| "loss": 0.8497, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 3.798457207492107, | |
| "learning_rate": 2.6515590948815934e-06, | |
| "loss": 0.7611, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.660299502636275, | |
| "learning_rate": 2.6475753995156746e-06, | |
| "loss": 0.8, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.930258601091212, | |
| "learning_rate": 2.643591328084309e-06, | |
| "loss": 1.0316, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.17822237801153, | |
| "learning_rate": 2.6396068907400784e-06, | |
| "loss": 0.8839, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.338544848126124, | |
| "learning_rate": 2.635622097636501e-06, | |
| "loss": 0.7836, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.40163809982145, | |
| "learning_rate": 2.631636958928e-06, | |
| "loss": 0.7965, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.411528020705482, | |
| "learning_rate": 2.6276514847698763e-06, | |
| "loss": 0.745, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.007219388292756, | |
| "learning_rate": 2.623665685318291e-06, | |
| "loss": 0.7881, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 3.914836989848749, | |
| "learning_rate": 2.6196795707302304e-06, | |
| "loss": 0.7929, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.063241782135146, | |
| "learning_rate": 2.6156931511634838e-06, | |
| "loss": 0.801, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.213904333151254, | |
| "learning_rate": 2.61170643677662e-06, | |
| "loss": 0.8638, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.285745911664518, | |
| "learning_rate": 2.607719437728957e-06, | |
| "loss": 0.77, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.689303741924254, | |
| "learning_rate": 2.603732164180539e-06, | |
| "loss": 0.7846, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.666129815883496, | |
| "learning_rate": 2.5997446262921105e-06, | |
| "loss": 0.7803, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.477070260564313, | |
| "learning_rate": 2.595756834225089e-06, | |
| "loss": 0.7998, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.184569743823347, | |
| "learning_rate": 2.5917687981415375e-06, | |
| "loss": 0.8212, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.242849593041545, | |
| "learning_rate": 2.5877805282041456e-06, | |
| "loss": 0.7881, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 3.735902947064209, | |
| "learning_rate": 2.583792034576194e-06, | |
| "loss": 0.7876, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.46938280982608, | |
| "learning_rate": 2.5798033274215363e-06, | |
| "loss": 0.8288, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.033690870298263, | |
| "learning_rate": 2.575814416904569e-06, | |
| "loss": 0.8134, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.597709987316143, | |
| "learning_rate": 2.5718253131902084e-06, | |
| "loss": 0.823, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.142434190114618, | |
| "learning_rate": 2.5678360264438608e-06, | |
| "loss": 0.797, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.1203821819631115, | |
| "learning_rate": 2.5638465668314006e-06, | |
| "loss": 0.7622, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 10.857001786231981, | |
| "learning_rate": 2.5598569445191418e-06, | |
| "loss": 0.7962, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 5.560092742577767, | |
| "learning_rate": 2.5558671696738145e-06, | |
| "loss": 0.7627, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.847492024501584, | |
| "learning_rate": 2.5518772524625356e-06, | |
| "loss": 0.812, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.635222745766512, | |
| "learning_rate": 2.547887203052786e-06, | |
| "loss": 0.8278, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.436016580970295, | |
| "learning_rate": 2.543897031612382e-06, | |
| "loss": 0.7846, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 3.7357290599674355, | |
| "learning_rate": 2.539906748309454e-06, | |
| "loss": 0.8013, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.288805373285202, | |
| "learning_rate": 2.535916363312414e-06, | |
| "loss": 0.8952, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.097132254798519, | |
| "learning_rate": 2.531925886789935e-06, | |
| "loss": 0.7853, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.933595710527326, | |
| "learning_rate": 2.5279353289109227e-06, | |
| "loss": 0.7475, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.626882278982255, | |
| "learning_rate": 2.52394469984449e-06, | |
| "loss": 0.7467, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.637647959979263, | |
| "learning_rate": 2.5199540097599316e-06, | |
| "loss": 0.8511, | |
| "step": 989 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1978, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 989, | |
| "total_flos": 1585526184345600.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |