diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,70117 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.984008395592314,
+  "eval_steps": 1500,
+  "global_step": 10004,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019989505509607455,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 1.476,
+      "step": 1
+    },
+    {
+      "epoch": 0.00019989505509607455,
+      "eval_loss": 1.7460540533065796,
+      "eval_runtime": 595.8959,
+      "eval_samples_per_second": 3.588,
+      "eval_steps_per_second": 3.588,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003997901101921491,
+      "grad_norm": 12.625,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.6568,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005996851652882237,
+      "grad_norm": 13.125,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.5199,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007995802203842982,
+      "grad_norm": 11.75,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.5059,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009994752754803728,
+      "grad_norm": 10.125,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.568,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011993703305764475,
+      "grad_norm": 10.5625,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.5257,
+      "step": 6
+    },
+    {
+      "epoch": 0.001399265385672522,
+      "grad_norm": 10.6875,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.4327,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015991604407685964,
+      "grad_norm": 9.1875,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.5052,
+      "step": 8
+    },
+    {
+      "epoch": 0.001799055495864671,
+      "grad_norm": 9.5625,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.5233,
+      "step": 9
+    },
+    {
+      "epoch": 0.0019989505509607455,
+      "grad_norm": 8.875,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.5549,
+      "step": 10
+    },
+    {
+      "epoch": 0.0021988456060568202,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1e-06,
+      "loss": 1.5139,
+      "step": 11
+    },
+    {
+      "epoch": 0.002398740661152895,
+      "grad_norm": 6.625,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.4168,
+      "step": 12
+    },
+    {
+      "epoch": 0.002598635716248969,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.3e-06,
+      "loss": 1.452,
+      "step": 13
+    },
+    {
+      "epoch": 0.002798530771345044,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.384,
+      "step": 14
+    },
+    {
+      "epoch": 0.0029984258264411185,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.5e-06,
+      "loss": 1.4018,
+      "step": 15
+    },
+    {
+      "epoch": 0.003198320881537193,
+      "grad_norm": 5.09375,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.3572,
+      "step": 16
+    },
+    {
+      "epoch": 0.0033982159366332675,
+      "grad_norm": 4.9375,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.3639,
+      "step": 17
+    },
+    {
+      "epoch": 0.003598110991729342,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.4174,
+      "step": 18
+    },
+    {
+      "epoch": 0.003798006046825417,
+      "grad_norm": 4.625,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.3139,
+      "step": 19
+    },
+    {
+      "epoch": 0.003997901101921491,
+      "grad_norm": 4.4375,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.468,
+      "step": 20
+    },
+    {
+      "epoch": 0.004197796157017566,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.3801,
+      "step": 21
+    },
+    {
+      "epoch": 0.0043976912121136405,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.2e-06,
+      "loss": 1.3115,
+      "step": 22
+    },
+    {
+      "epoch": 0.004597586267209715,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 1.4058,
+      "step": 23
+    },
+    {
+      "epoch": 0.00479748132230579,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.3053,
+      "step": 24
+    },
+    {
+      "epoch": 0.004997376377401864,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.5e-06,
+      "loss": 1.285,
+      "step": 25
+    },
+    {
+      "epoch": 0.005197271432497938,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.6e-06,
+      "loss": 1.4823,
+      "step": 26
+    },
+    {
+      "epoch": 0.005397166487594013,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 1.3546,
+      "step": 27
+    },
+    {
+      "epoch": 0.005597061542690088,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.3501,
+      "step": 28
+    },
+    {
+      "epoch": 0.005796956597786162,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.9e-06,
+      "loss": 1.3478,
+      "step": 29
+    },
+    {
+      "epoch": 0.005996851652882237,
+      "grad_norm": 3.28125,
+      "learning_rate": 3e-06,
+      "loss": 1.2747,
+      "step": 30
+    },
+    {
+      "epoch": 0.006196746707978312,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 1.4141,
+      "step": 31
+    },
+    {
+      "epoch": 0.006396641763074386,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.3632,
+      "step": 32
+    },
+    {
+      "epoch": 0.00659653681817046,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 1.2251,
+      "step": 33
+    },
+    {
+      "epoch": 0.006796431873266535,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.3892,
+      "step": 34
+    },
+    {
+      "epoch": 0.00699632692836261,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.5e-06,
+      "loss": 1.3128,
+      "step": 35
+    },
+    {
+      "epoch": 0.007196221983458684,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.2615,
+      "step": 36
+    },
+    {
+      "epoch": 0.007396117038554759,
+      "grad_norm": 3.25,
+      "learning_rate": 3.7e-06,
+      "loss": 1.3802,
+      "step": 37
+    },
+    {
+      "epoch": 0.007596012093650834,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.3061,
+      "step": 38
+    },
+    {
+      "epoch": 0.0077959071487469075,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 1.2044,
+      "step": 39
+    },
+    {
+      "epoch": 0.007995802203842982,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3376,
+      "step": 40
+    },
+    {
+      "epoch": 0.008195697258939058,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.1e-06,
+      "loss": 1.3461,
+      "step": 41
+    },
+    {
+      "epoch": 0.008395592314035132,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.2514,
+      "step": 42
+    },
+    {
+      "epoch": 0.008595487369131205,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.3e-06,
+      "loss": 1.2735,
+      "step": 43
+    },
+    {
+      "epoch": 0.008795382424227281,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4e-06,
+      "loss": 1.2665,
+      "step": 44
+    },
+    {
+      "epoch": 0.008995277479323355,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5e-06,
+      "loss": 1.3754,
+      "step": 45
+    },
+    {
+      "epoch": 0.00919517253441943,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.2434,
+      "step": 46
+    },
+    {
+      "epoch": 0.009395067589515504,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7e-06,
+      "loss": 1.3023,
+      "step": 47
+    },
+    {
+      "epoch": 0.00959496264461158,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.2093,
+      "step": 48
+    },
+    {
+      "epoch": 0.009794857699707653,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.3093,
+      "step": 49
+    },
+    {
+      "epoch": 0.009994752754803727,
+      "grad_norm": 2.5,
+      "learning_rate": 5e-06,
+      "loss": 1.299,
+      "step": 50
+    },
+    {
+      "epoch": 0.010194647809899803,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.1e-06,
+      "loss": 1.2736,
+      "step": 51
+    },
+    {
+      "epoch": 0.010394542864995877,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.2e-06,
+      "loss": 1.2607,
+      "step": 52
+    },
+    {
+      "epoch": 0.010594437920091952,
+      "grad_norm": 2.5,
+      "learning_rate": 5.300000000000001e-06,
+      "loss": 1.2765,
+      "step": 53
+    },
+    {
+      "epoch": 0.010794332975188026,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.2538,
+      "step": 54
+    },
+    {
+      "epoch": 0.010994228030284102,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.2165,
+      "step": 55
+    },
+    {
+      "epoch": 0.011194123085380175,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.2152,
+      "step": 56
+    },
+    {
+      "epoch": 0.01139401814047625,
+      "grad_norm": 2.375,
+      "learning_rate": 5.7e-06,
+      "loss": 1.2222,
+      "step": 57
+    },
+    {
+      "epoch": 0.011593913195572325,
+      "grad_norm": 2.375,
+      "learning_rate": 5.8e-06,
+      "loss": 1.2779,
+      "step": 58
+    },
+    {
+      "epoch": 0.011793808250668399,
+      "grad_norm": 2.375,
+      "learning_rate": 5.9e-06,
+      "loss": 1.239,
+      "step": 59
+    },
+    {
+      "epoch": 0.011993703305764474,
+      "grad_norm": 2.34375,
+      "learning_rate": 6e-06,
+      "loss": 1.3348,
+      "step": 60
+    },
+    {
+      "epoch": 0.012193598360860548,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.1e-06,
+      "loss": 1.2105,
+      "step": 61
+    },
+    {
+      "epoch": 0.012393493415956624,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.2447,
+      "step": 62
+    },
+    {
+      "epoch": 0.012593388471052697,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 1.2812,
+      "step": 63
+    },
+    {
+      "epoch": 0.012793283526148771,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.1675,
+      "step": 64
+    },
+    {
+      "epoch": 0.012993178581244847,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.3107,
+      "step": 65
+    },
+    {
+      "epoch": 0.01319307363634092,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.2161,
+      "step": 66
+    },
+    {
+      "epoch": 0.013392968691436996,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 1.2651,
+      "step": 67
+    },
+    {
+      "epoch": 0.01359286374653307,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.2281,
+      "step": 68
+    },
+    {
+      "epoch": 0.013792758801629145,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.9e-06,
+      "loss": 1.3876,
+      "step": 69
+    },
+    {
+      "epoch": 0.01399265385672522,
+      "grad_norm": 2.578125,
+      "learning_rate": 7e-06,
+      "loss": 1.3301,
+      "step": 70
+    },
+    {
+      "epoch": 0.014192548911821293,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.100000000000001e-06,
+      "loss": 1.234,
+      "step": 71
+    },
+    {
+      "epoch": 0.014392443966917369,
+      "grad_norm": 2.5,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.2861,
+      "step": 72
+    },
+    {
+      "epoch": 0.014592339022013442,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.3e-06,
+      "loss": 1.2072,
+      "step": 73
+    },
+    {
+      "epoch": 0.014792234077109518,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.4e-06,
+      "loss": 1.2894,
+      "step": 74
+    },
+    {
+      "epoch": 0.014992129132205592,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.1828,
+      "step": 75
+    },
+    {
+      "epoch": 0.015192024187301667,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.1208,
+      "step": 76
+    },
+    {
+      "epoch": 0.015391919242397741,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.7e-06,
+      "loss": 1.1869,
+      "step": 77
+    },
+    {
+      "epoch": 0.015591814297493815,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.232,
+      "step": 78
+    },
+    {
+      "epoch": 0.01579170935258989,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.9e-06,
+      "loss": 1.1701,
+      "step": 79
+    },
+    {
+      "epoch": 0.015991604407685964,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.1675,
+      "step": 80
+    },
+    {
+      "epoch": 0.01619149946278204,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.1e-06,
+      "loss": 1.3752,
+      "step": 81
+    },
+    {
+      "epoch": 0.016391394517878115,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.2e-06,
+      "loss": 1.1496,
+      "step": 82
+    },
+    {
+      "epoch": 0.016591289572974188,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.3e-06,
+      "loss": 1.1925,
+      "step": 83
+    },
+    {
+      "epoch": 0.016791184628070263,
+      "grad_norm": 2.5,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.2526,
+      "step": 84
+    },
+    {
+      "epoch": 0.01699107968316634,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.5e-06,
+      "loss": 1.2725,
+      "step": 85
+    },
+    {
+      "epoch": 0.01719097473826241,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.6e-06,
+      "loss": 1.2927,
+      "step": 86
+    },
+    {
+      "epoch": 0.017390869793358486,
+      "grad_norm": 2.375,
+      "learning_rate": 8.700000000000001e-06,
+      "loss": 1.2508,
+      "step": 87
+    },
+    {
+      "epoch": 0.017590764848454562,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.8e-06,
+      "loss": 1.164,
+      "step": 88
+    },
+    {
+      "epoch": 0.017790659903550637,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 1.2274,
+      "step": 89
+    },
+    {
+      "epoch": 0.01799055495864671,
+      "grad_norm": 2.234375,
+      "learning_rate": 9e-06,
+      "loss": 1.221,
+      "step": 90
+    },
+    {
+      "epoch": 0.018190450013742785,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 1.2804,
+      "step": 91
+    },
+    {
+      "epoch": 0.01839034506883886,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.1848,
+      "step": 92
+    },
+    {
+      "epoch": 0.018590240123934933,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.3e-06,
+      "loss": 1.2262,
+      "step": 93
+    },
+    {
+      "epoch": 0.018790135179031008,
+      "grad_norm": 2.25,
+      "learning_rate": 9.4e-06,
+      "loss": 1.2016,
+      "step": 94
+    },
+    {
+      "epoch": 0.018990030234127084,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.5e-06,
+      "loss": 1.2149,
+      "step": 95
+    },
+    {
+      "epoch": 0.01918992528922316,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.2341,
+      "step": 96
+    },
+    {
+      "epoch": 0.01938982034431923,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.7e-06,
+      "loss": 1.2795,
+      "step": 97
+    },
+    {
+      "epoch": 0.019589715399415307,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.2028,
+      "step": 98
+    },
+    {
+      "epoch": 0.019789610454511383,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.9e-06,
+      "loss": 1.1827,
+      "step": 99
+    },
+    {
+      "epoch": 0.019989505509607455,
+      "grad_norm": 2.328125,
+      "learning_rate": 1e-05,
+      "loss": 1.2169,
+      "step": 100
+    },
+    {
+      "epoch": 0.02018940056470353,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.999999888950268e-06,
+      "loss": 1.1576,
+      "step": 101
+    },
+    {
+      "epoch": 0.020389295619799606,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.999999555801075e-06,
+      "loss": 1.1927,
+      "step": 102
+    },
+    {
+      "epoch": 0.02058919067489568,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.999999000552435e-06,
+      "loss": 1.1975,
+      "step": 103
+    },
+    {
+      "epoch": 0.020789085729991753,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.999998223204373e-06,
+      "loss": 1.2024,
+      "step": 104
+    },
+    {
+      "epoch": 0.02098898078508783,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.999997223756924e-06,
+      "loss": 1.267,
+      "step": 105
+    },
+    {
+      "epoch": 0.021188875840183904,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.99999600221013e-06,
+      "loss": 1.2559,
+      "step": 106
+    },
+    {
+      "epoch": 0.021388770895279977,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.999994558564052e-06,
+      "loss": 1.179,
+      "step": 107
+    },
+    {
+      "epoch": 0.021588665950376052,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.999992892818746e-06,
+      "loss": 1.0606,
+      "step": 108
+    },
+    {
+      "epoch": 0.021788561005472128,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.999991004974292e-06,
+      "loss": 1.1812,
+      "step": 109
+    },
+    {
+      "epoch": 0.021988456060568203,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.999988895030772e-06,
+      "loss": 1.1652,
+      "step": 110
+    },
+    {
+      "epoch": 0.022188351115664275,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.999986562988278e-06,
+      "loss": 1.1498,
+      "step": 111
+    },
+    {
+      "epoch": 0.02238824617076035,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.999984008846914e-06,
+      "loss": 1.3272,
+      "step": 112
+    },
+    {
+      "epoch": 0.022588141225856426,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.999981232606796e-06,
+      "loss": 1.2065,
+      "step": 113
+    },
+    {
+      "epoch": 0.0227880362809525,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.999978234268047e-06,
+      "loss": 1.1944,
+      "step": 114
+    },
+    {
+      "epoch": 0.022987931336048574,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.999975013830797e-06,
+      "loss": 1.1866,
+      "step": 115
+    },
+    {
+      "epoch": 0.02318782639114465,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.99997157129519e-06,
+      "loss": 1.1906,
+      "step": 116
+    },
+    {
+      "epoch": 0.023387721446240725,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.99996790666138e-06,
+      "loss": 1.1623,
+      "step": 117
+    },
+    {
+      "epoch": 0.023587616501336797,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.99996401992953e-06,
+      "loss": 1.4069,
+      "step": 118
+    },
+    {
+      "epoch": 0.023787511556432873,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.999959911099814e-06,
+      "loss": 1.1261,
+      "step": 119
+    },
+    {
+      "epoch": 0.02398740661152895,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.999955580172411e-06,
+      "loss": 1.2397,
+      "step": 120
+    },
+    {
+      "epoch": 0.02418730166662502,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.999951027147514e-06,
+      "loss": 1.3055,
+      "step": 121
+    },
+    {
+      "epoch": 0.024387196721721096,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.999946252025329e-06,
+      "loss": 1.1119,
+      "step": 122
+    },
+    {
+      "epoch": 0.02458709177681717,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.999941254806065e-06,
+      "loss": 1.2665,
+      "step": 123
+    },
+    {
+      "epoch": 0.024786986831913247,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.999936035489943e-06,
+      "loss": 1.1914,
+      "step": 124
+    },
+    {
+      "epoch": 0.02498688188700932,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.999930594077199e-06,
+      "loss": 1.2627,
+      "step": 125
+    },
+    {
+      "epoch": 0.025186776942105395,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.99992493056807e-06,
+      "loss": 1.2579,
+      "step": 126
+    },
+    {
+      "epoch": 0.02538667199720147,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.999919044962809e-06,
+      "loss": 1.2031,
+      "step": 127
+    },
+    {
+      "epoch": 0.025586567052297542,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.999912937261679e-06,
+      "loss": 1.2567,
+      "step": 128
+    },
+    {
+      "epoch": 0.025786462107393618,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.99990660746495e-06,
+      "loss": 1.1828,
+      "step": 129
+    },
+    {
+      "epoch": 0.025986357162489693,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.999900055572904e-06,
+      "loss": 1.2224,
+      "step": 130
+    },
+    {
+      "epoch": 0.02618625221758577,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.99989328158583e-06,
+      "loss": 1.1482,
+      "step": 131
+    },
+    {
+      "epoch": 0.02638614727268184,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.999886285504033e-06,
+      "loss": 1.1253,
+      "step": 132
+    },
+    {
+      "epoch": 0.026586042327777917,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.99987906732782e-06,
+      "loss": 1.1303,
+      "step": 133
+    },
+    {
+      "epoch": 0.026785937382873992,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.999871627057511e-06,
+      "loss": 1.2155,
+      "step": 134
+    },
+    {
+      "epoch": 0.026985832437970064,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.999863964693441e-06,
+      "loss": 1.1278,
+      "step": 135
+    },
+    {
+      "epoch": 0.02718572749306614,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.999856080235947e-06,
+      "loss": 1.1915,
+      "step": 136
+    },
+    {
+      "epoch": 0.027385622548162215,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.99984797368538e-06,
+      "loss": 1.1856,
+      "step": 137
+    },
+    {
+      "epoch": 0.02758551760325829,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.9998396450421e-06,
+      "loss": 1.2587,
+      "step": 138
+    },
+    {
+      "epoch": 0.027785412658354363,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.999831094306475e-06,
+      "loss": 1.2584,
+      "step": 139
+    },
+    {
+      "epoch": 0.02798530771345044,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.999822321478889e-06,
+      "loss": 1.2475,
+      "step": 140
+    },
+    {
+      "epoch": 0.028185202768546514,
+      "grad_norm": 2.375,
+      "learning_rate": 9.999813326559728e-06,
+      "loss": 1.149,
+      "step": 141
+    },
+    {
+      "epoch": 0.028385097823642586,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.999804109549397e-06,
+      "loss": 1.2239,
+      "step": 142
+    },
+    {
+      "epoch": 0.028584992878738662,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.999794670448298e-06,
+      "loss": 1.1599,
+      "step": 143
+    },
+    {
+      "epoch": 0.028784887933834737,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.999785009256853e-06,
+      "loss": 1.2722,
+      "step": 144
+    },
+    {
+      "epoch": 0.028984782988930813,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.999775125975492e-06,
+      "loss": 1.1292,
+      "step": 145
+    },
+    {
+      "epoch": 0.029184678044026885,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.999765020604655e-06,
+      "loss": 1.1786,
+      "step": 146
+    },
+    {
+      "epoch": 0.02938457309912296,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.99975469314479e-06,
+      "loss": 1.1699,
+      "step": 147
+    },
+    {
+      "epoch": 0.029584468154219036,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.999744143596354e-06,
+      "loss": 1.25,
+      "step": 148
+    },
+    {
+      "epoch": 0.029784363209315108,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.99973337195982e-06,
+      "loss": 1.2246,
+      "step": 149
+    },
+    {
+      "epoch": 0.029984258264411184,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.999722378235661e-06,
+      "loss": 1.173,
+      "step": 150
+    },
+    {
+      "epoch": 0.03018415331950726,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.99971116242437e-06,
+      "loss": 1.1347,
+      "step": 151
+    },
+    {
+      "epoch": 0.030384048374603335,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.99969972452644e-06,
+      "loss": 1.1403,
+      "step": 152
+    },
+    {
+      "epoch": 0.030583943429699407,
+      "grad_norm": 2.25,
+      "learning_rate": 9.999688064542385e-06,
+      "loss": 1.2141,
+      "step": 153
+    },
+    {
+      "epoch": 0.030783838484795482,
+      "grad_norm": 2.375,
+      "learning_rate": 9.99967618247272e-06,
+      "loss": 1.1917,
+      "step": 154
+    },
+    {
+      "epoch": 0.030983733539891558,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.999664078317972e-06,
+      "loss": 1.1414,
+      "step": 155
+    },
+    {
+      "epoch": 0.03118362859498763,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.999651752078681e-06,
+      "loss": 1.2256,
+      "step": 156
+    },
+    {
+      "epoch": 0.031383523650083706,
+      "grad_norm": 2.25,
+      "learning_rate": 9.999639203755392e-06,
+      "loss": 1.1904,
+      "step": 157
+    },
+    {
+      "epoch": 0.03158341870517978,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.999626433348664e-06,
+      "loss": 1.2102,
+      "step": 158
+    },
+    {
+      "epoch": 0.03178331376027586,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.999613440859064e-06,
+      "loss": 1.1549,
+      "step": 159
+    },
+    {
+      "epoch": 0.03198320881537193,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.999600226287168e-06,
+      "loss": 1.1984,
+      "step": 160
+    },
+    {
+      "epoch": 0.032183103870468,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.999586789633565e-06,
+      "loss": 1.1532,
+      "step": 161
+    },
+    {
+      "epoch": 0.03238299892556408,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.99957313089885e-06,
+      "loss": 1.2764,
+      "step": 162
+    },
+    {
+      "epoch": 0.03258289398066015,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.999559250083631e-06,
+      "loss": 1.2102,
+      "step": 163
+    },
+    {
+      "epoch": 0.03278278903575623,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.999545147188523e-06,
+      "loss": 1.1573,
+      "step": 164
+    },
+    {
+      "epoch": 0.0329826840908523,
+      "grad_norm": 2.25,
+      "learning_rate": 9.999530822214154e-06,
+      "loss": 1.2516,
+      "step": 165
+    },
+    {
+      "epoch": 0.033182579145948375,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.99951627516116e-06,
+      "loss": 1.2498,
+      "step": 166
+    },
+    {
+      "epoch": 0.033382474201044454,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.999501506030187e-06,
+      "loss": 1.0588,
+      "step": 167
+    },
+    {
+      "epoch": 0.033582369256140526,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.999486514821892e-06,
+      "loss": 1.1712,
+      "step": 168
+    },
+    {
+      "epoch": 0.0337822643112366,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.999471301536938e-06,
+      "loss": 1.0943,
+      "step": 169
+    },
+    {
+      "epoch": 0.03398215936633268,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.999455866176004e-06,
+      "loss": 1.1701,
+      "step": 170
+    },
+    {
+      "epoch": 0.03418205442142875,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.999440208739775e-06,
+      "loss": 1.216,
+      "step": 171
+    },
+    {
+      "epoch": 0.03438194947652482,
+      "grad_norm": 2.25,
+      "learning_rate": 9.999424329228944e-06,
+      "loss": 1.1877,
+      "step": 172
+    },
+    {
+      "epoch": 0.0345818445316209,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.999408227644221e-06,
+      "loss": 1.191,
+      "step": 173
+    },
+    {
+      "epoch": 0.03478173958671697,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.999391903986316e-06,
+      "loss": 1.1894,
+      "step": 174
+    },
+    {
+      "epoch": 0.034981634641813045,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.999375358255959e-06,
+      "loss": 1.2301,
+      "step": 175
+    },
+    {
+      "epoch": 0.035181529696909124,
+      "grad_norm": 2.25,
+      "learning_rate": 9.99935859045388e-06,
+      "loss": 1.1464,
+      "step": 176
+    },
+    {
+      "epoch": 0.035381424752005196,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.999341600580827e-06,
+      "loss": 1.1863,
+      "step": 177
+    },
+    {
+      "epoch": 0.035581319807101275,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.999324388637554e-06,
+      "loss": 1.1387,
+      "step": 178
+    },
+    {
+      "epoch": 0.03578121486219735,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.999306954624826e-06,
+      "loss": 1.2182,
+      "step": 179
+    },
+    {
+      "epoch": 0.03598110991729342,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.999289298543417e-06,
+      "loss": 1.1852,
+      "step": 180
+    },
+    {
+      "epoch": 0.0361810049723895,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.999271420394112e-06,
+      "loss": 1.1018,
+      "step": 181
+    },
+    {
+      "epoch": 0.03638090002748557,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.999253320177704e-06,
+      "loss": 1.3003,
+      "step": 182
+    },
+    {
+      "epoch": 0.03658079508258164,
+      "grad_norm": 2.25,
+      "learning_rate": 9.999234997894998e-06,
+      "loss": 1.15,
+      "step": 183
+    },
+    {
+      "epoch": 0.03678069013767772,
+      "grad_norm": 2.375,
+      "learning_rate": 9.999216453546805e-06,
+      "loss": 1.1683,
+      "step": 184
+    },
+    {
+      "epoch": 0.03698058519277379,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.999197687133953e-06,
+      "loss": 1.1442,
+      "step": 185
+    },
+    {
+      "epoch": 0.037180480247869865,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.999178698657273e-06,
+      "loss": 1.3008,
+      "step": 186
+    },
+    {
+      "epoch": 0.037380375302965944,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.99915948811761e-06,
+      "loss": 1.1976,
+      "step": 187
+    },
+    {
+      "epoch": 0.037580270358062016,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.999140055515817e-06,
+      "loss": 1.2784,
+      "step": 188
+    },
+    {
+      "epoch": 0.03778016541315809,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.999120400852756e-06,
+      "loss": 1.251,
+      "step": 189
+    },
+    {
+      "epoch": 0.03798006046825417,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.999100524129299e-06,
+      "loss": 1.2041,
+      "step": 190
+    },
+    {
+      "epoch": 0.03817995552335024,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.999080425346333e-06,
+      "loss": 1.2354,
+      "step": 191
+    },
+    {
+      "epoch": 0.03837985057844632,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.999060104504746e-06,
+      "loss": 1.1894,
+      "step": 192
+    },
+    {
+      "epoch": 0.03857974563354239,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.999039561605445e-06,
+      "loss": 1.2374,
+      "step": 193
+    },
+    {
+      "epoch": 0.03877964068863846,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.99901879664934e-06,
+      "loss": 1.1506,
+      "step": 194
+    },
+    {
+      "epoch": 0.03897953574373454,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.998997809637354e-06,
+      "loss": 1.1861,
+      "step": 195
+    },
+    {
+      "epoch": 0.039179430798830614,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.998976600570418e-06,
+      "loss": 1.1206,
+      "step": 196
+    },
+    {
+      "epoch": 0.039379325853926686,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.998955169449479e-06,
+      "loss": 1.1798,
+      "step": 197
+    },
+    {
+      "epoch": 0.039579220909022765,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.998933516275482e-06,
+      "loss": 1.1726,
+      "step": 198
+    },
+    {
+      "epoch": 0.03977911596411884,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.998911641049393e-06,
+      "loss": 1.2349,
+      "step": 199
+    },
+    {
+      "epoch": 0.03997901101921491,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.998889543772182e-06,
+      "loss": 1.0874,
+      "step": 200
+    },
+    {
+      "epoch": 0.04017890607431099,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.998867224444832e-06,
+      "loss": 1.141,
+      "step": 201
+    },
+    {
+      "epoch": 0.04037880112940706,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.998844683068335e-06,
+      "loss": 1.2468,
+      "step": 202
+    },
+    {
+      "epoch": 0.04057869618450313,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.998821919643689e-06,
+      "loss": 1.0987,
+      "step": 203
+    },
+    {
+      "epoch": 0.04077859123959921,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.998798934171908e-06,
+      "loss": 1.127,
+      "step": 204
+    },
+    {
+      "epoch": 0.040978486294695284,
+      "grad_norm": 2.375,
+      "learning_rate": 9.998775726654014e-06,
+      "loss": 1.1358,
+      "step": 205
+    },
+    {
+      "epoch": 0.04117838134979136,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.998752297091033e-06,
+      "loss": 1.195,
+      "step": 206
+    },
+    {
+      "epoch": 0.041378276404887435,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.99872864548401e-06,
+      "loss": 1.142,
+      "step": 207
+    },
+    {
+      "epoch": 0.04157817145998351,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.998704771833996e-06,
+      "loss": 1.1762,
+      "step": 208
+    },
+    {
+      "epoch": 0.041778066515079586,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.998680676142048e-06,
+      "loss": 1.162,
+      "step": 209
+    },
+    {
+      "epoch": 0.04197796157017566,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.99865635840924e-06,
+      "loss": 1.2461,
+      "step": 210
+    },
+    {
+      "epoch": 0.04217785662527173,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.998631818636648e-06,
+      "loss": 1.159,
+      "step": 211
+    },
+    {
+      "epoch": 0.04237775168036781,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.998607056825367e-06,
+      "loss": 1.1312,
+      "step": 212
+    },
+    {
+      "epoch": 0.04257764673546388,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.998582072976493e-06,
+      "loss": 1.1496,
+      "step": 213
+    },
+    {
+      "epoch": 0.04277754179055995,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.998556867091137e-06,
+      "loss": 1.1292,
+      "step": 214
+    },
+    {
+      "epoch": 0.04297743684565603,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.99853143917042e-06,
+      "loss": 1.1355,
+      "step": 215
+    },
+    {
+      "epoch": 0.043177331900752104,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.998505789215469e-06,
+      "loss": 1.2097,
+      "step": 216
+    },
+    {
+      "epoch": 0.043377226955848176,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.998479917227427e-06,
+      "loss": 1.2135,
+      "step": 217
+    },
+    {
+      "epoch": 0.043577122010944255,
+      "grad_norm": 2.25,
+      "learning_rate": 9.998453823207437e-06,
+      "loss": 1.1332,
+      "step": 218
+    },
+    {
+      "epoch": 0.04377701706604033,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.998427507156665e-06,
+      "loss": 1.1971,
+      "step": 219
+    },
+    {
+      "epoch": 0.043976912121136406,
+      "grad_norm": 2.25,
+      "learning_rate": 9.998400969076276e-06,
+      "loss": 1.1738,
+      "step": 220
+    },
+    {
+      "epoch": 0.04417680717623248,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.998374208967451e-06,
+      "loss": 1.2697,
+      "step": 221
+    },
+    {
+      "epoch": 0.04437670223132855,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.998347226831375e-06,
+      "loss": 1.1589,
+      "step": 222
+    },
+    {
+      "epoch": 0.04457659728642463,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.998320022669253e-06,
+      "loss": 1.2381,
+      "step": 223
+    },
+    {
+      "epoch": 0.0447764923415207,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.998292596482286e-06,
+      "loss": 1.1596,
+      "step": 224
+    },
+    {
+      "epoch": 0.044976387396616774,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.998264948271696e-06,
+      "loss": 1.2429,
+      "step": 225
+    },
+    {
+      "epoch": 0.04517628245171285,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.998237078038712e-06,
+      "loss": 1.1608,
+      "step": 226
+    },
+    {
+      "epoch": 0.045376177506808925,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.99820898578457e-06,
+      "loss": 1.1371,
+      "step": 227
+    },
+    {
+      "epoch": 0.045576072561905,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.99818067151052e-06,
+      "loss": 1.1286,
+      "step": 228
+    },
+    {
+      "epoch": 0.045775967617001076,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.998152135217816e-06,
+      "loss": 1.2278,
+      "step": 229
+    },
+    {
+      "epoch": 0.04597586267209715,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.99812337690773e-06,
+      "loss": 1.1407,
+      "step": 230
+    },
+    {
+      "epoch": 0.04617575772719322,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.998094396581538e-06,
+      "loss": 1.1912,
+      "step": 231
+    },
+    {
+      "epoch": 0.0463756527822893,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.998065194240524e-06,
+      "loss": 1.252,
+      "step": 232
+    },
+    {
+      "epoch": 0.04657554783738537,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.99803576988599e-06,
+      "loss": 1.166,
+      "step": 233
+    },
+    {
+      "epoch": 0.04677544289248145,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.99800612351924e-06,
+      "loss": 1.1465,
+      "step": 234
+    },
+    {
+      "epoch": 0.04697533794757752,
+      "grad_norm": 2.25,
+      "learning_rate": 9.997976255141593e-06,
+      "loss": 1.23,
+      "step": 235
+    },
+    {
+      "epoch": 0.047175233002673594,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.997946164754373e-06,
+      "loss": 1.1665,
+      "step": 236
+    },
+    {
+      "epoch": 0.04737512805776967,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.99791585235892e-06,
+      "loss": 1.1391,
+      "step": 237
+    },
+    {
+      "epoch": 0.047575023112865746,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.997885317956577e-06,
+      "loss": 1.152,
+      "step": 238
+    },
+    {
+      "epoch": 0.04777491816796182,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.997854561548702e-06,
+      "loss": 1.153,
+      "step": 239
+    },
+    {
+      "epoch": 0.0479748132230579,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.99782358313666e-06,
+      "loss": 1.189,
+      "step": 240
+    },
+    {
+      "epoch": 0.04817470827815397,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.997792382721831e-06,
+      "loss": 1.1307,
+      "step": 241
+    },
+    {
+      "epoch": 0.04837460333325004,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.997760960305598e-06,
+      "loss": 1.0662,
+      "step": 242
+    },
+    {
+      "epoch": 0.04857449838834612,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.997729315889356e-06,
+      "loss": 1.0934,
+      "step": 243
+    },
+    {
+      "epoch": 0.04877439344344219,
+      "grad_norm": 2.125,
+      "learning_rate": 9.99769744947451e-06,
+      "loss": 1.2033,
+      "step": 244
+    },
+    {
+      "epoch": 0.048974288498538264,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.99766536106248e-06,
+      "loss": 1.153,
+      "step": 245
+    },
+    {
+      "epoch": 0.04917418355363434,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.997633050654687e-06,
+      "loss": 1.172,
+      "step": 246
+    },
+    {
+      "epoch": 0.049374078608730415,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.997600518252568e-06,
+      "loss": 1.23,
+      "step": 247
+    },
+    {
+      "epoch": 0.049573973663826494,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.997567763857566e-06,
+      "loss": 1.1873,
+      "step": 248
+    },
+    {
+      "epoch": 0.049773868718922566,
+      "grad_norm": 2.375,
+      "learning_rate": 9.997534787471139e-06,
+      "loss": 1.1962,
+      "step": 249
+    },
+    {
+      "epoch": 0.04997376377401864,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.997501589094752e-06,
+      "loss": 1.1827,
+      "step": 250
+    },
+    {
+      "epoch": 0.05017365882911472,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.997468168729876e-06,
+      "loss": 1.0892,
+      "step": 251
+    },
+    {
+      "epoch": 0.05037355388421079,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.997434526377998e-06,
+      "loss": 1.1739,
+      "step": 252
+    },
+    {
+      "epoch": 0.05057344893930686,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.997400662040613e-06,
+      "loss": 1.0799,
+      "step": 253
+    },
+    {
+      "epoch": 0.05077334399440294,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.997366575719223e-06,
+      "loss": 1.1721,
+      "step": 254
+    },
+    {
+      "epoch": 0.05097323904949901,
+      "grad_norm": 2.25,
+      "learning_rate": 9.997332267415345e-06,
+      "loss": 1.2029,
+      "step": 255
+    },
+    {
+      "epoch": 0.051173134104595085,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.9972977371305e-06,
+      "loss": 1.1735,
+      "step": 256
+    },
+    {
+      "epoch": 0.051373029159691164,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.997262984866225e-06,
+      "loss": 1.1807,
+      "step": 257
+    },
+    {
+      "epoch": 0.051572924214787236,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.997228010624061e-06,
+      "loss": 1.2564,
+      "step": 258
+    },
+    {
+      "epoch": 0.05177281926988331,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.997192814405561e-06,
+      "loss": 1.1182,
+      "step": 259
+    },
+    {
+      "epoch": 0.05197271432497939,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.99715739621229e-06,
+      "loss": 1.2241,
+      "step": 260
+    },
+    {
+      "epoch": 0.05217260938007546,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.997121756045824e-06,
+      "loss": 1.1293,
+      "step": 261
+    },
+    {
+      "epoch": 0.05237250443517154,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.997085893907742e-06,
+      "loss": 1.1547,
+      "step": 262
+    },
+    {
+      "epoch": 0.05257239949026761,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.997049809799639e-06,
+      "loss": 1.1602,
+      "step": 263
+    },
+    {
+      "epoch": 0.05277229454536368,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.997013503723116e-06,
+      "loss": 1.157,
+      "step": 264
+    },
+    {
+      "epoch": 0.05297218960045976,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.996976975679786e-06,
+      "loss": 1.186,
+      "step": 265
+    },
+    {
+      "epoch": 0.05317208465555583,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.996940225671275e-06,
+      "loss": 1.1212,
+      "step": 266
+    },
+    {
+      "epoch": 0.053371979710651905,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.996903253699211e-06,
+      "loss": 1.1478,
+      "step": 267
+    },
+    {
+      "epoch": 0.053571874765747984,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.996866059765238e-06,
+      "loss": 1.1511,
+      "step": 268
+    },
+    {
+      "epoch": 0.053771769820844056,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.996828643871008e-06,
+      "loss": 1.119,
+      "step": 269
+    },
+    {
+      "epoch": 0.05397166487594013,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.996791006018185e-06,
+      "loss": 1.1387,
+      "step": 270
+    },
+    {
+      "epoch": 0.05417155993103621,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.996753146208438e-06,
+      "loss": 1.1503,
+      "step": 271
+    },
+    {
+      "epoch": 0.05437145498613228,
+      "grad_norm": 2.25,
+      "learning_rate": 9.996715064443449e-06,
+      "loss": 1.2386,
+      "step": 272
+    },
+    {
+      "epoch": 0.05457135004122835,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.996676760724913e-06,
+      "loss": 1.1869,
+      "step": 273
+    },
+    {
+      "epoch": 0.05477124509632443,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.996638235054527e-06,
+      "loss": 1.1408,
+      "step": 274
+    },
+    {
+      "epoch": 0.0549711401514205,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.996599487434007e-06,
+      "loss": 1.2918,
+      "step": 275
+    },
+    {
+      "epoch": 0.05517103520651658,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.996560517865069e-06,
+      "loss": 1.1994,
+      "step": 276
+    },
+    {
+      "epoch": 0.055370930261612654,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.996521326349447e-06,
+      "loss": 1.2154,
+      "step": 277
+    },
+    {
+      "epoch": 0.055570825316708726,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.996481912888881e-06,
+      "loss": 1.2332,
+      "step": 278
+    },
+    {
+      "epoch": 0.055770720371804805,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.996442277485122e-06,
+      "loss": 1.1615,
+      "step": 279
+    },
+    {
+      "epoch": 0.05597061542690088,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.99640242013993e-06,
+      "loss": 1.2607,
+      "step": 280
+    },
+    {
+      "epoch": 0.05617051048199695,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.996362340855076e-06,
+      "loss": 1.1624,
+      "step": 281
+    },
+    {
+      "epoch": 0.05637040553709303,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.996322039632343e-06,
+      "loss": 1.1052,
+      "step": 282
+    },
+    {
+      "epoch": 0.0565703005921891,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.996281516473517e-06,
+      "loss": 1.1284,
+      "step": 283
+    },
+    {
+      "epoch": 0.05677019564728517,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.9962407713804e-06,
+      "loss": 1.1039,
+      "step": 284
+    },
+    {
+      "epoch": 0.05697009070238125,
+      "grad_norm": 2.125,
+      "learning_rate": 9.9961998043548e-06,
+      "loss": 1.1431,
+      "step": 285
+    },
+    {
+      "epoch": 0.057169985757477323,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.996158615398541e-06,
+      "loss": 1.1113,
+      "step": 286
+    },
+    {
+      "epoch": 0.057369880812573396,
+      "grad_norm": 2.5,
+      "learning_rate": 9.996117204513448e-06,
+      "loss": 1.1783,
+      "step": 287
+    },
+    {
+      "epoch": 0.057569775867669475,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.996075571701363e-06,
+      "loss": 1.0627,
+      "step": 288
+    },
+    {
+      "epoch": 0.05776967092276555,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.996033716964136e-06,
+      "loss": 1.1035,
+      "step": 289
+    },
+    {
+      "epoch": 0.057969565977861626,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.995991640303622e-06,
+      "loss": 1.1161,
+      "step": 290
+    },
+    {
+      "epoch": 0.0581694610329577,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.995949341721695e-06,
+      "loss": 1.1651,
+      "step": 291
+    },
+    {
+      "epoch": 0.05836935608805377,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.995906821220231e-06,
+      "loss": 1.1721,
+      "step": 292
+    },
+    {
+      "epoch": 0.05856925114314985,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.99586407880112e-06,
+      "loss": 1.1163,
+      "step": 293
+    },
+    {
+      "epoch": 0.05876914619824592,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.99582111446626e-06,
+      "loss": 1.1774,
+      "step": 294
+    },
+    {
+      "epoch": 0.05896904125334199,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.99577792821756e-06,
+      "loss": 1.2022,
+      "step": 295
+    },
+    {
+      "epoch": 0.05916893630843807,
+      "grad_norm": 2.25,
+      "learning_rate": 9.995734520056936e-06,
+      "loss": 1.1483,
+      "step": 296
+    },
+    {
+      "epoch": 0.059368831363534144,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.995690889986321e-06,
+      "loss": 1.1886,
+      "step": 297
+    },
+    {
+      "epoch": 0.059568726418630216,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.99564703800765e-06,
+      "loss": 1.0558,
+      "step": 298
+    },
+    {
+      "epoch": 0.059768621473726295,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.99560296412287e-06,
+      "loss": 1.1211,
+      "step": 299
+    },
+    {
+      "epoch": 0.05996851652882237,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.99555866833394e-06,
+      "loss": 1.1445,
+      "step": 300
+    },
+    {
+      "epoch": 0.06016841158391844,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.995514150642827e-06,
+      "loss": 1.122,
+      "step": 301
+    },
+    {
+      "epoch": 0.06036830663901452,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.995469411051511e-06,
+      "loss": 1.0816,
+      "step": 302
+    },
+    {
+      "epoch": 0.06056820169411059,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.995424449561974e-06,
+      "loss": 1.2716,
+      "step": 303
+    },
+    {
+      "epoch": 0.06076809674920667,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.99537926617622e-06,
+      "loss": 1.1145,
+      "step": 304
+    },
+    {
+      "epoch": 0.06096799180430274,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.99533386089625e-06,
+      "loss": 1.112,
+      "step": 305
+    },
+    {
+      "epoch": 0.061167886859398814,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.995288233724084e-06,
+      "loss": 1.1509,
+      "step": 306
+    },
+    {
+      "epoch": 0.06136778191449489,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.995242384661748e-06,
+      "loss": 1.186,
+      "step": 307
+    },
+    {
+      "epoch": 0.061567676969590965,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.995196313711279e-06,
+      "loss": 1.1512,
+      "step": 308
+    },
+    {
+      "epoch": 0.06176757202468704,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.995150020874724e-06,
+      "loss": 1.136,
+      "step": 309
+    },
+    {
+      "epoch": 0.061967467079783116,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.995103506154138e-06,
+      "loss": 1.2276,
+      "step": 310
+    },
+    {
+      "epoch": 0.06216736213487919,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.995056769551587e-06,
+      "loss": 1.1692,
+      "step": 311
+    },
+    {
+      "epoch": 0.06236725718997526,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.99500981106915e-06,
+      "loss": 1.3003,
+      "step": 312
+    },
+    {
+      "epoch": 0.06256715224507134,
+      "grad_norm": 2.25,
+      "learning_rate": 9.99496263070891e-06,
+      "loss": 1.0791,
+      "step": 313
+    },
+    {
+      "epoch": 0.06276704730016741,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.99491522847296e-06,
+      "loss": 1.2507,
+      "step": 314
+    },
+    {
+      "epoch": 0.06296694235526348,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.994867604363415e-06,
+      "loss": 1.3307,
+      "step": 315
+    },
+    {
+      "epoch": 0.06316683741035956,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.994819758382379e-06,
+      "loss": 1.0752,
+      "step": 316
+    },
+    {
+      "epoch": 0.06336673246545564,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.994771690531986e-06,
+      "loss": 1.137,
+      "step": 317
+    },
+    {
+      "epoch": 0.06356662752055171,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.994723400814367e-06,
+      "loss": 1.1592,
+      "step": 318
+    },
+    {
+      "epoch": 0.06376652257564779,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.994674889231668e-06,
+      "loss": 1.1348,
+      "step": 319
+    },
+    {
+      "epoch": 0.06396641763074386,
+      "grad_norm": 2.375,
+      "learning_rate": 9.994626155786044e-06,
+      "loss": 1.1495,
+      "step": 320
+    },
+    {
+      "epoch": 0.06416631268583993,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.994577200479659e-06,
+      "loss": 1.1524,
+      "step": 321
+    },
+    {
+      "epoch": 0.064366207740936,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.994528023314689e-06,
+      "loss": 1.1618,
+      "step": 322
+    },
+    {
+      "epoch": 0.06456610279603209,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.994478624293317e-06,
+      "loss": 1.1259,
+      "step": 323
+    },
+    {
+      "epoch": 0.06476599785112816,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.994429003417739e-06,
+      "loss": 1.2054,
+      "step": 324
+    },
+    {
+      "epoch": 0.06496589290622423,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.994379160690156e-06,
+      "loss": 1.164,
+      "step": 325
+    },
+    {
+      "epoch": 0.0651657879613203,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.994329096112786e-06,
+      "loss": 1.1282,
+      "step": 326
+    },
+    {
+      "epoch": 0.06536568301641638,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.994278809687849e-06,
+      "loss": 1.1085,
+      "step": 327
+    },
+    {
+      "epoch": 0.06556557807151246,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.994228301417584e-06,
+      "loss": 1.1176,
+      "step": 328
+    },
+    {
+      "epoch": 0.06576547312660853,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.994177571304228e-06,
+      "loss": 1.1428,
+      "step": 329
+    },
+    {
+      "epoch": 0.0659653681817046,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.99412661935004e-06,
+      "loss": 1.0369,
+      "step": 330
+    },
+    {
+      "epoch": 0.06616526323680068,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.99407544555728e-06,
+      "loss": 1.1183,
+      "step": 331
+    },
+    {
+      "epoch": 0.06636515829189675,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.994024049928222e-06,
+      "loss": 1.2004,
+      "step": 332
+    },
+    {
+      "epoch": 0.06656505334699282,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.99397243246515e-06,
+      "loss": 1.0809,
+      "step": 333
+    },
+    {
+      "epoch": 0.06676494840208891,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.993920593170355e-06,
+      "loss": 1.153,
+      "step": 334
+    },
+    {
+      "epoch": 0.06696484345718498,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.99386853204614e-06,
+      "loss": 1.1371,
+      "step": 335
+    },
+    {
+      "epoch": 0.06716473851228105,
+      "grad_norm": 2.25,
+      "learning_rate": 9.993816249094818e-06,
+      "loss": 1.0658,
+      "step": 336
+    },
+    {
+      "epoch": 0.06736463356737712,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.993763744318711e-06,
+      "loss": 1.1894,
+      "step": 337
+    },
+    {
+      "epoch": 0.0675645286224732,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.993711017720155e-06,
+      "loss": 1.1231,
+      "step": 338
+    },
+    {
+      "epoch": 0.06776442367756928,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.993658069301487e-06,
+      "loss": 1.1764,
+      "step": 339
+    },
+    {
+      "epoch": 0.06796431873266535,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.993604899065061e-06,
+      "loss": 1.1511,
+      "step": 340
+    },
+    {
+      "epoch": 0.06816421378776143,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.993551507013239e-06,
+      "loss": 1.1341,
+      "step": 341
+    },
+    {
+      "epoch": 0.0683641088428575,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.993497893148391e-06,
+      "loss": 1.1729,
+      "step": 342
+    },
+    {
+      "epoch": 0.06856400389795357,
+      "grad_norm": 2.25,
+      "learning_rate": 9.993444057472902e-06,
+      "loss": 1.2644,
+      "step": 343
+    },
+    {
+      "epoch": 0.06876389895304964,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.99338999998916e-06,
+      "loss": 1.1547,
+      "step": 344
+    },
+    {
+      "epoch": 0.06896379400814573,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.99333572069957e-06,
+      "loss": 1.0723,
+      "step": 345
+    },
+    {
+      "epoch": 0.0691636890632418,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.993281219606537e-06,
+      "loss": 1.1183,
+      "step": 346
+    },
+    {
+      "epoch": 0.06936358411833787,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.993226496712488e-06,
+      "loss": 1.1073,
+      "step": 347
+    },
+    {
+      "epoch": 0.06956347917343395,
+      "grad_norm": 2.25,
+      "learning_rate": 9.993171552019849e-06,
+      "loss": 1.1318,
+      "step": 348
+    },
+    {
+      "epoch": 0.06976337422853002,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.993116385531064e-06,
+      "loss": 1.202,
+      "step": 349
+    },
+    {
+      "epoch": 0.06996326928362609,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.993060997248582e-06,
+      "loss": 1.1376,
+      "step": 350
+    },
+    {
+      "epoch": 0.07016316433872218,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.993005387174865e-06,
+      "loss": 1.1328,
+      "step": 351
+    },
+    {
+      "epoch": 0.07036305939381825,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.99294955531238e-06,
+      "loss": 1.2125,
+      "step": 352
+    },
+    {
+      "epoch": 0.07056295444891432,
+      "grad_norm": 2.375,
+      "learning_rate": 9.992893501663613e-06,
+      "loss": 1.2495,
+      "step": 353
+    },
+    {
+      "epoch": 0.07076284950401039,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.992837226231046e-06,
+      "loss": 1.0926,
+      "step": 354
+    },
+    {
+      "epoch": 0.07096274455910646,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.992780729017184e-06,
+      "loss": 1.1878,
+      "step": 355
+    },
+    {
+      "epoch": 0.07116263961420255,
+      "grad_norm": 2.25,
+      "learning_rate": 9.992724010024536e-06,
+      "loss": 1.2096,
+      "step": 356
+    },
+    {
+      "epoch": 0.07136253466929862,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.99266706925562e-06,
+      "loss": 1.1596,
+      "step": 357
+    },
+    {
+      "epoch": 0.0715624297243947,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.992609906712967e-06,
+      "loss": 1.1814,
+      "step": 358
+    },
+    {
+      "epoch": 0.07176232477949077,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.992552522399112e-06,
+      "loss": 1.213,
+      "step": 359
+    },
+    {
+      "epoch": 0.07196221983458684,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.99249491631661e-06,
+      "loss": 1.1729,
+      "step": 360
+    },
+    {
+      "epoch": 0.07216211488968291,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.992437088468016e-06,
+      "loss": 1.058,
+      "step": 361
+    },
+    {
+      "epoch": 0.072362009944779,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.992379038855902e-06,
+      "loss": 1.131,
+      "step": 362
+    },
+    {
+      "epoch": 0.07256190499987507,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.992320767482842e-06,
+      "loss": 1.0648,
+      "step": 363
+    },
+    {
+      "epoch": 0.07276180005497114,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.992262274351427e-06,
+      "loss": 1.1781,
+      "step": 364
+    },
+    {
+      "epoch": 0.07296169511006721,
+      "grad_norm": 2.625,
+      "learning_rate": 9.992203559464256e-06,
+      "loss": 1.1641,
+      "step": 365
+    },
+    {
+      "epoch": 0.07316159016516328,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.992144622823933e-06,
+      "loss": 1.144,
+      "step": 366
+    },
+    {
+      "epoch": 0.07336148522025937,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.992085464433084e-06,
+      "loss": 1.0778,
+      "step": 367
+    },
+    {
+      "epoch": 0.07356138027535544,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.992026084294328e-06,
+      "loss": 1.1035,
+      "step": 368
+    },
+    {
+      "epoch": 0.07376127533045151,
+      "grad_norm": 2.25,
+      "learning_rate": 9.991966482410307e-06,
+      "loss": 1.1741,
+      "step": 369
+    },
+    {
+      "epoch": 0.07396117038554759,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.99190665878367e-06,
+      "loss": 1.1478,
+      "step": 370
+    },
+    {
+      "epoch": 0.07416106544064366,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.99184661341707e-06,
+      "loss": 1.1733,
+      "step": 371
+    },
+    {
+      "epoch": 0.07436096049573973,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.991786346313178e-06,
+      "loss": 1.2256,
+      "step": 372
+    },
+    {
+      "epoch": 0.07456085555083582,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.991725857474668e-06,
+      "loss": 1.1197,
+      "step": 373
+    },
+    {
+      "epoch": 0.07476075060593189,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.99166514690423e-06,
+      "loss": 1.0933,
+      "step": 374
+    },
+    {
+      "epoch": 0.07496064566102796,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.99160421460456e-06,
+      "loss": 1.1109,
+      "step": 375
+    },
+    {
+      "epoch": 0.07516054071612403,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.991543060578363e-06,
+      "loss": 1.1302,
+      "step": 376
+    },
+    {
+      "epoch": 0.0753604357712201,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.991481684828356e-06,
+      "loss": 1.0812,
+      "step": 377
+    },
+    {
+      "epoch": 0.07556033082631618,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.991420087357266e-06,
+      "loss": 1.1587,
+      "step": 378
+    },
+    {
+      "epoch": 0.07576022588141226,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.991358268167828e-06,
+      "loss": 1.0641,
+      "step": 379
+    },
+    {
+      "epoch": 0.07596012093650834,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.99129622726279e-06,
+      "loss": 1.0763,
+      "step": 380
+    },
+    {
+      "epoch": 0.07616001599160441,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.991233964644906e-06,
+      "loss": 1.2373,
+      "step": 381
+    },
+    {
+      "epoch": 0.07635991104670048,
+      "grad_norm": 2.375,
+      "learning_rate": 9.991171480316944e-06,
+      "loss": 1.2241,
+      "step": 382
+    },
+    {
+      "epoch": 0.07655980610179655,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.991108774281676e-06,
+      "loss": 1.2159,
+      "step": 383
+    },
+    {
+      "epoch": 0.07675970115689264,
+      "grad_norm": 2.25,
+      "learning_rate": 9.99104584654189e-06,
+      "loss": 1.0728,
+      "step": 384
+    },
+    {
+      "epoch": 0.07695959621198871,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.99098269710038e-06,
+      "loss": 1.0239,
+      "step": 385
+    },
+    {
+      "epoch": 0.07715949126708478,
+      "grad_norm": 2.125,
+      "learning_rate": 9.990919325959952e-06,
+      "loss": 1.1313,
+      "step": 386
+    },
+    {
+      "epoch": 0.07735938632218085,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.990855733123421e-06,
+      "loss": 1.0394,
+      "step": 387
+    },
+    {
+      "epoch": 0.07755928137727693,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.99079191859361e-06,
+      "loss": 1.1339,
+      "step": 388
+    },
+    {
+      "epoch": 0.077759176432373,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.990727882373357e-06,
+      "loss": 1.1582,
+      "step": 389
+    },
+    {
+      "epoch": 0.07795907148746908,
+      "grad_norm": 2.375,
+      "learning_rate": 9.990663624465504e-06,
+      "loss": 1.1049,
+      "step": 390
+    },
+    {
+      "epoch": 0.07815896654256516,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.990599144872905e-06,
+      "loss": 1.1503,
+      "step": 391
+    },
+    {
+      "epoch": 0.07835886159766123,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.990534443598425e-06,
+      "loss": 1.1675,
+      "step": 392
+    },
+    {
+      "epoch": 0.0785587566527573,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.99046952064494e-06,
+      "loss": 1.084,
+      "step": 393
+    },
+    {
+      "epoch": 0.07875865170785337,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.99040437601533e-06,
+      "loss": 1.1882,
+      "step": 394
+    },
+    {
+      "epoch": 0.07895854676294946,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.99033900971249e-06,
+      "loss": 1.0875,
+      "step": 395
+    },
+    {
+      "epoch": 0.07915844181804553,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.990273421739325e-06,
+      "loss": 1.0732,
+      "step": 396
+    },
+    {
+      "epoch": 0.0793583368731416,
+      "grad_norm": 2.25,
+      "learning_rate": 9.990207612098748e-06,
+      "loss": 1.2069,
+      "step": 397
+    },
+    {
+      "epoch": 0.07955823192823767,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.990141580793682e-06,
+      "loss": 1.1131,
+      "step": 398
+    },
+    {
+      "epoch": 0.07975812698333375,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.990075327827058e-06,
+      "loss": 1.0508,
+      "step": 399
+    },
+    {
+      "epoch": 0.07995802203842982,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.990008853201823e-06,
+      "loss": 1.1659,
+      "step": 400
+    },
+    {
+      "epoch": 0.0801579170935259,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.989942156920926e-06,
+      "loss": 1.1182,
+      "step": 401
+    },
+    {
+      "epoch": 0.08035781214862198,
+      "grad_norm": 2.25,
+      "learning_rate": 9.989875238987333e-06,
+      "loss": 1.1216,
+      "step": 402
+    },
+    {
+      "epoch": 0.08055770720371805,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.989808099404015e-06,
+      "loss": 1.104,
+      "step": 403
+    },
+    {
+      "epoch": 0.08075760225881412,
+      "grad_norm": 2.25,
+      "learning_rate": 9.989740738173953e-06,
+      "loss": 1.1707,
+      "step": 404
+    },
+    {
+      "epoch": 0.08095749731391019,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.989673155300141e-06,
+      "loss": 1.1789,
+      "step": 405
+    },
+    {
+      "epoch": 0.08115739236900626,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.98960535078558e-06,
+      "loss": 1.2305,
+      "step": 406
+    },
+    {
+      "epoch": 0.08135728742410235,
+      "grad_norm": 2.25,
+      "learning_rate": 9.989537324633283e-06,
+      "loss": 1.0805,
+      "step": 407
+    },
+    {
+      "epoch": 0.08155718247919842,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.98946907684627e-06,
+      "loss": 1.2271,
+      "step": 408
+    },
+    {
+      "epoch": 0.0817570775342945,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.989400607427574e-06,
+      "loss": 1.0816,
+      "step": 409
+    },
+    {
+      "epoch": 0.08195697258939057,
+      "grad_norm": 2.25,
+      "learning_rate": 9.989331916380233e-06,
+      "loss": 1.1939,
+      "step": 410
+    },
+    {
+      "epoch": 0.08215686764448664,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.989263003707306e-06,
+      "loss": 1.0495,
+      "step": 411
+    },
+    {
+      "epoch": 0.08235676269958273,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.989193869411846e-06,
+      "loss": 1.0046,
+      "step": 412
+    },
+    {
+      "epoch": 0.0825566577546788,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.989124513496929e-06,
+      "loss": 1.1798,
+      "step": 413
+    },
+    {
+      "epoch": 0.08275655280977487,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.989054935965631e-06,
+      "loss": 1.1185,
+      "step": 414
+    },
+    {
+      "epoch": 0.08295644786487094,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.98898513682105e-06,
+      "loss": 1.142,
+      "step": 415
+    },
+    {
+      "epoch": 0.08315634291996701,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.98891511606628e-06,
+      "loss": 1.2098,
+      "step": 416
+    },
+    {
+      "epoch": 0.08335623797506309,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.988844873704433e-06,
+      "loss": 1.1389,
+      "step": 417
+    },
+    {
+      "epoch": 0.08355613303015917,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.988774409738628e-06,
+      "loss": 1.1318,
+      "step": 418
+    },
+    {
+      "epoch": 0.08375602808525524,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.988703724172e-06,
+      "loss": 1.1519,
+      "step": 419
+    },
+    {
+      "epoch": 0.08395592314035132,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.988632817007683e-06,
+      "loss": 1.1853,
+      "step": 420
+    },
+    {
+      "epoch": 0.08415581819544739,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.98856168824883e-06,
+      "loss": 1.0061,
+      "step": 421
+    },
+    {
+      "epoch": 0.08435571325054346,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.9884903378986e-06,
+      "loss": 1.1319,
+      "step": 422
+    },
+    {
+      "epoch": 0.08455560830563955,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.988418765960161e-06,
+      "loss": 1.1209,
+      "step": 423
+    },
+    {
+      "epoch": 0.08475550336073562,
+      "grad_norm": 2.125,
+      "learning_rate": 9.988346972436693e-06,
+      "loss": 1.1002,
+      "step": 424
+    },
+    {
+      "epoch": 0.08495539841583169,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.988274957331385e-06,
+      "loss": 1.0286,
+      "step": 425
+    },
+    {
+      "epoch": 0.08515529347092776,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.988202720647438e-06,
+      "loss": 0.9752,
+      "step": 426
+    },
+    {
+      "epoch": 0.08535518852602383,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.988130262388058e-06,
+      "loss": 1.131,
+      "step": 427
+    },
+    {
+      "epoch": 0.0855550835811199,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.988057582556465e-06,
+      "loss": 1.1002,
+      "step": 428
+    },
+    {
+      "epoch": 0.08575497863621599,
+      "grad_norm": 2.375,
+      "learning_rate": 9.987984681155884e-06,
+      "loss": 1.2404,
+      "step": 429
+    },
+    {
+      "epoch": 0.08595487369131206,
+      "grad_norm": 2.25,
+      "learning_rate": 9.98791155818956e-06,
+      "loss": 1.13,
+      "step": 430
+    },
+    {
+      "epoch": 0.08615476874640814,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.987838213660736e-06,
+      "loss": 1.1198,
+      "step": 431
+    },
+    {
+      "epoch": 0.08635466380150421,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.987764647572671e-06,
+      "loss": 1.1485,
+      "step": 432
+    },
+    {
+      "epoch": 0.08655455885660028,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.987690859928633e-06,
+      "loss": 1.1426,
+      "step": 433
+    },
+    {
+      "epoch": 0.08675445391169635,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.987616850731899e-06,
+      "loss": 1.1378,
+      "step": 434
+    },
+    {
+      "epoch": 0.08695434896679244,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.987542619985758e-06,
+      "loss": 1.2654,
+      "step": 435
+    },
+    {
+      "epoch": 0.08715424402188851,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.987468167693507e-06,
+      "loss": 1.1183,
+      "step": 436
+    },
+    {
+      "epoch": 0.08735413907698458,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.987393493858453e-06,
+      "loss": 1.1122,
+      "step": 437
+    },
+    {
+      "epoch": 0.08755403413208065,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.98731859848391e-06,
+      "loss": 1.1731,
+      "step": 438
+    },
+    {
+      "epoch": 0.08775392918717673,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.98724348157321e-06,
+      "loss": 1.0737,
+      "step": 439
+    },
+    {
+      "epoch": 0.08795382424227281,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.987168143129687e-06,
+      "loss": 1.1597,
+      "step": 440
+    },
+    {
+      "epoch": 0.08815371929736888,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.987092583156688e-06,
+      "loss": 1.1576,
+      "step": 441
+    },
+    {
+      "epoch": 0.08835361435246496,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.987016801657568e-06,
+      "loss": 1.0758,
+      "step": 442
+    },
+    {
+      "epoch": 0.08855350940756103,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.986940798635694e-06,
+      "loss": 1.1427,
+      "step": 443
+    },
+    {
+      "epoch": 0.0887534044626571,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.986864574094443e-06,
+      "loss": 1.1078,
+      "step": 444
+    },
+    {
+      "epoch": 0.08895329951775317,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.986788128037202e-06,
+      "loss": 1.1726,
+      "step": 445
+    },
+    {
+      "epoch": 0.08915319457284926,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.986711460467362e-06,
+      "loss": 1.158,
+      "step": 446
+    },
+    {
+      "epoch": 0.08935308962794533,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.986634571388333e-06,
+      "loss": 1.211,
+      "step": 447
+    },
+    {
+      "epoch": 0.0895529846830414,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.986557460803527e-06,
+      "loss": 1.0706,
+      "step": 448
+    },
+    {
+      "epoch": 0.08975287973813748,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.986480128716374e-06,
+      "loss": 1.0745,
+      "step": 449
+    },
+    {
+      "epoch": 0.08995277479323355,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.986402575130305e-06,
+      "loss": 1.1669,
+      "step": 450
+    },
+    {
+      "epoch": 0.09015266984832963,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.986324800048767e-06,
+      "loss": 1.0876,
+      "step": 451
+    },
+    {
+      "epoch": 0.0903525649034257,
+      "grad_norm": 2.25,
+      "learning_rate": 9.986246803475213e-06,
+      "loss": 1.2014,
+      "step": 452
+    },
+    {
+      "epoch": 0.09055245995852178,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.986168585413108e-06,
+      "loss": 1.2562,
+      "step": 453
+    },
+    {
+      "epoch": 0.09075235501361785,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.98609014586593e-06,
+      "loss": 1.1761,
+      "step": 454
+    },
+    {
+      "epoch": 0.09095225006871392,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.986011484837157e-06,
+      "loss": 1.1152,
+      "step": 455
+    },
+    {
+      "epoch": 0.09115214512381,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.985932602330287e-06,
+      "loss": 1.047,
+      "step": 456
+    },
+    {
+      "epoch": 0.09135204017890608,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.985853498348823e-06,
+      "loss": 1.1492,
+      "step": 457
+    },
+    {
+      "epoch": 0.09155193523400215,
+      "grad_norm": 2.25,
+      "learning_rate": 9.985774172896281e-06,
+      "loss": 1.1138,
+      "step": 458
+    },
+    {
+      "epoch": 0.09175183028909822,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.985694625976181e-06,
+      "loss": 1.106,
+      "step": 459
+    },
+    {
+      "epoch": 0.0919517253441943,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.985614857592058e-06,
+      "loss": 1.1019,
+      "step": 460
+    },
+    {
+      "epoch": 0.09215162039929037,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.985534867747457e-06,
+      "loss": 1.1848,
+      "step": 461
+    },
+    {
+      "epoch": 0.09235151545438644,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.985454656445928e-06,
+      "loss": 1.1093,
+      "step": 462
+    },
+    {
+      "epoch": 0.09255141050948253,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.985374223691039e-06,
+      "loss": 1.1284,
+      "step": 463
+    },
+    {
+      "epoch": 0.0927513055645786,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.985293569486356e-06,
+      "loss": 1.2787,
+      "step": 464
+    },
+    {
+      "epoch": 0.09295120061967467,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.985212693835465e-06,
+      "loss": 1.1167,
+      "step": 465
+    },
+    {
+      "epoch": 0.09315109567477074,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.985131596741959e-06,
+      "loss": 1.158,
+      "step": 466
+    },
+    {
+      "epoch": 0.09335099072986681,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.985050278209442e-06,
+      "loss": 1.2403,
+      "step": 467
+    },
+    {
+      "epoch": 0.0935508857849629,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.984968738241522e-06,
+      "loss": 1.1573,
+      "step": 468
+    },
+    {
+      "epoch": 0.09375078084005897,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.984886976841824e-06,
+      "loss": 1.0895,
+      "step": 469
+    },
+    {
+      "epoch": 0.09395067589515504,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.98480499401398e-06,
+      "loss": 1.0453,
+      "step": 470
+    },
+    {
+      "epoch": 0.09415057095025112,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.984722789761627e-06,
+      "loss": 1.2329,
+      "step": 471
+    },
+    {
+      "epoch": 0.09435046600534719,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.984640364088422e-06,
+      "loss": 1.213,
+      "step": 472
+    },
+    {
+      "epoch": 0.09455036106044326,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.984557716998026e-06,
+      "loss": 1.0911,
+      "step": 473
+    },
+    {
+      "epoch": 0.09475025611553935,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.984474848494107e-06,
+      "loss": 1.0302,
+      "step": 474
+    },
+    {
+      "epoch": 0.09495015117063542,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.984391758580347e-06,
+      "loss": 1.0809,
+      "step": 475
+    },
+    {
+      "epoch": 0.09515004622573149,
+      "grad_norm": 2.25,
+      "learning_rate": 9.984308447260437e-06,
+      "loss": 1.1552,
+      "step": 476
+    },
+    {
+      "epoch": 0.09534994128082756,
+      "grad_norm": 2.25,
+      "learning_rate": 9.98422491453808e-06,
+      "loss": 1.1293,
+      "step": 477
+    },
+    {
+      "epoch": 0.09554983633592364,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.984141160416985e-06,
+      "loss": 1.0692,
+      "step": 478
+    },
+    {
+      "epoch": 0.09574973139101972,
+      "grad_norm": 2.375,
+      "learning_rate": 9.98405718490087e-06,
+      "loss": 1.2213,
+      "step": 479
+    },
+    {
+      "epoch": 0.0959496264461158,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.983972987993468e-06,
+      "loss": 1.1687,
+      "step": 480
+    },
+    {
+      "epoch": 0.09614952150121187,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.983888569698519e-06,
+      "loss": 1.1588,
+      "step": 481
+    },
+    {
+      "epoch": 0.09634941655630794,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.983803930019771e-06,
+      "loss": 1.1676,
+      "step": 482
+    },
+    {
+      "epoch": 0.09654931161140401,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.983719068960985e-06,
+      "loss": 1.1553,
+      "step": 483
+    },
+    {
+      "epoch": 0.09674920666650008,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.983633986525932e-06,
+      "loss": 1.2282,
+      "step": 484
+    },
+    {
+      "epoch": 0.09694910172159617,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.983548682718388e-06,
+      "loss": 1.17,
+      "step": 485
+    },
+    {
+      "epoch": 0.09714899677669224,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.983463157542142e-06,
+      "loss": 1.1835,
+      "step": 486
+    },
+    {
+      "epoch": 0.09734889183178831,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.983377411000996e-06,
+      "loss": 1.1484,
+      "step": 487
+    },
+    {
+      "epoch": 0.09754878688688438,
+      "grad_norm": 2.25,
+      "learning_rate": 9.983291443098759e-06,
+      "loss": 1.1656,
+      "step": 488
+    },
+    {
+      "epoch": 0.09774868194198046,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.983205253839247e-06,
+      "loss": 1.1873,
+      "step": 489
+    },
+    {
+      "epoch": 0.09794857699707653,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.98311884322629e-06,
+      "loss": 1.0371,
+      "step": 490
+    },
+    {
+      "epoch": 0.09814847205217261,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.983032211263725e-06,
+      "loss": 1.1887,
+      "step": 491
+    },
+    {
+      "epoch": 0.09834836710726869,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.982945357955406e-06,
+      "loss": 1.0655,
+      "step": 492
+    },
+    {
+      "epoch": 0.09854826216236476,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.982858283305181e-06,
+      "loss": 1.1893,
+      "step": 493
+    },
+    {
+      "epoch": 0.09874815721746083,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.982770987316926e-06,
+      "loss": 1.1584,
+      "step": 494
+    },
+    {
+      "epoch": 0.0989480522725569,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.982683469994515e-06,
+      "loss": 1.1144,
+      "step": 495
+    },
+    {
+      "epoch": 0.09914794732765299,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.982595731341838e-06,
+      "loss": 1.153,
+      "step": 496
+    },
+    {
+      "epoch": 0.09934784238274906,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.982507771362789e-06,
+      "loss": 1.0859,
+      "step": 497
+    },
+    {
+      "epoch": 0.09954773743784513,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.982419590061277e-06,
+      "loss": 1.1539,
+      "step": 498
+    },
+    {
+      "epoch": 0.0997476324929412,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.98233118744122e-06,
+      "loss": 1.1645,
+      "step": 499
+    },
+    {
+      "epoch": 0.09994752754803728,
+      "grad_norm": 2.25,
+      "learning_rate": 9.982242563506543e-06,
+      "loss": 1.2078,
+      "step": 500
+    },
+    {
+      "epoch": 0.10014742260313335,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.982153718261183e-06,
+      "loss": 1.1374,
+      "step": 501
+    },
+    {
+      "epoch": 0.10034731765822943,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.982064651709088e-06,
+      "loss": 1.1488,
+      "step": 502
+    },
+    {
+      "epoch": 0.1005472127133255,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.981975363854212e-06,
+      "loss": 1.11,
+      "step": 503
+    },
+    {
+      "epoch": 0.10074710776842158,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.981885854700524e-06,
+      "loss": 1.1718,
+      "step": 504
+    },
+    {
+      "epoch": 0.10094700282351765,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.981796124251999e-06,
+      "loss": 1.1763,
+      "step": 505
+    },
+    {
+      "epoch": 0.10114689787861372,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.98170617251262e-06,
+      "loss": 1.103,
+      "step": 506
+    },
+    {
+      "epoch": 0.10134679293370981,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.981615999486386e-06,
+      "loss": 1.1321,
+      "step": 507
+    },
+    {
+      "epoch": 0.10154668798880588,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.981525605177301e-06,
+      "loss": 1.1236,
+      "step": 508
+    },
+    {
+      "epoch": 0.10174658304390195,
+      "grad_norm": 2.125,
+      "learning_rate": 9.981434989589382e-06,
+      "loss": 1.0633,
+      "step": 509
+    },
+    {
+      "epoch": 0.10194647809899803,
+      "grad_norm": 2.25,
+      "learning_rate": 9.981344152726651e-06,
+      "loss": 1.2384,
+      "step": 510
+    },
+    {
+      "epoch": 0.1021463731540941,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.981253094593147e-06,
+      "loss": 1.0622,
+      "step": 511
+    },
+    {
+      "epoch": 0.10234626820919017,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.98116181519291e-06,
+      "loss": 1.1921,
+      "step": 512
+    },
+    {
+      "epoch": 0.10254616326428626,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.981070314529998e-06,
+      "loss": 1.0312,
+      "step": 513
+    },
+    {
+      "epoch": 0.10274605831938233,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.980978592608475e-06,
+      "loss": 1.1928,
+      "step": 514
+    },
+    {
+      "epoch": 0.1029459533744784,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.980886649432413e-06,
+      "loss": 1.2021,
+      "step": 515
+    },
+    {
+      "epoch": 0.10314584842957447,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.9807944850059e-06,
+      "loss": 1.1377,
+      "step": 516
+    },
+    {
+      "epoch": 0.10334574348467054,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.980702099333029e-06,
+      "loss": 1.0915,
+      "step": 517
+    },
+    {
+      "epoch": 0.10354563853976662,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.980609492417901e-06,
+      "loss": 1.1397,
+      "step": 518
+    },
+    {
+      "epoch": 0.1037455335948627,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.980516664264632e-06,
+      "loss": 1.0924,
+      "step": 519
+    },
+    {
+      "epoch": 0.10394542864995877,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.980423614877344e-06,
+      "loss": 1.2099,
+      "step": 520
+    },
+    {
+      "epoch": 0.10414532370505485,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.980330344260172e-06,
+      "loss": 1.124,
+      "step": 521
+    },
+    {
+      "epoch": 0.10434521876015092,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.980236852417256e-06,
+      "loss": 1.1106,
+      "step": 522
+    },
+    {
+      "epoch": 0.10454511381524699,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.980143139352753e-06,
+      "loss": 1.1668,
+      "step": 523
+    },
+    {
+      "epoch": 0.10474500887034308,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.980049205070824e-06,
+      "loss": 1.0663,
+      "step": 524
+    },
+    {
+      "epoch": 0.10494490392543915,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.979955049575639e-06,
+      "loss": 1.1818,
+      "step": 525
+    },
+    {
+      "epoch": 0.10514479898053522,
+      "grad_norm": 2.25,
+      "learning_rate": 9.979860672871384e-06,
+      "loss": 1.1244,
+      "step": 526
+    },
+    {
+      "epoch": 0.10534469403563129,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.979766074962249e-06,
+      "loss": 1.1311,
+      "step": 527
+    },
+    {
+      "epoch": 0.10554458909072736,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.979671255852437e-06,
+      "loss": 1.1327,
+      "step": 528
+    },
+    {
+      "epoch": 0.10574448414582344,
+      "grad_norm": 2.25,
+      "learning_rate": 9.979576215546161e-06,
+      "loss": 1.1418,
+      "step": 529
+    },
+    {
+      "epoch": 0.10594437920091952,
+      "grad_norm": 2.125,
+      "learning_rate": 9.97948095404764e-06,
+      "loss": 1.1011,
+      "step": 530
+    },
+    {
+      "epoch": 0.1061442742560156,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.979385471361108e-06,
+      "loss": 1.0591,
+      "step": 531
+    },
+    {
+      "epoch": 0.10634416931111167,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.979289767490803e-06,
+      "loss": 1.1741,
+      "step": 532
+    },
+    {
+      "epoch": 0.10654406436620774,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.97919384244098e-06,
+      "loss": 1.1312,
+      "step": 533
+    },
+    {
+      "epoch": 0.10674395942130381,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.979097696215898e-06,
+      "loss": 1.1005,
+      "step": 534
+    },
+    {
+      "epoch": 0.1069438544763999,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.979001328819828e-06,
+      "loss": 1.0767,
+      "step": 535
+    },
+    {
+      "epoch": 0.10714374953149597,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.978904740257051e-06,
+      "loss": 1.2502,
+      "step": 536
+    },
+    {
+      "epoch": 0.10734364458659204,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.978807930531857e-06,
+      "loss": 1.0616,
+      "step": 537
+    },
+    {
+      "epoch": 0.10754353964168811,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.978710899648547e-06,
+      "loss": 1.1741,
+      "step": 538
+    },
+    {
+      "epoch": 0.10774343469678418,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.978613647611429e-06,
+      "loss": 1.1575,
+      "step": 539
+    },
+    {
+      "epoch": 0.10794332975188026,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.978516174424826e-06,
+      "loss": 1.2128,
+      "step": 540
+    },
+    {
+      "epoch": 0.10814322480697634,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.978418480093065e-06,
+      "loss": 1.0435,
+      "step": 541
+    },
+    {
+      "epoch": 0.10834311986207242,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.97832056462049e-06,
+      "loss": 1.1376,
+      "step": 542
+    },
+    {
+      "epoch": 0.10854301491716849,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.978222428011444e-06,
+      "loss": 1.2298,
+      "step": 543
+    },
+    {
+      "epoch": 0.10874290997226456,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.97812407027029e-06,
+      "loss": 1.168,
+      "step": 544
+    },
+    {
+      "epoch": 0.10894280502736063,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.978025491401397e-06,
+      "loss": 1.1013,
+      "step": 545
+    },
+    {
+      "epoch": 0.1091427000824567,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.977926691409143e-06,
+      "loss": 1.194,
+      "step": 546
+    },
+    {
+      "epoch": 0.10934259513755279,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.977827670297917e-06,
+      "loss": 1.0775,
+      "step": 547
+    },
+    {
+      "epoch": 0.10954249019264886,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.977728428072118e-06,
+      "loss": 1.1678,
+      "step": 548
+    },
+    {
+      "epoch": 0.10974238524774493,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.977628964736153e-06,
+      "loss": 1.0741,
+      "step": 549
+    },
+    {
+      "epoch": 0.109942280302841,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.977529280294442e-06,
+      "loss": 1.2779,
+      "step": 550
+    },
+    {
+      "epoch": 0.11014217535793708,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.977429374751411e-06,
+      "loss": 1.2253,
+      "step": 551
+    },
+    {
+      "epoch": 0.11034207041303316,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.9773292481115e-06,
+      "loss": 1.065,
+      "step": 552
+    },
+    {
+      "epoch": 0.11054196546812924,
+      "grad_norm": 2.125,
+      "learning_rate": 9.977228900379155e-06,
+      "loss": 1.154,
+      "step": 553
+    },
+    {
+      "epoch": 0.11074186052322531,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.977128331558834e-06,
+      "loss": 1.1185,
+      "step": 554
+    },
+    {
+      "epoch": 0.11094175557832138,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.977027541655003e-06,
+      "loss": 1.1429,
+      "step": 555
+    },
+    {
+      "epoch": 0.11114165063341745,
+      "grad_norm": 2.25,
+      "learning_rate": 9.97692653067214e-06,
+      "loss": 1.1406,
+      "step": 556
+    },
+    {
+      "epoch": 0.11134154568851352,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.976825298614734e-06,
+      "loss": 1.1413,
+      "step": 557
+    },
+    {
+      "epoch": 0.11154144074360961,
+      "grad_norm": 2.125,
+      "learning_rate": 9.976723845487278e-06,
+      "loss": 1.0157,
+      "step": 558
+    },
+    {
+      "epoch": 0.11174133579870568,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.97662217129428e-06,
+      "loss": 1.2004,
+      "step": 559
+    },
+    {
+      "epoch": 0.11194123085380175,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.97652027604026e-06,
+      "loss": 1.159,
+      "step": 560
+    },
+    {
+      "epoch": 0.11214112590889783,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.976418159729737e-06,
+      "loss": 1.0831,
+      "step": 561
+    },
+    {
+      "epoch": 0.1123410209639939,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.976315822367254e-06,
+      "loss": 1.1762,
+      "step": 562
+    },
+    {
+      "epoch": 0.11254091601908998,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.97621326395735e-06,
+      "loss": 1.1847,
+      "step": 563
+    },
+    {
+      "epoch": 0.11274081107418606,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.976110484504587e-06,
+      "loss": 1.0756,
+      "step": 564
+    },
+    {
+      "epoch": 0.11294070612928213,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.976007484013528e-06,
+      "loss": 1.0894,
+      "step": 565
+    },
+    {
+      "epoch": 0.1131406011843782,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.975904262488747e-06,
+      "loss": 1.0933,
+      "step": 566
+    },
+    {
+      "epoch": 0.11334049623947427,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.97580081993483e-06,
+      "loss": 1.0876,
+      "step": 567
+    },
+    {
+      "epoch": 0.11354039129457034,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.975697156356372e-06,
+      "loss": 1.0503,
+      "step": 568
+    },
+    {
+      "epoch": 0.11374028634966643,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.975593271757977e-06,
+      "loss": 1.0898,
+      "step": 569
+    },
+    {
+      "epoch": 0.1139401814047625,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.975489166144262e-06,
+      "loss": 1.1363,
+      "step": 570
+    },
+    {
+      "epoch": 0.11414007645985857,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.975384839519849e-06,
+      "loss": 1.1002,
+      "step": 571
+    },
+    {
+      "epoch": 0.11433997151495465,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.975280291889373e-06,
+      "loss": 1.0651,
+      "step": 572
+    },
+    {
+      "epoch": 0.11453986657005072,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.975175523257477e-06,
+      "loss": 1.0237,
+      "step": 573
+    },
+    {
+      "epoch": 0.11473976162514679,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.975070533628817e-06,
+      "loss": 1.072,
+      "step": 574
+    },
+    {
+      "epoch": 0.11493965668024288,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.974965323008055e-06,
+      "loss": 1.0854,
+      "step": 575
+    },
+    {
+      "epoch": 0.11513955173533895,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.974859891399863e-06,
+      "loss": 1.1208,
+      "step": 576
+    },
+    {
+      "epoch": 0.11533944679043502,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.974754238808927e-06,
+      "loss": 1.1068,
+      "step": 577
+    },
+    {
+      "epoch": 0.1155393418455311,
+      "grad_norm": 2.125,
+      "learning_rate": 9.974648365239938e-06,
+      "loss": 1.0612,
+      "step": 578
+    },
+    {
+      "epoch": 0.11573923690062717,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.974542270697602e-06,
+      "loss": 1.1479,
+      "step": 579
+    },
+    {
+      "epoch": 0.11593913195572325,
+      "grad_norm": 2.125,
+      "learning_rate": 9.974435955186628e-06,
+      "loss": 1.1138,
+      "step": 580
+    },
+    {
+      "epoch": 0.11613902701081932,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.974329418711742e-06,
+      "loss": 1.1786,
+      "step": 581
+    },
+    {
+      "epoch": 0.1163389220659154,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.974222661277672e-06,
+      "loss": 1.1402,
+      "step": 582
+    },
+    {
+      "epoch": 0.11653881712101147,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.974115682889164e-06,
+      "loss": 1.0779,
+      "step": 583
+    },
+    {
+      "epoch": 0.11673871217610754,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.97400848355097e-06,
+      "loss": 1.1407,
+      "step": 584
+    },
+    {
+      "epoch": 0.11693860723120361,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.973901063267848e-06,
+      "loss": 1.1509,
+      "step": 585
+    },
+    {
+      "epoch": 0.1171385022862997,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.973793422044573e-06,
+      "loss": 1.1145,
+      "step": 586
+    },
+    {
+      "epoch": 0.11733839734139577,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.973685559885927e-06,
+      "loss": 1.1153,
+      "step": 587
+    },
+    {
+      "epoch": 0.11753829239649184,
+      "grad_norm": 2.25,
+      "learning_rate": 9.973577476796697e-06,
+      "loss": 1.1684,
+      "step": 588
+    },
+    {
+      "epoch": 0.11773818745158791,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.973469172781688e-06,
+      "loss": 1.1161,
+      "step": 589
+    },
+    {
+      "epoch": 0.11793808250668399,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.973360647845708e-06,
+      "loss": 1.1282,
+      "step": 590
+    },
+    {
+      "epoch": 0.11813797756178007,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.973251901993582e-06,
+      "loss": 1.0917,
+      "step": 591
+    },
+    {
+      "epoch": 0.11833787261687614,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.973142935230135e-06,
+      "loss": 1.222,
+      "step": 592
+    },
+    {
+      "epoch": 0.11853776767197222,
+      "grad_norm": 2.25,
+      "learning_rate": 9.97303374756021e-06,
+      "loss": 1.1674,
+      "step": 593
+    },
+    {
+      "epoch": 0.11873766272706829,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.972924338988658e-06,
+      "loss": 1.095,
+      "step": 594
+    },
+    {
+      "epoch": 0.11893755778216436,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.972814709520339e-06,
+      "loss": 1.0656,
+      "step": 595
+    },
+    {
+      "epoch": 0.11913745283726043,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.97270485916012e-06,
+      "loss": 1.0735,
+      "step": 596
+    },
+    {
+      "epoch": 0.11933734789235652,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.972594787912884e-06,
+      "loss": 0.9428,
+      "step": 597
+    },
+    {
+      "epoch": 0.11953724294745259,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.972484495783518e-06,
+      "loss": 1.092,
+      "step": 598
+    },
+    {
+      "epoch": 0.11973713800254866,
+      "grad_norm": 2.125,
+      "learning_rate": 9.972373982776922e-06,
+      "loss": 1.0373,
+      "step": 599
+    },
+    {
+      "epoch": 0.11993703305764473,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.972263248898004e-06,
+      "loss": 1.1532,
+      "step": 600
+    },
+    {
+      "epoch": 0.1201369281127408,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.972152294151682e-06,
+      "loss": 1.0678,
+      "step": 601
+    },
+    {
+      "epoch": 0.12033682316783688,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.972041118542889e-06,
+      "loss": 1.0426,
+      "step": 602
+    },
+    {
+      "epoch": 0.12053671822293296,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.97192972207656e-06,
+      "loss": 1.1312,
+      "step": 603
+    },
+    {
+      "epoch": 0.12073661327802904,
+      "grad_norm": 2.125,
+      "learning_rate": 9.971818104757643e-06,
+      "loss": 1.0671,
+      "step": 604
+    },
+    {
+      "epoch": 0.12093650833312511,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.971706266591097e-06,
+      "loss": 1.1811,
+      "step": 605
+    },
+    {
+      "epoch": 0.12113640338822118,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.97159420758189e-06,
+      "loss": 1.1151,
+      "step": 606
+    },
+    {
+      "epoch": 0.12133629844331725,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.971481927734998e-06,
+      "loss": 1.0453,
+      "step": 607
+    },
+    {
+      "epoch": 0.12153619349841334,
+      "grad_norm": 2.125,
+      "learning_rate": 9.97136942705541e-06,
+      "loss": 1.0407,
+      "step": 608
+    },
+    {
+      "epoch": 0.12173608855350941,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.971256705548125e-06,
+      "loss": 1.0148,
+      "step": 609
+    },
+    {
+      "epoch": 0.12193598360860548,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.971143763218145e-06,
+      "loss": 1.1297,
+      "step": 610
+    },
+    {
+      "epoch": 0.12213587866370156,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.971030600070493e-06,
+      "loss": 1.0892,
+      "step": 611
+    },
+    {
+      "epoch": 0.12233577371879763,
+      "grad_norm": 2.25,
+      "learning_rate": 9.970917216110192e-06,
+      "loss": 1.1809,
+      "step": 612
+    },
+    {
+      "epoch": 0.1225356687738937,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.970803611342278e-06,
+      "loss": 1.1111,
+      "step": 613
+    },
+    {
+      "epoch": 0.12273556382898979,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.970689785771798e-06,
+      "loss": 1.0206,
+      "step": 614
+    },
+    {
+      "epoch": 0.12293545888408586,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.97057573940381e-06,
+      "loss": 1.2189,
+      "step": 615
+    },
+    {
+      "epoch": 0.12313535393918193,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.970461472243378e-06,
+      "loss": 1.114,
+      "step": 616
+    },
+    {
+      "epoch": 0.123335248994278,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.97034698429558e-06,
+      "loss": 1.0699,
+      "step": 617
+    },
+    {
+      "epoch": 0.12353514404937407,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.970232275565497e-06,
+      "loss": 1.1188,
+      "step": 618
+    },
+    {
+      "epoch": 0.12373503910447016,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.970117346058229e-06,
+      "loss": 1.138,
+      "step": 619
+    },
+    {
+      "epoch": 0.12393493415956623,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.970002195778879e-06,
+      "loss": 1.2651,
+      "step": 620
+    },
+    {
+      "epoch": 0.1241348292146623,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.969886824732561e-06,
+      "loss": 1.1851,
+      "step": 621
+    },
+    {
+      "epoch": 0.12433472426975838,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.969771232924404e-06,
+      "loss": 1.0908,
+      "step": 622
+    },
+    {
+      "epoch": 0.12453461932485445,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.969655420359537e-06,
+      "loss": 1.0961,
+      "step": 623
+    },
+    {
+      "epoch": 0.12473451437995052,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.969539387043106e-06,
+      "loss": 1.0848,
+      "step": 624
+    },
+    {
+      "epoch": 0.1249344094350466,
+      "grad_norm": 2.125,
+      "learning_rate": 9.969423132980269e-06,
+      "loss": 1.1229,
+      "step": 625
+    },
+    {
+      "epoch": 0.12513430449014268,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.969306658176185e-06,
+      "loss": 1.1251,
+      "step": 626
+    },
+    {
+      "epoch": 0.12533419954523875,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.969189962636032e-06,
+      "loss": 1.0888,
+      "step": 627
+    },
+    {
+      "epoch": 0.12553409460033482,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.96907304636499e-06,
+      "loss": 1.1758,
+      "step": 628
+    },
+    {
+      "epoch": 0.1257339896554309,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.968955909368256e-06,
+      "loss": 1.2216,
+      "step": 629
+    },
+    {
+      "epoch": 0.12593388471052697,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.96883855165103e-06,
+      "loss": 1.1126,
+      "step": 630
+    },
+    {
+      "epoch": 0.12613377976562304,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.968720973218525e-06,
+      "loss": 1.0975,
+      "step": 631
+    },
+    {
+      "epoch": 0.1263336748207191,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.968603174075967e-06,
+      "loss": 1.1615,
+      "step": 632
+    },
+    {
+      "epoch": 0.1265335698758152,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.968485154228584e-06,
+      "loss": 1.0805,
+      "step": 633
+    },
+    {
+      "epoch": 0.12673346493091128,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.968366913681624e-06,
+      "loss": 1.0741,
+      "step": 634
+    },
+    {
+      "epoch": 0.12693335998600735,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.968248452440335e-06,
+      "loss": 1.1183,
+      "step": 635
+    },
+    {
+      "epoch": 0.12713325504110343,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.96812977050998e-06,
+      "loss": 1.069,
+      "step": 636
+    },
+    {
+      "epoch": 0.1273331500961995,
+      "grad_norm": 2.125,
+      "learning_rate": 9.968010867895832e-06,
+      "loss": 1.1027,
+      "step": 637
+    },
+    {
+      "epoch": 0.12753304515129557,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.967891744603173e-06,
+      "loss": 1.1573,
+      "step": 638
+    },
+    {
+      "epoch": 0.12773294020639164,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.967772400637292e-06,
+      "loss": 1.1458,
+      "step": 639
+    },
+    {
+      "epoch": 0.12793283526148772,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.96765283600349e-06,
+      "loss": 1.1447,
+      "step": 640
+    },
+    {
+      "epoch": 0.1281327303165838,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.967533050707081e-06,
+      "loss": 1.1196,
+      "step": 641
+    },
+    {
+      "epoch": 0.12833262537167986,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.967413044753385e-06,
+      "loss": 1.1622,
+      "step": 642
+    },
+    {
+      "epoch": 0.12853252042677593,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.967292818147731e-06,
+      "loss": 1.1221,
+      "step": 643
+    },
+    {
+      "epoch": 0.128732415481872,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.967172370895462e-06,
+      "loss": 1.0577,
+      "step": 644
+    },
+    {
+      "epoch": 0.1289323105369681,
+      "grad_norm": 2.125,
+      "learning_rate": 9.967051703001926e-06,
+      "loss": 0.9921,
+      "step": 645
+    },
+    {
+      "epoch": 0.12913220559206418,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.966930814472484e-06,
+      "loss": 1.2149,
+      "step": 646
+    },
+    {
+      "epoch": 0.12933210064716025,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.966809705312506e-06,
+      "loss": 1.1856,
+      "step": 647
+    },
+    {
+      "epoch": 0.12953199570225632,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.96668837552737e-06,
+      "loss": 1.1351,
+      "step": 648
+    },
+    {
+      "epoch": 0.1297318907573524,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.966566825122467e-06,
+      "loss": 1.059,
+      "step": 649
+    },
+    {
+      "epoch": 0.12993178581244846,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.966445054103198e-06,
+      "loss": 1.1092,
+      "step": 650
+    },
+    {
+      "epoch": 0.13013168086754454,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.96632306247497e-06,
+      "loss": 1.126,
+      "step": 651
+    },
+    {
+      "epoch": 0.1303315759226406,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.9662008502432e-06,
+      "loss": 1.1809,
+      "step": 652
+    },
+    {
+      "epoch": 0.13053147097773668,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.96607841741332e-06,
+      "loss": 1.1616,
+      "step": 653
+    },
+    {
+      "epoch": 0.13073136603283275,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.965955763990765e-06,
+      "loss": 1.0585,
+      "step": 654
+    },
+    {
+      "epoch": 0.13093126108792882,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.965832889980987e-06,
+      "loss": 1.0355,
+      "step": 655
+    },
+    {
+      "epoch": 0.13113115614302492,
+      "grad_norm": 2.25,
+      "learning_rate": 9.965709795389441e-06,
+      "loss": 1.2128,
+      "step": 656
+    },
+    {
+      "epoch": 0.131331051198121,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.965586480221599e-06,
+      "loss": 1.1607,
+      "step": 657
+    },
+    {
+      "epoch": 0.13153094625321707,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.965462944482935e-06,
+      "loss": 1.2712,
+      "step": 658
+    },
+    {
+      "epoch": 0.13173084130831314,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.965339188178936e-06,
+      "loss": 1.143,
+      "step": 659
+    },
+    {
+      "epoch": 0.1319307363634092,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.965215211315103e-06,
+      "loss": 1.1531,
+      "step": 660
+    },
+    {
+      "epoch": 0.13213063141850528,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.965091013896937e-06,
+      "loss": 1.0669,
+      "step": 661
+    },
+    {
+      "epoch": 0.13233052647360136,
+      "grad_norm": 2.125,
+      "learning_rate": 9.964966595929961e-06,
+      "loss": 1.0308,
+      "step": 662
+    },
+    {
+      "epoch": 0.13253042152869743,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.9648419574197e-06,
+      "loss": 1.1347,
+      "step": 663
+    },
+    {
+      "epoch": 0.1327303165837935,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.964717098371687e-06,
+      "loss": 1.1258,
+      "step": 664
+    },
+    {
+      "epoch": 0.13293021163888957,
+      "grad_norm": 2.125,
+      "learning_rate": 9.964592018791473e-06,
+      "loss": 1.2204,
+      "step": 665
+    },
+    {
+      "epoch": 0.13313010669398564,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.964466718684609e-06,
+      "loss": 1.1216,
+      "step": 666
+    },
+    {
+      "epoch": 0.13333000174908174,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.964341198056665e-06,
+      "loss": 1.1338,
+      "step": 667
+    },
+    {
+      "epoch": 0.13352989680417782,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.964215456913215e-06,
+      "loss": 1.1389,
+      "step": 668
+    },
+    {
+      "epoch": 0.1337297918592739,
+      "grad_norm": 2.25,
+      "learning_rate": 9.964089495259846e-06,
+      "loss": 1.2297,
+      "step": 669
+    },
+    {
+      "epoch": 0.13392968691436996,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.96396331310215e-06,
+      "loss": 1.1715,
+      "step": 670
+    },
+    {
+      "epoch": 0.13412958196946603,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.963836910445735e-06,
+      "loss": 1.0458,
+      "step": 671
+    },
+    {
+      "epoch": 0.1343294770245621,
+      "grad_norm": 2.25,
+      "learning_rate": 9.963710287296213e-06,
+      "loss": 1.1225,
+      "step": 672
+    },
+    {
+      "epoch": 0.13452937207965818,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.963583443659213e-06,
+      "loss": 1.1095,
+      "step": 673
+    },
+    {
+      "epoch": 0.13472926713475425,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.963456379540364e-06,
+      "loss": 1.1456,
+      "step": 674
+    },
+    {
+      "epoch": 0.13492916218985032,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.963329094945313e-06,
+      "loss": 1.1126,
+      "step": 675
+    },
+    {
+      "epoch": 0.1351290572449464,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.963201589879715e-06,
+      "loss": 1.0257,
+      "step": 676
+    },
+    {
+      "epoch": 0.13532895230004247,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.963073864349232e-06,
+      "loss": 1.0276,
+      "step": 677
+    },
+    {
+      "epoch": 0.13552884735513857,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.962945918359537e-06,
+      "loss": 1.031,
+      "step": 678
+    },
+    {
+      "epoch": 0.13572874241023464,
+      "grad_norm": 2.125,
+      "learning_rate": 9.962817751916316e-06,
+      "loss": 1.0606,
+      "step": 679
+    },
+    {
+      "epoch": 0.1359286374653307,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.962689365025259e-06,
+      "loss": 1.0507,
+      "step": 680
+    },
+    {
+      "epoch": 0.13612853252042678,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.962560757692072e-06,
+      "loss": 1.2604,
+      "step": 681
+    },
+    {
+      "epoch": 0.13632842757552285,
+      "grad_norm": 2.125,
+      "learning_rate": 9.962431929922464e-06,
+      "loss": 1.1347,
+      "step": 682
+    },
+    {
+      "epoch": 0.13652832263061893,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.962302881722162e-06,
+      "loss": 1.0511,
+      "step": 683
+    },
+    {
+      "epoch": 0.136728217685715,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.962173613096895e-06,
+      "loss": 1.0503,
+      "step": 684
+    },
+    {
+      "epoch": 0.13692811274081107,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.962044124052406e-06,
+      "loss": 1.0673,
+      "step": 685
+    },
+    {
+      "epoch": 0.13712800779590714,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.961914414594447e-06,
+      "loss": 1.2253,
+      "step": 686
+    },
+    {
+      "epoch": 0.13732790285100321,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.96178448472878e-06,
+      "loss": 1.0849,
+      "step": 687
+    },
+    {
+      "epoch": 0.13752779790609929,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.961654334461175e-06,
+      "loss": 1.0873,
+      "step": 688
+    },
+    {
+      "epoch": 0.13772769296119539,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.961523963797415e-06,
+      "loss": 1.1177,
+      "step": 689
+    },
+    {
+      "epoch": 0.13792758801629146,
+      "grad_norm": 2.25,
+      "learning_rate": 9.961393372743291e-06,
+      "loss": 1.1586,
+      "step": 690
+    },
+    {
+      "epoch": 0.13812748307138753,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.961262561304604e-06,
+      "loss": 1.1439,
+      "step": 691
+    },
+    {
+      "epoch": 0.1383273781264836,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.961131529487161e-06,
+      "loss": 1.143,
+      "step": 692
+    },
+    {
+      "epoch": 0.13852727318157967,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.961000277296788e-06,
+      "loss": 1.121,
+      "step": 693
+    },
+    {
+      "epoch": 0.13872716823667575,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.960868804739312e-06,
+      "loss": 1.1314,
+      "step": 694
+    },
+    {
+      "epoch": 0.13892706329177182,
+      "grad_norm": 2.25,
+      "learning_rate": 9.960737111820572e-06,
+      "loss": 1.1114,
+      "step": 695
+    },
+    {
+      "epoch": 0.1391269583468679,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.96060519854642e-06,
+      "loss": 1.22,
+      "step": 696
+    },
+    {
+      "epoch": 0.13932685340196396,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.960473064922716e-06,
+      "loss": 1.2051,
+      "step": 697
+    },
+    {
+      "epoch": 0.13952674845706003,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.960340710955327e-06,
+      "loss": 1.0331,
+      "step": 698
+    },
+    {
+      "epoch": 0.1397266435121561,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.960208136650137e-06,
+      "loss": 1.1898,
+      "step": 699
+    },
+    {
+      "epoch": 0.13992653856725218,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.960075342013027e-06,
+      "loss": 1.1642,
+      "step": 700
+    },
+    {
+      "epoch": 0.14012643362234828,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.959942327049901e-06,
+      "loss": 1.1635,
+      "step": 701
+    },
+    {
+      "epoch": 0.14032632867744435,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.959809091766667e-06,
+      "loss": 1.1644,
+      "step": 702
+    },
+    {
+      "epoch": 0.14052622373254042,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.959675636169242e-06,
+      "loss": 1.1439,
+      "step": 703
+    },
+    {
+      "epoch": 0.1407261187876365,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.959541960263557e-06,
+      "loss": 1.0681,
+      "step": 704
+    },
+    {
+      "epoch": 0.14092601384273257,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.959408064055547e-06,
+      "loss": 1.1353,
+      "step": 705
+    },
+    {
+      "epoch": 0.14112590889782864,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.95927394755116e-06,
+      "loss": 1.1704,
+      "step": 706
+    },
+    {
+      "epoch": 0.1413258039529247,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.959139610756353e-06,
+      "loss": 1.2734,
+      "step": 707
+    },
+    {
+      "epoch": 0.14152569900802078,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.959005053677096e-06,
+      "loss": 1.1302,
+      "step": 708
+    },
+    {
+      "epoch": 0.14172559406311686,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.958870276319364e-06,
+      "loss": 1.1517,
+      "step": 709
+    },
+    {
+      "epoch": 0.14192548911821293,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.958735278689143e-06,
+      "loss": 1.0484,
+      "step": 710
+    },
+    {
+      "epoch": 0.142125384173309,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.958600060792429e-06,
+      "loss": 1.1174,
+      "step": 711
+    },
+    {
+      "epoch": 0.1423252792284051,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.958464622635233e-06,
+      "loss": 1.1346,
+      "step": 712
+    },
+    {
+      "epoch": 0.14252517428350117,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.958328964223566e-06,
+      "loss": 1.1224,
+      "step": 713
+    },
+    {
+      "epoch": 0.14272506933859724,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.958193085563456e-06,
+      "loss": 1.1046,
+      "step": 714
+    },
+    {
+      "epoch": 0.14292496439369332,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.95805698666094e-06,
+      "loss": 1.0429,
+      "step": 715
+    },
+    {
+      "epoch": 0.1431248594487894,
+      "grad_norm": 2.25,
+      "learning_rate": 9.957920667522063e-06,
+      "loss": 1.2122,
+      "step": 716
+    },
+    {
+      "epoch": 0.14332475450388546,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.957784128152877e-06,
+      "loss": 1.1542,
+      "step": 717
+    },
+    {
+      "epoch": 0.14352464955898153,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.957647368559451e-06,
+      "loss": 1.1105,
+      "step": 718
+    },
+    {
+      "epoch": 0.1437245446140776,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.957510388747858e-06,
+      "loss": 1.2298,
+      "step": 719
+    },
+    {
+      "epoch": 0.14392443966917368,
+      "grad_norm": 2.125,
+      "learning_rate": 9.957373188724184e-06,
+      "loss": 1.0803,
+      "step": 720
+    },
+    {
+      "epoch": 0.14412433472426975,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.95723576849452e-06,
+      "loss": 1.0925,
+      "step": 721
+    },
+    {
+      "epoch": 0.14432422977936582,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.957098128064974e-06,
+      "loss": 1.1767,
+      "step": 722
+    },
+    {
+      "epoch": 0.14452412483446192,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.95696026744166e-06,
+      "loss": 1.0755,
+      "step": 723
+    },
+    {
+      "epoch": 0.144724019889558,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.956822186630697e-06,
+      "loss": 1.1333,
+      "step": 724
+    },
+    {
+      "epoch": 0.14492391494465406,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.956683885638225e-06,
+      "loss": 1.1836,
+      "step": 725
+    },
+    {
+      "epoch": 0.14512380999975014,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.956545364470383e-06,
+      "loss": 1.1527,
+      "step": 726
+    },
+    {
+      "epoch": 0.1453237050548462,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.956406623133326e-06,
+      "loss": 1.0995,
+      "step": 727
+    },
+    {
+      "epoch": 0.14552360010994228,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.956267661633215e-06,
+      "loss": 1.1751,
+      "step": 728
+    },
+    {
+      "epoch": 0.14572349516503835,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.956128479976223e-06,
+      "loss": 1.2085,
+      "step": 729
+    },
+    {
+      "epoch": 0.14592339022013442,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.955989078168535e-06,
+      "loss": 1.1981,
+      "step": 730
+    },
+    {
+      "epoch": 0.1461232852752305,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.955849456216343e-06,
+      "loss": 1.109,
+      "step": 731
+    },
+    {
+      "epoch": 0.14632318033032657,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.955709614125844e-06,
+      "loss": 1.1483,
+      "step": 732
+    },
+    {
+      "epoch": 0.14652307538542264,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.955569551903255e-06,
+      "loss": 1.1024,
+      "step": 733
+    },
+    {
+      "epoch": 0.14672297044051874,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.955429269554795e-06,
+      "loss": 1.0162,
+      "step": 734
+    },
+    {
+      "epoch": 0.1469228654956148,
+      "grad_norm": 2.125,
+      "learning_rate": 9.955288767086698e-06,
+      "loss": 1.1791,
+      "step": 735
+    },
+    {
+      "epoch": 0.14712276055071088,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.9551480445052e-06,
+      "loss": 1.156,
+      "step": 736
+    },
+    {
+      "epoch": 0.14732265560580696,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.955007101816558e-06,
+      "loss": 1.1466,
+      "step": 737
+    },
+    {
+      "epoch": 0.14752255066090303,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.954865939027028e-06,
+      "loss": 1.0374,
+      "step": 738
+    },
+    {
+      "epoch": 0.1477224457159991,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.954724556142884e-06,
+      "loss": 1.0774,
+      "step": 739
+    },
+    {
+      "epoch": 0.14792234077109517,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.954582953170403e-06,
+      "loss": 1.0355,
+      "step": 740
+    },
+    {
+      "epoch": 0.14812223582619125,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.954441130115876e-06,
+      "loss": 1.1776,
+      "step": 741
+    },
+    {
+      "epoch": 0.14832213088128732,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.954299086985604e-06,
+      "loss": 1.1553,
+      "step": 742
+    },
+    {
+      "epoch": 0.1485220259363834,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.954156823785896e-06,
+      "loss": 1.0897,
+      "step": 743
+    },
+    {
+      "epoch": 0.14872192099147946,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.95401434052307e-06,
+      "loss": 1.141,
+      "step": 744
+    },
+    {
+      "epoch": 0.14892181604657556,
+      "grad_norm": 2.125,
+      "learning_rate": 9.953871637203456e-06,
+      "loss": 1.0888,
+      "step": 745
+    },
+    {
+      "epoch": 0.14912171110167163,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.953728713833395e-06,
+      "loss": 1.0699,
+      "step": 746
+    },
+    {
+      "epoch": 0.1493216061567677,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.95358557041923e-06,
+      "loss": 1.0551,
+      "step": 747
+    },
+    {
+      "epoch": 0.14952150121186378,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.953442206967327e-06,
+      "loss": 1.0822,
+      "step": 748
+    },
+    {
+      "epoch": 0.14972139626695985,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.953298623484049e-06,
+      "loss": 1.0417,
+      "step": 749
+    },
+    {
+      "epoch": 0.14992129132205592,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.953154819975773e-06,
+      "loss": 1.2122,
+      "step": 750
+    },
+    {
+      "epoch": 0.150121186377152,
+      "grad_norm": 2.125,
+      "learning_rate": 9.95301079644889e-06,
+      "loss": 1.0791,
+      "step": 751
+    },
+    {
+      "epoch": 0.15032108143224807,
+      "grad_norm": 2.25,
+      "learning_rate": 9.952866552909797e-06,
+      "loss": 1.1611,
+      "step": 752
+    },
+    {
+      "epoch": 0.15052097648734414,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.9527220893649e-06,
+      "loss": 1.1831,
+      "step": 753
+    },
+    {
+      "epoch": 0.1507208715424402,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.952577405820618e-06,
+      "loss": 1.1212,
+      "step": 754
+    },
+    {
+      "epoch": 0.15092076659753628,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.952432502283378e-06,
+      "loss": 1.0424,
+      "step": 755
+    },
+    {
+      "epoch": 0.15112066165263235,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.952287378759613e-06,
+      "loss": 1.0381,
+      "step": 756
+    },
+    {
+      "epoch": 0.15132055670772845,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.95214203525577e-06,
+      "loss": 1.1073,
+      "step": 757
+    },
+    {
+      "epoch": 0.15152045176282453,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.951996471778308e-06,
+      "loss": 1.092,
+      "step": 758
+    },
+    {
+      "epoch": 0.1517203468179206,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.951850688333693e-06,
+      "loss": 1.1301,
+      "step": 759
+    },
+    {
+      "epoch": 0.15192024187301667,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.951704684928398e-06,
+      "loss": 1.123,
+      "step": 760
+    },
+    {
+      "epoch": 0.15212013692811274,
+      "grad_norm": 2.25,
+      "learning_rate": 9.95155846156891e-06,
+      "loss": 1.0566,
+      "step": 761
+    },
+    {
+      "epoch": 0.15232003198320881,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.951412018261724e-06,
+      "loss": 1.0931,
+      "step": 762
+    },
+    {
+      "epoch": 0.1525199270383049,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.951265355013345e-06,
+      "loss": 0.9994,
+      "step": 763
+    },
+    {
+      "epoch": 0.15271982209340096,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.951118471830287e-06,
+      "loss": 1.0814,
+      "step": 764
+    },
+    {
+      "epoch": 0.15291971714849703,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.950971368719077e-06,
+      "loss": 1.1147,
+      "step": 765
+    },
+    {
+      "epoch": 0.1531196122035931,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.950824045686247e-06,
+      "loss": 1.0999,
+      "step": 766
+    },
+    {
+      "epoch": 0.15331950725868917,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.95067650273834e-06,
+      "loss": 1.1821,
+      "step": 767
+    },
+    {
+      "epoch": 0.15351940231378527,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.950528739881915e-06,
+      "loss": 1.1256,
+      "step": 768
+    },
+    {
+      "epoch": 0.15371929736888135,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.95038075712353e-06,
+      "loss": 1.1463,
+      "step": 769
+    },
+    {
+      "epoch": 0.15391919242397742,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.950232554469759e-06,
+      "loss": 1.0449,
+      "step": 770
+    },
+    {
+      "epoch": 0.1541190874790735,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.950084131927188e-06,
+      "loss": 1.0059,
+      "step": 771
+    },
+    {
+      "epoch": 0.15431898253416956,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.949935489502409e-06,
+      "loss": 1.0697,
+      "step": 772
+    },
+    {
+      "epoch": 0.15451887758926564,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.949786627202023e-06,
+      "loss": 1.1426,
+      "step": 773
+    },
+    {
+      "epoch": 0.1547187726443617,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.949637545032644e-06,
+      "loss": 1.2014,
+      "step": 774
+    },
+    {
+      "epoch": 0.15491866769945778,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.949488243000895e-06,
+      "loss": 1.0977,
+      "step": 775
+    },
+    {
+      "epoch": 0.15511856275455385,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.949338721113406e-06,
+      "loss": 1.0963,
+      "step": 776
+    },
+    {
+      "epoch": 0.15531845780964992,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.94918897937682e-06,
+      "loss": 1.0902,
+      "step": 777
+    },
+    {
+      "epoch": 0.155518352864746,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.949039017797788e-06,
+      "loss": 1.0116,
+      "step": 778
+    },
+    {
+      "epoch": 0.1557182479198421,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.948888836382971e-06,
+      "loss": 1.1572,
+      "step": 779
+    },
+    {
+      "epoch": 0.15591814297493817,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.948738435139042e-06,
+      "loss": 1.0033,
+      "step": 780
+    },
+    {
+      "epoch": 0.15611803803003424,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.948587814072679e-06,
+      "loss": 0.9917,
+      "step": 781
+    },
+    {
+      "epoch": 0.1563179330851303,
+      "grad_norm": 2.125,
+      "learning_rate": 9.948436973190574e-06,
+      "loss": 1.1254,
+      "step": 782
+    },
+    {
+      "epoch": 0.15651782814022638,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.948285912499427e-06,
+      "loss": 1.04,
+      "step": 783
+    },
+    {
+      "epoch": 0.15671772319532246,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.948134632005948e-06,
+      "loss": 1.1461,
+      "step": 784
+    },
+    {
+      "epoch": 0.15691761825041853,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.947983131716858e-06,
+      "loss": 1.1382,
+      "step": 785
+    },
+    {
+      "epoch": 0.1571175133055146,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.947831411638884e-06,
+      "loss": 1.0689,
+      "step": 786
+    },
+    {
+      "epoch": 0.15731740836061067,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.947679471778768e-06,
+      "loss": 1.0443,
+      "step": 787
+    },
+    {
+      "epoch": 0.15751730341570674,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.947527312143259e-06,
+      "loss": 1.0321,
+      "step": 788
+    },
+    {
+      "epoch": 0.15771719847080282,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.947374932739115e-06,
+      "loss": 1.233,
+      "step": 789
+    },
+    {
+      "epoch": 0.15791709352589892,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.947222333573105e-06,
+      "loss": 1.1441,
+      "step": 790
+    },
+    {
+      "epoch": 0.158116988580995,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.947069514652006e-06,
+      "loss": 1.055,
+      "step": 791
+    },
+    {
+      "epoch": 0.15831688363609106,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.94691647598261e-06,
+      "loss": 1.1694,
+      "step": 792
+    },
+    {
+      "epoch": 0.15851677869118713,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.946763217571712e-06,
+      "loss": 0.9969,
+      "step": 793
+    },
+    {
+      "epoch": 0.1587166737462832,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.946609739426119e-06,
+      "loss": 1.1176,
+      "step": 794
+    },
+    {
+      "epoch": 0.15891656880137928,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.94645604155265e-06,
+      "loss": 1.1539,
+      "step": 795
+    },
+    {
+      "epoch": 0.15911646385647535,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.94630212395813e-06,
+      "loss": 1.1059,
+      "step": 796
+    },
+    {
+      "epoch": 0.15931635891157142,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.9461479866494e-06,
+      "loss": 1.1263,
+      "step": 797
+    },
+    {
+      "epoch": 0.1595162539666675,
+      "grad_norm": 2.125,
+      "learning_rate": 9.945993629633305e-06,
+      "loss": 1.1279,
+      "step": 798
+    },
+    {
+      "epoch": 0.15971614902176356,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.945839052916702e-06,
+      "loss": 1.0915,
+      "step": 799
+    },
+    {
+      "epoch": 0.15991604407685964,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.945684256506454e-06,
+      "loss": 1.2039,
+      "step": 800
+    },
+    {
+      "epoch": 0.16011593913195574,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.945529240409442e-06,
+      "loss": 1.1241,
+      "step": 801
+    },
+    {
+      "epoch": 0.1603158341870518,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.945374004632547e-06,
+      "loss": 1.1659,
+      "step": 802
+    },
+    {
+      "epoch": 0.16051572924214788,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.945218549182668e-06,
+      "loss": 1.1306,
+      "step": 803
+    },
+    {
+      "epoch": 0.16071562429724395,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.945062874066709e-06,
+      "loss": 1.0898,
+      "step": 804
+    },
+    {
+      "epoch": 0.16091551935234003,
+      "grad_norm": 2.25,
+      "learning_rate": 9.944906979291587e-06,
+      "loss": 1.1502,
+      "step": 805
+    },
+    {
+      "epoch": 0.1611154144074361,
+      "grad_norm": 2.25,
+      "learning_rate": 9.944750864864224e-06,
+      "loss": 1.1441,
+      "step": 806
+    },
+    {
+      "epoch": 0.16131530946253217,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.944594530791553e-06,
+      "loss": 1.0863,
+      "step": 807
+    },
+    {
+      "epoch": 0.16151520451762824,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.944437977080525e-06,
+      "loss": 1.1074,
+      "step": 808
+    },
+    {
+      "epoch": 0.1617150995727243,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.944281203738087e-06,
+      "loss": 1.0624,
+      "step": 809
+    },
+    {
+      "epoch": 0.16191499462782039,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.944124210771209e-06,
+      "loss": 1.2287,
+      "step": 810
+    },
+    {
+      "epoch": 0.16211488968291646,
+      "grad_norm": 2.25,
+      "learning_rate": 9.94396699818686e-06,
+      "loss": 1.1621,
+      "step": 811
+    },
+    {
+      "epoch": 0.16231478473801253,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.943809565992024e-06,
+      "loss": 1.1585,
+      "step": 812
+    },
+    {
+      "epoch": 0.16251467979310863,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.943651914193697e-06,
+      "loss": 1.0682,
+      "step": 813
+    },
+    {
+      "epoch": 0.1627145748482047,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.943494042798878e-06,
+      "loss": 1.101,
+      "step": 814
+    },
+    {
+      "epoch": 0.16291446990330077,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.943335951814583e-06,
+      "loss": 1.1353,
+      "step": 815
+    },
+    {
+      "epoch": 0.16311436495839685,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.943177641247833e-06,
+      "loss": 1.0897,
+      "step": 816
+    },
+    {
+      "epoch": 0.16331426001349292,
+      "grad_norm": 5.25,
+      "learning_rate": 9.94301911110566e-06,
+      "loss": 0.9894,
+      "step": 817
+    },
+    {
+      "epoch": 0.163514155068589,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.942860361395106e-06,
+      "loss": 1.095,
+      "step": 818
+    },
+    {
+      "epoch": 0.16371405012368506,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.942701392123222e-06,
+      "loss": 1.1018,
+      "step": 819
+    },
+    {
+      "epoch": 0.16391394517878113,
+      "grad_norm": 2.125,
+      "learning_rate": 9.942542203297072e-06,
+      "loss": 1.1963,
+      "step": 820
+    },
+    {
+      "epoch": 0.1641138402338772,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.942382794923723e-06,
+      "loss": 1.0891,
+      "step": 821
+    },
+    {
+      "epoch": 0.16431373528897328,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.94222316701026e-06,
+      "loss": 1.1304,
+      "step": 822
+    },
+    {
+      "epoch": 0.16451363034406935,
+      "grad_norm": 2.125,
+      "learning_rate": 9.942063319563769e-06,
+      "loss": 1.0862,
+      "step": 823
+    },
+    {
+      "epoch": 0.16471352539916545,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.941903252591356e-06,
+      "loss": 1.1904,
+      "step": 824
+    },
+    {
+      "epoch": 0.16491342045426152,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.941742966100128e-06,
+      "loss": 1.1039,
+      "step": 825
+    },
+    {
+      "epoch": 0.1651133155093576,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.941582460097203e-06,
+      "loss": 1.0461,
+      "step": 826
+    },
+    {
+      "epoch": 0.16531321056445367,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.941421734589715e-06,
+      "loss": 1.0681,
+      "step": 827
+    },
+    {
+      "epoch": 0.16551310561954974,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.9412607895848e-06,
+      "loss": 1.0501,
+      "step": 828
+    },
+    {
+      "epoch": 0.1657130006746458,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.94109962508961e-06,
+      "loss": 1.1685,
+      "step": 829
+    },
+    {
+      "epoch": 0.16591289572974188,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.9409382411113e-06,
+      "loss": 1.0765,
+      "step": 830
+    },
+    {
+      "epoch": 0.16611279078483795,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.940776637657044e-06,
+      "loss": 1.0488,
+      "step": 831
+    },
+    {
+      "epoch": 0.16631268583993403,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.940614814734015e-06,
+      "loss": 1.0233,
+      "step": 832
+    },
+    {
+      "epoch": 0.1665125808950301,
+      "grad_norm": 2.125,
+      "learning_rate": 9.940452772349405e-06,
+      "loss": 1.1514,
+      "step": 833
+    },
+    {
+      "epoch": 0.16671247595012617,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.940290510510411e-06,
+      "loss": 1.0156,
+      "step": 834
+    },
+    {
+      "epoch": 0.16691237100522227,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.940128029224239e-06,
+      "loss": 1.0412,
+      "step": 835
+    },
+    {
+      "epoch": 0.16711226606031834,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.939965328498107e-06,
+      "loss": 1.1049,
+      "step": 836
+    },
+    {
+      "epoch": 0.16731216111541442,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.939802408339244e-06,
+      "loss": 1.173,
+      "step": 837
+    },
+    {
+      "epoch": 0.1675120561705105,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.939639268754886e-06,
+      "loss": 1.1309,
+      "step": 838
+    },
+    {
+      "epoch": 0.16771195122560656,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.939475909752278e-06,
+      "loss": 1.0505,
+      "step": 839
+    },
+    {
+      "epoch": 0.16791184628070263,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.939312331338678e-06,
+      "loss": 1.0639,
+      "step": 840
+    },
+    {
+      "epoch": 0.1681117413357987,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.939148533521353e-06,
+      "loss": 1.0412,
+      "step": 841
+    },
+    {
+      "epoch": 0.16831163639089478,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.938984516307575e-06,
+      "loss": 1.063,
+      "step": 842
+    },
+    {
+      "epoch": 0.16851153144599085,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.938820279704635e-06,
+      "loss": 1.1225,
+      "step": 843
+    },
+    {
+      "epoch": 0.16871142650108692,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.938655823719823e-06,
+      "loss": 1.1079,
+      "step": 844
+    },
+    {
+      "epoch": 0.168911321556183,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.938491148360448e-06,
+      "loss": 1.1533,
+      "step": 845
+    },
+    {
+      "epoch": 0.1691112166112791,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.938326253633825e-06,
+      "loss": 1.1815,
+      "step": 846
+    },
+    {
+      "epoch": 0.16931111166637516,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.938161139547276e-06,
+      "loss": 1.0423,
+      "step": 847
+    },
+    {
+      "epoch": 0.16951100672147124,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.937995806108135e-06,
+      "loss": 1.1081,
+      "step": 848
+    },
+    {
+      "epoch": 0.1697109017765673,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.93783025332375e-06,
+      "loss": 1.0819,
+      "step": 849
+    },
+    {
+      "epoch": 0.16991079683166338,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.937664481201472e-06,
+      "loss": 1.0859,
+      "step": 850
+    },
+    {
+      "epoch": 0.17011069188675945,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.937498489748665e-06,
+      "loss": 1.1142,
+      "step": 851
+    },
+    {
+      "epoch": 0.17031058694185552,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.937332278972703e-06,
+      "loss": 1.1683,
+      "step": 852
+    },
+    {
+      "epoch": 0.1705104819969516,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.937165848880968e-06,
+      "loss": 1.0265,
+      "step": 853
+    },
+    {
+      "epoch": 0.17071037705204767,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.936999199480854e-06,
+      "loss": 1.0466,
+      "step": 854
+    },
+    {
+      "epoch": 0.17091027210714374,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.936832330779761e-06,
+      "loss": 1.0995,
+      "step": 855
+    },
+    {
+      "epoch": 0.1711101671622398,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.936665242785105e-06,
+      "loss": 1.256,
+      "step": 856
+    },
+    {
+      "epoch": 0.1713100622173359,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.936497935504306e-06,
+      "loss": 1.0735,
+      "step": 857
+    },
+    {
+      "epoch": 0.17150995727243198,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.936330408944794e-06,
+      "loss": 1.1574,
+      "step": 858
+    },
+    {
+      "epoch": 0.17170985232752806,
+      "grad_norm": 2.125,
+      "learning_rate": 9.936162663114014e-06,
+      "loss": 1.0946,
+      "step": 859
+    },
+    {
+      "epoch": 0.17190974738262413,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.935994698019416e-06,
+      "loss": 1.1149,
+      "step": 860
+    },
+    {
+      "epoch": 0.1721096424377202,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.93582651366846e-06,
+      "loss": 1.0987,
+      "step": 861
+    },
+    {
+      "epoch": 0.17230953749281627,
+      "grad_norm": 2.125,
+      "learning_rate": 9.935658110068618e-06,
+      "loss": 1.0275,
+      "step": 862
+    },
+    {
+      "epoch": 0.17250943254791234,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.93548948722737e-06,
+      "loss": 1.0396,
+      "step": 863
+    },
+    {
+      "epoch": 0.17270932760300842,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.935320645152205e-06,
+      "loss": 1.0945,
+      "step": 864
+    },
+    {
+      "epoch": 0.1729092226581045,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.935151583850624e-06,
+      "loss": 1.1701,
+      "step": 865
+    },
+    {
+      "epoch": 0.17310911771320056,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.934982303330138e-06,
+      "loss": 1.0033,
+      "step": 866
+    },
+    {
+      "epoch": 0.17330901276829663,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.934812803598265e-06,
+      "loss": 1.027,
+      "step": 867
+    },
+    {
+      "epoch": 0.1735089078233927,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.934643084662533e-06,
+      "loss": 1.1306,
+      "step": 868
+    },
+    {
+      "epoch": 0.1737088028784888,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.934473146530483e-06,
+      "loss": 1.1985,
+      "step": 869
+    },
+    {
+      "epoch": 0.17390869793358488,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.934302989209663e-06,
+      "loss": 1.2504,
+      "step": 870
+    },
+    {
+      "epoch": 0.17410859298868095,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.934132612707631e-06,
+      "loss": 1.0985,
+      "step": 871
+    },
+    {
+      "epoch": 0.17430848804377702,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.933962017031957e-06,
+      "loss": 1.1471,
+      "step": 872
+    },
+    {
+      "epoch": 0.1745083830988731,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.933791202190215e-06,
+      "loss": 1.0599,
+      "step": 873
+    },
+    {
+      "epoch": 0.17470827815396917,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.933620168189995e-06,
+      "loss": 1.162,
+      "step": 874
+    },
+    {
+      "epoch": 0.17490817320906524,
+      "grad_norm": 2.125,
+      "learning_rate": 9.933448915038895e-06,
+      "loss": 1.083,
+      "step": 875
+    },
+    {
+      "epoch": 0.1751080682641613,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.93327744274452e-06,
+      "loss": 1.116,
+      "step": 876
+    },
+    {
+      "epoch": 0.17530796331925738,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.933105751314489e-06,
+      "loss": 0.98,
+      "step": 877
+    },
+    {
+      "epoch": 0.17550785837435345,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.932933840756428e-06,
+      "loss": 1.0147,
+      "step": 878
+    },
+    {
+      "epoch": 0.17570775342944953,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.93276171107797e-06,
+      "loss": 1.0976,
+      "step": 879
+    },
+    {
+      "epoch": 0.17590764848454563,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.932589362286766e-06,
+      "loss": 1.0766,
+      "step": 880
+    },
+    {
+      "epoch": 0.1761075435396417,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.932416794390467e-06,
+      "loss": 1.0134,
+      "step": 881
+    },
+    {
+      "epoch": 0.17630743859473777,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.932244007396742e-06,
+      "loss": 1.0746,
+      "step": 882
+    },
+    {
+      "epoch": 0.17650733364983384,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.932071001313265e-06,
+      "loss": 1.066,
+      "step": 883
+    },
+    {
+      "epoch": 0.17670722870492991,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.931897776147724e-06,
+      "loss": 1.1709,
+      "step": 884
+    },
+    {
+      "epoch": 0.17690712376002599,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.931724331907806e-06,
+      "loss": 1.15,
+      "step": 885
+    },
+    {
+      "epoch": 0.17710701881512206,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.931550668601222e-06,
+      "loss": 1.1492,
+      "step": 886
+    },
+    {
+      "epoch": 0.17730691387021813,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.931376786235684e-06,
+      "loss": 1.1116,
+      "step": 887
+    },
+    {
+      "epoch": 0.1775068089253142,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.931202684818914e-06,
+      "loss": 1.1235,
+      "step": 888
+    },
+    {
+      "epoch": 0.17770670398041027,
+      "grad_norm": 2.125,
+      "learning_rate": 9.931028364358651e-06,
+      "loss": 1.0914,
+      "step": 889
+    },
+    {
+      "epoch": 0.17790659903550635,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.930853824862632e-06,
+      "loss": 1.1155,
+      "step": 890
+    },
+    {
+      "epoch": 0.17810649409060245,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.930679066338613e-06,
+      "loss": 1.0528,
+      "step": 891
+    },
+    {
+      "epoch": 0.17830638914569852,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.930504088794356e-06,
+      "loss": 1.1009,
+      "step": 892
+    },
+    {
+      "epoch": 0.1785062842007946,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.930328892237636e-06,
+      "loss": 1.1056,
+      "step": 893
+    },
+    {
+      "epoch": 0.17870617925589066,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.930153476676231e-06,
+      "loss": 1.0138,
+      "step": 894
+    },
+    {
+      "epoch": 0.17890607431098673,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.929977842117935e-06,
+      "loss": 1.0252,
+      "step": 895
+    },
+    {
+      "epoch": 0.1791059693660828,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.92980198857055e-06,
+      "loss": 1.0429,
+      "step": 896
+    },
+    {
+      "epoch": 0.17930586442117888,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.92962591604189e-06,
+      "loss": 1.0689,
+      "step": 897
+    },
+    {
+      "epoch": 0.17950575947627495,
+      "grad_norm": 2.125,
+      "learning_rate": 9.929449624539772e-06,
+      "loss": 1.0321,
+      "step": 898
+    },
+    {
+      "epoch": 0.17970565453137102,
+      "grad_norm": 2.125,
+      "learning_rate": 9.929273114072027e-06,
+      "loss": 1.0742,
+      "step": 899
+    },
+    {
+      "epoch": 0.1799055495864671,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.929096384646498e-06,
+      "loss": 1.1502,
+      "step": 900
+    },
+    {
+      "epoch": 0.18010544464156317,
+      "grad_norm": 2.125,
+      "learning_rate": 9.928919436271032e-06,
+      "loss": 1.1121,
+      "step": 901
+    },
+    {
+      "epoch": 0.18030533969665927,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.928742268953493e-06,
+      "loss": 1.1333,
+      "step": 902
+    },
+    {
+      "epoch": 0.18050523475175534,
+      "grad_norm": 2.125,
+      "learning_rate": 9.928564882701749e-06,
+      "loss": 1.1531,
+      "step": 903
+    },
+    {
+      "epoch": 0.1807051298068514,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.928387277523676e-06,
+      "loss": 1.1334,
+      "step": 904
+    },
+    {
+      "epoch": 0.18090502486194748,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.92820945342717e-06,
+      "loss": 1.2134,
+      "step": 905
+    },
+    {
+      "epoch": 0.18110491991704356,
+      "grad_norm": 2.25,
+      "learning_rate": 9.928031410420125e-06,
+      "loss": 1.0337,
+      "step": 906
+    },
+    {
+      "epoch": 0.18130481497213963,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.927853148510451e-06,
+      "loss": 1.0733,
+      "step": 907
+    },
+    {
+      "epoch": 0.1815047100272357,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.92767466770607e-06,
+      "loss": 1.065,
+      "step": 908
+    },
+    {
+      "epoch": 0.18170460508233177,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.927495968014903e-06,
+      "loss": 1.1061,
+      "step": 909
+    },
+    {
+      "epoch": 0.18190450013742784,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.92731704944489e-06,
+      "loss": 1.1142,
+      "step": 910
+    },
+    {
+      "epoch": 0.18210439519252392,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.92713791200398e-06,
+      "loss": 1.0194,
+      "step": 911
+    },
+    {
+      "epoch": 0.18230429024762,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.926958555700134e-06,
+      "loss": 1.1316,
+      "step": 912
+    },
+    {
+      "epoch": 0.1825041853027161,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.926778980541314e-06,
+      "loss": 1.0904,
+      "step": 913
+    },
+    {
+      "epoch": 0.18270408035781216,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.926599186535496e-06,
+      "loss": 1.1301,
+      "step": 914
+    },
+    {
+      "epoch": 0.18290397541290823,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.92641917369067e-06,
+      "loss": 1.0447,
+      "step": 915
+    },
+    {
+      "epoch": 0.1831038704680043,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.92623894201483e-06,
+      "loss": 1.0149,
+      "step": 916
+    },
+    {
+      "epoch": 0.18330376552310038,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.926058491515982e-06,
+      "loss": 1.1744,
+      "step": 917
+    },
+    {
+      "epoch": 0.18350366057819645,
+      "grad_norm": 2.125,
+      "learning_rate": 9.92587782220214e-06,
+      "loss": 1.0773,
+      "step": 918
+    },
+    {
+      "epoch": 0.18370355563329252,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.925696934081335e-06,
+      "loss": 1.0894,
+      "step": 919
+    },
+    {
+      "epoch": 0.1839034506883886,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.925515827161596e-06,
+      "loss": 1.1613,
+      "step": 920
+    },
+    {
+      "epoch": 0.18410334574348466,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.925334501450972e-06,
+      "loss": 1.1258,
+      "step": 921
+    },
+    {
+      "epoch": 0.18430324079858074,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.925152956957513e-06,
+      "loss": 1.0427,
+      "step": 922
+    },
+    {
+      "epoch": 0.1845031358536768,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.924971193689287e-06,
+      "loss": 1.0592,
+      "step": 923
+    },
+    {
+      "epoch": 0.18470303090877288,
+      "grad_norm": 2.125,
+      "learning_rate": 9.924789211654367e-06,
+      "loss": 1.1031,
+      "step": 924
+    },
+    {
+      "epoch": 0.18490292596386898,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.924607010860833e-06,
+      "loss": 1.077,
+      "step": 925
+    },
+    {
+      "epoch": 0.18510282101896505,
+      "grad_norm": 2.25,
+      "learning_rate": 9.924424591316785e-06,
+      "loss": 1.1194,
+      "step": 926
+    },
+    {
+      "epoch": 0.18530271607406112,
+      "grad_norm": 2.125,
+      "learning_rate": 9.924241953030323e-06,
+      "loss": 1.0626,
+      "step": 927
+    },
+    {
+      "epoch": 0.1855026111291572,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.924059096009556e-06,
+      "loss": 1.1046,
+      "step": 928
+    },
+    {
+      "epoch": 0.18570250618425327,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.923876020262613e-06,
+      "loss": 1.1724,
+      "step": 929
+    },
+    {
+      "epoch": 0.18590240123934934,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.923692725797622e-06,
+      "loss": 1.1293,
+      "step": 930
+    },
+    {
+      "epoch": 0.1861022962944454,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.923509212622726e-06,
+      "loss": 1.1245,
+      "step": 931
+    },
+    {
+      "epoch": 0.18630219134954148,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.923325480746077e-06,
+      "loss": 1.0509,
+      "step": 932
+    },
+    {
+      "epoch": 0.18650208640463756,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.923141530175835e-06,
+      "loss": 1.0779,
+      "step": 933
+    },
+    {
+      "epoch": 0.18670198145973363,
+      "grad_norm": 2.125,
+      "learning_rate": 9.922957360920173e-06,
+      "loss": 1.1274,
+      "step": 934
+    },
+    {
+      "epoch": 0.1869018765148297,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.922772972987271e-06,
+      "loss": 1.1678,
+      "step": 935
+    },
+    {
+      "epoch": 0.1871017715699258,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.922588366385319e-06,
+      "loss": 1.1291,
+      "step": 936
+    },
+    {
+      "epoch": 0.18730166662502187,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.922403541122516e-06,
+      "loss": 0.9955,
+      "step": 937
+    },
+    {
+      "epoch": 0.18750156168011795,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.922218497207075e-06,
+      "loss": 1.0822,
+      "step": 938
+    },
+    {
+      "epoch": 0.18770145673521402,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.922033234647213e-06,
+      "loss": 1.0365,
+      "step": 939
+    },
+    {
+      "epoch": 0.1879013517903101,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.921847753451162e-06,
+      "loss": 1.0929,
+      "step": 940
+    },
+    {
+      "epoch": 0.18810124684540616,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.92166205362716e-06,
+      "loss": 1.1478,
+      "step": 941
+    },
+    {
+      "epoch": 0.18830114190050223,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.921476135183452e-06,
+      "loss": 1.037,
+      "step": 942
+    },
+    {
+      "epoch": 0.1885010369555983,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.921289998128303e-06,
+      "loss": 1.0561,
+      "step": 943
+    },
+    {
+      "epoch": 0.18870093201069438,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.921103642469976e-06,
+      "loss": 1.0575,
+      "step": 944
+    },
+    {
+      "epoch": 0.18890082706579045,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.92091706821675e-06,
+      "loss": 1.1551,
+      "step": 945
+    },
+    {
+      "epoch": 0.18910072212088652,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.920730275376915e-06,
+      "loss": 1.09,
+      "step": 946
+    },
+    {
+      "epoch": 0.18930061717598262,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.920543263958767e-06,
+      "loss": 1.0898,
+      "step": 947
+    },
+    {
+      "epoch": 0.1895005122310787,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.920356033970613e-06,
+      "loss": 1.1994,
+      "step": 948
+    },
+    {
+      "epoch": 0.18970040728617477,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.920168585420768e-06,
+      "loss": 1.2403,
+      "step": 949
+    },
+    {
+      "epoch": 0.18990030234127084,
+      "grad_norm": 2.125,
+      "learning_rate": 9.91998091831756e-06,
+      "loss": 1.2184,
+      "step": 950
+    },
+    {
+      "epoch": 0.1901001973963669,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.919793032669324e-06,
+      "loss": 1.2009,
+      "step": 951
+    },
+    {
+      "epoch": 0.19030009245146298,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.91960492848441e-06,
+      "loss": 1.0286,
+      "step": 952
+    },
+    {
+      "epoch": 0.19049998750655905,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.91941660577117e-06,
+      "loss": 1.0413,
+      "step": 953
+    },
+    {
+      "epoch": 0.19069988256165513,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.919228064537968e-06,
+      "loss": 1.0637,
+      "step": 954
+    },
+    {
+      "epoch": 0.1908997776167512,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.91903930479318e-06,
+      "loss": 1.206,
+      "step": 955
+    },
+    {
+      "epoch": 0.19109967267184727,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.918850326545195e-06,
+      "loss": 1.0584,
+      "step": 956
+    },
+    {
+      "epoch": 0.19129956772694334,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.918661129802402e-06,
+      "loss": 1.1496,
+      "step": 957
+    },
+    {
+      "epoch": 0.19149946278203944,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.918471714573205e-06,
+      "loss": 1.0116,
+      "step": 958
+    },
+    {
+      "epoch": 0.19169935783713551,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.918282080866022e-06,
+      "loss": 1.1686,
+      "step": 959
+    },
+    {
+      "epoch": 0.1918992528922316,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.918092228689276e-06,
+      "loss": 1.0539,
+      "step": 960
+    },
+    {
+      "epoch": 0.19209914794732766,
+      "grad_norm": 2.125,
+      "learning_rate": 9.917902158051395e-06,
+      "loss": 1.1964,
+      "step": 961
+    },
+    {
+      "epoch": 0.19229904300242373,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.917711868960826e-06,
+      "loss": 1.1666,
+      "step": 962
+    },
+    {
+      "epoch": 0.1924989380575198,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.917521361426023e-06,
+      "loss": 1.0984,
+      "step": 963
+    },
+    {
+      "epoch": 0.19269883311261587,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.917330635455445e-06,
+      "loss": 1.137,
+      "step": 964
+    },
+    {
+      "epoch": 0.19289872816771195,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.917139691057565e-06,
+      "loss": 1.084,
+      "step": 965
+    },
+    {
+      "epoch": 0.19309862322280802,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.916948528240865e-06,
+      "loss": 1.056,
+      "step": 966
+    },
+    {
+      "epoch": 0.1932985182779041,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.916757147013837e-06,
+      "loss": 1.0956,
+      "step": 967
+    },
+    {
+      "epoch": 0.19349841333300016,
+      "grad_norm": 2.25,
+      "learning_rate": 9.916565547384981e-06,
+      "loss": 1.1427,
+      "step": 968
+    },
+    {
+      "epoch": 0.19369830838809626,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.91637372936281e-06,
+      "loss": 1.0469,
+      "step": 969
+    },
+    {
+      "epoch": 0.19389820344319234,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.916181692955841e-06,
+      "loss": 1.0015,
+      "step": 970
+    },
+    {
+      "epoch": 0.1940980984982884,
+      "grad_norm": 2.375,
+      "learning_rate": 9.915989438172608e-06,
+      "loss": 1.1941,
+      "step": 971
+    },
+    {
+      "epoch": 0.19429799355338448,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.915796965021648e-06,
+      "loss": 1.109,
+      "step": 972
+    },
+    {
+      "epoch": 0.19449788860848055,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.915604273511514e-06,
+      "loss": 1.0788,
+      "step": 973
+    },
+    {
+      "epoch": 0.19469778366357662,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.915411363650762e-06,
+      "loss": 1.0593,
+      "step": 974
+    },
+    {
+      "epoch": 0.1948976787186727,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.915218235447962e-06,
+      "loss": 1.1721,
+      "step": 975
+    },
+    {
+      "epoch": 0.19509757377376877,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.915024888911692e-06,
+      "loss": 1.1182,
+      "step": 976
+    },
+    {
+      "epoch": 0.19529746882886484,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.914831324050542e-06,
+      "loss": 1.1052,
+      "step": 977
+    },
+    {
+      "epoch": 0.1954973638839609,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.914637540873112e-06,
+      "loss": 1.1633,
+      "step": 978
+    },
+    {
+      "epoch": 0.19569725893905698,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.914443539388003e-06,
+      "loss": 1.1462,
+      "step": 979
+    },
+    {
+      "epoch": 0.19589715399415306,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.914249319603839e-06,
+      "loss": 1.0912,
+      "step": 980
+    },
+    {
+      "epoch": 0.19609704904924916,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.914054881529245e-06,
+      "loss": 1.0723,
+      "step": 981
+    },
+    {
+      "epoch": 0.19629694410434523,
+      "grad_norm": 2.125,
+      "learning_rate": 9.91386022517286e-06,
+      "loss": 1.1248,
+      "step": 982
+    },
+    {
+      "epoch": 0.1964968391594413,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.913665350543324e-06,
+      "loss": 1.0391,
+      "step": 983
+    },
+    {
+      "epoch": 0.19669673421453737,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.913470257649303e-06,
+      "loss": 1.1673,
+      "step": 984
+    },
+    {
+      "epoch": 0.19689662926963344,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.913274946499453e-06,
+      "loss": 1.1034,
+      "step": 985
+    },
+    {
+      "epoch": 0.19709652432472952,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.913079417102458e-06,
+      "loss": 1.1082,
+      "step": 986
+    },
+    {
+      "epoch": 0.1972964193798256,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.912883669467e-06,
+      "loss": 1.1261,
+      "step": 987
+    },
+    {
+      "epoch": 0.19749631443492166,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.912687703601774e-06,
+      "loss": 1.1113,
+      "step": 988
+    },
+    {
+      "epoch": 0.19769620949001773,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.912491519515484e-06,
+      "loss": 1.1534,
+      "step": 989
+    },
+    {
+      "epoch": 0.1978961045451138,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.912295117216844e-06,
+      "loss": 1.0584,
+      "step": 990
+    },
+    {
+      "epoch": 0.19809599960020988,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.912098496714582e-06,
+      "loss": 1.0466,
+      "step": 991
+    },
+    {
+      "epoch": 0.19829589465530598,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.911901658017428e-06,
+      "loss": 1.0825,
+      "step": 992
+    },
+    {
+      "epoch": 0.19849578971040205,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.911704601134127e-06,
+      "loss": 1.1456,
+      "step": 993
+    },
+    {
+      "epoch": 0.19869568476549812,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.911507326073433e-06,
+      "loss": 1.1568,
+      "step": 994
+    },
+    {
+      "epoch": 0.1988955798205942,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.911309832844108e-06,
+      "loss": 1.1335,
+      "step": 995
+    },
+    {
+      "epoch": 0.19909547487569026,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.911112121454925e-06,
+      "loss": 1.0891,
+      "step": 996
+    },
+    {
+      "epoch": 0.19929536993078634,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.910914191914664e-06,
+      "loss": 1.1071,
+      "step": 997
+    },
+    {
+      "epoch": 0.1994952649858824,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.910716044232122e-06,
+      "loss": 1.0985,
+      "step": 998
+    },
+    {
+      "epoch": 0.19969516004097848,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.910517678416097e-06,
+      "loss": 1.055,
+      "step": 999
+    },
+    {
+      "epoch": 0.19989505509607455,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.9103190944754e-06,
+      "loss": 1.092,
+      "step": 1000
+    },
+    {
+      "epoch": 0.20009495015117063,
+      "grad_norm": 2.125,
+      "learning_rate": 9.910120292418855e-06,
+      "loss": 1.1513,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2002948452062667,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.909921272255289e-06,
+      "loss": 1.0961,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2004947402613628,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.909722033993546e-06,
+      "loss": 1.1312,
+      "step": 1003
+    },
+    {
+      "epoch": 0.20069463531645887,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.909522577642474e-06,
+      "loss": 1.088,
+      "step": 1004
+    },
+    {
+      "epoch": 0.20089453037155494,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.909322903210934e-06,
+      "loss": 1.0347,
+      "step": 1005
+    },
+    {
+      "epoch": 0.201094425426651,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.909123010707793e-06,
+      "loss": 1.097,
+      "step": 1006
+    },
+    {
+      "epoch": 0.20129432048174709,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.908922900141935e-06,
+      "loss": 1.158,
+      "step": 1007
+    },
+    {
+      "epoch": 0.20149421553684316,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.908722571522244e-06,
+      "loss": 1.1167,
+      "step": 1008
+    },
+    {
+      "epoch": 0.20169411059193923,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.90852202485762e-06,
+      "loss": 1.0482,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2018940056470353,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.908321260156975e-06,
+      "loss": 1.0526,
+      "step": 1010
+    },
+    {
+      "epoch": 0.20209390070213137,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.908120277429224e-06,
+      "loss": 1.1366,
+      "step": 1011
+    },
+    {
+      "epoch": 0.20229379575722745,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.90791907668329e-06,
+      "loss": 1.0403,
+      "step": 1012
+    },
+    {
+      "epoch": 0.20249369081232352,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.907717657928117e-06,
+      "loss": 1.0606,
+      "step": 1013
+    },
+    {
+      "epoch": 0.20269358586741962,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.907516021172652e-06,
+      "loss": 1.0486,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2028934809225157,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.907314166425847e-06,
+      "loss": 1.1866,
+      "step": 1015
+    },
+    {
+      "epoch": 0.20309337597761176,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.907112093696672e-06,
+      "loss": 1.0322,
+      "step": 1016
+    },
+    {
+      "epoch": 0.20329327103270783,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.906909802994101e-06,
+      "loss": 1.1012,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2034931660878039,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.90670729432712e-06,
+      "loss": 1.0504,
+      "step": 1018
+    },
+    {
+      "epoch": 0.20369306114289998,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.906504567704727e-06,
+      "loss": 1.0983,
+      "step": 1019
+    },
+    {
+      "epoch": 0.20389295619799605,
+      "grad_norm": 2.125,
+      "learning_rate": 9.906301623135925e-06,
+      "loss": 1.1327,
+      "step": 1020
+    },
+    {
+      "epoch": 0.20409285125309212,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.906098460629728e-06,
+      "loss": 1.0782,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2042927463081882,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.905895080195161e-06,
+      "loss": 1.0306,
+      "step": 1022
+    },
+    {
+      "epoch": 0.20449264136328427,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.90569148184126e-06,
+      "loss": 1.0627,
+      "step": 1023
+    },
+    {
+      "epoch": 0.20469253641838034,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.905487665577067e-06,
+      "loss": 1.0469,
+      "step": 1024
+    },
+    {
+      "epoch": 0.20489243147347644,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.905283631411635e-06,
+      "loss": 0.9674,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2050923265285725,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.905079379354028e-06,
+      "loss": 1.1038,
+      "step": 1026
+    },
+    {
+      "epoch": 0.20529222158366858,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.904874909413318e-06,
+      "loss": 1.0563,
+      "step": 1027
+    },
+    {
+      "epoch": 0.20549211663876465,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.90467022159859e-06,
+      "loss": 1.079,
+      "step": 1028
+    },
+    {
+      "epoch": 0.20569201169386073,
+      "grad_norm": 2.125,
+      "learning_rate": 9.904465315918934e-06,
+      "loss": 1.0519,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2058919067489568,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.904260192383452e-06,
+      "loss": 1.1527,
+      "step": 1030
+    },
+    {
+      "epoch": 0.20609180180405287,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.904054851001257e-06,
+      "loss": 1.0425,
+      "step": 1031
+    },
+    {
+      "epoch": 0.20629169685914894,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.903849291781468e-06,
+      "loss": 1.0826,
+      "step": 1032
+    },
+    {
+      "epoch": 0.20649159191424502,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.903643514733218e-06,
+      "loss": 1.0684,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2066914869693411,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.903437519865648e-06,
+      "loss": 1.0938,
+      "step": 1034
+    },
+    {
+      "epoch": 0.20689138202443716,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.903231307187906e-06,
+      "loss": 1.1274,
+      "step": 1035
+    },
+    {
+      "epoch": 0.20709127707953323,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.903024876709154e-06,
+      "loss": 1.1545,
+      "step": 1036
+    },
+    {
+      "epoch": 0.20729117213462933,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.902818228438557e-06,
+      "loss": 1.0743,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2074910671897254,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.902611362385302e-06,
+      "loss": 1.0972,
+      "step": 1038
+    },
+    {
+      "epoch": 0.20769096224482148,
+      "grad_norm": 4.625,
+      "learning_rate": 9.902404278558573e-06,
+      "loss": 1.0203,
+      "step": 1039
+    },
+    {
+      "epoch": 0.20789085729991755,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.902196976967568e-06,
+      "loss": 1.0272,
+      "step": 1040
+    },
+    {
+      "epoch": 0.20809075235501362,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.901989457621497e-06,
+      "loss": 1.2083,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2082906474101097,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.90178172052958e-06,
+      "loss": 1.0203,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20849054246520576,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.90157376570104e-06,
+      "loss": 1.1863,
+      "step": 1043
+    },
+    {
+      "epoch": 0.20869043752030184,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.901365593145119e-06,
+      "loss": 1.2206,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2088903325753979,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.90115720287106e-06,
+      "loss": 1.1041,
+      "step": 1045
+    },
+    {
+      "epoch": 0.20909022763049398,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.900948594888122e-06,
+      "loss": 1.1631,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20929012268559005,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.900739769205571e-06,
+      "loss": 1.0053,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20949001774068615,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.900530725832685e-06,
+      "loss": 1.0736,
+      "step": 1048
+    },
+    {
+      "epoch": 0.20968991279578222,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.900321464778745e-06,
+      "loss": 1.1255,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2098898078508783,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.90011198605305e-06,
+      "loss": 1.0491,
+      "step": 1050
+    },
+    {
+      "epoch": 0.21008970290597437,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.899902289664902e-06,
+      "loss": 1.0678,
+      "step": 1051
+    },
+    {
+      "epoch": 0.21028959796107044,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.899692375623619e-06,
+      "loss": 1.0399,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2104894930161665,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.899482243938526e-06,
+      "loss": 1.0988,
+      "step": 1053
+    },
+    {
+      "epoch": 0.21068938807126258,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.89927189461895e-06,
+      "loss": 1.0065,
+      "step": 1054
+    },
+    {
+      "epoch": 0.21088928312635866,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.899061327674245e-06,
+      "loss": 1.0968,
+      "step": 1055
+    },
+    {
+      "epoch": 0.21108917818145473,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.898850543113756e-06,
+      "loss": 1.1262,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2112890732365508,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.89863954094685e-06,
+      "loss": 0.9881,
+      "step": 1057
+    },
+    {
+      "epoch": 0.21148896829164687,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.8984283211829e-06,
+      "loss": 1.1341,
+      "step": 1058
+    },
+    {
+      "epoch": 0.21168886334674297,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.898216883831284e-06,
+      "loss": 1.0557,
+      "step": 1059
+    },
+    {
+      "epoch": 0.21188875840183904,
+      "grad_norm": 2.125,
+      "learning_rate": 9.898005228901399e-06,
+      "loss": 1.1488,
+      "step": 1060
+    },
+    {
+      "epoch": 0.21208865345693512,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.897793356402646e-06,
+      "loss": 1.1246,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2122885485120312,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.897581266344434e-06,
+      "loss": 1.1322,
+      "step": 1062
+    },
+    {
+      "epoch": 0.21248844356712726,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.897368958736185e-06,
+      "loss": 1.0576,
+      "step": 1063
+    },
+    {
+      "epoch": 0.21268833862222333,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.89715643358733e-06,
+      "loss": 1.0027,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2128882336773194,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.89694369090731e-06,
+      "loss": 1.091,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21308812873241548,
+      "grad_norm": 2.125,
+      "learning_rate": 9.896730730705574e-06,
+      "loss": 1.0713,
+      "step": 1066
+    },
+    {
+      "epoch": 0.21328802378751155,
+      "grad_norm": 2.25,
+      "learning_rate": 9.896517552991581e-06,
+      "loss": 1.1078,
+      "step": 1067
+    },
+    {
+      "epoch": 0.21348791884260762,
+      "grad_norm": 2.25,
+      "learning_rate": 9.896304157774802e-06,
+      "loss": 1.1173,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2136878138977037,
+      "grad_norm": 2.125,
+      "learning_rate": 9.896090545064717e-06,
+      "loss": 1.0837,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2138877089527998,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.895876714870809e-06,
+      "loss": 1.1325,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21408760400789587,
+      "grad_norm": 2.25,
+      "learning_rate": 9.895662667202583e-06,
+      "loss": 1.0625,
+      "step": 1071
+    },
+    {
+      "epoch": 0.21428749906299194,
+      "grad_norm": 2.0,
+      "learning_rate": 9.895448402069543e-06,
+      "loss": 1.0352,
+      "step": 1072
+    },
+    {
+      "epoch": 0.214487394118088,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.895233919481208e-06,
+      "loss": 1.1547,
+      "step": 1073
+    },
+    {
+      "epoch": 0.21468728917318408,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.895019219447107e-06,
+      "loss": 1.1479,
+      "step": 1074
+    },
+    {
+      "epoch": 0.21488718422828015,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.894804301976773e-06,
+      "loss": 0.9864,
+      "step": 1075
+    },
+    {
+      "epoch": 0.21508707928337623,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.894589167079754e-06,
+      "loss": 1.0687,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2152869743384723,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.894373814765609e-06,
+      "loss": 1.063,
+      "step": 1077
+    },
+    {
+      "epoch": 0.21548686939356837,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.894158245043902e-06,
+      "loss": 1.114,
+      "step": 1078
+    },
+    {
+      "epoch": 0.21568676444866444,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.893942457924206e-06,
+      "loss": 1.1266,
+      "step": 1079
+    },
+    {
+      "epoch": 0.21588665950376051,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.89372645341611e-06,
+      "loss": 1.0792,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21608655455885661,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.893510231529209e-06,
+      "loss": 1.1263,
+      "step": 1081
+    },
+    {
+      "epoch": 0.21628644961395269,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.893293792273104e-06,
+      "loss": 1.0622,
+      "step": 1082
+    },
+    {
+      "epoch": 0.21648634466904876,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.893077135657413e-06,
+      "loss": 1.1443,
+      "step": 1083
+    },
+    {
+      "epoch": 0.21668623972414483,
+      "grad_norm": 2.125,
+      "learning_rate": 9.892860261691756e-06,
+      "loss": 1.0848,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2168861347792409,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.892643170385771e-06,
+      "loss": 1.0369,
+      "step": 1085
+    },
+    {
+      "epoch": 0.21708602983433697,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.8924258617491e-06,
+      "loss": 1.0839,
+      "step": 1086
+    },
+    {
+      "epoch": 0.21728592488943305,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.892208335791392e-06,
+      "loss": 1.0948,
+      "step": 1087
+    },
+    {
+      "epoch": 0.21748581994452912,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.891990592522314e-06,
+      "loss": 1.0512,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2176857149996252,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.891772631951535e-06,
+      "loss": 1.0416,
+      "step": 1089
+    },
+    {
+      "epoch": 0.21788561005472126,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.891554454088738e-06,
+      "loss": 1.1323,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21808550510981733,
+      "grad_norm": 2.125,
+      "learning_rate": 9.891336058943617e-06,
+      "loss": 1.1509,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2182854001649134,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.891117446525869e-06,
+      "loss": 1.0703,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2184852952200095,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.890898616845206e-06,
+      "loss": 1.1602,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21868519027510558,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.890679569911349e-06,
+      "loss": 1.0828,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21888508533020165,
+      "grad_norm": 2.0,
+      "learning_rate": 9.890460305734028e-06,
+      "loss": 1.0106,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21908498038529772,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.890240824322983e-06,
+      "loss": 1.1057,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2192848754403938,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.890021125687962e-06,
+      "loss": 1.1892,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21948477049548987,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.889801209838725e-06,
+      "loss": 1.0643,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21968466555058594,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.889581076785042e-06,
+      "loss": 1.1837,
+      "step": 1099
+    },
+    {
+      "epoch": 0.219884560605682,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.889360726536687e-06,
+      "loss": 1.2431,
+      "step": 1100
+    },
+    {
+      "epoch": 0.22008445566077808,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.889140159103454e-06,
+      "loss": 1.1512,
+      "step": 1101
+    },
+    {
+      "epoch": 0.22028435071587416,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.888919374495134e-06,
+      "loss": 1.1054,
+      "step": 1102
+    },
+    {
+      "epoch": 0.22048424577097023,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.88869837272154e-06,
+      "loss": 1.0885,
+      "step": 1103
+    },
+    {
+      "epoch": 0.22068414082606633,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.888477153792486e-06,
+      "loss": 1.0958,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2208840358811624,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.888255717717798e-06,
+      "loss": 1.1226,
+      "step": 1105
+    },
+    {
+      "epoch": 0.22108393093625847,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.888034064507314e-06,
+      "loss": 1.1548,
+      "step": 1106
+    },
+    {
+      "epoch": 0.22128382599135454,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.887812194170878e-06,
+      "loss": 1.0003,
+      "step": 1107
+    },
+    {
+      "epoch": 0.22148372104645062,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.887590106718348e-06,
+      "loss": 1.0862,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2216836161015467,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.887367802159587e-06,
+      "loss": 1.0139,
+      "step": 1109
+    },
+    {
+      "epoch": 0.22188351115664276,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.887145280504468e-06,
+      "loss": 1.072,
+      "step": 1110
+    },
+    {
+      "epoch": 0.22208340621173883,
+      "grad_norm": 2.125,
+      "learning_rate": 9.88692254176288e-06,
+      "loss": 0.973,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2222833012668349,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.886699585944715e-06,
+      "loss": 1.1539,
+      "step": 1112
+    },
+    {
+      "epoch": 0.22248319632193098,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.886476413059874e-06,
+      "loss": 0.9878,
+      "step": 1113
+    },
+    {
+      "epoch": 0.22268309137702705,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.886253023118276e-06,
+      "loss": 1.1039,
+      "step": 1114
+    },
+    {
+      "epoch": 0.22288298643212315,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.886029416129837e-06,
+      "loss": 1.0656,
+      "step": 1115
+    },
+    {
+      "epoch": 0.22308288148721922,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.885805592104494e-06,
+      "loss": 1.0074,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2232827765423153,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.88558155105219e-06,
+      "loss": 1.1455,
+      "step": 1117
+    },
+    {
+      "epoch": 0.22348267159741136,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.885357292982873e-06,
+      "loss": 1.1776,
+      "step": 1118
+    },
+    {
+      "epoch": 0.22368256665250744,
+      "grad_norm": 2.125,
+      "learning_rate": 9.885132817906509e-06,
+      "loss": 1.1606,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2238824617076035,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.884908125833066e-06,
+      "loss": 1.1144,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22408235676269958,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.884683216772527e-06,
+      "loss": 1.1148,
+      "step": 1121
+    },
+    {
+      "epoch": 0.22428225181779565,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.88445809073488e-06,
+      "loss": 1.0931,
+      "step": 1122
+    },
+    {
+      "epoch": 0.22448214687289172,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.884232747730125e-06,
+      "loss": 1.1479,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2246820419279878,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.884007187768275e-06,
+      "loss": 0.9995,
+      "step": 1124
+    },
+    {
+      "epoch": 0.22488193698308387,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.883781410859347e-06,
+      "loss": 1.1488,
+      "step": 1125
+    },
+    {
+      "epoch": 0.22508183203817997,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.88355541701337e-06,
+      "loss": 1.0807,
+      "step": 1126
+    },
+    {
+      "epoch": 0.22528172709327604,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.883329206240383e-06,
+      "loss": 1.0489,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2254816221483721,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.883102778550434e-06,
+      "loss": 1.0662,
+      "step": 1128
+    },
+    {
+      "epoch": 0.22568151720346818,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.882876133953582e-06,
+      "loss": 1.1357,
+      "step": 1129
+    },
+    {
+      "epoch": 0.22588141225856426,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.882649272459892e-06,
+      "loss": 1.0505,
+      "step": 1130
+    },
+    {
+      "epoch": 0.22608130731366033,
+      "grad_norm": 2.125,
+      "learning_rate": 9.882422194079444e-06,
+      "loss": 1.1083,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2262812023687564,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.882194898822324e-06,
+      "loss": 1.1381,
+      "step": 1132
+    },
+    {
+      "epoch": 0.22648109742385247,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.881967386698627e-06,
+      "loss": 1.0367,
+      "step": 1133
+    },
+    {
+      "epoch": 0.22668099247894855,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.881739657718462e-06,
+      "loss": 1.2138,
+      "step": 1134
+    },
+    {
+      "epoch": 0.22688088753404462,
+      "grad_norm": 2.125,
+      "learning_rate": 9.881511711891941e-06,
+      "loss": 1.1267,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2270807825891407,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.88128354922919e-06,
+      "loss": 1.0884,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2272806776442368,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.881055169740347e-06,
+      "loss": 1.1466,
+      "step": 1137
+    },
+    {
+      "epoch": 0.22748057269933286,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.880826573435555e-06,
+      "loss": 1.16,
+      "step": 1138
+    },
+    {
+      "epoch": 0.22768046775442893,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.880597760324966e-06,
+      "loss": 1.1587,
+      "step": 1139
+    },
+    {
+      "epoch": 0.227880362809525,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.880368730418749e-06,
+      "loss": 1.1319,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22808025786462108,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.880139483727071e-06,
+      "loss": 1.1318,
+      "step": 1141
+    },
+    {
+      "epoch": 0.22828015291971715,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.879910020260119e-06,
+      "loss": 1.0074,
+      "step": 1142
+    },
+    {
+      "epoch": 0.22848004797481322,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.879680340028087e-06,
+      "loss": 1.0707,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2286799430299093,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.879450443041172e-06,
+      "loss": 1.0146,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22887983808500537,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.879220329309591e-06,
+      "loss": 1.1047,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22907973314010144,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.878989998843565e-06,
+      "loss": 1.0841,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2292796281951975,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.878759451653323e-06,
+      "loss": 1.1131,
+      "step": 1147
+    },
+    {
+      "epoch": 0.22947952325029358,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.87852868774911e-06,
+      "loss": 1.1771,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22967941830538968,
+      "grad_norm": 2.25,
+      "learning_rate": 9.878297707141172e-06,
+      "loss": 1.1596,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22987931336048575,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.87806650983977e-06,
+      "loss": 1.1953,
+      "step": 1150
+    },
+    {
+      "epoch": 0.23007920841558183,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.877835095855174e-06,
+      "loss": 1.0653,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2302791034706779,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.877603465197667e-06,
+      "loss": 1.0236,
+      "step": 1152
+    },
+    {
+      "epoch": 0.23047899852577397,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.877371617877533e-06,
+      "loss": 1.1288,
+      "step": 1153
+    },
+    {
+      "epoch": 0.23067889358087004,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.877139553905072e-06,
+      "loss": 1.0238,
+      "step": 1154
+    },
+    {
+      "epoch": 0.23087878863596611,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.876907273290594e-06,
+      "loss": 1.0817,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2310786836910622,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.876674776044417e-06,
+      "loss": 1.1418,
+      "step": 1156
+    },
+    {
+      "epoch": 0.23127857874615826,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.876442062176866e-06,
+      "loss": 1.1305,
+      "step": 1157
+    },
+    {
+      "epoch": 0.23147847380125433,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.87620913169828e-06,
+      "loss": 1.0216,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2316783688563504,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.875975984619004e-06,
+      "loss": 1.0736,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2318782639114465,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.875742620949395e-06,
+      "loss": 1.0892,
+      "step": 1160
+    },
+    {
+      "epoch": 0.23207815896654257,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.875509040699821e-06,
+      "loss": 1.0612,
+      "step": 1161
+    },
+    {
+      "epoch": 0.23227805402163865,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.875275243880657e-06,
+      "loss": 1.1469,
+      "step": 1162
+    },
+    {
+      "epoch": 0.23247794907673472,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.875041230502286e-06,
+      "loss": 1.1195,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2326778441318308,
+      "grad_norm": 2.125,
+      "learning_rate": 9.874807000575105e-06,
+      "loss": 1.0738,
+      "step": 1164
+    },
+    {
+      "epoch": 0.23287773918692686,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.874572554109517e-06,
+      "loss": 1.0748,
+      "step": 1165
+    },
+    {
+      "epoch": 0.23307763424202294,
+      "grad_norm": 2.125,
+      "learning_rate": 9.874337891115938e-06,
+      "loss": 1.0728,
+      "step": 1166
+    },
+    {
+      "epoch": 0.233277529297119,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.874103011604788e-06,
+      "loss": 1.0373,
+      "step": 1167
+    },
+    {
+      "epoch": 0.23347742435221508,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.873867915586504e-06,
+      "loss": 1.0788,
+      "step": 1168
+    },
+    {
+      "epoch": 0.23367731940731115,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.873632603071528e-06,
+      "loss": 1.1423,
+      "step": 1169
+    },
+    {
+      "epoch": 0.23387721446240722,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.873397074070312e-06,
+      "loss": 1.1078,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23407710951750332,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.873161328593319e-06,
+      "loss": 1.1371,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2342770045725994,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.87292536665102e-06,
+      "loss": 1.0824,
+      "step": 1172
+    },
+    {
+      "epoch": 0.23447689962769547,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.872689188253895e-06,
+      "loss": 1.191,
+      "step": 1173
+    },
+    {
+      "epoch": 0.23467679468279154,
+      "grad_norm": 2.0,
+      "learning_rate": 9.872452793412439e-06,
+      "loss": 1.0575,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2348766897378876,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.872216182137148e-06,
+      "loss": 1.1062,
+      "step": 1175
+    },
+    {
+      "epoch": 0.23507658479298368,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.871979354438539e-06,
+      "loss": 1.1298,
+      "step": 1176
+    },
+    {
+      "epoch": 0.23527647984807976,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.871742310327124e-06,
+      "loss": 1.1012,
+      "step": 1177
+    },
+    {
+      "epoch": 0.23547637490317583,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.871505049813436e-06,
+      "loss": 1.0347,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2356762699582719,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.871267572908015e-06,
+      "loss": 1.0454,
+      "step": 1179
+    },
+    {
+      "epoch": 0.23587616501336797,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.871029879621408e-06,
+      "loss": 1.1042,
+      "step": 1180
+    },
+    {
+      "epoch": 0.23607606006846404,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.870791969964173e-06,
+      "loss": 1.0733,
+      "step": 1181
+    },
+    {
+      "epoch": 0.23627595512356014,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.870553843946879e-06,
+      "loss": 1.0674,
+      "step": 1182
+    },
+    {
+      "epoch": 0.23647585017865622,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.870315501580106e-06,
+      "loss": 1.0334,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2366757452337523,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.870076942874435e-06,
+      "loss": 1.0535,
+      "step": 1184
+    },
+    {
+      "epoch": 0.23687564028884836,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.86983816784047e-06,
+      "loss": 1.093,
+      "step": 1185
+    },
+    {
+      "epoch": 0.23707553534394443,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.869599176488812e-06,
+      "loss": 1.0887,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2372754303990405,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.869359968830078e-06,
+      "loss": 1.0941,
+      "step": 1187
+    },
+    {
+      "epoch": 0.23747532545413658,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.869120544874895e-06,
+      "loss": 1.0711,
+      "step": 1188
+    },
+    {
+      "epoch": 0.23767522050923265,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.868880904633898e-06,
+      "loss": 1.2008,
+      "step": 1189
+    },
+    {
+      "epoch": 0.23787511556432872,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.86864104811773e-06,
+      "loss": 1.1192,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2380750106194248,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.868400975337046e-06,
+      "loss": 1.0715,
+      "step": 1191
+    },
+    {
+      "epoch": 0.23827490567452086,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.868160686302513e-06,
+      "loss": 1.1605,
+      "step": 1192
+    },
+    {
+      "epoch": 0.23847480072961696,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.867920181024802e-06,
+      "loss": 1.1909,
+      "step": 1193
+    },
+    {
+      "epoch": 0.23867469578471304,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.867679459514596e-06,
+      "loss": 1.0285,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2388745908398091,
+      "grad_norm": 2.125,
+      "learning_rate": 9.867438521782586e-06,
+      "loss": 1.1216,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23907448589490518,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.867197367839481e-06,
+      "loss": 1.0404,
+      "step": 1196
+    },
+    {
+      "epoch": 0.23927438095000125,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.866955997695984e-06,
+      "loss": 1.1369,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23947427600509733,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.866714411362825e-06,
+      "loss": 1.0819,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2396741710601934,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.86647260885073e-06,
+      "loss": 1.0858,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23987406611528947,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.866230590170442e-06,
+      "loss": 1.081,
+      "step": 1200
+    },
+    {
+      "epoch": 0.24007396117038554,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.86598835533271e-06,
+      "loss": 1.0809,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2402738562254816,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.865745904348296e-06,
+      "loss": 1.1058,
+      "step": 1202
+    },
+    {
+      "epoch": 0.24047375128057769,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.86550323722797e-06,
+      "loss": 1.1048,
+      "step": 1203
+    },
+    {
+      "epoch": 0.24067364633567376,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.865260353982506e-06,
+      "loss": 1.0469,
+      "step": 1204
+    },
+    {
+      "epoch": 0.24087354139076986,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.8650172546227e-06,
+      "loss": 1.082,
+      "step": 1205
+    },
+    {
+      "epoch": 0.24107343644586593,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.864773939159346e-06,
+      "loss": 1.1258,
+      "step": 1206
+    },
+    {
+      "epoch": 0.241273331500962,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.864530407603253e-06,
+      "loss": 1.1412,
+      "step": 1207
+    },
+    {
+      "epoch": 0.24147322655605807,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.86428665996524e-06,
+      "loss": 0.947,
+      "step": 1208
+    },
+    {
+      "epoch": 0.24167312161115415,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.864042696256132e-06,
+      "loss": 1.1225,
+      "step": 1209
+    },
+    {
+      "epoch": 0.24187301666625022,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.863798516486767e-06,
+      "loss": 1.1101,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2420729117213463,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.86355412066799e-06,
+      "loss": 1.0778,
+      "step": 1211
+    },
+    {
+      "epoch": 0.24227280677644236,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.86330950881066e-06,
+      "loss": 1.151,
+      "step": 1212
+    },
+    {
+      "epoch": 0.24247270183153843,
+      "grad_norm": 2.125,
+      "learning_rate": 9.863064680925643e-06,
+      "loss": 1.0395,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2426725968866345,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.86281963702381e-06,
+      "loss": 1.03,
+      "step": 1214
+    },
+    {
+      "epoch": 0.24287249194173058,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.862574377116048e-06,
+      "loss": 1.0622,
+      "step": 1215
+    },
+    {
+      "epoch": 0.24307238699682668,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.862328901213253e-06,
+      "loss": 1.0321,
+      "step": 1216
+    },
+    {
+      "epoch": 0.24327228205192275,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.862083209326326e-06,
+      "loss": 1.1011,
+      "step": 1217
+    },
+    {
+      "epoch": 0.24347217710701882,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.861837301466182e-06,
+      "loss": 1.1446,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2436720721621149,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.861591177643744e-06,
+      "loss": 1.1048,
+      "step": 1219
+    },
+    {
+      "epoch": 0.24387196721721097,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.861344837869947e-06,
+      "loss": 1.1738,
+      "step": 1220
+    },
+    {
+      "epoch": 0.24407186227230704,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.86109828215573e-06,
+      "loss": 1.0178,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2442717573274031,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.860851510512046e-06,
+      "loss": 1.0646,
+      "step": 1222
+    },
+    {
+      "epoch": 0.24447165238249918,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.860604522949859e-06,
+      "loss": 1.0765,
+      "step": 1223
+    },
+    {
+      "epoch": 0.24467154743759525,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.860357319480137e-06,
+      "loss": 1.1155,
+      "step": 1224
+    },
+    {
+      "epoch": 0.24487144249269133,
+      "grad_norm": 2.25,
+      "learning_rate": 9.860109900113861e-06,
+      "loss": 1.1207,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2450713375477874,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.859862264862023e-06,
+      "loss": 1.1647,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2452712326028835,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.859614413735623e-06,
+      "loss": 1.0827,
+      "step": 1227
+    },
+    {
+      "epoch": 0.24547112765797957,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.85936634674567e-06,
+      "loss": 1.0661,
+      "step": 1228
+    },
+    {
+      "epoch": 0.24567102271307564,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.859118063903182e-06,
+      "loss": 1.1036,
+      "step": 1229
+    },
+    {
+      "epoch": 0.24587091776817172,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.858869565219189e-06,
+      "loss": 1.0841,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2460708128232678,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.85862085070473e-06,
+      "loss": 1.0565,
+      "step": 1231
+    },
+    {
+      "epoch": 0.24627070787836386,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.85837192037085e-06,
+      "loss": 1.1141,
+      "step": 1232
+    },
+    {
+      "epoch": 0.24647060293345993,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.858122774228609e-06,
+      "loss": 1.0912,
+      "step": 1233
+    },
+    {
+      "epoch": 0.246670497988556,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.857873412289071e-06,
+      "loss": 0.9702,
+      "step": 1234
+    },
+    {
+      "epoch": 0.24687039304365208,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.85762383456332e-06,
+      "loss": 1.1655,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24707028809874815,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.857374041062433e-06,
+      "loss": 1.0594,
+      "step": 1236
+    },
+    {
+      "epoch": 0.24727018315384422,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.85712403179751e-06,
+      "loss": 1.0415,
+      "step": 1237
+    },
+    {
+      "epoch": 0.24747007820894032,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.856873806779656e-06,
+      "loss": 1.0599,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2476699732640364,
+      "grad_norm": 2.125,
+      "learning_rate": 9.856623366019988e-06,
+      "loss": 1.0679,
+      "step": 1239
+    },
+    {
+      "epoch": 0.24786986831913246,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.85637270952963e-06,
+      "loss": 1.1581,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24806976337422854,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.856121837319712e-06,
+      "loss": 0.9953,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2482696584293246,
+      "grad_norm": 2.125,
+      "learning_rate": 9.855870749401381e-06,
+      "loss": 1.0956,
+      "step": 1242
+    },
+    {
+      "epoch": 0.24846955348442068,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.855619445785791e-06,
+      "loss": 1.0473,
+      "step": 1243
+    },
+    {
+      "epoch": 0.24866944853951675,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.855367926484103e-06,
+      "loss": 1.1015,
+      "step": 1244
+    },
+    {
+      "epoch": 0.24886934359461282,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.85511619150749e-06,
+      "loss": 1.1346,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2490692386497089,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.854864240867137e-06,
+      "loss": 1.1165,
+      "step": 1246
+    },
+    {
+      "epoch": 0.24926913370480497,
+      "grad_norm": 2.25,
+      "learning_rate": 9.85461207457423e-06,
+      "loss": 1.1865,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24946902875990104,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.854359692639974e-06,
+      "loss": 1.0496,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24966892381499714,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.854107095075578e-06,
+      "loss": 1.0782,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2498688188700932,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.853854281892265e-06,
+      "loss": 1.0533,
+      "step": 1250
+    },
+    {
+      "epoch": 0.25006871392518926,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.853601253101262e-06,
+      "loss": 1.1048,
+      "step": 1251
+    },
+    {
+      "epoch": 0.25026860898028536,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.85334800871381e-06,
+      "loss": 1.0275,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2504685040353814,
+      "grad_norm": 2.125,
+      "learning_rate": 9.853094548741158e-06,
+      "loss": 1.1582,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2506683990904775,
+      "grad_norm": 2.125,
+      "learning_rate": 9.852840873194565e-06,
+      "loss": 1.0982,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2508682941455736,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.852586982085298e-06,
+      "loss": 1.0824,
+      "step": 1255
+    },
+    {
+      "epoch": 0.25106818920066964,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.852332875424636e-06,
+      "loss": 1.1468,
+      "step": 1256
+    },
+    {
+      "epoch": 0.25126808425576574,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.852078553223865e-06,
+      "loss": 1.0221,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2514679793108618,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.851824015494284e-06,
+      "loss": 1.049,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2516678743659579,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.851569262247198e-06,
+      "loss": 1.1261,
+      "step": 1259
+    },
+    {
+      "epoch": 0.25186776942105393,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.851314293493923e-06,
+      "loss": 1.1281,
+      "step": 1260
+    },
+    {
+      "epoch": 0.25206766447615003,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.851059109245785e-06,
+      "loss": 1.022,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2522675595312461,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.850803709514121e-06,
+      "loss": 1.1081,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2524674545863422,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.850548094310273e-06,
+      "loss": 1.0569,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2526673496414382,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.850292263645597e-06,
+      "loss": 1.0977,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2528672446965343,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.850036217531457e-06,
+      "loss": 1.06,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2530671397516304,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.849779955979226e-06,
+      "loss": 1.0829,
+      "step": 1266
+    },
+    {
+      "epoch": 0.25326703480672647,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.849523479000287e-06,
+      "loss": 1.1455,
+      "step": 1267
+    },
+    {
+      "epoch": 0.25346692986182257,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.849266786606033e-06,
+      "loss": 1.1681,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2536668249169186,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.849009878807867e-06,
+      "loss": 1.07,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2538667199720147,
+      "grad_norm": 2.25,
+      "learning_rate": 9.848752755617201e-06,
+      "loss": 1.1013,
+      "step": 1270
+    },
+    {
+      "epoch": 0.25406661502711075,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.848495417045454e-06,
+      "loss": 1.1068,
+      "step": 1271
+    },
+    {
+      "epoch": 0.25426651008220685,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.848237863104057e-06,
+      "loss": 0.9974,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2544664051373029,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.847980093804455e-06,
+      "loss": 1.1419,
+      "step": 1273
+    },
+    {
+      "epoch": 0.254666300192399,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.847722109158094e-06,
+      "loss": 1.0713,
+      "step": 1274
+    },
+    {
+      "epoch": 0.25486619524749504,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.847463909176433e-06,
+      "loss": 1.0939,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25506609030259114,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.847205493870944e-06,
+      "loss": 1.1118,
+      "step": 1276
+    },
+    {
+      "epoch": 0.25526598535768724,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.846946863253104e-06,
+      "loss": 1.052,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2554658804127833,
+      "grad_norm": 2.125,
+      "learning_rate": 9.846688017334405e-06,
+      "loss": 1.0839,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2556657754678794,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.846428956126338e-06,
+      "loss": 1.1021,
+      "step": 1279
+    },
+    {
+      "epoch": 0.25586567052297543,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.846169679640417e-06,
+      "loss": 1.011,
+      "step": 1280
+    },
+    {
+      "epoch": 0.25606556557807153,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.845910187888155e-06,
+      "loss": 1.0385,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2562654606331676,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.84565048088108e-06,
+      "loss": 1.1017,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2564653556882637,
+      "grad_norm": 2.125,
+      "learning_rate": 9.84539055863073e-06,
+      "loss": 1.1327,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2566652507433597,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.845130421148646e-06,
+      "loss": 1.1224,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2568651457984558,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.844870068446389e-06,
+      "loss": 1.1337,
+      "step": 1285
+    },
+    {
+      "epoch": 0.25706504085355186,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.84460950053552e-06,
+      "loss": 1.0371,
+      "step": 1286
+    },
+    {
+      "epoch": 0.25726493590864796,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.844348717427614e-06,
+      "loss": 1.1558,
+      "step": 1287
+    },
+    {
+      "epoch": 0.257464830963744,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.844087719134254e-06,
+      "loss": 1.0198,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2576647260188401,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.843826505667038e-06,
+      "loss": 1.1347,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2578646210739362,
+      "grad_norm": 2.125,
+      "learning_rate": 9.843565077037563e-06,
+      "loss": 1.2295,
+      "step": 1290
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.843303433257447e-06,
+      "loss": 1.1543,
+      "step": 1291
+    },
+    {
+      "epoch": 0.25826441118412835,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.843041574338307e-06,
+      "loss": 1.1631,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2584643062392244,
+      "grad_norm": 2.25,
+      "learning_rate": 9.84277950029178e-06,
+      "loss": 1.0916,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2586642012943205,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.842517211129504e-06,
+      "loss": 1.1489,
+      "step": 1294
+    },
+    {
+      "epoch": 0.25886409634941654,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.842254706863131e-06,
+      "loss": 0.9759,
+      "step": 1295
+    },
+    {
+      "epoch": 0.25906399140451264,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.84199198750432e-06,
+      "loss": 1.0633,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2592638864596087,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.841729053064744e-06,
+      "loss": 1.0705,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2594637815147048,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.84146590355608e-06,
+      "loss": 1.1205,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2596636765698008,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.841202538990016e-06,
+      "loss": 1.1211,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2598635716248969,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.840938959378254e-06,
+      "loss": 1.1488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.260063466679993,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.8406751647325e-06,
+      "loss": 1.1001,
+      "step": 1301
+    },
+    {
+      "epoch": 0.26026336173508907,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.840411155064472e-06,
+      "loss": 1.0636,
+      "step": 1302
+    },
+    {
+      "epoch": 0.26046325679018517,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.840146930385897e-06,
+      "loss": 1.0858,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2606631518452812,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.839882490708512e-06,
+      "loss": 0.9908,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2608630469003773,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.839617836044065e-06,
+      "loss": 1.1681,
+      "step": 1305
+    },
+    {
+      "epoch": 0.26106294195547336,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.83935296640431e-06,
+      "loss": 1.1194,
+      "step": 1306
+    },
+    {
+      "epoch": 0.26126283701056946,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.839087881801012e-06,
+      "loss": 1.1615,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2614627320656655,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.83882258224595e-06,
+      "loss": 1.0381,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2616626271207616,
+      "grad_norm": 2.125,
+      "learning_rate": 9.838557067750903e-06,
+      "loss": 1.1391,
+      "step": 1309
+    },
+    {
+      "epoch": 0.26186252217585765,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.83829133832767e-06,
+      "loss": 1.1343,
+      "step": 1310
+    },
+    {
+      "epoch": 0.26206241723095375,
+      "grad_norm": 2.125,
+      "learning_rate": 9.838025393988051e-06,
+      "loss": 1.0177,
+      "step": 1311
+    },
+    {
+      "epoch": 0.26226231228604985,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.83775923474386e-06,
+      "loss": 1.1457,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2624622073411459,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.837492860606923e-06,
+      "loss": 1.1076,
+      "step": 1313
+    },
+    {
+      "epoch": 0.262662102396242,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.837226271589067e-06,
+      "loss": 1.0448,
+      "step": 1314
+    },
+    {
+      "epoch": 0.26286199745133804,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.836959467702139e-06,
+      "loss": 1.0799,
+      "step": 1315
+    },
+    {
+      "epoch": 0.26306189250643414,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.836692448957987e-06,
+      "loss": 1.0002,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2632617875615302,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.836425215368472e-06,
+      "loss": 1.1429,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2634616826166263,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.836157766945467e-06,
+      "loss": 1.0986,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2636615776717223,
+      "grad_norm": 2.25,
+      "learning_rate": 9.835890103700849e-06,
+      "loss": 1.1768,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2638614727268184,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.83562222564651e-06,
+      "loss": 1.0491,
+      "step": 1320
+    },
+    {
+      "epoch": 0.26406136778191447,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.835354132794349e-06,
+      "loss": 1.0405,
+      "step": 1321
+    },
+    {
+      "epoch": 0.26426126283701057,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.835085825156274e-06,
+      "loss": 1.1763,
+      "step": 1322
+    },
+    {
+      "epoch": 0.26446115789210667,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.834817302744201e-06,
+      "loss": 1.11,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2646610529472027,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.834548565570062e-06,
+      "loss": 1.1882,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2648609480022988,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.834279613645791e-06,
+      "loss": 1.1312,
+      "step": 1325
+    },
+    {
+      "epoch": 0.26506084305739486,
+      "grad_norm": 2.375,
+      "learning_rate": 9.834010446983335e-06,
+      "loss": 1.1019,
+      "step": 1326
+    },
+    {
+      "epoch": 0.26526073811249096,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.833741065594652e-06,
+      "loss": 1.1003,
+      "step": 1327
+    },
+    {
+      "epoch": 0.265460633167587,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.833471469491708e-06,
+      "loss": 1.0937,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2656605282226831,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.833201658686478e-06,
+      "loss": 1.0801,
+      "step": 1329
+    },
+    {
+      "epoch": 0.26586042327777915,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.832931633190943e-06,
+      "loss": 1.0232,
+      "step": 1330
+    },
+    {
+      "epoch": 0.26606031833287525,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.832661393017104e-06,
+      "loss": 1.1369,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2662602133879713,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.83239093817696e-06,
+      "loss": 0.9929,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2664601084430674,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.83212026868253e-06,
+      "loss": 1.1219,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2666600034981635,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.831849384545831e-06,
+      "loss": 1.0917,
+      "step": 1334
+    },
+    {
+      "epoch": 0.26685989855325953,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.8315782857789e-06,
+      "loss": 1.2081,
+      "step": 1335
+    },
+    {
+      "epoch": 0.26705979360835563,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.831306972393778e-06,
+      "loss": 1.0511,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2672596886634517,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.831035444402514e-06,
+      "loss": 1.1524,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2674595837185478,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.830763701817173e-06,
+      "loss": 1.0655,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2676594787736438,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.830491744649824e-06,
+      "loss": 1.1691,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2678593738287399,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.830219572912546e-06,
+      "loss": 1.0408,
+      "step": 1340
+    },
+    {
+      "epoch": 0.26805926888383597,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.829947186617432e-06,
+      "loss": 1.0452,
+      "step": 1341
+    },
+    {
+      "epoch": 0.26825916393893207,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.82967458577658e-06,
+      "loss": 1.1355,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2684590589940281,
+      "grad_norm": 2.125,
+      "learning_rate": 9.829401770402099e-06,
+      "loss": 1.1821,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2686589540491242,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.829128740506107e-06,
+      "loss": 1.0181,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2688588491042203,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.828855496100733e-06,
+      "loss": 1.1149,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26905874415931635,
+      "grad_norm": 2.125,
+      "learning_rate": 9.828582037198111e-06,
+      "loss": 1.0257,
+      "step": 1346
+    },
+    {
+      "epoch": 0.26925863921441245,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.828308363810392e-06,
+      "loss": 1.1245,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2694585342695085,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.828034475949732e-06,
+      "loss": 1.066,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2696584293246046,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.827760373628295e-06,
+      "loss": 1.1373,
+      "step": 1349
+    },
+    {
+      "epoch": 0.26985832437970064,
+      "grad_norm": 2.25,
+      "learning_rate": 9.82748605685826e-06,
+      "loss": 1.2466,
+      "step": 1350
+    },
+    {
+      "epoch": 0.27005821943479674,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.827211525651808e-06,
+      "loss": 1.1199,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2702581144898928,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.826936780021134e-06,
+      "loss": 0.9884,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2704580095449889,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.826661819978446e-06,
+      "loss": 1.0505,
+      "step": 1353
+    },
+    {
+      "epoch": 0.27065790460008493,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.826386645535955e-06,
+      "loss": 1.048,
+      "step": 1354
+    },
+    {
+      "epoch": 0.27085779965518103,
+      "grad_norm": 2.25,
+      "learning_rate": 9.826111256705885e-06,
+      "loss": 1.2033,
+      "step": 1355
+    },
+    {
+      "epoch": 0.27105769471027713,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.825835653500468e-06,
+      "loss": 1.0404,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2712575897653732,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.825559835931948e-06,
+      "loss": 1.077,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2714574848204693,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.825283804012573e-06,
+      "loss": 1.1831,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2716573798755653,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.825007557754608e-06,
+      "loss": 0.9796,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2718572749306614,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.824731097170323e-06,
+      "loss": 1.0301,
+      "step": 1360
+    },
+    {
+      "epoch": 0.27205716998575746,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.824454422271999e-06,
+      "loss": 1.1681,
+      "step": 1361
+    },
+    {
+      "epoch": 0.27225706504085356,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.824177533071922e-06,
+      "loss": 1.0397,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2724569600959496,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.823900429582396e-06,
+      "loss": 1.0641,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2726568551510457,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.823623111815728e-06,
+      "loss": 1.0351,
+      "step": 1364
+    },
+    {
+      "epoch": 0.27285675020614175,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.823345579784236e-06,
+      "loss": 1.076,
+      "step": 1365
+    },
+    {
+      "epoch": 0.27305664526123785,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.823067833500248e-06,
+      "loss": 1.0386,
+      "step": 1366
+    },
+    {
+      "epoch": 0.27325654031633395,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.822789872976105e-06,
+      "loss": 1.1059,
+      "step": 1367
+    },
+    {
+      "epoch": 0.27345643537143,
+      "grad_norm": 2.25,
+      "learning_rate": 9.822511698224147e-06,
+      "loss": 1.1233,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2736563304265261,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.822233309256738e-06,
+      "loss": 1.1211,
+      "step": 1369
+    },
+    {
+      "epoch": 0.27385622548162214,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.821954706086237e-06,
+      "loss": 1.0413,
+      "step": 1370
+    },
+    {
+      "epoch": 0.27405612053671824,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.821675888725025e-06,
+      "loss": 1.0646,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2742560155918143,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.821396857185484e-06,
+      "loss": 1.0904,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2744559106469104,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.821117611480011e-06,
+      "loss": 1.0808,
+      "step": 1373
+    },
+    {
+      "epoch": 0.27465580570200643,
+      "grad_norm": 2.125,
+      "learning_rate": 9.820838151621008e-06,
+      "loss": 1.0683,
+      "step": 1374
+    },
+    {
+      "epoch": 0.27485570075710253,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.820558477620888e-06,
+      "loss": 1.198,
+      "step": 1375
+    },
+    {
+      "epoch": 0.27505559581219857,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.820278589492076e-06,
+      "loss": 0.9376,
+      "step": 1376
+    },
+    {
+      "epoch": 0.27525549086729467,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.819998487247004e-06,
+      "loss": 1.0398,
+      "step": 1377
+    },
+    {
+      "epoch": 0.27545538592239077,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.819718170898116e-06,
+      "loss": 1.0014,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2756552809774868,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.819437640457858e-06,
+      "loss": 1.0894,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2758551760325829,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.819156895938697e-06,
+      "loss": 1.0558,
+      "step": 1380
+    },
+    {
+      "epoch": 0.27605507108767896,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.8188759373531e-06,
+      "loss": 1.0198,
+      "step": 1381
+    },
+    {
+      "epoch": 0.27625496614277506,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.81859476471355e-06,
+      "loss": 1.0753,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2764548611978711,
+      "grad_norm": 2.125,
+      "learning_rate": 9.818313378032535e-06,
+      "loss": 1.1915,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2766547562529672,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.818031777322554e-06,
+      "loss": 1.081,
+      "step": 1384
+    },
+    {
+      "epoch": 0.27685465130806325,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.817749962596115e-06,
+      "loss": 1.0739,
+      "step": 1385
+    },
+    {
+      "epoch": 0.27705454636315935,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.817467933865739e-06,
+      "loss": 1.1438,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2772544414182554,
+      "grad_norm": 2.125,
+      "learning_rate": 9.81718569114395e-06,
+      "loss": 1.0289,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2774543364733515,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.81690323444329e-06,
+      "loss": 1.1005,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2776542315284476,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.8166205637763e-06,
+      "loss": 1.0965,
+      "step": 1389
+    },
+    {
+      "epoch": 0.27785412658354364,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.81633767915554e-06,
+      "loss": 1.1429,
+      "step": 1390
+    },
+    {
+      "epoch": 0.27805402163863974,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.816054580593575e-06,
+      "loss": 1.0227,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2782539166937358,
+      "grad_norm": 2.125,
+      "learning_rate": 9.81577126810298e-06,
+      "loss": 1.1028,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2784538117488319,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.815487741696339e-06,
+      "loss": 1.244,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2786537068039279,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.815204001386245e-06,
+      "loss": 1.0981,
+      "step": 1394
+    },
+    {
+      "epoch": 0.278853601859024,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.814920047185306e-06,
+      "loss": 1.1177,
+      "step": 1395
+    },
+    {
+      "epoch": 0.27905349691412007,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.814635879106134e-06,
+      "loss": 1.1771,
+      "step": 1396
+    },
+    {
+      "epoch": 0.27925339196921617,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.814351497161348e-06,
+      "loss": 1.1397,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2794532870243122,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.814066901363584e-06,
+      "loss": 1.0615,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2796531820794083,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.81378209172548e-06,
+      "loss": 1.0988,
+      "step": 1399
+    },
+    {
+      "epoch": 0.27985307713450436,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.813497068259692e-06,
+      "loss": 1.0661,
+      "step": 1400
+    },
+    {
+      "epoch": 0.28005297218960046,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.813211830978879e-06,
+      "loss": 1.1077,
+      "step": 1401
+    },
+    {
+      "epoch": 0.28025286724469656,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.812926379895708e-06,
+      "loss": 1.122,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2804527622997926,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.812640715022863e-06,
+      "loss": 1.173,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2806526573548887,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.812354836373031e-06,
+      "loss": 1.0561,
+      "step": 1404
+    },
+    {
+      "epoch": 0.28085255240998475,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.812068743958912e-06,
+      "loss": 1.1136,
+      "step": 1405
+    },
+    {
+      "epoch": 0.28105244746508085,
+      "grad_norm": 2.125,
+      "learning_rate": 9.811782437793211e-06,
+      "loss": 1.0279,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2812523425201769,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.81149591788865e-06,
+      "loss": 1.0633,
+      "step": 1407
+    },
+    {
+      "epoch": 0.281452237575273,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.811209184257953e-06,
+      "loss": 1.1056,
+      "step": 1408
+    },
+    {
+      "epoch": 0.28165213263036903,
+      "grad_norm": 2.125,
+      "learning_rate": 9.81092223691386e-06,
+      "loss": 1.0893,
+      "step": 1409
+    },
+    {
+      "epoch": 0.28185202768546513,
+      "grad_norm": 2.125,
+      "learning_rate": 9.810635075869113e-06,
+      "loss": 0.9967,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2820519227405612,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.81034770113647e-06,
+      "loss": 1.0597,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2822518177956573,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.810060112728696e-06,
+      "loss": 0.9984,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2824517128507534,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.809772310658567e-06,
+      "loss": 1.0279,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2826516079058494,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.809484294938864e-06,
+      "loss": 1.1348,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2828515029609455,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.809196065582383e-06,
+      "loss": 1.0826,
+      "step": 1415
+    },
+    {
+      "epoch": 0.28305139801604157,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.808907622601926e-06,
+      "loss": 1.0565,
+      "step": 1416
+    },
+    {
+      "epoch": 0.28325129307113767,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.808618966010306e-06,
+      "loss": 1.1655,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2834511881262337,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.808330095820346e-06,
+      "loss": 1.1342,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2836510831813298,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.808041012044875e-06,
+      "loss": 1.07,
+      "step": 1419
+    },
+    {
+      "epoch": 0.28385097823642585,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.807751714696737e-06,
+      "loss": 1.0354,
+      "step": 1420
+    },
+    {
+      "epoch": 0.28405087329152195,
+      "grad_norm": 2.125,
+      "learning_rate": 9.807462203788782e-06,
+      "loss": 1.0195,
+      "step": 1421
+    },
+    {
+      "epoch": 0.284250768346618,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.807172479333868e-06,
+      "loss": 1.1112,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2844506634017141,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.806882541344867e-06,
+      "loss": 1.0059,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2846505584568102,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.806592389834654e-06,
+      "loss": 1.1527,
+      "step": 1424
+    },
+    {
+      "epoch": 0.28485045351190624,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.806302024816124e-06,
+      "loss": 1.0392,
+      "step": 1425
+    },
+    {
+      "epoch": 0.28505034856700234,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.806011446302169e-06,
+      "loss": 1.1161,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2852502436220984,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.8057206543057e-06,
+      "loss": 1.0734,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2854501386771945,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.805429648839633e-06,
+      "loss": 1.0341,
+      "step": 1428
+    },
+    {
+      "epoch": 0.28565003373229053,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.805138429916894e-06,
+      "loss": 1.1391,
+      "step": 1429
+    },
+    {
+      "epoch": 0.28584992878738663,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.80484699755042e-06,
+      "loss": 1.1245,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2860498238424827,
+      "grad_norm": 2.125,
+      "learning_rate": 9.804555351753153e-06,
+      "loss": 1.1387,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2862497188975788,
+      "grad_norm": 2.75,
+      "learning_rate": 9.804263492538054e-06,
+      "loss": 1.0452,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2864496139526748,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.80397141991808e-06,
+      "loss": 1.0985,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2866495090077709,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.80367913390621e-06,
+      "loss": 0.9788,
+      "step": 1434
+    },
+    {
+      "epoch": 0.286849404062867,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.803386634515427e-06,
+      "loss": 1.053,
+      "step": 1435
+    },
+    {
+      "epoch": 0.28704929911796306,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.803093921758721e-06,
+      "loss": 1.109,
+      "step": 1436
+    },
+    {
+      "epoch": 0.28724919417305916,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.802800995649098e-06,
+      "loss": 1.2099,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2874490892281552,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.802507856199567e-06,
+      "loss": 1.0597,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2876489842832513,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.802214503423149e-06,
+      "loss": 1.0489,
+      "step": 1439
+    },
+    {
+      "epoch": 0.28784887933834735,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.801920937332876e-06,
+      "loss": 1.1361,
+      "step": 1440
+    },
+    {
+      "epoch": 0.28804877439344345,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.801627157941788e-06,
+      "loss": 1.0418,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2882486694485395,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.801333165262936e-06,
+      "loss": 1.2393,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2884485645036356,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.801038959309376e-06,
+      "loss": 1.1047,
+      "step": 1443
+    },
+    {
+      "epoch": 0.28864845955873164,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.800744540094178e-06,
+      "loss": 0.9954,
+      "step": 1444
+    },
+    {
+      "epoch": 0.28884835461382774,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.80044990763042e-06,
+      "loss": 1.1797,
+      "step": 1445
+    },
+    {
+      "epoch": 0.28904824966892384,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.800155061931192e-06,
+      "loss": 1.0957,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2892481447240199,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.799860003009587e-06,
+      "loss": 1.1394,
+      "step": 1447
+    },
+    {
+      "epoch": 0.289448039779116,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.799564730878713e-06,
+      "loss": 1.0629,
+      "step": 1448
+    },
+    {
+      "epoch": 0.28964793483421203,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.799269245551688e-06,
+      "loss": 1.1368,
+      "step": 1449
+    },
+    {
+      "epoch": 0.28984782988930813,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.798973547041633e-06,
+      "loss": 1.0617,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2900477249444042,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.79867763536169e-06,
+      "loss": 1.0883,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2902476199995003,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.798381510524995e-06,
+      "loss": 1.042,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2904475150545963,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.798085172544707e-06,
+      "loss": 1.1617,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2906474101096924,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.797788621433987e-06,
+      "loss": 1.0817,
+      "step": 1454
+    },
+    {
+      "epoch": 0.29084730516478846,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.797491857206009e-06,
+      "loss": 1.2027,
+      "step": 1455
+    },
+    {
+      "epoch": 0.29104720021988456,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.797194879873956e-06,
+      "loss": 1.1056,
+      "step": 1456
+    },
+    {
+      "epoch": 0.29124709527498066,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.796897689451019e-06,
+      "loss": 1.1966,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2914469903300767,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.7966002859504e-06,
+      "loss": 1.1577,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2916468853851728,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.796302669385307e-06,
+      "loss": 1.0255,
+      "step": 1459
+    },
+    {
+      "epoch": 0.29184678044026885,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.796004839768962e-06,
+      "loss": 1.1133,
+      "step": 1460
+    },
+    {
+      "epoch": 0.29204667549536495,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.795706797114593e-06,
+      "loss": 1.0865,
+      "step": 1461
+    },
+    {
+      "epoch": 0.292246570550461,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.795408541435443e-06,
+      "loss": 1.0332,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2924464656055571,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.795110072744756e-06,
+      "loss": 1.1121,
+      "step": 1463
+    },
+    {
+      "epoch": 0.29264636066065314,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.794811391055793e-06,
+      "loss": 1.0971,
+      "step": 1464
+    },
+    {
+      "epoch": 0.29284625571574924,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.79451249638182e-06,
+      "loss": 1.0235,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2930461507708453,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.79421338873611e-06,
+      "loss": 1.0476,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2932460458259414,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.793914068131959e-06,
+      "loss": 1.1475,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2934459408810375,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.793614534582653e-06,
+      "loss": 1.1651,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2936458359361335,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.793314788101502e-06,
+      "loss": 1.1456,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2938457309912296,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.793014828701822e-06,
+      "loss": 1.0554,
+      "step": 1470
+    },
+    {
+      "epoch": 0.29404562604632567,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.792714656396934e-06,
+      "loss": 1.0414,
+      "step": 1471
+    },
+    {
+      "epoch": 0.29424552110142177,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.792414271200173e-06,
+      "loss": 1.0277,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2944454161565178,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.79211367312488e-06,
+      "loss": 1.1543,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2946453112116139,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.791812862184413e-06,
+      "loss": 1.1505,
+      "step": 1474
+    },
+    {
+      "epoch": 0.29484520626670996,
+      "grad_norm": 2.375,
+      "learning_rate": 9.791511838392128e-06,
+      "loss": 1.0876,
+      "step": 1475
+    },
+    {
+      "epoch": 0.29504510132180606,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.7912106017614e-06,
+      "loss": 1.1306,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2952449963769021,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.790909152305609e-06,
+      "loss": 1.0793,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2954448914319982,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.790607490038145e-06,
+      "loss": 1.0108,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2956447864870943,
+      "grad_norm": 2.125,
+      "learning_rate": 9.790305614972407e-06,
+      "loss": 1.0751,
+      "step": 1479
+    },
+    {
+      "epoch": 0.29584468154219035,
+      "grad_norm": 2.0,
+      "learning_rate": 9.790003527121806e-06,
+      "loss": 1.0636,
+      "step": 1480
+    },
+    {
+      "epoch": 0.29604457659728645,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.78970122649976e-06,
+      "loss": 1.1285,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2962444716523825,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.789398713119696e-06,
+      "loss": 1.0562,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2964443667074786,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.789095986995052e-06,
+      "loss": 1.1148,
+      "step": 1483
+    },
+    {
+      "epoch": 0.29664426176257463,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.788793048139277e-06,
+      "loss": 1.0267,
+      "step": 1484
+    },
+    {
+      "epoch": 0.29684415681767073,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.788489896565827e-06,
+      "loss": 1.0022,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2970440518727668,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.788186532288166e-06,
+      "loss": 1.0756,
+      "step": 1486
+    },
+    {
+      "epoch": 0.2972439469278629,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.787882955319771e-06,
+      "loss": 1.1432,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2974438419829589,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.787579165674129e-06,
+      "loss": 1.1975,
+      "step": 1488
+    },
+    {
+      "epoch": 0.297643737038055,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.787275163364729e-06,
+      "loss": 1.087,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2978436320931511,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.786970948405077e-06,
+      "loss": 1.0913,
+      "step": 1490
+    },
+    {
+      "epoch": 0.29804352714824717,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.786666520808688e-06,
+      "loss": 1.0329,
+      "step": 1491
+    },
+    {
+      "epoch": 0.29824342220334327,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.786361880589084e-06,
+      "loss": 1.1613,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2984433172584393,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.786057027759796e-06,
+      "loss": 1.0602,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2986432123135354,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.785751962334365e-06,
+      "loss": 1.125,
+      "step": 1494
+    },
+    {
+      "epoch": 0.29884310736863146,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.785446684326345e-06,
+      "loss": 1.1915,
+      "step": 1495
+    },
+    {
+      "epoch": 0.29904300242372756,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.785141193749292e-06,
+      "loss": 1.1184,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2992428974788236,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.78483549061678e-06,
+      "loss": 1.0091,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2994427925339197,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.784529574942385e-06,
+      "loss": 0.9844,
+      "step": 1498
+    },
+    {
+      "epoch": 0.29964268758901574,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.784223446739698e-06,
+      "loss": 1.1051,
+      "step": 1499
+    },
+    {
+      "epoch": 0.29984258264411184,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.783917106022316e-06,
+      "loss": 1.1233,
+      "step": 1500
+    },
+    {
+      "epoch": 0.29984258264411184,
+      "eval_loss": 0.954179048538208,
+      "eval_runtime": 595.9323,
+      "eval_samples_per_second": 3.588,
+      "eval_steps_per_second": 3.588,
+      "step": 1500
+    },
+    {
+      "epoch": 0.30004247769920794,
+      "grad_norm": 2.125,
+      "learning_rate": 9.783610552803849e-06,
+      "loss": 1.0654,
+      "step": 1501
+    },
+    {
+      "epoch": 0.300242372754304,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.78330378709791e-06,
+      "loss": 1.0456,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3004422678094001,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.782996808918128e-06,
+      "loss": 1.1971,
+      "step": 1503
+    },
+    {
+      "epoch": 0.30064216286449613,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.782689618278139e-06,
+      "loss": 1.1156,
+      "step": 1504
+    },
+    {
+      "epoch": 0.30084205791959223,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.782382215191589e-06,
+      "loss": 1.1085,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3010419529746883,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.782074599672131e-06,
+      "loss": 1.0369,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3012418480297844,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.78176677173343e-06,
+      "loss": 1.2029,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3014417430848804,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.78145873138916e-06,
+      "loss": 1.1371,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3016416381399765,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.781150478653003e-06,
+      "loss": 1.1452,
+      "step": 1509
+    },
+    {
+      "epoch": 0.30184153319507256,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.780842013538652e-06,
+      "loss": 1.0721,
+      "step": 1510
+    },
+    {
+      "epoch": 0.30204142825016866,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.78053333605981e-06,
+      "loss": 1.1146,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3022413233052647,
+      "grad_norm": 2.125,
+      "learning_rate": 9.780224446230188e-06,
+      "loss": 1.1378,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3024412183603608,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.779915344063506e-06,
+      "loss": 1.0507,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3026411134154569,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.779606029573496e-06,
+      "loss": 1.1384,
+      "step": 1514
+    },
+    {
+      "epoch": 0.30284100847055295,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.779296502773896e-06,
+      "loss": 1.0204,
+      "step": 1515
+    },
+    {
+      "epoch": 0.30304090352564905,
+      "grad_norm": 2.125,
+      "learning_rate": 9.778986763678455e-06,
+      "loss": 1.0589,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3032407985807451,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.778676812300935e-06,
+      "loss": 1.11,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3034406936358412,
+      "grad_norm": 2.125,
+      "learning_rate": 9.778366648655098e-06,
+      "loss": 1.0446,
+      "step": 1518
+    },
+    {
+      "epoch": 0.30364058869093724,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.778056272754728e-06,
+      "loss": 1.0326,
+      "step": 1519
+    },
+    {
+      "epoch": 0.30384048374603334,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.777745684613606e-06,
+      "loss": 1.0558,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3040403788011294,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.777434884245533e-06,
+      "loss": 1.0122,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3042402738562255,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.77712387166431e-06,
+      "loss": 1.0617,
+      "step": 1522
+    },
+    {
+      "epoch": 0.30444016891132153,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.776812646883758e-06,
+      "loss": 1.1102,
+      "step": 1523
+    },
+    {
+      "epoch": 0.30464006396641763,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.776501209917697e-06,
+      "loss": 1.1396,
+      "step": 1524
+    },
+    {
+      "epoch": 0.30483995902151373,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.776189560779963e-06,
+      "loss": 0.9904,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3050398540766098,
+      "grad_norm": 2.125,
+      "learning_rate": 9.775877699484397e-06,
+      "loss": 1.0014,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3052397491317059,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.775565626044856e-06,
+      "loss": 1.0889,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3054396441868019,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.775253340475199e-06,
+      "loss": 1.0932,
+      "step": 1528
+    },
+    {
+      "epoch": 0.305639539241898,
+      "grad_norm": 2.125,
+      "learning_rate": 9.774940842789298e-06,
+      "loss": 1.0992,
+      "step": 1529
+    },
+    {
+      "epoch": 0.30583943429699406,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.774628133001037e-06,
+      "loss": 1.0882,
+      "step": 1530
+    },
+    {
+      "epoch": 0.30603932935209016,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.7743152111243e-06,
+      "loss": 1.1073,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3062392244071862,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.774002077172994e-06,
+      "loss": 1.0974,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3064391194622823,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.773688731161027e-06,
+      "loss": 1.029,
+      "step": 1533
+    },
+    {
+      "epoch": 0.30663901451737835,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.773375173102315e-06,
+      "loss": 1.1065,
+      "step": 1534
+    },
+    {
+      "epoch": 0.30683890957247445,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.773061403010786e-06,
+      "loss": 1.0404,
+      "step": 1535
+    },
+    {
+      "epoch": 0.30703880462757055,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.772747420900381e-06,
+      "loss": 1.0275,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3072386996826666,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.772433226785045e-06,
+      "loss": 1.0852,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3074385947377627,
+      "grad_norm": 2.125,
+      "learning_rate": 9.772118820678735e-06,
+      "loss": 1.0726,
+      "step": 1538
+    },
+    {
+      "epoch": 0.30763848979285874,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.771804202595417e-06,
+      "loss": 1.1003,
+      "step": 1539
+    },
+    {
+      "epoch": 0.30783838484795484,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.771489372549064e-06,
+      "loss": 1.0986,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3080382799030509,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.771174330553665e-06,
+      "loss": 1.0746,
+      "step": 1541
+    },
+    {
+      "epoch": 0.308238174958147,
+      "grad_norm": 2.125,
+      "learning_rate": 9.770859076623211e-06,
+      "loss": 0.9782,
+      "step": 1542
+    },
+    {
+      "epoch": 0.308438070013243,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.770543610771706e-06,
+      "loss": 1.1435,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3086379650683391,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.770227933013163e-06,
+      "loss": 1.0373,
+      "step": 1544
+    },
+    {
+      "epoch": 0.30883786012343517,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.769912043361606e-06,
+      "loss": 1.0527,
+      "step": 1545
+    },
+    {
+      "epoch": 0.30903775517853127,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.769595941831066e-06,
+      "loss": 1.185,
+      "step": 1546
+    },
+    {
+      "epoch": 0.30923765023362737,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.76927962843558e-06,
+      "loss": 1.0581,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3094375452887234,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.768963103189206e-06,
+      "loss": 1.1129,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3096374403438195,
+      "grad_norm": 2.125,
+      "learning_rate": 9.768646366105997e-06,
+      "loss": 1.1202,
+      "step": 1549
+    },
+    {
+      "epoch": 0.30983733539891556,
+      "grad_norm": 2.125,
+      "learning_rate": 9.768329417200029e-06,
+      "loss": 1.1176,
+      "step": 1550
+    },
+    {
+      "epoch": 0.31003723045401166,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.768012256485376e-06,
+      "loss": 1.0759,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3102371255091077,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.767694883976128e-06,
+      "loss": 1.1666,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3104370205642038,
+      "grad_norm": 1.921875,
+      "learning_rate": 9.767377299686382e-06,
+      "loss": 1.0544,
+      "step": 1553
+    },
+    {
+      "epoch": 0.31063691561929985,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.767059503630247e-06,
+      "loss": 1.1891,
+      "step": 1554
+    },
+    {
+      "epoch": 0.31083681067439595,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.766741495821838e-06,
+      "loss": 1.1982,
+      "step": 1555
+    },
+    {
+      "epoch": 0.311036705729492,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.76642327627528e-06,
+      "loss": 1.0769,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3112366007845881,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.766104845004709e-06,
+      "loss": 1.1212,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3114364958396842,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.76578620202427e-06,
+      "loss": 1.0862,
+      "step": 1558
+    },
+    {
+      "epoch": 0.31163639089478024,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.765467347348116e-06,
+      "loss": 1.0631,
+      "step": 1559
+    },
+    {
+      "epoch": 0.31183628594987634,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.765148280990412e-06,
+      "loss": 1.0448,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3120361810049724,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.76482900296533e-06,
+      "loss": 1.0488,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3122360760600685,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.764509513287054e-06,
+      "loss": 1.0261,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3124359711151645,
+      "grad_norm": 2.125,
+      "learning_rate": 9.764189811969773e-06,
+      "loss": 1.0469,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3126358661702606,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.763869899027689e-06,
+      "loss": 1.0626,
+      "step": 1564
+    },
+    {
+      "epoch": 0.31283576122535667,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.763549774475014e-06,
+      "loss": 1.0377,
+      "step": 1565
+    },
+    {
+      "epoch": 0.31303565628045277,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.763229438325968e-06,
+      "loss": 0.9987,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3132355513355488,
+      "grad_norm": 2.125,
+      "learning_rate": 9.762908890594777e-06,
+      "loss": 1.0444,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3134354463906449,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.762588131295681e-06,
+      "loss": 1.1318,
+      "step": 1568
+    },
+    {
+      "epoch": 0.313635341445741,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.76226716044293e-06,
+      "loss": 1.0587,
+      "step": 1569
+    },
+    {
+      "epoch": 0.31383523650083706,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.761945978050782e-06,
+      "loss": 1.1694,
+      "step": 1570
+    },
+    {
+      "epoch": 0.31403513155593316,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.7616245841335e-06,
+      "loss": 1.0964,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3142350266110292,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.761302978705364e-06,
+      "loss": 1.0968,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3144349216661253,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.760981161780657e-06,
+      "loss": 1.0919,
+      "step": 1573
+    },
+    {
+      "epoch": 0.31463481672122134,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.760659133373675e-06,
+      "loss": 1.0793,
+      "step": 1574
+    },
+    {
+      "epoch": 0.31483471177631744,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.760336893498724e-06,
+      "loss": 1.1369,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3150346068314135,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.760014442170116e-06,
+      "loss": 1.2367,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3152345018865096,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.759691779402175e-06,
+      "loss": 1.0322,
+      "step": 1577
+    },
+    {
+      "epoch": 0.31543439694160563,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.759368905209234e-06,
+      "loss": 1.2766,
+      "step": 1578
+    },
+    {
+      "epoch": 0.31563429199670173,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.759045819605635e-06,
+      "loss": 1.1224,
+      "step": 1579
+    },
+    {
+      "epoch": 0.31583418705179783,
+      "grad_norm": 2.125,
+      "learning_rate": 9.758722522605727e-06,
+      "loss": 1.1049,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3160340821068939,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.758399014223874e-06,
+      "loss": 1.1412,
+      "step": 1581
+    },
+    {
+      "epoch": 0.31623397716199,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.758075294474445e-06,
+      "loss": 1.0146,
+      "step": 1582
+    },
+    {
+      "epoch": 0.316433872217086,
+      "grad_norm": 2.125,
+      "learning_rate": 9.75775136337182e-06,
+      "loss": 1.0551,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3166337672721821,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.757427220930387e-06,
+      "loss": 1.1061,
+      "step": 1584
+    },
+    {
+      "epoch": 0.31683366232727816,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.757102867164544e-06,
+      "loss": 1.0484,
+      "step": 1585
+    },
+    {
+      "epoch": 0.31703355738237426,
+      "grad_norm": 2.625,
+      "learning_rate": 9.756778302088701e-06,
+      "loss": 1.0879,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3172334524374703,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.756453525717274e-06,
+      "loss": 1.0491,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3174333474925664,
+      "grad_norm": 2.125,
+      "learning_rate": 9.75612853806469e-06,
+      "loss": 1.049,
+      "step": 1588
+    },
+    {
+      "epoch": 0.31763324254766245,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.755803339145382e-06,
+      "loss": 1.1043,
+      "step": 1589
+    },
+    {
+      "epoch": 0.31783313760275855,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.755477928973797e-06,
+      "loss": 1.0747,
+      "step": 1590
+    },
+    {
+      "epoch": 0.31803303265785465,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.755152307564393e-06,
+      "loss": 1.031,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3182329277129507,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.75482647493163e-06,
+      "loss": 1.1383,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3184328227680468,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.754500431089984e-06,
+      "loss": 1.1015,
+      "step": 1593
+    },
+    {
+      "epoch": 0.31863271782314284,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.754174176053936e-06,
+      "loss": 1.083,
+      "step": 1594
+    },
+    {
+      "epoch": 0.31883261287823894,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.75384770983798e-06,
+      "loss": 1.004,
+      "step": 1595
+    },
+    {
+      "epoch": 0.319032507933335,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.753521032456615e-06,
+      "loss": 1.0074,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3192324029884311,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.753194143924354e-06,
+      "loss": 1.1477,
+      "step": 1597
+    },
+    {
+      "epoch": 0.31943229804352713,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.752867044255716e-06,
+      "loss": 1.0784,
+      "step": 1598
+    },
+    {
+      "epoch": 0.31963219309862323,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.752539733465231e-06,
+      "loss": 1.0531,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3198320881537193,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.75221221156744e-06,
+      "loss": 1.08,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3200319832088154,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.75188447857689e-06,
+      "loss": 1.093,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3202318782639115,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.75155653450814e-06,
+      "loss": 1.1182,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3204317733190075,
+      "grad_norm": 2.125,
+      "learning_rate": 9.751228379375754e-06,
+      "loss": 1.058,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3206316683741036,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.750900013194312e-06,
+      "loss": 1.0517,
+      "step": 1604
+    },
+    {
+      "epoch": 0.32083156342919966,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.750571435978399e-06,
+      "loss": 0.9927,
+      "step": 1605
+    },
+    {
+      "epoch": 0.32103145848429576,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.750242647742609e-06,
+      "loss": 1.1342,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3212313535393918,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.74991364850155e-06,
+      "loss": 1.0087,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3214312485944879,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.749584438269833e-06,
+      "loss": 1.1029,
+      "step": 1608
+    },
+    {
+      "epoch": 0.32163114364958395,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.749255017062081e-06,
+      "loss": 1.1581,
+      "step": 1609
+    },
+    {
+      "epoch": 0.32183103870468005,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.74892538489293e-06,
+      "loss": 1.0755,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3220309337597761,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.748595541777021e-06,
+      "loss": 1.1912,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3222308288148722,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.748265487729003e-06,
+      "loss": 1.165,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3224307238699683,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.747935222763542e-06,
+      "loss": 0.9882,
+      "step": 1613
+    },
+    {
+      "epoch": 0.32263061892506434,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.747604746895303e-06,
+      "loss": 1.0114,
+      "step": 1614
+    },
+    {
+      "epoch": 0.32283051398016044,
+      "grad_norm": 2.125,
+      "learning_rate": 9.747274060138971e-06,
+      "loss": 1.0804,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3230304090352565,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.74694316250923e-06,
+      "loss": 0.9537,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3232303040903526,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.74661205402078e-06,
+      "loss": 1.1196,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3234301991454486,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.746280734688332e-06,
+      "loss": 1.01,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3236300942005447,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.7459492045266e-06,
+      "loss": 1.1002,
+      "step": 1619
+    },
+    {
+      "epoch": 0.32382998925564077,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.74561746355031e-06,
+      "loss": 1.0775,
+      "step": 1620
+    },
+    {
+      "epoch": 0.32402988431073687,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.7452855117742e-06,
+      "loss": 1.1096,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3242297793658329,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.744953349213016e-06,
+      "loss": 1.0339,
+      "step": 1622
+    },
+    {
+      "epoch": 0.324429674420929,
+      "grad_norm": 2.125,
+      "learning_rate": 9.74462097588151e-06,
+      "loss": 1.0775,
+      "step": 1623
+    },
+    {
+      "epoch": 0.32462956947602506,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.744288391794446e-06,
+      "loss": 1.091,
+      "step": 1624
+    },
+    {
+      "epoch": 0.32482946453112116,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.743955596966597e-06,
+      "loss": 1.037,
+      "step": 1625
+    },
+    {
+      "epoch": 0.32502935958621726,
+      "grad_norm": 2.125,
+      "learning_rate": 9.743622591412749e-06,
+      "loss": 1.0736,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3252292546413133,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.743289375147693e-06,
+      "loss": 1.0584,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3254291496964094,
+      "grad_norm": 2.125,
+      "learning_rate": 9.742955948186228e-06,
+      "loss": 1.0357,
+      "step": 1628
+    },
+    {
+      "epoch": 0.32562904475150545,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.742622310543165e-06,
+      "loss": 1.0085,
+      "step": 1629
+    },
+    {
+      "epoch": 0.32582893980660155,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.742288462233329e-06,
+      "loss": 1.0853,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3260288348616976,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.741954403271543e-06,
+      "loss": 1.0205,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3262287299167937,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.741620133672651e-06,
+      "loss": 1.0081,
+      "step": 1632
+    },
+    {
+      "epoch": 0.32642862497188974,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.741285653451497e-06,
+      "loss": 1.0474,
+      "step": 1633
+    },
+    {
+      "epoch": 0.32662852002698584,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.740950962622943e-06,
+      "loss": 1.1232,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3268284150820819,
+      "grad_norm": 2.125,
+      "learning_rate": 9.740616061201852e-06,
+      "loss": 1.0578,
+      "step": 1635
+    },
+    {
+      "epoch": 0.327028310137178,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.740280949203102e-06,
+      "loss": 1.1024,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3272282051922741,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.739945626641579e-06,
+      "loss": 1.1667,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3274281002473701,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.739610093532176e-06,
+      "loss": 1.0927,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3276279953024662,
+      "grad_norm": 2.125,
+      "learning_rate": 9.739274349889802e-06,
+      "loss": 1.0491,
+      "step": 1639
+    },
+    {
+      "epoch": 0.32782789035756227,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.738938395729364e-06,
+      "loss": 0.9977,
+      "step": 1640
+    },
+    {
+      "epoch": 0.32802778541265837,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.738602231065793e-06,
+      "loss": 0.9795,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3282276804677544,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.738265855914014e-06,
+      "loss": 1.1082,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3284275755228505,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.73792927028897e-06,
+      "loss": 1.046,
+      "step": 1643
+    },
+    {
+      "epoch": 0.32862747057794656,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.737592474205617e-06,
+      "loss": 1.0181,
+      "step": 1644
+    },
+    {
+      "epoch": 0.32882736563304266,
+      "grad_norm": 2.125,
+      "learning_rate": 9.73725546767891e-06,
+      "loss": 1.0592,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3290272606881387,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.736918250723823e-06,
+      "loss": 1.0517,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3292271557432348,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.736580823355333e-06,
+      "loss": 1.0778,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3294270507983309,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.736243185588428e-06,
+      "loss": 1.038,
+      "step": 1648
+    },
+    {
+      "epoch": 0.32962694585342694,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.735905337438107e-06,
+      "loss": 1.0171,
+      "step": 1649
+    },
+    {
+      "epoch": 0.32982684090852304,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.735567278919376e-06,
+      "loss": 1.1386,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3300267359636191,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.735229010047253e-06,
+      "loss": 1.0989,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3302266310187152,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.734890530836763e-06,
+      "loss": 1.0239,
+      "step": 1652
+    },
+    {
+      "epoch": 0.33042652607381123,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.734551841302941e-06,
+      "loss": 1.0697,
+      "step": 1653
+    },
+    {
+      "epoch": 0.33062642112890733,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.734212941460833e-06,
+      "loss": 1.1375,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3308263161840034,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.73387383132549e-06,
+      "loss": 0.9819,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3310262112390995,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.733534510911977e-06,
+      "loss": 0.9949,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3312261062941955,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.733194980235367e-06,
+      "loss": 1.0075,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3314260013492916,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.732855239310743e-06,
+      "loss": 1.1145,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3316258964043877,
+      "grad_norm": 2.25,
+      "learning_rate": 9.732515288153193e-06,
+      "loss": 1.0824,
+      "step": 1659
+    },
+    {
+      "epoch": 0.33182579145948377,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.732175126777821e-06,
+      "loss": 1.1481,
+      "step": 1660
+    },
+    {
+      "epoch": 0.33202568651457987,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.731834755199734e-06,
+      "loss": 0.9968,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3322255815696759,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.731494173434053e-06,
+      "loss": 1.053,
+      "step": 1662
+    },
+    {
+      "epoch": 0.332425476624772,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.731153381495905e-06,
+      "loss": 1.0219,
+      "step": 1663
+    },
+    {
+      "epoch": 0.33262537167986805,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.730812379400432e-06,
+      "loss": 1.0924,
+      "step": 1664
+    },
+    {
+      "epoch": 0.33282526673496415,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.730471167162776e-06,
+      "loss": 1.0849,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3330251617900602,
+      "grad_norm": 2.125,
+      "learning_rate": 9.730129744798096e-06,
+      "loss": 1.2105,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3332250568451563,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.729788112321558e-06,
+      "loss": 1.0258,
+      "step": 1667
+    },
+    {
+      "epoch": 0.33342495190025234,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.729446269748338e-06,
+      "loss": 1.1197,
+      "step": 1668
+    },
+    {
+      "epoch": 0.33362484695534844,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.729104217093618e-06,
+      "loss": 1.0676,
+      "step": 1669
+    },
+    {
+      "epoch": 0.33382474201044454,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.728761954372597e-06,
+      "loss": 1.0351,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3340246370655406,
+      "grad_norm": 2.25,
+      "learning_rate": 9.728419481600472e-06,
+      "loss": 1.1705,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3342245321206367,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.728076798792461e-06,
+      "loss": 1.1647,
+      "step": 1672
+    },
+    {
+      "epoch": 0.33442442717573273,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.727733905963783e-06,
+      "loss": 1.006,
+      "step": 1673
+    },
+    {
+      "epoch": 0.33462432223082883,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.727390803129668e-06,
+      "loss": 1.0953,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3348242172859249,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.72704749030536e-06,
+      "loss": 1.0583,
+      "step": 1675
+    },
+    {
+      "epoch": 0.335024112341021,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.726703967506107e-06,
+      "loss": 1.0887,
+      "step": 1676
+    },
+    {
+      "epoch": 0.335224007396117,
+      "grad_norm": 2.125,
+      "learning_rate": 9.726360234747168e-06,
+      "loss": 0.9653,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3354239024512131,
+      "grad_norm": 2.125,
+      "learning_rate": 9.726016292043814e-06,
+      "loss": 0.9459,
+      "step": 1678
+    },
+    {
+      "epoch": 0.33562379750630916,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.725672139411319e-06,
+      "loss": 1.0816,
+      "step": 1679
+    },
+    {
+      "epoch": 0.33582369256140526,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.725327776864974e-06,
+      "loss": 1.0573,
+      "step": 1680
+    },
+    {
+      "epoch": 0.33602358761650136,
+      "grad_norm": 2.25,
+      "learning_rate": 9.724983204420073e-06,
+      "loss": 1.0713,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3362234826715974,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.724638422091922e-06,
+      "loss": 1.0342,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3364233777266935,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.724293429895836e-06,
+      "loss": 1.111,
+      "step": 1683
+    },
+    {
+      "epoch": 0.33662327278178955,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.723948227847145e-06,
+      "loss": 0.9647,
+      "step": 1684
+    },
+    {
+      "epoch": 0.33682316783688565,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.723602815961173e-06,
+      "loss": 1.0749,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3370230628919817,
+      "grad_norm": 2.125,
+      "learning_rate": 9.723257194253272e-06,
+      "loss": 1.0911,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3372229579470778,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.72291136273879e-06,
+      "loss": 1.0592,
+      "step": 1687
+    },
+    {
+      "epoch": 0.33742285300217384,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.722565321433088e-06,
+      "loss": 1.043,
+      "step": 1688
+    },
+    {
+      "epoch": 0.33762274805726994,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.722219070351541e-06,
+      "loss": 1.1058,
+      "step": 1689
+    },
+    {
+      "epoch": 0.337822643112366,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.721872609509526e-06,
+      "loss": 1.1402,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3380225381674621,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.721525938922434e-06,
+      "loss": 1.1047,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3382224332225582,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.721179058605664e-06,
+      "loss": 1.0193,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3384223282776542,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.720831968574625e-06,
+      "loss": 1.0501,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3386222233327503,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.720484668844734e-06,
+      "loss": 1.0406,
+      "step": 1694
+    },
+    {
+      "epoch": 0.33882211838784637,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.720137159431418e-06,
+      "loss": 1.0744,
+      "step": 1695
+    },
+    {
+      "epoch": 0.33902201344294247,
+      "grad_norm": 2.125,
+      "learning_rate": 9.719789440350113e-06,
+      "loss": 1.0918,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3392219084980385,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.719441511616266e-06,
+      "loss": 1.0955,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3394218035531346,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.719093373245331e-06,
+      "loss": 1.0195,
+      "step": 1698
+    },
+    {
+      "epoch": 0.33962169860823066,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.718745025252773e-06,
+      "loss": 1.0618,
+      "step": 1699
+    },
+    {
+      "epoch": 0.33982159366332676,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.718396467654064e-06,
+      "loss": 1.1286,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3400214887184228,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.718047700464688e-06,
+      "loss": 1.0813,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3402213837735189,
+      "grad_norm": 2.125,
+      "learning_rate": 9.717698723700137e-06,
+      "loss": 1.0753,
+      "step": 1702
+    },
+    {
+      "epoch": 0.340421278828615,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.717349537375913e-06,
+      "loss": 1.1358,
+      "step": 1703
+    },
+    {
+      "epoch": 0.34062117388371105,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.717000141507527e-06,
+      "loss": 0.9364,
+      "step": 1704
+    },
+    {
+      "epoch": 0.34082106893880715,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.716650536110496e-06,
+      "loss": 1.0648,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3410209639939032,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.716300721200354e-06,
+      "loss": 1.0959,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3412208590489993,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.715950696792638e-06,
+      "loss": 1.009,
+      "step": 1707
+    },
+    {
+      "epoch": 0.34142075410409534,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.715600462902895e-06,
+      "loss": 1.051,
+      "step": 1708
+    },
+    {
+      "epoch": 0.34162064915919144,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.715250019546683e-06,
+      "loss": 0.999,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3418205442142875,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.714899366739569e-06,
+      "loss": 0.9646,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3420204392693836,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.714548504497128e-06,
+      "loss": 1.0364,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3422203343244796,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.714197432834947e-06,
+      "loss": 1.0647,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3424202293795757,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.713846151768618e-06,
+      "loss": 1.12,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3426201244346718,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.713494661313746e-06,
+      "loss": 1.033,
+      "step": 1714
+    },
+    {
+      "epoch": 0.34282001948976787,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.713142961485947e-06,
+      "loss": 1.0881,
+      "step": 1715
+    },
+    {
+      "epoch": 0.34301991454486397,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.71279105230084e-06,
+      "loss": 1.1303,
+      "step": 1716
+    },
+    {
+      "epoch": 0.34321980959996,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.712438933774057e-06,
+      "loss": 1.1371,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3434197046550561,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.71208660592124e-06,
+      "loss": 0.968,
+      "step": 1718
+    },
+    {
+      "epoch": 0.34361959971015216,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.71173406875804e-06,
+      "loss": 1.0274,
+      "step": 1719
+    },
+    {
+      "epoch": 0.34381949476524826,
+      "grad_norm": 2.25,
+      "learning_rate": 9.711381322300117e-06,
+      "loss": 0.9893,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3440193898203443,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.711028366563138e-06,
+      "loss": 1.171,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3442192848754404,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.71067520156278e-06,
+      "loss": 1.1283,
+      "step": 1722
+    },
+    {
+      "epoch": 0.34441917993053645,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.710321827314735e-06,
+      "loss": 1.0001,
+      "step": 1723
+    },
+    {
+      "epoch": 0.34461907498563255,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.709968243834698e-06,
+      "loss": 1.0845,
+      "step": 1724
+    },
+    {
+      "epoch": 0.34481897004072865,
+      "grad_norm": 2.125,
+      "learning_rate": 9.709614451138373e-06,
+      "loss": 1.0823,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3450188650958247,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.709260449241478e-06,
+      "loss": 1.058,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3452187601509208,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.708906238159736e-06,
+      "loss": 1.1112,
+      "step": 1727
+    },
+    {
+      "epoch": 0.34541865520601683,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.708551817908883e-06,
+      "loss": 1.1158,
+      "step": 1728
+    },
+    {
+      "epoch": 0.34561855026111293,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.70819718850466e-06,
+      "loss": 1.0908,
+      "step": 1729
+    },
+    {
+      "epoch": 0.345818445316209,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.707842349962821e-06,
+      "loss": 1.1913,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3460183403713051,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.707487302299128e-06,
+      "loss": 1.0317,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3462182354264011,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.707132045529352e-06,
+      "loss": 1.1217,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3464181304814972,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.706776579669272e-06,
+      "loss": 1.0175,
+      "step": 1733
+    },
+    {
+      "epoch": 0.34661802553659327,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.70642090473468e-06,
+      "loss": 1.1629,
+      "step": 1734
+    },
+    {
+      "epoch": 0.34681792059168937,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.706065020741375e-06,
+      "loss": 1.0368,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3470178156467854,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.705708927705163e-06,
+      "loss": 1.0294,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3472177107018815,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.705352625641863e-06,
+      "loss": 1.0309,
+      "step": 1737
+    },
+    {
+      "epoch": 0.3474176057569776,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.704996114567302e-06,
+      "loss": 1.079,
+      "step": 1738
+    },
+    {
+      "epoch": 0.34761750081207365,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.704639394497317e-06,
+      "loss": 1.0315,
+      "step": 1739
+    },
+    {
+      "epoch": 0.34781739586716975,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.704282465447752e-06,
+      "loss": 1.1318,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3480172909222658,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.703925327434462e-06,
+      "loss": 1.0676,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3482171859773619,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.703567980473311e-06,
+      "loss": 1.0814,
+      "step": 1742
+    },
+    {
+      "epoch": 0.34841708103245794,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.703210424580174e-06,
+      "loss": 1.0813,
+      "step": 1743
+    },
+    {
+      "epoch": 0.34861697608755404,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.702852659770932e-06,
+      "loss": 1.1354,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3488168711426501,
+      "grad_norm": 2.125,
+      "learning_rate": 9.702494686061477e-06,
+      "loss": 0.9704,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3490167661977462,
+      "grad_norm": 2.125,
+      "learning_rate": 9.70213650346771e-06,
+      "loss": 1.0792,
+      "step": 1746
+    },
+    {
+      "epoch": 0.34921666125284223,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.70177811200554e-06,
+      "loss": 1.0346,
+      "step": 1747
+    },
+    {
+      "epoch": 0.34941655630793833,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.701419511690891e-06,
+      "loss": 1.1517,
+      "step": 1748
+    },
+    {
+      "epoch": 0.34961645136303443,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.701060702539689e-06,
+      "loss": 1.1122,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3498163464181305,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.700701684567872e-06,
+      "loss": 0.995,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3500162414732266,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.700342457791387e-06,
+      "loss": 1.0151,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3502161365283226,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.699983022226194e-06,
+      "loss": 1.1168,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3504160315834187,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.699623377888256e-06,
+      "loss": 1.0673,
+      "step": 1753
+    },
+    {
+      "epoch": 0.35061592663851476,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.699263524793551e-06,
+      "loss": 1.0326,
+      "step": 1754
+    },
+    {
+      "epoch": 0.35081582169361086,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.69890346295806e-06,
+      "loss": 1.1526,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3510157167487069,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.69854319239778e-06,
+      "loss": 1.101,
+      "step": 1756
+    },
+    {
+      "epoch": 0.351215611803803,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.698182713128714e-06,
+      "loss": 1.1748,
+      "step": 1757
+    },
+    {
+      "epoch": 0.35141550685889905,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.697822025166874e-06,
+      "loss": 1.1522,
+      "step": 1758
+    },
+    {
+      "epoch": 0.35161540191399515,
+      "grad_norm": 2.25,
+      "learning_rate": 9.697461128528281e-06,
+      "loss": 1.0798,
+      "step": 1759
+    },
+    {
+      "epoch": 0.35181529696909125,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.697100023228966e-06,
+      "loss": 1.0741,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3520151920241873,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.69673870928497e-06,
+      "loss": 1.0234,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3522150870792834,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.69637718671234e-06,
+      "loss": 1.1072,
+      "step": 1762
+    },
+    {
+      "epoch": 0.35241498213437944,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.696015455527139e-06,
+      "loss": 1.0557,
+      "step": 1763
+    },
+    {
+      "epoch": 0.35261487718947554,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.695653515745433e-06,
+      "loss": 1.0192,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3528147722445716,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.6952913673833e-06,
+      "loss": 1.0337,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3530146672996677,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.694929010456824e-06,
+      "loss": 1.0215,
+      "step": 1766
+    },
+    {
+      "epoch": 0.35321456235476373,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.694566444982104e-06,
+      "loss": 1.1584,
+      "step": 1767
+    },
+    {
+      "epoch": 0.35341445740985983,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.694203670975244e-06,
+      "loss": 1.0854,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3536143524649559,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.693840688452358e-06,
+      "loss": 1.1391,
+      "step": 1769
+    },
+    {
+      "epoch": 0.35381424752005197,
+      "grad_norm": 2.25,
+      "learning_rate": 9.69347749742957e-06,
+      "loss": 0.9615,
+      "step": 1770
+    },
+    {
+      "epoch": 0.35401414257514807,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.693114097923013e-06,
+      "loss": 1.0806,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3542140376302441,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.692750489948829e-06,
+      "loss": 1.1137,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3544139326853402,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.692386673523168e-06,
+      "loss": 1.004,
+      "step": 1773
+    },
+    {
+      "epoch": 0.35461382774043626,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.692022648662193e-06,
+      "loss": 1.0561,
+      "step": 1774
+    },
+    {
+      "epoch": 0.35481372279553236,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.691658415382073e-06,
+      "loss": 1.003,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3550136178506284,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.691293973698988e-06,
+      "loss": 1.0291,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3552135129057245,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.690929323629126e-06,
+      "loss": 1.0943,
+      "step": 1777
+    },
+    {
+      "epoch": 0.35541340796082055,
+      "grad_norm": 2.125,
+      "learning_rate": 9.690564465188684e-06,
+      "loss": 1.1177,
+      "step": 1778
+    },
+    {
+      "epoch": 0.35561330301591665,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.69019939839387e-06,
+      "loss": 1.0692,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3558131980710127,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.689834123260896e-06,
+      "loss": 1.1017,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3560130931261088,
+      "grad_norm": 2.25,
+      "learning_rate": 9.689468639805996e-06,
+      "loss": 1.0774,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3562129881812049,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.689102948045398e-06,
+      "loss": 1.0658,
+      "step": 1782
+    },
+    {
+      "epoch": 0.35641288323630094,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.688737047995349e-06,
+      "loss": 1.0039,
+      "step": 1783
+    },
+    {
+      "epoch": 0.35661277829139704,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.688370939672099e-06,
+      "loss": 1.0924,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3568126733464931,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.688004623091914e-06,
+      "loss": 1.1315,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3570125684015892,
+      "grad_norm": 2.25,
+      "learning_rate": 9.687638098271064e-06,
+      "loss": 1.1243,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3572124634566852,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.68727136522583e-06,
+      "loss": 1.0848,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3574123585117813,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.686904423972502e-06,
+      "loss": 1.0546,
+      "step": 1788
+    },
+    {
+      "epoch": 0.35761225356687737,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.68653727452738e-06,
+      "loss": 1.0223,
+      "step": 1789
+    },
+    {
+      "epoch": 0.35781214862197347,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.686169916906773e-06,
+      "loss": 1.0758,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3580120436770695,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.685802351127e-06,
+      "loss": 1.0396,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3582119387321656,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.685434577204387e-06,
+      "loss": 1.0373,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3584118337872617,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.68506659515527e-06,
+      "loss": 1.0745,
+      "step": 1793
+    },
+    {
+      "epoch": 0.35861172884235776,
+      "grad_norm": 2.25,
+      "learning_rate": 9.684698404995995e-06,
+      "loss": 1.0851,
+      "step": 1794
+    },
+    {
+      "epoch": 0.35881162389745386,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.684330006742916e-06,
+      "loss": 1.0869,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3590115189525499,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.6839614004124e-06,
+      "loss": 1.1189,
+      "step": 1796
+    },
+    {
+      "epoch": 0.359211414007646,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.68359258602082e-06,
+      "loss": 1.0061,
+      "step": 1797
+    },
+    {
+      "epoch": 0.35941130906274205,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.683223563584556e-06,
+      "loss": 1.0797,
+      "step": 1798
+    },
+    {
+      "epoch": 0.35961120411783815,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.682854333120002e-06,
+      "loss": 1.0724,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3598110991729342,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.682484894643558e-06,
+      "loss": 1.0145,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3600109942280303,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.682115248171635e-06,
+      "loss": 1.1822,
+      "step": 1801
+    },
+    {
+      "epoch": 0.36021088928312633,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.681745393720654e-06,
+      "loss": 1.1737,
+      "step": 1802
+    },
+    {
+      "epoch": 0.36041078433822243,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.68137533130704e-06,
+      "loss": 1.1308,
+      "step": 1803
+    },
+    {
+      "epoch": 0.36061067939331853,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.681005060947237e-06,
+      "loss": 0.9999,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3608105744484146,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.680634582657688e-06,
+      "loss": 1.0311,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3610104695035107,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.68026389645485e-06,
+      "loss": 1.0794,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3612103645586067,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.67989300235519e-06,
+      "loss": 1.0,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3614102596137028,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.679521900375183e-06,
+      "loss": 1.0749,
+      "step": 1808
+    },
+    {
+      "epoch": 0.36161015466879887,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.679150590531312e-06,
+      "loss": 1.1228,
+      "step": 1809
+    },
+    {
+      "epoch": 0.36181004972389497,
+      "grad_norm": 2.125,
+      "learning_rate": 9.678779072840072e-06,
+      "loss": 1.0755,
+      "step": 1810
+    },
+    {
+      "epoch": 0.362009944778991,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.678407347317967e-06,
+      "loss": 1.1757,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3622098398340871,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.678035413981505e-06,
+      "loss": 0.97,
+      "step": 1812
+    },
+    {
+      "epoch": 0.36240973488918315,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.677663272847211e-06,
+      "loss": 0.9848,
+      "step": 1813
+    },
+    {
+      "epoch": 0.36260962994427925,
+      "grad_norm": 2.125,
+      "learning_rate": 9.677290923931613e-06,
+      "loss": 1.1265,
+      "step": 1814
+    },
+    {
+      "epoch": 0.36280952499937535,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.676918367251251e-06,
+      "loss": 1.0995,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3630094200544714,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.676545602822675e-06,
+      "loss": 1.0715,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3632093151095675,
+      "grad_norm": 2.125,
+      "learning_rate": 9.676172630662446e-06,
+      "loss": 1.0607,
+      "step": 1817
+    },
+    {
+      "epoch": 0.36340921016466354,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.675799450787125e-06,
+      "loss": 1.0595,
+      "step": 1818
+    },
+    {
+      "epoch": 0.36360910521975964,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.675426063213291e-06,
+      "loss": 1.0687,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3638090002748557,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.675052467957533e-06,
+      "loss": 1.0903,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3640088953299518,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.674678665036443e-06,
+      "loss": 1.1392,
+      "step": 1821
+    },
+    {
+      "epoch": 0.36420879038504783,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.674304654466625e-06,
+      "loss": 1.0846,
+      "step": 1822
+    },
+    {
+      "epoch": 0.36440868544014393,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.673930436264695e-06,
+      "loss": 1.1428,
+      "step": 1823
+    },
+    {
+      "epoch": 0.36460858049524,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.673556010447274e-06,
+      "loss": 1.0965,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3648084755503361,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.673181377030993e-06,
+      "loss": 1.0701,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3650083706054322,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.672806536032495e-06,
+      "loss": 1.0332,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3652082656605282,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.67243148746843e-06,
+      "loss": 1.0372,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3654081607156243,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.672056231355455e-06,
+      "loss": 1.1678,
+      "step": 1828
+    },
+    {
+      "epoch": 0.36560805577072036,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.671680767710244e-06,
+      "loss": 1.0088,
+      "step": 1829
+    },
+    {
+      "epoch": 0.36580795082581646,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.671305096549473e-06,
+      "loss": 1.1343,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3660078458809125,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.670929217889825e-06,
+      "loss": 0.9761,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3662077409360086,
+      "grad_norm": 2.125,
+      "learning_rate": 9.670553131748003e-06,
+      "loss": 1.0866,
+      "step": 1832
+    },
+    {
+      "epoch": 0.36640763599110465,
+      "grad_norm": 2.25,
+      "learning_rate": 9.670176838140708e-06,
+      "loss": 1.1274,
+      "step": 1833
+    },
+    {
+      "epoch": 0.36660753104620075,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.669800337084658e-06,
+      "loss": 1.1767,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3668074261012968,
+      "grad_norm": 2.25,
+      "learning_rate": 9.669423628596574e-06,
+      "loss": 1.0194,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3670073211563929,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.669046712693192e-06,
+      "loss": 0.9397,
+      "step": 1836
+    },
+    {
+      "epoch": 0.367207216211489,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.668669589391256e-06,
+      "loss": 1.0062,
+      "step": 1837
+    },
+    {
+      "epoch": 0.36740711126658504,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.668292258707513e-06,
+      "loss": 1.0522,
+      "step": 1838
+    },
+    {
+      "epoch": 0.36760700632168114,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.667914720658726e-06,
+      "loss": 1.0841,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3678069013767772,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.667536975261667e-06,
+      "loss": 1.1204,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3680067964318733,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.667159022533115e-06,
+      "loss": 1.0638,
+      "step": 1841
+    },
+    {
+      "epoch": 0.36820669148696933,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.666780862489856e-06,
+      "loss": 1.035,
+      "step": 1842
+    },
+    {
+      "epoch": 0.36840658654206543,
+      "grad_norm": 2.125,
+      "learning_rate": 9.66640249514869e-06,
+      "loss": 1.2266,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3686064815971615,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.666023920526423e-06,
+      "loss": 1.0598,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3688063766522576,
+      "grad_norm": 2.125,
+      "learning_rate": 9.665645138639872e-06,
+      "loss": 1.0532,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3690062717073536,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.665266149505863e-06,
+      "loss": 1.0488,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3692061667624497,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.664886953141228e-06,
+      "loss": 1.0693,
+      "step": 1847
+    },
+    {
+      "epoch": 0.36940606181754576,
+      "grad_norm": 2.125,
+      "learning_rate": 9.664507549562814e-06,
+      "loss": 1.016,
+      "step": 1848
+    },
+    {
+      "epoch": 0.36960595687264186,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.664127938787473e-06,
+      "loss": 1.0602,
+      "step": 1849
+    },
+    {
+      "epoch": 0.36980585192773796,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.663748120832069e-06,
+      "loss": 1.1527,
+      "step": 1850
+    },
+    {
+      "epoch": 0.370005746982834,
+      "grad_norm": 2.125,
+      "learning_rate": 9.663368095713468e-06,
+      "loss": 0.9813,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3702056420379301,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.662987863448556e-06,
+      "loss": 1.065,
+      "step": 1852
+    },
+    {
+      "epoch": 0.37040553709302615,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.662607424054221e-06,
+      "loss": 1.0928,
+      "step": 1853
+    },
+    {
+      "epoch": 0.37060543214812225,
+      "grad_norm": 2.125,
+      "learning_rate": 9.662226777547363e-06,
+      "loss": 1.0619,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3708053272032183,
+      "grad_norm": 2.125,
+      "learning_rate": 9.661845923944888e-06,
+      "loss": 1.1729,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3710052222583144,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.661464863263717e-06,
+      "loss": 1.0977,
+      "step": 1856
+    },
+    {
+      "epoch": 0.37120511731341044,
+      "grad_norm": 2.25,
+      "learning_rate": 9.661083595520772e-06,
+      "loss": 1.0552,
+      "step": 1857
+    },
+    {
+      "epoch": 0.37140501236850654,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.660702120732992e-06,
+      "loss": 1.1044,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3716049074236026,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.660320438917323e-06,
+      "loss": 1.0601,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3718048024786987,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.659938550090717e-06,
+      "loss": 1.0558,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3720046975337948,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.659556454270139e-06,
+      "loss": 1.0663,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3722045925888908,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.659174151472562e-06,
+      "loss": 1.0182,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3724044876439869,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.658791641714965e-06,
+      "loss": 1.0283,
+      "step": 1863
+    },
+    {
+      "epoch": 0.37260438269908297,
+      "grad_norm": 2.125,
+      "learning_rate": 9.65840892501434e-06,
+      "loss": 1.0972,
+      "step": 1864
+    },
+    {
+      "epoch": 0.37280427775417907,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.65802600138769e-06,
+      "loss": 1.0751,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3730041728092751,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.65764287085202e-06,
+      "loss": 1.076,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3732040678643712,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.657259533424355e-06,
+      "loss": 1.0158,
+      "step": 1867
+    },
+    {
+      "epoch": 0.37340396291946726,
+      "grad_norm": 2.125,
+      "learning_rate": 9.656875989121715e-06,
+      "loss": 1.0534,
+      "step": 1868
+    },
+    {
+      "epoch": 0.37360385797456336,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.656492237961143e-06,
+      "loss": 1.062,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3738037530296594,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.656108279959684e-06,
+      "loss": 1.111,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3740036480847555,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.65572411513439e-06,
+      "loss": 1.1052,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3742035431398516,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.65533974350233e-06,
+      "loss": 1.0701,
+      "step": 1872
+    },
+    {
+      "epoch": 0.37440343819494765,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.654955165080575e-06,
+      "loss": 1.0559,
+      "step": 1873
+    },
+    {
+      "epoch": 0.37460333325004375,
+      "grad_norm": 1.890625,
+      "learning_rate": 9.65457037988621e-06,
+      "loss": 0.9771,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3748032283051398,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.654185387936323e-06,
+      "loss": 1.0589,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3750031233602359,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.653800189248022e-06,
+      "loss": 1.0885,
+      "step": 1876
+    },
+    {
+      "epoch": 0.37520301841533193,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.65341478383841e-06,
+      "loss": 0.9952,
+      "step": 1877
+    },
+    {
+      "epoch": 0.37540291347042803,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.653029171724612e-06,
+      "loss": 1.0742,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3756028085255241,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.652643352923756e-06,
+      "loss": 1.0288,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3758027035806202,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.652257327452978e-06,
+      "loss": 1.1885,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3760025986357162,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.651871095329425e-06,
+      "loss": 1.1295,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3762024936908123,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.651484656570257e-06,
+      "loss": 1.0479,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3764023887459084,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.651098011192638e-06,
+      "loss": 1.0396,
+      "step": 1883
+    },
+    {
+      "epoch": 0.37660228380100447,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.65071115921374e-06,
+      "loss": 1.1473,
+      "step": 1884
+    },
+    {
+      "epoch": 0.37680217885610057,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.65032410065075e-06,
+      "loss": 1.0292,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3770020739111966,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.64993683552086e-06,
+      "loss": 1.1086,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3772019689662927,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.649549363841273e-06,
+      "loss": 0.9923,
+      "step": 1887
+    },
+    {
+      "epoch": 0.37740186402138876,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.649161685629199e-06,
+      "loss": 1.0775,
+      "step": 1888
+    },
+    {
+      "epoch": 0.37760175907648486,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.64877380090186e-06,
+      "loss": 1.1234,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3778016541315809,
+      "grad_norm": 2.0,
+      "learning_rate": 9.648385709676485e-06,
+      "loss": 1.0188,
+      "step": 1890
+    },
+    {
+      "epoch": 0.378001549186677,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.647997411970313e-06,
+      "loss": 0.9725,
+      "step": 1891
+    },
+    {
+      "epoch": 0.37820144424177304,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.647608907800593e-06,
+      "loss": 1.0811,
+      "step": 1892
+    },
+    {
+      "epoch": 0.37840133929686914,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.647220197184582e-06,
+      "loss": 0.9877,
+      "step": 1893
+    },
+    {
+      "epoch": 0.37860123435196524,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.646831280139544e-06,
+      "loss": 1.1433,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3788011294070613,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.646442156682758e-06,
+      "loss": 1.1211,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3790010244621574,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.646052826831509e-06,
+      "loss": 0.996,
+      "step": 1896
+    },
+    {
+      "epoch": 0.37920091951725343,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.64566329060309e-06,
+      "loss": 1.088,
+      "step": 1897
+    },
+    {
+      "epoch": 0.37940081457234953,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.6452735480148e-06,
+      "loss": 1.0102,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3796007096274456,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.644883599083959e-06,
+      "loss": 1.072,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3798006046825417,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.644493443827883e-06,
+      "loss": 1.113,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3800004997376377,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.644103082263904e-06,
+      "loss": 1.0365,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3802003947927338,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.643712514409362e-06,
+      "loss": 1.1374,
+      "step": 1902
+    },
+    {
+      "epoch": 0.38040028984782986,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.643321740281606e-06,
+      "loss": 1.091,
+      "step": 1903
+    },
+    {
+      "epoch": 0.38060018490292596,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.642930759897995e-06,
+      "loss": 1.1503,
+      "step": 1904
+    },
+    {
+      "epoch": 0.38080007995802206,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.642539573275895e-06,
+      "loss": 1.1326,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3809999750131181,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.642148180432685e-06,
+      "loss": 1.0329,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3811998700682142,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.641756581385745e-06,
+      "loss": 1.1499,
+      "step": 1907
+    },
+    {
+      "epoch": 0.38139976512331025,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.641364776152477e-06,
+      "loss": 1.1574,
+      "step": 1908
+    },
+    {
+      "epoch": 0.38159966017840635,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.64097276475028e-06,
+      "loss": 1.0513,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3817995552335024,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.640580547196568e-06,
+      "loss": 1.0772,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3819994502885985,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.640188123508764e-06,
+      "loss": 1.0826,
+      "step": 1911
+    },
+    {
+      "epoch": 0.38219934534369454,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.639795493704299e-06,
+      "loss": 1.1318,
+      "step": 1912
+    },
+    {
+      "epoch": 0.38239924039879064,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.639402657800613e-06,
+      "loss": 1.0901,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3825991354538867,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.639009615815158e-06,
+      "loss": 1.07,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3827990305089828,
+      "grad_norm": 2.0,
+      "learning_rate": 9.63861636776539e-06,
+      "loss": 1.103,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3829989255640789,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.63822291366878e-06,
+      "loss": 1.1312,
+      "step": 1916
+    },
+    {
+      "epoch": 0.38319882061917493,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.637829253542801e-06,
+      "loss": 1.0867,
+      "step": 1917
+    },
+    {
+      "epoch": 0.38339871567427103,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.637435387404943e-06,
+      "loss": 1.1388,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3835986107293671,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.637041315272702e-06,
+      "loss": 1.1109,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3837985057844632,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.63664703716358e-06,
+      "loss": 0.9584,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3839984008395592,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.636252553095091e-06,
+      "loss": 1.1159,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3841982958946553,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.635857863084758e-06,
+      "loss": 1.0773,
+      "step": 1922
+    },
+    {
+      "epoch": 0.38439819094975136,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.635462967150116e-06,
+      "loss": 1.0794,
+      "step": 1923
+    },
+    {
+      "epoch": 0.38459808600484746,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.635067865308703e-06,
+      "loss": 1.0855,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3847979810599435,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.634672557578072e-06,
+      "loss": 1.0506,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3849978761150396,
+      "grad_norm": 2.125,
+      "learning_rate": 9.63427704397578e-06,
+      "loss": 0.9756,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3851977711701357,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.633881324519397e-06,
+      "loss": 1.0076,
+      "step": 1927
+    },
+    {
+      "epoch": 0.38539766622523175,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.633485399226499e-06,
+      "loss": 1.0761,
+      "step": 1928
+    },
+    {
+      "epoch": 0.38559756128032785,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.633089268114675e-06,
+      "loss": 1.0176,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3857974563354239,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.632692931201522e-06,
+      "loss": 1.0506,
+      "step": 1930
+    },
+    {
+      "epoch": 0.38599735139052,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.632296388504641e-06,
+      "loss": 1.0752,
+      "step": 1931
+    },
+    {
+      "epoch": 0.38619724644561604,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.631899640041652e-06,
+      "loss": 1.1649,
+      "step": 1932
+    },
+    {
+      "epoch": 0.38639714150071214,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.631502685830175e-06,
+      "loss": 1.0854,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3865970365558082,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.631105525887841e-06,
+      "loss": 1.0688,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3867969316109043,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.630708160232297e-06,
+      "loss": 1.0149,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3869968266660003,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.63031058888119e-06,
+      "loss": 1.0706,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3871967217210964,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.629912811852181e-06,
+      "loss": 0.984,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3873966167761925,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.62951482916294e-06,
+      "loss": 1.07,
+      "step": 1938
+    },
+    {
+      "epoch": 0.38759651183128857,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.629116640831144e-06,
+      "loss": 1.1219,
+      "step": 1939
+    },
+    {
+      "epoch": 0.38779640688638467,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.628718246874482e-06,
+      "loss": 1.1367,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3879963019414807,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.628319647310648e-06,
+      "loss": 1.0883,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3881961969965768,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.627920842157352e-06,
+      "loss": 1.1226,
+      "step": 1942
+    },
+    {
+      "epoch": 0.38839609205167286,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.627521831432308e-06,
+      "loss": 1.1547,
+      "step": 1943
+    },
+    {
+      "epoch": 0.38859598710676896,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.627122615153234e-06,
+      "loss": 1.0724,
+      "step": 1944
+    },
+    {
+      "epoch": 0.388795882161865,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.62672319333787e-06,
+      "loss": 1.1552,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3889957772169611,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.626323566003955e-06,
+      "loss": 1.1128,
+      "step": 1946
+    },
+    {
+      "epoch": 0.38919567227205715,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.625923733169242e-06,
+      "loss": 1.1447,
+      "step": 1947
+    },
+    {
+      "epoch": 0.38939556732715325,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.62552369485149e-06,
+      "loss": 1.0015,
+      "step": 1948
+    },
+    {
+      "epoch": 0.38959546238224935,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.625123451068472e-06,
+      "loss": 1.267,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3897953574373454,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.62472300183796e-06,
+      "loss": 1.0929,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3899952524924415,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.62432234717775e-06,
+      "loss": 0.9937,
+      "step": 1951
+    },
+    {
+      "epoch": 0.39019514754753754,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.623921487105634e-06,
+      "loss": 1.1026,
+      "step": 1952
+    },
+    {
+      "epoch": 0.39039504260263364,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.62352042163942e-06,
+      "loss": 1.0733,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3905949376577297,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.623119150796923e-06,
+      "loss": 1.0681,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3907948327128258,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.622717674595965e-06,
+      "loss": 1.0314,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3909947277679218,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.622315993054384e-06,
+      "loss": 1.1474,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3911946228230179,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.621914106190019e-06,
+      "loss": 1.0736,
+      "step": 1957
+    },
+    {
+      "epoch": 0.39139451787811397,
+      "grad_norm": 2.125,
+      "learning_rate": 9.621512014020722e-06,
+      "loss": 1.0334,
+      "step": 1958
+    },
+    {
+      "epoch": 0.39159441293321007,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.621109716564357e-06,
+      "loss": 1.1237,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3917943079883061,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.620707213838789e-06,
+      "loss": 1.0922,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3919942030434022,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.620304505861902e-06,
+      "loss": 1.0834,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3921940980984983,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.619901592651582e-06,
+      "loss": 1.0662,
+      "step": 1962
+    },
+    {
+      "epoch": 0.39239399315359436,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.619498474225729e-06,
+      "loss": 1.0044,
+      "step": 1963
+    },
+    {
+      "epoch": 0.39259388820869046,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.619095150602243e-06,
+      "loss": 1.0423,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3927937832637865,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.618691621799047e-06,
+      "loss": 1.0499,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3929936783188826,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.618287887834062e-06,
+      "loss": 0.9927,
+      "step": 1966
+    },
+    {
+      "epoch": 0.39319357337397864,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.617883948725219e-06,
+      "loss": 0.9855,
+      "step": 1967
+    },
+    {
+      "epoch": 0.39339346842907474,
+      "grad_norm": 2.25,
+      "learning_rate": 9.617479804490468e-06,
+      "loss": 1.0272,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3935933634841708,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.617075455147757e-06,
+      "loss": 1.1507,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3937932585392669,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.616670900715046e-06,
+      "loss": 1.0953,
+      "step": 1970
+    },
+    {
+      "epoch": 0.39399315359436293,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.616266141210306e-06,
+      "loss": 1.0569,
+      "step": 1971
+    },
+    {
+      "epoch": 0.39419304864945903,
+      "grad_norm": 2.125,
+      "learning_rate": 9.615861176651519e-06,
+      "loss": 1.1121,
+      "step": 1972
+    },
+    {
+      "epoch": 0.39439294370455513,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.61545600705667e-06,
+      "loss": 1.1301,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3945928387596512,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.615050632443759e-06,
+      "loss": 1.0539,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3947927338147473,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.614645052830791e-06,
+      "loss": 1.0629,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3949926288698433,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.614239268235783e-06,
+      "loss": 1.0353,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3951925239249394,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.613833278676762e-06,
+      "loss": 1.1782,
+      "step": 1977
+    },
+    {
+      "epoch": 0.39539241898003546,
+      "grad_norm": 2.25,
+      "learning_rate": 9.613427084171755e-06,
+      "loss": 1.0323,
+      "step": 1978
+    },
+    {
+      "epoch": 0.39559231403513156,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.613020684738813e-06,
+      "loss": 1.146,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3957922090902276,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.612614080395983e-06,
+      "loss": 1.0603,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3959921041453237,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.612207271161328e-06,
+      "loss": 1.0065,
+      "step": 1981
+    },
+    {
+      "epoch": 0.39619199920041975,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.61180025705292e-06,
+      "loss": 1.0248,
+      "step": 1982
+    },
+    {
+      "epoch": 0.39639189425551585,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.611393038088839e-06,
+      "loss": 1.0946,
+      "step": 1983
+    },
+    {
+      "epoch": 0.39659178931061195,
+      "grad_norm": 2.25,
+      "learning_rate": 9.610985614287168e-06,
+      "loss": 1.049,
+      "step": 1984
+    },
+    {
+      "epoch": 0.396791684365708,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.61057798566601e-06,
+      "loss": 1.0451,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3969915794208041,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.61017015224347e-06,
+      "loss": 1.1273,
+      "step": 1986
+    },
+    {
+      "epoch": 0.39719147447590014,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.609762114037665e-06,
+      "loss": 1.1022,
+      "step": 1987
+    },
+    {
+      "epoch": 0.39739136953099624,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.609353871066719e-06,
+      "loss": 0.9584,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3975912645860923,
+      "grad_norm": 2.25,
+      "learning_rate": 9.608945423348766e-06,
+      "loss": 1.1338,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3977911596411884,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.60853677090195e-06,
+      "loss": 1.0808,
+      "step": 1990
+    },
+    {
+      "epoch": 0.39799105469628443,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.608127913744421e-06,
+      "loss": 1.0717,
+      "step": 1991
+    },
+    {
+      "epoch": 0.39819094975138053,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.607718851894346e-06,
+      "loss": 1.066,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3983908448064766,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.607309585369889e-06,
+      "loss": 1.1569,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3985907398615727,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.606900114189231e-06,
+      "loss": 1.0631,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3987906349166688,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.606490438370563e-06,
+      "loss": 1.0811,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3989905299717648,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.606080557932084e-06,
+      "loss": 1.15,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3991904250268609,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.605670472891998e-06,
+      "loss": 1.0575,
+      "step": 1997
+    },
+    {
+      "epoch": 0.39939032008195696,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.605260183268519e-06,
+      "loss": 0.9726,
+      "step": 1998
+    },
+    {
+      "epoch": 0.39959021513705306,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.604849689079875e-06,
+      "loss": 1.0419,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3997901101921491,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.604438990344303e-06,
+      "loss": 1.0341,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3999900052472452,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.60402808708004e-06,
+      "loss": 1.0606,
+      "step": 2001
+    },
+    {
+      "epoch": 0.40018990030234125,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.60361697930534e-06,
+      "loss": 1.011,
+      "step": 2002
+    },
+    {
+      "epoch": 0.40038979535743735,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.603205667038468e-06,
+      "loss": 1.0795,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4005896904125334,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.60279415029769e-06,
+      "loss": 0.9848,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4007895854676295,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.60238242910129e-06,
+      "loss": 0.9808,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4009894805227256,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.601970503467551e-06,
+      "loss": 1.0222,
+      "step": 2006
+    },
+    {
+      "epoch": 0.40118937557782164,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.601558373414776e-06,
+      "loss": 1.0831,
+      "step": 2007
+    },
+    {
+      "epoch": 0.40138927063291774,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.601146038961267e-06,
+      "loss": 0.9672,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4015891656880138,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.600733500125345e-06,
+      "loss": 1.0311,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4017890607431099,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.600320756925332e-06,
+      "loss": 1.1057,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4019889557982059,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.59990780937956e-06,
+      "loss": 1.1822,
+      "step": 2011
+    },
+    {
+      "epoch": 0.402188850853302,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.599494657506376e-06,
+      "loss": 1.1193,
+      "step": 2012
+    },
+    {
+      "epoch": 0.40238874590839807,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.599081301324132e-06,
+      "loss": 1.1015,
+      "step": 2013
+    },
+    {
+      "epoch": 0.40258864096349417,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.598667740851187e-06,
+      "loss": 1.0014,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4027885360185902,
+      "grad_norm": 2.125,
+      "learning_rate": 9.59825397610591e-06,
+      "loss": 1.0768,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4029884310736863,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.597840007106685e-06,
+      "loss": 1.0659,
+      "step": 2016
+    },
+    {
+      "epoch": 0.4031883261287824,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.597425833871896e-06,
+      "loss": 1.0412,
+      "step": 2017
+    },
+    {
+      "epoch": 0.40338822118387846,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.597011456419943e-06,
+      "loss": 1.0398,
+      "step": 2018
+    },
+    {
+      "epoch": 0.40358811623897456,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.596596874769232e-06,
+      "loss": 1.1117,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4037880112940706,
+      "grad_norm": 2.0,
+      "learning_rate": 9.59618208893818e-06,
+      "loss": 1.0276,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4039879063491667,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.59576709894521e-06,
+      "loss": 1.1292,
+      "step": 2021
+    },
+    {
+      "epoch": 0.40418780140426275,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.595351904808757e-06,
+      "loss": 1.0813,
+      "step": 2022
+    },
+    {
+      "epoch": 0.40438769645935885,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.594936506547262e-06,
+      "loss": 1.1305,
+      "step": 2023
+    },
+    {
+      "epoch": 0.4045875915144549,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.594520904179176e-06,
+      "loss": 1.1371,
+      "step": 2024
+    },
+    {
+      "epoch": 0.404787486569551,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.594105097722966e-06,
+      "loss": 0.9341,
+      "step": 2025
+    },
+    {
+      "epoch": 0.40498738162464704,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.593689087197096e-06,
+      "loss": 1.1288,
+      "step": 2026
+    },
+    {
+      "epoch": 0.40518727667974314,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.59327287262005e-06,
+      "loss": 1.0997,
+      "step": 2027
+    },
+    {
+      "epoch": 0.40538717173483924,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.592856454010309e-06,
+      "loss": 1.0629,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4055870667899353,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.592439831386378e-06,
+      "loss": 1.0659,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4057869618450314,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.59202300476676e-06,
+      "loss": 1.1291,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4059868569001274,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.59160597416997e-06,
+      "loss": 1.0006,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4061867519552235,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.591188739614534e-06,
+      "loss": 1.1197,
+      "step": 2032
+    },
+    {
+      "epoch": 0.40638664701031957,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.590771301118983e-06,
+      "loss": 1.1276,
+      "step": 2033
+    },
+    {
+      "epoch": 0.40658654206541567,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.590353658701863e-06,
+      "loss": 1.0585,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4067864371205117,
+      "grad_norm": 2.125,
+      "learning_rate": 9.589935812381722e-06,
+      "loss": 1.0337,
+      "step": 2035
+    },
+    {
+      "epoch": 0.4069863321756078,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.589517762177122e-06,
+      "loss": 1.0947,
+      "step": 2036
+    },
+    {
+      "epoch": 0.40718622723070386,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.589099508106637e-06,
+      "loss": 1.167,
+      "step": 2037
+    },
+    {
+      "epoch": 0.40738612228579996,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.588681050188837e-06,
+      "loss": 1.1224,
+      "step": 2038
+    },
+    {
+      "epoch": 0.40758601734089606,
+      "grad_norm": 2.125,
+      "learning_rate": 9.588262388442317e-06,
+      "loss": 1.0462,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4077859123959921,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.587843522885674e-06,
+      "loss": 1.0847,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4079858074510882,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.587424453537508e-06,
+      "loss": 1.0729,
+      "step": 2041
+    },
+    {
+      "epoch": 0.40818570250618424,
+      "grad_norm": 2.125,
+      "learning_rate": 9.587005180416439e-06,
+      "loss": 1.0238,
+      "step": 2042
+    },
+    {
+      "epoch": 0.40838559756128034,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.586585703541092e-06,
+      "loss": 1.0113,
+      "step": 2043
+    },
+    {
+      "epoch": 0.4085854926163764,
+      "grad_norm": 2.125,
+      "learning_rate": 9.586166022930095e-06,
+      "loss": 1.0278,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4087853876714725,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.585746138602095e-06,
+      "loss": 1.0631,
+      "step": 2045
+    },
+    {
+      "epoch": 0.40898528272656853,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.58532605057574e-06,
+      "loss": 1.1036,
+      "step": 2046
+    },
+    {
+      "epoch": 0.40918517778166463,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.584905758869691e-06,
+      "loss": 1.1368,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4093850728367607,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.584485263502619e-06,
+      "loss": 1.1605,
+      "step": 2048
+    },
+    {
+      "epoch": 0.4095849678918568,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.5840645644932e-06,
+      "loss": 1.0755,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4097848629469529,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.583643661860125e-06,
+      "loss": 1.0774,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4099847580020489,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.583222555622087e-06,
+      "loss": 0.9607,
+      "step": 2051
+    },
+    {
+      "epoch": 0.410184653057145,
+      "grad_norm": 2.125,
+      "learning_rate": 9.582801245797793e-06,
+      "loss": 1.1346,
+      "step": 2052
+    },
+    {
+      "epoch": 0.41038454811224107,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.582379732405955e-06,
+      "loss": 1.0721,
+      "step": 2053
+    },
+    {
+      "epoch": 0.41058444316733717,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.5819580154653e-06,
+      "loss": 1.1277,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4107843382224332,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.581536094994562e-06,
+      "loss": 1.0082,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4109842332775293,
+      "grad_norm": 2.125,
+      "learning_rate": 9.581113971012476e-06,
+      "loss": 1.0861,
+      "step": 2056
+    },
+    {
+      "epoch": 0.41118412833262535,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.580691643537798e-06,
+      "loss": 1.1013,
+      "step": 2057
+    },
+    {
+      "epoch": 0.41138402338772145,
+      "grad_norm": 2.125,
+      "learning_rate": 9.580269112589287e-06,
+      "loss": 1.0299,
+      "step": 2058
+    },
+    {
+      "epoch": 0.4115839184428175,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.579846378185712e-06,
+      "loss": 1.0958,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4117838134979136,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.579423440345847e-06,
+      "loss": 1.1165,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4119837085530097,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.579000299088485e-06,
+      "loss": 1.0798,
+      "step": 2061
+    },
+    {
+      "epoch": 0.41218360360810574,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.578576954432416e-06,
+      "loss": 1.0372,
+      "step": 2062
+    },
+    {
+      "epoch": 0.41238349866320184,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.57815340639645e-06,
+      "loss": 1.0837,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4125833937182979,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.577729654999396e-06,
+      "loss": 1.0542,
+      "step": 2064
+    },
+    {
+      "epoch": 0.412783288773394,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.577305700260084e-06,
+      "loss": 1.1162,
+      "step": 2065
+    },
+    {
+      "epoch": 0.41298318382849003,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.576881542197337e-06,
+      "loss": 1.0897,
+      "step": 2066
+    },
+    {
+      "epoch": 0.41318307888358613,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.576457180830004e-06,
+      "loss": 1.16,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4133829739386822,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.576032616176932e-06,
+      "loss": 1.0592,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4135828689937783,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.575607848256979e-06,
+      "loss": 1.0421,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4137827640488743,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.575182877089014e-06,
+      "loss": 1.0748,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4139826591039704,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.574757702691912e-06,
+      "loss": 1.1479,
+      "step": 2071
+    },
+    {
+      "epoch": 0.41418255415906646,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.574332325084564e-06,
+      "loss": 0.9352,
+      "step": 2072
+    },
+    {
+      "epoch": 0.41438244921416256,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.573906744285862e-06,
+      "loss": 1.0908,
+      "step": 2073
+    },
+    {
+      "epoch": 0.41458234426925866,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.573480960314711e-06,
+      "loss": 1.0818,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4147822393243547,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.573054973190023e-06,
+      "loss": 1.0848,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4149821343794508,
+      "grad_norm": 2.125,
+      "learning_rate": 9.572628782930724e-06,
+      "loss": 1.0214,
+      "step": 2076
+    },
+    {
+      "epoch": 0.41518202943454685,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.572202389555741e-06,
+      "loss": 1.1551,
+      "step": 2077
+    },
+    {
+      "epoch": 0.41538192448964295,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.571775793084017e-06,
+      "loss": 1.0506,
+      "step": 2078
+    },
+    {
+      "epoch": 0.415581819544739,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.5713489935345e-06,
+      "loss": 1.0266,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4157817145998351,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.57092199092615e-06,
+      "loss": 1.0646,
+      "step": 2080
+    },
+    {
+      "epoch": 0.41598160965493114,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.570494785277931e-06,
+      "loss": 1.0513,
+      "step": 2081
+    },
+    {
+      "epoch": 0.41618150471002724,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.570067376608826e-06,
+      "loss": 1.0574,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4163813997651233,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.569639764937813e-06,
+      "loss": 1.0695,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4165812948202194,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.56921195028389e-06,
+      "loss": 1.1231,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4167811898753155,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.56878393266606e-06,
+      "loss": 1.135,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4169810849304115,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.568355712103336e-06,
+      "loss": 1.1205,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4171809799855076,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.56792728861474e-06,
+      "loss": 1.1319,
+      "step": 2087
+    },
+    {
+      "epoch": 0.41738087504060367,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.5674986622193e-06,
+      "loss": 1.0088,
+      "step": 2088
+    },
+    {
+      "epoch": 0.41758077009569977,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.567069832936058e-06,
+      "loss": 1.076,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4177806651507958,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.566640800784061e-06,
+      "loss": 1.1271,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4179805602058919,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.56621156578237e-06,
+      "loss": 1.0794,
+      "step": 2091
+    },
+    {
+      "epoch": 0.41818045526098796,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.565782127950047e-06,
+      "loss": 1.0835,
+      "step": 2092
+    },
+    {
+      "epoch": 0.41838035031608406,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.56535248730617e-06,
+      "loss": 1.1816,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4185802453711801,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.564922643869822e-06,
+      "loss": 1.0854,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4187801404262762,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.5644925976601e-06,
+      "loss": 1.0796,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4189800354813723,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.564062348696103e-06,
+      "loss": 1.1448,
+      "step": 2096
+    },
+    {
+      "epoch": 0.41917993053646835,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.563631896996943e-06,
+      "loss": 1.0599,
+      "step": 2097
+    },
+    {
+      "epoch": 0.41937982559156445,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.563201242581743e-06,
+      "loss": 1.0835,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4195797206466605,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.562770385469631e-06,
+      "loss": 1.1198,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4197796157017566,
+      "grad_norm": 2.125,
+      "learning_rate": 9.562339325679747e-06,
+      "loss": 1.0479,
+      "step": 2100
+    },
+    {
+      "epoch": 0.41997951075685264,
+      "grad_norm": 2.125,
+      "learning_rate": 9.561908063231234e-06,
+      "loss": 1.1437,
+      "step": 2101
+    },
+    {
+      "epoch": 0.42017940581194874,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.561476598143255e-06,
+      "loss": 1.0515,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4203793008670448,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.56104493043497e-06,
+      "loss": 1.0693,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4205791959221409,
+      "grad_norm": 2.125,
+      "learning_rate": 9.560613060125558e-06,
+      "loss": 1.1833,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4207790909772369,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.560180987234202e-06,
+      "loss": 1.058,
+      "step": 2105
+    },
+    {
+      "epoch": 0.420978986032333,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.559748711780092e-06,
+      "loss": 1.0376,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4211788810874291,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.559316233782432e-06,
+      "loss": 1.096,
+      "step": 2107
+    },
+    {
+      "epoch": 0.42137877614252517,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.558883553260431e-06,
+      "loss": 1.059,
+      "step": 2108
+    },
+    {
+      "epoch": 0.42157867119762127,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.55845067023331e-06,
+      "loss": 1.0785,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4217785662527173,
+      "grad_norm": 2.0,
+      "learning_rate": 9.558017584720298e-06,
+      "loss": 1.014,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4219784613078134,
+      "grad_norm": 2.25,
+      "learning_rate": 9.55758429674063e-06,
+      "loss": 1.0571,
+      "step": 2111
+    },
+    {
+      "epoch": 0.42217835636290946,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.557150806313555e-06,
+      "loss": 1.0677,
+      "step": 2112
+    },
+    {
+      "epoch": 0.42237825141800556,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.55671711345833e-06,
+      "loss": 1.0337,
+      "step": 2113
+    },
+    {
+      "epoch": 0.4225781464731016,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.556283218194214e-06,
+      "loss": 1.0861,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4227780415281977,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.555849120540486e-06,
+      "loss": 1.0313,
+      "step": 2115
+    },
+    {
+      "epoch": 0.42297793658329375,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.555414820516427e-06,
+      "loss": 0.9798,
+      "step": 2116
+    },
+    {
+      "epoch": 0.42317783163838985,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.554980318141327e-06,
+      "loss": 1.1204,
+      "step": 2117
+    },
+    {
+      "epoch": 0.42337772669348595,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.554545613434489e-06,
+      "loss": 1.0616,
+      "step": 2118
+    },
+    {
+      "epoch": 0.423577621748582,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.55411070641522e-06,
+      "loss": 1.0907,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4237775168036781,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.55367559710284e-06,
+      "loss": 1.1348,
+      "step": 2120
+    },
+    {
+      "epoch": 0.42397741185877413,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.553240285516676e-06,
+      "loss": 0.9642,
+      "step": 2121
+    },
+    {
+      "epoch": 0.42417730691387023,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.552804771676067e-06,
+      "loss": 1.1451,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4243772019689663,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.552369055600354e-06,
+      "loss": 1.0617,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4245770970240624,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.551933137308895e-06,
+      "loss": 1.111,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4247769920791584,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.551497016821051e-06,
+      "loss": 1.0553,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4249768871342545,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.551060694156197e-06,
+      "loss": 1.1344,
+      "step": 2126
+    },
+    {
+      "epoch": 0.42517678218935057,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.550624169333713e-06,
+      "loss": 0.9748,
+      "step": 2127
+    },
+    {
+      "epoch": 0.42537667724444667,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.550187442372987e-06,
+      "loss": 1.0298,
+      "step": 2128
+    },
+    {
+      "epoch": 0.42557657229954277,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.549750513293422e-06,
+      "loss": 1.0584,
+      "step": 2129
+    },
+    {
+      "epoch": 0.4257764673546388,
+      "grad_norm": 2.125,
+      "learning_rate": 9.549313382114427e-06,
+      "loss": 1.0891,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4259763624097349,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.548876048855417e-06,
+      "loss": 1.1657,
+      "step": 2131
+    },
+    {
+      "epoch": 0.42617625746483095,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.548438513535819e-06,
+      "loss": 1.0,
+      "step": 2132
+    },
+    {
+      "epoch": 0.42637615251992705,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.548000776175066e-06,
+      "loss": 1.2088,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4265760475750231,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.547562836792606e-06,
+      "loss": 1.1251,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4267759426301192,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.547124695407888e-06,
+      "loss": 1.1375,
+      "step": 2135
+    },
+    {
+      "epoch": 0.42697583768521524,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.546686352040379e-06,
+      "loss": 1.0466,
+      "step": 2136
+    },
+    {
+      "epoch": 0.42717573274031134,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.546247806709548e-06,
+      "loss": 1.0535,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4273756277954074,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.545809059434874e-06,
+      "loss": 1.0398,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4275755228505035,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.545370110235847e-06,
+      "loss": 1.0722,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4277754179055996,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.544930959131967e-06,
+      "loss": 1.0896,
+      "step": 2140
+    },
+    {
+      "epoch": 0.42797531296069563,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.544491606142737e-06,
+      "loss": 1.0248,
+      "step": 2141
+    },
+    {
+      "epoch": 0.42817520801579173,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.544052051287675e-06,
+      "loss": 1.0491,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4283751030708878,
+      "grad_norm": 2.125,
+      "learning_rate": 9.543612294586306e-06,
+      "loss": 1.0873,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4285749981259839,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.543172336058166e-06,
+      "loss": 1.0051,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4287748931810799,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.542732175722796e-06,
+      "loss": 1.1224,
+      "step": 2145
+    },
+    {
+      "epoch": 0.428974788236176,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.542291813599746e-06,
+      "loss": 1.1181,
+      "step": 2146
+    },
+    {
+      "epoch": 0.42917468329127206,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.541851249708581e-06,
+      "loss": 1.0761,
+      "step": 2147
+    },
+    {
+      "epoch": 0.42937457834636816,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.541410484068867e-06,
+      "loss": 1.0671,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4295744734014642,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.540969516700185e-06,
+      "loss": 1.1324,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4297743684565603,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.540528347622123e-06,
+      "loss": 1.0483,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4299742635116564,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.540086976854274e-06,
+      "loss": 1.0794,
+      "step": 2151
+    },
+    {
+      "epoch": 0.43017415856675245,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.539645404416249e-06,
+      "loss": 1.108,
+      "step": 2152
+    },
+    {
+      "epoch": 0.43037405362184855,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.53920363032766e-06,
+      "loss": 0.9816,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4305739486769446,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.538761654608128e-06,
+      "loss": 1.0624,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4307738437320407,
+      "grad_norm": 2.0,
+      "learning_rate": 9.53831947727729e-06,
+      "loss": 1.0004,
+      "step": 2155
+    },
+    {
+      "epoch": 0.43097373878713674,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.537877098354787e-06,
+      "loss": 1.0317,
+      "step": 2156
+    },
+    {
+      "epoch": 0.43117363384223284,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.537434517860265e-06,
+      "loss": 1.058,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4313735288973289,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.536991735813388e-06,
+      "loss": 1.0414,
+      "step": 2158
+    },
+    {
+      "epoch": 0.431573423952425,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.536548752233822e-06,
+      "loss": 1.1674,
+      "step": 2159
+    },
+    {
+      "epoch": 0.43177331900752103,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.536105567141246e-06,
+      "loss": 1.0639,
+      "step": 2160
+    },
+    {
+      "epoch": 0.43197321406261713,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.535662180555342e-06,
+      "loss": 1.0177,
+      "step": 2161
+    },
+    {
+      "epoch": 0.43217310911771323,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.535218592495812e-06,
+      "loss": 1.145,
+      "step": 2162
+    },
+    {
+      "epoch": 0.43237300417280927,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.534774802982356e-06,
+      "loss": 1.0808,
+      "step": 2163
+    },
+    {
+      "epoch": 0.43257289922790537,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.534330812034686e-06,
+      "loss": 1.0745,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4327727942830014,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.533886619672527e-06,
+      "loss": 0.9411,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4329726893380975,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.533442225915607e-06,
+      "loss": 1.0548,
+      "step": 2166
+    },
+    {
+      "epoch": 0.43317258439319356,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.532997630783669e-06,
+      "loss": 1.0255,
+      "step": 2167
+    },
+    {
+      "epoch": 0.43337247944828966,
+      "grad_norm": 2.25,
+      "learning_rate": 9.53255283429646e-06,
+      "loss": 1.0143,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4335723745033857,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.532107836473739e-06,
+      "loss": 1.0807,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4337722695584818,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.53166263733527e-06,
+      "loss": 1.1443,
+      "step": 2170
+    },
+    {
+      "epoch": 0.43397216461357785,
+      "grad_norm": 2.25,
+      "learning_rate": 9.531217236900832e-06,
+      "loss": 1.1225,
+      "step": 2171
+    },
+    {
+      "epoch": 0.43417205966867395,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.530771635190208e-06,
+      "loss": 1.0479,
+      "step": 2172
+    },
+    {
+      "epoch": 0.43437195472377005,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.53032583222319e-06,
+      "loss": 1.0765,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4345718497788661,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.529879828019586e-06,
+      "loss": 1.1234,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4347717448339622,
+      "grad_norm": 2.125,
+      "learning_rate": 9.5294336225992e-06,
+      "loss": 1.088,
+      "step": 2175
+    },
+    {
+      "epoch": 0.43497163988905824,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.528987215981859e-06,
+      "loss": 1.065,
+      "step": 2176
+    },
+    {
+      "epoch": 0.43517153494415434,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.52854060818739e-06,
+      "loss": 1.0823,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4353714299992504,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.528093799235629e-06,
+      "loss": 1.0597,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4355713250543465,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.527646789146427e-06,
+      "loss": 1.0214,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4357712201094425,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.527199577939636e-06,
+      "loss": 1.0689,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4359711151645386,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.526752165635125e-06,
+      "loss": 1.0881,
+      "step": 2181
+    },
+    {
+      "epoch": 0.43617101021963467,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.526304552252766e-06,
+      "loss": 0.9681,
+      "step": 2182
+    },
+    {
+      "epoch": 0.43637090527473077,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.52585673781244e-06,
+      "loss": 1.0983,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4365708003298268,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.525408722334045e-06,
+      "loss": 1.1078,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4367706953849229,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.524960505837475e-06,
+      "loss": 0.992,
+      "step": 2185
+    },
+    {
+      "epoch": 0.436970590440019,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.524512088342642e-06,
+      "loss": 1.0459,
+      "step": 2186
+    },
+    {
+      "epoch": 0.43717048549511506,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.524063469869467e-06,
+      "loss": 1.0761,
+      "step": 2187
+    },
+    {
+      "epoch": 0.43737038055021116,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.523614650437876e-06,
+      "loss": 1.1383,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4375702756053072,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.523165630067804e-06,
+      "loss": 1.0518,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4377701706604033,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.522716408779198e-06,
+      "loss": 1.0729,
+      "step": 2190
+    },
+    {
+      "epoch": 0.43797006571549935,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.522266986592012e-06,
+      "loss": 1.0924,
+      "step": 2191
+    },
+    {
+      "epoch": 0.43816996077059545,
+      "grad_norm": 2.25,
+      "learning_rate": 9.521817363526211e-06,
+      "loss": 1.0138,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4383698558256915,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.521367539601766e-06,
+      "loss": 1.0787,
+      "step": 2193
+    },
+    {
+      "epoch": 0.4385697508807876,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.520917514838657e-06,
+      "loss": 0.9883,
+      "step": 2194
+    },
+    {
+      "epoch": 0.43876964593588363,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.520467289256874e-06,
+      "loss": 0.9786,
+      "step": 2195
+    },
+    {
+      "epoch": 0.43896954099097973,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.520016862876416e-06,
+      "loss": 1.03,
+      "step": 2196
+    },
+    {
+      "epoch": 0.43916943604607583,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.519566235717295e-06,
+      "loss": 1.0743,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4393693311011719,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.519115407799523e-06,
+      "loss": 1.0433,
+      "step": 2198
+    },
+    {
+      "epoch": 0.439569226156268,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.518664379143125e-06,
+      "loss": 1.0033,
+      "step": 2199
+    },
+    {
+      "epoch": 0.439769121211364,
+      "grad_norm": 2.0,
+      "learning_rate": 9.518213149768141e-06,
+      "loss": 1.0444,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4399690162664601,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.517761719694609e-06,
+      "loss": 1.0285,
+      "step": 2201
+    },
+    {
+      "epoch": 0.44016891132155617,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.517310088942585e-06,
+      "loss": 1.1743,
+      "step": 2202
+    },
+    {
+      "epoch": 0.44036880637665227,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.516858257532128e-06,
+      "loss": 0.973,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4405687014317483,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.516406225483312e-06,
+      "loss": 1.074,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4407685964868444,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.515953992816215e-06,
+      "loss": 1.0519,
+      "step": 2205
+    },
+    {
+      "epoch": 0.44096849154194045,
+      "grad_norm": 2.125,
+      "learning_rate": 9.515501559550919e-06,
+      "loss": 1.0592,
+      "step": 2206
+    },
+    {
+      "epoch": 0.44116838659703655,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.51504892570753e-06,
+      "loss": 1.0736,
+      "step": 2207
+    },
+    {
+      "epoch": 0.44136828165213265,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.514596091306148e-06,
+      "loss": 1.1223,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4415681767072287,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.514143056366892e-06,
+      "loss": 1.0031,
+      "step": 2209
+    },
+    {
+      "epoch": 0.4417680717623248,
+      "grad_norm": 2.0,
+      "learning_rate": 9.513689820909882e-06,
+      "loss": 1.0427,
+      "step": 2210
+    },
+    {
+      "epoch": 0.44196796681742084,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.51323638495525e-06,
+      "loss": 1.0908,
+      "step": 2211
+    },
+    {
+      "epoch": 0.44216786187251694,
+      "grad_norm": 2.0,
+      "learning_rate": 9.512782748523143e-06,
+      "loss": 0.9979,
+      "step": 2212
+    },
+    {
+      "epoch": 0.442367756927613,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.512328911633708e-06,
+      "loss": 1.104,
+      "step": 2213
+    },
+    {
+      "epoch": 0.4425676519827091,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.511874874307103e-06,
+      "loss": 1.0095,
+      "step": 2214
+    },
+    {
+      "epoch": 0.44276754703780513,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.5114206365635e-06,
+      "loss": 1.1195,
+      "step": 2215
+    },
+    {
+      "epoch": 0.44296744209290123,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.510966198423073e-06,
+      "loss": 1.0611,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4431673371479973,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.51051155990601e-06,
+      "loss": 1.0874,
+      "step": 2217
+    },
+    {
+      "epoch": 0.4433672322030934,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.510056721032504e-06,
+      "loss": 1.1258,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4435671272581895,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.509601681822761e-06,
+      "loss": 1.0456,
+      "step": 2219
+    },
+    {
+      "epoch": 0.4437670223132855,
+      "grad_norm": 2.125,
+      "learning_rate": 9.509146442296992e-06,
+      "loss": 1.048,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4439669173683816,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.508691002475421e-06,
+      "loss": 1.1,
+      "step": 2221
+    },
+    {
+      "epoch": 0.44416681242347766,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.508235362378278e-06,
+      "loss": 1.1121,
+      "step": 2222
+    },
+    {
+      "epoch": 0.44436670747857376,
+      "grad_norm": 2.125,
+      "learning_rate": 9.507779522025799e-06,
+      "loss": 0.9907,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4445666025336698,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.507323481438236e-06,
+      "loss": 1.071,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4447664975887659,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.506867240635847e-06,
+      "loss": 1.0096,
+      "step": 2225
+    },
+    {
+      "epoch": 0.44496639264386195,
+      "grad_norm": 2.125,
+      "learning_rate": 9.506410799638894e-06,
+      "loss": 1.1819,
+      "step": 2226
+    },
+    {
+      "epoch": 0.44516628769895805,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.505954158467656e-06,
+      "loss": 1.1682,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4453661827540541,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.505497317142416e-06,
+      "loss": 1.0369,
+      "step": 2228
+    },
+    {
+      "epoch": 0.4455660778091502,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.505040275683465e-06,
+      "loss": 1.1164,
+      "step": 2229
+    },
+    {
+      "epoch": 0.4457659728642463,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.504583034111108e-06,
+      "loss": 1.1338,
+      "step": 2230
+    },
+    {
+      "epoch": 0.44596586791934234,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.504125592445653e-06,
+      "loss": 1.074,
+      "step": 2231
+    },
+    {
+      "epoch": 0.44616576297443844,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.50366795070742e-06,
+      "loss": 1.0827,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4463656580295345,
+      "grad_norm": 2.125,
+      "learning_rate": 9.503210108916736e-06,
+      "loss": 0.9933,
+      "step": 2233
+    },
+    {
+      "epoch": 0.4465655530846306,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.502752067093942e-06,
+      "loss": 1.0505,
+      "step": 2234
+    },
+    {
+      "epoch": 0.44676544813972663,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.50229382525938e-06,
+      "loss": 1.0627,
+      "step": 2235
+    },
+    {
+      "epoch": 0.44696534319482273,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.50183538343341e-06,
+      "loss": 1.0896,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4471652382499188,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.501376741636392e-06,
+      "loss": 1.0012,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4473651333050149,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.5009178998887e-06,
+      "loss": 1.1222,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4475650283601109,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.500458858210714e-06,
+      "loss": 1.1579,
+      "step": 2239
+    },
+    {
+      "epoch": 0.447764923415207,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.499999616622828e-06,
+      "loss": 1.105,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4479648184703031,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.499540175145438e-06,
+      "loss": 1.0628,
+      "step": 2241
+    },
+    {
+      "epoch": 0.44816471352539916,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.499080533798956e-06,
+      "loss": 0.9828,
+      "step": 2242
+    },
+    {
+      "epoch": 0.44836460858049526,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.498620692603797e-06,
+      "loss": 1.0454,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4485645036355913,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.498160651580387e-06,
+      "loss": 1.063,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4487643986906874,
+      "grad_norm": 2.0,
+      "learning_rate": 9.49770041074916e-06,
+      "loss": 1.0674,
+      "step": 2245
+    },
+    {
+      "epoch": 0.44896429374578345,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.497239970130561e-06,
+      "loss": 0.9574,
+      "step": 2246
+    },
+    {
+      "epoch": 0.44916418880087955,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.496779329745045e-06,
+      "loss": 1.0144,
+      "step": 2247
+    },
+    {
+      "epoch": 0.4493640838559756,
+      "grad_norm": 2.25,
+      "learning_rate": 9.49631848961307e-06,
+      "loss": 1.1661,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4495639789110717,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.495857449755109e-06,
+      "loss": 0.9927,
+      "step": 2249
+    },
+    {
+      "epoch": 0.44976387396616774,
+      "grad_norm": 2.125,
+      "learning_rate": 9.495396210191639e-06,
+      "loss": 1.0636,
+      "step": 2250
+    },
+    {
+      "epoch": 0.44996376902126384,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.49493477094315e-06,
+      "loss": 1.0666,
+      "step": 2251
+    },
+    {
+      "epoch": 0.45016366407635994,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.494473132030137e-06,
+      "loss": 1.0508,
+      "step": 2252
+    },
+    {
+      "epoch": 0.450363559131456,
+      "grad_norm": 2.125,
+      "learning_rate": 9.494011293473109e-06,
+      "loss": 1.0578,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4505634541865521,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.49354925529258e-06,
+      "loss": 1.0574,
+      "step": 2254
+    },
+    {
+      "epoch": 0.4507633492416481,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.493087017509072e-06,
+      "loss": 1.1154,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4509632442967442,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.49262458014312e-06,
+      "loss": 1.1417,
+      "step": 2256
+    },
+    {
+      "epoch": 0.45116313935184027,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.492161943215262e-06,
+      "loss": 1.1611,
+      "step": 2257
+    },
+    {
+      "epoch": 0.45136303440693637,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.491699106746051e-06,
+      "loss": 1.0836,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4515629294620324,
+      "grad_norm": 2.125,
+      "learning_rate": 9.491236070756045e-06,
+      "loss": 1.1178,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4517628245171285,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.490772835265814e-06,
+      "loss": 1.0072,
+      "step": 2260
+    },
+    {
+      "epoch": 0.45196271957222456,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.490309400295932e-06,
+      "loss": 1.0635,
+      "step": 2261
+    },
+    {
+      "epoch": 0.45216261462732066,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.489845765866986e-06,
+      "loss": 0.9716,
+      "step": 2262
+    },
+    {
+      "epoch": 0.45236250968241676,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.48938193199957e-06,
+      "loss": 1.0294,
+      "step": 2263
+    },
+    {
+      "epoch": 0.4525624047375128,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.48891789871429e-06,
+      "loss": 1.0198,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4527622997926089,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.488453666031755e-06,
+      "loss": 1.0776,
+      "step": 2265
+    },
+    {
+      "epoch": 0.45296219484770495,
+      "grad_norm": 2.0,
+      "learning_rate": 9.487989233972587e-06,
+      "loss": 1.0584,
+      "step": 2266
+    },
+    {
+      "epoch": 0.45316208990280105,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.487524602557417e-06,
+      "loss": 1.0575,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4533619849578971,
+      "grad_norm": 2.125,
+      "learning_rate": 9.487059771806883e-06,
+      "loss": 1.0741,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4535618800129932,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.486594741741634e-06,
+      "loss": 1.0547,
+      "step": 2269
+    },
+    {
+      "epoch": 0.45376177506808923,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.486129512382327e-06,
+      "loss": 1.0859,
+      "step": 2270
+    },
+    {
+      "epoch": 0.45396167012318533,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.485664083749623e-06,
+      "loss": 1.1005,
+      "step": 2271
+    },
+    {
+      "epoch": 0.4541615651782814,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.485198455864203e-06,
+      "loss": 1.2026,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4543614602333775,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.484732628746744e-06,
+      "loss": 1.0519,
+      "step": 2273
+    },
+    {
+      "epoch": 0.4545613552884736,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.484266602417942e-06,
+      "loss": 1.0333,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4547612503435696,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.483800376898496e-06,
+      "loss": 1.1083,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4549611453986657,
+      "grad_norm": 2.125,
+      "learning_rate": 9.483333952209118e-06,
+      "loss": 1.071,
+      "step": 2276
+    },
+    {
+      "epoch": 0.45516104045376177,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.482867328370521e-06,
+      "loss": 0.9757,
+      "step": 2277
+    },
+    {
+      "epoch": 0.45536093550885787,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.482400505403439e-06,
+      "loss": 1.1338,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4555608305639539,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.481933483328604e-06,
+      "loss": 1.1481,
+      "step": 2279
+    },
+    {
+      "epoch": 0.45576072561905,
+      "grad_norm": 2.125,
+      "learning_rate": 9.481466262166763e-06,
+      "loss": 0.9722,
+      "step": 2280
+    },
+    {
+      "epoch": 0.45596062067414606,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.480998841938668e-06,
+      "loss": 1.0343,
+      "step": 2281
+    },
+    {
+      "epoch": 0.45616051572924216,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.480531222665084e-06,
+      "loss": 1.089,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4563604107843382,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.480063404366781e-06,
+      "loss": 1.1038,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4565603058394343,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.479595387064542e-06,
+      "loss": 1.0458,
+      "step": 2284
+    },
+    {
+      "epoch": 0.4567602008945304,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.479127170779151e-06,
+      "loss": 1.1301,
+      "step": 2285
+    },
+    {
+      "epoch": 0.45696009594962644,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.478658755531413e-06,
+      "loss": 1.1147,
+      "step": 2286
+    },
+    {
+      "epoch": 0.45715999100472254,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.47819014134213e-06,
+      "loss": 1.0767,
+      "step": 2287
+    },
+    {
+      "epoch": 0.4573598860598186,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.47772132823212e-06,
+      "loss": 1.1587,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4575597811149147,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.477252316222204e-06,
+      "loss": 1.1138,
+      "step": 2289
+    },
+    {
+      "epoch": 0.45775967617001073,
+      "grad_norm": 2.125,
+      "learning_rate": 9.47678310533322e-06,
+      "loss": 0.9616,
+      "step": 2290
+    },
+    {
+      "epoch": 0.45795957122510683,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.47631369558601e-06,
+      "loss": 1.029,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4581594662802029,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.475844087001423e-06,
+      "loss": 1.0689,
+      "step": 2292
+    },
+    {
+      "epoch": 0.458359361335299,
+      "grad_norm": 2.25,
+      "learning_rate": 9.475374279600317e-06,
+      "loss": 1.1476,
+      "step": 2293
+    },
+    {
+      "epoch": 0.458559256390395,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.474904273403567e-06,
+      "loss": 1.0776,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4587591514454911,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.474434068432046e-06,
+      "loss": 1.0983,
+      "step": 2295
+    },
+    {
+      "epoch": 0.45895904650058716,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.47396366470664e-06,
+      "loss": 1.0944,
+      "step": 2296
+    },
+    {
+      "epoch": 0.45915894155568326,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.473493062248247e-06,
+      "loss": 1.0984,
+      "step": 2297
+    },
+    {
+      "epoch": 0.45935883661077936,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.473022261077771e-06,
+      "loss": 0.881,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4595587316658754,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.472551261216124e-06,
+      "loss": 0.9679,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4597586267209715,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.472080062684225e-06,
+      "loss": 1.0331,
+      "step": 2300
+    },
+    {
+      "epoch": 0.45995852177606755,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.471608665503008e-06,
+      "loss": 1.1127,
+      "step": 2301
+    },
+    {
+      "epoch": 0.46015841683116365,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.471137069693415e-06,
+      "loss": 1.1487,
+      "step": 2302
+    },
+    {
+      "epoch": 0.4603583118862597,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.470665275276387e-06,
+      "loss": 1.0939,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4605582069413558,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.470193282272886e-06,
+      "loss": 1.0531,
+      "step": 2304
+    },
+    {
+      "epoch": 0.46075810199645184,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.469721090703879e-06,
+      "loss": 1.0568,
+      "step": 2305
+    },
+    {
+      "epoch": 0.46095799705154794,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.469248700590336e-06,
+      "loss": 1.0505,
+      "step": 2306
+    },
+    {
+      "epoch": 0.461157892106644,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.468776111953243e-06,
+      "loss": 1.0995,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4613577871617401,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.468303324813595e-06,
+      "loss": 1.1443,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4615576822168362,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.467830339192387e-06,
+      "loss": 1.101,
+      "step": 2309
+    },
+    {
+      "epoch": 0.46175757727193223,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.467357155110636e-06,
+      "loss": 1.08,
+      "step": 2310
+    },
+    {
+      "epoch": 0.46195747232702833,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.466883772589355e-06,
+      "loss": 1.0662,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4621573673821244,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.466410191649575e-06,
+      "loss": 1.0828,
+      "step": 2312
+    },
+    {
+      "epoch": 0.4623572624372205,
+      "grad_norm": 2.125,
+      "learning_rate": 9.46593641231233e-06,
+      "loss": 1.0318,
+      "step": 2313
+    },
+    {
+      "epoch": 0.4625571574923165,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.465462434598669e-06,
+      "loss": 1.0425,
+      "step": 2314
+    },
+    {
+      "epoch": 0.4627570525474126,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.464988258529642e-06,
+      "loss": 1.0506,
+      "step": 2315
+    },
+    {
+      "epoch": 0.46295694760250866,
+      "grad_norm": 2.125,
+      "learning_rate": 9.464513884126312e-06,
+      "loss": 1.1342,
+      "step": 2316
+    },
+    {
+      "epoch": 0.46315684265760476,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.464039311409753e-06,
+      "loss": 1.0277,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4633567377127008,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.463564540401046e-06,
+      "loss": 1.0571,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4635566327677969,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.463089571121278e-06,
+      "loss": 1.1149,
+      "step": 2319
+    },
+    {
+      "epoch": 0.463756527822893,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.462614403591548e-06,
+      "loss": 1.0607,
+      "step": 2320
+    },
+    {
+      "epoch": 0.46395642287798905,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.462139037832963e-06,
+      "loss": 1.0649,
+      "step": 2321
+    },
+    {
+      "epoch": 0.46415631793308515,
+      "grad_norm": 2.375,
+      "learning_rate": 9.461663473866638e-06,
+      "loss": 1.1467,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4643562129881812,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.461187711713697e-06,
+      "loss": 1.1511,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4645561080432773,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.460711751395276e-06,
+      "loss": 1.0512,
+      "step": 2324
+    },
+    {
+      "epoch": 0.46475600309837334,
+      "grad_norm": 2.125,
+      "learning_rate": 9.460235592932515e-06,
+      "loss": 1.0769,
+      "step": 2325
+    },
+    {
+      "epoch": 0.46495589815346944,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.459759236346565e-06,
+      "loss": 1.1492,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4651557932085655,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.459282681658585e-06,
+      "loss": 1.0587,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4653556882636616,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.458805928889747e-06,
+      "loss": 1.0467,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4655555833187576,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.458328978061225e-06,
+      "loss": 1.0781,
+      "step": 2329
+    },
+    {
+      "epoch": 0.4657554783738537,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.457851829194205e-06,
+      "loss": 1.0948,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4659553734289498,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.457374482309885e-06,
+      "loss": 1.141,
+      "step": 2331
+    },
+    {
+      "epoch": 0.46615526848404587,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.456896937429465e-06,
+      "loss": 0.9686,
+      "step": 2332
+    },
+    {
+      "epoch": 0.46635516353914197,
+      "grad_norm": 2.125,
+      "learning_rate": 9.456419194574158e-06,
+      "loss": 1.0933,
+      "step": 2333
+    },
+    {
+      "epoch": 0.466555058594238,
+      "grad_norm": 2.125,
+      "learning_rate": 9.455941253765188e-06,
+      "loss": 1.048,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4667549536493341,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.455463115023783e-06,
+      "loss": 1.105,
+      "step": 2335
+    },
+    {
+      "epoch": 0.46695484870443016,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.45498477837118e-06,
+      "loss": 1.0048,
+      "step": 2336
+    },
+    {
+      "epoch": 0.46715474375952626,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.454506243828633e-06,
+      "loss": 0.9989,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4673546388146223,
+      "grad_norm": 2.25,
+      "learning_rate": 9.454027511417392e-06,
+      "loss": 1.0726,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4675545338697184,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.453548581158726e-06,
+      "loss": 1.1318,
+      "step": 2339
+    },
+    {
+      "epoch": 0.46775442892481445,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.453069453073906e-06,
+      "loss": 1.064,
+      "step": 2340
+    },
+    {
+      "epoch": 0.46795432397991055,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.452590127184217e-06,
+      "loss": 1.0699,
+      "step": 2341
+    },
+    {
+      "epoch": 0.46815421903500665,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.45211060351095e-06,
+      "loss": 1.0564,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4683541140901027,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.451630882075407e-06,
+      "loss": 1.086,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4685540091451988,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.451150962898894e-06,
+      "loss": 1.1009,
+      "step": 2344
+    },
+    {
+      "epoch": 0.46875390420029484,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.450670846002732e-06,
+      "loss": 1.1218,
+      "step": 2345
+    },
+    {
+      "epoch": 0.46895379925539094,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.450190531408245e-06,
+      "loss": 1.067,
+      "step": 2346
+    },
+    {
+      "epoch": 0.469153694310487,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.44971001913677e-06,
+      "loss": 1.0177,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4693535893655831,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.449229309209654e-06,
+      "loss": 0.9842,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4695534844206791,
+      "grad_norm": 2.25,
+      "learning_rate": 9.448748401648244e-06,
+      "loss": 1.0315,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4697533794757752,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.448267296473905e-06,
+      "loss": 1.1507,
+      "step": 2350
+    },
+    {
+      "epoch": 0.46995327453087127,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.44778599370801e-06,
+      "loss": 1.0273,
+      "step": 2351
+    },
+    {
+      "epoch": 0.47015316958596737,
+      "grad_norm": 2.125,
+      "learning_rate": 9.447304493371934e-06,
+      "loss": 1.0988,
+      "step": 2352
+    },
+    {
+      "epoch": 0.47035306464106347,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.44682279548707e-06,
+      "loss": 1.1306,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4705529596961595,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.44634090007481e-06,
+      "loss": 0.9867,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4707528547512556,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.445858807156563e-06,
+      "loss": 1.0614,
+      "step": 2355
+    },
+    {
+      "epoch": 0.47095274980635166,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.445376516753743e-06,
+      "loss": 0.9955,
+      "step": 2356
+    },
+    {
+      "epoch": 0.47115264486144776,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.444894028887773e-06,
+      "loss": 1.0919,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4713525399165438,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.444411343580083e-06,
+      "loss": 0.9668,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4715524349716399,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.443928460852118e-06,
+      "loss": 1.089,
+      "step": 2359
+    },
+    {
+      "epoch": 0.47175233002673594,
+      "grad_norm": 2.375,
+      "learning_rate": 9.443445380725324e-06,
+      "loss": 1.1675,
+      "step": 2360
+    },
+    {
+      "epoch": 0.47195222508183204,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.442962103221161e-06,
+      "loss": 1.0798,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4721521201369281,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.442478628361098e-06,
+      "loss": 1.0142,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4723520151920242,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.441994956166607e-06,
+      "loss": 1.0302,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4725519102471203,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.441511086659175e-06,
+      "loss": 1.0224,
+      "step": 2364
+    },
+    {
+      "epoch": 0.47275180530221633,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.441027019860294e-06,
+      "loss": 1.013,
+      "step": 2365
+    },
+    {
+      "epoch": 0.47295170035731243,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.440542755791467e-06,
+      "loss": 1.1531,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4731515954124085,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.440058294474206e-06,
+      "loss": 1.1263,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4733514904675046,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.439573635930029e-06,
+      "loss": 1.2652,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4735513855226006,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.439088780180465e-06,
+      "loss": 1.0479,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4737512805776967,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.438603727247053e-06,
+      "loss": 0.9835,
+      "step": 2370
+    },
+    {
+      "epoch": 0.47395117563279276,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.438118477151336e-06,
+      "loss": 1.105,
+      "step": 2371
+    },
+    {
+      "epoch": 0.47415107068788886,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.43763302991487e-06,
+      "loss": 1.094,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4743509657429849,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.43714738555922e-06,
+      "loss": 1.0631,
+      "step": 2373
+    },
+    {
+      "epoch": 0.474550860798081,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.436661544105958e-06,
+      "loss": 1.0402,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4747507558531771,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.436175505576663e-06,
+      "loss": 1.0733,
+      "step": 2375
+    },
+    {
+      "epoch": 0.47495065090827315,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.435689269992924e-06,
+      "loss": 0.9708,
+      "step": 2376
+    },
+    {
+      "epoch": 0.47515054596336925,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.435202837376344e-06,
+      "loss": 1.0567,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4753504410184653,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.434716207748527e-06,
+      "loss": 1.0168,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4755503360735614,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.434229381131088e-06,
+      "loss": 1.0316,
+      "step": 2379
+    },
+    {
+      "epoch": 0.47575023112865744,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.433742357545655e-06,
+      "loss": 1.0635,
+      "step": 2380
+    },
+    {
+      "epoch": 0.47595012618375354,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.433255137013861e-06,
+      "loss": 1.0482,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4761500212388496,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.432767719557345e-06,
+      "loss": 1.0387,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4763499162939457,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.432280105197761e-06,
+      "loss": 1.0403,
+      "step": 2383
+    },
+    {
+      "epoch": 0.47654981134904173,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.43179229395677e-06,
+      "loss": 1.186,
+      "step": 2384
+    },
+    {
+      "epoch": 0.47674970640413783,
+      "grad_norm": 2.125,
+      "learning_rate": 9.431304285856037e-06,
+      "loss": 1.0996,
+      "step": 2385
+    },
+    {
+      "epoch": 0.47694960145923393,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.43081608091724e-06,
+      "loss": 1.0019,
+      "step": 2386
+    },
+    {
+      "epoch": 0.47714949651433,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.430327679162068e-06,
+      "loss": 1.0486,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4773493915694261,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.429839080612213e-06,
+      "loss": 1.0941,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4775492866245221,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.429350285289378e-06,
+      "loss": 1.1118,
+      "step": 2389
+    },
+    {
+      "epoch": 0.4777491816796182,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.428861293215278e-06,
+      "loss": 1.0182,
+      "step": 2390
+    },
+    {
+      "epoch": 0.47794907673471426,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.428372104411632e-06,
+      "loss": 1.0864,
+      "step": 2391
+    },
+    {
+      "epoch": 0.47814897178981036,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.427882718900168e-06,
+      "loss": 1.0834,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4783488668449064,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.42739313670263e-06,
+      "loss": 1.0204,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4785487619000025,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.42690335784076e-06,
+      "loss": 0.9799,
+      "step": 2394
+    },
+    {
+      "epoch": 0.47874865695509855,
+      "grad_norm": 2.25,
+      "learning_rate": 9.426413382336314e-06,
+      "loss": 1.16,
+      "step": 2395
+    },
+    {
+      "epoch": 0.47894855201019465,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.42592321021106e-06,
+      "loss": 1.1539,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4791484470652907,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.42543284148677e-06,
+      "loss": 1.108,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4793483421203868,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.424942276185226e-06,
+      "loss": 1.0172,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4795482371754829,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.424451514328218e-06,
+      "loss": 1.0551,
+      "step": 2399
+    },
+    {
+      "epoch": 0.47974813223057894,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.423960555937546e-06,
+      "loss": 1.054,
+      "step": 2400
+    },
+    {
+      "epoch": 0.47994802728567504,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.423469401035019e-06,
+      "loss": 0.9714,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4801479223407711,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.422978049642456e-06,
+      "loss": 1.084,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4803478173958672,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.422486501781676e-06,
+      "loss": 1.0797,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4805477124509632,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.421994757474522e-06,
+      "loss": 1.0522,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4807476075060593,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.421502816742829e-06,
+      "loss": 1.1385,
+      "step": 2405
+    },
+    {
+      "epoch": 0.48094750256115537,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.421010679608455e-06,
+      "loss": 1.0394,
+      "step": 2406
+    },
+    {
+      "epoch": 0.48114739761625147,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.42051834609326e-06,
+      "loss": 1.0059,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4813472926713475,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.420025816219111e-06,
+      "loss": 1.1555,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4815471877264436,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.419533090007888e-06,
+      "loss": 1.0863,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4817470827815397,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.419040167481477e-06,
+      "loss": 1.1842,
+      "step": 2410
+    },
+    {
+      "epoch": 0.48194697783663576,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.418547048661772e-06,
+      "loss": 1.1178,
+      "step": 2411
+    },
+    {
+      "epoch": 0.48214687289173186,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.418053733570682e-06,
+      "loss": 1.0866,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4823467679468279,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.417560222230115e-06,
+      "loss": 1.1088,
+      "step": 2413
+    },
+    {
+      "epoch": 0.482546663001924,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.417066514661995e-06,
+      "loss": 1.0889,
+      "step": 2414
+    },
+    {
+      "epoch": 0.48274655805702005,
+      "grad_norm": 2.125,
+      "learning_rate": 9.416572610888253e-06,
+      "loss": 1.0295,
+      "step": 2415
+    },
+    {
+      "epoch": 0.48294645311211615,
+      "grad_norm": 2.0,
+      "learning_rate": 9.416078510930827e-06,
+      "loss": 1.0452,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4831463481672122,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.415584214811665e-06,
+      "loss": 1.0287,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4833462432223083,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.415089722552723e-06,
+      "loss": 1.0204,
+      "step": 2418
+    },
+    {
+      "epoch": 0.48354613827740434,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.414595034175968e-06,
+      "loss": 1.031,
+      "step": 2419
+    },
+    {
+      "epoch": 0.48374603333250044,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.414100149703373e-06,
+      "loss": 1.1251,
+      "step": 2420
+    },
+    {
+      "epoch": 0.48394592838759654,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.413605069156921e-06,
+      "loss": 1.0836,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4841458234426926,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.413109792558603e-06,
+      "loss": 1.0684,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4843457184977887,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.41261431993042e-06,
+      "loss": 1.091,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4845456135528847,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.41211865129438e-06,
+      "loss": 1.1022,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4847455086079808,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.411622786672499e-06,
+      "loss": 1.1228,
+      "step": 2425
+    },
+    {
+      "epoch": 0.48494540366307687,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.411126726086807e-06,
+      "loss": 1.1744,
+      "step": 2426
+    },
+    {
+      "epoch": 0.48514529871817297,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.410630469559336e-06,
+      "loss": 1.0278,
+      "step": 2427
+    },
+    {
+      "epoch": 0.485345193773269,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.41013401711213e-06,
+      "loss": 1.0577,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4855450888283651,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.409637368767244e-06,
+      "loss": 1.0304,
+      "step": 2429
+    },
+    {
+      "epoch": 0.48574498388346116,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.409140524546736e-06,
+      "loss": 0.9207,
+      "step": 2430
+    },
+    {
+      "epoch": 0.48594487893855726,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.408643484472676e-06,
+      "loss": 1.0671,
+      "step": 2431
+    },
+    {
+      "epoch": 0.48614477399365336,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.408146248567143e-06,
+      "loss": 1.044,
+      "step": 2432
+    },
+    {
+      "epoch": 0.4863446690487494,
+      "grad_norm": 2.25,
+      "learning_rate": 9.407648816852226e-06,
+      "loss": 1.0707,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4865445641038455,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.407151189350019e-06,
+      "loss": 1.0925,
+      "step": 2434
+    },
+    {
+      "epoch": 0.48674445915894154,
+      "grad_norm": 2.125,
+      "learning_rate": 9.406653366082626e-06,
+      "loss": 1.0222,
+      "step": 2435
+    },
+    {
+      "epoch": 0.48694435421403764,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.406155347072162e-06,
+      "loss": 1.0815,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4871442492691337,
+      "grad_norm": 2.25,
+      "learning_rate": 9.405657132340746e-06,
+      "loss": 1.0399,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4873441443242298,
+      "grad_norm": 2.125,
+      "learning_rate": 9.405158721910514e-06,
+      "loss": 1.0846,
+      "step": 2438
+    },
+    {
+      "epoch": 0.48754403937932583,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.4046601158036e-06,
+      "loss": 1.0979,
+      "step": 2439
+    },
+    {
+      "epoch": 0.48774393443442193,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.404161314042155e-06,
+      "loss": 1.0644,
+      "step": 2440
+    },
+    {
+      "epoch": 0.487943829489518,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.403662316648335e-06,
+      "loss": 1.0271,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4881437245446141,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.403163123644303e-06,
+      "loss": 1.0608,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4883436195997102,
+      "grad_norm": 2.125,
+      "learning_rate": 9.402663735052238e-06,
+      "loss": 1.0577,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4885435146548062,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.402164150894318e-06,
+      "loss": 1.1031,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4887434097099023,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.401664371192738e-06,
+      "loss": 0.9917,
+      "step": 2445
+    },
+    {
+      "epoch": 0.48894330476499837,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.401164395969697e-06,
+      "loss": 1.1124,
+      "step": 2446
+    },
+    {
+      "epoch": 0.48914319982009447,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.400664225247402e-06,
+      "loss": 1.1387,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4893430948751905,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.400163859048073e-06,
+      "loss": 1.0916,
+      "step": 2448
+    },
+    {
+      "epoch": 0.4895429899302866,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.399663297393937e-06,
+      "loss": 1.0525,
+      "step": 2449
+    },
+    {
+      "epoch": 0.48974288498538265,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.399162540307225e-06,
+      "loss": 1.0592,
+      "step": 2450
+    },
+    {
+      "epoch": 0.48994278004047875,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.398661587810183e-06,
+      "loss": 1.0823,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4901426750955748,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.398160439925064e-06,
+      "loss": 0.9808,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4903425701506709,
+      "grad_norm": 2.25,
+      "learning_rate": 9.397659096674128e-06,
+      "loss": 1.1753,
+      "step": 2453
+    },
+    {
+      "epoch": 0.490542465205767,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.397157558079644e-06,
+      "loss": 1.1111,
+      "step": 2454
+    },
+    {
+      "epoch": 0.49074236026086304,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.39665582416389e-06,
+      "loss": 0.9791,
+      "step": 2455
+    },
+    {
+      "epoch": 0.49094225531595914,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.396153894949155e-06,
+      "loss": 1.0467,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4911421503710552,
+      "grad_norm": 2.125,
+      "learning_rate": 9.395651770457735e-06,
+      "loss": 1.0843,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4913420454261513,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.39514945071193e-06,
+      "loss": 1.0924,
+      "step": 2458
+    },
+    {
+      "epoch": 0.49154194048124733,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.394646935734057e-06,
+      "loss": 1.1448,
+      "step": 2459
+    },
+    {
+      "epoch": 0.49174183553634343,
+      "grad_norm": 2.25,
+      "learning_rate": 9.394144225546436e-06,
+      "loss": 1.0852,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4919417305914395,
+      "grad_norm": 2.25,
+      "learning_rate": 9.393641320171398e-06,
+      "loss": 1.1342,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4921416256465356,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.393138219631283e-06,
+      "loss": 1.0038,
+      "step": 2462
+    },
+    {
+      "epoch": 0.4923415207016316,
+      "grad_norm": 2.125,
+      "learning_rate": 9.392634923948437e-06,
+      "loss": 0.9803,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4925414157567277,
+      "grad_norm": 2.125,
+      "learning_rate": 9.392131433145216e-06,
+      "loss": 1.0426,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4927413108118238,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.391627747243986e-06,
+      "loss": 1.0738,
+      "step": 2465
+    },
+    {
+      "epoch": 0.49294120586691986,
+      "grad_norm": 2.0,
+      "learning_rate": 9.391123866267121e-06,
+      "loss": 0.9984,
+      "step": 2466
+    },
+    {
+      "epoch": 0.49314110092201596,
+      "grad_norm": 2.125,
+      "learning_rate": 9.390619790237003e-06,
+      "loss": 1.0881,
+      "step": 2467
+    },
+    {
+      "epoch": 0.493340995977112,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.390115519176022e-06,
+      "loss": 1.0613,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4935408910322081,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.38961105310658e-06,
+      "loss": 1.0991,
+      "step": 2469
+    },
+    {
+      "epoch": 0.49374078608730415,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.389106392051083e-06,
+      "loss": 1.0358,
+      "step": 2470
+    },
+    {
+      "epoch": 0.49394068114240025,
+      "grad_norm": 2.125,
+      "learning_rate": 9.388601536031949e-06,
+      "loss": 1.0046,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4941405761974963,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.388096485071603e-06,
+      "loss": 1.0651,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4943404712525924,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.387591239192479e-06,
+      "loss": 1.0863,
+      "step": 2473
+    },
+    {
+      "epoch": 0.49454036630768844,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.387085798417021e-06,
+      "loss": 1.1749,
+      "step": 2474
+    },
+    {
+      "epoch": 0.49474026136278454,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.386580162767682e-06,
+      "loss": 1.025,
+      "step": 2475
+    },
+    {
+      "epoch": 0.49494015641788064,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.38607433226692e-06,
+      "loss": 1.0639,
+      "step": 2476
+    },
+    {
+      "epoch": 0.4951400514729767,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.385568306937204e-06,
+      "loss": 1.1536,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4953399465280728,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.385062086801013e-06,
+      "loss": 1.0838,
+      "step": 2478
+    },
+    {
+      "epoch": 0.4955398415831688,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.384555671880834e-06,
+      "loss": 1.0367,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4957397366382649,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.384049062199157e-06,
+      "loss": 1.0263,
+      "step": 2480
+    },
+    {
+      "epoch": 0.49593963169336097,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.383542257778491e-06,
+      "loss": 0.9039,
+      "step": 2481
+    },
+    {
+      "epoch": 0.49613952674845707,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.383035258641345e-06,
+      "loss": 1.0273,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4963394218035531,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.382528064810242e-06,
+      "loss": 1.0644,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4965393168586492,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.382020676307708e-06,
+      "loss": 1.1447,
+      "step": 2484
+    },
+    {
+      "epoch": 0.49673921191374526,
+      "grad_norm": 2.25,
+      "learning_rate": 9.381513093156286e-06,
+      "loss": 1.0842,
+      "step": 2485
+    },
+    {
+      "epoch": 0.49693910696884136,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.381005315378519e-06,
+      "loss": 1.1256,
+      "step": 2486
+    },
+    {
+      "epoch": 0.49713900202393746,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.380497342996966e-06,
+      "loss": 1.0525,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4973388970790335,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.379989176034187e-06,
+      "loss": 1.0753,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4975387921341296,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.379480814512756e-06,
+      "loss": 1.0823,
+      "step": 2489
+    },
+    {
+      "epoch": 0.49773868718922565,
+      "grad_norm": 2.125,
+      "learning_rate": 9.378972258455256e-06,
+      "loss": 1.0258,
+      "step": 2490
+    },
+    {
+      "epoch": 0.49793858224432175,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.378463507884276e-06,
+      "loss": 1.1388,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4981384772994178,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.377954562822416e-06,
+      "loss": 1.1564,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4983383723545139,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.37744542329228e-06,
+      "loss": 1.1278,
+      "step": 2493
+    },
+    {
+      "epoch": 0.49853826740960994,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.376936089316487e-06,
+      "loss": 1.0647,
+      "step": 2494
+    },
+    {
+      "epoch": 0.49873816246470604,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.376426560917659e-06,
+      "loss": 1.1887,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4989380575198021,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.37591683811843e-06,
+      "loss": 1.0494,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4991379525748982,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.375406920941444e-06,
+      "loss": 1.09,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4993378476299943,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.37489680940935e-06,
+      "loss": 1.0159,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4995377426850903,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.374386503544805e-06,
+      "loss": 1.0739,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4997376377401864,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.37387600337048e-06,
+      "loss": 1.0217,
+      "step": 2500
+    },
+    {
+      "epoch": 0.49993753279528247,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.373365308909052e-06,
+      "loss": 1.0345,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5001374278503785,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.372854420183201e-06,
+      "loss": 1.0266,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5003373229054746,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.372343337215627e-06,
+      "loss": 1.1541,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5005372179605707,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.371832060029027e-06,
+      "loss": 1.0033,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5007371130156668,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.371320588646113e-06,
+      "loss": 1.0522,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5009370080707628,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.370808923089606e-06,
+      "loss": 1.1124,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5011369031258589,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.370297063382235e-06,
+      "loss": 1.0216,
+      "step": 2507
+    },
+    {
+      "epoch": 0.501336798180955,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.369785009546732e-06,
+      "loss": 1.0626,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5015366932360511,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.369272761605848e-06,
+      "loss": 1.076,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5017365882911472,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.368760319582334e-06,
+      "loss": 1.0161,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5019364833462432,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.368247683498952e-06,
+      "loss": 1.1571,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5021363784013393,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.367734853378476e-06,
+      "loss": 1.0536,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5023362734564354,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.367221829243685e-06,
+      "loss": 0.9218,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5025361685115315,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.366708611117366e-06,
+      "loss": 0.9943,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5027360635666275,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.366195199022315e-06,
+      "loss": 1.0951,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5029359586217236,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.365681592981341e-06,
+      "loss": 1.0603,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5031358536768197,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.365167793017258e-06,
+      "loss": 1.0842,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5033357487319158,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.364653799152887e-06,
+      "loss": 1.1236,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5035356437870118,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.36413961141106e-06,
+      "loss": 0.9942,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5037355388421079,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.363625229814617e-06,
+      "loss": 1.0299,
+      "step": 2520
+    },
+    {
+      "epoch": 0.503935433897204,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.363110654386409e-06,
+      "loss": 0.9862,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5041353289523001,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.36259588514929e-06,
+      "loss": 1.0143,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5043352240073962,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.36208092212613e-06,
+      "loss": 1.0184,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5045351190624922,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.361565765339799e-06,
+      "loss": 1.0772,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5047350141175883,
+      "grad_norm": 2.125,
+      "learning_rate": 9.361050414813184e-06,
+      "loss": 1.0264,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5049349091726844,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.360534870569175e-06,
+      "loss": 1.0159,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5051348042277805,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.360019132630672e-06,
+      "loss": 1.1623,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5053346992828764,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.359503201020587e-06,
+      "loss": 0.9668,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5055345943379725,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.358987075761834e-06,
+      "loss": 1.0134,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5057344893930686,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.35847075687734e-06,
+      "loss": 1.116,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5059343844481647,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.35795424439004e-06,
+      "loss": 1.011,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5061342795032608,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.35743753832288e-06,
+      "loss": 1.0904,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5063341745583568,
+      "grad_norm": 2.125,
+      "learning_rate": 9.356920638698809e-06,
+      "loss": 1.0521,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5065340696134529,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.35640354554079e-06,
+      "loss": 0.9193,
+      "step": 2534
+    },
+    {
+      "epoch": 0.506733964668549,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.355886258871786e-06,
+      "loss": 1.0674,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5069338597236451,
+      "grad_norm": 2.0,
+      "learning_rate": 9.355368778714784e-06,
+      "loss": 1.0886,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5071337547787411,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.354851105092765e-06,
+      "loss": 1.099,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5073336498338372,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.354333238028726e-06,
+      "loss": 1.1305,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5075335448889333,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.353815177545666e-06,
+      "loss": 1.0836,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5077334399440294,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.353296923666605e-06,
+      "loss": 1.0657,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5079333349991254,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.352778476414556e-06,
+      "loss": 1.0458,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5081332300542215,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.352259835812556e-06,
+      "loss": 1.059,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5083331251093176,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.351741001883636e-06,
+      "loss": 1.1243,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5085330201644137,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.351221974650846e-06,
+      "loss": 1.1725,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5087329152195098,
+      "grad_norm": 2.0,
+      "learning_rate": 9.350702754137242e-06,
+      "loss": 1.0211,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5089328102746058,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.350183340365884e-06,
+      "loss": 1.0711,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5091327053297019,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.349663733359848e-06,
+      "loss": 1.0382,
+      "step": 2547
+    },
+    {
+      "epoch": 0.509332600384798,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.349143933142214e-06,
+      "loss": 1.1329,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5095324954398941,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.34862393973607e-06,
+      "loss": 1.0075,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5097323904949901,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.348103753164515e-06,
+      "loss": 1.0336,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5099322855500862,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.347583373450657e-06,
+      "loss": 1.1463,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5101321806051823,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.347062800617609e-06,
+      "loss": 1.0115,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5103320756602784,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.346542034688495e-06,
+      "loss": 1.1991,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5105319707153745,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.346021075686448e-06,
+      "loss": 0.978,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5107318657704705,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.345499923634612e-06,
+      "loss": 1.0984,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5109317608255666,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.34497857855613e-06,
+      "loss": 1.0057,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5111316558806627,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.344457040474164e-06,
+      "loss": 1.1259,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5113315509357588,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.343935309411882e-06,
+      "loss": 0.9573,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5115314459908548,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.343413385392457e-06,
+      "loss": 1.0034,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5117313410459509,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.342891268439071e-06,
+      "loss": 1.0915,
+      "step": 2560
+    },
+    {
+      "epoch": 0.511931236101047,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.342368958574921e-06,
+      "loss": 1.0353,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5121311311561431,
+      "grad_norm": 2.125,
+      "learning_rate": 9.341846455823206e-06,
+      "loss": 1.1109,
+      "step": 2562
+    },
+    {
+      "epoch": 0.512331026211239,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.341323760207135e-06,
+      "loss": 1.1072,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5125309212663351,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.340800871749925e-06,
+      "loss": 1.0226,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5127308163214312,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.340277790474804e-06,
+      "loss": 1.0575,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5129307113765273,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.33975451640501e-06,
+      "loss": 1.0841,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5131306064316234,
+      "grad_norm": 2.125,
+      "learning_rate": 9.339231049563779e-06,
+      "loss": 1.141,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5133305014867194,
+      "grad_norm": 2.25,
+      "learning_rate": 9.33870738997437e-06,
+      "loss": 1.1355,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5135303965418155,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.338183537660043e-06,
+      "loss": 1.0538,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5137302915969116,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.337659492644067e-06,
+      "loss": 1.1484,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5139301866520077,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.337135254949719e-06,
+      "loss": 1.0157,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5141300817071037,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.336610824600288e-06,
+      "loss": 1.0814,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5143299767621998,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.336086201619065e-06,
+      "loss": 0.9951,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5145298718172959,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.335561386029356e-06,
+      "loss": 1.0877,
+      "step": 2574
+    },
+    {
+      "epoch": 0.514729766872392,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.335036377854474e-06,
+      "loss": 1.0248,
+      "step": 2575
+    },
+    {
+      "epoch": 0.514929661927488,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.334511177117739e-06,
+      "loss": 1.0022,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5151295569825841,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.333985783842482e-06,
+      "loss": 1.0916,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5153294520376802,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.333460198052036e-06,
+      "loss": 1.0654,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5155293470927763,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.332934419769752e-06,
+      "loss": 1.1317,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5157292421478724,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.332408449018987e-06,
+      "loss": 1.0809,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5159291372029684,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.331882285823098e-06,
+      "loss": 1.1266,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 1.921875,
+      "learning_rate": 9.33135593020546e-06,
+      "loss": 1.0306,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5163289273131606,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.330829382189456e-06,
+      "loss": 1.1622,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5165288223682567,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.330302641798473e-06,
+      "loss": 1.014,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5167287174233527,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.32977570905591e-06,
+      "loss": 1.0604,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5169286124784488,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.329248583985171e-06,
+      "loss": 1.0563,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5171285075335449,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.328721266609673e-06,
+      "loss": 1.1403,
+      "step": 2587
+    },
+    {
+      "epoch": 0.517328402588641,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.328193756952837e-06,
+      "loss": 1.0036,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5175282976437371,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.3276660550381e-06,
+      "loss": 0.8862,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5177281926988331,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.327138160888897e-06,
+      "loss": 1.2041,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5179280877539292,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.32661007452868e-06,
+      "loss": 1.0425,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5181279828090253,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.326081795980903e-06,
+      "loss": 1.0795,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5183278778641214,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.325553325269036e-06,
+      "loss": 1.0518,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5185277729192174,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.325024662416553e-06,
+      "loss": 0.9939,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5187276679743135,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.324495807446935e-06,
+      "loss": 0.9843,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5189275630294096,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.323966760383679e-06,
+      "loss": 1.183,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5191274580845057,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.323437521250278e-06,
+      "loss": 1.0339,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5193273531396017,
+      "grad_norm": 2.375,
+      "learning_rate": 9.322908090070243e-06,
+      "loss": 1.0489,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5195272481946978,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.322378466867095e-06,
+      "loss": 1.2156,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5197271432497939,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.321848651664357e-06,
+      "loss": 1.0942,
+      "step": 2600
+    },
+    {
+      "epoch": 0.51992703830489,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.321318644485561e-06,
+      "loss": 0.9874,
+      "step": 2601
+    },
+    {
+      "epoch": 0.520126933359986,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.320788445354255e-06,
+      "loss": 0.9747,
+      "step": 2602
+    },
+    {
+      "epoch": 0.520326828415082,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.320258054293987e-06,
+      "loss": 1.0756,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5205267234701781,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.319727471328318e-06,
+      "loss": 1.1056,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5207266185252742,
+      "grad_norm": 2.125,
+      "learning_rate": 9.319196696480814e-06,
+      "loss": 1.0712,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5209265135803703,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.318665729775056e-06,
+      "loss": 1.1167,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5211264086354663,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.318134571234626e-06,
+      "loss": 1.0241,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5213263036905624,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.317603220883121e-06,
+      "loss": 1.0953,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5215261987456585,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.317071678744143e-06,
+      "loss": 0.9876,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5217260938007546,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.3165399448413e-06,
+      "loss": 1.016,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5219259888558507,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.316008019198216e-06,
+      "loss": 0.9751,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5221258839109467,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.315475901838514e-06,
+      "loss": 1.0808,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5223257789660428,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.314943592785834e-06,
+      "loss": 1.0874,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5225256740211389,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.314411092063822e-06,
+      "loss": 1.1452,
+      "step": 2614
+    },
+    {
+      "epoch": 0.522725569076235,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.313878399696127e-06,
+      "loss": 1.0408,
+      "step": 2615
+    },
+    {
+      "epoch": 0.522925464131331,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.313345515706417e-06,
+      "loss": 0.955,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5231253591864271,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.31281244011836e-06,
+      "loss": 0.9659,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5233252542415232,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.312279172955634e-06,
+      "loss": 1.0329,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5235251492966193,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.311745714241926e-06,
+      "loss": 1.0165,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5237250443517153,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.311212064000936e-06,
+      "loss": 1.134,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5239249394068114,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.310678222256367e-06,
+      "loss": 1.1248,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5241248344619075,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.31014418903193e-06,
+      "loss": 1.0595,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5243247295170036,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.30960996435135e-06,
+      "loss": 1.0898,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5245246245720997,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.309075548238355e-06,
+      "loss": 1.1028,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5247245196271957,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.308540940716685e-06,
+      "loss": 1.0883,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5249244146822918,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.308006141810086e-06,
+      "loss": 0.973,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5251243097373879,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.307471151542315e-06,
+      "loss": 1.081,
+      "step": 2627
+    },
+    {
+      "epoch": 0.525324204792484,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.306935969937135e-06,
+      "loss": 1.0576,
+      "step": 2628
+    },
+    {
+      "epoch": 0.52552409984758,
+      "grad_norm": 2.125,
+      "learning_rate": 9.30640059701832e-06,
+      "loss": 1.1406,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5257239949026761,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.30586503280965e-06,
+      "loss": 1.162,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5259238899577722,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.305329277334914e-06,
+      "loss": 1.106,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5261237850128683,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.304793330617912e-06,
+      "loss": 1.0658,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5263236800679644,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.304257192682449e-06,
+      "loss": 1.0368,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5265235751230604,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.303720863552343e-06,
+      "loss": 1.0142,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5267234701781565,
+      "grad_norm": 2.25,
+      "learning_rate": 9.303184343251415e-06,
+      "loss": 1.0655,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5269233652332526,
+      "grad_norm": 2.125,
+      "learning_rate": 9.302647631803498e-06,
+      "loss": 1.0741,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5271232602883487,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.302110729232432e-06,
+      "loss": 1.0317,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5273231553434446,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.301573635562068e-06,
+      "loss": 1.0812,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5275230503985407,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.301036350816264e-06,
+      "loss": 1.0591,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5277229454536368,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.300498875018882e-06,
+      "loss": 1.0668,
+      "step": 2640
+    },
+    {
+      "epoch": 0.527922840508733,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.299961208193801e-06,
+      "loss": 1.1023,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5281227355638289,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.299423350364903e-06,
+      "loss": 1.0111,
+      "step": 2642
+    },
+    {
+      "epoch": 0.528322630618925,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.298885301556075e-06,
+      "loss": 1.1069,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5285225256740211,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.298347061791224e-06,
+      "loss": 0.9699,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5287224207291172,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.297808631094257e-06,
+      "loss": 1.0813,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5289223157842133,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.297270009489088e-06,
+      "loss": 0.9616,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5291222108393093,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.296731196999643e-06,
+      "loss": 1.0316,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5293221058944054,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.296192193649857e-06,
+      "loss": 0.9889,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5295220009495015,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.295652999463675e-06,
+      "loss": 0.9873,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5297218960045976,
+      "grad_norm": 2.125,
+      "learning_rate": 9.295113614465045e-06,
+      "loss": 1.1143,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5299217910596936,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.294574038677926e-06,
+      "loss": 1.0498,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5301216861147897,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.294034272126286e-06,
+      "loss": 1.0765,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5303215811698858,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.293494314834105e-06,
+      "loss": 1.0264,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5305214762249819,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.292954166825363e-06,
+      "loss": 1.0044,
+      "step": 2654
+    },
+    {
+      "epoch": 0.530721371280078,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.292413828124056e-06,
+      "loss": 1.0188,
+      "step": 2655
+    },
+    {
+      "epoch": 0.530921266335174,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.291873298754187e-06,
+      "loss": 1.0454,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5311211613902701,
+      "grad_norm": 2.25,
+      "learning_rate": 9.291332578739762e-06,
+      "loss": 1.067,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5313210564453662,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.290791668104802e-06,
+      "loss": 1.0037,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5315209515004623,
+      "grad_norm": 2.125,
+      "learning_rate": 9.290250566873335e-06,
+      "loss": 1.0557,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5317208465555583,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.289709275069396e-06,
+      "loss": 1.0282,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5319207416106544,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.28916779271703e-06,
+      "loss": 1.0812,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5321206366657505,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.288626119840287e-06,
+      "loss": 1.0249,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5323205317208466,
+      "grad_norm": 2.125,
+      "learning_rate": 9.288084256463233e-06,
+      "loss": 1.0339,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5325204267759426,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.287542202609932e-06,
+      "loss": 1.0646,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5327203218310387,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.286999958304464e-06,
+      "loss": 1.0251,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5329202168861348,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.286457523570915e-06,
+      "loss": 1.1246,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5331201119412309,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.285914898433384e-06,
+      "loss": 1.0116,
+      "step": 2667
+    },
+    {
+      "epoch": 0.533320006996327,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.285372082915968e-06,
+      "loss": 1.0878,
+      "step": 2668
+    },
+    {
+      "epoch": 0.533519902051423,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.284829077042784e-06,
+      "loss": 1.0951,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5337197971065191,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.284285880837947e-06,
+      "loss": 1.1074,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5339196921616152,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.283742494325591e-06,
+      "loss": 1.0042,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5341195872167113,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.28319891752985e-06,
+      "loss": 1.064,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5343194822718073,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.282655150474871e-06,
+      "loss": 0.9645,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5345193773269034,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.282111193184806e-06,
+      "loss": 1.0087,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5347192723819995,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.281567045683822e-06,
+      "loss": 1.1043,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5349191674370956,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.281022707996085e-06,
+      "loss": 1.0959,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5351190624921915,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.280478180145778e-06,
+      "loss": 1.1068,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5353189575472876,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.279933462157088e-06,
+      "loss": 1.129,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5355188526023837,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.279388554054207e-06,
+      "loss": 0.9898,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5357187476574798,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.278843455861346e-06,
+      "loss": 1.0103,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5359186427125759,
+      "grad_norm": 2.5,
+      "learning_rate": 9.278298167602716e-06,
+      "loss": 1.1809,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5361185377676719,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.277752689302537e-06,
+      "loss": 0.9956,
+      "step": 2682
+    },
+    {
+      "epoch": 0.536318432822768,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.277207020985042e-06,
+      "loss": 1.0903,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5365183278778641,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.276661162674467e-06,
+      "loss": 0.9674,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5367182229329602,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.27611511439506e-06,
+      "loss": 1.0355,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5369181179880562,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.275568876171078e-06,
+      "loss": 0.9648,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5371180130431523,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.275022448026782e-06,
+      "loss": 0.9565,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5373179080982484,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.274475829986444e-06,
+      "loss": 1.0516,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5375178031533445,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.273929022074348e-06,
+      "loss": 1.0077,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5377176982084406,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.273382024314781e-06,
+      "loss": 1.0389,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5379175932635366,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.272834836732039e-06,
+      "loss": 0.9891,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5381174883186327,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.272287459350432e-06,
+      "loss": 1.0165,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5383173833737288,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.271739892194272e-06,
+      "loss": 0.9821,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5385172784288249,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.271192135287882e-06,
+      "loss": 1.102,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5387171734839209,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.270644188655594e-06,
+      "loss": 0.9087,
+      "step": 2695
+    },
+    {
+      "epoch": 0.538917068539017,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.270096052321747e-06,
+      "loss": 1.1452,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5391169635941131,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.269547726310688e-06,
+      "loss": 1.0612,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5393168586492092,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.268999210646777e-06,
+      "loss": 0.9784,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5395167537043052,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.268450505354375e-06,
+      "loss": 0.9654,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5397166487594013,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.267901610457859e-06,
+      "loss": 1.121,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5399165438144974,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.26735252598161e-06,
+      "loss": 1.073,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5401164388695935,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.266803251950017e-06,
+      "loss": 1.082,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5403163339246896,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.266253788387479e-06,
+      "loss": 1.0063,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5405162289797856,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.2657041353184e-06,
+      "loss": 1.1217,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5407161240348817,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.265154292767204e-06,
+      "loss": 0.9752,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5409160190899778,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.264604260758307e-06,
+      "loss": 1.02,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5411159141450739,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.264054039316146e-06,
+      "loss": 1.0283,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5413158092001699,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.263503628465159e-06,
+      "loss": 1.1819,
+      "step": 2708
+    },
+    {
+      "epoch": 0.541515704255266,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.262953028229794e-06,
+      "loss": 1.0463,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5417155993103621,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.262402238634514e-06,
+      "loss": 1.2011,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5419154943654582,
+      "grad_norm": 2.0,
+      "learning_rate": 9.261851259703782e-06,
+      "loss": 0.9946,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5421153894205543,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.261300091462071e-06,
+      "loss": 1.0925,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5423152844756502,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.260748733933865e-06,
+      "loss": 1.0614,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5425151795307463,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.260197187143656e-06,
+      "loss": 1.0966,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5427150745858424,
+      "grad_norm": 2.125,
+      "learning_rate": 9.259645451115941e-06,
+      "loss": 1.1616,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5429149696409385,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.259093525875232e-06,
+      "loss": 1.1168,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5431148646960345,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.258541411446042e-06,
+      "loss": 1.0294,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5433147597511306,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.257989107852898e-06,
+      "loss": 1.0106,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5435146548062267,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.257436615120335e-06,
+      "loss": 0.9188,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5437145498613228,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.256883933272888e-06,
+      "loss": 1.0516,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5439144449164188,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.256331062335114e-06,
+      "loss": 1.1339,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5441143399715149,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.255778002331569e-06,
+      "loss": 1.0622,
+      "step": 2722
+    },
+    {
+      "epoch": 0.544314235026611,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.255224753286818e-06,
+      "loss": 1.0338,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5445141300817071,
+      "grad_norm": 2.0,
+      "learning_rate": 9.25467131522544e-06,
+      "loss": 1.0663,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5447140251368032,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.254117688172014e-06,
+      "loss": 1.0436,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5449139201918992,
+      "grad_norm": 2.125,
+      "learning_rate": 9.253563872151136e-06,
+      "loss": 1.0577,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5451138152469953,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.253009867187406e-06,
+      "loss": 1.0489,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5453137103020914,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.252455673305431e-06,
+      "loss": 0.9571,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5455136053571875,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.251901290529829e-06,
+      "loss": 1.1165,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5457135004122835,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.251346718885226e-06,
+      "loss": 1.0281,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5459133954673796,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.250791958396255e-06,
+      "loss": 1.0421,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5461132905224757,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.25023700908756e-06,
+      "loss": 1.0919,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5463131855775718,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.249681870983792e-06,
+      "loss": 1.0162,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5465130806326679,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.249126544109608e-06,
+      "loss": 1.0894,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5467129756877639,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.248571028489678e-06,
+      "loss": 1.039,
+      "step": 2735
+    },
+    {
+      "epoch": 0.54691287074286,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.248015324148673e-06,
+      "loss": 1.1551,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5471127657979561,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.247459431111285e-06,
+      "loss": 1.0632,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5473126608530522,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.246903349402201e-06,
+      "loss": 1.1491,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5475125559081482,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.246347079046124e-06,
+      "loss": 1.0886,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5477124509632443,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.245790620067763e-06,
+      "loss": 1.0163,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5479123460183404,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.245233972491836e-06,
+      "loss": 1.1689,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5481122410734365,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.24467713634307e-06,
+      "loss": 0.9898,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5483121361285325,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.2441201116462e-06,
+      "loss": 1.0118,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5485120311836286,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.243562898425967e-06,
+      "loss": 1.0321,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5487119262387247,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.243005496707123e-06,
+      "loss": 1.1106,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5489118212938208,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.242447906514428e-06,
+      "loss": 1.123,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5491117163489169,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.24189012787265e-06,
+      "loss": 1.0215,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5493116114040129,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.241332160806566e-06,
+      "loss": 1.0798,
+      "step": 2748
+    },
+    {
+      "epoch": 0.549511506459109,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.24077400534096e-06,
+      "loss": 1.0293,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5497114015142051,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.240215661500625e-06,
+      "loss": 1.129,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5499112965693012,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.239657129310364e-06,
+      "loss": 1.1994,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5501111916243971,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.239098408794986e-06,
+      "loss": 1.0474,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5503110866794932,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.23853949997931e-06,
+      "loss": 1.0514,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5505109817345893,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.23798040288816e-06,
+      "loss": 1.0656,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5507108767896854,
+      "grad_norm": 2.125,
+      "learning_rate": 9.237421117546375e-06,
+      "loss": 1.0498,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5509107718447815,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.236861643978797e-06,
+      "loss": 0.9487,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5511106668998775,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.236301982210276e-06,
+      "loss": 1.0643,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5513105619549736,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.235742132265676e-06,
+      "loss": 1.0824,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5515104570100697,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.235182094169859e-06,
+      "loss": 1.0396,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5517103520651658,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.234621867947709e-06,
+      "loss": 1.1565,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5519102471202618,
+      "grad_norm": 2.0,
+      "learning_rate": 9.234061453624106e-06,
+      "loss": 0.9588,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5521101421753579,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.233500851223945e-06,
+      "loss": 1.0023,
+      "step": 2762
+    },
+    {
+      "epoch": 0.552310037230454,
+      "grad_norm": 2.125,
+      "learning_rate": 9.23294006077213e-06,
+      "loss": 1.1064,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5525099322855501,
+      "grad_norm": 2.125,
+      "learning_rate": 9.232379082293568e-06,
+      "loss": 1.1066,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5527098273406461,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.23181791581318e-06,
+      "loss": 1.0331,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5529097223957422,
+      "grad_norm": 2.125,
+      "learning_rate": 9.231256561355892e-06,
+      "loss": 0.9992,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5531096174508383,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.23069501894664e-06,
+      "loss": 1.0246,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5533095125059344,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.230133288610366e-06,
+      "loss": 0.978,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5535094075610305,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.229571370372023e-06,
+      "loss": 1.0375,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5537093026161265,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.229009264256574e-06,
+      "loss": 1.0868,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5539091976712226,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.228446970288983e-06,
+      "loss": 1.1305,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5541090927263187,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.22788448849423e-06,
+      "loss": 1.0305,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5543089877814148,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.2273218188973e-06,
+      "loss": 1.1207,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5545088828365108,
+      "grad_norm": 2.125,
+      "learning_rate": 9.226758961523185e-06,
+      "loss": 1.0737,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5547087778916069,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.226195916396888e-06,
+      "loss": 1.1121,
+      "step": 2775
+    },
+    {
+      "epoch": 0.554908672946703,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.225632683543421e-06,
+      "loss": 1.1163,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5551085680017991,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.2250692629878e-06,
+      "loss": 1.1029,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5553084630568952,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.224505654755055e-06,
+      "loss": 1.0431,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5555083581119912,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.22394185887022e-06,
+      "loss": 1.0452,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5557082531670873,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.22337787535834e-06,
+      "loss": 0.9932,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5559081482221834,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.222813704244464e-06,
+      "loss": 1.145,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5561080432772795,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.222249345553654e-06,
+      "loss": 1.0276,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5563079383323755,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.221684799310979e-06,
+      "loss": 0.9823,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5565078333874716,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.221120065541517e-06,
+      "loss": 1.0601,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5567077284425677,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.220555144270353e-06,
+      "loss": 1.0849,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5569076234976638,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.219990035522579e-06,
+      "loss": 0.9661,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5571075185527598,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.2194247393233e-06,
+      "loss": 1.128,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5573074136078559,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.218859255697622e-06,
+      "loss": 1.0929,
+      "step": 2788
+    },
+    {
+      "epoch": 0.557507308662952,
+      "grad_norm": 2.25,
+      "learning_rate": 9.218293584670667e-06,
+      "loss": 0.9534,
+      "step": 2789
+    },
+    {
+      "epoch": 0.557707203718048,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.217727726267564e-06,
+      "loss": 1.1144,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5579070987731441,
+      "grad_norm": 2.125,
+      "learning_rate": 9.217161680513445e-06,
+      "loss": 1.0807,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5581069938282401,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.216595447433454e-06,
+      "loss": 1.1086,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5583068888833362,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.216029027052744e-06,
+      "loss": 1.0106,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5585067839384323,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.215462419396474e-06,
+      "loss": 1.0926,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5587066789935284,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.214895624489813e-06,
+      "loss": 1.0694,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5589065740486244,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.21432864235794e-06,
+      "loss": 1.0344,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5591064691037205,
+      "grad_norm": 2.125,
+      "learning_rate": 9.213761473026039e-06,
+      "loss": 1.1513,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5593063641588166,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.213194116519301e-06,
+      "loss": 1.0874,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5595062592139127,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.212626572862933e-06,
+      "loss": 1.0074,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5597061542690087,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.212058842082141e-06,
+      "loss": 1.0354,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5599060493241048,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.211490924202147e-06,
+      "loss": 0.9803,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5601059443792009,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.210922819248173e-06,
+      "loss": 1.0824,
+      "step": 2802
+    },
+    {
+      "epoch": 0.560305839434297,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.21035452724546e-06,
+      "loss": 1.1036,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5605057344893931,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.209786048219247e-06,
+      "loss": 1.1111,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5607056295444891,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.209217382194788e-06,
+      "loss": 1.1009,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5609055245995852,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.208648529197342e-06,
+      "loss": 1.2159,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5611054196546813,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.208079489252178e-06,
+      "loss": 1.0909,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5613053147097774,
+      "grad_norm": 2.125,
+      "learning_rate": 9.207510262384573e-06,
+      "loss": 1.0151,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5615052097648734,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.20694084861981e-06,
+      "loss": 1.0352,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5617051048199695,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.206371247983186e-06,
+      "loss": 1.0656,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5619049998750656,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.205801460499999e-06,
+      "loss": 1.1071,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5621048949301617,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.205231486195561e-06,
+      "loss": 1.0241,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5623047899852578,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.204661325095189e-06,
+      "loss": 1.1143,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5625046850403538,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.20409097722421e-06,
+      "loss": 1.1146,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5627045800954499,
+      "grad_norm": 2.0,
+      "learning_rate": 9.203520442607958e-06,
+      "loss": 1.01,
+      "step": 2815
+    },
+    {
+      "epoch": 0.562904475150546,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.202949721271779e-06,
+      "loss": 0.963,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5631043702056421,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.20237881324102e-06,
+      "loss": 1.0704,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5633042652607381,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.201807718541046e-06,
+      "loss": 1.0934,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5635041603158342,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.201236437197219e-06,
+      "loss": 1.1211,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5637040553709303,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.200664969234919e-06,
+      "loss": 1.0596,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5639039504260264,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.20009331467953e-06,
+      "loss": 1.0909,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5641038454811224,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.199521473556444e-06,
+      "loss": 1.0574,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5643037405362185,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.198949445891064e-06,
+      "loss": 1.0427,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5645036355913146,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.198377231708795e-06,
+      "loss": 1.0195,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5647035306464107,
+      "grad_norm": 2.25,
+      "learning_rate": 9.197804831035061e-06,
+      "loss": 1.1618,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5649034257015068,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.197232243895285e-06,
+      "loss": 1.0371,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5651033207566027,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.196659470314899e-06,
+      "loss": 1.0595,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5653032158116988,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.196086510319348e-06,
+      "loss": 1.014,
+      "step": 2828
+    },
+    {
+      "epoch": 0.565503110866795,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.195513363934083e-06,
+      "loss": 1.1328,
+      "step": 2829
+    },
+    {
+      "epoch": 0.565703005921891,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.194940031184562e-06,
+      "loss": 1.0148,
+      "step": 2830
+    },
+    {
+      "epoch": 0.565902900976987,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.194366512096252e-06,
+      "loss": 1.1294,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5661027960320831,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.193792806694631e-06,
+      "loss": 1.0612,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5663026910871792,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.19321891500518e-06,
+      "loss": 1.0499,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5665025861422753,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.192644837053394e-06,
+      "loss": 1.1479,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5667024811973714,
+      "grad_norm": 2.125,
+      "learning_rate": 9.192070572864772e-06,
+      "loss": 1.0428,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5669023762524674,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.191496122464822e-06,
+      "loss": 1.0941,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5671022713075635,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.190921485879064e-06,
+      "loss": 1.0604,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5673021663626596,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.19034666313302e-06,
+      "loss": 1.1028,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5675020614177557,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.189771654252221e-06,
+      "loss": 1.0805,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5677019564728517,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.189196459262217e-06,
+      "loss": 1.1518,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5679018515279478,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.188621078188554e-06,
+      "loss": 1.1624,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5681017465830439,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.188045511056787e-06,
+      "loss": 1.0701,
+      "step": 2842
+    },
+    {
+      "epoch": 0.56830164163814,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.187469757892489e-06,
+      "loss": 1.1878,
+      "step": 2843
+    },
+    {
+      "epoch": 0.568501536693236,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.18689381872123e-06,
+      "loss": 1.0141,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5687014317483321,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.186317693568594e-06,
+      "loss": 0.999,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5689013268034282,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.185741382460173e-06,
+      "loss": 1.0444,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5691012218585243,
+      "grad_norm": 2.125,
+      "learning_rate": 9.18516488542157e-06,
+      "loss": 1.078,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5693011169136204,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.184588202478387e-06,
+      "loss": 1.1302,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5695010119687164,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.184011333656243e-06,
+      "loss": 1.0853,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5697009070238125,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.183434278980763e-06,
+      "loss": 1.0982,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5699008020789086,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.18285703847758e-06,
+      "loss": 1.1812,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5701006971340047,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.182279612172332e-06,
+      "loss": 1.0367,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5703005921891007,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.18170200009067e-06,
+      "loss": 1.0112,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5705004872441968,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.181124202258254e-06,
+      "loss": 0.9669,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5707003822992929,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.180546218700748e-06,
+      "loss": 0.9886,
+      "step": 2855
+    },
+    {
+      "epoch": 0.570900277354389,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.179968049443823e-06,
+      "loss": 1.0388,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5711001724094851,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.179389694513166e-06,
+      "loss": 1.0461,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5713000674645811,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.178811153934463e-06,
+      "loss": 1.0419,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5714999625196772,
+      "grad_norm": 2.125,
+      "learning_rate": 9.178232427733418e-06,
+      "loss": 1.0207,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5716998575747733,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.177653515935732e-06,
+      "loss": 1.052,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5718997526298694,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.177074418567124e-06,
+      "loss": 1.0429,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5720996476849654,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.176495135653315e-06,
+      "loss": 0.9709,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5722995427400615,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.175915667220039e-06,
+      "loss": 1.084,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5724994377951576,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.175336013293037e-06,
+      "loss": 1.029,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5726993328502537,
+      "grad_norm": 2.0,
+      "learning_rate": 9.174756173898053e-06,
+      "loss": 1.0011,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5728992279053496,
+      "grad_norm": 2.125,
+      "learning_rate": 9.174176149060847e-06,
+      "loss": 1.1444,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5730991229604457,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.173595938807181e-06,
+      "loss": 1.0194,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5732990180155418,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.173015543162828e-06,
+      "loss": 1.0937,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5734989130706379,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.172434962153573e-06,
+      "loss": 1.0366,
+      "step": 2869
+    },
+    {
+      "epoch": 0.573698808125734,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.171854195805202e-06,
+      "loss": 1.0427,
+      "step": 2870
+    },
+    {
+      "epoch": 0.57389870318083,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.171273244143512e-06,
+      "loss": 1.0721,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5740985982359261,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.170692107194311e-06,
+      "loss": 1.0075,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5742984932910222,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.170110784983413e-06,
+      "loss": 1.0556,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5744983883461183,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.169529277536637e-06,
+      "loss": 1.0713,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5746982834012143,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.168947584879818e-06,
+      "loss": 0.9023,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5748981784563104,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.16836570703879e-06,
+      "loss": 1.1496,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5750980735114065,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.167783644039405e-06,
+      "loss": 1.145,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5752979685665026,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.167201395907515e-06,
+      "loss": 0.9597,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5754978636215987,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.166618962668983e-06,
+      "loss": 1.0683,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5756977586766947,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.166036344349683e-06,
+      "loss": 0.936,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5758976537317908,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.165453540975492e-06,
+      "loss": 1.0168,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5760975487868869,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.1648705525723e-06,
+      "loss": 1.0268,
+      "step": 2882
+    },
+    {
+      "epoch": 0.576297443841983,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.164287379166004e-06,
+      "loss": 1.0488,
+      "step": 2883
+    },
+    {
+      "epoch": 0.576497338897079,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.163704020782507e-06,
+      "loss": 1.0319,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5766972339521751,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.163120477447722e-06,
+      "loss": 1.0859,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5768971290072712,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.16253674918757e-06,
+      "loss": 1.0748,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5770970240623673,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.161952836027978e-06,
+      "loss": 1.0619,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5772969191174633,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.16136873799489e-06,
+      "loss": 1.0363,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5774968141725594,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.160784455114245e-06,
+      "loss": 0.9707,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5776967092276555,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.160199987411997e-06,
+      "loss": 1.1169,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5778966042827516,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.159615334914112e-06,
+      "loss": 1.0517,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5780964993378477,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.159030497646558e-06,
+      "loss": 1.1246,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5782963943929437,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.158445475635312e-06,
+      "loss": 0.9835,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5784962894480398,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.157860268906364e-06,
+      "loss": 0.9527,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5786961845031359,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.157274877485706e-06,
+      "loss": 1.0521,
+      "step": 2895
+    },
+    {
+      "epoch": 0.578896079558232,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.156689301399342e-06,
+      "loss": 1.0526,
+      "step": 2896
+    },
+    {
+      "epoch": 0.579095974613328,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.156103540673282e-06,
+      "loss": 1.0322,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5792958696684241,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.15551759533355e-06,
+      "loss": 1.0069,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5794957647235202,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.154931465406167e-06,
+      "loss": 1.1428,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5796956597786163,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.154345150917173e-06,
+      "loss": 1.1547,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5798955548337122,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.15375865189261e-06,
+      "loss": 1.1787,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5800954498888083,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.153171968358534e-06,
+      "loss": 1.0503,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5802953449439044,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.152585100341e-06,
+      "loss": 1.0329,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5804952399990005,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.151998047866082e-06,
+      "loss": 1.0738,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5806951350540966,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.151410810959853e-06,
+      "loss": 1.0141,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5808950301091926,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.150823389648398e-06,
+      "loss": 1.1399,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5810949251642887,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.150235783957812e-06,
+      "loss": 1.0416,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5812948202193848,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.149647993914196e-06,
+      "loss": 1.0442,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5814947152744809,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.149060019543661e-06,
+      "loss": 1.0853,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5816946103295769,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.148471860872322e-06,
+      "loss": 0.9844,
+      "step": 2910
+    },
+    {
+      "epoch": 0.581894505384673,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.147883517926303e-06,
+      "loss": 1.0472,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5820944004397691,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.147294990731746e-06,
+      "loss": 1.1004,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5822942954948652,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.146706279314786e-06,
+      "loss": 1.075,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5824941905499613,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.146117383701575e-06,
+      "loss": 0.9788,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5826940856050573,
+      "grad_norm": 2.125,
+      "learning_rate": 9.145528303918273e-06,
+      "loss": 1.0786,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5828939806601534,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.144939039991047e-06,
+      "loss": 1.0987,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5830938757152495,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.144349591946072e-06,
+      "loss": 1.0612,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5832937707703456,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.143759959809531e-06,
+      "loss": 1.1392,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5834936658254416,
+      "grad_norm": 2.0,
+      "learning_rate": 9.143170143607613e-06,
+      "loss": 1.0307,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5836935608805377,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.142580143366523e-06,
+      "loss": 1.1968,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5838934559356338,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.141989959112464e-06,
+      "loss": 1.0975,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5840933509907299,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.141399590871652e-06,
+      "loss": 1.0722,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5842932460458259,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.140809038670314e-06,
+      "loss": 1.0535,
+      "step": 2923
+    },
+    {
+      "epoch": 0.584493141100922,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.14021830253468e-06,
+      "loss": 1.1045,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5846930361560181,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.139627382490992e-06,
+      "loss": 1.0238,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5848929312111142,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.1390362785655e-06,
+      "loss": 1.1183,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5850928262662103,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.138444990784455e-06,
+      "loss": 1.0808,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5852927213213063,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.137853519174124e-06,
+      "loss": 1.0787,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5854926163764024,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.137261863760784e-06,
+      "loss": 1.0622,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5856925114314985,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.136670024570715e-06,
+      "loss": 1.0653,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5858924064865946,
+      "grad_norm": 2.125,
+      "learning_rate": 9.136078001630205e-06,
+      "loss": 1.1184,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5860923015416906,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.13548579496555e-06,
+      "loss": 0.9863,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5862921965967867,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.13489340460306e-06,
+      "loss": 1.1144,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5864920916518828,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.134300830569046e-06,
+      "loss": 1.1486,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5866919867069789,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.133708072889828e-06,
+      "loss": 1.0694,
+      "step": 2935
+    },
+    {
+      "epoch": 0.586891881762075,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.133115131591744e-06,
+      "loss": 1.1313,
+      "step": 2936
+    },
+    {
+      "epoch": 0.587091776817171,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.132522006701123e-06,
+      "loss": 0.9837,
+      "step": 2937
+    },
+    {
+      "epoch": 0.587291671872267,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.131928698244317e-06,
+      "loss": 1.2114,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5874915669273632,
+      "grad_norm": 2.125,
+      "learning_rate": 9.13133520624768e-06,
+      "loss": 0.9378,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5876914619824593,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.130741530737573e-06,
+      "loss": 1.0948,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5878913570375552,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.130147671740371e-06,
+      "loss": 1.0223,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5880912520926513,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.129553629282448e-06,
+      "loss": 1.1147,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5882911471477474,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.128959403390195e-06,
+      "loss": 1.0553,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5884910422028435,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.128364994090007e-06,
+      "loss": 1.0579,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5886909372579395,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.127770401408284e-06,
+      "loss": 1.097,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5888908323130356,
+      "grad_norm": 2.125,
+      "learning_rate": 9.127175625371443e-06,
+      "loss": 1.0519,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5890907273681317,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.126580666005901e-06,
+      "loss": 1.0373,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5892906224232278,
+      "grad_norm": 2.125,
+      "learning_rate": 9.125985523338088e-06,
+      "loss": 1.0299,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5894905174783239,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.125390197394437e-06,
+      "loss": 1.0801,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5896904125334199,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.124794688201394e-06,
+      "loss": 1.1661,
+      "step": 2950
+    },
+    {
+      "epoch": 0.589890307588516,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.124198995785414e-06,
+      "loss": 1.1123,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5900902026436121,
+      "grad_norm": 2.0,
+      "learning_rate": 9.123603120172952e-06,
+      "loss": 0.9848,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5902900976987082,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.123007061390481e-06,
+      "loss": 0.9996,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5904899927538042,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.122410819464476e-06,
+      "loss": 1.1073,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5906898878089003,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.121814394421423e-06,
+      "loss": 0.9456,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5908897828639964,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.121217786287815e-06,
+      "loss": 1.0559,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5910896779190925,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.120620995090154e-06,
+      "loss": 0.9933,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5912895729741886,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.120024020854947e-06,
+      "loss": 1.0785,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5914894680292846,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.119426863608713e-06,
+      "loss": 1.1198,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5916893630843807,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.11882952337798e-06,
+      "loss": 0.9539,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5918892581394768,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.118232000189277e-06,
+      "loss": 0.9188,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5920891531945729,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.117634294069148e-06,
+      "loss": 1.1255,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5922890482496689,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.117036405044146e-06,
+      "loss": 1.1125,
+      "step": 2963
+    },
+    {
+      "epoch": 0.592488943304765,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.116438333140825e-06,
+      "loss": 1.0321,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5926888383598611,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.115840078385753e-06,
+      "loss": 1.0814,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5928887334149572,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.115241640805506e-06,
+      "loss": 1.0344,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5930886284700532,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.114643020426662e-06,
+      "loss": 1.0181,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5932885235251493,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.114044217275816e-06,
+      "loss": 1.0138,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5934884185802454,
+      "grad_norm": 2.25,
+      "learning_rate": 9.113445231379565e-06,
+      "loss": 1.1072,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5936883136353415,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.112846062764516e-06,
+      "loss": 0.9869,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5938882086904376,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.112246711457284e-06,
+      "loss": 0.998,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5940881037455336,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.111647177484493e-06,
+      "loss": 1.0892,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5942879988006297,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.111047460872773e-06,
+      "loss": 1.1074,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5944878938557258,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.110447561648766e-06,
+      "loss": 1.0628,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5946877889108219,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.109847479839114e-06,
+      "loss": 1.0733,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5948876839659178,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.109247215470478e-06,
+      "loss": 1.0749,
+      "step": 2976
+    },
+    {
+      "epoch": 0.595087579021014,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.108646768569518e-06,
+      "loss": 0.9976,
+      "step": 2977
+    },
+    {
+      "epoch": 0.59528747407611,
+      "grad_norm": 2.125,
+      "learning_rate": 9.108046139162908e-06,
+      "loss": 1.0781,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5954873691312061,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.107445327277327e-06,
+      "loss": 1.1034,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5956872641863022,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.106844332939464e-06,
+      "loss": 0.968,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5958871592413982,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.106243156176015e-06,
+      "loss": 1.0788,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5960870542964943,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.105641797013682e-06,
+      "loss": 1.058,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5962869493515904,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.10504025547918e-06,
+      "loss": 1.0087,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5964868444066865,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.104438531599227e-06,
+      "loss": 1.013,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5966867394617825,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.103836625400554e-06,
+      "loss": 0.9962,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5968866345168786,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.103234536909895e-06,
+      "loss": 1.0629,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5970865295719747,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.102632266153997e-06,
+      "loss": 1.1041,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5972864246270708,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.102029813159613e-06,
+      "loss": 0.9685,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5974863196821668,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.101427177953502e-06,
+      "loss": 0.9442,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5976862147372629,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.100824360562432e-06,
+      "loss": 0.9826,
+      "step": 2990
+    },
+    {
+      "epoch": 0.597886109792359,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.100221361013185e-06,
+      "loss": 0.9846,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5980860048474551,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.099618179332541e-06,
+      "loss": 1.1631,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5982858999025512,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.099014815547296e-06,
+      "loss": 1.1305,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5984857949576472,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.098411269684251e-06,
+      "loss": 0.9717,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5986856900127433,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.097807541770214e-06,
+      "loss": 1.1103,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5988855850678394,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.097203631832006e-06,
+      "loss": 1.0348,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5990854801229355,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.096599539896447e-06,
+      "loss": 1.1009,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5992853751780315,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.095995265990375e-06,
+      "loss": 1.0759,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5994852702331276,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.095390810140633e-06,
+      "loss": 1.0881,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5996851652882237,
+      "grad_norm": 2.125,
+      "learning_rate": 9.094786172374066e-06,
+      "loss": 1.0333,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5996851652882237,
+      "eval_loss": 0.9266427159309387,
+      "eval_runtime": 594.7722,
+      "eval_samples_per_second": 3.595,
+      "eval_steps_per_second": 3.595,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5998850603433198,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.094181352717535e-06,
+      "loss": 1.0592,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6000849553984159,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.093576351197907e-06,
+      "loss": 1.1619,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6002848504535119,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.092971167842053e-06,
+      "loss": 1.0548,
+      "step": 3003
+    },
+    {
+      "epoch": 0.600484745508608,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.092365802676858e-06,
+      "loss": 1.1491,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6006846405637041,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.091760255729212e-06,
+      "loss": 1.0809,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6008845356188002,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.09115452702601e-06,
+      "loss": 1.0501,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6010844306738962,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.090548616594163e-06,
+      "loss": 1.0777,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6012843257289923,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.089942524460582e-06,
+      "loss": 1.0544,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6014842207840884,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.08933625065219e-06,
+      "loss": 1.0639,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6016841158391845,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.088729795195921e-06,
+      "loss": 1.0678,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6018840108942805,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.08812315811871e-06,
+      "loss": 1.0587,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6020839059493766,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.087516339447504e-06,
+      "loss": 1.0833,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6022838010044727,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.08690933920926e-06,
+      "loss": 1.1271,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6024836960595688,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.08630215743094e-06,
+      "loss": 0.9931,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6026835911146649,
+      "grad_norm": 2.25,
+      "learning_rate": 9.085694794139514e-06,
+      "loss": 1.0981,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6028834861697608,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.085087249361961e-06,
+      "loss": 1.0819,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6030833812248569,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.08447952312527e-06,
+      "loss": 0.9643,
+      "step": 3017
+    },
+    {
+      "epoch": 0.603283276279953,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.083871615456433e-06,
+      "loss": 1.0773,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6034831713350491,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.083263526382457e-06,
+      "loss": 0.9463,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6036830663901451,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.08265525593035e-06,
+      "loss": 1.0375,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6038829614452412,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.082046804127133e-06,
+      "loss": 0.9687,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6040828565003373,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.081438170999833e-06,
+      "loss": 1.1131,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6042827515554334,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.080829356575484e-06,
+      "loss": 0.963,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6044826466105294,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.080220360881133e-06,
+      "loss": 1.1095,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6046825416656255,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.079611183943828e-06,
+      "loss": 1.0107,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6048824367207216,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.079001825790632e-06,
+      "loss": 1.0581,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6050823317758177,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.078392286448607e-06,
+      "loss": 1.1935,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6052822268309138,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.077782565944836e-06,
+      "loss": 1.0421,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6054821218860098,
+      "grad_norm": 2.25,
+      "learning_rate": 9.077172664306396e-06,
+      "loss": 1.1599,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6056820169411059,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.076562581560384e-06,
+      "loss": 1.0622,
+      "step": 3030
+    },
+    {
+      "epoch": 0.605881911996202,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.075952317733894e-06,
+      "loss": 1.0438,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6060818070512981,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.07534187285404e-06,
+      "loss": 0.9847,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6062817021063941,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.074731246947936e-06,
+      "loss": 1.0763,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6064815971614902,
+      "grad_norm": 2.0,
+      "learning_rate": 9.074120440042705e-06,
+      "loss": 0.954,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6066814922165863,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.073509452165476e-06,
+      "loss": 1.0371,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6068813872716824,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.072898283343395e-06,
+      "loss": 1.0441,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6070812823267785,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.072286933603607e-06,
+      "loss": 1.1834,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6072811773818745,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.071675402973268e-06,
+      "loss": 1.1353,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6074810724369706,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.071063691479542e-06,
+      "loss": 1.0175,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6076809674920667,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.070451799149604e-06,
+      "loss": 1.003,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6078808625471628,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.069839726010629e-06,
+      "loss": 1.2073,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6080807576022588,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.06922747208981e-06,
+      "loss": 1.0623,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6082806526573549,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.068615037414339e-06,
+      "loss": 1.0947,
+      "step": 3043
+    },
+    {
+      "epoch": 0.608480547712451,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.068002422011426e-06,
+      "loss": 0.8832,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6086804427675471,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.067389625908277e-06,
+      "loss": 1.0155,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6088803378226431,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.066776649132116e-06,
+      "loss": 1.1535,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6090802328777392,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.06616349171017e-06,
+      "loss": 1.0368,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6092801279328353,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.065550153669676e-06,
+      "loss": 1.1035,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6094800229879314,
+      "grad_norm": 2.0,
+      "learning_rate": 9.06493663503788e-06,
+      "loss": 1.0604,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6096799180430275,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.064322935842032e-06,
+      "loss": 0.999,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6098798130981234,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.063709056109393e-06,
+      "loss": 1.0718,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6100797081532195,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.063094995867232e-06,
+      "loss": 1.0119,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6102796032083156,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.062480755142824e-06,
+      "loss": 0.9728,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6104794982634117,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.061866333963455e-06,
+      "loss": 1.0766,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6106793933185077,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.06125173235642e-06,
+      "loss": 1.0861,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6108792883736038,
+      "grad_norm": 2.0,
+      "learning_rate": 9.060636950349015e-06,
+      "loss": 1.0806,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6110791834286999,
+      "grad_norm": 2.0,
+      "learning_rate": 9.06002198796855e-06,
+      "loss": 1.0579,
+      "step": 3057
+    },
+    {
+      "epoch": 0.611279078483796,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.059406845242343e-06,
+      "loss": 1.0949,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6114789735388921,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.058791522197717e-06,
+      "loss": 0.9956,
+      "step": 3059
+    },
+    {
+      "epoch": 0.6116788685939881,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.058176018862004e-06,
+      "loss": 1.0662,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6118787636490842,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.057560335262546e-06,
+      "loss": 1.0139,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6120786587041803,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.056944471426692e-06,
+      "loss": 1.0547,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6122785537592764,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.056328427381798e-06,
+      "loss": 1.1331,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6124784488143724,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.055712203155226e-06,
+      "loss": 1.0904,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6126783438694685,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.055095798774353e-06,
+      "loss": 1.0879,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6128782389245646,
+      "grad_norm": 2.0,
+      "learning_rate": 9.05447921426656e-06,
+      "loss": 1.0473,
+      "step": 3066
+    },
+    {
+      "epoch": 0.6130781339796607,
+      "grad_norm": 1.9453125,
+      "learning_rate": 9.05386244965923e-06,
+      "loss": 1.0045,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6132780290347567,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.053245504979764e-06,
+      "loss": 0.9839,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6134779240898528,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.052628380255565e-06,
+      "loss": 1.0137,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6136778191449489,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.052011075514049e-06,
+      "loss": 0.985,
+      "step": 3070
+    },
+    {
+      "epoch": 0.613877714200045,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.051393590782631e-06,
+      "loss": 0.9909,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6140776092551411,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.050775926088743e-06,
+      "loss": 1.0162,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6142775043102371,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.050158081459821e-06,
+      "loss": 1.0008,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6144773993653332,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.049540056923309e-06,
+      "loss": 1.0254,
+      "step": 3074
+    },
+    {
+      "epoch": 0.6146772944204293,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.048921852506662e-06,
+      "loss": 1.0035,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6148771894755254,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.048303468237337e-06,
+      "loss": 1.0047,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6150770845306214,
+      "grad_norm": 2.25,
+      "learning_rate": 9.047684904142806e-06,
+      "loss": 1.0216,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6152769795857175,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.047066160250542e-06,
+      "loss": 1.0792,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6154768746408136,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.046447236588032e-06,
+      "loss": 0.9194,
+      "step": 3079
+    },
+    {
+      "epoch": 0.6156767696959097,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.045828133182769e-06,
+      "loss": 1.0588,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6158766647510058,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.045208850062252e-06,
+      "loss": 0.926,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6160765598061018,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.044589387253988e-06,
+      "loss": 1.0148,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6162764548611979,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.043969744785498e-06,
+      "loss": 1.0506,
+      "step": 3083
+    },
+    {
+      "epoch": 0.616476349916294,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.043349922684302e-06,
+      "loss": 1.0651,
+      "step": 3084
+    },
+    {
+      "epoch": 0.6166762449713901,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.042729920977936e-06,
+      "loss": 1.0557,
+      "step": 3085
+    },
+    {
+      "epoch": 0.616876140026486,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.042109739693938e-06,
+      "loss": 1.0763,
+      "step": 3086
+    },
+    {
+      "epoch": 0.6170760350815822,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.041489378859856e-06,
+      "loss": 1.0005,
+      "step": 3087
+    },
+    {
+      "epoch": 0.6172759301366783,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.040868838503247e-06,
+      "loss": 1.0142,
+      "step": 3088
+    },
+    {
+      "epoch": 0.6174758251917744,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.040248118651677e-06,
+      "loss": 1.0612,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6176757202468703,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.039627219332718e-06,
+      "loss": 1.0389,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6178756153019664,
+      "grad_norm": 2.0,
+      "learning_rate": 9.03900614057395e-06,
+      "loss": 0.9776,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6180755103570625,
+      "grad_norm": 2.125,
+      "learning_rate": 9.038384882402957e-06,
+      "loss": 0.9351,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6182754054121586,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.037763444847342e-06,
+      "loss": 1.0853,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6184753004672547,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.037141827934705e-06,
+      "loss": 1.0996,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6186751955223507,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.036520031692658e-06,
+      "loss": 0.9808,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6188750905774468,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.035898056148824e-06,
+      "loss": 1.0767,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6190749856325429,
+      "grad_norm": 2.125,
+      "learning_rate": 9.035275901330828e-06,
+      "loss": 0.9893,
+      "step": 3097
+    },
+    {
+      "epoch": 0.619274880687639,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.03465356726631e-06,
+      "loss": 1.0129,
+      "step": 3098
+    },
+    {
+      "epoch": 0.619474775742735,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.034031053982909e-06,
+      "loss": 1.0106,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6196746707978311,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.03340836150828e-06,
+      "loss": 0.9799,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6198745658529272,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.032785489870084e-06,
+      "loss": 0.9461,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6200744609080233,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.032162439095984e-06,
+      "loss": 1.1415,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6202743559631194,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.031539209213662e-06,
+      "loss": 1.1099,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6204742510182154,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.030915800250797e-06,
+      "loss": 1.03,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6206741460733115,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.030292212235083e-06,
+      "loss": 1.0174,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6208740411284076,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.02966844519422e-06,
+      "loss": 1.0381,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6210739361835037,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.029044499155914e-06,
+      "loss": 0.9533,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6212738312385997,
+      "grad_norm": 2.125,
+      "learning_rate": 9.028420374147885e-06,
+      "loss": 0.9208,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6214737262936958,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.02779607019785e-06,
+      "loss": 1.0589,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6216736213487919,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.027171587333543e-06,
+      "loss": 1.0516,
+      "step": 3110
+    },
+    {
+      "epoch": 0.621873516403888,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.026546925582707e-06,
+      "loss": 1.038,
+      "step": 3111
+    },
+    {
+      "epoch": 0.622073411458984,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.025922084973084e-06,
+      "loss": 1.0842,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6222733065140801,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.025297065532435e-06,
+      "loss": 1.1177,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6224732015691762,
+      "grad_norm": 2.25,
+      "learning_rate": 9.024671867288518e-06,
+      "loss": 1.1309,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6226730966242723,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.024046490269107e-06,
+      "loss": 1.1373,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6228729916793684,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.023420934501981e-06,
+      "loss": 1.0819,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6230728867344644,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.022795200014927e-06,
+      "loss": 0.9935,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6232727817895605,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.022169286835737e-06,
+      "loss": 0.9575,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6234726768446566,
+      "grad_norm": 2.125,
+      "learning_rate": 9.02154319499222e-06,
+      "loss": 1.0777,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6236725718997527,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.020916924512183e-06,
+      "loss": 1.0401,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6238724669548487,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.020290475423447e-06,
+      "loss": 0.9766,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6240723620099448,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.019663847753837e-06,
+      "loss": 1.0157,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6242722570650409,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.019037041531187e-06,
+      "loss": 0.9897,
+      "step": 3123
+    },
+    {
+      "epoch": 0.624472152120137,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.01841005678334e-06,
+      "loss": 0.974,
+      "step": 3124
+    },
+    {
+      "epoch": 0.624672047175233,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.017782893538149e-06,
+      "loss": 1.0121,
+      "step": 3125
+    },
+    {
+      "epoch": 0.624871942230329,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.01715555182347e-06,
+      "loss": 1.0159,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6250718372854251,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.016528031667173e-06,
+      "loss": 1.0378,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6252717323405212,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.015900333097127e-06,
+      "loss": 1.0614,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6254716273956173,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.015272456141218e-06,
+      "loss": 1.041,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6256715224507133,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.014644400827336e-06,
+      "loss": 1.0148,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6258714175058094,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.014016167183378e-06,
+      "loss": 1.0268,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6260713125609055,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.013387755237251e-06,
+      "loss": 0.9141,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6262712076160016,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.012759165016867e-06,
+      "loss": 1.0393,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6264711026710976,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.01213039655015e-06,
+      "loss": 1.0668,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6266709977261937,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.01150144986503e-06,
+      "loss": 1.0435,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6268708927812898,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.010872324989444e-06,
+      "loss": 1.0377,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6270707878363859,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.010243021951338e-06,
+      "loss": 1.021,
+      "step": 3137
+    },
+    {
+      "epoch": 0.627270682891482,
+      "grad_norm": 2.0,
+      "learning_rate": 9.009613540778666e-06,
+      "loss": 0.9202,
+      "step": 3138
+    },
+    {
+      "epoch": 0.627470577946578,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.008983881499387e-06,
+      "loss": 1.0746,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6276704730016741,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.008354044141471e-06,
+      "loss": 0.9922,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6278703680567702,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.0077240287329e-06,
+      "loss": 1.1474,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6280702631118663,
+      "grad_norm": 2.0,
+      "learning_rate": 9.007093835301652e-06,
+      "loss": 1.0599,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6282701581669623,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.006463463875728e-06,
+      "loss": 1.044,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6284700532220584,
+      "grad_norm": 2.0,
+      "learning_rate": 9.005832914483121e-06,
+      "loss": 0.971,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6286699482771545,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.005202187151845e-06,
+      "loss": 1.0,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6288698433322506,
+      "grad_norm": 2.0,
+      "learning_rate": 9.004571281909918e-06,
+      "loss": 1.0634,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6290697383873466,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.00394019878536e-06,
+      "loss": 0.9944,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6292696334424427,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.003308937806206e-06,
+      "loss": 1.077,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6294695284975388,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.002677499000496e-06,
+      "loss": 1.1407,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6296694235526349,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.002045882396279e-06,
+      "loss": 1.0637,
+      "step": 3150
+    },
+    {
+      "epoch": 0.629869318607731,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.001414088021612e-06,
+      "loss": 1.0534,
+      "step": 3151
+    },
+    {
+      "epoch": 0.630069213662827,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.00078211590456e-06,
+      "loss": 1.0054,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6302691087179231,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.000149966073192e-06,
+      "loss": 0.9175,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6304690037730192,
+      "grad_norm": 2.125,
+      "learning_rate": 8.99951763855559e-06,
+      "loss": 1.1064,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6306688988281153,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.998885133379842e-06,
+      "loss": 1.07,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6308687938832113,
+      "grad_norm": 2.125,
+      "learning_rate": 8.998252450574044e-06,
+      "loss": 1.0637,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6310686889383074,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.997619590166298e-06,
+      "loss": 1.1576,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6312685839934035,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.996986552184716e-06,
+      "loss": 1.0527,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6314684790484996,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.996353336657421e-06,
+      "loss": 1.0808,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6316683741035957,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.995719943612535e-06,
+      "loss": 1.2639,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6318682691586917,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.995086373078197e-06,
+      "loss": 0.948,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6320681642137878,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.99445262508255e-06,
+      "loss": 0.9227,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6322680592688839,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.993818699653742e-06,
+      "loss": 1.0668,
+      "step": 3163
+    },
+    {
+      "epoch": 0.63246795432398,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.993184596819935e-06,
+      "loss": 1.103,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6326678493790759,
+      "grad_norm": 2.125,
+      "learning_rate": 8.992550316609294e-06,
+      "loss": 1.0526,
+      "step": 3165
+    },
+    {
+      "epoch": 0.632867744434172,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.991915859049996e-06,
+      "loss": 1.105,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6330676394892681,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.99128122417022e-06,
+      "loss": 0.9941,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6332675345443642,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.990646411998161e-06,
+      "loss": 1.0805,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6334674295994602,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.990011422562012e-06,
+      "loss": 0.986,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6336673246545563,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.989376255889982e-06,
+      "loss": 1.0714,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6338672197096524,
+      "grad_norm": 2.125,
+      "learning_rate": 8.988740912010285e-06,
+      "loss": 1.1068,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6340671147647485,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.988105390951143e-06,
+      "loss": 1.0174,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6342670098198446,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.987469692740787e-06,
+      "loss": 1.1122,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6344669048749406,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.98683381740745e-06,
+      "loss": 1.0848,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6346667999300367,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.986197764979382e-06,
+      "loss": 1.025,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6348666949851328,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.985561535484836e-06,
+      "loss": 0.9516,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6350665900402289,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.984925128952072e-06,
+      "loss": 1.1565,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6352664850953249,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.984288545409358e-06,
+      "loss": 1.0739,
+      "step": 3178
+    },
+    {
+      "epoch": 0.635466380150421,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.983651784884974e-06,
+      "loss": 1.0221,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6356662752055171,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.983014847407202e-06,
+      "loss": 1.0373,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6358661702606132,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.982377733004338e-06,
+      "loss": 1.0651,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6360660653157093,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.981740441704677e-06,
+      "loss": 1.1086,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6362659603708053,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.981102973536533e-06,
+      "loss": 1.1332,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6364658554259014,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.98046532852822e-06,
+      "loss": 1.0953,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6366657504809975,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.97982750670806e-06,
+      "loss": 1.0265,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6368656455360936,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.979189508104391e-06,
+      "loss": 1.089,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6370655405911896,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.978551332745546e-06,
+      "loss": 1.0756,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6372654356462857,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.977912980659878e-06,
+      "loss": 1.0064,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6374653307013818,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.97727445187574e-06,
+      "loss": 1.0146,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6376652257564779,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.976635746421493e-06,
+      "loss": 1.0238,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6378651208115739,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.975996864325514e-06,
+      "loss": 0.9988,
+      "step": 3191
+    },
+    {
+      "epoch": 0.63806501586667,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.975357805616176e-06,
+      "loss": 1.0451,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6382649109217661,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.974718570321873e-06,
+      "loss": 1.0624,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6384648059768622,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.974079158470991e-06,
+      "loss": 0.9772,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6386647010319583,
+      "grad_norm": 2.125,
+      "learning_rate": 8.97343957009194e-06,
+      "loss": 1.0772,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6388645960870543,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.972799805213125e-06,
+      "loss": 0.962,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6390644911421504,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.97215986386297e-06,
+      "loss": 1.1486,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6392643861972465,
+      "grad_norm": 2.125,
+      "learning_rate": 8.971519746069897e-06,
+      "loss": 1.0017,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6394642812523426,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.970879451862341e-06,
+      "loss": 1.0565,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6396641763074385,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.970238981268745e-06,
+      "loss": 1.0777,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6398640713625346,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.969598334317556e-06,
+      "loss": 0.9181,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6400639664176307,
+      "grad_norm": 2.125,
+      "learning_rate": 8.968957511037233e-06,
+      "loss": 1.0161,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6402638614727268,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.968316511456241e-06,
+      "loss": 1.1003,
+      "step": 3203
+    },
+    {
+      "epoch": 0.640463756527823,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.967675335603055e-06,
+      "loss": 1.0365,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6406636515829189,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.967033983506153e-06,
+      "loss": 0.9613,
+      "step": 3205
+    },
+    {
+      "epoch": 0.640863546638015,
+      "grad_norm": 2.0,
+      "learning_rate": 8.966392455194026e-06,
+      "loss": 1.041,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6410634416931111,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.965750750695168e-06,
+      "loss": 1.0774,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6412633367482072,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.965108870038088e-06,
+      "loss": 1.0497,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6414632318033032,
+      "grad_norm": 2.125,
+      "learning_rate": 8.964466813251294e-06,
+      "loss": 1.1386,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6416631268583993,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.963824580363307e-06,
+      "loss": 1.098,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6418630219134954,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.963182171402656e-06,
+      "loss": 1.1158,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6420629169685915,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.962539586397876e-06,
+      "loss": 1.1141,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6422628120236875,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.961896825377512e-06,
+      "loss": 1.0352,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6424627070787836,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.961253888370113e-06,
+      "loss": 0.9783,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6426626021338797,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.960610775404239e-06,
+      "loss": 0.9919,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6428624971889758,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.95996748650846e-06,
+      "loss": 0.9799,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6430623922440719,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.959324021711345e-06,
+      "loss": 0.9817,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6432622872991679,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.95868038104148e-06,
+      "loss": 1.13,
+      "step": 3218
+    },
+    {
+      "epoch": 0.643462182354264,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.958036564527457e-06,
+      "loss": 1.1042,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6436620774093601,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.957392572197871e-06,
+      "loss": 0.9991,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6438619724644562,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.95674840408133e-06,
+      "loss": 0.9395,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6440618675195522,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.95610406020645e-06,
+      "loss": 1.0837,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6442617625746483,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.955459540601847e-06,
+      "loss": 1.0176,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6444616576297444,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.954814845296153e-06,
+      "loss": 1.0846,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6446615526848405,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.95416997431801e-06,
+      "loss": 0.9247,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6448614477399366,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.953524927696056e-06,
+      "loss": 1.0291,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6450613427950326,
+      "grad_norm": 2.125,
+      "learning_rate": 8.952879705458949e-06,
+      "loss": 1.0674,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6452612378501287,
+      "grad_norm": 2.125,
+      "learning_rate": 8.952234307635346e-06,
+      "loss": 1.0894,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6454611329052248,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.951588734253917e-06,
+      "loss": 1.023,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6456610279603209,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.950942985343339e-06,
+      "loss": 0.9906,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6458609230154169,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.950297060932294e-06,
+      "loss": 0.9879,
+      "step": 3231
+    },
+    {
+      "epoch": 0.646060818070513,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.949650961049479e-06,
+      "loss": 1.0749,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6462607131256091,
+      "grad_norm": 2.125,
+      "learning_rate": 8.949004685723587e-06,
+      "loss": 1.0911,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6464606081807052,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.94835823498333e-06,
+      "loss": 0.9759,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6466605032358012,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.94771160885742e-06,
+      "loss": 1.0999,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6468603982908973,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.947064807374586e-06,
+      "loss": 0.9683,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6470602933459934,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.946417830563551e-06,
+      "loss": 1.0466,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6472601884010895,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.94577067845306e-06,
+      "loss": 1.0536,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6474600834561856,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.945123351071856e-06,
+      "loss": 1.0053,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6476599785112815,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.944475848448692e-06,
+      "loss": 1.0653,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6478598735663776,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.943828170612335e-06,
+      "loss": 1.0131,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6480597686214737,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.94318031759155e-06,
+      "loss": 1.125,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6482596636765698,
+      "grad_norm": 2.0,
+      "learning_rate": 8.942532289415117e-06,
+      "loss": 1.025,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6484595587316658,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.941884086111824e-06,
+      "loss": 1.0884,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6486594537867619,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.941235707710457e-06,
+      "loss": 1.0583,
+      "step": 3245
+    },
+    {
+      "epoch": 0.648859348841858,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.940587154239822e-06,
+      "loss": 1.0255,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6490592438969541,
+      "grad_norm": 2.125,
+      "learning_rate": 8.939938425728725e-06,
+      "loss": 1.1508,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6492591389520501,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.939289522205986e-06,
+      "loss": 1.1196,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6494590340071462,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.938640443700426e-06,
+      "loss": 0.9891,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6496589290622423,
+      "grad_norm": 2.125,
+      "learning_rate": 8.937991190240878e-06,
+      "loss": 1.0844,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6498588241173384,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.937341761856184e-06,
+      "loss": 1.0354,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6500587191724345,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.936692158575186e-06,
+      "loss": 1.0588,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6502586142275305,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.936042380426746e-06,
+      "loss": 0.9619,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6504585092826266,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.935392427439723e-06,
+      "loss": 0.9899,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6506584043377227,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.934742299642987e-06,
+      "loss": 1.1772,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6508582993928188,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.93409199706542e-06,
+      "loss": 1.025,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6510581944479148,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.933441519735907e-06,
+      "loss": 1.0423,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6512580895030109,
+      "grad_norm": 2.125,
+      "learning_rate": 8.932790867683339e-06,
+      "loss": 1.0719,
+      "step": 3258
+    },
+    {
+      "epoch": 0.651457984558107,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.932140040936623e-06,
+      "loss": 1.0208,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6516578796132031,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.931489039524667e-06,
+      "loss": 1.0917,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6518577746682992,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.930837863476386e-06,
+      "loss": 1.0357,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6520576697233952,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.930186512820707e-06,
+      "loss": 1.02,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6522575647784913,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.929534987586565e-06,
+      "loss": 1.1055,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6524574598335874,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.928883287802897e-06,
+      "loss": 1.0013,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6526573548886835,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.928231413498652e-06,
+      "loss": 0.9495,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6528572499437795,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.92757936470279e-06,
+      "loss": 0.9701,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6530571449988756,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.92692714144427e-06,
+      "loss": 1.0784,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6532570400539717,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.926274743752065e-06,
+      "loss": 1.0928,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6534569351090678,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.925622171655157e-06,
+      "loss": 1.0559,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6536568301641638,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.92496942518253e-06,
+      "loss": 0.9591,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6538567252192599,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.924316504363182e-06,
+      "loss": 1.0161,
+      "step": 3271
+    },
+    {
+      "epoch": 0.654056620274356,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.923663409226112e-06,
+      "loss": 1.0892,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6542565153294521,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.923010139800335e-06,
+      "loss": 1.0097,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6544564103845482,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.922356696114865e-06,
+      "loss": 1.0528,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6546563054396441,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.921703078198728e-06,
+      "loss": 1.0536,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6548562004947402,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.92104928608096e-06,
+      "loss": 1.0184,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6550560955498363,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.920395319790604e-06,
+      "loss": 0.9432,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6552559906049324,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.919741179356705e-06,
+      "loss": 1.126,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6554558856600284,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.919086864808319e-06,
+      "loss": 1.0793,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6556557807151245,
+      "grad_norm": 2.125,
+      "learning_rate": 8.918432376174516e-06,
+      "loss": 1.0164,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6558556757702206,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.917777713484366e-06,
+      "loss": 1.0069,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6560555708253167,
+      "grad_norm": 2.0,
+      "learning_rate": 8.917122876766946e-06,
+      "loss": 1.0158,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6562554658804128,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.916467866051347e-06,
+      "loss": 1.0721,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6564553609355088,
+      "grad_norm": 2.125,
+      "learning_rate": 8.915812681366665e-06,
+      "loss": 1.0576,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6566552559906049,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.915157322742e-06,
+      "loss": 1.0276,
+      "step": 3285
+    },
+    {
+      "epoch": 0.656855151045701,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.914501790206466e-06,
+      "loss": 0.9683,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6570550461007971,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.913846083789179e-06,
+      "loss": 0.9412,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6572549411558931,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.91319020351927e-06,
+      "loss": 1.0381,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6574548362109892,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.912534149425868e-06,
+      "loss": 1.049,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6576547312660853,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.911877921538117e-06,
+      "loss": 1.0618,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6578546263211814,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.911221519885167e-06,
+      "loss": 1.0587,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6580545213762774,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.910564944496174e-06,
+      "loss": 0.9946,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6582544164313735,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.909908195400305e-06,
+      "loss": 1.0538,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6584543114864696,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.909251272626731e-06,
+      "loss": 1.0824,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6586542065415657,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.908594176204632e-06,
+      "loss": 1.0578,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6588541015966618,
+      "grad_norm": 2.125,
+      "learning_rate": 8.9079369061632e-06,
+      "loss": 1.0767,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6590539966517578,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.907279462531625e-06,
+      "loss": 1.0766,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6592538917068539,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.906621845339115e-06,
+      "loss": 0.9306,
+      "step": 3298
+    },
+    {
+      "epoch": 0.65945378676195,
+      "grad_norm": 2.0,
+      "learning_rate": 8.90596405461488e-06,
+      "loss": 0.8914,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6596536818170461,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.905306090388137e-06,
+      "loss": 1.0862,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6598535768721421,
+      "grad_norm": 2.0,
+      "learning_rate": 8.904647952688117e-06,
+      "loss": 0.9762,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6600534719272382,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.903989641544052e-06,
+      "loss": 1.0451,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6602533669823343,
+      "grad_norm": 2.125,
+      "learning_rate": 8.903331156985181e-06,
+      "loss": 1.044,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6604532620374304,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.902672499040759e-06,
+      "loss": 1.0062,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6606531570925265,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.902013667740043e-06,
+      "loss": 1.0294,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6608530521476225,
+      "grad_norm": 2.125,
+      "learning_rate": 8.901354663112294e-06,
+      "loss": 1.0725,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6610529472027186,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.900695485186788e-06,
+      "loss": 1.0517,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6612528422578147,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.900036133992807e-06,
+      "loss": 0.9954,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6614527373129108,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.899376609559636e-06,
+      "loss": 1.0294,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6616526323680068,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.898716911916571e-06,
+      "loss": 1.0569,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6618525274231029,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.89805704109292e-06,
+      "loss": 0.9953,
+      "step": 3311
+    },
+    {
+      "epoch": 0.662052422478199,
+      "grad_norm": 2.0,
+      "learning_rate": 8.897396997117991e-06,
+      "loss": 1.005,
+      "step": 3312
+    },
+    {
+      "epoch": 0.662252317533295,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.896736780021102e-06,
+      "loss": 1.0816,
+      "step": 3313
+    },
+    {
+      "epoch": 0.662452212588391,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.896076389831583e-06,
+      "loss": 0.9675,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6626521076434871,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.895415826578766e-06,
+      "loss": 1.1988,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6628520026985832,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.894755090291996e-06,
+      "loss": 1.1341,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6630518977536793,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.89409418100062e-06,
+      "loss": 1.0202,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6632517928087754,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.893433098733995e-06,
+      "loss": 1.0561,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6634516878638714,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.892771843521487e-06,
+      "loss": 1.1008,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6636515829189675,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.89211041539247e-06,
+      "loss": 1.0711,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6638514779740636,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.891448814376326e-06,
+      "loss": 1.0149,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6640513730291597,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.89078704050244e-06,
+      "loss": 0.9886,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6642512680842557,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.890125093800208e-06,
+      "loss": 1.0175,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6644511631393518,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.889462974299037e-06,
+      "loss": 1.0867,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6646510581944479,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.888800682028334e-06,
+      "loss": 1.0379,
+      "step": 3325
+    },
+    {
+      "epoch": 0.664850953249544,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.888138217017521e-06,
+      "loss": 1.0689,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6650508483046401,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.887475579296025e-06,
+      "loss": 1.0328,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6652507433597361,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.886812768893277e-06,
+      "loss": 1.0361,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6654506384148322,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.886149785838722e-06,
+      "loss": 1.0431,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6656505334699283,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.885486630161808e-06,
+      "loss": 1.0075,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6658504285250244,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.884823301891993e-06,
+      "loss": 1.0585,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6660503235801204,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.884159801058743e-06,
+      "loss": 1.0838,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6662502186352165,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.88349612769153e-06,
+      "loss": 1.0646,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6664501136903126,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.88283228181983e-06,
+      "loss": 0.8971,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6666500087454087,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.882168263473137e-06,
+      "loss": 1.0699,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6668499038005047,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.881504072680945e-06,
+      "loss": 0.9898,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6670497988556008,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.880839709472755e-06,
+      "loss": 1.0643,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6672496939106969,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.88017517387808e-06,
+      "loss": 1.1073,
+      "step": 3338
+    },
+    {
+      "epoch": 0.667449588965793,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.87951046592644e-06,
+      "loss": 1.0848,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6676494840208891,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.878845585647357e-06,
+      "loss": 1.0101,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6678493790759851,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.87818053307037e-06,
+      "loss": 1.055,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6680492741310812,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.877515308225015e-06,
+      "loss": 0.9716,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6682491691861773,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.876849911140846e-06,
+      "loss": 0.9588,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6684490642412734,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.876184341847418e-06,
+      "loss": 1.0359,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6686489592963694,
+      "grad_norm": 2.125,
+      "learning_rate": 8.875518600374296e-06,
+      "loss": 1.0397,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6688488543514655,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.874852686751051e-06,
+      "loss": 1.0957,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6690487494065616,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.874186601007263e-06,
+      "loss": 1.0045,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6692486444616577,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.87352034317252e-06,
+      "loss": 1.0593,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6694485395167536,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.872853913276418e-06,
+      "loss": 1.045,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6696484345718497,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.872187311348558e-06,
+      "loss": 1.1227,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6698483296269458,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.871520537418552e-06,
+      "loss": 0.9508,
+      "step": 3351
+    },
+    {
+      "epoch": 0.670048224682042,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.870853591516016e-06,
+      "loss": 0.9973,
+      "step": 3352
+    },
+    {
+      "epoch": 0.670248119737138,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.870186473670577e-06,
+      "loss": 1.0057,
+      "step": 3353
+    },
+    {
+      "epoch": 0.670448014792234,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.86951918391187e-06,
+      "loss": 1.0588,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6706479098473301,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.868851722269531e-06,
+      "loss": 0.987,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6708478049024262,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.868184088773216e-06,
+      "loss": 1.1464,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6710476999575223,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.867516283452572e-06,
+      "loss": 1.0607,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6712475950126183,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.866848306337272e-06,
+      "loss": 0.99,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6714474900677144,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.866180157456981e-06,
+      "loss": 1.0594,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6716473851228105,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.865511836841381e-06,
+      "loss": 1.0414,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6718472801779066,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.864843344520158e-06,
+      "loss": 1.16,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6720471752330027,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.864174680523005e-06,
+      "loss": 1.0834,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6722470702880987,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.863505844879628e-06,
+      "loss": 1.1152,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6724469653431948,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.862836837619732e-06,
+      "loss": 0.9614,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6726468603982909,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.862167658773037e-06,
+      "loss": 0.9623,
+      "step": 3365
+    },
+    {
+      "epoch": 0.672846755453387,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.861498308369267e-06,
+      "loss": 1.1579,
+      "step": 3366
+    },
+    {
+      "epoch": 0.673046650508483,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.860828786438155e-06,
+      "loss": 0.9972,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6732465455635791,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.86015909300944e-06,
+      "loss": 1.0825,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6734464406186752,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.85948922811287e-06,
+      "loss": 1.0788,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6736463356737713,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.858819191778201e-06,
+      "loss": 1.1012,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6738462307288673,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.858148984035196e-06,
+      "loss": 0.9939,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6740461257839634,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.857478604913625e-06,
+      "loss": 1.1226,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6742460208390595,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.856808054443266e-06,
+      "loss": 0.9636,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6744459158941556,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.856137332653907e-06,
+      "loss": 0.9492,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6746458109492517,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.855466439575338e-06,
+      "loss": 1.0091,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6748457060043477,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.85479537523736e-06,
+      "loss": 1.0131,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6750456010594438,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.854124139669786e-06,
+      "loss": 1.013,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6752454961145399,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.853452732902428e-06,
+      "loss": 1.0919,
+      "step": 3378
+    },
+    {
+      "epoch": 0.675445391169636,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.85278115496511e-06,
+      "loss": 1.0912,
+      "step": 3379
+    },
+    {
+      "epoch": 0.675645286224732,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.852109405887667e-06,
+      "loss": 1.1148,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6758451812798281,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.851437485699935e-06,
+      "loss": 0.9866,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6760450763349242,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.85076539443176e-06,
+      "loss": 0.9758,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6762449713900203,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.850093132112999e-06,
+      "loss": 0.9487,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6764448664451164,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.849420698773513e-06,
+      "loss": 1.0743,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6766447615002124,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.848748094443167e-06,
+      "loss": 1.088,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6768446565553085,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.848075319151844e-06,
+      "loss": 1.0013,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6770445516104046,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.847402372929426e-06,
+      "loss": 1.1631,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6772444466655007,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.846729255805806e-06,
+      "loss": 1.0131,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6774443417205966,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.846055967810882e-06,
+      "loss": 1.0088,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6776442367756927,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.845382508974565e-06,
+      "loss": 1.0365,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6778441318307888,
+      "grad_norm": 2.0,
+      "learning_rate": 8.844708879326767e-06,
+      "loss": 0.9699,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6780440268858849,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.84403507889741e-06,
+      "loss": 0.9918,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6782439219409809,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.843361107716427e-06,
+      "loss": 1.0832,
+      "step": 3393
+    },
+    {
+      "epoch": 0.678443816996077,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.842686965813752e-06,
+      "loss": 0.992,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6786437120511731,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.842012653219333e-06,
+      "loss": 1.0853,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6788436071062692,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.841338169963122e-06,
+      "loss": 1.0101,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6790435021613653,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.840663516075081e-06,
+      "loss": 0.9916,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6792433972164613,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.839988691585177e-06,
+      "loss": 1.0407,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6794432922715574,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.839313696523384e-06,
+      "loss": 1.0267,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6796431873266535,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.838638530919688e-06,
+      "loss": 1.0569,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6798430823817496,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.837963194804077e-06,
+      "loss": 1.0449,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6800429774368456,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.837287688206552e-06,
+      "loss": 1.0707,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6802428724919417,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.836612011157117e-06,
+      "loss": 1.0324,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6804427675470378,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.835936163685786e-06,
+      "loss": 1.0641,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6806426626021339,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.835260145822582e-06,
+      "loss": 1.0306,
+      "step": 3405
+    },
+    {
+      "epoch": 0.68084255765723,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.83458395759753e-06,
+      "loss": 1.1774,
+      "step": 3406
+    },
+    {
+      "epoch": 0.681042452712326,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.833907599040668e-06,
+      "loss": 1.0588,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6812423477674221,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.833231070182042e-06,
+      "loss": 1.1366,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6814422428225182,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.8325543710517e-06,
+      "loss": 1.0346,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6816421378776143,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.831877501679701e-06,
+      "loss": 1.1177,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6818420329327103,
+      "grad_norm": 2.125,
+      "learning_rate": 8.831200462096115e-06,
+      "loss": 1.0492,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6820419279878064,
+      "grad_norm": 2.0,
+      "learning_rate": 8.83052325233101e-06,
+      "loss": 1.0085,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6822418230429025,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.829845872414477e-06,
+      "loss": 1.0229,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6824417180979986,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.829168322376595e-06,
+      "loss": 0.9829,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6826416131530946,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.828490602247466e-06,
+      "loss": 1.0217,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6828415082081907,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.827812712057195e-06,
+      "loss": 1.0444,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6830414032632868,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.827134651835889e-06,
+      "loss": 1.0498,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6832412983183829,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.826456421613674e-06,
+      "loss": 1.0302,
+      "step": 3418
+    },
+    {
+      "epoch": 0.683441193373479,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.82577802142067e-06,
+      "loss": 1.0086,
+      "step": 3419
+    },
+    {
+      "epoch": 0.683641088428575,
+      "grad_norm": 2.125,
+      "learning_rate": 8.825099451287018e-06,
+      "loss": 1.0617,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6838409834836711,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.824420711242855e-06,
+      "loss": 1.027,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6840408785387672,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.823741801318332e-06,
+      "loss": 1.1092,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6842407735938633,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.82306272154361e-06,
+      "loss": 1.0552,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6844406686489592,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.822383471948846e-06,
+      "loss": 0.9296,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6846405637040553,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.821704052564218e-06,
+      "loss": 1.0257,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6848404587591514,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.821024463419904e-06,
+      "loss": 0.9842,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6850403538142475,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.820344704546093e-06,
+      "loss": 1.0134,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6852402488693436,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.819664775972976e-06,
+      "loss": 1.1109,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6854401439244396,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.818984677730759e-06,
+      "loss": 1.0232,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6856400389795357,
+      "grad_norm": 2.0,
+      "learning_rate": 8.81830440984965e-06,
+      "loss": 1.0117,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6858399340346318,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.817623972359867e-06,
+      "loss": 1.0454,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6860398290897279,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.816943365291635e-06,
+      "loss": 1.0472,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6862397241448239,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.816262588675186e-06,
+      "loss": 1.1059,
+      "step": 3433
+    },
+    {
+      "epoch": 0.68643961919992,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.815581642540763e-06,
+      "loss": 0.9828,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6866395142550161,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.814900526918608e-06,
+      "loss": 1.0322,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6868394093101122,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.814219241838979e-06,
+      "loss": 1.0228,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6870393043652082,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.81353778733214e-06,
+      "loss": 1.0842,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6872391994203043,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.812856163428358e-06,
+      "loss": 1.0629,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6874390944754004,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.812174370157915e-06,
+      "loss": 1.0816,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6876389895304965,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.811492407551092e-06,
+      "loss": 1.0512,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6878388845855926,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.810810275638183e-06,
+      "loss": 1.0826,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6880387796406886,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.810127974449489e-06,
+      "loss": 1.0436,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6882386746957847,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.809445504015318e-06,
+      "loss": 0.9917,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6884385697508808,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.808762864365985e-06,
+      "loss": 1.106,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6886384648059769,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.80808005553181e-06,
+      "loss": 1.0241,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6888383598610729,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.807397077543127e-06,
+      "loss": 1.1441,
+      "step": 3446
+    },
+    {
+      "epoch": 0.689038254916169,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.806713930430273e-06,
+      "loss": 1.0712,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6892381499712651,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.806030614223592e-06,
+      "loss": 1.082,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6894380450263612,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.805347128953438e-06,
+      "loss": 1.0949,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6896379400814573,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.80466347465017e-06,
+      "loss": 0.9658,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6898378351365533,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.803979651344159e-06,
+      "loss": 0.9569,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6900377301916494,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.803295659065776e-06,
+      "loss": 1.0401,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6902376252467455,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.802611497845407e-06,
+      "loss": 1.0324,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6904375203018416,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.801927167713442e-06,
+      "loss": 1.0611,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6906374153569376,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.801242668700277e-06,
+      "loss": 1.005,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6908373104120337,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.800558000836318e-06,
+      "loss": 1.0715,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6910372054671298,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.799873164151981e-06,
+      "loss": 1.0608,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6912371005222259,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.799188158677683e-06,
+      "loss": 1.1111,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6914369955773219,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.79850298444385e-06,
+      "loss": 1.0775,
+      "step": 3459
+    },
+    {
+      "epoch": 0.691636890632418,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.797817641480923e-06,
+      "loss": 1.0977,
+      "step": 3460
+    },
+    {
+      "epoch": 0.691836785687514,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.797132129819343e-06,
+      "loss": 1.0691,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6920366807426102,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.796446449489557e-06,
+      "loss": 1.0476,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6922365757977063,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.795760600522025e-06,
+      "loss": 1.0863,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6924364708528022,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.795074582947214e-06,
+      "loss": 1.073,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6926363659078983,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.794388396795595e-06,
+      "loss": 1.0367,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6928362609629944,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.79370204209765e-06,
+      "loss": 0.9947,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6930361560180905,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.793015518883862e-06,
+      "loss": 1.1298,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6932360510731865,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.792328827184733e-06,
+      "loss": 1.0406,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6934359461282826,
+      "grad_norm": 2.25,
+      "learning_rate": 8.791641967030761e-06,
+      "loss": 1.0598,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6936358411833787,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.790954938452458e-06,
+      "loss": 1.0779,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6938357362384748,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.790267741480342e-06,
+      "loss": 1.0949,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6940356312935708,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.789580376144938e-06,
+      "loss": 1.0611,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6942355263486669,
+      "grad_norm": 2.125,
+      "learning_rate": 8.788892842476777e-06,
+      "loss": 1.1123,
+      "step": 3473
+    },
+    {
+      "epoch": 0.694435421403763,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.7882051405064e-06,
+      "loss": 1.0101,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6946353164588591,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.78751727026436e-06,
+      "loss": 1.0699,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6948352115139552,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.786829231781203e-06,
+      "loss": 0.986,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6950351065690512,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.786141025087496e-06,
+      "loss": 1.0022,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6952350016241473,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.78545265021381e-06,
+      "loss": 1.0105,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6954348966792434,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.784764107190723e-06,
+      "loss": 1.059,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6956347917343395,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.784075396048814e-06,
+      "loss": 0.9632,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6958346867894355,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.783386516818684e-06,
+      "loss": 1.001,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6960345818445316,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.782697469530929e-06,
+      "loss": 1.0463,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6962344768996277,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.782008254216155e-06,
+      "loss": 1.0061,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6964343719547238,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.78131887090498e-06,
+      "loss": 1.1041,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6966342670098199,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.780629319628023e-06,
+      "loss": 0.9527,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6968341620649159,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.779939600415917e-06,
+      "loss": 1.0341,
+      "step": 3486
+    },
+    {
+      "epoch": 0.697034057120012,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.779249713299296e-06,
+      "loss": 0.8984,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6972339521751081,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.778559658308806e-06,
+      "loss": 1.0076,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6974338472302042,
+      "grad_norm": 2.125,
+      "learning_rate": 8.777869435475101e-06,
+      "loss": 1.0677,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6976337422853002,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.777179044828838e-06,
+      "loss": 1.0401,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6978336373403963,
+      "grad_norm": 2.0,
+      "learning_rate": 8.776488486400688e-06,
+      "loss": 0.942,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6980335323954924,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.775797760221318e-06,
+      "loss": 1.1208,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6982334274505885,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.775106866321419e-06,
+      "loss": 1.0702,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6984333225056845,
+      "grad_norm": 2.25,
+      "learning_rate": 8.774415804731674e-06,
+      "loss": 1.153,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6986332175607806,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.773724575482783e-06,
+      "loss": 1.0306,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6988331126158767,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.77303317860545e-06,
+      "loss": 0.9907,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6990330076709728,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.772341614130384e-06,
+      "loss": 1.0426,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6992329027260689,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.771649882088309e-06,
+      "loss": 1.0738,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6994327977811649,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.770957982509947e-06,
+      "loss": 1.0714,
+      "step": 3499
+    },
+    {
+      "epoch": 0.699632692836261,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.770265915426035e-06,
+      "loss": 1.0174,
+      "step": 3500
+    },
+    {
+      "epoch": 0.699832587891357,
+      "grad_norm": 2.125,
+      "learning_rate": 8.769573680867314e-06,
+      "loss": 0.9708,
+      "step": 3501
+    },
+    {
+      "epoch": 0.7000324829464531,
+      "grad_norm": 2.0,
+      "learning_rate": 8.768881278864532e-06,
+      "loss": 1.025,
+      "step": 3502
+    },
+    {
+      "epoch": 0.7002323780015491,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.768188709448446e-06,
+      "loss": 1.0102,
+      "step": 3503
+    },
+    {
+      "epoch": 0.7004322730566452,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.76749597264982e-06,
+      "loss": 1.0685,
+      "step": 3504
+    },
+    {
+      "epoch": 0.7006321681117413,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.766803068499426e-06,
+      "loss": 1.0231,
+      "step": 3505
+    },
+    {
+      "epoch": 0.7008320631668374,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.766109997028042e-06,
+      "loss": 1.0792,
+      "step": 3506
+    },
+    {
+      "epoch": 0.7010319582219335,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.765416758266454e-06,
+      "loss": 1.0519,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7012318532770295,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.764723352245455e-06,
+      "loss": 1.057,
+      "step": 3508
+    },
+    {
+      "epoch": 0.7014317483321256,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.764029778995848e-06,
+      "loss": 0.9859,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7016316433872217,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.76333603854844e-06,
+      "loss": 1.0624,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7018315384423178,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.762642130934048e-06,
+      "loss": 1.0676,
+      "step": 3511
+    },
+    {
+      "epoch": 0.7020314334974138,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.761948056183492e-06,
+      "loss": 1.0841,
+      "step": 3512
+    },
+    {
+      "epoch": 0.7022313285525099,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.761253814327606e-06,
+      "loss": 1.0859,
+      "step": 3513
+    },
+    {
+      "epoch": 0.702431223607606,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.760559405397228e-06,
+      "loss": 1.0258,
+      "step": 3514
+    },
+    {
+      "epoch": 0.7026311186627021,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.759864829423202e-06,
+      "loss": 0.9817,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7028310137177981,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.759170086436382e-06,
+      "loss": 1.0133,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7030309087728942,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.75847517646763e-06,
+      "loss": 1.1056,
+      "step": 3517
+    },
+    {
+      "epoch": 0.7032308038279903,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.75778009954781e-06,
+      "loss": 1.0127,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7034306988830864,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.757084855707799e-06,
+      "loss": 1.1148,
+      "step": 3519
+    },
+    {
+      "epoch": 0.7036305939381825,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.756389444978482e-06,
+      "loss": 1.0803,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7038304889932785,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.755693867390746e-06,
+      "loss": 1.0793,
+      "step": 3521
+    },
+    {
+      "epoch": 0.7040303840483746,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.754998122975489e-06,
+      "loss": 1.1372,
+      "step": 3522
+    },
+    {
+      "epoch": 0.7042302791034707,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.754302211763616e-06,
+      "loss": 1.0156,
+      "step": 3523
+    },
+    {
+      "epoch": 0.7044301741585668,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.753606133786042e-06,
+      "loss": 0.9808,
+      "step": 3524
+    },
+    {
+      "epoch": 0.7046300692136628,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.752909889073681e-06,
+      "loss": 1.0435,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7048299642687589,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.752213477657467e-06,
+      "loss": 1.017,
+      "step": 3526
+    },
+    {
+      "epoch": 0.705029859323855,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.751516899568329e-06,
+      "loss": 1.0034,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7052297543789511,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.750820154837213e-06,
+      "loss": 1.0923,
+      "step": 3528
+    },
+    {
+      "epoch": 0.7054296494340472,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.750123243495066e-06,
+      "loss": 1.1833,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7056295444891432,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.749426165572843e-06,
+      "loss": 1.0742,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7058294395442393,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.748728921101511e-06,
+      "loss": 0.9526,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7060293345993354,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.748031510112041e-06,
+      "loss": 1.0627,
+      "step": 3532
+    },
+    {
+      "epoch": 0.7062292296544315,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.747333932635412e-06,
+      "loss": 0.9998,
+      "step": 3533
+    },
+    {
+      "epoch": 0.7064291247095275,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.746636188702609e-06,
+      "loss": 0.9563,
+      "step": 3534
+    },
+    {
+      "epoch": 0.7066290197646236,
+      "grad_norm": 2.125,
+      "learning_rate": 8.745938278344628e-06,
+      "loss": 1.1783,
+      "step": 3535
+    },
+    {
+      "epoch": 0.7068289148197197,
+      "grad_norm": 2.0,
+      "learning_rate": 8.745240201592466e-06,
+      "loss": 1.0853,
+      "step": 3536
+    },
+    {
+      "epoch": 0.7070288098748158,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.744541958477138e-06,
+      "loss": 1.1799,
+      "step": 3537
+    },
+    {
+      "epoch": 0.7072287049299117,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.743843549029653e-06,
+      "loss": 0.994,
+      "step": 3538
+    },
+    {
+      "epoch": 0.7074285999850078,
+      "grad_norm": 2.125,
+      "learning_rate": 8.74314497328104e-06,
+      "loss": 1.1054,
+      "step": 3539
+    },
+    {
+      "epoch": 0.7076284950401039,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.742446231262324e-06,
+      "loss": 1.0607,
+      "step": 3540
+    },
+    {
+      "epoch": 0.7078283900952,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.741747323004549e-06,
+      "loss": 1.0312,
+      "step": 3541
+    },
+    {
+      "epoch": 0.7080282851502961,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.741048248538757e-06,
+      "loss": 1.0552,
+      "step": 3542
+    },
+    {
+      "epoch": 0.7082281802053921,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.740349007896001e-06,
+      "loss": 1.0946,
+      "step": 3543
+    },
+    {
+      "epoch": 0.7084280752604882,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.73964960110734e-06,
+      "loss": 1.029,
+      "step": 3544
+    },
+    {
+      "epoch": 0.7086279703155843,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.738950028203845e-06,
+      "loss": 1.1373,
+      "step": 3545
+    },
+    {
+      "epoch": 0.7088278653706804,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.738250289216588e-06,
+      "loss": 1.1068,
+      "step": 3546
+    },
+    {
+      "epoch": 0.7090277604257764,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.737550384176654e-06,
+      "loss": 0.9384,
+      "step": 3547
+    },
+    {
+      "epoch": 0.7092276554808725,
+      "grad_norm": 2.125,
+      "learning_rate": 8.73685031311513e-06,
+      "loss": 1.1025,
+      "step": 3548
+    },
+    {
+      "epoch": 0.7094275505359686,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.736150076063114e-06,
+      "loss": 1.1502,
+      "step": 3549
+    },
+    {
+      "epoch": 0.7096274455910647,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.735449673051711e-06,
+      "loss": 1.0809,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7098273406461608,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.734749104112032e-06,
+      "loss": 1.0088,
+      "step": 3551
+    },
+    {
+      "epoch": 0.7100272357012568,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.734048369275199e-06,
+      "loss": 1.1559,
+      "step": 3552
+    },
+    {
+      "epoch": 0.7102271307563529,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.733347468572333e-06,
+      "loss": 0.9297,
+      "step": 3553
+    },
+    {
+      "epoch": 0.710427025811449,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.732646402034572e-06,
+      "loss": 0.9721,
+      "step": 3554
+    },
+    {
+      "epoch": 0.7106269208665451,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.731945169693058e-06,
+      "loss": 1.0645,
+      "step": 3555
+    },
+    {
+      "epoch": 0.7108268159216411,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.731243771578937e-06,
+      "loss": 1.0229,
+      "step": 3556
+    },
+    {
+      "epoch": 0.7110267109767372,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.730542207723367e-06,
+      "loss": 1.0102,
+      "step": 3557
+    },
+    {
+      "epoch": 0.7112266060318333,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.72984047815751e-06,
+      "loss": 1.0751,
+      "step": 3558
+    },
+    {
+      "epoch": 0.7114265010869294,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.729138582912538e-06,
+      "loss": 1.0856,
+      "step": 3559
+    },
+    {
+      "epoch": 0.7116263961420254,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.728436522019627e-06,
+      "loss": 1.0536,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7118262911971215,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.727734295509964e-06,
+      "loss": 1.0875,
+      "step": 3561
+    },
+    {
+      "epoch": 0.7120261862522176,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.727031903414743e-06,
+      "loss": 1.0707,
+      "step": 3562
+    },
+    {
+      "epoch": 0.7122260813073137,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.72632934576516e-06,
+      "loss": 1.0657,
+      "step": 3563
+    },
+    {
+      "epoch": 0.7124259763624098,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.72562662259243e-06,
+      "loss": 1.0626,
+      "step": 3564
+    },
+    {
+      "epoch": 0.7126258714175058,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.72492373392776e-06,
+      "loss": 0.9411,
+      "step": 3565
+    },
+    {
+      "epoch": 0.7128257664726019,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.724220679802377e-06,
+      "loss": 1.157,
+      "step": 3566
+    },
+    {
+      "epoch": 0.713025661527698,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.723517460247509e-06,
+      "loss": 1.0013,
+      "step": 3567
+    },
+    {
+      "epoch": 0.7132255565827941,
+      "grad_norm": 2.0,
+      "learning_rate": 8.722814075294392e-06,
+      "loss": 1.0098,
+      "step": 3568
+    },
+    {
+      "epoch": 0.7134254516378901,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.722110524974273e-06,
+      "loss": 1.0333,
+      "step": 3569
+    },
+    {
+      "epoch": 0.7136253466929862,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.7214068093184e-06,
+      "loss": 1.0358,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7138252417480823,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.720702928358036e-06,
+      "loss": 1.0748,
+      "step": 3571
+    },
+    {
+      "epoch": 0.7140251368031784,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.719998882124446e-06,
+      "loss": 1.0685,
+      "step": 3572
+    },
+    {
+      "epoch": 0.7142250318582744,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.7192946706489e-06,
+      "loss": 1.0813,
+      "step": 3573
+    },
+    {
+      "epoch": 0.7144249269133705,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.718590293962684e-06,
+      "loss": 1.0034,
+      "step": 3574
+    },
+    {
+      "epoch": 0.7146248219684666,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.717885752097084e-06,
+      "loss": 1.0424,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7148247170235627,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.717181045083396e-06,
+      "loss": 1.0401,
+      "step": 3576
+    },
+    {
+      "epoch": 0.7150246120786588,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.716476172952921e-06,
+      "loss": 1.0595,
+      "step": 3577
+    },
+    {
+      "epoch": 0.7152245071337547,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.715771135736975e-06,
+      "loss": 1.0123,
+      "step": 3578
+    },
+    {
+      "epoch": 0.7154244021888508,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.715065933466869e-06,
+      "loss": 1.1616,
+      "step": 3579
+    },
+    {
+      "epoch": 0.7156242972439469,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.714360566173932e-06,
+      "loss": 1.057,
+      "step": 3580
+    },
+    {
+      "epoch": 0.715824192299043,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.713655033889495e-06,
+      "loss": 1.0771,
+      "step": 3581
+    },
+    {
+      "epoch": 0.716024087354139,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.712949336644898e-06,
+      "loss": 1.0603,
+      "step": 3582
+    },
+    {
+      "epoch": 0.7162239824092351,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.71224347447149e-06,
+      "loss": 1.0811,
+      "step": 3583
+    },
+    {
+      "epoch": 0.7164238774643312,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.711537447400622e-06,
+      "loss": 1.0414,
+      "step": 3584
+    },
+    {
+      "epoch": 0.7166237725194273,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.710831255463656e-06,
+      "loss": 1.0759,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7168236675745234,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.710124898691963e-06,
+      "loss": 1.0057,
+      "step": 3586
+    },
+    {
+      "epoch": 0.7170235626296194,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.709418377116918e-06,
+      "loss": 1.0743,
+      "step": 3587
+    },
+    {
+      "epoch": 0.7172234576847155,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.708711690769904e-06,
+      "loss": 1.1095,
+      "step": 3588
+    },
+    {
+      "epoch": 0.7174233527398116,
+      "grad_norm": 2.125,
+      "learning_rate": 8.708004839682315e-06,
+      "loss": 1.1079,
+      "step": 3589
+    },
+    {
+      "epoch": 0.7176232477949077,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.707297823885545e-06,
+      "loss": 1.0356,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7178231428500037,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.706590643411002e-06,
+      "loss": 1.057,
+      "step": 3591
+    },
+    {
+      "epoch": 0.7180230379050998,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.7058832982901e-06,
+      "loss": 1.0344,
+      "step": 3592
+    },
+    {
+      "epoch": 0.7182229329601959,
+      "grad_norm": 2.125,
+      "learning_rate": 8.705175788554256e-06,
+      "loss": 0.9713,
+      "step": 3593
+    },
+    {
+      "epoch": 0.718422828015292,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.7044681142349e-06,
+      "loss": 1.0609,
+      "step": 3594
+    },
+    {
+      "epoch": 0.718622723070388,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.703760275363466e-06,
+      "loss": 1.0574,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7188226181254841,
+      "grad_norm": 2.0,
+      "learning_rate": 8.703052271971395e-06,
+      "loss": 1.0209,
+      "step": 3596
+    },
+    {
+      "epoch": 0.7190225131805802,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.702344104090139e-06,
+      "loss": 0.9954,
+      "step": 3597
+    },
+    {
+      "epoch": 0.7192224082356763,
+      "grad_norm": 2.125,
+      "learning_rate": 8.701635771751153e-06,
+      "loss": 0.9903,
+      "step": 3598
+    },
+    {
+      "epoch": 0.7194223032907724,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.700927274985903e-06,
+      "loss": 1.092,
+      "step": 3599
+    },
+    {
+      "epoch": 0.7196221983458684,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.700218613825855e-06,
+      "loss": 1.0449,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7198220934009645,
+      "grad_norm": 2.0,
+      "learning_rate": 8.699509788302493e-06,
+      "loss": 1.0659,
+      "step": 3601
+    },
+    {
+      "epoch": 0.7200219884560606,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.698800798447302e-06,
+      "loss": 0.9256,
+      "step": 3602
+    },
+    {
+      "epoch": 0.7202218835111567,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.698091644291774e-06,
+      "loss": 1.0132,
+      "step": 3603
+    },
+    {
+      "epoch": 0.7204217785662527,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.69738232586741e-06,
+      "loss": 1.0542,
+      "step": 3604
+    },
+    {
+      "epoch": 0.7206216736213488,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.696672843205718e-06,
+      "loss": 1.046,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7208215686764449,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.695963196338214e-06,
+      "loss": 0.9889,
+      "step": 3606
+    },
+    {
+      "epoch": 0.721021463731541,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.69525338529642e-06,
+      "loss": 1.0458,
+      "step": 3607
+    },
+    {
+      "epoch": 0.7212213587866371,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.694543410111864e-06,
+      "loss": 1.0359,
+      "step": 3608
+    },
+    {
+      "epoch": 0.7214212538417331,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.693833270816083e-06,
+      "loss": 1.0612,
+      "step": 3609
+    },
+    {
+      "epoch": 0.7216211488968292,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.693122967440626e-06,
+      "loss": 1.0821,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7218210439519253,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.69241250001704e-06,
+      "loss": 1.0275,
+      "step": 3611
+    },
+    {
+      "epoch": 0.7220209390070214,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.691701868576883e-06,
+      "loss": 1.0511,
+      "step": 3612
+    },
+    {
+      "epoch": 0.7222208340621173,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.690991073151724e-06,
+      "loss": 0.9708,
+      "step": 3613
+    },
+    {
+      "epoch": 0.7224207291172134,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.690280113773138e-06,
+      "loss": 0.9171,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7226206241723095,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.689568990472701e-06,
+      "loss": 1.1249,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7228205192274056,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.688857703282005e-06,
+      "loss": 1.0374,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7230204142825016,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.688146252232644e-06,
+      "loss": 0.9733,
+      "step": 3617
+    },
+    {
+      "epoch": 0.7232203093375977,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.68743463735622e-06,
+      "loss": 1.0975,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7234202043926938,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.686722858684342e-06,
+      "loss": 1.0447,
+      "step": 3619
+    },
+    {
+      "epoch": 0.7236200994477899,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.68601091624863e-06,
+      "loss": 1.1732,
+      "step": 3620
+    },
+    {
+      "epoch": 0.723819994502886,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.685298810080706e-06,
+      "loss": 1.0777,
+      "step": 3621
+    },
+    {
+      "epoch": 0.724019889557982,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.684586540212203e-06,
+      "loss": 1.1034,
+      "step": 3622
+    },
+    {
+      "epoch": 0.7242197846130781,
+      "grad_norm": 2.0,
+      "learning_rate": 8.683874106674759e-06,
+      "loss": 1.0943,
+      "step": 3623
+    },
+    {
+      "epoch": 0.7244196796681742,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.68316150950002e-06,
+      "loss": 1.0627,
+      "step": 3624
+    },
+    {
+      "epoch": 0.7246195747232703,
+      "grad_norm": 2.0,
+      "learning_rate": 8.68244874871964e-06,
+      "loss": 0.9761,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7248194697783663,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.681735824365281e-06,
+      "loss": 0.9976,
+      "step": 3626
+    },
+    {
+      "epoch": 0.7250193648334624,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.681022736468609e-06,
+      "loss": 0.9816,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7252192598885585,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.680309485061302e-06,
+      "loss": 1.0761,
+      "step": 3628
+    },
+    {
+      "epoch": 0.7254191549436546,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.679596070175038e-06,
+      "loss": 1.0507,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7256190499987507,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.678882491841512e-06,
+      "loss": 1.0754,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7258189450538467,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.678168750092419e-06,
+      "loss": 1.0151,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7260188401089428,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.67745484495946e-06,
+      "loss": 1.0596,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7262187351640389,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.676740776474351e-06,
+      "loss": 1.0023,
+      "step": 3633
+    },
+    {
+      "epoch": 0.726418630219135,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.67602654466881e-06,
+      "loss": 1.0468,
+      "step": 3634
+    },
+    {
+      "epoch": 0.726618525274231,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.675312149574562e-06,
+      "loss": 1.0281,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7268184203293271,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.67459759122334e-06,
+      "loss": 1.0334,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7270183153844232,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.673882869646888e-06,
+      "loss": 1.1007,
+      "step": 3637
+    },
+    {
+      "epoch": 0.7272182104395193,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.67316798487695e-06,
+      "loss": 1.1333,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7274181054946153,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.672452936945282e-06,
+      "loss": 1.0483,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7276180005497114,
+      "grad_norm": 2.125,
+      "learning_rate": 8.671737725883646e-06,
+      "loss": 1.0026,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7278178956048075,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.671022351723813e-06,
+      "loss": 0.9676,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7280177906599036,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.67030681449756e-06,
+      "loss": 1.0058,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7282176857149997,
+      "grad_norm": 2.25,
+      "learning_rate": 8.66959111423667e-06,
+      "loss": 1.0806,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7284175807700957,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.668875250972934e-06,
+      "loss": 1.0737,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7286174758251918,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.668159224738153e-06,
+      "loss": 1.0112,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7288173708802879,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.667443035564129e-06,
+      "loss": 1.073,
+      "step": 3646
+    },
+    {
+      "epoch": 0.729017265935384,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.666726683482678e-06,
+      "loss": 1.1717,
+      "step": 3647
+    },
+    {
+      "epoch": 0.72921716099048,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.666010168525618e-06,
+      "loss": 1.072,
+      "step": 3648
+    },
+    {
+      "epoch": 0.729417056045576,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.66529349072478e-06,
+      "loss": 1.0895,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7296169511006722,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.664576650111995e-06,
+      "loss": 1.0967,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7298168461557683,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.663859646719106e-06,
+      "loss": 1.024,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7300167412108644,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.663142480577965e-06,
+      "loss": 0.999,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7302166362659603,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.662425151720425e-06,
+      "loss": 1.0337,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7304165313210564,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.661707660178351e-06,
+      "loss": 1.0501,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7306164263761525,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.660990005983613e-06,
+      "loss": 1.0929,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7308163214312486,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.660272189168093e-06,
+      "loss": 1.002,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7310162164863446,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.659554209763669e-06,
+      "loss": 1.0552,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7312161115414407,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.65883606780224e-06,
+      "loss": 1.031,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7314160065965368,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.658117763315705e-06,
+      "loss": 0.9722,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7316159016516329,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.657399296335967e-06,
+      "loss": 1.0115,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7318157967067289,
+      "grad_norm": 2.125,
+      "learning_rate": 8.656680666894945e-06,
+      "loss": 0.9965,
+      "step": 3661
+    },
+    {
+      "epoch": 0.732015691761825,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.655961875024557e-06,
+      "loss": 1.065,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7322155868169211,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.655242920756733e-06,
+      "loss": 1.1328,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7324154818720172,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.65452380412341e-06,
+      "loss": 1.085,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7326153769271133,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.653804525156529e-06,
+      "loss": 1.0886,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7328152719822093,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.653085083888042e-06,
+      "loss": 1.0906,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7330151670373054,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.652365480349904e-06,
+      "loss": 1.0874,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7332150620924015,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.651645714574082e-06,
+      "loss": 0.971,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7334149571474976,
+      "grad_norm": 1.875,
+      "learning_rate": 8.65092578659255e-06,
+      "loss": 1.0021,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7336148522025936,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.650205696437282e-06,
+      "loss": 0.9673,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7338147472576897,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.649485444140267e-06,
+      "loss": 1.0164,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7340146423127858,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.6487650297335e-06,
+      "loss": 1.0038,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7342145373678819,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.648044453248978e-06,
+      "loss": 0.9634,
+      "step": 3673
+    },
+    {
+      "epoch": 0.734414432422978,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.647323714718712e-06,
+      "loss": 1.1127,
+      "step": 3674
+    },
+    {
+      "epoch": 0.734614327478074,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.646602814174715e-06,
+      "loss": 0.8807,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7348142225331701,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.645881751649012e-06,
+      "loss": 1.0312,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7350141175882662,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.64516052717363e-06,
+      "loss": 1.0603,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7352140126433623,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.644439140780608e-06,
+      "loss": 1.1072,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7354139076984583,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.643717592501988e-06,
+      "loss": 1.0585,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7356138027535544,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.64299588236982e-06,
+      "loss": 1.0328,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7358136978086505,
+      "grad_norm": 2.125,
+      "learning_rate": 8.642274010416165e-06,
+      "loss": 1.0931,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7360135928637466,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.641551976673088e-06,
+      "loss": 1.0921,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7362134879188426,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.64082978117266e-06,
+      "loss": 1.0605,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7364133829739387,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.640107423946964e-06,
+      "loss": 0.9517,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7366132780290348,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.639384905028084e-06,
+      "loss": 1.0521,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7368131730841309,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.638662224448115e-06,
+      "loss": 0.9472,
+      "step": 3686
+    },
+    {
+      "epoch": 0.737013068139227,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.63793938223916e-06,
+      "loss": 1.0221,
+      "step": 3687
+    },
+    {
+      "epoch": 0.737212963194323,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.637216378433324e-06,
+      "loss": 1.121,
+      "step": 3688
+    },
+    {
+      "epoch": 0.737412858249419,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.636493213062725e-06,
+      "loss": 1.0409,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7376127533045151,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.635769886159488e-06,
+      "loss": 1.1295,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7378126483596112,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.63504639775574e-06,
+      "loss": 1.0737,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7380125434147072,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.634322747883619e-06,
+      "loss": 1.025,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7382124384698033,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.63359893657527e-06,
+      "loss": 1.011,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7384123335248994,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.632874963862844e-06,
+      "loss": 1.0802,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7386122285799955,
+      "grad_norm": 2.125,
+      "learning_rate": 8.632150829778498e-06,
+      "loss": 1.0525,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7388121236350915,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.631426534354404e-06,
+      "loss": 0.9998,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7390120186901876,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.630702077622728e-06,
+      "loss": 1.0091,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7392119137452837,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.629977459615655e-06,
+      "loss": 1.0374,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7394118088003798,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.62925268036537e-06,
+      "loss": 0.9763,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7396117038554759,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.62852773990407e-06,
+      "loss": 1.0981,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7398115989105719,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.627802638263955e-06,
+      "loss": 0.9995,
+      "step": 3701
+    },
+    {
+      "epoch": 0.740011493965668,
+      "grad_norm": 2.0,
+      "learning_rate": 8.627077375477233e-06,
+      "loss": 1.0346,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7402113890207641,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.626351951576122e-06,
+      "loss": 1.0345,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7404112840758602,
+      "grad_norm": 2.125,
+      "learning_rate": 8.625626366592844e-06,
+      "loss": 0.99,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7406111791309562,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.624900620559633e-06,
+      "loss": 1.1051,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7408110741860523,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.624174713508722e-06,
+      "loss": 1.1394,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7410109692411484,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.623448645472356e-06,
+      "loss": 1.0432,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7412108642962445,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.62272241648279e-06,
+      "loss": 0.9488,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7414107593513406,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.62199602657228e-06,
+      "loss": 1.0239,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7416106544064366,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.621269475773092e-06,
+      "loss": 1.1023,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7418105494615327,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.620542764117503e-06,
+      "loss": 0.9405,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7420104445166288,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.61981589163779e-06,
+      "loss": 1.129,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7422103395717249,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.619088858366242e-06,
+      "loss": 0.9819,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7424102346268209,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.618361664335153e-06,
+      "loss": 1.0968,
+      "step": 3714
+    },
+    {
+      "epoch": 0.742610129681917,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.617634309576827e-06,
+      "loss": 1.0922,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7428100247370131,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.61690679412357e-06,
+      "loss": 1.0316,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7430099197921092,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.6161791180077e-06,
+      "loss": 1.0087,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7432098148472052,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.615451281261539e-06,
+      "loss": 0.9273,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7434097099023013,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.614723283917418e-06,
+      "loss": 0.9929,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7436096049573974,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.613995126007674e-06,
+      "loss": 0.9644,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7438095000124935,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.613266807564656e-06,
+      "loss": 1.0416,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7440093950675896,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.61253832862071e-06,
+      "loss": 1.0855,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7442092901226856,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.611809689208197e-06,
+      "loss": 1.073,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7444091851777817,
+      "grad_norm": 2.0,
+      "learning_rate": 8.611080889359485e-06,
+      "loss": 1.0305,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7446090802328778,
+      "grad_norm": 2.125,
+      "learning_rate": 8.610351929106944e-06,
+      "loss": 1.0055,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7448089752879739,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.609622808482956e-06,
+      "loss": 1.0343,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7450088703430698,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.608893527519908e-06,
+      "loss": 1.0539,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7452087653981659,
+      "grad_norm": 2.125,
+      "learning_rate": 8.608164086250197e-06,
+      "loss": 1.0643,
+      "step": 3728
+    },
+    {
+      "epoch": 0.745408660453262,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.607434484706221e-06,
+      "loss": 1.0237,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7456085555083581,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.60670472292039e-06,
+      "loss": 1.0754,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7458084505634542,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.605974800925121e-06,
+      "loss": 1.0677,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7460083456185502,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.605244718752837e-06,
+      "loss": 0.9748,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7462082406736463,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.604514476435969e-06,
+      "loss": 0.9695,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7464081357287424,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.60378407400695e-06,
+      "loss": 0.9913,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7466080307838385,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.603053511498228e-06,
+      "loss": 1.0384,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7468079258389345,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.602322788942255e-06,
+      "loss": 0.9894,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7470078208940306,
+      "grad_norm": 2.125,
+      "learning_rate": 8.601591906371487e-06,
+      "loss": 1.0653,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7472077159491267,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.600860863818392e-06,
+      "loss": 1.0099,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7474076110042228,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.600129661315443e-06,
+      "loss": 1.1134,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7476075060593188,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.599398298895117e-06,
+      "loss": 1.0463,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7478074011144149,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.598666776589904e-06,
+      "loss": 1.0142,
+      "step": 3741
+    },
+    {
+      "epoch": 0.748007296169511,
+      "grad_norm": 2.0,
+      "learning_rate": 8.597935094432298e-06,
+      "loss": 1.0762,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7482071912246071,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.597203252454798e-06,
+      "loss": 1.1143,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7484070862797032,
+      "grad_norm": 2.125,
+      "learning_rate": 8.596471250689915e-06,
+      "loss": 1.078,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7486069813347992,
+      "grad_norm": 2.25,
+      "learning_rate": 8.595739089170162e-06,
+      "loss": 1.0361,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7488068763898953,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.595006767928064e-06,
+      "loss": 1.0813,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7490067714449914,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.594274286996147e-06,
+      "loss": 1.1081,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7492066665000875,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.593541646406952e-06,
+      "loss": 1.0592,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7494065615551835,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.59280884619302e-06,
+      "loss": 0.9439,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7496064566102796,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.592075886386903e-06,
+      "loss": 1.0597,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7498063516653757,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.59134276702116e-06,
+      "loss": 0.9952,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7500062467204718,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.590609488128354e-06,
+      "loss": 0.9731,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7502061417755679,
+      "grad_norm": 2.125,
+      "learning_rate": 8.589876049741058e-06,
+      "loss": 0.9752,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7504060368306639,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.589142451891849e-06,
+      "loss": 1.0571,
+      "step": 3754
+    },
+    {
+      "epoch": 0.75060593188576,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.58840869461332e-06,
+      "loss": 1.0112,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7508058269408561,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.587674777938057e-06,
+      "loss": 0.9903,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7510057219959522,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.586940701898665e-06,
+      "loss": 1.1016,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7512056170510482,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.586206466527749e-06,
+      "loss": 0.9849,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7514055121061443,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.585472071857924e-06,
+      "loss": 1.0105,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7516054071612404,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.584737517921815e-06,
+      "loss": 0.9755,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7518053022163365,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.584002804752046e-06,
+      "loss": 1.0461,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7520051972714324,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.583267932381257e-06,
+      "loss": 1.1102,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7522050923265285,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.582532900842088e-06,
+      "loss": 1.0866,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7524049873816246,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.58179771016719e-06,
+      "loss": 1.0478,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7526048824367207,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.581062360389222e-06,
+      "loss": 0.9944,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7528047774918168,
+      "grad_norm": 2.0,
+      "learning_rate": 8.580326851540844e-06,
+      "loss": 1.1228,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7530046725469128,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.579591183654731e-06,
+      "loss": 1.0448,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7532045676020089,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.578855356763559e-06,
+      "loss": 1.1004,
+      "step": 3768
+    },
+    {
+      "epoch": 0.753404462657105,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.578119370900016e-06,
+      "loss": 1.0422,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7536043577122011,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.577383226096792e-06,
+      "loss": 0.9927,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7538042527672971,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.576646922386587e-06,
+      "loss": 1.0948,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7540041478223932,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.575910459802107e-06,
+      "loss": 1.0204,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7542040428774893,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.575173838376067e-06,
+      "loss": 1.0898,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7544039379325854,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.574437058141187e-06,
+      "loss": 0.9686,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7546038329876815,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.573700119130194e-06,
+      "loss": 1.0663,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7548037280427775,
+      "grad_norm": 2.125,
+      "learning_rate": 8.572963021375825e-06,
+      "loss": 0.9369,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7550036230978736,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.572225764910818e-06,
+      "loss": 1.0403,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7552035181529697,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.571488349767925e-06,
+      "loss": 1.069,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7554034132080658,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.570750775979901e-06,
+      "loss": 0.9747,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7556033082631618,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.57001304357951e-06,
+      "loss": 1.1118,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7558032033182579,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.56927515259952e-06,
+      "loss": 1.0418,
+      "step": 3781
+    },
+    {
+      "epoch": 0.756003098373354,
+      "grad_norm": 2.25,
+      "learning_rate": 8.568537103072707e-06,
+      "loss": 1.1046,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7562029934284501,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.56779889503186e-06,
+      "loss": 1.0775,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7564028884835461,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.567060528509767e-06,
+      "loss": 1.1559,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7566027835386422,
+      "grad_norm": 2.125,
+      "learning_rate": 8.566322003539225e-06,
+      "loss": 1.0536,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7568026785937383,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.56558332015304e-06,
+      "loss": 1.0553,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7570025736488344,
+      "grad_norm": 2.0,
+      "learning_rate": 8.564844478384027e-06,
+      "loss": 1.0333,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7572024687039305,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.564105478265e-06,
+      "loss": 0.9435,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7574023637590265,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.563366319828789e-06,
+      "loss": 1.1762,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7576022588141226,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.562627003108229e-06,
+      "loss": 1.0673,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7578021538692187,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.561887528136157e-06,
+      "loss": 0.942,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7580020489243148,
+      "grad_norm": 2.125,
+      "learning_rate": 8.56114789494542e-06,
+      "loss": 1.0512,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7582019439794108,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.560408103568875e-06,
+      "loss": 1.0053,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7584018390345069,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.559668154039382e-06,
+      "loss": 1.0991,
+      "step": 3794
+    },
+    {
+      "epoch": 0.758601734089603,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.55892804638981e-06,
+      "loss": 0.9974,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7588016291446991,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.558187780653033e-06,
+      "loss": 0.9253,
+      "step": 3796
+    },
+    {
+      "epoch": 0.759001524199795,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.557447356861937e-06,
+      "loss": 1.0197,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7592014192548912,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.556706775049408e-06,
+      "loss": 1.0748,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7594013143099873,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.555966035248344e-06,
+      "loss": 1.0473,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7596012093650834,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.555225137491649e-06,
+      "loss": 1.1592,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7598011044201795,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.554484081812233e-06,
+      "loss": 1.1061,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7600009994752754,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.553742868243014e-06,
+      "loss": 1.1416,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7602008945303715,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.553001496816915e-06,
+      "loss": 1.0807,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7604007895854676,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.552259967566871e-06,
+      "loss": 1.1007,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7606006846405637,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.551518280525816e-06,
+      "loss": 0.9757,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7608005796956597,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.550776435726701e-06,
+      "loss": 1.0683,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7610004747507558,
+      "grad_norm": 2.125,
+      "learning_rate": 8.550034433202474e-06,
+      "loss": 1.1383,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7612003698058519,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.549292272986098e-06,
+      "loss": 1.1601,
+      "step": 3808
+    },
+    {
+      "epoch": 0.761400264860948,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.548549955110538e-06,
+      "loss": 0.9979,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7616001599160441,
+      "grad_norm": 2.125,
+      "learning_rate": 8.547807479608768e-06,
+      "loss": 1.0703,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7618000549711401,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.547064846513768e-06,
+      "loss": 1.0299,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7619999500262362,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.546322055858526e-06,
+      "loss": 0.9472,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7621998450813323,
+      "grad_norm": 2.125,
+      "learning_rate": 8.545579107676039e-06,
+      "loss": 1.095,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7623997401364284,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.544836001999306e-06,
+      "loss": 1.1145,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7625996351915244,
+      "grad_norm": 2.125,
+      "learning_rate": 8.544092738861336e-06,
+      "loss": 1.0481,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7627995302466205,
+      "grad_norm": 2.25,
+      "learning_rate": 8.543349318295145e-06,
+      "loss": 1.0895,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7629994253017166,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.542605740333754e-06,
+      "loss": 1.1261,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7631993203568127,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.541862005010198e-06,
+      "loss": 0.9861,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7633992154119087,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.541118112357507e-06,
+      "loss": 0.97,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7635991104670048,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.54037406240873e-06,
+      "loss": 1.0524,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7637990055221009,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.539629855196913e-06,
+      "loss": 1.0944,
+      "step": 3821
+    },
+    {
+      "epoch": 0.763998900577197,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.538885490755117e-06,
+      "loss": 0.9729,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7641987956322931,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.538140969116406e-06,
+      "loss": 1.0453,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7643986906873891,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.537396290313849e-06,
+      "loss": 1.0251,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7645985857424852,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.53665145438053e-06,
+      "loss": 1.1079,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7647984807975813,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.535906461349528e-06,
+      "loss": 1.0166,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7649983758526774,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.53516131125394e-06,
+      "loss": 1.0555,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7651982709077734,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.534416004126863e-06,
+      "loss": 1.1303,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7653981659628695,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.533670540001406e-06,
+      "loss": 1.0398,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7655980610179656,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.53292491891068e-06,
+      "loss": 1.0753,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7657979560730617,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.532179140887807e-06,
+      "loss": 0.973,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7659978511281578,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.531433205965913e-06,
+      "loss": 0.9727,
+      "step": 3832
+    },
+    {
+      "epoch": 0.7661977461832538,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.530687114178134e-06,
+      "loss": 1.0441,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7663976412383499,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.529940865557611e-06,
+      "loss": 0.9827,
+      "step": 3834
+    },
+    {
+      "epoch": 0.766597536293446,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.52919446013749e-06,
+      "loss": 1.0333,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7667974313485421,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.528447897950932e-06,
+      "loss": 1.0204,
+      "step": 3836
+    },
+    {
+      "epoch": 0.766997326403638,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.527701179031092e-06,
+      "loss": 1.1111,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7671972214587341,
+      "grad_norm": 2.0,
+      "learning_rate": 8.526954303411141e-06,
+      "loss": 1.0995,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7673971165138302,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.526207271124258e-06,
+      "loss": 0.9218,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7675970115689263,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.525460082203626e-06,
+      "loss": 1.113,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7677969066240223,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.524712736682433e-06,
+      "loss": 0.9998,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7679968016791184,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.523965234593877e-06,
+      "loss": 1.0899,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7681966967342145,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.523217575971159e-06,
+      "loss": 1.0126,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7683965917893106,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.522469760847496e-06,
+      "loss": 0.9371,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7685964868444067,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.5217217892561e-06,
+      "loss": 1.0394,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7687963818995027,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.520973661230198e-06,
+      "loss": 1.0531,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7689962769545988,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.520225376803024e-06,
+      "loss": 1.0576,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7691961720096949,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.519476936007814e-06,
+      "loss": 1.0065,
+      "step": 3848
+    },
+    {
+      "epoch": 0.769396067064791,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.518728338877814e-06,
+      "loss": 1.0778,
+      "step": 3849
+    },
+    {
+      "epoch": 0.769595962119887,
+      "grad_norm": 2.0,
+      "learning_rate": 8.51797958544628e-06,
+      "loss": 0.9625,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7697958571749831,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.517230675746464e-06,
+      "loss": 1.035,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7699957522300792,
+      "grad_norm": 2.125,
+      "learning_rate": 8.51648160981164e-06,
+      "loss": 1.0866,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7701956472851753,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.515732387675077e-06,
+      "loss": 1.1333,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7703955423402714,
+      "grad_norm": 2.125,
+      "learning_rate": 8.514983009370057e-06,
+      "loss": 1.0626,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7705954373953674,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.514233474929867e-06,
+      "loss": 1.0494,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7707953324504635,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.5134837843878e-06,
+      "loss": 1.1193,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7709952275055596,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.512733937777162e-06,
+      "loss": 0.9475,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7711951225606557,
+      "grad_norm": 2.0,
+      "learning_rate": 8.511983935131256e-06,
+      "loss": 1.052,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7713950176157517,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.511233776483398e-06,
+      "loss": 1.0843,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7715949126708478,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.510483461866911e-06,
+      "loss": 1.0252,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7717948077259439,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.509732991315125e-06,
+      "loss": 1.0261,
+      "step": 3861
+    },
+    {
+      "epoch": 0.77199470278104,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.508982364861373e-06,
+      "loss": 1.0702,
+      "step": 3862
+    },
+    {
+      "epoch": 0.772194597836136,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.508231582538999e-06,
+      "loss": 1.0123,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7723944928912321,
+      "grad_norm": 2.125,
+      "learning_rate": 8.507480644381355e-06,
+      "loss": 1.0134,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7725943879463282,
+      "grad_norm": 2.125,
+      "learning_rate": 8.506729550421791e-06,
+      "loss": 1.1005,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7727942830014243,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.505978300693678e-06,
+      "loss": 1.0609,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7729941780565204,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.505226895230383e-06,
+      "loss": 1.0536,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7731940731116164,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.504475334065283e-06,
+      "loss": 1.0487,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7733939681667125,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.503723617231764e-06,
+      "loss": 1.0884,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7735938632218086,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.502971744763216e-06,
+      "loss": 1.0808,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7737937582769047,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.502219716693037e-06,
+      "loss": 1.0959,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7739936533320007,
+      "grad_norm": 2.125,
+      "learning_rate": 8.501467533054632e-06,
+      "loss": 1.1245,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.500715193881412e-06,
+      "loss": 0.8819,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7743934434421929,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.499962699206798e-06,
+      "loss": 0.9738,
+      "step": 3874
+    },
+    {
+      "epoch": 0.774593338497289,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.499210049064216e-06,
+      "loss": 1.1011,
+      "step": 3875
+    },
+    {
+      "epoch": 0.774793233552385,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.498457243487096e-06,
+      "loss": 1.0569,
+      "step": 3876
+    },
+    {
+      "epoch": 0.774993128607481,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.497704282508879e-06,
+      "loss": 1.0651,
+      "step": 3877
+    },
+    {
+      "epoch": 0.7751930236625771,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.496951166163012e-06,
+      "loss": 1.0435,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7753929187176732,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.496197894482947e-06,
+      "loss": 0.9878,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7755928137727693,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.495444467502144e-06,
+      "loss": 1.1107,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7757927088278653,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.494690885254073e-06,
+      "loss": 1.0232,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7759926038829614,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.493937147772204e-06,
+      "loss": 1.058,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7761924989380575,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.493183255090022e-06,
+      "loss": 1.043,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7763923939931536,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.492429207241013e-06,
+      "loss": 0.9438,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7765922890482496,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.49167500425867e-06,
+      "loss": 1.0213,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7767921841033457,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.490920646176497e-06,
+      "loss": 1.0034,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7769920791584418,
+      "grad_norm": 2.125,
+      "learning_rate": 8.490166133028002e-06,
+      "loss": 1.0945,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7771919742135379,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.4894114648467e-06,
+      "loss": 0.9811,
+      "step": 3888
+    },
+    {
+      "epoch": 0.777391869268634,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.488656641666113e-06,
+      "loss": 1.1282,
+      "step": 3889
+    },
+    {
+      "epoch": 0.77759176432373,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.487901663519771e-06,
+      "loss": 0.9224,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7777916593788261,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.48714653044121e-06,
+      "loss": 1.0431,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7779915544339222,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.486391242463972e-06,
+      "loss": 1.0122,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7781914494890183,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.485635799621607e-06,
+      "loss": 1.112,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7783913445441143,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.484880201947674e-06,
+      "loss": 0.95,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7785912395992104,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.484124449475733e-06,
+      "loss": 1.0991,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7787911346543065,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.483368542239356e-06,
+      "loss": 1.0603,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7789910297094026,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.482612480272121e-06,
+      "loss": 1.0628,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7791909247644987,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.481856263607611e-06,
+      "loss": 1.004,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7793908198195947,
+      "grad_norm": 2.25,
+      "learning_rate": 8.481099892279418e-06,
+      "loss": 1.0244,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7795907148746908,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.480343366321138e-06,
+      "loss": 1.0226,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7797906099297869,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.47958668576638e-06,
+      "loss": 0.9943,
+      "step": 3901
+    },
+    {
+      "epoch": 0.779990504984883,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.478829850648752e-06,
+      "loss": 1.0683,
+      "step": 3902
+    },
+    {
+      "epoch": 0.780190400039979,
+      "grad_norm": 2.125,
+      "learning_rate": 8.478072861001872e-06,
+      "loss": 1.0354,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7803902950950751,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.477315716859369e-06,
+      "loss": 1.1009,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7805901901501712,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.476558418254872e-06,
+      "loss": 1.1659,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7807900852052673,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.475800965222024e-06,
+      "loss": 1.0932,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7809899802603633,
+      "grad_norm": 2.0,
+      "learning_rate": 8.475043357794466e-06,
+      "loss": 0.9968,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7811898753154594,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.474285596005853e-06,
+      "loss": 1.0099,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7813897703705555,
+      "grad_norm": 2.25,
+      "learning_rate": 8.473527679889847e-06,
+      "loss": 1.0626,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7815896654256516,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.472769609480109e-06,
+      "loss": 1.1036,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7817895604807477,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.472011384810317e-06,
+      "loss": 1.0819,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7819894555358436,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.471253005914153e-06,
+      "loss": 1.0333,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7821893505909397,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.470494472825299e-06,
+      "loss": 1.0854,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7823892456460358,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.469735785577451e-06,
+      "loss": 0.9313,
+      "step": 3914
+    },
+    {
+      "epoch": 0.782589140701132,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.46897694420431e-06,
+      "loss": 1.0547,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7827890357562279,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.468217948739583e-06,
+      "loss": 1.0774,
+      "step": 3916
+    },
+    {
+      "epoch": 0.782988930811324,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.467458799216987e-06,
+      "loss": 1.0402,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7831888258664201,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.46669949567024e-06,
+      "loss": 0.9801,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7833887209215162,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.465940038133072e-06,
+      "loss": 1.0642,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7835886159766122,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.465180426639218e-06,
+      "loss": 1.0513,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7837885110317083,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.464420661222419e-06,
+      "loss": 1.0154,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7839884060868044,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.463660741916425e-06,
+      "loss": 1.0165,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7841883011419005,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.46290066875499e-06,
+      "loss": 1.0002,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7843881961969966,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.462140441771878e-06,
+      "loss": 1.0972,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7845880912520926,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.461380061000857e-06,
+      "loss": 1.0545,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7847879863071887,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.460619526475704e-06,
+      "loss": 1.0847,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7849878813622848,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.4598588382302e-06,
+      "loss": 0.9939,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7851877764173809,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.459097996298137e-06,
+      "loss": 1.1484,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7853876714724769,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.45833700071331e-06,
+      "loss": 1.029,
+      "step": 3929
+    },
+    {
+      "epoch": 0.785587566527573,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.457575851509523e-06,
+      "loss": 0.9141,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7857874615826691,
+      "grad_norm": 2.125,
+      "learning_rate": 8.456814548720584e-06,
+      "loss": 1.0615,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7859873566377652,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.456053092380313e-06,
+      "loss": 0.8921,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7861872516928613,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.455291482522536e-06,
+      "loss": 1.1657,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7863871467479573,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.454529719181077e-06,
+      "loss": 1.0102,
+      "step": 3934
+    },
+    {
+      "epoch": 0.7865870418030534,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.453767802389776e-06,
+      "loss": 1.0866,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7867869368581495,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.453005732182482e-06,
+      "loss": 1.0282,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7869868319132456,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.452243508593038e-06,
+      "loss": 1.1174,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7871867269683416,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.451481131655308e-06,
+      "loss": 0.9955,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7873866220234377,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.450718601403155e-06,
+      "loss": 0.9518,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7875865170785338,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.449955917870449e-06,
+      "loss": 1.1177,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7877864121336299,
+      "grad_norm": 2.125,
+      "learning_rate": 8.449193081091071e-06,
+      "loss": 1.1051,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7879863071887259,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.448430091098904e-06,
+      "loss": 1.0499,
+      "step": 3942
+    },
+    {
+      "epoch": 0.788186202243822,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.447666947927842e-06,
+      "loss": 1.0631,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7883860972989181,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.446903651611782e-06,
+      "loss": 1.0635,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7885859923540142,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.446140202184628e-06,
+      "loss": 0.8986,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7887858874091103,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.445376599680295e-06,
+      "loss": 1.0746,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7889857824642063,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.444612844132703e-06,
+      "loss": 0.9498,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7891856775193024,
+      "grad_norm": 2.125,
+      "learning_rate": 8.443848935575776e-06,
+      "loss": 1.0482,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7893855725743985,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.443084874043446e-06,
+      "loss": 1.0781,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7895854676294946,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.442320659569654e-06,
+      "loss": 1.1076,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7897853626845905,
+      "grad_norm": 2.0,
+      "learning_rate": 8.441556292188347e-06,
+      "loss": 0.984,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7899852577396866,
+      "grad_norm": 2.125,
+      "learning_rate": 8.440791771933474e-06,
+      "loss": 1.0832,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7901851527947827,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.440027098839002e-06,
+      "loss": 0.9624,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7903850478498788,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.439262272938891e-06,
+      "loss": 1.1171,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7905849429049749,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.438497294267117e-06,
+      "loss": 1.1415,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7907848379600709,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.43773216285766e-06,
+      "loss": 1.0359,
+      "step": 3956
+    },
+    {
+      "epoch": 0.790984733015167,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.43696687874451e-06,
+      "loss": 1.0812,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7911846280702631,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.436201441961658e-06,
+      "loss": 1.0503,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7913845231253592,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.435435852543103e-06,
+      "loss": 1.1193,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7915844181804552,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.434670110522855e-06,
+      "loss": 0.9925,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7917843132355513,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.433904215934929e-06,
+      "loss": 1.1037,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7919842082906474,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.433138168813344e-06,
+      "loss": 0.9896,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7921841033457435,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.432371969192127e-06,
+      "loss": 1.1006,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7923839984008395,
+      "grad_norm": 2.0,
+      "learning_rate": 8.431605617105314e-06,
+      "loss": 1.0182,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7925838934559356,
+      "grad_norm": 2.125,
+      "learning_rate": 8.430839112586947e-06,
+      "loss": 1.0864,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7927837885110317,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.430072455671072e-06,
+      "loss": 1.0649,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7929836835661278,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.429305646391746e-06,
+      "loss": 1.0821,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7931835786212239,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.428538684783027e-06,
+      "loss": 1.0131,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7933834736763199,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.427771570878988e-06,
+      "loss": 1.1314,
+      "step": 3969
+    },
+    {
+      "epoch": 0.793583368731416,
+      "grad_norm": 2.25,
+      "learning_rate": 8.427004304713701e-06,
+      "loss": 1.1188,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7937832637865121,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.42623688632125e-06,
+      "loss": 0.9961,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7939831588416082,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.425469315735722e-06,
+      "loss": 1.0934,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7941830538967042,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.424701592991212e-06,
+      "loss": 0.9757,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7943829489518003,
+      "grad_norm": 2.125,
+      "learning_rate": 8.423933718121823e-06,
+      "loss": 1.0581,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7945828440068964,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.423165691161665e-06,
+      "loss": 1.033,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7947827390619925,
+      "grad_norm": 2.125,
+      "learning_rate": 8.422397512144854e-06,
+      "loss": 1.0961,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7949826341170886,
+      "grad_norm": 2.25,
+      "learning_rate": 8.421629181105509e-06,
+      "loss": 1.08,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7951825291721846,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.42086069807776e-06,
+      "loss": 1.0658,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7953824242272807,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.420092063095746e-06,
+      "loss": 0.9352,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7955823192823768,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.419323276193607e-06,
+      "loss": 1.0642,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7957822143374729,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.418554337405493e-06,
+      "loss": 1.0289,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7959821093925689,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.417785246765561e-06,
+      "loss": 1.1825,
+      "step": 3982
+    },
+    {
+      "epoch": 0.796182004447665,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.417016004307974e-06,
+      "loss": 0.9411,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7963818995027611,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.4162466100669e-06,
+      "loss": 1.015,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7965817945578572,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.415477064076518e-06,
+      "loss": 1.0249,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7967816896129531,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.414707366371006e-06,
+      "loss": 1.1011,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7969815846680492,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.41393751698456e-06,
+      "loss": 1.15,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7971814797231453,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.413167515951374e-06,
+      "loss": 1.0764,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7973813747782414,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.412397363305653e-06,
+      "loss": 1.0558,
+      "step": 3989
+    },
+    {
+      "epoch": 0.7975812698333375,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.411627059081603e-06,
+      "loss": 1.0267,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7977811648884335,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.410856603313446e-06,
+      "loss": 1.0978,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7979810599435296,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.410085996035402e-06,
+      "loss": 1.0463,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7981809549986257,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.409315237281701e-06,
+      "loss": 1.0431,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7983808500537218,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.408544327086583e-06,
+      "loss": 1.2017,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7985807451088178,
+      "grad_norm": 2.125,
+      "learning_rate": 8.407773265484289e-06,
+      "loss": 1.1155,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7987806401639139,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.40700205250907e-06,
+      "loss": 1.0254,
+      "step": 3996
+    },
+    {
+      "epoch": 0.79898053521901,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.406230688195184e-06,
+      "loss": 0.9988,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7991804302741061,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.405459172576895e-06,
+      "loss": 0.9203,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7993803253292022,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.404687505688474e-06,
+      "loss": 0.9984,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7995802203842982,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.403915687564198e-06,
+      "loss": 0.9678,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7997801154393943,
+      "grad_norm": 2.0,
+      "learning_rate": 8.403143718238347e-06,
+      "loss": 0.986,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7999800104944904,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.402371597745218e-06,
+      "loss": 1.0025,
+      "step": 4002
+    },
+    {
+      "epoch": 0.8001799055495865,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.401599326119108e-06,
+      "loss": 1.0755,
+      "step": 4003
+    },
+    {
+      "epoch": 0.8003798006046825,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.400826903394317e-06,
+      "loss": 0.8787,
+      "step": 4004
+    },
+    {
+      "epoch": 0.8005796956597786,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.400054329605159e-06,
+      "loss": 0.9739,
+      "step": 4005
+    },
+    {
+      "epoch": 0.8007795907148747,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.39928160478595e-06,
+      "loss": 0.9824,
+      "step": 4006
+    },
+    {
+      "epoch": 0.8009794857699708,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.398508728971016e-06,
+      "loss": 1.004,
+      "step": 4007
+    },
+    {
+      "epoch": 0.8011793808250668,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.397735702194686e-06,
+      "loss": 0.9618,
+      "step": 4008
+    },
+    {
+      "epoch": 0.8013792758801629,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.396962524491299e-06,
+      "loss": 1.1228,
+      "step": 4009
+    },
+    {
+      "epoch": 0.801579170935259,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.3961891958952e-06,
+      "loss": 0.9377,
+      "step": 4010
+    },
+    {
+      "epoch": 0.8017790659903551,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.39541571644074e-06,
+      "loss": 1.0748,
+      "step": 4011
+    },
+    {
+      "epoch": 0.8019789610454512,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.394642086162278e-06,
+      "loss": 1.0429,
+      "step": 4012
+    },
+    {
+      "epoch": 0.8021788561005472,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.393868305094173e-06,
+      "loss": 1.099,
+      "step": 4013
+    },
+    {
+      "epoch": 0.8023787511556433,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.393094373270804e-06,
+      "loss": 1.0019,
+      "step": 4014
+    },
+    {
+      "epoch": 0.8025786462107394,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.392320290726543e-06,
+      "loss": 1.059,
+      "step": 4015
+    },
+    {
+      "epoch": 0.8027785412658355,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.391546057495778e-06,
+      "loss": 1.0901,
+      "step": 4016
+    },
+    {
+      "epoch": 0.8029784363209315,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.3907716736129e-06,
+      "loss": 1.0134,
+      "step": 4017
+    },
+    {
+      "epoch": 0.8031783313760276,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.389997139112306e-06,
+      "loss": 0.9359,
+      "step": 4018
+    },
+    {
+      "epoch": 0.8033782264311237,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.3892224540284e-06,
+      "loss": 1.0603,
+      "step": 4019
+    },
+    {
+      "epoch": 0.8035781214862198,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.388447618395598e-06,
+      "loss": 0.9355,
+      "step": 4020
+    },
+    {
+      "epoch": 0.8037780165413158,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.387672632248312e-06,
+      "loss": 1.1244,
+      "step": 4021
+    },
+    {
+      "epoch": 0.8039779115964119,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.38689749562097e-06,
+      "loss": 1.0161,
+      "step": 4022
+    },
+    {
+      "epoch": 0.804177806651508,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.386122208548002e-06,
+      "loss": 0.9934,
+      "step": 4023
+    },
+    {
+      "epoch": 0.804377701706604,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.385346771063848e-06,
+      "loss": 1.0326,
+      "step": 4024
+    },
+    {
+      "epoch": 0.8045775967617002,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.384571183202952e-06,
+      "loss": 1.059,
+      "step": 4025
+    },
+    {
+      "epoch": 0.8047774918167961,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.383795444999766e-06,
+      "loss": 1.0174,
+      "step": 4026
+    },
+    {
+      "epoch": 0.8049773868718922,
+      "grad_norm": 2.0,
+      "learning_rate": 8.383019556488747e-06,
+      "loss": 1.0145,
+      "step": 4027
+    },
+    {
+      "epoch": 0.8051772819269883,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.38224351770436e-06,
+      "loss": 1.0756,
+      "step": 4028
+    },
+    {
+      "epoch": 0.8053771769820844,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.381467328681078e-06,
+      "loss": 1.0536,
+      "step": 4029
+    },
+    {
+      "epoch": 0.8055770720371804,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.380690989453379e-06,
+      "loss": 1.0156,
+      "step": 4030
+    },
+    {
+      "epoch": 0.8057769670922765,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.379914500055745e-06,
+      "loss": 1.0135,
+      "step": 4031
+    },
+    {
+      "epoch": 0.8059768621473726,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.379137860522672e-06,
+      "loss": 1.0025,
+      "step": 4032
+    },
+    {
+      "epoch": 0.8061767572024687,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.378361070888656e-06,
+      "loss": 1.1067,
+      "step": 4033
+    },
+    {
+      "epoch": 0.8063766522575648,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.3775841311882e-06,
+      "loss": 1.0107,
+      "step": 4034
+    },
+    {
+      "epoch": 0.8065765473126608,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.376807041455822e-06,
+      "loss": 0.9656,
+      "step": 4035
+    },
+    {
+      "epoch": 0.8067764423677569,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.376029801726033e-06,
+      "loss": 1.0475,
+      "step": 4036
+    },
+    {
+      "epoch": 0.806976337422853,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.375252412033361e-06,
+      "loss": 1.0904,
+      "step": 4037
+    },
+    {
+      "epoch": 0.8071762324779491,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.374474872412338e-06,
+      "loss": 1.0797,
+      "step": 4038
+    },
+    {
+      "epoch": 0.8073761275330451,
+      "grad_norm": 2.0,
+      "learning_rate": 8.3736971828975e-06,
+      "loss": 1.0428,
+      "step": 4039
+    },
+    {
+      "epoch": 0.8075760225881412,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.372919343523395e-06,
+      "loss": 0.9768,
+      "step": 4040
+    },
+    {
+      "epoch": 0.8077759176432373,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.372141354324573e-06,
+      "loss": 1.057,
+      "step": 4041
+    },
+    {
+      "epoch": 0.8079758126983334,
+      "grad_norm": 2.125,
+      "learning_rate": 8.37136321533559e-06,
+      "loss": 0.9486,
+      "step": 4042
+    },
+    {
+      "epoch": 0.8081757077534294,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.370584926591015e-06,
+      "loss": 1.0627,
+      "step": 4043
+    },
+    {
+      "epoch": 0.8083756028085255,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.369806488125418e-06,
+      "loss": 1.0046,
+      "step": 4044
+    },
+    {
+      "epoch": 0.8085754978636216,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.369027899973377e-06,
+      "loss": 0.9561,
+      "step": 4045
+    },
+    {
+      "epoch": 0.8087753929187177,
+      "grad_norm": 2.125,
+      "learning_rate": 8.368249162169474e-06,
+      "loss": 1.1483,
+      "step": 4046
+    },
+    {
+      "epoch": 0.8089752879738138,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.367470274748303e-06,
+      "loss": 1.0364,
+      "step": 4047
+    },
+    {
+      "epoch": 0.8091751830289098,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.366691237744465e-06,
+      "loss": 1.0638,
+      "step": 4048
+    },
+    {
+      "epoch": 0.8093750780840059,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.365912051192559e-06,
+      "loss": 1.0725,
+      "step": 4049
+    },
+    {
+      "epoch": 0.809574973139102,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.365132715127201e-06,
+      "loss": 1.0622,
+      "step": 4050
+    },
+    {
+      "epoch": 0.8097748681941981,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.364353229583007e-06,
+      "loss": 1.115,
+      "step": 4051
+    },
+    {
+      "epoch": 0.8099747632492941,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.363573594594603e-06,
+      "loss": 0.9759,
+      "step": 4052
+    },
+    {
+      "epoch": 0.8101746583043902,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.362793810196616e-06,
+      "loss": 0.9862,
+      "step": 4053
+    },
+    {
+      "epoch": 0.8103745533594863,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.362013876423689e-06,
+      "loss": 1.0371,
+      "step": 4054
+    },
+    {
+      "epoch": 0.8105744484145824,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.361233793310466e-06,
+      "loss": 1.0631,
+      "step": 4055
+    },
+    {
+      "epoch": 0.8107743434696785,
+      "grad_norm": 2.125,
+      "learning_rate": 8.360453560891594e-06,
+      "loss": 1.0111,
+      "step": 4056
+    },
+    {
+      "epoch": 0.8109742385247745,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.359673179201734e-06,
+      "loss": 1.0972,
+      "step": 4057
+    },
+    {
+      "epoch": 0.8111741335798706,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.358892648275554e-06,
+      "loss": 0.9885,
+      "step": 4058
+    },
+    {
+      "epoch": 0.8113740286349667,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.358111968147717e-06,
+      "loss": 0.994,
+      "step": 4059
+    },
+    {
+      "epoch": 0.8115739236900628,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.357331138852907e-06,
+      "loss": 0.9565,
+      "step": 4060
+    },
+    {
+      "epoch": 0.8117738187451587,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.356550160425806e-06,
+      "loss": 1.1143,
+      "step": 4061
+    },
+    {
+      "epoch": 0.8119737138002548,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.355769032901105e-06,
+      "loss": 1.0377,
+      "step": 4062
+    },
+    {
+      "epoch": 0.812173608855351,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.354987756313501e-06,
+      "loss": 0.9248,
+      "step": 4063
+    },
+    {
+      "epoch": 0.812373503910447,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.354206330697702e-06,
+      "loss": 0.9865,
+      "step": 4064
+    },
+    {
+      "epoch": 0.812573398965543,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.353424756088415e-06,
+      "loss": 0.9604,
+      "step": 4065
+    },
+    {
+      "epoch": 0.8127732940206391,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.352643032520357e-06,
+      "loss": 1.1125,
+      "step": 4066
+    },
+    {
+      "epoch": 0.8129731890757352,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.351861160028256e-06,
+      "loss": 1.1097,
+      "step": 4067
+    },
+    {
+      "epoch": 0.8131730841308313,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.351079138646838e-06,
+      "loss": 1.0526,
+      "step": 4068
+    },
+    {
+      "epoch": 0.8133729791859274,
+      "grad_norm": 2.125,
+      "learning_rate": 8.350296968410845e-06,
+      "loss": 1.0931,
+      "step": 4069
+    },
+    {
+      "epoch": 0.8135728742410234,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.349514649355016e-06,
+      "loss": 1.0257,
+      "step": 4070
+    },
+    {
+      "epoch": 0.8137727692961195,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.348732181514105e-06,
+      "loss": 1.0811,
+      "step": 4071
+    },
+    {
+      "epoch": 0.8139726643512156,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.347949564922869e-06,
+      "loss": 1.1355,
+      "step": 4072
+    },
+    {
+      "epoch": 0.8141725594063117,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.347166799616069e-06,
+      "loss": 1.0506,
+      "step": 4073
+    },
+    {
+      "epoch": 0.8143724544614077,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.346383885628478e-06,
+      "loss": 1.0503,
+      "step": 4074
+    },
+    {
+      "epoch": 0.8145723495165038,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.345600822994872e-06,
+      "loss": 1.0772,
+      "step": 4075
+    },
+    {
+      "epoch": 0.8147722445715999,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.344817611750036e-06,
+      "loss": 1.0449,
+      "step": 4076
+    },
+    {
+      "epoch": 0.814972139626696,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.344034251928759e-06,
+      "loss": 0.9987,
+      "step": 4077
+    },
+    {
+      "epoch": 0.8151720346817921,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.343250743565837e-06,
+      "loss": 1.0547,
+      "step": 4078
+    },
+    {
+      "epoch": 0.8153719297368881,
+      "grad_norm": 2.125,
+      "learning_rate": 8.342467086696073e-06,
+      "loss": 1.1069,
+      "step": 4079
+    },
+    {
+      "epoch": 0.8155718247919842,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.341683281354277e-06,
+      "loss": 1.0543,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8157717198470803,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.34089932757527e-06,
+      "loss": 1.0877,
+      "step": 4081
+    },
+    {
+      "epoch": 0.8159716149021764,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.34011522539387e-06,
+      "loss": 1.1096,
+      "step": 4082
+    },
+    {
+      "epoch": 0.8161715099572724,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.339330974844908e-06,
+      "loss": 1.0155,
+      "step": 4083
+    },
+    {
+      "epoch": 0.8163714050123685,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.338546575963223e-06,
+      "loss": 1.0979,
+      "step": 4084
+    },
+    {
+      "epoch": 0.8165713000674646,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.337762028783653e-06,
+      "loss": 1.0712,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8167711951225607,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.336977333341052e-06,
+      "loss": 1.034,
+      "step": 4086
+    },
+    {
+      "epoch": 0.8169710901776567,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.336192489670273e-06,
+      "loss": 1.0542,
+      "step": 4087
+    },
+    {
+      "epoch": 0.8171709852327528,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.33540749780618e-06,
+      "loss": 0.9744,
+      "step": 4088
+    },
+    {
+      "epoch": 0.8173708802878489,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.334622357783642e-06,
+      "loss": 1.0116,
+      "step": 4089
+    },
+    {
+      "epoch": 0.817570775342945,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.333837069637536e-06,
+      "loss": 1.1014,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8177706703980411,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.333051633402743e-06,
+      "loss": 0.9645,
+      "step": 4091
+    },
+    {
+      "epoch": 0.8179705654531371,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.332266049114152e-06,
+      "loss": 0.9681,
+      "step": 4092
+    },
+    {
+      "epoch": 0.8181704605082332,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.33148031680666e-06,
+      "loss": 1.0724,
+      "step": 4093
+    },
+    {
+      "epoch": 0.8183703555633293,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.330694436515168e-06,
+      "loss": 1.0307,
+      "step": 4094
+    },
+    {
+      "epoch": 0.8185702506184254,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.329908408274583e-06,
+      "loss": 1.1436,
+      "step": 4095
+    },
+    {
+      "epoch": 0.8187701456735214,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.329122232119824e-06,
+      "loss": 1.0197,
+      "step": 4096
+    },
+    {
+      "epoch": 0.8189700407286175,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.32833590808581e-06,
+      "loss": 0.9844,
+      "step": 4097
+    },
+    {
+      "epoch": 0.8191699357837136,
+      "grad_norm": 2.0,
+      "learning_rate": 8.327549436207472e-06,
+      "loss": 0.9957,
+      "step": 4098
+    },
+    {
+      "epoch": 0.8193698308388097,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.326762816519743e-06,
+      "loss": 0.9574,
+      "step": 4099
+    },
+    {
+      "epoch": 0.8195697258939058,
+      "grad_norm": 2.0,
+      "learning_rate": 8.325976049057565e-06,
+      "loss": 0.9851,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8197696209490017,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.325189133855884e-06,
+      "loss": 1.1136,
+      "step": 4101
+    },
+    {
+      "epoch": 0.8199695160040978,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.324402070949658e-06,
+      "loss": 1.0151,
+      "step": 4102
+    },
+    {
+      "epoch": 0.8201694110591939,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.323614860373848e-06,
+      "loss": 1.0295,
+      "step": 4103
+    },
+    {
+      "epoch": 0.82036930611429,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.322827502163422e-06,
+      "loss": 1.0286,
+      "step": 4104
+    },
+    {
+      "epoch": 0.820569201169386,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.32203999635335e-06,
+      "loss": 1.1238,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8207690962244821,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.321252342978617e-06,
+      "loss": 1.1182,
+      "step": 4106
+    },
+    {
+      "epoch": 0.8209689912795782,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.32046454207421e-06,
+      "loss": 1.0793,
+      "step": 4107
+    },
+    {
+      "epoch": 0.8211688863346743,
+      "grad_norm": 2.0,
+      "learning_rate": 8.319676593675124e-06,
+      "loss": 0.9489,
+      "step": 4108
+    },
+    {
+      "epoch": 0.8213687813897703,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.318888497816357e-06,
+      "loss": 0.9854,
+      "step": 4109
+    },
+    {
+      "epoch": 0.8215686764448664,
+      "grad_norm": 2.0,
+      "learning_rate": 8.318100254532917e-06,
+      "loss": 0.8775,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8217685714999625,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.31731186385982e-06,
+      "loss": 1.0759,
+      "step": 4111
+    },
+    {
+      "epoch": 0.8219684665550586,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.316523325832083e-06,
+      "loss": 0.9916,
+      "step": 4112
+    },
+    {
+      "epoch": 0.8221683616101547,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.315734640484734e-06,
+      "loss": 1.062,
+      "step": 4113
+    },
+    {
+      "epoch": 0.8223682566652507,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.314945807852808e-06,
+      "loss": 1.061,
+      "step": 4114
+    },
+    {
+      "epoch": 0.8225681517203468,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.31415682797134e-06,
+      "loss": 1.0866,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8227680467754429,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.313367700875381e-06,
+      "loss": 1.17,
+      "step": 4116
+    },
+    {
+      "epoch": 0.822967941830539,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.312578426599984e-06,
+      "loss": 0.9527,
+      "step": 4117
+    },
+    {
+      "epoch": 0.823167836885635,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.311789005180207e-06,
+      "loss": 1.085,
+      "step": 4118
+    },
+    {
+      "epoch": 0.8233677319407311,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.310999436651115e-06,
+      "loss": 1.0659,
+      "step": 4119
+    },
+    {
+      "epoch": 0.8235676269958272,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.310209721047782e-06,
+      "loss": 1.0947,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8237675220509233,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.309419858405287e-06,
+      "loss": 0.9207,
+      "step": 4121
+    },
+    {
+      "epoch": 0.8239674171060194,
+      "grad_norm": 2.0,
+      "learning_rate": 8.308629848758714e-06,
+      "loss": 1.0384,
+      "step": 4122
+    },
+    {
+      "epoch": 0.8241673121611154,
+      "grad_norm": 2.25,
+      "learning_rate": 8.307839692143158e-06,
+      "loss": 1.0301,
+      "step": 4123
+    },
+    {
+      "epoch": 0.8243672072162115,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.307049388593716e-06,
+      "loss": 1.0894,
+      "step": 4124
+    },
+    {
+      "epoch": 0.8245671022713076,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.306258938145493e-06,
+      "loss": 0.9666,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8247669973264037,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.3054683408336e-06,
+      "loss": 1.0584,
+      "step": 4126
+    },
+    {
+      "epoch": 0.8249668923814997,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.304677596693158e-06,
+      "loss": 1.0143,
+      "step": 4127
+    },
+    {
+      "epoch": 0.8251667874365958,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.30388670575929e-06,
+      "loss": 0.9847,
+      "step": 4128
+    },
+    {
+      "epoch": 0.8253666824916919,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.303095668067127e-06,
+      "loss": 0.9561,
+      "step": 4129
+    },
+    {
+      "epoch": 0.825566577546788,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.302304483651806e-06,
+      "loss": 1.0133,
+      "step": 4130
+    },
+    {
+      "epoch": 0.825766472601884,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.301513152548474e-06,
+      "loss": 1.0884,
+      "step": 4131
+    },
+    {
+      "epoch": 0.8259663676569801,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.30072167479228e-06,
+      "loss": 1.0631,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8261662627120762,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.299930050418383e-06,
+      "loss": 1.0408,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8263661577671723,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.299138279461945e-06,
+      "loss": 1.0644,
+      "step": 4134
+    },
+    {
+      "epoch": 0.8265660528222684,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.298346361958136e-06,
+      "loss": 0.9942,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8267659478773643,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.297554297942134e-06,
+      "loss": 1.0154,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8269658429324604,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.296762087449122e-06,
+      "loss": 1.07,
+      "step": 4137
+    },
+    {
+      "epoch": 0.8271657379875565,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.29596973051429e-06,
+      "loss": 1.014,
+      "step": 4138
+    },
+    {
+      "epoch": 0.8273656330426526,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.295177227172837e-06,
+      "loss": 0.9717,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8275655280977486,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.294384577459961e-06,
+      "loss": 1.0929,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8277654231528447,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.293591781410874e-06,
+      "loss": 0.992,
+      "step": 4141
+    },
+    {
+      "epoch": 0.8279653182079408,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.292798839060794e-06,
+      "loss": 1.1135,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8281652132630369,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.29200575044494e-06,
+      "loss": 1.0303,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8283651083181329,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.29121251559854e-06,
+      "loss": 0.9511,
+      "step": 4144
+    },
+    {
+      "epoch": 0.828565003373229,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.290419134556835e-06,
+      "loss": 1.0883,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8287648984283251,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.289625607355062e-06,
+      "loss": 1.0834,
+      "step": 4146
+    },
+    {
+      "epoch": 0.8289647934834212,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.288831934028471e-06,
+      "loss": 1.0117,
+      "step": 4147
+    },
+    {
+      "epoch": 0.8291646885385173,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.288038114612316e-06,
+      "loss": 0.9673,
+      "step": 4148
+    },
+    {
+      "epoch": 0.8293645835936133,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.287244149141861e-06,
+      "loss": 1.0559,
+      "step": 4149
+    },
+    {
+      "epoch": 0.8295644786487094,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.286450037652369e-06,
+      "loss": 1.0355,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8297643737038055,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.28565578017912e-06,
+      "loss": 1.0386,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8299642687589016,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.284861376757391e-06,
+      "loss": 1.0309,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8301641638139976,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.28406682742247e-06,
+      "loss": 0.9418,
+      "step": 4153
+    },
+    {
+      "epoch": 0.8303640588690937,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.283272132209653e-06,
+      "loss": 1.0459,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8305639539241898,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.282477291154238e-06,
+      "loss": 1.0016,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8307638489792859,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.281682304291531e-06,
+      "loss": 1.0282,
+      "step": 4156
+    },
+    {
+      "epoch": 0.830963744034382,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.280887171656848e-06,
+      "loss": 1.0862,
+      "step": 4157
+    },
+    {
+      "epoch": 0.831163639089478,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.280091893285508e-06,
+      "loss": 1.0085,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8313635341445741,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.279296469212833e-06,
+      "loss": 1.0787,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8315634291996702,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.278500899474162e-06,
+      "loss": 1.0031,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8317633242547663,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.277705184104831e-06,
+      "loss": 0.9732,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8319632193098623,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.276909323140186e-06,
+      "loss": 0.9604,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8321631143649584,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.276113316615577e-06,
+      "loss": 1.0415,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8323630094200545,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.275317164566365e-06,
+      "loss": 1.0282,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8325629044751506,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.274520867027915e-06,
+      "loss": 1.0074,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8327627995302466,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.273724424035599e-06,
+      "loss": 1.022,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8329626945853427,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.272927835624791e-06,
+      "loss": 1.0232,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8331625896404388,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.272131101830878e-06,
+      "loss": 1.0433,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8333624846955349,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.271334222689254e-06,
+      "loss": 1.0848,
+      "step": 4169
+    },
+    {
+      "epoch": 0.833562379750631,
+      "grad_norm": 2.0,
+      "learning_rate": 8.270537198235311e-06,
+      "loss": 1.024,
+      "step": 4170
+    },
+    {
+      "epoch": 0.833762274805727,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.269740028504455e-06,
+      "loss": 1.0799,
+      "step": 4171
+    },
+    {
+      "epoch": 0.833962169860823,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.268942713532098e-06,
+      "loss": 1.02,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8341620649159192,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.268145253353653e-06,
+      "loss": 0.9191,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8343619599710153,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.267347648004545e-06,
+      "loss": 1.1071,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8345618550261112,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.266549897520204e-06,
+      "loss": 1.0778,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8347617500812073,
+      "grad_norm": 2.0,
+      "learning_rate": 8.265752001936067e-06,
+      "loss": 1.0233,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8349616451363034,
+      "grad_norm": 2.0,
+      "learning_rate": 8.264953961287573e-06,
+      "loss": 1.0217,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8351615401913995,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.264155775610172e-06,
+      "loss": 1.0185,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8353614352464956,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.263357444939321e-06,
+      "loss": 1.0463,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8355613303015916,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.26255896931048e-06,
+      "loss": 1.0506,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8357612253566877,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.26176034875912e-06,
+      "loss": 1.0384,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8359611204117838,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.26096158332071e-06,
+      "loss": 1.03,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8361610154668799,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.260162673030739e-06,
+      "loss": 1.1151,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8363609105219759,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.259363617924689e-06,
+      "loss": 0.9808,
+      "step": 4184
+    },
+    {
+      "epoch": 0.836560805577072,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.258564418038053e-06,
+      "loss": 1.0657,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8367607006321681,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.257765073406337e-06,
+      "loss": 1.0732,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8369605956872642,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.256965584065042e-06,
+      "loss": 0.9965,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8371604907423602,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.256165950049684e-06,
+      "loss": 1.0115,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8373603857974563,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.255366171395783e-06,
+      "loss": 1.1086,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8375602808525524,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.254566248138865e-06,
+      "loss": 0.9886,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8377601759076485,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.25376618031446e-06,
+      "loss": 1.0307,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8379600709627446,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.252965967958108e-06,
+      "loss": 1.1474,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8381599660178406,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.252165611105358e-06,
+      "loss": 1.0694,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8383598610729367,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.251365109791758e-06,
+      "loss": 0.9362,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8385597561280328,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.250564464052865e-06,
+      "loss": 0.9577,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8387596511831289,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.249763673924248e-06,
+      "loss": 0.9991,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8389595462382249,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.248962739441475e-06,
+      "loss": 0.9452,
+      "step": 4197
+    },
+    {
+      "epoch": 0.839159441293321,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.248161660640123e-06,
+      "loss": 1.0779,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8393593363484171,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.24736043755578e-06,
+      "loss": 1.0068,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8395592314035132,
+      "grad_norm": 2.0,
+      "learning_rate": 8.24655907022403e-06,
+      "loss": 0.9692,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8397591264586093,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.245757558680474e-06,
+      "loss": 0.9732,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8399590215137053,
+      "grad_norm": 2.0,
+      "learning_rate": 8.244955902960713e-06,
+      "loss": 0.9686,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8401589165688014,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.24415410310036e-06,
+      "loss": 0.9976,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8403588116238975,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.243352159135026e-06,
+      "loss": 1.1032,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8405587066789936,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.242550071100336e-06,
+      "loss": 1.012,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8407586017340896,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.24174783903192e-06,
+      "loss": 1.1062,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8409584967891857,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.240945462965408e-06,
+      "loss": 1.0424,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8411583918442818,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.240142942936446e-06,
+      "loss": 1.0927,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8413582868993779,
+      "grad_norm": 2.125,
+      "learning_rate": 8.239340278980681e-06,
+      "loss": 1.0483,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8415581819544738,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.238537471133768e-06,
+      "loss": 1.0106,
+      "step": 4210
+    },
+    {
+      "epoch": 0.84175807700957,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.237734519431365e-06,
+      "loss": 0.9823,
+      "step": 4211
+    },
+    {
+      "epoch": 0.841957972064666,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.23693142390914e-06,
+      "loss": 1.0696,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8421578671197621,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.236128184602766e-06,
+      "loss": 0.9739,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8423577621748582,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.235324801547926e-06,
+      "loss": 1.0381,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8425576572299542,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.234521274780302e-06,
+      "loss": 0.9209,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8427575522850503,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.233717604335589e-06,
+      "loss": 1.057,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8429574473401464,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.232913790249486e-06,
+      "loss": 0.9963,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8431573423952425,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.232109832557696e-06,
+      "loss": 1.087,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8433572374503385,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.231305731295935e-06,
+      "loss": 1.128,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8435571325054346,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.230501486499915e-06,
+      "loss": 1.0614,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8437570275605307,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.22969709820537e-06,
+      "loss": 1.0407,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8439569226156268,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.228892566448018e-06,
+      "loss": 1.0015,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8441568176707229,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.228087891263608e-06,
+      "loss": 1.0536,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8443567127258189,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.227283072687877e-06,
+      "loss": 1.0252,
+      "step": 4224
+    },
+    {
+      "epoch": 0.844556607780915,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.226478110756574e-06,
+      "loss": 1.0179,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8447565028360111,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.225673005505461e-06,
+      "loss": 1.0577,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8449563978911072,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.224867756970298e-06,
+      "loss": 0.9411,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8451562929462032,
+      "grad_norm": 2.125,
+      "learning_rate": 8.224062365186852e-06,
+      "loss": 1.1332,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8453561880012993,
+      "grad_norm": 2.125,
+      "learning_rate": 8.223256830190901e-06,
+      "loss": 1.0812,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8455560830563954,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.222451152018225e-06,
+      "loss": 0.9662,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8457559781114915,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.221645330704615e-06,
+      "loss": 1.0622,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8459558731665875,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.220839366285862e-06,
+      "loss": 0.948,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8461557682216836,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.220033258797767e-06,
+      "loss": 1.0634,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8463556632767797,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.21922700827614e-06,
+      "loss": 1.0048,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8465555583318758,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.218420614756793e-06,
+      "loss": 1.0041,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8467554533869719,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.217614078275547e-06,
+      "loss": 1.008,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8469553484420679,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.216807398868225e-06,
+      "loss": 0.9848,
+      "step": 4237
+    },
+    {
+      "epoch": 0.847155243497164,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.216000576570664e-06,
+      "loss": 1.0451,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8473551385522601,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.2151936114187e-06,
+      "loss": 0.9849,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8475550336073562,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.21438650344818e-06,
+      "loss": 1.1296,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8477549286624522,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.213579252694954e-06,
+      "loss": 1.0473,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8479548237175483,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.212771859194881e-06,
+      "loss": 1.056,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8481547187726444,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.211964322983824e-06,
+      "loss": 1.0194,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8483546138277405,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.211156644097656e-06,
+      "loss": 1.0683,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8485545088828365,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.210348822572253e-06,
+      "loss": 1.0429,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8487544039379326,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.209540858443499e-06,
+      "loss": 1.0126,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8489542989930287,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.208732751747281e-06,
+      "loss": 1.0767,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8491541940481248,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.207924502519498e-06,
+      "loss": 1.0443,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8493540891032209,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.20711611079605e-06,
+      "loss": 1.0341,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8495539841583168,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.206307576612848e-06,
+      "loss": 0.9718,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8497538792134129,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.205498900005806e-06,
+      "loss": 1.1404,
+      "step": 4251
+    },
+    {
+      "epoch": 0.849953774268509,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.204690081010845e-06,
+      "loss": 0.9937,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8501536693236051,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.203881119663893e-06,
+      "loss": 1.0616,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8503535643787011,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.203072016000884e-06,
+      "loss": 1.064,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8505534594337972,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.202262770057756e-06,
+      "loss": 1.0711,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8507533544888933,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.201453381870461e-06,
+      "loss": 0.9932,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8509532495439894,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.200643851474947e-06,
+      "loss": 1.0243,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8511531445990855,
+      "grad_norm": 2.125,
+      "learning_rate": 8.199834178907174e-06,
+      "loss": 1.0337,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8513530396541815,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.19902436420311e-06,
+      "loss": 1.0912,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8515529347092776,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.198214407398726e-06,
+      "loss": 0.9725,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8517528297643737,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.197404308529997e-06,
+      "loss": 1.0013,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8519527248194698,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.196594067632913e-06,
+      "loss": 1.0067,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8521526198745658,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.195783684743461e-06,
+      "loss": 0.9922,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8523525149296619,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.19497315989764e-06,
+      "loss": 0.9992,
+      "step": 4264
+    },
+    {
+      "epoch": 0.852552409984758,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.19416249313145e-06,
+      "loss": 1.0969,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8527523050398541,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.193351684480904e-06,
+      "loss": 1.0002,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8529522000949501,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.192540733982017e-06,
+      "loss": 0.8837,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8531520951500462,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.191729641670813e-06,
+      "loss": 1.0419,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8533519902051423,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.190918407583319e-06,
+      "loss": 0.9997,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8535518852602384,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.190107031755569e-06,
+      "loss": 1.0607,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8537517803153345,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.189295514223607e-06,
+      "loss": 1.0516,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8539516753704305,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.188483855023476e-06,
+      "loss": 0.9929,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8541515704255266,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.187672054191236e-06,
+      "loss": 1.0549,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8543514654806227,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.186860111762941e-06,
+      "loss": 1.1324,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8545513605357188,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.186048027774661e-06,
+      "loss": 1.1121,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8547512555908148,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.18523580226247e-06,
+      "loss": 0.9482,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8549511506459109,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.184423435262442e-06,
+      "loss": 0.979,
+      "step": 4277
+    },
+    {
+      "epoch": 0.855151045701007,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.183610926810667e-06,
+      "loss": 1.0637,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8553509407561031,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.182798276943236e-06,
+      "loss": 1.1295,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8555508358111992,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.181985485696242e-06,
+      "loss": 1.0336,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8557507308662952,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.181172553105793e-06,
+      "loss": 1.0707,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8559506259213913,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.180359479208002e-06,
+      "loss": 1.1241,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8561505209764874,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.179546264038982e-06,
+      "loss": 1.0029,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8563504160315835,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.178732907634854e-06,
+      "loss": 1.0113,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8565503110866795,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.177919410031752e-06,
+      "loss": 1.0303,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8567502061417755,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.177105771265808e-06,
+      "loss": 0.9785,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8569501011968716,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.176291991373164e-06,
+      "loss": 0.9925,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8571499962519677,
+      "grad_norm": 2.125,
+      "learning_rate": 8.17547807038997e-06,
+      "loss": 1.0772,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8573498913070637,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.17466400835238e-06,
+      "loss": 1.0025,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8575497863621598,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.173849805296553e-06,
+      "loss": 1.0628,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8577496814172559,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.173035461258658e-06,
+      "loss": 1.0494,
+      "step": 4291
+    },
+    {
+      "epoch": 0.857949576472352,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.172220976274865e-06,
+      "loss": 1.0009,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8581494715274481,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.171406350381354e-06,
+      "loss": 1.0389,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8583493665825441,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.170591583614313e-06,
+      "loss": 0.9508,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8585492616376402,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.169776676009935e-06,
+      "loss": 1.0055,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8587491566927363,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.168961627604413e-06,
+      "loss": 0.9748,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8589490517478324,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.168146438433952e-06,
+      "loss": 1.0029,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8591489468029284,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.167331108534769e-06,
+      "loss": 1.1574,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8593488418580245,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.166515637943072e-06,
+      "loss": 0.9615,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8595487369131206,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.165700026695094e-06,
+      "loss": 1.0508,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8597486319682167,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.164884274827055e-06,
+      "loss": 1.0879,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8599485270233128,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.164068382375195e-06,
+      "loss": 0.9528,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8601484220784088,
+      "grad_norm": 2.125,
+      "learning_rate": 8.163252349375755e-06,
+      "loss": 1.0222,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8603483171335049,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.162436175864985e-06,
+      "loss": 0.9972,
+      "step": 4304
+    },
+    {
+      "epoch": 0.860548212188601,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.161619861879136e-06,
+      "loss": 1.0338,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8607481072436971,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.160803407454472e-06,
+      "loss": 1.0123,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8609480022987931,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.159986812627258e-06,
+      "loss": 1.0679,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8611478973538892,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.159170077433766e-06,
+      "loss": 1.1495,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8613477924089853,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.158353201910279e-06,
+      "loss": 1.0023,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8615476874640814,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.157536186093079e-06,
+      "loss": 1.037,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8617475825191774,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.156719030018456e-06,
+      "loss": 1.0871,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8619474775742735,
+      "grad_norm": 2.0,
+      "learning_rate": 8.155901733722714e-06,
+      "loss": 1.0027,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8621473726293696,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.155084297242152e-06,
+      "loss": 1.0906,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8623472676844657,
+      "grad_norm": 2.125,
+      "learning_rate": 8.154266720613086e-06,
+      "loss": 1.0564,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8625471627395618,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.153449003871828e-06,
+      "loss": 1.1208,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8627470577946578,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.1526311470547e-06,
+      "loss": 0.9666,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8629469528497539,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.151813150198033e-06,
+      "loss": 1.0289,
+      "step": 4317
+    },
+    {
+      "epoch": 0.86314684790485,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.150995013338165e-06,
+      "loss": 1.0856,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8633467429599461,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.150176736511432e-06,
+      "loss": 1.0077,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8635466380150421,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.149358319754188e-06,
+      "loss": 1.0339,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8637465330701382,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.148539763102782e-06,
+      "loss": 1.0569,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8639464281252343,
+      "grad_norm": 2.125,
+      "learning_rate": 8.147721066593577e-06,
+      "loss": 1.0126,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8641463231803304,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.146902230262936e-06,
+      "loss": 1.0008,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8643462182354265,
+      "grad_norm": 2.125,
+      "learning_rate": 8.146083254147237e-06,
+      "loss": 1.0532,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8645461132905224,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.145264138282853e-06,
+      "loss": 1.0592,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8647460083456185,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.144444882706175e-06,
+      "loss": 0.9666,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8649459034007146,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.14362548745359e-06,
+      "loss": 1.0502,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8651457984558107,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.142805952561495e-06,
+      "loss": 1.0719,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8653456935109067,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.141986278066296e-06,
+      "loss": 1.0825,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8655455885660028,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.141166464004404e-06,
+      "loss": 1.0434,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8657454836210989,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.140346510412232e-06,
+      "loss": 1.0814,
+      "step": 4331
+    },
+    {
+      "epoch": 0.865945378676195,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.139526417326203e-06,
+      "loss": 0.9666,
+      "step": 4332
+    },
+    {
+      "epoch": 0.866145273731291,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.138706184782745e-06,
+      "loss": 1.0682,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8663451687863871,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.137885812818296e-06,
+      "loss": 1.0596,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8665450638414832,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.137065301469292e-06,
+      "loss": 0.9758,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8667449588965793,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.136244650772183e-06,
+      "loss": 1.0406,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8669448539516754,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.135423860763422e-06,
+      "loss": 1.0339,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8671447490067714,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.134602931479468e-06,
+      "loss": 0.9688,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8673446440618675,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.133781862956787e-06,
+      "loss": 0.9505,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8675445391169636,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.132960655231849e-06,
+      "loss": 1.0011,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8677444341720597,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.132139308341134e-06,
+      "loss": 0.988,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8679443292271557,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.131317822321125e-06,
+      "loss": 1.0682,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8681442242822518,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.130496197208313e-06,
+      "loss": 1.0255,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8683441193373479,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.129674433039196e-06,
+      "loss": 1.0647,
+      "step": 4344
+    },
+    {
+      "epoch": 0.868544014392444,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.128852529850272e-06,
+      "loss": 0.9956,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8687439094475401,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.128030487678055e-06,
+      "loss": 1.0578,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8689438045026361,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.127208306559058e-06,
+      "loss": 1.0504,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8691436995577322,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.126385986529802e-06,
+      "loss": 1.038,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8693435946128283,
+      "grad_norm": 2.125,
+      "learning_rate": 8.125563527626812e-06,
+      "loss": 0.996,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8695434896679244,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.124740929886625e-06,
+      "loss": 1.0464,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8697433847230204,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.12391819334578e-06,
+      "loss": 0.9245,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8699432797781165,
+      "grad_norm": 2.125,
+      "learning_rate": 8.123095318040824e-06,
+      "loss": 1.0916,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8701431748332126,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.122272304008306e-06,
+      "loss": 1.0831,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8703430698883087,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.121449151284784e-06,
+      "loss": 0.9706,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8705429649434047,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.120625859906825e-06,
+      "loss": 1.0293,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8707428599985008,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.119802429911002e-06,
+      "loss": 0.9694,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8709427550535969,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.118978861333883e-06,
+      "loss": 0.9881,
+      "step": 4357
+    },
+    {
+      "epoch": 0.871142650108693,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.11815515421206e-06,
+      "loss": 1.0694,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8713425451637891,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.117331308582116e-06,
+      "loss": 0.9553,
+      "step": 4359
+    },
+    {
+      "epoch": 0.871542440218885,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.116507324480651e-06,
+      "loss": 1.0958,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8717423352739812,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.115683201944262e-06,
+      "loss": 1.1264,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8719422303290773,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.114858941009556e-06,
+      "loss": 0.9975,
+      "step": 4362
+    },
+    {
+      "epoch": 0.8721421253841734,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.114034541713152e-06,
+      "loss": 1.0681,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8723420204392693,
+      "grad_norm": 2.25,
+      "learning_rate": 8.113210004091663e-06,
+      "loss": 1.083,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8725419154943654,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.11238532818172e-06,
+      "loss": 1.044,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8727418105494615,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.111560514019951e-06,
+      "loss": 0.9365,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8729417056045576,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.110735561643e-06,
+      "loss": 1.128,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8731416006596536,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.109910471087505e-06,
+      "loss": 1.0742,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8733414957147497,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.109085242390118e-06,
+      "loss": 1.0137,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8735413907698458,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.108259875587498e-06,
+      "loss": 1.023,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8737412858249419,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.107434370716307e-06,
+      "loss": 1.027,
+      "step": 4371
+    },
+    {
+      "epoch": 0.873941180880038,
+      "grad_norm": 2.0,
+      "learning_rate": 8.106608727813212e-06,
+      "loss": 0.9632,
+      "step": 4372
+    },
+    {
+      "epoch": 0.874141075935134,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.105782946914891e-06,
+      "loss": 1.0354,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8743409709902301,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.10495702805802e-06,
+      "loss": 0.9849,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8745408660453262,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.104130971279292e-06,
+      "loss": 1.0414,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8747407611004223,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.103304776615399e-06,
+      "loss": 1.1185,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8749406561555183,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.102478444103037e-06,
+      "loss": 1.1865,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8751405512106144,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.101651973778914e-06,
+      "loss": 1.025,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8753404462657105,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.100825365679741e-06,
+      "loss": 0.9985,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8755403413208066,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.099998619842238e-06,
+      "loss": 1.0752,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8757402363759027,
+      "grad_norm": 2.125,
+      "learning_rate": 8.099171736303126e-06,
+      "loss": 0.985,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8759401314309987,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.098344715099136e-06,
+      "loss": 1.0516,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8761400264860948,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.097517556267007e-06,
+      "loss": 1.0916,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8763399215411909,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.096690259843478e-06,
+      "loss": 0.8999,
+      "step": 4384
+    },
+    {
+      "epoch": 0.876539816596287,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.095862825865297e-06,
+      "loss": 1.0837,
+      "step": 4385
+    },
+    {
+      "epoch": 0.876739711651383,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.09503525436922e-06,
+      "loss": 0.9613,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8769396067064791,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.09420754539201e-06,
+      "loss": 1.0288,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8771395017615752,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.09337969897043e-06,
+      "loss": 0.987,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8773393968166713,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.092551715141254e-06,
+      "loss": 1.0139,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8775392918717673,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.091723593941261e-06,
+      "loss": 1.0515,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8777391869268634,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.090895335407238e-06,
+      "loss": 1.0595,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8779390819819595,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.090066939575972e-06,
+      "loss": 1.1062,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8781389770370556,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.089238406484263e-06,
+      "loss": 0.975,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8783388720921517,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.088409736168915e-06,
+      "loss": 0.9842,
+      "step": 4394
+    },
+    {
+      "epoch": 0.8785387671472477,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.087580928666736e-06,
+      "loss": 1.084,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8787386622023438,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.08675198401454e-06,
+      "loss": 0.9676,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8789385572574399,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.085922902249153e-06,
+      "loss": 0.9899,
+      "step": 4397
+    },
+    {
+      "epoch": 0.879138452312536,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.085093683407399e-06,
+      "loss": 1.0496,
+      "step": 4398
+    },
+    {
+      "epoch": 0.879338347367632,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.084264327526112e-06,
+      "loss": 1.014,
+      "step": 4399
+    },
+    {
+      "epoch": 0.879538242422728,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.083434834642133e-06,
+      "loss": 1.085,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8797381374778241,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.08260520479231e-06,
+      "loss": 1.0846,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8799380325329202,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.081775438013493e-06,
+      "loss": 1.0657,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8801379275880163,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.08094553434254e-06,
+      "loss": 0.9896,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8803378226431123,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.080115493816314e-06,
+      "loss": 1.0088,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8805377176982084,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.079285316471688e-06,
+      "loss": 1.0624,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8807376127533045,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.078455002345538e-06,
+      "loss": 0.9994,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8809375078084006,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.077624551474744e-06,
+      "loss": 1.0085,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8811374028634966,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.076793963896197e-06,
+      "loss": 0.9928,
+      "step": 4408
+    },
+    {
+      "epoch": 0.8813372979185927,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.07596323964679e-06,
+      "loss": 1.0614,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8815371929736888,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.075132378763424e-06,
+      "loss": 1.0672,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8817370880287849,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.074301381283007e-06,
+      "loss": 0.9155,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8819369830838809,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.073470247242452e-06,
+      "loss": 0.9595,
+      "step": 4412
+    },
+    {
+      "epoch": 0.882136878138977,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.072638976678675e-06,
+      "loss": 0.9704,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8823367731940731,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.071807569628602e-06,
+      "loss": 1.0088,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8825366682491692,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.070976026129166e-06,
+      "loss": 0.9518,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8827365633042653,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.070144346217305e-06,
+      "loss": 1.0652,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8829364583593613,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.069312529929958e-06,
+      "loss": 0.9716,
+      "step": 4417
+    },
+    {
+      "epoch": 0.8831363534144574,
+      "grad_norm": 2.125,
+      "learning_rate": 8.068480577304076e-06,
+      "loss": 1.0798,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8833362484695535,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.067648488376616e-06,
+      "loss": 0.9723,
+      "step": 4419
+    },
+    {
+      "epoch": 0.8835361435246496,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.066816263184535e-06,
+      "loss": 0.9742,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8837360385797456,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.065983901764807e-06,
+      "loss": 0.9657,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8839359336348417,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.0651514041544e-06,
+      "loss": 1.02,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8841358286899378,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.064318770390293e-06,
+      "loss": 1.0419,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8843357237450339,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.063486000509475e-06,
+      "loss": 1.085,
+      "step": 4424
+    },
+    {
+      "epoch": 0.88453561880013,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.062653094548936e-06,
+      "loss": 1.083,
+      "step": 4425
+    },
+    {
+      "epoch": 0.884735513855226,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.061820052545675e-06,
+      "loss": 1.1178,
+      "step": 4426
+    },
+    {
+      "epoch": 0.8849354089103221,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.060986874536691e-06,
+      "loss": 1.0527,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8851353039654182,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.060153560559e-06,
+      "loss": 1.0529,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8853351990205143,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.059320110649614e-06,
+      "loss": 0.948,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8855350940756103,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.058486524845554e-06,
+      "loss": 1.0805,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8857349891307064,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.05765280318385e-06,
+      "loss": 1.0965,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8859348841858025,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.056818945701537e-06,
+      "loss": 1.0667,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8861347792408986,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.05598495243565e-06,
+      "loss": 1.0619,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8863346742959946,
+      "grad_norm": 2.0,
+      "learning_rate": 8.055150823423239e-06,
+      "loss": 0.8879,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8865345693510907,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.054316558701355e-06,
+      "loss": 1.0322,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8867344644061868,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.053482158307055e-06,
+      "loss": 1.1139,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8869343594612829,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.052647622277405e-06,
+      "loss": 1.0347,
+      "step": 4437
+    },
+    {
+      "epoch": 0.887134254516379,
+      "grad_norm": 2.125,
+      "learning_rate": 8.051812950649474e-06,
+      "loss": 1.0791,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8873341495714749,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.050978143460335e-06,
+      "loss": 0.9837,
+      "step": 4439
+    },
+    {
+      "epoch": 0.887534044626571,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.050143200747073e-06,
+      "loss": 0.9943,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8877339396816671,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.049308122546776e-06,
+      "loss": 1.16,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8879338347367632,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.04847290889654e-06,
+      "loss": 1.0046,
+      "step": 4442
+    },
+    {
+      "epoch": 0.8881337297918592,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.047637559833464e-06,
+      "loss": 1.0639,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8883336248469553,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.04680207539465e-06,
+      "loss": 1.0163,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8885335199020514,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.045966455617214e-06,
+      "loss": 1.0285,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8887334149571475,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.045130700538273e-06,
+      "loss": 1.0341,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8889333100122436,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.044294810194953e-06,
+      "loss": 0.9628,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8891332050673396,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.04345878462438e-06,
+      "loss": 1.0712,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8893331001224357,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.042622623863694e-06,
+      "loss": 1.027,
+      "step": 4449
+    },
+    {
+      "epoch": 0.8895329951775318,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.041786327950037e-06,
+      "loss": 1.0139,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8897328902326279,
+      "grad_norm": 2.125,
+      "learning_rate": 8.040949896920556e-06,
+      "loss": 1.0845,
+      "step": 4451
+    },
+    {
+      "epoch": 0.8899327852877239,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.040113330812404e-06,
+      "loss": 1.0785,
+      "step": 4452
+    },
+    {
+      "epoch": 0.89013268034282,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.039276629662745e-06,
+      "loss": 1.0703,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8903325753979161,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.038439793508741e-06,
+      "loss": 1.0732,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8905324704530122,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.037602822387566e-06,
+      "loss": 1.0146,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8907323655081082,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.036765716336399e-06,
+      "loss": 0.9116,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8909322605632043,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.035928475392422e-06,
+      "loss": 0.9984,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8911321556183004,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.035091099592827e-06,
+      "loss": 1.0102,
+      "step": 4458
+    },
+    {
+      "epoch": 0.8913320506733965,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.034253588974809e-06,
+      "loss": 0.8845,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8915319457284926,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.033415943575572e-06,
+      "loss": 0.9525,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8917318407835886,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.03257816343232e-06,
+      "loss": 1.005,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8919317358386847,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.031740248582272e-06,
+      "loss": 1.0035,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8921316308937808,
+      "grad_norm": 2.125,
+      "learning_rate": 8.030902199062646e-06,
+      "loss": 1.0726,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8923315259488769,
+      "grad_norm": 2.0,
+      "learning_rate": 8.030064014910668e-06,
+      "loss": 1.0157,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8925314210039729,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.02922569616357e-06,
+      "loss": 1.0501,
+      "step": 4465
+    },
+    {
+      "epoch": 0.892731316059069,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.028387242858588e-06,
+      "loss": 0.9781,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8929312111141651,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.02754865503297e-06,
+      "loss": 1.0365,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8931311061692612,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.026709932723964e-06,
+      "loss": 0.9437,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8933310012243572,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.025871075968828e-06,
+      "loss": 0.9995,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8935308962794533,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.02503208480482e-06,
+      "loss": 1.0265,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8937307913345494,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.024192959269209e-06,
+      "loss": 1.1485,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8939306863896455,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.02335369939927e-06,
+      "loss": 1.0292,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8941305814447416,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.022514305232283e-06,
+      "loss": 0.9181,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8943304764998375,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.021674776805534e-06,
+      "loss": 1.0503,
+      "step": 4474
+    },
+    {
+      "epoch": 0.8945303715549336,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.020835114156313e-06,
+      "loss": 1.0704,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8947302666100297,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.019995317321921e-06,
+      "loss": 1.0324,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8949301616651258,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.019155386339657e-06,
+      "loss": 0.9435,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8951300567202218,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.018315321246834e-06,
+      "loss": 1.0742,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8953299517753179,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.017475122080767e-06,
+      "loss": 1.0427,
+      "step": 4479
+    },
+    {
+      "epoch": 0.895529846830414,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.016634788878779e-06,
+      "loss": 1.0984,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8957297418855101,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.015794321678194e-06,
+      "loss": 0.9826,
+      "step": 4481
+    },
+    {
+      "epoch": 0.8959296369406062,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.014953720516347e-06,
+      "loss": 1.0419,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8961295319957022,
+      "grad_norm": 2.0,
+      "learning_rate": 8.014112985430578e-06,
+      "loss": 0.9784,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8963294270507983,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.013272116458233e-06,
+      "loss": 1.0174,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8965293221058944,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.012431113636662e-06,
+      "loss": 1.0536,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8967292171609905,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.011589977003222e-06,
+      "loss": 1.0687,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8969291122160865,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.01074870659528e-06,
+      "loss": 1.0259,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8971290072711826,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.009907302450199e-06,
+      "loss": 1.0712,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8973289023262787,
+      "grad_norm": 2.0,
+      "learning_rate": 8.009065764605358e-06,
+      "loss": 1.0412,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8975287973813748,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.008224093098136e-06,
+      "loss": 1.1454,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8977286924364708,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.007382287965921e-06,
+      "loss": 1.054,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8979285874915669,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.006540349246107e-06,
+      "loss": 0.8789,
+      "step": 4492
+    },
+    {
+      "epoch": 0.898128482546663,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.005698276976092e-06,
+      "loss": 1.0072,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8983283776017591,
+      "grad_norm": 2.125,
+      "learning_rate": 8.004856071193278e-06,
+      "loss": 1.0211,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8985282726568552,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.004013731935082e-06,
+      "loss": 1.0182,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8987281677119512,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.003171259238915e-06,
+      "loss": 1.0249,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8989280627670473,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.002328653142203e-06,
+      "loss": 1.0605,
+      "step": 4497
+    },
+    {
+      "epoch": 0.8991279578221434,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.00148591368237e-06,
+      "loss": 0.9666,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8993278528772395,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.000643040896855e-06,
+      "loss": 1.0767,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8995277479323355,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.999800034823097e-06,
+      "loss": 1.0541,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8995277479323355,
+      "eval_loss": 0.9116131067276001,
+      "eval_runtime": 594.896,
+      "eval_samples_per_second": 3.594,
+      "eval_steps_per_second": 3.594,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8997276429874316,
+      "grad_norm": 2.125,
+      "learning_rate": 7.998956895498542e-06,
+      "loss": 1.1048,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8999275380425277,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.99811362296064e-06,
+      "loss": 0.9938,
+      "step": 4502
+    },
+    {
+      "epoch": 0.9001274330976238,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.997270217246853e-06,
+      "loss": 1.1029,
+      "step": 4503
+    },
+    {
+      "epoch": 0.9003273281527199,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.996426678394642e-06,
+      "loss": 1.1396,
+      "step": 4504
+    },
+    {
+      "epoch": 0.9005272232078159,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.99558300644148e-06,
+      "loss": 0.9531,
+      "step": 4505
+    },
+    {
+      "epoch": 0.900727118262912,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.994739201424836e-06,
+      "loss": 1.0054,
+      "step": 4506
+    },
+    {
+      "epoch": 0.9009270133180081,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.993895263382201e-06,
+      "loss": 1.0793,
+      "step": 4507
+    },
+    {
+      "epoch": 0.9011269083731042,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.993051192351056e-06,
+      "loss": 1.0196,
+      "step": 4508
+    },
+    {
+      "epoch": 0.9013268034282002,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.992206988368898e-06,
+      "loss": 0.954,
+      "step": 4509
+    },
+    {
+      "epoch": 0.9015266984832963,
+      "grad_norm": 2.0,
+      "learning_rate": 7.991362651473225e-06,
+      "loss": 1.0146,
+      "step": 4510
+    },
+    {
+      "epoch": 0.9017265935383924,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.990518181701542e-06,
+      "loss": 1.0859,
+      "step": 4511
+    },
+    {
+      "epoch": 0.9019264885934885,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.989673579091361e-06,
+      "loss": 1.0375,
+      "step": 4512
+    },
+    {
+      "epoch": 0.9021263836485844,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.988828843680198e-06,
+      "loss": 1.0987,
+      "step": 4513
+    },
+    {
+      "epoch": 0.9023262787036805,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.987983975505579e-06,
+      "loss": 0.984,
+      "step": 4514
+    },
+    {
+      "epoch": 0.9025261737587766,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.98713897460503e-06,
+      "loss": 1.049,
+      "step": 4515
+    },
+    {
+      "epoch": 0.9027260688138727,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.986293841016087e-06,
+      "loss": 1.0108,
+      "step": 4516
+    },
+    {
+      "epoch": 0.9029259638689688,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.98544857477629e-06,
+      "loss": 1.0748,
+      "step": 4517
+    },
+    {
+      "epoch": 0.9031258589240648,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.984603175923186e-06,
+      "loss": 1.043,
+      "step": 4518
+    },
+    {
+      "epoch": 0.9033257539791609,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.983757644494327e-06,
+      "loss": 1.0418,
+      "step": 4519
+    },
+    {
+      "epoch": 0.903525649034257,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.982911980527276e-06,
+      "loss": 1.1628,
+      "step": 4520
+    },
+    {
+      "epoch": 0.9037255440893531,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.98206618405959e-06,
+      "loss": 0.9724,
+      "step": 4521
+    },
+    {
+      "epoch": 0.9039254391444491,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.981220255128842e-06,
+      "loss": 1.0297,
+      "step": 4522
+    },
+    {
+      "epoch": 0.9041253341995452,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.98037419377261e-06,
+      "loss": 0.9934,
+      "step": 4523
+    },
+    {
+      "epoch": 0.9043252292546413,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.979528000028474e-06,
+      "loss": 1.0693,
+      "step": 4524
+    },
+    {
+      "epoch": 0.9045251243097374,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.978681673934023e-06,
+      "loss": 1.015,
+      "step": 4525
+    },
+    {
+      "epoch": 0.9047250193648335,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.97783521552685e-06,
+      "loss": 0.9598,
+      "step": 4526
+    },
+    {
+      "epoch": 0.9049249144199295,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.976988624844556e-06,
+      "loss": 1.0851,
+      "step": 4527
+    },
+    {
+      "epoch": 0.9051248094750256,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.976141901924743e-06,
+      "loss": 0.9594,
+      "step": 4528
+    },
+    {
+      "epoch": 0.9053247045301217,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.975295046805026e-06,
+      "loss": 0.9813,
+      "step": 4529
+    },
+    {
+      "epoch": 0.9055245995852178,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.974448059523018e-06,
+      "loss": 0.9971,
+      "step": 4530
+    },
+    {
+      "epoch": 0.9057244946403138,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.97360094011635e-06,
+      "loss": 1.1101,
+      "step": 4531
+    },
+    {
+      "epoch": 0.9059243896954099,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.972753688622644e-06,
+      "loss": 1.0363,
+      "step": 4532
+    },
+    {
+      "epoch": 0.906124284750506,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.971906305079535e-06,
+      "loss": 1.0802,
+      "step": 4533
+    },
+    {
+      "epoch": 0.9063241798056021,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.971058789524666e-06,
+      "loss": 1.0146,
+      "step": 4534
+    },
+    {
+      "epoch": 0.9065240748606981,
+      "grad_norm": 2.0,
+      "learning_rate": 7.970211141995682e-06,
+      "loss": 0.9591,
+      "step": 4535
+    },
+    {
+      "epoch": 0.9067239699157942,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.969363362530238e-06,
+      "loss": 1.0927,
+      "step": 4536
+    },
+    {
+      "epoch": 0.9069238649708903,
+      "grad_norm": 2.0,
+      "learning_rate": 7.96851545116599e-06,
+      "loss": 0.9781,
+      "step": 4537
+    },
+    {
+      "epoch": 0.9071237600259864,
+      "grad_norm": 2.0,
+      "learning_rate": 7.967667407940603e-06,
+      "loss": 0.9595,
+      "step": 4538
+    },
+    {
+      "epoch": 0.9073236550810825,
+      "grad_norm": 2.125,
+      "learning_rate": 7.966819232891744e-06,
+      "loss": 1.0559,
+      "step": 4539
+    },
+    {
+      "epoch": 0.9075235501361785,
+      "grad_norm": 2.125,
+      "learning_rate": 7.965970926057095e-06,
+      "loss": 1.0765,
+      "step": 4540
+    },
+    {
+      "epoch": 0.9077234451912746,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.965122487474333e-06,
+      "loss": 1.1197,
+      "step": 4541
+    },
+    {
+      "epoch": 0.9079233402463707,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.964273917181147e-06,
+      "loss": 1.0798,
+      "step": 4542
+    },
+    {
+      "epoch": 0.9081232353014668,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.96342521521523e-06,
+      "loss": 0.9719,
+      "step": 4543
+    },
+    {
+      "epoch": 0.9083231303565628,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.962576381614282e-06,
+      "loss": 1.0806,
+      "step": 4544
+    },
+    {
+      "epoch": 0.9085230254116589,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.961727416416007e-06,
+      "loss": 1.099,
+      "step": 4545
+    },
+    {
+      "epoch": 0.908722920466755,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.960878319658117e-06,
+      "loss": 1.1292,
+      "step": 4546
+    },
+    {
+      "epoch": 0.9089228155218511,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.960029091378327e-06,
+      "loss": 0.9665,
+      "step": 4547
+    },
+    {
+      "epoch": 0.9091227105769472,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.959179731614363e-06,
+      "loss": 1.1101,
+      "step": 4548
+    },
+    {
+      "epoch": 0.9093226056320431,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.95833024040395e-06,
+      "loss": 1.0884,
+      "step": 4549
+    },
+    {
+      "epoch": 0.9095225006871392,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.957480617784823e-06,
+      "loss": 0.9976,
+      "step": 4550
+    },
+    {
+      "epoch": 0.9097223957422353,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.956630863794723e-06,
+      "loss": 1.0861,
+      "step": 4551
+    },
+    {
+      "epoch": 0.9099222907973314,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.955780978471396e-06,
+      "loss": 1.0039,
+      "step": 4552
+    },
+    {
+      "epoch": 0.9101221858524274,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.954930961852594e-06,
+      "loss": 1.0688,
+      "step": 4553
+    },
+    {
+      "epoch": 0.9103220809075235,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.954080813976075e-06,
+      "loss": 1.0032,
+      "step": 4554
+    },
+    {
+      "epoch": 0.9105219759626196,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.953230534879601e-06,
+      "loss": 1.0197,
+      "step": 4555
+    },
+    {
+      "epoch": 0.9107218710177157,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.952380124600943e-06,
+      "loss": 0.9938,
+      "step": 4556
+    },
+    {
+      "epoch": 0.9109217660728117,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.951529583177874e-06,
+      "loss": 1.0392,
+      "step": 4557
+    },
+    {
+      "epoch": 0.9111216611279078,
+      "grad_norm": 2.0,
+      "learning_rate": 7.950678910648176e-06,
+      "loss": 1.0021,
+      "step": 4558
+    },
+    {
+      "epoch": 0.9113215561830039,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.949828107049638e-06,
+      "loss": 1.0592,
+      "step": 4559
+    },
+    {
+      "epoch": 0.9115214512381,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.948977172420046e-06,
+      "loss": 0.975,
+      "step": 4560
+    },
+    {
+      "epoch": 0.9117213462931961,
+      "grad_norm": 2.0,
+      "learning_rate": 7.948126106797208e-06,
+      "loss": 1.0879,
+      "step": 4561
+    },
+    {
+      "epoch": 0.9119212413482921,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.94727491021892e-06,
+      "loss": 0.9099,
+      "step": 4562
+    },
+    {
+      "epoch": 0.9121211364033882,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.946423582722998e-06,
+      "loss": 1.1033,
+      "step": 4563
+    },
+    {
+      "epoch": 0.9123210314584843,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.945572124347253e-06,
+      "loss": 0.9591,
+      "step": 4564
+    },
+    {
+      "epoch": 0.9125209265135804,
+      "grad_norm": 2.0,
+      "learning_rate": 7.944720535129509e-06,
+      "loss": 1.0181,
+      "step": 4565
+    },
+    {
+      "epoch": 0.9127208215686764,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.943868815107594e-06,
+      "loss": 1.0219,
+      "step": 4566
+    },
+    {
+      "epoch": 0.9129207166237725,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.94301696431934e-06,
+      "loss": 1.0529,
+      "step": 4567
+    },
+    {
+      "epoch": 0.9131206116788686,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.942164982802588e-06,
+      "loss": 1.0201,
+      "step": 4568
+    },
+    {
+      "epoch": 0.9133205067339647,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.941312870595179e-06,
+      "loss": 0.9623,
+      "step": 4569
+    },
+    {
+      "epoch": 0.9135204017890608,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.940460627734969e-06,
+      "loss": 1.0634,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9137202968441568,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.939608254259812e-06,
+      "loss": 1.0753,
+      "step": 4571
+    },
+    {
+      "epoch": 0.9139201918992529,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.938755750207569e-06,
+      "loss": 1.0382,
+      "step": 4572
+    },
+    {
+      "epoch": 0.914120086954349,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.93790311561611e-06,
+      "loss": 1.0883,
+      "step": 4573
+    },
+    {
+      "epoch": 0.9143199820094451,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.937050350523308e-06,
+      "loss": 1.0178,
+      "step": 4574
+    },
+    {
+      "epoch": 0.9145198770645411,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.936197454967043e-06,
+      "loss": 0.9342,
+      "step": 4575
+    },
+    {
+      "epoch": 0.9147197721196372,
+      "grad_norm": 2.125,
+      "learning_rate": 7.935344428985202e-06,
+      "loss": 1.1515,
+      "step": 4576
+    },
+    {
+      "epoch": 0.9149196671747333,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.934491272615674e-06,
+      "loss": 1.0595,
+      "step": 4577
+    },
+    {
+      "epoch": 0.9151195622298294,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.933637985896356e-06,
+      "loss": 1.0235,
+      "step": 4578
+    },
+    {
+      "epoch": 0.9153194572849254,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.932784568865155e-06,
+      "loss": 0.9725,
+      "step": 4579
+    },
+    {
+      "epoch": 0.9155193523400215,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.931931021559973e-06,
+      "loss": 1.1227,
+      "step": 4580
+    },
+    {
+      "epoch": 0.9157192473951176,
+      "grad_norm": 2.0,
+      "learning_rate": 7.931077344018731e-06,
+      "loss": 1.0332,
+      "step": 4581
+    },
+    {
+      "epoch": 0.9159191424502137,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.930223536279344e-06,
+      "loss": 1.025,
+      "step": 4582
+    },
+    {
+      "epoch": 0.9161190375053098,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.929369598379743e-06,
+      "loss": 1.0924,
+      "step": 4583
+    },
+    {
+      "epoch": 0.9163189325604058,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.928515530357857e-06,
+      "loss": 1.0261,
+      "step": 4584
+    },
+    {
+      "epoch": 0.9165188276155019,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.927661332251622e-06,
+      "loss": 1.0325,
+      "step": 4585
+    },
+    {
+      "epoch": 0.916718722670598,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.926807004098985e-06,
+      "loss": 0.9855,
+      "step": 4586
+    },
+    {
+      "epoch": 0.916918617725694,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.925952545937892e-06,
+      "loss": 1.0322,
+      "step": 4587
+    },
+    {
+      "epoch": 0.91711851278079,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.9250979578063e-06,
+      "loss": 0.9784,
+      "step": 4588
+    },
+    {
+      "epoch": 0.9173184078358861,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.924243239742171e-06,
+      "loss": 1.0395,
+      "step": 4589
+    },
+    {
+      "epoch": 0.9175183028909822,
+      "grad_norm": 2.125,
+      "learning_rate": 7.923388391783467e-06,
+      "loss": 1.1187,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9177181979460783,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.922533413968164e-06,
+      "loss": 0.9561,
+      "step": 4591
+    },
+    {
+      "epoch": 0.9179180930011743,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.92167830633424e-06,
+      "loss": 1.0588,
+      "step": 4592
+    },
+    {
+      "epoch": 0.9181179880562704,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.920823068919676e-06,
+      "loss": 0.9722,
+      "step": 4593
+    },
+    {
+      "epoch": 0.9183178831113665,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.919967701762464e-06,
+      "loss": 1.0818,
+      "step": 4594
+    },
+    {
+      "epoch": 0.9185177781664626,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.919112204900597e-06,
+      "loss": 0.9887,
+      "step": 4595
+    },
+    {
+      "epoch": 0.9187176732215587,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.918256578372079e-06,
+      "loss": 1.0106,
+      "step": 4596
+    },
+    {
+      "epoch": 0.9189175682766547,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.917400822214916e-06,
+      "loss": 0.9737,
+      "step": 4597
+    },
+    {
+      "epoch": 0.9191174633317508,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.916544936467119e-06,
+      "loss": 1.1294,
+      "step": 4598
+    },
+    {
+      "epoch": 0.9193173583868469,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.915688921166709e-06,
+      "loss": 1.0798,
+      "step": 4599
+    },
+    {
+      "epoch": 0.919517253441943,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.914832776351707e-06,
+      "loss": 1.0665,
+      "step": 4600
+    },
+    {
+      "epoch": 0.919717148497039,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.913976502060143e-06,
+      "loss": 0.9593,
+      "step": 4601
+    },
+    {
+      "epoch": 0.9199170435521351,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.913120098330056e-06,
+      "loss": 1.0382,
+      "step": 4602
+    },
+    {
+      "epoch": 0.9201169386072312,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.912263565199486e-06,
+      "loss": 0.9877,
+      "step": 4603
+    },
+    {
+      "epoch": 0.9203168336623273,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.911406902706478e-06,
+      "loss": 0.9745,
+      "step": 4604
+    },
+    {
+      "epoch": 0.9205167287174234,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.910550110889086e-06,
+      "loss": 0.9713,
+      "step": 4605
+    },
+    {
+      "epoch": 0.9207166237725194,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.909693189785371e-06,
+      "loss": 1.0932,
+      "step": 4606
+    },
+    {
+      "epoch": 0.9209165188276155,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.908836139433393e-06,
+      "loss": 1.0144,
+      "step": 4607
+    },
+    {
+      "epoch": 0.9211164138827116,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.907978959871228e-06,
+      "loss": 0.9774,
+      "step": 4608
+    },
+    {
+      "epoch": 0.9213163089378077,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.907121651136944e-06,
+      "loss": 1.0085,
+      "step": 4609
+    },
+    {
+      "epoch": 0.9215162039929037,
+      "grad_norm": 2.125,
+      "learning_rate": 7.90626421326863e-06,
+      "loss": 1.0132,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9217160990479998,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.905406646304367e-06,
+      "loss": 1.053,
+      "step": 4611
+    },
+    {
+      "epoch": 0.9219159941030959,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.904548950282254e-06,
+      "loss": 1.0387,
+      "step": 4612
+    },
+    {
+      "epoch": 0.922115889158192,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.903691125240385e-06,
+      "loss": 1.0253,
+      "step": 4613
+    },
+    {
+      "epoch": 0.922315784213288,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.902833171216867e-06,
+      "loss": 1.0571,
+      "step": 4614
+    },
+    {
+      "epoch": 0.9225156792683841,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.901975088249808e-06,
+      "loss": 1.0589,
+      "step": 4615
+    },
+    {
+      "epoch": 0.9227155743234802,
+      "grad_norm": 2.125,
+      "learning_rate": 7.901116876377326e-06,
+      "loss": 1.0856,
+      "step": 4616
+    },
+    {
+      "epoch": 0.9229154693785763,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.900258535637544e-06,
+      "loss": 1.1088,
+      "step": 4617
+    },
+    {
+      "epoch": 0.9231153644336724,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.899400066068588e-06,
+      "loss": 1.0745,
+      "step": 4618
+    },
+    {
+      "epoch": 0.9233152594887684,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.898541467708588e-06,
+      "loss": 0.9503,
+      "step": 4619
+    },
+    {
+      "epoch": 0.9235151545438645,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.897682740595686e-06,
+      "loss": 0.9778,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9237150495989606,
+      "grad_norm": 2.25,
+      "learning_rate": 7.896823884768028e-06,
+      "loss": 0.9761,
+      "step": 4621
+    },
+    {
+      "epoch": 0.9239149446540567,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.895964900263762e-06,
+      "loss": 1.0342,
+      "step": 4622
+    },
+    {
+      "epoch": 0.9241148397091526,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.895105787121045e-06,
+      "loss": 1.0463,
+      "step": 4623
+    },
+    {
+      "epoch": 0.9243147347642487,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.894246545378037e-06,
+      "loss": 1.0387,
+      "step": 4624
+    },
+    {
+      "epoch": 0.9245146298193448,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.893387175072907e-06,
+      "loss": 1.055,
+      "step": 4625
+    },
+    {
+      "epoch": 0.924714524874441,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.892527676243825e-06,
+      "loss": 0.9812,
+      "step": 4626
+    },
+    {
+      "epoch": 0.924914419929537,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.891668048928975e-06,
+      "loss": 1.0316,
+      "step": 4627
+    },
+    {
+      "epoch": 0.925114314984633,
+      "grad_norm": 2.0,
+      "learning_rate": 7.89080829316654e-06,
+      "loss": 0.9835,
+      "step": 4628
+    },
+    {
+      "epoch": 0.9253142100397291,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.889948408994707e-06,
+      "loss": 1.0214,
+      "step": 4629
+    },
+    {
+      "epoch": 0.9255141050948252,
+      "grad_norm": 2.0,
+      "learning_rate": 7.889088396451676e-06,
+      "loss": 1.0421,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9257140001499213,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.888228255575648e-06,
+      "loss": 0.9934,
+      "step": 4631
+    },
+    {
+      "epoch": 0.9259138952050173,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.887367986404827e-06,
+      "loss": 0.9687,
+      "step": 4632
+    },
+    {
+      "epoch": 0.9261137902601134,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.88650758897743e-06,
+      "loss": 1.1221,
+      "step": 4633
+    },
+    {
+      "epoch": 0.9263136853152095,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.885647063331674e-06,
+      "loss": 1.0123,
+      "step": 4634
+    },
+    {
+      "epoch": 0.9265135803703056,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.884786409505782e-06,
+      "loss": 0.9704,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9267134754254016,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.883925627537987e-06,
+      "loss": 1.0636,
+      "step": 4636
+    },
+    {
+      "epoch": 0.9269133704804977,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.883064717466524e-06,
+      "loss": 1.0541,
+      "step": 4637
+    },
+    {
+      "epoch": 0.9271132655355938,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.882203679329635e-06,
+      "loss": 1.0097,
+      "step": 4638
+    },
+    {
+      "epoch": 0.9273131605906899,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.881342513165567e-06,
+      "loss": 1.0571,
+      "step": 4639
+    },
+    {
+      "epoch": 0.927513055645786,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.88048121901257e-06,
+      "loss": 0.9967,
+      "step": 4640
+    },
+    {
+      "epoch": 0.927712950700882,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.879619796908905e-06,
+      "loss": 0.9924,
+      "step": 4641
+    },
+    {
+      "epoch": 0.9279128457559781,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.878758246892836e-06,
+      "loss": 0.9907,
+      "step": 4642
+    },
+    {
+      "epoch": 0.9281127408110742,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.877896569002634e-06,
+      "loss": 1.0302,
+      "step": 4643
+    },
+    {
+      "epoch": 0.9283126358661703,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.877034763276575e-06,
+      "loss": 1.0495,
+      "step": 4644
+    },
+    {
+      "epoch": 0.9285125309212663,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.876172829752937e-06,
+      "loss": 1.1101,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9287124259763624,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.87531076847001e-06,
+      "loss": 1.0882,
+      "step": 4646
+    },
+    {
+      "epoch": 0.9289123210314585,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.874448579466085e-06,
+      "loss": 1.0235,
+      "step": 4647
+    },
+    {
+      "epoch": 0.9291122160865546,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.873586262779462e-06,
+      "loss": 1.0154,
+      "step": 4648
+    },
+    {
+      "epoch": 0.9293121111416507,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.872723818448443e-06,
+      "loss": 1.083,
+      "step": 4649
+    },
+    {
+      "epoch": 0.9295120061967467,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.87186124651134e-06,
+      "loss": 1.0428,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9297119012518428,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.870998547006467e-06,
+      "loss": 1.0099,
+      "step": 4651
+    },
+    {
+      "epoch": 0.9299117963069389,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.870135719972146e-06,
+      "loss": 0.9716,
+      "step": 4652
+    },
+    {
+      "epoch": 0.930111691362035,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.869272765446701e-06,
+      "loss": 1.0331,
+      "step": 4653
+    },
+    {
+      "epoch": 0.930311586417131,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.868409683468466e-06,
+      "loss": 0.9574,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9305114814722271,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.867546474075782e-06,
+      "loss": 1.1178,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9307113765273232,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.866683137306987e-06,
+      "loss": 1.0142,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9309112715824193,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.865819673200435e-06,
+      "loss": 0.9947,
+      "step": 4657
+    },
+    {
+      "epoch": 0.9311111666375153,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.864956081794477e-06,
+      "loss": 0.9642,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9313110616926114,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.864092363127478e-06,
+      "loss": 1.0742,
+      "step": 4659
+    },
+    {
+      "epoch": 0.9315109567477075,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.8632285172378e-06,
+      "loss": 0.8985,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9317108518028036,
+      "grad_norm": 2.0,
+      "learning_rate": 7.86236454416382e-06,
+      "loss": 0.9271,
+      "step": 4661
+    },
+    {
+      "epoch": 0.9319107468578997,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.86150044394391e-06,
+      "loss": 1.0636,
+      "step": 4662
+    },
+    {
+      "epoch": 0.9321106419129956,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.860636216616458e-06,
+      "loss": 1.073,
+      "step": 4663
+    },
+    {
+      "epoch": 0.9323105369680917,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.85977186221985e-06,
+      "loss": 1.0081,
+      "step": 4664
+    },
+    {
+      "epoch": 0.9325104320231878,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.85890738079248e-06,
+      "loss": 1.0321,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9327103270782839,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.858042772372751e-06,
+      "loss": 1.0386,
+      "step": 4666
+    },
+    {
+      "epoch": 0.9329102221333799,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.857178036999066e-06,
+      "loss": 0.9717,
+      "step": 4667
+    },
+    {
+      "epoch": 0.933110117188476,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.85631317470984e-06,
+      "loss": 1.0604,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9333100122435721,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.855448185543486e-06,
+      "loss": 0.9417,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9335099072986682,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.854583069538431e-06,
+      "loss": 1.0199,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9337098023537643,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.853717826733098e-06,
+      "loss": 1.1401,
+      "step": 4671
+    },
+    {
+      "epoch": 0.9339096974088603,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.852852457165924e-06,
+      "loss": 1.0263,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9341095924639564,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.851986960875351e-06,
+      "loss": 0.9771,
+      "step": 4673
+    },
+    {
+      "epoch": 0.9343094875190525,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.851121337899819e-06,
+      "loss": 1.0479,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9345093825741486,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.850255588277784e-06,
+      "loss": 0.9217,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9347092776292446,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.8493897120477e-06,
+      "loss": 0.9912,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9349091726843407,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.848523709248026e-06,
+      "loss": 1.0977,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9351090677394368,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.847657579917237e-06,
+      "loss": 0.9783,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9353089627945329,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.8467913240938e-06,
+      "loss": 1.0672,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9355088578496289,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.845924941816198e-06,
+      "loss": 0.968,
+      "step": 4680
+    },
+    {
+      "epoch": 0.935708752904725,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.845058433122914e-06,
+      "loss": 1.0014,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9359086479598211,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.844191798052438e-06,
+      "loss": 1.1018,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9361085430149172,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.843325036643265e-06,
+      "loss": 1.003,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9363084380700133,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.842458148933898e-06,
+      "loss": 1.0554,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9365083331251093,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.841591134962845e-06,
+      "loss": 0.9854,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9367082281802054,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.840723994768616e-06,
+      "loss": 0.9764,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9369081232353015,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.83985672838973e-06,
+      "loss": 1.0718,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9371080182903976,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.838989335864714e-06,
+      "loss": 0.9564,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9373079133454936,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.838121817232093e-06,
+      "loss": 1.0227,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9375078084005897,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.837254172530404e-06,
+      "loss": 1.0139,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9377077034556858,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.836386401798188e-06,
+      "loss": 1.0004,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9379075985107819,
+      "grad_norm": 2.25,
+      "learning_rate": 7.83551850507399e-06,
+      "loss": 1.1292,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9381074935658779,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.834650482396364e-06,
+      "loss": 1.0614,
+      "step": 4693
+    },
+    {
+      "epoch": 0.938307388620974,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.833782333803865e-06,
+      "loss": 1.076,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9385072836760701,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.83291405933506e-06,
+      "loss": 0.9859,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9387071787311662,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.832045659028513e-06,
+      "loss": 1.0348,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9389070737862623,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.831177132922801e-06,
+      "loss": 1.0461,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9391069688413582,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.830308481056503e-06,
+      "loss": 1.0873,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9393068638964543,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.829439703468203e-06,
+      "loss": 0.9891,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9395067589515504,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.828570800196495e-06,
+      "loss": 0.9237,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9397066540066465,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.827701771279976e-06,
+      "loss": 1.0243,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9399065490617425,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.826832616757244e-06,
+      "loss": 1.0948,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9401064441168386,
+      "grad_norm": 2.125,
+      "learning_rate": 7.825963336666909e-06,
+      "loss": 1.0013,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9403063391719347,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.825093931047585e-06,
+      "loss": 1.0162,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9405062342270308,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.824224399937891e-06,
+      "loss": 0.9432,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9407061292821269,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.82335474337645e-06,
+      "loss": 1.0542,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9409060243372229,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.822484961401893e-06,
+      "loss": 1.0713,
+      "step": 4707
+    },
+    {
+      "epoch": 0.941105919392319,
+      "grad_norm": 2.125,
+      "learning_rate": 7.821615054052856e-06,
+      "loss": 0.9936,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9413058144474151,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.820745021367977e-06,
+      "loss": 0.9805,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9415057095025112,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.819874863385908e-06,
+      "loss": 1.0279,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9417056045576072,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.819004580145298e-06,
+      "loss": 1.1394,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9419054996127033,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.818134171684805e-06,
+      "loss": 0.972,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9421053946677994,
+      "grad_norm": 2.25,
+      "learning_rate": 7.817263638043096e-06,
+      "loss": 1.0544,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9423052897228955,
+      "grad_norm": 2.25,
+      "learning_rate": 7.816392979258834e-06,
+      "loss": 1.0741,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9425051847779915,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.815522195370697e-06,
+      "loss": 0.9426,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9427050798330876,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.814651286417367e-06,
+      "loss": 1.0312,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9429049748881837,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.813780252437526e-06,
+      "loss": 1.0476,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9431048699432798,
+      "grad_norm": 2.125,
+      "learning_rate": 7.812909093469868e-06,
+      "loss": 1.0199,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9433047649983759,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.812037809553086e-06,
+      "loss": 1.0714,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9435046600534719,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.811166400725884e-06,
+      "loss": 1.1525,
+      "step": 4720
+    },
+    {
+      "epoch": 0.943704555108568,
+      "grad_norm": 2.125,
+      "learning_rate": 7.810294867026974e-06,
+      "loss": 1.0328,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9439044501636641,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.809423208495064e-06,
+      "loss": 0.9166,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9441043452187602,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.808551425168878e-06,
+      "loss": 1.0082,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9443042402738562,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.807679517087135e-06,
+      "loss": 1.0474,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9445041353289523,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.806807484288567e-06,
+      "loss": 0.9546,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9447040303840484,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.805935326811913e-06,
+      "loss": 1.0596,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9449039254391445,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.805063044695909e-06,
+      "loss": 1.0443,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9451038204942406,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.804190637979305e-06,
+      "loss": 0.9845,
+      "step": 4728
+    },
+    {
+      "epoch": 0.9453037155493366,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.803318106700853e-06,
+      "loss": 1.0737,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9455036106044327,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.80244545089931e-06,
+      "loss": 1.0044,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9457035056595288,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.80157267061344e-06,
+      "loss": 1.0396,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9459034007146249,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.800699765882009e-06,
+      "loss": 1.052,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9461032957697209,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.799826736743796e-06,
+      "loss": 1.0078,
+      "step": 4733
+    },
+    {
+      "epoch": 0.946303190824817,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.798953583237578e-06,
+      "loss": 1.017,
+      "step": 4734
+    },
+    {
+      "epoch": 0.946503085879913,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.79808030540214e-06,
+      "loss": 1.0441,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9467029809350092,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.797206903276274e-06,
+      "loss": 0.9931,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9469028759901051,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.796333376898774e-06,
+      "loss": 0.9818,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9471027710452012,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.795459726308446e-06,
+      "loss": 1.0046,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9473026661002973,
+      "grad_norm": 2.0,
+      "learning_rate": 7.794585951544096e-06,
+      "loss": 1.0416,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9475025611553934,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.793712052644535e-06,
+      "loss": 1.0064,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9477024562104895,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.792838029648584e-06,
+      "loss": 0.9996,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9479023512655855,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.791963882595066e-06,
+      "loss": 1.101,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9481022463206816,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.791089611522811e-06,
+      "loss": 1.0511,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9483021413757777,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.790215216470654e-06,
+      "loss": 1.1489,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9485020364308738,
+      "grad_norm": 2.0,
+      "learning_rate": 7.789340697477432e-06,
+      "loss": 1.0727,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9487019314859698,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.788466054581997e-06,
+      "loss": 1.0585,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9489018265410659,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.787591287823197e-06,
+      "loss": 0.9812,
+      "step": 4747
+    },
+    {
+      "epoch": 0.949101721596162,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.78671639723989e-06,
+      "loss": 0.9217,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9493016166512581,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.785841382870938e-06,
+      "loss": 1.0322,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9495015117063542,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.78496624475521e-06,
+      "loss": 1.0459,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9497014067614502,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.784090982931577e-06,
+      "loss": 1.0993,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9499013018165463,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.78321559743892e-06,
+      "loss": 0.9609,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9501011968716424,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.782340088316125e-06,
+      "loss": 1.0295,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9503010919267385,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.78146445560208e-06,
+      "loss": 1.0306,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9505009869818345,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.78058869933568e-06,
+      "loss": 0.9982,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9507008820369306,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.77971281955583e-06,
+      "loss": 1.0818,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9509007770920267,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.778836816301429e-06,
+      "loss": 0.9594,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9511006721471228,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.777960689611396e-06,
+      "loss": 1.085,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9513005672022188,
+      "grad_norm": 2.0,
+      "learning_rate": 7.777084439524644e-06,
+      "loss": 1.0684,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9515004622573149,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.7762080660801e-06,
+      "loss": 1.0481,
+      "step": 4760
+    },
+    {
+      "epoch": 0.951700357312411,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.775331569316688e-06,
+      "loss": 0.9095,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9519002523675071,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.774454949273348e-06,
+      "loss": 0.9383,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9521001474226032,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.773578205989013e-06,
+      "loss": 0.9806,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9523000424776992,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.77270133950263e-06,
+      "loss": 1.0464,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9524999375327953,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.77182434985315e-06,
+      "loss": 1.2045,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9526998325878914,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.770947237079528e-06,
+      "loss": 1.0449,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9528997276429875,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.770070001220727e-06,
+      "loss": 1.0402,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9530996226980835,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.76919264231571e-06,
+      "loss": 1.1098,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9532995177531796,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.768315160403453e-06,
+      "loss": 1.0016,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9534994128082757,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.767437555522934e-06,
+      "loss": 0.996,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9536993078633718,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.766559827713131e-06,
+      "loss": 1.0092,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9538992029184679,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.765681977013037e-06,
+      "loss": 1.0211,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9540990979735638,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.764804003461646e-06,
+      "loss": 0.9899,
+      "step": 4773
+    },
+    {
+      "epoch": 0.95429899302866,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.763925907097956e-06,
+      "loss": 1.0812,
+      "step": 4774
+    },
+    {
+      "epoch": 0.954498888083756,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.763047687960971e-06,
+      "loss": 0.984,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9546987831388521,
+      "grad_norm": 2.125,
+      "learning_rate": 7.762169346089705e-06,
+      "loss": 1.0208,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9548986781939481,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.76129088152317e-06,
+      "loss": 0.9695,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9550985732490442,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.760412294300392e-06,
+      "loss": 1.0673,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9552984683041403,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.759533584460392e-06,
+      "loss": 0.9945,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9554983633592364,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.758654752042205e-06,
+      "loss": 1.0414,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9556982584143324,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.75777579708487e-06,
+      "loss": 1.0086,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9558981534694285,
+      "grad_norm": 2.25,
+      "learning_rate": 7.756896719627428e-06,
+      "loss": 1.0577,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9560980485245246,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.756017519708926e-06,
+      "loss": 1.0469,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9562979435796207,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.755138197368423e-06,
+      "loss": 1.0342,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9564978386347168,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.754258752644974e-06,
+      "loss": 0.9765,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9566977336898128,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.753379185577645e-06,
+      "loss": 1.0252,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9568976287449089,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.752499496205509e-06,
+      "loss": 0.975,
+      "step": 4787
+    },
+    {
+      "epoch": 0.957097523800005,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.751619684567638e-06,
+      "loss": 1.1005,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9572974188551011,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.750739750703114e-06,
+      "loss": 1.0138,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9574973139101971,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.749859694651023e-06,
+      "loss": 1.1132,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9576972089652932,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.74897951645046e-06,
+      "loss": 1.0403,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9578971040203893,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.74809921614052e-06,
+      "loss": 0.9768,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9580969990754854,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.747218793760308e-06,
+      "loss": 1.0403,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9582968941305814,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.746338249348928e-06,
+      "loss": 1.0055,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9584967891856775,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.745457582945497e-06,
+      "loss": 1.0358,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9586966842407736,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.744576794589132e-06,
+      "loss": 1.0853,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9588965792958697,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.743695884318961e-06,
+      "loss": 1.0774,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9590964743509658,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.742814852174112e-06,
+      "loss": 1.0557,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9592963694060618,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.741933698193719e-06,
+      "loss": 0.9809,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9594962644611579,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.741052422416923e-06,
+      "loss": 0.9666,
+      "step": 4800
+    },
+    {
+      "epoch": 0.959696159516254,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.740171024882875e-06,
+      "loss": 1.0036,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9598960545713501,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.73928950563072e-06,
+      "loss": 1.0543,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9600959496264461,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.738407864699618e-06,
+      "loss": 1.0295,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9602958446815422,
+      "grad_norm": 2.125,
+      "learning_rate": 7.737526102128729e-06,
+      "loss": 1.0324,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9604957397366383,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.736644217957226e-06,
+      "loss": 0.9947,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9606956347917344,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.735762212224278e-06,
+      "loss": 1.0961,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9608955298468305,
+      "grad_norm": 2.125,
+      "learning_rate": 7.734880084969065e-06,
+      "loss": 0.999,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9610954249019265,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.733997836230771e-06,
+      "loss": 1.0628,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9612953199570226,
+      "grad_norm": 2.0,
+      "learning_rate": 7.733115466048585e-06,
+      "loss": 1.0227,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9614952150121187,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.732232974461701e-06,
+      "loss": 1.0994,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9616951100672148,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.731350361509322e-06,
+      "loss": 1.0277,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9618950051223107,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.730467627230648e-06,
+      "loss": 1.0349,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9620949001774068,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.729584771664897e-06,
+      "loss": 0.9012,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9622947952325029,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.728701794851281e-06,
+      "loss": 0.9579,
+      "step": 4814
+    },
+    {
+      "epoch": 0.962494690287599,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.727818696829023e-06,
+      "loss": 1.1154,
+      "step": 4815
+    },
+    {
+      "epoch": 0.962694585342695,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.72693547763735e-06,
+      "loss": 0.9688,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9628944803977911,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.726052137315493e-06,
+      "loss": 1.043,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9630943754528872,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.725168675902692e-06,
+      "loss": 1.0939,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9632942705079833,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.72428509343819e-06,
+      "loss": 1.0627,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9634941655630794,
+      "grad_norm": 2.125,
+      "learning_rate": 7.723401389961235e-06,
+      "loss": 1.0583,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9636940606181754,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.72251756551108e-06,
+      "loss": 1.067,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9638939556732715,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.721633620126987e-06,
+      "loss": 1.0691,
+      "step": 4822
+    },
+    {
+      "epoch": 0.9640938507283676,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.72074955384822e-06,
+      "loss": 1.027,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9642937457834637,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.719865366714046e-06,
+      "loss": 1.1455,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9644936408385597,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.718981058763744e-06,
+      "loss": 0.9386,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9646935358936558,
+      "grad_norm": 2.25,
+      "learning_rate": 7.718096630036593e-06,
+      "loss": 1.0428,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9648934309487519,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.71721208057188e-06,
+      "loss": 0.941,
+      "step": 4827
+    },
+    {
+      "epoch": 0.965093326003848,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.7163274104089e-06,
+      "loss": 0.9998,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9652932210589441,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.715442619586943e-06,
+      "loss": 1.0023,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9654931161140401,
+      "grad_norm": 1.875,
+      "learning_rate": 7.714557708145315e-06,
+      "loss": 0.9066,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9656930111691362,
+      "grad_norm": 2.125,
+      "learning_rate": 7.713672676123324e-06,
+      "loss": 1.0072,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9658929062242323,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.712787523560283e-06,
+      "loss": 1.0759,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9660928012793284,
+      "grad_norm": 2.0,
+      "learning_rate": 7.711902250495508e-06,
+      "loss": 0.9749,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9662926963344244,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.711016856968327e-06,
+      "loss": 1.0362,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9664925913895205,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.710131343018066e-06,
+      "loss": 1.0392,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9666924864446166,
+      "grad_norm": 2.0,
+      "learning_rate": 7.709245708684061e-06,
+      "loss": 0.9769,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9668923814997127,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.708359954005651e-06,
+      "loss": 1.0357,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9670922765548087,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.70747407902218e-06,
+      "loss": 1.1927,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9672921716099048,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.706588083772999e-06,
+      "loss": 0.9573,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9674920666650009,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.705701968297466e-06,
+      "loss": 0.9817,
+      "step": 4840
+    },
+    {
+      "epoch": 0.967691961720097,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.704815732634941e-06,
+      "loss": 1.0487,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9678918567751931,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.703929376824787e-06,
+      "loss": 0.9766,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9680917518302891,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.703042900906383e-06,
+      "loss": 1.0385,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9682916468853852,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.702156304919098e-06,
+      "loss": 0.9849,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9684915419404813,
+      "grad_norm": 2.125,
+      "learning_rate": 7.70126958890232e-06,
+      "loss": 1.0976,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9686914369955774,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.700382752895436e-06,
+      "loss": 0.9467,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9688913320506733,
+      "grad_norm": 2.125,
+      "learning_rate": 7.69949579693784e-06,
+      "loss": 0.989,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9690912271057694,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.698608721068926e-06,
+      "loss": 1.0281,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9692911221608655,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.697721525328104e-06,
+      "loss": 1.0828,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9694910172159616,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.696834209754777e-06,
+      "loss": 0.9711,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9696909122710577,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.695946774388364e-06,
+      "loss": 1.026,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9698908073261537,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.695059219268281e-06,
+      "loss": 1.0395,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9700907023812498,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.694171544433958e-06,
+      "loss": 0.9939,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9702905974363459,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.693283749924821e-06,
+      "loss": 0.9857,
+      "step": 4854
+    },
+    {
+      "epoch": 0.970490492491442,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.69239583578031e-06,
+      "loss": 1.0404,
+      "step": 4855
+    },
+    {
+      "epoch": 0.970690387546538,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.691507802039861e-06,
+      "loss": 1.0722,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9708902826016341,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.690619648742923e-06,
+      "loss": 0.989,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9710901776567302,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.68973137592895e-06,
+      "loss": 1.0558,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9712900727118263,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.688842983637395e-06,
+      "loss": 1.0041,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9714899677669223,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.687954471907719e-06,
+      "loss": 1.0483,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9716898628220184,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.687065840779397e-06,
+      "loss": 1.0264,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9718897578771145,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.686177090291896e-06,
+      "loss": 1.079,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9720896529322106,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.685288220484693e-06,
+      "loss": 1.1362,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9722895479873067,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.684399231397278e-06,
+      "loss": 1.0843,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9724894430424027,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.683510123069133e-06,
+      "loss": 1.0176,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9726893380974988,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.682620895539756e-06,
+      "loss": 1.0219,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9728892331525949,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.681731548848645e-06,
+      "loss": 1.1454,
+      "step": 4867
+    },
+    {
+      "epoch": 0.973089128207691,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.680842083035305e-06,
+      "loss": 1.0198,
+      "step": 4868
+    },
+    {
+      "epoch": 0.973289023262787,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.679952498139248e-06,
+      "loss": 0.9693,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9734889183178831,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.679062794199982e-06,
+      "loss": 1.0717,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9736888133729792,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.678172971257038e-06,
+      "loss": 1.0067,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9738887084280753,
+      "grad_norm": 2.0,
+      "learning_rate": 7.677283029349936e-06,
+      "loss": 1.0556,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9740886034831714,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.676392968518205e-06,
+      "loss": 0.964,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9742884985382674,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.675502788801387e-06,
+      "loss": 1.1312,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9744883935933635,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.67461249023902e-06,
+      "loss": 0.971,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9746882886484596,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.67372207287065e-06,
+      "loss": 0.9664,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9748881837035557,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.672831536735832e-06,
+      "loss": 1.0971,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9750880787586517,
+      "grad_norm": 2.125,
+      "learning_rate": 7.671940881874124e-06,
+      "loss": 1.087,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9752879738137478,
+      "grad_norm": 2.125,
+      "learning_rate": 7.671050108325087e-06,
+      "loss": 0.9465,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9754878688688439,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.670159216128291e-06,
+      "loss": 0.9395,
+      "step": 4880
+    },
+    {
+      "epoch": 0.97568776392394,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.669268205323304e-06,
+      "loss": 0.9663,
+      "step": 4881
+    },
+    {
+      "epoch": 0.975887658979036,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.66837707594971e-06,
+      "loss": 1.026,
+      "step": 4882
+    },
+    {
+      "epoch": 0.976087554034132,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.667485828047091e-06,
+      "loss": 1.1083,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9762874490892282,
+      "grad_norm": 2.0,
+      "learning_rate": 7.666594461655039e-06,
+      "loss": 1.0118,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9764873441443243,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.665702976813142e-06,
+      "loss": 1.0811,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9766872391994204,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.664811373561008e-06,
+      "loss": 1.1536,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9768871342545163,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.663919651938234e-06,
+      "loss": 1.0617,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9770870293096124,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.663027811984433e-06,
+      "loss": 1.0567,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9772869243647085,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.662135853739224e-06,
+      "loss": 0.9078,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9774868194198046,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.661243777242223e-06,
+      "loss": 1.0036,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9776867144749006,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.660351582533057e-06,
+      "loss": 0.9038,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9778866095299967,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.65945926965136e-06,
+      "loss": 1.1392,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9780865045850928,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.658566838636762e-06,
+      "loss": 0.93,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9782863996401889,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.657674289528914e-06,
+      "loss": 0.9989,
+      "step": 4894
+    },
+    {
+      "epoch": 0.978486294695285,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.656781622367455e-06,
+      "loss": 1.0522,
+      "step": 4895
+    },
+    {
+      "epoch": 0.978686189750381,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.65588883719204e-06,
+      "loss": 1.0799,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9788860848054771,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.654995934042328e-06,
+      "loss": 0.9497,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9790859798605732,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.65410291295798e-06,
+      "loss": 1.1484,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9792858749156693,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.653209773978662e-06,
+      "loss": 1.06,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9794857699707653,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.652316517144052e-06,
+      "loss": 1.0536,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9796856650258614,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.651423142493824e-06,
+      "loss": 0.9357,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9798855600809575,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.650529650067665e-06,
+      "loss": 1.0424,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9800854551360536,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.64963603990526e-06,
+      "loss": 1.1032,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9802853501911496,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.648742312046306e-06,
+      "loss": 1.0987,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9804852452462457,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.6478484665305e-06,
+      "loss": 1.0874,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9806851403013418,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.64695450339755e-06,
+      "loss": 0.9831,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9808850353564379,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.64606042268716e-06,
+      "loss": 0.996,
+      "step": 4907
+    },
+    {
+      "epoch": 0.981084930411534,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.645166224439053e-06,
+      "loss": 1.0016,
+      "step": 4908
+    },
+    {
+      "epoch": 0.98128482546663,
+      "grad_norm": 2.0,
+      "learning_rate": 7.644271908692944e-06,
+      "loss": 0.9861,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9814847205217261,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.643377475488558e-06,
+      "loss": 1.081,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9816846155768222,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.642482924865627e-06,
+      "loss": 1.0244,
+      "step": 4911
+    },
+    {
+      "epoch": 0.9818845106319183,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.641588256863887e-06,
+      "loss": 1.0412,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9820844056870143,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.640693471523078e-06,
+      "loss": 1.1168,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9822843007421104,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.639798568882947e-06,
+      "loss": 1.0678,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9824841957972065,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.638903548983248e-06,
+      "loss": 1.0225,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9826840908523026,
+      "grad_norm": 2.0,
+      "learning_rate": 7.63800841186373e-06,
+      "loss": 0.9436,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9828839859073986,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.637113157564167e-06,
+      "loss": 1.0828,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9830838809624947,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.636217786124315e-06,
+      "loss": 0.9891,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9832837760175908,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.635322297583952e-06,
+      "loss": 1.1208,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9834836710726869,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.634426691982852e-06,
+      "loss": 0.9974,
+      "step": 4920
+    },
+    {
+      "epoch": 0.983683566127783,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.633530969360801e-06,
+      "loss": 1.0724,
+      "step": 4921
+    },
+    {
+      "epoch": 0.983883461182879,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.632635129757585e-06,
+      "loss": 1.0418,
+      "step": 4922
+    },
+    {
+      "epoch": 0.984083356237975,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.631739173212998e-06,
+      "loss": 0.9366,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9842832512930711,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.630843099766838e-06,
+      "loss": 1.0145,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9844831463481672,
+      "grad_norm": 2.125,
+      "learning_rate": 7.62994690945891e-06,
+      "loss": 1.0334,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9846830414032632,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.62905060232902e-06,
+      "loss": 1.0941,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9848829364583593,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.628154178416982e-06,
+      "loss": 1.0462,
+      "step": 4927
+    },
+    {
+      "epoch": 0.9850828315134554,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.627257637762617e-06,
+      "loss": 0.9218,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9852827265685515,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.626360980405748e-06,
+      "loss": 0.9681,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9854826216236476,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.625464206386205e-06,
+      "loss": 1.0727,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9856825166787436,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.624567315743823e-06,
+      "loss": 0.9793,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9858824117338397,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.623670308518441e-06,
+      "loss": 0.9675,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9860823067889358,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.622773184749903e-06,
+      "loss": 1.0283,
+      "step": 4933
+    },
+    {
+      "epoch": 0.9862822018440319,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.621875944478062e-06,
+      "loss": 1.1335,
+      "step": 4934
+    },
+    {
+      "epoch": 0.9864820968991279,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.62097858774277e-06,
+      "loss": 0.9563,
+      "step": 4935
+    },
+    {
+      "epoch": 0.986681991954224,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.6200811145838895e-06,
+      "loss": 0.9792,
+      "step": 4936
+    },
+    {
+      "epoch": 0.9868818870093201,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.619183525041286e-06,
+      "loss": 0.9916,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9870817820644162,
+      "grad_norm": 2.125,
+      "learning_rate": 7.618285819154829e-06,
+      "loss": 1.1046,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9872816771195122,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.617387996964396e-06,
+      "loss": 1.0579,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9874815721746083,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.616490058509869e-06,
+      "loss": 1.0783,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9876814672297044,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.615592003831132e-06,
+      "loss": 1.0153,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9878813622848005,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.614693832968079e-06,
+      "loss": 1.0465,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9880812573398966,
+      "grad_norm": 2.0,
+      "learning_rate": 7.613795545960602e-06,
+      "loss": 1.0529,
+      "step": 4943
+    },
+    {
+      "epoch": 0.9882811523949926,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.612897142848609e-06,
+      "loss": 1.1096,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9884810474500887,
+      "grad_norm": 2.125,
+      "learning_rate": 7.611998623672004e-06,
+      "loss": 0.9959,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9886809425051848,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.611099988470697e-06,
+      "loss": 0.986,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9888808375602809,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.610201237284608e-06,
+      "loss": 1.0304,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9890807326153769,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.60930237015366e-06,
+      "loss": 0.9567,
+      "step": 4948
+    },
+    {
+      "epoch": 0.989280627670473,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.608403387117779e-06,
+      "loss": 0.9946,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9894805227255691,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.607504288216898e-06,
+      "loss": 1.0602,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9896804177806652,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.606605073490955e-06,
+      "loss": 1.1024,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9898803128357613,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.605705742979892e-06,
+      "loss": 1.0524,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9900802078908573,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.604806296723659e-06,
+      "loss": 1.0077,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9902801029459534,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.603906734762209e-06,
+      "loss": 1.0811,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9904799980010495,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.6030070571354986e-06,
+      "loss": 1.0962,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9906798930561456,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.602107263883494e-06,
+      "loss": 0.9136,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9908797881112416,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.601207355046163e-06,
+      "loss": 1.1236,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9910796831663377,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.600307330663477e-06,
+      "loss": 0.9953,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9912795782214338,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.5994071907754185e-06,
+      "loss": 1.0416,
+      "step": 4959
+    },
+    {
+      "epoch": 0.9914794732765299,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.598506935421969e-06,
+      "loss": 1.0197,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9916793683316258,
+      "grad_norm": 2.0,
+      "learning_rate": 7.59760656464312e-06,
+      "loss": 0.9814,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9918792633867219,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.596706078478866e-06,
+      "loss": 0.9847,
+      "step": 4962
+    },
+    {
+      "epoch": 0.992079158441818,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.595805476969205e-06,
+      "loss": 1.0569,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9922790534969141,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.594904760154142e-06,
+      "loss": 1.0601,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9924789485520102,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.594003928073685e-06,
+      "loss": 1.1193,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9926788436071062,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.593102980767853e-06,
+      "loss": 1.0392,
+      "step": 4966
+    },
+    {
+      "epoch": 0.9928787386622023,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.592201918276661e-06,
+      "loss": 1.0762,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9930786337172984,
+      "grad_norm": 2.375,
+      "learning_rate": 7.591300740640138e-06,
+      "loss": 1.0531,
+      "step": 4968
+    },
+    {
+      "epoch": 0.9932785287723945,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.590399447898312e-06,
+      "loss": 0.9448,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9934784238274905,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.589498040091221e-06,
+      "loss": 0.9441,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9936783188825866,
+      "grad_norm": 2.125,
+      "learning_rate": 7.588596517258901e-06,
+      "loss": 1.1056,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9938782139376827,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.5876948794414015e-06,
+      "loss": 1.0203,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9940781089927788,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.586793126678773e-06,
+      "loss": 0.9404,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9942780040478749,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.585891259011069e-06,
+      "loss": 1.062,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9944778991029709,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.58498927647835e-06,
+      "loss": 1.132,
+      "step": 4975
+    },
+    {
+      "epoch": 0.994677794158067,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.5840871791206845e-06,
+      "loss": 1.0546,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9948776892131631,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.583184966978143e-06,
+      "loss": 1.0652,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9950775842682592,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.582282640090801e-06,
+      "loss": 1.0238,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9952774793233552,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.581380198498743e-06,
+      "loss": 1.0439,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9954773743784513,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.580477642242048e-06,
+      "loss": 0.9548,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9956772694335474,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.5795749713608125e-06,
+      "loss": 0.9202,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9958771644886435,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.578672185895133e-06,
+      "loss": 1.0457,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9960770595437395,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.57776928588511e-06,
+      "loss": 0.9742,
+      "step": 4983
+    },
+    {
+      "epoch": 0.9962769545988356,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.576866271370851e-06,
+      "loss": 1.0402,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9964768496539317,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.575963142392466e-06,
+      "loss": 0.9298,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9966767447090278,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.5750598989900745e-06,
+      "loss": 1.0118,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9968766397641239,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.574156541203799e-06,
+      "loss": 0.9927,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9970765348192199,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.573253069073763e-06,
+      "loss": 1.0184,
+      "step": 4988
+    },
+    {
+      "epoch": 0.997276429874316,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.5723494826401e-06,
+      "loss": 0.9645,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9974763249294121,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.571445781942948e-06,
+      "loss": 0.9988,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9976762199845082,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.570541967022449e-06,
+      "loss": 1.0216,
+      "step": 4991
+    },
+    {
+      "epoch": 0.9978761150396042,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.569638037918751e-06,
+      "loss": 1.0075,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9980760100947003,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.568733994672006e-06,
+      "loss": 1.1056,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9982759051497964,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.567829837322371e-06,
+      "loss": 0.9353,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9984758002048925,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.56692556591001e-06,
+      "loss": 1.153,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9986756952599886,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.566021180475088e-06,
+      "loss": 1.0002,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9988755903150845,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.565116681057779e-06,
+      "loss": 0.9701,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9990754853701806,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.564212067698262e-06,
+      "loss": 1.0824,
+      "step": 4998
+    },
+    {
+      "epoch": 0.9992753804252767,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.563307340436718e-06,
+      "loss": 0.9802,
+      "step": 4999
+    },
+    {
+      "epoch": 0.9994752754803728,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.562402499313336e-06,
+      "loss": 1.1098,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9996751705354688,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.561497544368309e-06,
+      "loss": 1.0144,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9998750655905649,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.560592475641835e-06,
+      "loss": 1.0494,
+      "step": 5002
+    },
+    {
+      "epoch": 1.000074960645661,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.559687293174115e-06,
+      "loss": 0.9346,
+      "step": 5003
+    },
+    {
+      "epoch": 1.000274855700757,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.55878199700536e-06,
+      "loss": 1.0637,
+      "step": 5004
+    },
+    {
+      "epoch": 1.0004747507558531,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.557876587175782e-06,
+      "loss": 0.9782,
+      "step": 5005
+    },
+    {
+      "epoch": 1.0006746458109492,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.5569710637255985e-06,
+      "loss": 0.9836,
+      "step": 5006
+    },
+    {
+      "epoch": 1.0008745408660453,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.556065426695035e-06,
+      "loss": 1.0265,
+      "step": 5007
+    },
+    {
+      "epoch": 1.0010744359211414,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.555159676124317e-06,
+      "loss": 1.1274,
+      "step": 5008
+    },
+    {
+      "epoch": 1.0012743309762375,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.55425381205368e-06,
+      "loss": 1.0692,
+      "step": 5009
+    },
+    {
+      "epoch": 1.0014742260313336,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.553347834523361e-06,
+      "loss": 0.9171,
+      "step": 5010
+    },
+    {
+      "epoch": 1.0016741210864297,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.552441743573604e-06,
+      "loss": 1.0443,
+      "step": 5011
+    },
+    {
+      "epoch": 1.0018740161415256,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.551535539244657e-06,
+      "loss": 1.0757,
+      "step": 5012
+    },
+    {
+      "epoch": 1.0020739111966217,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.550629221576774e-06,
+      "loss": 0.9581,
+      "step": 5013
+    },
+    {
+      "epoch": 1.0022738062517178,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.5497227906102125e-06,
+      "loss": 1.0106,
+      "step": 5014
+    },
+    {
+      "epoch": 1.002473701306814,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.5488162463852385e-06,
+      "loss": 1.0183,
+      "step": 5015
+    },
+    {
+      "epoch": 1.00267359636191,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.547909588942118e-06,
+      "loss": 0.9909,
+      "step": 5016
+    },
+    {
+      "epoch": 1.002873491417006,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.547002818321125e-06,
+      "loss": 0.9133,
+      "step": 5017
+    },
+    {
+      "epoch": 1.0030733864721022,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.54609593456254e-06,
+      "loss": 1.0139,
+      "step": 5018
+    },
+    {
+      "epoch": 1.0032732815271983,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.545188937706647e-06,
+      "loss": 1.036,
+      "step": 5019
+    },
+    {
+      "epoch": 1.0034731765822944,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.544281827793731e-06,
+      "loss": 1.0002,
+      "step": 5020
+    },
+    {
+      "epoch": 1.0036730716373903,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.543374604864089e-06,
+      "loss": 1.0068,
+      "step": 5021
+    },
+    {
+      "epoch": 1.0038729666924864,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.5424672689580185e-06,
+      "loss": 1.0954,
+      "step": 5022
+    },
+    {
+      "epoch": 1.0040728617475825,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.541559820115823e-06,
+      "loss": 1.0482,
+      "step": 5023
+    },
+    {
+      "epoch": 1.0042727568026786,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.540652258377812e-06,
+      "loss": 0.973,
+      "step": 5024
+    },
+    {
+      "epoch": 1.0044726518577747,
+      "grad_norm": 2.0,
+      "learning_rate": 7.539744583784299e-06,
+      "loss": 1.0635,
+      "step": 5025
+    },
+    {
+      "epoch": 1.0046725469128708,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.538836796375603e-06,
+      "loss": 0.9313,
+      "step": 5026
+    },
+    {
+      "epoch": 1.0048724419679669,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.537928896192048e-06,
+      "loss": 0.9956,
+      "step": 5027
+    },
+    {
+      "epoch": 1.005072337023063,
+      "grad_norm": 2.25,
+      "learning_rate": 7.5370208832739614e-06,
+      "loss": 1.0241,
+      "step": 5028
+    },
+    {
+      "epoch": 1.0052722320781589,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.536112757661678e-06,
+      "loss": 1.0751,
+      "step": 5029
+    },
+    {
+      "epoch": 1.005472127133255,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.535204519395538e-06,
+      "loss": 1.0667,
+      "step": 5030
+    },
+    {
+      "epoch": 1.005672022188351,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.534296168515883e-06,
+      "loss": 1.0316,
+      "step": 5031
+    },
+    {
+      "epoch": 1.0058719172434472,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.5333877050630645e-06,
+      "loss": 1.1334,
+      "step": 5032
+    },
+    {
+      "epoch": 1.0060718122985433,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.532479129077433e-06,
+      "loss": 1.0552,
+      "step": 5033
+    },
+    {
+      "epoch": 1.0062717073536394,
+      "grad_norm": 2.125,
+      "learning_rate": 7.53157044059935e-06,
+      "loss": 1.0294,
+      "step": 5034
+    },
+    {
+      "epoch": 1.0064716024087355,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.530661639669178e-06,
+      "loss": 0.9817,
+      "step": 5035
+    },
+    {
+      "epoch": 1.0066714974638316,
+      "grad_norm": 2.0,
+      "learning_rate": 7.529752726327286e-06,
+      "loss": 0.9589,
+      "step": 5036
+    },
+    {
+      "epoch": 1.0068713925189277,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.5288437006140484e-06,
+      "loss": 1.0801,
+      "step": 5037
+    },
+    {
+      "epoch": 1.0070712875740235,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.527934562569843e-06,
+      "loss": 0.9743,
+      "step": 5038
+    },
+    {
+      "epoch": 1.0072711826291196,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.5270253122350555e-06,
+      "loss": 0.9501,
+      "step": 5039
+    },
+    {
+      "epoch": 1.0074710776842157,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.526115949650073e-06,
+      "loss": 1.0456,
+      "step": 5040
+    },
+    {
+      "epoch": 1.0076709727393118,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.5252064748552915e-06,
+      "loss": 0.9935,
+      "step": 5041
+    },
+    {
+      "epoch": 1.007870867794408,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.524296887891105e-06,
+      "loss": 1.0365,
+      "step": 5042
+    },
+    {
+      "epoch": 1.008070762849504,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.5233871887979215e-06,
+      "loss": 1.073,
+      "step": 5043
+    },
+    {
+      "epoch": 1.0082706579046001,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.5224773776161495e-06,
+      "loss": 1.1247,
+      "step": 5044
+    },
+    {
+      "epoch": 1.0084705529596962,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.5215674543862004e-06,
+      "loss": 1.1517,
+      "step": 5045
+    },
+    {
+      "epoch": 1.0086704480147923,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.520657419148496e-06,
+      "loss": 1.0468,
+      "step": 5046
+    },
+    {
+      "epoch": 1.0088703430698882,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.519747271943457e-06,
+      "loss": 0.9336,
+      "step": 5047
+    },
+    {
+      "epoch": 1.0090702381249843,
+      "grad_norm": 2.125,
+      "learning_rate": 7.5188370128115125e-06,
+      "loss": 1.0277,
+      "step": 5048
+    },
+    {
+      "epoch": 1.0092701331800804,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.5179266417931e-06,
+      "loss": 0.9911,
+      "step": 5049
+    },
+    {
+      "epoch": 1.0094700282351765,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.517016158928652e-06,
+      "loss": 0.9861,
+      "step": 5050
+    },
+    {
+      "epoch": 1.0096699232902726,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.5161055642586154e-06,
+      "loss": 1.0165,
+      "step": 5051
+    },
+    {
+      "epoch": 1.0098698183453687,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.5151948578234405e-06,
+      "loss": 1.0301,
+      "step": 5052
+    },
+    {
+      "epoch": 1.0100697134004648,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.514284039663576e-06,
+      "loss": 0.9706,
+      "step": 5053
+    },
+    {
+      "epoch": 1.010269608455561,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.5133731098194854e-06,
+      "loss": 1.0205,
+      "step": 5054
+    },
+    {
+      "epoch": 1.010469503510657,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.5124620683316275e-06,
+      "loss": 1.0774,
+      "step": 5055
+    },
+    {
+      "epoch": 1.0106693985657529,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.511550915240475e-06,
+      "loss": 1.0181,
+      "step": 5056
+    },
+    {
+      "epoch": 1.010869293620849,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.510639650586498e-06,
+      "loss": 0.9683,
+      "step": 5057
+    },
+    {
+      "epoch": 1.011069188675945,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.509728274410175e-06,
+      "loss": 1.035,
+      "step": 5058
+    },
+    {
+      "epoch": 1.0112690837310412,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.508816786751991e-06,
+      "loss": 1.0364,
+      "step": 5059
+    },
+    {
+      "epoch": 1.0114689787861373,
+      "grad_norm": 2.125,
+      "learning_rate": 7.507905187652433e-06,
+      "loss": 1.0686,
+      "step": 5060
+    },
+    {
+      "epoch": 1.0116688738412334,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.506993477151996e-06,
+      "loss": 1.0212,
+      "step": 5061
+    },
+    {
+      "epoch": 1.0118687688963295,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.506081655291174e-06,
+      "loss": 1.054,
+      "step": 5062
+    },
+    {
+      "epoch": 1.0120686639514256,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.505169722110475e-06,
+      "loss": 1.0896,
+      "step": 5063
+    },
+    {
+      "epoch": 1.0122685590065217,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.504257677650402e-06,
+      "loss": 1.1003,
+      "step": 5064
+    },
+    {
+      "epoch": 1.0124684540616176,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.5033455219514705e-06,
+      "loss": 1.0835,
+      "step": 5065
+    },
+    {
+      "epoch": 1.0126683491167137,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.5024332550542e-06,
+      "loss": 0.909,
+      "step": 5066
+    },
+    {
+      "epoch": 1.0128682441718098,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.501520876999111e-06,
+      "loss": 1.0562,
+      "step": 5067
+    },
+    {
+      "epoch": 1.0130681392269059,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.500608387826731e-06,
+      "loss": 1.0792,
+      "step": 5068
+    },
+    {
+      "epoch": 1.013268034282002,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.499695787577596e-06,
+      "loss": 1.0013,
+      "step": 5069
+    },
+    {
+      "epoch": 1.013467929337098,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.498783076292238e-06,
+      "loss": 1.0855,
+      "step": 5070
+    },
+    {
+      "epoch": 1.0136678243921942,
+      "grad_norm": 2.0,
+      "learning_rate": 7.497870254011205e-06,
+      "loss": 1.0653,
+      "step": 5071
+    },
+    {
+      "epoch": 1.0138677194472903,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.496957320775042e-06,
+      "loss": 0.9759,
+      "step": 5072
+    },
+    {
+      "epoch": 1.0140676145023861,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.496044276624299e-06,
+      "loss": 1.0511,
+      "step": 5073
+    },
+    {
+      "epoch": 1.0142675095574822,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.495131121599537e-06,
+      "loss": 0.9811,
+      "step": 5074
+    },
+    {
+      "epoch": 1.0144674046125783,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.494217855741319e-06,
+      "loss": 0.9164,
+      "step": 5075
+    },
+    {
+      "epoch": 1.0146672996676744,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.493304479090208e-06,
+      "loss": 0.971,
+      "step": 5076
+    },
+    {
+      "epoch": 1.0148671947227705,
+      "grad_norm": 2.125,
+      "learning_rate": 7.49239099168678e-06,
+      "loss": 1.0386,
+      "step": 5077
+    },
+    {
+      "epoch": 1.0150670897778666,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.491477393571609e-06,
+      "loss": 0.9575,
+      "step": 5078
+    },
+    {
+      "epoch": 1.0152669848329627,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.490563684785277e-06,
+      "loss": 1.02,
+      "step": 5079
+    },
+    {
+      "epoch": 1.0154668798880588,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.489649865368375e-06,
+      "loss": 0.986,
+      "step": 5080
+    },
+    {
+      "epoch": 1.015666774943155,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.488735935361491e-06,
+      "loss": 1.0406,
+      "step": 5081
+    },
+    {
+      "epoch": 1.000124934409435,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.4878218948052206e-06,
+      "loss": 1.0103,
+      "step": 5082
+    },
+    {
+      "epoch": 1.000324829464531,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.4869077437401705e-06,
+      "loss": 0.9574,
+      "step": 5083
+    },
+    {
+      "epoch": 1.0005247245196272,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.485993482206941e-06,
+      "loss": 0.916,
+      "step": 5084
+    },
+    {
+      "epoch": 1.0007246195747233,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.48507911024615e-06,
+      "loss": 1.0239,
+      "step": 5085
+    },
+    {
+      "epoch": 1.0009245146298194,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.484164627898407e-06,
+      "loss": 1.0095,
+      "step": 5086
+    },
+    {
+      "epoch": 1.0011244096849155,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.483250035204338e-06,
+      "loss": 0.9091,
+      "step": 5087
+    },
+    {
+      "epoch": 1.0013243047400116,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.482335332204568e-06,
+      "loss": 0.9781,
+      "step": 5088
+    },
+    {
+      "epoch": 1.0015241997951077,
+      "grad_norm": 2.0,
+      "learning_rate": 7.481420518939727e-06,
+      "loss": 0.897,
+      "step": 5089
+    },
+    {
+      "epoch": 1.0017240948502038,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.480505595450451e-06,
+      "loss": 0.9291,
+      "step": 5090
+    },
+    {
+      "epoch": 1.0019239899052996,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.4795905617773834e-06,
+      "loss": 0.9006,
+      "step": 5091
+    },
+    {
+      "epoch": 1.0021238849603957,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.4786754179611666e-06,
+      "loss": 0.9878,
+      "step": 5092
+    },
+    {
+      "epoch": 1.0023237800154918,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.477760164042451e-06,
+      "loss": 0.9614,
+      "step": 5093
+    },
+    {
+      "epoch": 1.002523675070588,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.476844800061896e-06,
+      "loss": 0.9483,
+      "step": 5094
+    },
+    {
+      "epoch": 1.002723570125684,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.4759293260601585e-06,
+      "loss": 0.9782,
+      "step": 5095
+    },
+    {
+      "epoch": 1.0029234651807801,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.475013742077905e-06,
+      "loss": 0.9642,
+      "step": 5096
+    },
+    {
+      "epoch": 1.0031233602358762,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.474098048155806e-06,
+      "loss": 0.8765,
+      "step": 5097
+    },
+    {
+      "epoch": 1.0033232552909723,
+      "grad_norm": 2.125,
+      "learning_rate": 7.473182244334533e-06,
+      "loss": 1.0415,
+      "step": 5098
+    },
+    {
+      "epoch": 1.0035231503460682,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.472266330654773e-06,
+      "loss": 0.9978,
+      "step": 5099
+    },
+    {
+      "epoch": 1.0037230454011643,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.471350307157204e-06,
+      "loss": 1.0603,
+      "step": 5100
+    },
+    {
+      "epoch": 1.0039229404562604,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.470434173882519e-06,
+      "loss": 0.9902,
+      "step": 5101
+    },
+    {
+      "epoch": 1.0041228355113565,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.469517930871411e-06,
+      "loss": 0.8795,
+      "step": 5102
+    },
+    {
+      "epoch": 1.0043227305664526,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.468601578164582e-06,
+      "loss": 1.0094,
+      "step": 5103
+    },
+    {
+      "epoch": 1.0045226256215487,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.467685115802734e-06,
+      "loss": 0.9736,
+      "step": 5104
+    },
+    {
+      "epoch": 1.0047225206766448,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.466768543826577e-06,
+      "loss": 1.0023,
+      "step": 5105
+    },
+    {
+      "epoch": 1.004922415731741,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.465851862276824e-06,
+      "loss": 1.0279,
+      "step": 5106
+    },
+    {
+      "epoch": 1.005122310786837,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.4649350711941935e-06,
+      "loss": 1.079,
+      "step": 5107
+    },
+    {
+      "epoch": 1.0053222058419329,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.464018170619413e-06,
+      "loss": 0.9784,
+      "step": 5108
+    },
+    {
+      "epoch": 1.005522100897029,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.4631011605932065e-06,
+      "loss": 1.0549,
+      "step": 5109
+    },
+    {
+      "epoch": 1.005721995952125,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.462184041156309e-06,
+      "loss": 1.0397,
+      "step": 5110
+    },
+    {
+      "epoch": 1.0059218910072212,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.461266812349462e-06,
+      "loss": 1.0394,
+      "step": 5111
+    },
+    {
+      "epoch": 1.0061217860623173,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.460349474213404e-06,
+      "loss": 0.9565,
+      "step": 5112
+    },
+    {
+      "epoch": 1.0063216811174134,
+      "grad_norm": 2.125,
+      "learning_rate": 7.459432026788885e-06,
+      "loss": 0.9286,
+      "step": 5113
+    },
+    {
+      "epoch": 1.0065215761725095,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.45851447011666e-06,
+      "loss": 0.9695,
+      "step": 5114
+    },
+    {
+      "epoch": 1.0067214712276056,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.457596804237484e-06,
+      "loss": 0.9977,
+      "step": 5115
+    },
+    {
+      "epoch": 1.0069213662827017,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.45667902919212e-06,
+      "loss": 0.9384,
+      "step": 5116
+    },
+    {
+      "epoch": 1.0071212613377976,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.455761145021335e-06,
+      "loss": 0.9987,
+      "step": 5117
+    },
+    {
+      "epoch": 1.0073211563928937,
+      "grad_norm": 2.125,
+      "learning_rate": 7.454843151765904e-06,
+      "loss": 0.9157,
+      "step": 5118
+    },
+    {
+      "epoch": 1.0075210514479898,
+      "grad_norm": 2.125,
+      "learning_rate": 7.453925049466601e-06,
+      "loss": 0.9917,
+      "step": 5119
+    },
+    {
+      "epoch": 1.0077209465030859,
+      "grad_norm": 2.25,
+      "learning_rate": 7.453006838164211e-06,
+      "loss": 1.0897,
+      "step": 5120
+    },
+    {
+      "epoch": 1.007920841558182,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.452088517899518e-06,
+      "loss": 0.9977,
+      "step": 5121
+    },
+    {
+      "epoch": 1.008120736613278,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.451170088713315e-06,
+      "loss": 0.938,
+      "step": 5122
+    },
+    {
+      "epoch": 1.0083206316683742,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.450251550646398e-06,
+      "loss": 1.0497,
+      "step": 5123
+    },
+    {
+      "epoch": 1.0085205267234703,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.449332903739569e-06,
+      "loss": 1.0163,
+      "step": 5124
+    },
+    {
+      "epoch": 1.0087204217785664,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.4484141480336355e-06,
+      "loss": 1.0633,
+      "step": 5125
+    },
+    {
+      "epoch": 1.0089203168336622,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.447495283569406e-06,
+      "loss": 0.9094,
+      "step": 5126
+    },
+    {
+      "epoch": 1.0091202118887583,
+      "grad_norm": 2.125,
+      "learning_rate": 7.446576310387696e-06,
+      "loss": 0.9195,
+      "step": 5127
+    },
+    {
+      "epoch": 1.0093201069438544,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.44565722852933e-06,
+      "loss": 1.0892,
+      "step": 5128
+    },
+    {
+      "epoch": 1.0095200019989505,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.44473803803513e-06,
+      "loss": 0.9135,
+      "step": 5129
+    },
+    {
+      "epoch": 1.0097198970540466,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.443818738945927e-06,
+      "loss": 1.0406,
+      "step": 5130
+    },
+    {
+      "epoch": 1.0099197921091427,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.442899331302557e-06,
+      "loss": 0.9805,
+      "step": 5131
+    },
+    {
+      "epoch": 1.0101196871642388,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.44197981514586e-06,
+      "loss": 1.0349,
+      "step": 5132
+    },
+    {
+      "epoch": 1.010319582219335,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.44106019051668e-06,
+      "loss": 0.9732,
+      "step": 5133
+    },
+    {
+      "epoch": 1.010519477274431,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.440140457455869e-06,
+      "loss": 1.0351,
+      "step": 5134
+    },
+    {
+      "epoch": 1.010719372329527,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.439220616004277e-06,
+      "loss": 1.0287,
+      "step": 5135
+    },
+    {
+      "epoch": 1.010919267384623,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.438300666202767e-06,
+      "loss": 1.0394,
+      "step": 5136
+    },
+    {
+      "epoch": 1.011119162439719,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.4373806080922005e-06,
+      "loss": 0.9785,
+      "step": 5137
+    },
+    {
+      "epoch": 1.0113190574948152,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.436460441713448e-06,
+      "loss": 1.0159,
+      "step": 5138
+    },
+    {
+      "epoch": 1.0115189525499113,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.435540167107384e-06,
+      "loss": 0.9642,
+      "step": 5139
+    },
+    {
+      "epoch": 1.0117188476050074,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.434619784314885e-06,
+      "loss": 0.9716,
+      "step": 5140
+    },
+    {
+      "epoch": 1.0119187426601035,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.433699293376835e-06,
+      "loss": 0.9343,
+      "step": 5141
+    },
+    {
+      "epoch": 1.0121186377151996,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.432778694334124e-06,
+      "loss": 1.0314,
+      "step": 5142
+    },
+    {
+      "epoch": 1.0123185327702955,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.431857987227642e-06,
+      "loss": 1.0005,
+      "step": 5143
+    },
+    {
+      "epoch": 1.0125184278253916,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.430937172098288e-06,
+      "loss": 0.9713,
+      "step": 5144
+    },
+    {
+      "epoch": 1.0127183228804877,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.430016248986964e-06,
+      "loss": 1.0475,
+      "step": 5145
+    },
+    {
+      "epoch": 1.0129182179355838,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.429095217934578e-06,
+      "loss": 1.0999,
+      "step": 5146
+    },
+    {
+      "epoch": 1.0131181129906799,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.428174078982042e-06,
+      "loss": 1.0007,
+      "step": 5147
+    },
+    {
+      "epoch": 1.013318008045776,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.427252832170273e-06,
+      "loss": 0.9999,
+      "step": 5148
+    },
+    {
+      "epoch": 1.013517903100872,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.426331477540193e-06,
+      "loss": 1.0016,
+      "step": 5149
+    },
+    {
+      "epoch": 1.0137177981559682,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.425410015132728e-06,
+      "loss": 0.9672,
+      "step": 5150
+    },
+    {
+      "epoch": 1.0139176932110643,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.424488444988807e-06,
+      "loss": 0.9684,
+      "step": 5151
+    },
+    {
+      "epoch": 1.0141175882661602,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.42356676714937e-06,
+      "loss": 0.8617,
+      "step": 5152
+    },
+    {
+      "epoch": 1.0143174833212563,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.422644981655356e-06,
+      "loss": 0.9923,
+      "step": 5153
+    },
+    {
+      "epoch": 1.0145173783763524,
+      "grad_norm": 2.125,
+      "learning_rate": 7.42172308854771e-06,
+      "loss": 0.8722,
+      "step": 5154
+    },
+    {
+      "epoch": 1.0147172734314485,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.420801087867382e-06,
+      "loss": 0.9362,
+      "step": 5155
+    },
+    {
+      "epoch": 1.0149171684865446,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.419878979655331e-06,
+      "loss": 0.9697,
+      "step": 5156
+    },
+    {
+      "epoch": 1.0151170635416407,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.418956763952512e-06,
+      "loss": 1.011,
+      "step": 5157
+    },
+    {
+      "epoch": 1.0153169585967368,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.418034440799892e-06,
+      "loss": 1.0548,
+      "step": 5158
+    },
+    {
+      "epoch": 1.0155168536518329,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.417112010238442e-06,
+      "loss": 0.9735,
+      "step": 5159
+    },
+    {
+      "epoch": 1.015716748706929,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.416189472309133e-06,
+      "loss": 0.9858,
+      "step": 5160
+    },
+    {
+      "epoch": 1.0159166437620248,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.415266827052947e-06,
+      "loss": 0.9463,
+      "step": 5161
+    },
+    {
+      "epoch": 1.016116538817121,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.414344074510865e-06,
+      "loss": 0.9858,
+      "step": 5162
+    },
+    {
+      "epoch": 1.016316433872217,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.413421214723878e-06,
+      "loss": 1.0522,
+      "step": 5163
+    },
+    {
+      "epoch": 1.0165163289273131,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.412498247732979e-06,
+      "loss": 1.021,
+      "step": 5164
+    },
+    {
+      "epoch": 1.0167162239824092,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.4115751735791655e-06,
+      "loss": 0.9006,
+      "step": 5165
+    },
+    {
+      "epoch": 1.0169161190375053,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.410651992303439e-06,
+      "loss": 1.0421,
+      "step": 5166
+    },
+    {
+      "epoch": 1.0171160140926014,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.40972870394681e-06,
+      "loss": 1.002,
+      "step": 5167
+    },
+    {
+      "epoch": 1.0173159091476975,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.408805308550288e-06,
+      "loss": 0.9943,
+      "step": 5168
+    },
+    {
+      "epoch": 1.0175158042027936,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.407881806154892e-06,
+      "loss": 1.0892,
+      "step": 5169
+    },
+    {
+      "epoch": 1.0177156992578895,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.406958196801644e-06,
+      "loss": 0.9304,
+      "step": 5170
+    },
+    {
+      "epoch": 1.0179155943129856,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.406034480531568e-06,
+      "loss": 0.9051,
+      "step": 5171
+    },
+    {
+      "epoch": 1.0181154893680817,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.405110657385699e-06,
+      "loss": 0.9884,
+      "step": 5172
+    },
+    {
+      "epoch": 1.0183153844231778,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.4041867274050715e-06,
+      "loss": 1.0153,
+      "step": 5173
+    },
+    {
+      "epoch": 1.018515279478274,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.403262690630725e-06,
+      "loss": 0.8693,
+      "step": 5174
+    },
+    {
+      "epoch": 1.01871517453337,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.402338547103708e-06,
+      "loss": 1.012,
+      "step": 5175
+    },
+    {
+      "epoch": 1.0189150695884661,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.401414296865068e-06,
+      "loss": 1.0537,
+      "step": 5176
+    },
+    {
+      "epoch": 1.0191149646435622,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.400489939955862e-06,
+      "loss": 0.9194,
+      "step": 5177
+    },
+    {
+      "epoch": 1.019314859698658,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.39956547641715e-06,
+      "loss": 0.9411,
+      "step": 5178
+    },
+    {
+      "epoch": 1.0195147547537542,
+      "grad_norm": 2.125,
+      "learning_rate": 7.398640906289996e-06,
+      "loss": 0.9623,
+      "step": 5179
+    },
+    {
+      "epoch": 1.0197146498088503,
+      "grad_norm": 2.0,
+      "learning_rate": 7.397716229615468e-06,
+      "loss": 0.9034,
+      "step": 5180
+    },
+    {
+      "epoch": 1.0199145448639464,
+      "grad_norm": 2.125,
+      "learning_rate": 7.396791446434641e-06,
+      "loss": 0.9748,
+      "step": 5181
+    },
+    {
+      "epoch": 1.0201144399190425,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.3958665567885945e-06,
+      "loss": 0.9026,
+      "step": 5182
+    },
+    {
+      "epoch": 1.0203143349741386,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.394941560718412e-06,
+      "loss": 0.9915,
+      "step": 5183
+    },
+    {
+      "epoch": 1.0205142300292347,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.394016458265181e-06,
+      "loss": 0.9947,
+      "step": 5184
+    },
+    {
+      "epoch": 1.0207141250843308,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.3930912494699935e-06,
+      "loss": 1.004,
+      "step": 5185
+    },
+    {
+      "epoch": 1.020914020139427,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.3921659343739485e-06,
+      "loss": 0.9647,
+      "step": 5186
+    },
+    {
+      "epoch": 1.0211139151945228,
+      "grad_norm": 2.0,
+      "learning_rate": 7.391240513018149e-06,
+      "loss": 0.9818,
+      "step": 5187
+    },
+    {
+      "epoch": 1.0213138102496189,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.3903149854437e-06,
+      "loss": 0.8793,
+      "step": 5188
+    },
+    {
+      "epoch": 1.021513705304715,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.389389351691717e-06,
+      "loss": 1.012,
+      "step": 5189
+    },
+    {
+      "epoch": 1.021713600359811,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.3884636118033115e-06,
+      "loss": 0.9668,
+      "step": 5190
+    },
+    {
+      "epoch": 1.0219134954149072,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.387537765819609e-06,
+      "loss": 0.9573,
+      "step": 5191
+    },
+    {
+      "epoch": 1.0221133904700033,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.3866118137817344e-06,
+      "loss": 0.9249,
+      "step": 5192
+    },
+    {
+      "epoch": 1.0223132855250994,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.385685755730816e-06,
+      "loss": 1.0011,
+      "step": 5193
+    },
+    {
+      "epoch": 1.0225131805801955,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.384759591707993e-06,
+      "loss": 1.0457,
+      "step": 5194
+    },
+    {
+      "epoch": 1.0227130756352916,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.3838333217544035e-06,
+      "loss": 0.986,
+      "step": 5195
+    },
+    {
+      "epoch": 1.0229129706903874,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.3829069459111925e-06,
+      "loss": 1.0031,
+      "step": 5196
+    },
+    {
+      "epoch": 1.0231128657454835,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.381980464219508e-06,
+      "loss": 0.9511,
+      "step": 5197
+    },
+    {
+      "epoch": 1.0233127608005796,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.381053876720508e-06,
+      "loss": 1.0288,
+      "step": 5198
+    },
+    {
+      "epoch": 1.0235126558556757,
+      "grad_norm": 2.0,
+      "learning_rate": 7.380127183455345e-06,
+      "loss": 0.9307,
+      "step": 5199
+    },
+    {
+      "epoch": 1.0237125509107718,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.379200384465191e-06,
+      "loss": 0.9667,
+      "step": 5200
+    },
+    {
+      "epoch": 1.023912445965868,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.378273479791208e-06,
+      "loss": 0.968,
+      "step": 5201
+    },
+    {
+      "epoch": 1.024112341020964,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.377346469474571e-06,
+      "loss": 0.9898,
+      "step": 5202
+    },
+    {
+      "epoch": 1.0243122360760601,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.376419353556458e-06,
+      "loss": 0.9714,
+      "step": 5203
+    },
+    {
+      "epoch": 1.0245121311311562,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.375492132078051e-06,
+      "loss": 0.9354,
+      "step": 5204
+    },
+    {
+      "epoch": 1.0247120261862521,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.374564805080537e-06,
+      "loss": 1.0316,
+      "step": 5205
+    },
+    {
+      "epoch": 1.0249119212413482,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.373637372605111e-06,
+      "loss": 0.9557,
+      "step": 5206
+    },
+    {
+      "epoch": 1.0251118162964443,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.372709834692962e-06,
+      "loss": 0.9577,
+      "step": 5207
+    },
+    {
+      "epoch": 1.0253117113515404,
+      "grad_norm": 2.0,
+      "learning_rate": 7.371782191385297e-06,
+      "loss": 0.8865,
+      "step": 5208
+    },
+    {
+      "epoch": 1.0255116064066365,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.370854442723322e-06,
+      "loss": 0.9397,
+      "step": 5209
+    },
+    {
+      "epoch": 1.0257115014617326,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.369926588748244e-06,
+      "loss": 0.9066,
+      "step": 5210
+    },
+    {
+      "epoch": 1.0259113965168287,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.368998629501282e-06,
+      "loss": 1.0038,
+      "step": 5211
+    },
+    {
+      "epoch": 1.0261112915719248,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.368070565023653e-06,
+      "loss": 0.9404,
+      "step": 5212
+    },
+    {
+      "epoch": 1.026311186627021,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.367142395356581e-06,
+      "loss": 0.8999,
+      "step": 5213
+    },
+    {
+      "epoch": 1.0265110816821168,
+      "grad_norm": 2.125,
+      "learning_rate": 7.3662141205412975e-06,
+      "loss": 0.9461,
+      "step": 5214
+    },
+    {
+      "epoch": 1.026710976737213,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.365285740619036e-06,
+      "loss": 0.9219,
+      "step": 5215
+    },
+    {
+      "epoch": 1.026910871792309,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.364357255631034e-06,
+      "loss": 0.9997,
+      "step": 5216
+    },
+    {
+      "epoch": 1.027110766847405,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.363428665618535e-06,
+      "loss": 0.9891,
+      "step": 5217
+    },
+    {
+      "epoch": 1.0273106619025012,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.3624999706227885e-06,
+      "loss": 0.9606,
+      "step": 5218
+    },
+    {
+      "epoch": 1.0275105569575973,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.361571170685043e-06,
+      "loss": 1.0675,
+      "step": 5219
+    },
+    {
+      "epoch": 1.0277104520126934,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.360642265846562e-06,
+      "loss": 0.9213,
+      "step": 5220
+    },
+    {
+      "epoch": 1.0279103470677895,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.359713256148601e-06,
+      "loss": 0.8831,
+      "step": 5221
+    },
+    {
+      "epoch": 1.0281102421228854,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.358784141632429e-06,
+      "loss": 1.1255,
+      "step": 5222
+    },
+    {
+      "epoch": 1.0283101371779815,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.357854922339318e-06,
+      "loss": 0.9533,
+      "step": 5223
+    },
+    {
+      "epoch": 1.0285100322330776,
+      "grad_norm": 2.125,
+      "learning_rate": 7.356925598310544e-06,
+      "loss": 0.9819,
+      "step": 5224
+    },
+    {
+      "epoch": 1.0287099272881737,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.355996169587385e-06,
+      "loss": 0.8877,
+      "step": 5225
+    },
+    {
+      "epoch": 1.0289098223432698,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.35506663621113e-06,
+      "loss": 0.9553,
+      "step": 5226
+    },
+    {
+      "epoch": 1.0291097173983659,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.354136998223066e-06,
+      "loss": 1.0505,
+      "step": 5227
+    },
+    {
+      "epoch": 1.029309612453462,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.353207255664486e-06,
+      "loss": 0.9877,
+      "step": 5228
+    },
+    {
+      "epoch": 1.029509507508558,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.352277408576693e-06,
+      "loss": 1.1288,
+      "step": 5229
+    },
+    {
+      "epoch": 1.0297094025636542,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.3513474570009876e-06,
+      "loss": 0.9763,
+      "step": 5230
+    },
+    {
+      "epoch": 1.02990929761875,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.35041740097868e-06,
+      "loss": 0.975,
+      "step": 5231
+    },
+    {
+      "epoch": 1.0301091926738462,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.349487240551083e-06,
+      "loss": 1.0409,
+      "step": 5232
+    },
+    {
+      "epoch": 1.0303090877289423,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.348556975759512e-06,
+      "loss": 1.0961,
+      "step": 5233
+    },
+    {
+      "epoch": 1.0305089827840384,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.3476266066452925e-06,
+      "loss": 1.0103,
+      "step": 5234
+    },
+    {
+      "epoch": 1.0307088778391345,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.346696133249749e-06,
+      "loss": 0.9519,
+      "step": 5235
+    },
+    {
+      "epoch": 1.0309087728942306,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.345765555614214e-06,
+      "loss": 0.9152,
+      "step": 5236
+    },
+    {
+      "epoch": 1.0311086679493267,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.344834873780024e-06,
+      "loss": 1.0414,
+      "step": 5237
+    },
+    {
+      "epoch": 1.0313085630044228,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.3439040877885184e-06,
+      "loss": 0.9877,
+      "step": 5238
+    },
+    {
+      "epoch": 1.0315084580595189,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.342973197681045e-06,
+      "loss": 0.9529,
+      "step": 5239
+    },
+    {
+      "epoch": 1.0317083531146147,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.342042203498952e-06,
+      "loss": 0.9597,
+      "step": 5240
+    },
+    {
+      "epoch": 1.0319082481697108,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.341111105283594e-06,
+      "loss": 0.9566,
+      "step": 5241
+    },
+    {
+      "epoch": 1.032108143224807,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.34017990307633e-06,
+      "loss": 1.0736,
+      "step": 5242
+    },
+    {
+      "epoch": 1.032308038279903,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.339248596918526e-06,
+      "loss": 1.0523,
+      "step": 5243
+    },
+    {
+      "epoch": 1.0325079333349991,
+      "grad_norm": 2.0,
+      "learning_rate": 7.338317186851549e-06,
+      "loss": 1.0294,
+      "step": 5244
+    },
+    {
+      "epoch": 1.0327078283900952,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.337385672916772e-06,
+      "loss": 0.9691,
+      "step": 5245
+    },
+    {
+      "epoch": 1.0329077234451913,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.336454055155573e-06,
+      "loss": 1.0367,
+      "step": 5246
+    },
+    {
+      "epoch": 1.0331076185002874,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.335522333609334e-06,
+      "loss": 0.9095,
+      "step": 5247
+    },
+    {
+      "epoch": 1.0333075135553835,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.334590508319442e-06,
+      "loss": 0.9809,
+      "step": 5248
+    },
+    {
+      "epoch": 1.0335074086104794,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.3336585793272905e-06,
+      "loss": 0.9904,
+      "step": 5249
+    },
+    {
+      "epoch": 1.0337073036655755,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.3327265466742734e-06,
+      "loss": 0.941,
+      "step": 5250
+    },
+    {
+      "epoch": 1.0339071987206716,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.331794410401792e-06,
+      "loss": 1.0044,
+      "step": 5251
+    },
+    {
+      "epoch": 1.0341070937757677,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.330862170551253e-06,
+      "loss": 0.8835,
+      "step": 5252
+    },
+    {
+      "epoch": 1.0343069888308638,
+      "grad_norm": 2.125,
+      "learning_rate": 7.329929827164064e-06,
+      "loss": 0.9934,
+      "step": 5253
+    },
+    {
+      "epoch": 1.03450688388596,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.328997380281642e-06,
+      "loss": 0.9716,
+      "step": 5254
+    },
+    {
+      "epoch": 1.034706778941056,
+      "grad_norm": 2.25,
+      "learning_rate": 7.3280648299454035e-06,
+      "loss": 1.0728,
+      "step": 5255
+    },
+    {
+      "epoch": 1.034906673996152,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.3271321761967754e-06,
+      "loss": 0.9832,
+      "step": 5256
+    },
+    {
+      "epoch": 1.0351065690512482,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.326199419077185e-06,
+      "loss": 1.0008,
+      "step": 5257
+    },
+    {
+      "epoch": 1.035306464106344,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.325266558628064e-06,
+      "loss": 0.9786,
+      "step": 5258
+    },
+    {
+      "epoch": 1.0355063591614402,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.32433359489085e-06,
+      "loss": 0.9522,
+      "step": 5259
+    },
+    {
+      "epoch": 1.0357062542165363,
+      "grad_norm": 2.0,
+      "learning_rate": 7.323400527906988e-06,
+      "loss": 0.9269,
+      "step": 5260
+    },
+    {
+      "epoch": 1.0359061492716324,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.32246735771792e-06,
+      "loss": 0.8953,
+      "step": 5261
+    },
+    {
+      "epoch": 1.0361060443267285,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.321534084365101e-06,
+      "loss": 1.0412,
+      "step": 5262
+    },
+    {
+      "epoch": 1.0363059393818246,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.320600707889988e-06,
+      "loss": 1.0666,
+      "step": 5263
+    },
+    {
+      "epoch": 1.0365058344369207,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.3196672283340364e-06,
+      "loss": 0.9433,
+      "step": 5264
+    },
+    {
+      "epoch": 1.0367057294920168,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.318733645738716e-06,
+      "loss": 0.9937,
+      "step": 5265
+    },
+    {
+      "epoch": 1.0369056245471127,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.317799960145495e-06,
+      "loss": 1.0192,
+      "step": 5266
+    },
+    {
+      "epoch": 1.0371055196022088,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.316866171595846e-06,
+      "loss": 0.9529,
+      "step": 5267
+    },
+    {
+      "epoch": 1.0373054146573049,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.31593228013125e-06,
+      "loss": 1.0775,
+      "step": 5268
+    },
+    {
+      "epoch": 1.037505309712401,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.314998285793189e-06,
+      "loss": 1.0069,
+      "step": 5269
+    },
+    {
+      "epoch": 1.037705204767497,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.314064188623151e-06,
+      "loss": 0.9793,
+      "step": 5270
+    },
+    {
+      "epoch": 1.0379050998225932,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.313129988662631e-06,
+      "loss": 1.0244,
+      "step": 5271
+    },
+    {
+      "epoch": 1.0381049948776893,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.312195685953122e-06,
+      "loss": 1.026,
+      "step": 5272
+    },
+    {
+      "epoch": 1.0383048899327854,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.311261280536129e-06,
+      "loss": 0.9317,
+      "step": 5273
+    },
+    {
+      "epoch": 1.0385047849878815,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.310326772453156e-06,
+      "loss": 1.0624,
+      "step": 5274
+    },
+    {
+      "epoch": 1.0387046800429773,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.309392161745714e-06,
+      "loss": 0.9344,
+      "step": 5275
+    },
+    {
+      "epoch": 1.0389045750980734,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.3084574484553185e-06,
+      "loss": 0.9381,
+      "step": 5276
+    },
+    {
+      "epoch": 1.0391044701531695,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.307522632623491e-06,
+      "loss": 0.9804,
+      "step": 5277
+    },
+    {
+      "epoch": 1.0393043652082656,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.306587714291753e-06,
+      "loss": 1.0398,
+      "step": 5278
+    },
+    {
+      "epoch": 1.0395042602633617,
+      "grad_norm": 2.125,
+      "learning_rate": 7.305652693501637e-06,
+      "loss": 1.012,
+      "step": 5279
+    },
+    {
+      "epoch": 1.0397041553184578,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.304717570294674e-06,
+      "loss": 0.9732,
+      "step": 5280
+    },
+    {
+      "epoch": 1.039904050373554,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.303782344712401e-06,
+      "loss": 1.025,
+      "step": 5281
+    },
+    {
+      "epoch": 1.04010394542865,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.302847016796365e-06,
+      "loss": 0.9559,
+      "step": 5282
+    },
+    {
+      "epoch": 1.0403038404837461,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.301911586588108e-06,
+      "loss": 0.9812,
+      "step": 5283
+    },
+    {
+      "epoch": 1.040503735538842,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.300976054129185e-06,
+      "loss": 0.9592,
+      "step": 5284
+    },
+    {
+      "epoch": 1.0407036305939381,
+      "grad_norm": 2.125,
+      "learning_rate": 7.300040419461153e-06,
+      "loss": 0.9289,
+      "step": 5285
+    },
+    {
+      "epoch": 1.0409035256490342,
+      "grad_norm": 2.125,
+      "learning_rate": 7.2991046826255685e-06,
+      "loss": 0.9968,
+      "step": 5286
+    },
+    {
+      "epoch": 1.0411034207041303,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.298168843664001e-06,
+      "loss": 0.9789,
+      "step": 5287
+    },
+    {
+      "epoch": 1.0413033157592264,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.297232902618021e-06,
+      "loss": 0.8654,
+      "step": 5288
+    },
+    {
+      "epoch": 1.0415032108143225,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.296296859529199e-06,
+      "loss": 0.9042,
+      "step": 5289
+    },
+    {
+      "epoch": 1.0417031058694186,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.295360714439115e-06,
+      "loss": 0.9787,
+      "step": 5290
+    },
+    {
+      "epoch": 1.0419030009245147,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.294424467389354e-06,
+      "loss": 0.9877,
+      "step": 5291
+    },
+    {
+      "epoch": 1.0421028959796108,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.293488118421502e-06,
+      "loss": 1.0066,
+      "step": 5292
+    },
+    {
+      "epoch": 1.0423027910347067,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.292551667577153e-06,
+      "loss": 0.9006,
+      "step": 5293
+    },
+    {
+      "epoch": 1.0425026860898028,
+      "grad_norm": 2.125,
+      "learning_rate": 7.291615114897905e-06,
+      "loss": 0.9904,
+      "step": 5294
+    },
+    {
+      "epoch": 1.0427025811448989,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.290678460425358e-06,
+      "loss": 0.9638,
+      "step": 5295
+    },
+    {
+      "epoch": 1.042902476199995,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.289741704201119e-06,
+      "loss": 0.9075,
+      "step": 5296
+    },
+    {
+      "epoch": 1.043102371255091,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.288804846266796e-06,
+      "loss": 0.9738,
+      "step": 5297
+    },
+    {
+      "epoch": 1.0433022663101872,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.287867886664008e-06,
+      "loss": 0.9857,
+      "step": 5298
+    },
+    {
+      "epoch": 1.0435021613652833,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.286930825434372e-06,
+      "loss": 0.9809,
+      "step": 5299
+    },
+    {
+      "epoch": 1.0437020564203794,
+      "grad_norm": 3.0,
+      "learning_rate": 7.2859936626195126e-06,
+      "loss": 1.0613,
+      "step": 5300
+    },
+    {
+      "epoch": 1.0439019514754753,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.285056398261059e-06,
+      "loss": 0.9796,
+      "step": 5301
+    },
+    {
+      "epoch": 1.0441018465305714,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.2841190324006464e-06,
+      "loss": 0.8832,
+      "step": 5302
+    },
+    {
+      "epoch": 1.0443017415856675,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.283181565079907e-06,
+      "loss": 1.0052,
+      "step": 5303
+    },
+    {
+      "epoch": 1.0445016366407636,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.28224399634049e-06,
+      "loss": 0.9675,
+      "step": 5304
+    },
+    {
+      "epoch": 1.0447015316958597,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.2813063262240355e-06,
+      "loss": 1.0163,
+      "step": 5305
+    },
+    {
+      "epoch": 1.0449014267509558,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.280368554772199e-06,
+      "loss": 0.9621,
+      "step": 5306
+    },
+    {
+      "epoch": 1.0451013218060519,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.2794306820266335e-06,
+      "loss": 1.0418,
+      "step": 5307
+    },
+    {
+      "epoch": 1.045301216861148,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.2784927080290025e-06,
+      "loss": 1.0345,
+      "step": 5308
+    },
+    {
+      "epoch": 1.045501111916244,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.277554632820968e-06,
+      "loss": 1.0125,
+      "step": 5309
+    },
+    {
+      "epoch": 1.04570100697134,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.2766164564442e-06,
+      "loss": 0.9752,
+      "step": 5310
+    },
+    {
+      "epoch": 1.045900902026436,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.275678178940372e-06,
+      "loss": 1.078,
+      "step": 5311
+    },
+    {
+      "epoch": 1.0461007970815321,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.274739800351162e-06,
+      "loss": 0.9943,
+      "step": 5312
+    },
+    {
+      "epoch": 1.0463006921366282,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.273801320718254e-06,
+      "loss": 1.0436,
+      "step": 5313
+    },
+    {
+      "epoch": 1.0465005871917243,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.272862740083332e-06,
+      "loss": 1.027,
+      "step": 5314
+    },
+    {
+      "epoch": 1.0467004822468204,
+      "grad_norm": 2.125,
+      "learning_rate": 7.271924058488091e-06,
+      "loss": 1.0068,
+      "step": 5315
+    },
+    {
+      "epoch": 1.0469003773019165,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.270985275974227e-06,
+      "loss": 0.947,
+      "step": 5316
+    },
+    {
+      "epoch": 1.0471002723570126,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.270046392583438e-06,
+      "loss": 1.0863,
+      "step": 5317
+    },
+    {
+      "epoch": 1.0473001674121087,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.269107408357432e-06,
+      "loss": 0.9548,
+      "step": 5318
+    },
+    {
+      "epoch": 1.0475000624672046,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.2681683233379176e-06,
+      "loss": 0.9997,
+      "step": 5319
+    },
+    {
+      "epoch": 1.0476999575223007,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.267229137566607e-06,
+      "loss": 0.9421,
+      "step": 5320
+    },
+    {
+      "epoch": 1.0478998525773968,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.266289851085221e-06,
+      "loss": 0.9005,
+      "step": 5321
+    },
+    {
+      "epoch": 1.048099747632493,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.265350463935482e-06,
+      "loss": 1.0251,
+      "step": 5322
+    },
+    {
+      "epoch": 1.048299642687589,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.264410976159117e-06,
+      "loss": 1.008,
+      "step": 5323
+    },
+    {
+      "epoch": 1.0484995377426851,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.263471387797859e-06,
+      "loss": 1.0063,
+      "step": 5324
+    },
+    {
+      "epoch": 1.0486994327977812,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.262531698893443e-06,
+      "loss": 1.0123,
+      "step": 5325
+    },
+    {
+      "epoch": 1.0488993278528773,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.261591909487611e-06,
+      "loss": 0.9359,
+      "step": 5326
+    },
+    {
+      "epoch": 1.0490992229079734,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.260652019622108e-06,
+      "loss": 0.9853,
+      "step": 5327
+    },
+    {
+      "epoch": 1.0492991179630693,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.259712029338682e-06,
+      "loss": 0.9753,
+      "step": 5328
+    },
+    {
+      "epoch": 1.0494990130181654,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.25877193867909e-06,
+      "loss": 1.0207,
+      "step": 5329
+    },
+    {
+      "epoch": 1.0496989080732615,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.2578317476850915e-06,
+      "loss": 1.0005,
+      "step": 5330
+    },
+    {
+      "epoch": 1.0498988031283576,
+      "grad_norm": 2.125,
+      "learning_rate": 7.256891456398446e-06,
+      "loss": 0.9452,
+      "step": 5331
+    },
+    {
+      "epoch": 1.0500986981834537,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.2559510648609234e-06,
+      "loss": 1.0279,
+      "step": 5332
+    },
+    {
+      "epoch": 1.0502985932385498,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.255010573114296e-06,
+      "loss": 1.002,
+      "step": 5333
+    },
+    {
+      "epoch": 1.050498488293646,
+      "grad_norm": 2.0,
+      "learning_rate": 7.254069981200339e-06,
+      "loss": 0.8251,
+      "step": 5334
+    },
+    {
+      "epoch": 1.050698383348742,
+      "grad_norm": 2.125,
+      "learning_rate": 7.253129289160835e-06,
+      "loss": 1.0114,
+      "step": 5335
+    },
+    {
+      "epoch": 1.0508982784038379,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.252188497037569e-06,
+      "loss": 1.0005,
+      "step": 5336
+    },
+    {
+      "epoch": 1.051098173458934,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.251247604872329e-06,
+      "loss": 0.9921,
+      "step": 5337
+    },
+    {
+      "epoch": 1.05129806851403,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.250306612706912e-06,
+      "loss": 0.9755,
+      "step": 5338
+    },
+    {
+      "epoch": 1.0514979635691262,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.249365520583116e-06,
+      "loss": 0.942,
+      "step": 5339
+    },
+    {
+      "epoch": 1.0516978586242223,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.248424328542742e-06,
+      "loss": 1.055,
+      "step": 5340
+    },
+    {
+      "epoch": 1.0518977536793184,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.247483036627601e-06,
+      "loss": 1.0509,
+      "step": 5341
+    },
+    {
+      "epoch": 1.0520976487344145,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.246541644879502e-06,
+      "loss": 0.9881,
+      "step": 5342
+    },
+    {
+      "epoch": 1.0522975437895106,
+      "grad_norm": 2.125,
+      "learning_rate": 7.245600153340264e-06,
+      "loss": 0.9471,
+      "step": 5343
+    },
+    {
+      "epoch": 1.0524974388446067,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.244658562051708e-06,
+      "loss": 0.9631,
+      "step": 5344
+    },
+    {
+      "epoch": 1.0526973338997028,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.243716871055657e-06,
+      "loss": 0.9688,
+      "step": 5345
+    },
+    {
+      "epoch": 1.0528972289547986,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.242775080393942e-06,
+      "loss": 0.9997,
+      "step": 5346
+    },
+    {
+      "epoch": 1.0530971240098947,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.241833190108399e-06,
+      "loss": 0.947,
+      "step": 5347
+    },
+    {
+      "epoch": 1.0532970190649908,
+      "grad_norm": 2.125,
+      "learning_rate": 7.240891200240864e-06,
+      "loss": 1.0385,
+      "step": 5348
+    },
+    {
+      "epoch": 1.053496914120087,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.239949110833182e-06,
+      "loss": 1.0076,
+      "step": 5349
+    },
+    {
+      "epoch": 1.053696809175183,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.2390069219272e-06,
+      "loss": 1.033,
+      "step": 5350
+    },
+    {
+      "epoch": 1.0538967042302791,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.238064633564769e-06,
+      "loss": 1.0587,
+      "step": 5351
+    },
+    {
+      "epoch": 1.0540965992853752,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.2371222457877456e-06,
+      "loss": 1.0285,
+      "step": 5352
+    },
+    {
+      "epoch": 1.0542964943404713,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.236179758637991e-06,
+      "loss": 0.9545,
+      "step": 5353
+    },
+    {
+      "epoch": 1.0544963893955672,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.2352371721573715e-06,
+      "loss": 1.0356,
+      "step": 5354
+    },
+    {
+      "epoch": 1.0546962844506633,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.234294486387754e-06,
+      "loss": 0.9266,
+      "step": 5355
+    },
+    {
+      "epoch": 1.0548961795057594,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.233351701371015e-06,
+      "loss": 0.9389,
+      "step": 5356
+    },
+    {
+      "epoch": 1.0550960745608555,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.232408817149032e-06,
+      "loss": 0.9307,
+      "step": 5357
+    },
+    {
+      "epoch": 1.0552959696159516,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.231465833763687e-06,
+      "loss": 1.0046,
+      "step": 5358
+    },
+    {
+      "epoch": 1.0554958646710477,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.230522751256868e-06,
+      "loss": 0.9269,
+      "step": 5359
+    },
+    {
+      "epoch": 1.0556957597261438,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.229579569670467e-06,
+      "loss": 0.9892,
+      "step": 5360
+    },
+    {
+      "epoch": 1.05589565478124,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.2286362890463805e-06,
+      "loss": 0.9841,
+      "step": 5361
+    },
+    {
+      "epoch": 1.056095549836336,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.227692909426507e-06,
+      "loss": 1.094,
+      "step": 5362
+    },
+    {
+      "epoch": 1.056295444891432,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.226749430852753e-06,
+      "loss": 0.9837,
+      "step": 5363
+    },
+    {
+      "epoch": 1.056495339946528,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.225805853367027e-06,
+      "loss": 0.9187,
+      "step": 5364
+    },
+    {
+      "epoch": 1.056695235001624,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.224862177011241e-06,
+      "loss": 0.9311,
+      "step": 5365
+    },
+    {
+      "epoch": 1.0568951300567202,
+      "grad_norm": 2.125,
+      "learning_rate": 7.223918401827318e-06,
+      "loss": 0.9494,
+      "step": 5366
+    },
+    {
+      "epoch": 1.0570950251118163,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.222974527857176e-06,
+      "loss": 0.9531,
+      "step": 5367
+    },
+    {
+      "epoch": 1.0572949201669124,
+      "grad_norm": 2.125,
+      "learning_rate": 7.222030555142742e-06,
+      "loss": 1.0476,
+      "step": 5368
+    },
+    {
+      "epoch": 1.0574948152220085,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.2210864837259474e-06,
+      "loss": 0.9496,
+      "step": 5369
+    },
+    {
+      "epoch": 1.0576947102771046,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.22014231364873e-06,
+      "loss": 1.0252,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0578946053322007,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.219198044953026e-06,
+      "loss": 0.9492,
+      "step": 5371
+    },
+    {
+      "epoch": 1.0580945003872966,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.2182536776807845e-06,
+      "loss": 1.0375,
+      "step": 5372
+    },
+    {
+      "epoch": 1.0582943954423927,
+      "grad_norm": 2.125,
+      "learning_rate": 7.217309211873951e-06,
+      "loss": 0.9649,
+      "step": 5373
+    },
+    {
+      "epoch": 1.0584942904974888,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.2163646475744775e-06,
+      "loss": 0.9941,
+      "step": 5374
+    },
+    {
+      "epoch": 1.0586941855525849,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.215419984824325e-06,
+      "loss": 1.0115,
+      "step": 5375
+    },
+    {
+      "epoch": 1.058894080607681,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.214475223665452e-06,
+      "loss": 1.0192,
+      "step": 5376
+    },
+    {
+      "epoch": 1.059093975662777,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.213530364139826e-06,
+      "loss": 1.0203,
+      "step": 5377
+    },
+    {
+      "epoch": 1.0592938707178732,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.2125854062894184e-06,
+      "loss": 0.9952,
+      "step": 5378
+    },
+    {
+      "epoch": 1.0594937657729693,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.211640350156203e-06,
+      "loss": 1.0367,
+      "step": 5379
+    },
+    {
+      "epoch": 1.0596936608280654,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.21069519578216e-06,
+      "loss": 0.9191,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0598935558831613,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.209749943209273e-06,
+      "loss": 0.983,
+      "step": 5381
+    },
+    {
+      "epoch": 1.0600934509382574,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.208804592479528e-06,
+      "loss": 0.9328,
+      "step": 5382
+    },
+    {
+      "epoch": 1.0602933459933535,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.207859143634919e-06,
+      "loss": 0.9217,
+      "step": 5383
+    },
+    {
+      "epoch": 1.0604932410484496,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.206913596717444e-06,
+      "loss": 1.083,
+      "step": 5384
+    },
+    {
+      "epoch": 1.0606931361035457,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.205967951769101e-06,
+      "loss": 1.018,
+      "step": 5385
+    },
+    {
+      "epoch": 1.0608930311586418,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.2050222088318996e-06,
+      "loss": 0.9879,
+      "step": 5386
+    },
+    {
+      "epoch": 1.0610929262137379,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.204076367947846e-06,
+      "loss": 1.0441,
+      "step": 5387
+    },
+    {
+      "epoch": 1.061292821268834,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.203130429158954e-06,
+      "loss": 0.9985,
+      "step": 5388
+    },
+    {
+      "epoch": 1.0614927163239298,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.202184392507245e-06,
+      "loss": 0.9014,
+      "step": 5389
+    },
+    {
+      "epoch": 1.061692611379026,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.2012382580347405e-06,
+      "loss": 0.9595,
+      "step": 5390
+    },
+    {
+      "epoch": 1.061892506434122,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.200292025783467e-06,
+      "loss": 1.0465,
+      "step": 5391
+    },
+    {
+      "epoch": 1.0620924014892181,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.199345695795458e-06,
+      "loss": 0.9302,
+      "step": 5392
+    },
+    {
+      "epoch": 1.0622922965443142,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.198399268112747e-06,
+      "loss": 0.9104,
+      "step": 5393
+    },
+    {
+      "epoch": 1.0624921915994103,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.197452742777376e-06,
+      "loss": 0.9866,
+      "step": 5394
+    },
+    {
+      "epoch": 1.0626920866545064,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.196506119831388e-06,
+      "loss": 0.8295,
+      "step": 5395
+    },
+    {
+      "epoch": 1.0628919817096025,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.1955593993168335e-06,
+      "loss": 0.979,
+      "step": 5396
+    },
+    {
+      "epoch": 1.0630918767646986,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.194612581275765e-06,
+      "loss": 0.9389,
+      "step": 5397
+    },
+    {
+      "epoch": 1.0632917718197945,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.1936656657502405e-06,
+      "loss": 0.972,
+      "step": 5398
+    },
+    {
+      "epoch": 1.0634916668748906,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.19271865278232e-06,
+      "loss": 0.9406,
+      "step": 5399
+    },
+    {
+      "epoch": 1.0636915619299867,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.1917715424140736e-06,
+      "loss": 1.0155,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0638914569850828,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.190824334687567e-06,
+      "loss": 0.9612,
+      "step": 5401
+    },
+    {
+      "epoch": 1.064091352040179,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.1898770296448775e-06,
+      "loss": 0.9834,
+      "step": 5402
+    },
+    {
+      "epoch": 1.064291247095275,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.188929627328085e-06,
+      "loss": 0.9918,
+      "step": 5403
+    },
+    {
+      "epoch": 1.064491142150371,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.187982127779272e-06,
+      "loss": 0.8852,
+      "step": 5404
+    },
+    {
+      "epoch": 1.0646910372054672,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.187034531040526e-06,
+      "loss": 0.9535,
+      "step": 5405
+    },
+    {
+      "epoch": 1.0648909322605633,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.186086837153941e-06,
+      "loss": 0.9757,
+      "step": 5406
+    },
+    {
+      "epoch": 1.0650908273156592,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.185139046161611e-06,
+      "loss": 0.9962,
+      "step": 5407
+    },
+    {
+      "epoch": 1.0652907223707553,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.184191158105639e-06,
+      "loss": 1.019,
+      "step": 5408
+    },
+    {
+      "epoch": 1.0654906174258514,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.183243173028128e-06,
+      "loss": 0.9948,
+      "step": 5409
+    },
+    {
+      "epoch": 1.0656905124809475,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.182295090971189e-06,
+      "loss": 0.947,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0658904075360436,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.181346911976935e-06,
+      "loss": 0.9602,
+      "step": 5411
+    },
+    {
+      "epoch": 1.0660903025911397,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.180398636087485e-06,
+      "loss": 0.9464,
+      "step": 5412
+    },
+    {
+      "epoch": 1.0662901976462358,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.179450263344959e-06,
+      "loss": 1.0616,
+      "step": 5413
+    },
+    {
+      "epoch": 1.0664900927013319,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.178501793791487e-06,
+      "loss": 0.9518,
+      "step": 5414
+    },
+    {
+      "epoch": 1.066689987756428,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.1775532274691965e-06,
+      "loss": 0.895,
+      "step": 5415
+    },
+    {
+      "epoch": 1.0668898828115239,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.176604564420224e-06,
+      "loss": 0.9706,
+      "step": 5416
+    },
+    {
+      "epoch": 1.06708977786662,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.17565580468671e-06,
+      "loss": 0.976,
+      "step": 5417
+    },
+    {
+      "epoch": 1.067289672921716,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.174706948310797e-06,
+      "loss": 0.9806,
+      "step": 5418
+    },
+    {
+      "epoch": 1.0674895679768122,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.173757995334634e-06,
+      "loss": 1.0118,
+      "step": 5419
+    },
+    {
+      "epoch": 1.0676894630319083,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.172808945800372e-06,
+      "loss": 0.9701,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0678893580870044,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.171859799750169e-06,
+      "loss": 0.9854,
+      "step": 5421
+    },
+    {
+      "epoch": 1.0680892531421005,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.170910557226186e-06,
+      "loss": 0.9865,
+      "step": 5422
+    },
+    {
+      "epoch": 1.0682891481971966,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.1699612182705894e-06,
+      "loss": 1.0558,
+      "step": 5423
+    },
+    {
+      "epoch": 1.0684890432522924,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.169011782925545e-06,
+      "loss": 0.9415,
+      "step": 5424
+    },
+    {
+      "epoch": 1.0686889383073885,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.16806225123323e-06,
+      "loss": 1.0717,
+      "step": 5425
+    },
+    {
+      "epoch": 1.0688888333624846,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.167112623235821e-06,
+      "loss": 0.8841,
+      "step": 5426
+    },
+    {
+      "epoch": 1.0690887284175807,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.1661628989755016e-06,
+      "loss": 0.9353,
+      "step": 5427
+    },
+    {
+      "epoch": 1.0692886234726768,
+      "grad_norm": 2.25,
+      "learning_rate": 7.165213078494456e-06,
+      "loss": 1.0364,
+      "step": 5428
+    },
+    {
+      "epoch": 1.069488518527773,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.164263161834879e-06,
+      "loss": 0.9237,
+      "step": 5429
+    },
+    {
+      "epoch": 1.069688413582869,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.163313149038962e-06,
+      "loss": 0.9851,
+      "step": 5430
+    },
+    {
+      "epoch": 1.0698883086379651,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.162363040148905e-06,
+      "loss": 1.0218,
+      "step": 5431
+    },
+    {
+      "epoch": 1.0700882036930612,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.161412835206915e-06,
+      "loss": 0.9518,
+      "step": 5432
+    },
+    {
+      "epoch": 1.0702880987481573,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.160462534255195e-06,
+      "loss": 1.0077,
+      "step": 5433
+    },
+    {
+      "epoch": 1.0704879938032532,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.159512137335962e-06,
+      "loss": 0.9244,
+      "step": 5434
+    },
+    {
+      "epoch": 1.0706878888583493,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.15856164449143e-06,
+      "loss": 1.0028,
+      "step": 5435
+    },
+    {
+      "epoch": 1.0708877839134454,
+      "grad_norm": 2.125,
+      "learning_rate": 7.15761105576382e-06,
+      "loss": 1.0343,
+      "step": 5436
+    },
+    {
+      "epoch": 1.0710876789685415,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.156660371195357e-06,
+      "loss": 0.9361,
+      "step": 5437
+    },
+    {
+      "epoch": 1.0712875740236376,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.155709590828271e-06,
+      "loss": 0.986,
+      "step": 5438
+    },
+    {
+      "epoch": 1.0714874690787337,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.154758714704797e-06,
+      "loss": 1.0873,
+      "step": 5439
+    },
+    {
+      "epoch": 1.0716873641338298,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.153807742867169e-06,
+      "loss": 0.9848,
+      "step": 5440
+    },
+    {
+      "epoch": 1.071887259188926,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.152856675357631e-06,
+      "loss": 0.9704,
+      "step": 5441
+    },
+    {
+      "epoch": 1.0720871542440218,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.15190551221843e-06,
+      "loss": 0.9277,
+      "step": 5442
+    },
+    {
+      "epoch": 1.072287049299118,
+      "grad_norm": 2.25,
+      "learning_rate": 7.150954253491818e-06,
+      "loss": 0.9952,
+      "step": 5443
+    },
+    {
+      "epoch": 1.072486944354214,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.1500028992200445e-06,
+      "loss": 0.9912,
+      "step": 5444
+    },
+    {
+      "epoch": 1.07268683940931,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.1490514494453736e-06,
+      "loss": 0.964,
+      "step": 5445
+    },
+    {
+      "epoch": 1.0728867344644062,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.148099904210067e-06,
+      "loss": 1.1416,
+      "step": 5446
+    },
+    {
+      "epoch": 1.0730866295195023,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.14714826355639e-06,
+      "loss": 0.927,
+      "step": 5447
+    },
+    {
+      "epoch": 1.0732865245745984,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.146196527526617e-06,
+      "loss": 0.9822,
+      "step": 5448
+    },
+    {
+      "epoch": 1.0734864196296945,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.145244696163025e-06,
+      "loss": 1.0945,
+      "step": 5449
+    },
+    {
+      "epoch": 1.0736863146847906,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.144292769507891e-06,
+      "loss": 0.9773,
+      "step": 5450
+    },
+    {
+      "epoch": 1.0738862097398865,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.143340747603503e-06,
+      "loss": 0.9519,
+      "step": 5451
+    },
+    {
+      "epoch": 1.0740861047949826,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.142388630492147e-06,
+      "loss": 0.9781,
+      "step": 5452
+    },
+    {
+      "epoch": 1.0742859998500787,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.141436418216118e-06,
+      "loss": 0.954,
+      "step": 5453
+    },
+    {
+      "epoch": 1.0744858949051748,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.14048411081771e-06,
+      "loss": 0.8944,
+      "step": 5454
+    },
+    {
+      "epoch": 1.0746857899602709,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.139531708339227e-06,
+      "loss": 0.9898,
+      "step": 5455
+    },
+    {
+      "epoch": 1.074885685015367,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.138579210822976e-06,
+      "loss": 0.9674,
+      "step": 5456
+    },
+    {
+      "epoch": 1.075085580070463,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.137626618311262e-06,
+      "loss": 0.9289,
+      "step": 5457
+    },
+    {
+      "epoch": 1.0752854751255592,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.136673930846404e-06,
+      "loss": 0.9659,
+      "step": 5458
+    },
+    {
+      "epoch": 1.075485370180655,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.135721148470718e-06,
+      "loss": 1.0642,
+      "step": 5459
+    },
+    {
+      "epoch": 1.0756852652357511,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.134768271226525e-06,
+      "loss": 1.0505,
+      "step": 5460
+    },
+    {
+      "epoch": 1.0758851602908472,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.133815299156155e-06,
+      "loss": 0.9973,
+      "step": 5461
+    },
+    {
+      "epoch": 1.0760850553459433,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.132862232301937e-06,
+      "loss": 0.9298,
+      "step": 5462
+    },
+    {
+      "epoch": 1.0762849504010394,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.1319090707062065e-06,
+      "loss": 0.9508,
+      "step": 5463
+    },
+    {
+      "epoch": 1.0764848454561355,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.130955814411302e-06,
+      "loss": 1.013,
+      "step": 5464
+    },
+    {
+      "epoch": 1.0766847405112316,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.130002463459569e-06,
+      "loss": 1.0727,
+      "step": 5465
+    },
+    {
+      "epoch": 1.0768846355663277,
+      "grad_norm": 2.125,
+      "learning_rate": 7.129049017893352e-06,
+      "loss": 0.9824,
+      "step": 5466
+    },
+    {
+      "epoch": 1.0770845306214238,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.128095477755006e-06,
+      "loss": 0.8751,
+      "step": 5467
+    },
+    {
+      "epoch": 1.07728442567652,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.127141843086888e-06,
+      "loss": 0.9525,
+      "step": 5468
+    },
+    {
+      "epoch": 1.0774843207316158,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.126188113931353e-06,
+      "loss": 0.9577,
+      "step": 5469
+    },
+    {
+      "epoch": 1.077684215786712,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.125234290330774e-06,
+      "loss": 1.0297,
+      "step": 5470
+    },
+    {
+      "epoch": 1.077884110841808,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.124280372327511e-06,
+      "loss": 0.9516,
+      "step": 5471
+    },
+    {
+      "epoch": 1.0780840058969041,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.123326359963941e-06,
+      "loss": 0.978,
+      "step": 5472
+    },
+    {
+      "epoch": 1.0782839009520002,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.122372253282442e-06,
+      "loss": 0.924,
+      "step": 5473
+    },
+    {
+      "epoch": 1.0784837960070963,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.121418052325395e-06,
+      "loss": 0.9747,
+      "step": 5474
+    },
+    {
+      "epoch": 1.0786836910621924,
+      "grad_norm": 2.125,
+      "learning_rate": 7.1204637571351835e-06,
+      "loss": 0.9425,
+      "step": 5475
+    },
+    {
+      "epoch": 1.0788835861172885,
+      "grad_norm": 2.125,
+      "learning_rate": 7.119509367754198e-06,
+      "loss": 0.8835,
+      "step": 5476
+    },
+    {
+      "epoch": 1.0790834811723844,
+      "grad_norm": 2.125,
+      "learning_rate": 7.118554884224833e-06,
+      "loss": 0.9955,
+      "step": 5477
+    },
+    {
+      "epoch": 1.0792833762274805,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.117600306589486e-06,
+      "loss": 0.9089,
+      "step": 5478
+    },
+    {
+      "epoch": 1.0794832712825766,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.11664563489056e-06,
+      "loss": 0.9591,
+      "step": 5479
+    },
+    {
+      "epoch": 1.0796831663376727,
+      "grad_norm": 2.25,
+      "learning_rate": 7.1156908691704604e-06,
+      "loss": 1.0151,
+      "step": 5480
+    },
+    {
+      "epoch": 1.0798830613927688,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.114736009471599e-06,
+      "loss": 1.1079,
+      "step": 5481
+    },
+    {
+      "epoch": 1.080082956447865,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.113781055836391e-06,
+      "loss": 0.9985,
+      "step": 5482
+    },
+    {
+      "epoch": 1.080282851502961,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.112826008307252e-06,
+      "loss": 1.023,
+      "step": 5483
+    },
+    {
+      "epoch": 1.080482746558057,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.111870866926609e-06,
+      "loss": 1.0379,
+      "step": 5484
+    },
+    {
+      "epoch": 1.0806826416131532,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.110915631736887e-06,
+      "loss": 0.9583,
+      "step": 5485
+    },
+    {
+      "epoch": 1.080882536668249,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.109960302780518e-06,
+      "loss": 1.0009,
+      "step": 5486
+    },
+    {
+      "epoch": 1.0810824317233452,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.109004880099938e-06,
+      "loss": 0.9799,
+      "step": 5487
+    },
+    {
+      "epoch": 1.0812823267784413,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.108049363737586e-06,
+      "loss": 0.9076,
+      "step": 5488
+    },
+    {
+      "epoch": 1.0814822218335374,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.107093753735907e-06,
+      "loss": 0.9846,
+      "step": 5489
+    },
+    {
+      "epoch": 1.0816821168886335,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.106138050137349e-06,
+      "loss": 1.0191,
+      "step": 5490
+    },
+    {
+      "epoch": 1.0818820119437296,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.105182252984363e-06,
+      "loss": 0.9635,
+      "step": 5491
+    },
+    {
+      "epoch": 1.0820819069988257,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.104226362319405e-06,
+      "loss": 0.8561,
+      "step": 5492
+    },
+    {
+      "epoch": 1.0822818020539218,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.103270378184939e-06,
+      "loss": 0.9057,
+      "step": 5493
+    },
+    {
+      "epoch": 1.0824816971090176,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.102314300623425e-06,
+      "loss": 0.9439,
+      "step": 5494
+    },
+    {
+      "epoch": 1.0826815921641137,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.101358129677336e-06,
+      "loss": 0.9405,
+      "step": 5495
+    },
+    {
+      "epoch": 1.0828814872192098,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.100401865389144e-06,
+      "loss": 0.9559,
+      "step": 5496
+    },
+    {
+      "epoch": 1.083081382274306,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.099445507801324e-06,
+      "loss": 0.9675,
+      "step": 5497
+    },
+    {
+      "epoch": 1.083281277329402,
+      "grad_norm": 2.25,
+      "learning_rate": 7.0984890569563595e-06,
+      "loss": 0.9406,
+      "step": 5498
+    },
+    {
+      "epoch": 1.0834811723844981,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.097532512896734e-06,
+      "loss": 1.0121,
+      "step": 5499
+    },
+    {
+      "epoch": 1.0836810674395942,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.096575875664939e-06,
+      "loss": 0.9781,
+      "step": 5500
+    },
+    {
+      "epoch": 1.0838809624946903,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.095619145303469e-06,
+      "loss": 1.0666,
+      "step": 5501
+    },
+    {
+      "epoch": 1.0840808575497864,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.094662321854818e-06,
+      "loss": 0.9576,
+      "step": 5502
+    },
+    {
+      "epoch": 1.0842807526048825,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.09370540536149e-06,
+      "loss": 0.9628,
+      "step": 5503
+    },
+    {
+      "epoch": 1.0844806476599784,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.092748395865995e-06,
+      "loss": 0.8464,
+      "step": 5504
+    },
+    {
+      "epoch": 1.0846805427150745,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.091791293410838e-06,
+      "loss": 1.0179,
+      "step": 5505
+    },
+    {
+      "epoch": 1.0848804377701706,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.090834098038535e-06,
+      "loss": 1.0201,
+      "step": 5506
+    },
+    {
+      "epoch": 1.0850803328252667,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.0898768097916045e-06,
+      "loss": 0.9504,
+      "step": 5507
+    },
+    {
+      "epoch": 1.0852802278803628,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.08891942871257e-06,
+      "loss": 0.8916,
+      "step": 5508
+    },
+    {
+      "epoch": 1.085480122935459,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.087961954843956e-06,
+      "loss": 0.9202,
+      "step": 5509
+    },
+    {
+      "epoch": 1.085680017990555,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.087004388228297e-06,
+      "loss": 0.9722,
+      "step": 5510
+    },
+    {
+      "epoch": 1.0858799130456511,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.086046728908125e-06,
+      "loss": 1.0817,
+      "step": 5511
+    },
+    {
+      "epoch": 1.086079808100747,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.085088976925979e-06,
+      "loss": 0.8969,
+      "step": 5512
+    },
+    {
+      "epoch": 1.086279703155843,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.084131132324405e-06,
+      "loss": 0.9292,
+      "step": 5513
+    },
+    {
+      "epoch": 1.0864795982109392,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.083173195145947e-06,
+      "loss": 0.8921,
+      "step": 5514
+    },
+    {
+      "epoch": 1.0866794932660353,
+      "grad_norm": 2.125,
+      "learning_rate": 7.08221516543316e-06,
+      "loss": 0.9571,
+      "step": 5515
+    },
+    {
+      "epoch": 1.0868793883211314,
+      "grad_norm": 2.125,
+      "learning_rate": 7.081257043228597e-06,
+      "loss": 1.0,
+      "step": 5516
+    },
+    {
+      "epoch": 1.0870792833762275,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.080298828574818e-06,
+      "loss": 0.9582,
+      "step": 5517
+    },
+    {
+      "epoch": 1.0872791784313236,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.079340521514389e-06,
+      "loss": 0.9338,
+      "step": 5518
+    },
+    {
+      "epoch": 1.0874790734864197,
+      "grad_norm": 2.125,
+      "learning_rate": 7.078382122089873e-06,
+      "loss": 1.0542,
+      "step": 5519
+    },
+    {
+      "epoch": 1.0876789685415158,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.0774236303438485e-06,
+      "loss": 0.9537,
+      "step": 5520
+    },
+    {
+      "epoch": 1.0878788635966117,
+      "grad_norm": 2.25,
+      "learning_rate": 7.076465046318886e-06,
+      "loss": 0.903,
+      "step": 5521
+    },
+    {
+      "epoch": 1.0880787586517078,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.075506370057569e-06,
+      "loss": 0.9471,
+      "step": 5522
+    },
+    {
+      "epoch": 1.0882786537068039,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.074547601602479e-06,
+      "loss": 0.96,
+      "step": 5523
+    },
+    {
+      "epoch": 1.0884785487619,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.073588740996208e-06,
+      "loss": 0.9324,
+      "step": 5524
+    },
+    {
+      "epoch": 1.088678443816996,
+      "grad_norm": 2.125,
+      "learning_rate": 7.072629788281345e-06,
+      "loss": 1.002,
+      "step": 5525
+    },
+    {
+      "epoch": 1.0888783388720922,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.07167074350049e-06,
+      "loss": 0.9232,
+      "step": 5526
+    },
+    {
+      "epoch": 1.0890782339271883,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.0707116066962415e-06,
+      "loss": 0.9671,
+      "step": 5527
+    },
+    {
+      "epoch": 1.0892781289822844,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.069752377911203e-06,
+      "loss": 0.9804,
+      "step": 5528
+    },
+    {
+      "epoch": 1.0894780240373805,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.068793057187986e-06,
+      "loss": 0.8772,
+      "step": 5529
+    },
+    {
+      "epoch": 1.0896779190924764,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.067833644569202e-06,
+      "loss": 1.0187,
+      "step": 5530
+    },
+    {
+      "epoch": 1.0898778141475725,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.066874140097468e-06,
+      "loss": 0.9763,
+      "step": 5531
+    },
+    {
+      "epoch": 1.0900777092026686,
+      "grad_norm": 2.125,
+      "learning_rate": 7.065914543815408e-06,
+      "loss": 0.9646,
+      "step": 5532
+    },
+    {
+      "epoch": 1.0902776042577647,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.064954855765641e-06,
+      "loss": 0.9707,
+      "step": 5533
+    },
+    {
+      "epoch": 1.0904774993128608,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.063995075990801e-06,
+      "loss": 1.0343,
+      "step": 5534
+    },
+    {
+      "epoch": 1.0906773943679569,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.063035204533522e-06,
+      "loss": 0.9079,
+      "step": 5535
+    },
+    {
+      "epoch": 1.090877289423053,
+      "grad_norm": 2.0,
+      "learning_rate": 7.062075241436439e-06,
+      "loss": 0.9172,
+      "step": 5536
+    },
+    {
+      "epoch": 1.091077184478149,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.061115186742192e-06,
+      "loss": 1.0221,
+      "step": 5537
+    },
+    {
+      "epoch": 1.0912770795332452,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.060155040493431e-06,
+      "loss": 0.9661,
+      "step": 5538
+    },
+    {
+      "epoch": 1.091476974588341,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.059194802732802e-06,
+      "loss": 1.0385,
+      "step": 5539
+    },
+    {
+      "epoch": 1.0916768696434371,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.0582344735029585e-06,
+      "loss": 0.9009,
+      "step": 5540
+    },
+    {
+      "epoch": 1.0918767646985332,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.0572740528465625e-06,
+      "loss": 0.9973,
+      "step": 5541
+    },
+    {
+      "epoch": 1.0920766597536293,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.0563135408062696e-06,
+      "loss": 1.022,
+      "step": 5542
+    },
+    {
+      "epoch": 1.0922765548087254,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.055352937424751e-06,
+      "loss": 0.947,
+      "step": 5543
+    },
+    {
+      "epoch": 1.0924764498638215,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.0543922427446734e-06,
+      "loss": 0.9117,
+      "step": 5544
+    },
+    {
+      "epoch": 1.0926763449189176,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.053431456808712e-06,
+      "loss": 0.9254,
+      "step": 5545
+    },
+    {
+      "epoch": 1.0928762399740137,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.052470579659545e-06,
+      "loss": 1.0449,
+      "step": 5546
+    },
+    {
+      "epoch": 1.0930761350291096,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.051509611339853e-06,
+      "loss": 0.9985,
+      "step": 5547
+    },
+    {
+      "epoch": 1.0932760300842057,
+      "grad_norm": 2.125,
+      "learning_rate": 7.050548551892325e-06,
+      "loss": 0.9993,
+      "step": 5548
+    },
+    {
+      "epoch": 1.0934759251393018,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.049587401359647e-06,
+      "loss": 0.9773,
+      "step": 5549
+    },
+    {
+      "epoch": 1.093675820194398,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.048626159784517e-06,
+      "loss": 1.0189,
+      "step": 5550
+    },
+    {
+      "epoch": 1.093875715249494,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.047664827209633e-06,
+      "loss": 0.9613,
+      "step": 5551
+    },
+    {
+      "epoch": 1.09407561030459,
+      "grad_norm": 2.125,
+      "learning_rate": 7.0467034036776945e-06,
+      "loss": 0.938,
+      "step": 5552
+    },
+    {
+      "epoch": 1.0942755053596862,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.04574188923141e-06,
+      "loss": 0.9479,
+      "step": 5553
+    },
+    {
+      "epoch": 1.0944754004147823,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.044780283913488e-06,
+      "loss": 1.0025,
+      "step": 5554
+    },
+    {
+      "epoch": 1.0946752954698784,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.043818587766645e-06,
+      "loss": 1.0156,
+      "step": 5555
+    },
+    {
+      "epoch": 1.0948751905249743,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.0428568008336e-06,
+      "loss": 1.1055,
+      "step": 5556
+    },
+    {
+      "epoch": 1.0950750855800704,
+      "grad_norm": 2.375,
+      "learning_rate": 7.041894923157071e-06,
+      "loss": 1.0323,
+      "step": 5557
+    },
+    {
+      "epoch": 1.0952749806351665,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.04093295477979e-06,
+      "loss": 1.035,
+      "step": 5558
+    },
+    {
+      "epoch": 1.0954748756902626,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.039970895744485e-06,
+      "loss": 1.0006,
+      "step": 5559
+    },
+    {
+      "epoch": 1.0956747707453587,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.039008746093889e-06,
+      "loss": 1.0166,
+      "step": 5560
+    },
+    {
+      "epoch": 1.0958746658004548,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.038046505870744e-06,
+      "loss": 0.9481,
+      "step": 5561
+    },
+    {
+      "epoch": 1.0960745608555509,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.03708417511779e-06,
+      "loss": 0.9159,
+      "step": 5562
+    },
+    {
+      "epoch": 1.096274455910647,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.036121753877776e-06,
+      "loss": 1.0279,
+      "step": 5563
+    },
+    {
+      "epoch": 1.096474350965743,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.035159242193449e-06,
+      "loss": 1.049,
+      "step": 5564
+    },
+    {
+      "epoch": 1.096674246020839,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.0341966401075664e-06,
+      "loss": 0.9342,
+      "step": 5565
+    },
+    {
+      "epoch": 1.096874141075935,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.033233947662887e-06,
+      "loss": 0.8556,
+      "step": 5566
+    },
+    {
+      "epoch": 1.0970740361310312,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.0322711649021735e-06,
+      "loss": 0.9793,
+      "step": 5567
+    },
+    {
+      "epoch": 1.0972739311861273,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.031308291868191e-06,
+      "loss": 1.0418,
+      "step": 5568
+    },
+    {
+      "epoch": 1.0974738262412234,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.030345328603711e-06,
+      "loss": 0.9986,
+      "step": 5569
+    },
+    {
+      "epoch": 1.0976737212963195,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.029382275151508e-06,
+      "loss": 0.9284,
+      "step": 5570
+    },
+    {
+      "epoch": 1.0978736163514156,
+      "grad_norm": 2.125,
+      "learning_rate": 7.0284191315543606e-06,
+      "loss": 1.0405,
+      "step": 5571
+    },
+    {
+      "epoch": 1.0980735114065117,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.027455897855053e-06,
+      "loss": 1.0036,
+      "step": 5572
+    },
+    {
+      "epoch": 1.0982734064616078,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.026492574096371e-06,
+      "loss": 0.999,
+      "step": 5573
+    },
+    {
+      "epoch": 1.0984733015167036,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.025529160321107e-06,
+      "loss": 0.9254,
+      "step": 5574
+    },
+    {
+      "epoch": 1.0986731965717997,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.024565656572051e-06,
+      "loss": 1.0104,
+      "step": 5575
+    },
+    {
+      "epoch": 1.0988730916268958,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.023602062892005e-06,
+      "loss": 1.0095,
+      "step": 5576
+    },
+    {
+      "epoch": 1.099072986681992,
+      "grad_norm": 2.125,
+      "learning_rate": 7.022638379323774e-06,
+      "loss": 0.9809,
+      "step": 5577
+    },
+    {
+      "epoch": 1.099272881737088,
+      "grad_norm": 2.125,
+      "learning_rate": 7.021674605910161e-06,
+      "loss": 0.9851,
+      "step": 5578
+    },
+    {
+      "epoch": 1.0994727767921841,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.0207107426939755e-06,
+      "loss": 0.9673,
+      "step": 5579
+    },
+    {
+      "epoch": 1.0996726718472802,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.019746789718038e-06,
+      "loss": 0.9563,
+      "step": 5580
+    },
+    {
+      "epoch": 1.0998725669023763,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.018782747025161e-06,
+      "loss": 0.9989,
+      "step": 5581
+    },
+    {
+      "epoch": 1.1000724619574722,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.0178186146581725e-06,
+      "loss": 0.9575,
+      "step": 5582
+    },
+    {
+      "epoch": 1.1002723570125683,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.0168543926598965e-06,
+      "loss": 0.9197,
+      "step": 5583
+    },
+    {
+      "epoch": 1.1004722520676644,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.015890081073162e-06,
+      "loss": 0.9936,
+      "step": 5584
+    },
+    {
+      "epoch": 1.1006721471227605,
+      "grad_norm": 2.0,
+      "learning_rate": 7.014925679940807e-06,
+      "loss": 0.8853,
+      "step": 5585
+    },
+    {
+      "epoch": 1.1008720421778566,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.013961189305668e-06,
+      "loss": 1.0785,
+      "step": 5586
+    },
+    {
+      "epoch": 1.1010719372329527,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.012996609210587e-06,
+      "loss": 0.9489,
+      "step": 5587
+    },
+    {
+      "epoch": 1.1012718322880488,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.012031939698414e-06,
+      "loss": 0.9351,
+      "step": 5588
+    },
+    {
+      "epoch": 1.101471727343145,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.011067180811994e-06,
+      "loss": 0.9849,
+      "step": 5589
+    },
+    {
+      "epoch": 1.101671622398241,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.010102332594186e-06,
+      "loss": 1.059,
+      "step": 5590
+    },
+    {
+      "epoch": 1.1018715174533371,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.009137395087848e-06,
+      "loss": 1.0314,
+      "step": 5591
+    },
+    {
+      "epoch": 1.102071412508433,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.008172368335842e-06,
+      "loss": 0.8892,
+      "step": 5592
+    },
+    {
+      "epoch": 1.102271307563529,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.007207252381032e-06,
+      "loss": 0.9998,
+      "step": 5593
+    },
+    {
+      "epoch": 1.1024712026186252,
+      "grad_norm": 2.25,
+      "learning_rate": 7.006242047266292e-06,
+      "loss": 0.9962,
+      "step": 5594
+    },
+    {
+      "epoch": 1.1026710976737213,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.005276753034492e-06,
+      "loss": 0.9656,
+      "step": 5595
+    },
+    {
+      "epoch": 1.1028709927288174,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.004311369728514e-06,
+      "loss": 1.0528,
+      "step": 5596
+    },
+    {
+      "epoch": 1.1030708877839135,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.003345897391241e-06,
+      "loss": 0.94,
+      "step": 5597
+    },
+    {
+      "epoch": 1.1032707828390096,
+      "grad_norm": 2.25,
+      "learning_rate": 7.002380336065555e-06,
+      "loss": 1.0038,
+      "step": 5598
+    },
+    {
+      "epoch": 1.1034706778941057,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.0014146857943486e-06,
+      "loss": 0.9046,
+      "step": 5599
+    },
+    {
+      "epoch": 1.1036705729492016,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.000448946620517e-06,
+      "loss": 1.0184,
+      "step": 5600
+    },
+    {
+      "epoch": 1.1038704680042977,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.999483118586955e-06,
+      "loss": 0.9515,
+      "step": 5601
+    },
+    {
+      "epoch": 1.1040703630593938,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.998517201736566e-06,
+      "loss": 0.9786,
+      "step": 5602
+    },
+    {
+      "epoch": 1.1042702581144899,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.997551196112259e-06,
+      "loss": 1.0022,
+      "step": 5603
+    },
+    {
+      "epoch": 1.104470153169586,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.996585101756938e-06,
+      "loss": 0.9757,
+      "step": 5604
+    },
+    {
+      "epoch": 1.104670048224682,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.9956189187135226e-06,
+      "loss": 0.8901,
+      "step": 5605
+    },
+    {
+      "epoch": 1.1048699432797782,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.994652647024927e-06,
+      "loss": 0.9721,
+      "step": 5606
+    },
+    {
+      "epoch": 1.1050698383348743,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.9936862867340736e-06,
+      "loss": 1.022,
+      "step": 5607
+    },
+    {
+      "epoch": 1.1052697333899704,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.99271983788389e-06,
+      "loss": 1.0285,
+      "step": 5608
+    },
+    {
+      "epoch": 1.1054696284450662,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.991753300517302e-06,
+      "loss": 1.0873,
+      "step": 5609
+    },
+    {
+      "epoch": 1.1056695235001623,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.990786674677246e-06,
+      "loss": 0.9022,
+      "step": 5610
+    },
+    {
+      "epoch": 1.1058694185552584,
+      "grad_norm": 1.96875,
+      "learning_rate": 6.989819960406659e-06,
+      "loss": 0.8544,
+      "step": 5611
+    },
+    {
+      "epoch": 1.1060693136103545,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.988853157748479e-06,
+      "loss": 0.9286,
+      "step": 5612
+    },
+    {
+      "epoch": 1.1062692086654506,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.987886266745658e-06,
+      "loss": 0.9779,
+      "step": 5613
+    },
+    {
+      "epoch": 1.1064691037205467,
+      "grad_norm": 2.125,
+      "learning_rate": 6.986919287441141e-06,
+      "loss": 0.8984,
+      "step": 5614
+    },
+    {
+      "epoch": 1.1066689987756428,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.985952219877879e-06,
+      "loss": 0.9477,
+      "step": 5615
+    },
+    {
+      "epoch": 1.106868893830739,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.984985064098833e-06,
+      "loss": 1.0035,
+      "step": 5616
+    },
+    {
+      "epoch": 1.1070687888858348,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.984017820146962e-06,
+      "loss": 0.8997,
+      "step": 5617
+    },
+    {
+      "epoch": 1.107268683940931,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.983050488065231e-06,
+      "loss": 1.0193,
+      "step": 5618
+    },
+    {
+      "epoch": 1.107468578996027,
+      "grad_norm": 1.9921875,
+      "learning_rate": 6.98208306789661e-06,
+      "loss": 0.9445,
+      "step": 5619
+    },
+    {
+      "epoch": 1.1076684740511231,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.98111555968407e-06,
+      "loss": 0.9414,
+      "step": 5620
+    },
+    {
+      "epoch": 1.1078683691062192,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.9801479634705885e-06,
+      "loss": 1.0249,
+      "step": 5621
+    },
+    {
+      "epoch": 1.1080682641613153,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.979180279299147e-06,
+      "loss": 0.9569,
+      "step": 5622
+    },
+    {
+      "epoch": 1.1082681592164114,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.978212507212727e-06,
+      "loss": 1.0551,
+      "step": 5623
+    },
+    {
+      "epoch": 1.1084680542715075,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.97724464725432e-06,
+      "loss": 0.875,
+      "step": 5624
+    },
+    {
+      "epoch": 1.1086679493266036,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.9762766994669175e-06,
+      "loss": 0.9271,
+      "step": 5625
+    },
+    {
+      "epoch": 1.1088678443816997,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.9753086638935144e-06,
+      "loss": 0.9936,
+      "step": 5626
+    },
+    {
+      "epoch": 1.1090677394367956,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.974340540577111e-06,
+      "loss": 0.9048,
+      "step": 5627
+    },
+    {
+      "epoch": 1.1092676344918917,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.973372329560713e-06,
+      "loss": 0.9632,
+      "step": 5628
+    },
+    {
+      "epoch": 1.1094675295469878,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.972404030887325e-06,
+      "loss": 0.8766,
+      "step": 5629
+    },
+    {
+      "epoch": 1.109667424602084,
+      "grad_norm": 2.125,
+      "learning_rate": 6.971435644599961e-06,
+      "loss": 0.8904,
+      "step": 5630
+    },
+    {
+      "epoch": 1.10986731965718,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.970467170741637e-06,
+      "loss": 0.9855,
+      "step": 5631
+    },
+    {
+      "epoch": 1.110067214712276,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.969498609355372e-06,
+      "loss": 0.9869,
+      "step": 5632
+    },
+    {
+      "epoch": 1.1102671097673722,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.968529960484189e-06,
+      "loss": 0.9859,
+      "step": 5633
+    },
+    {
+      "epoch": 1.1104670048224683,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.967561224171114e-06,
+      "loss": 0.9741,
+      "step": 5634
+    },
+    {
+      "epoch": 1.1106668998775642,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.96659240045918e-06,
+      "loss": 0.8432,
+      "step": 5635
+    },
+    {
+      "epoch": 1.1108667949326603,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.965623489391423e-06,
+      "loss": 1.0741,
+      "step": 5636
+    },
+    {
+      "epoch": 1.1110666899877564,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.964654491010879e-06,
+      "loss": 0.9718,
+      "step": 5637
+    },
+    {
+      "epoch": 1.1112665850428525,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.963685405360594e-06,
+      "loss": 1.0086,
+      "step": 5638
+    },
+    {
+      "epoch": 1.1114664800979486,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.962716232483612e-06,
+      "loss": 0.9935,
+      "step": 5639
+    },
+    {
+      "epoch": 1.1116663751530447,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.961746972422985e-06,
+      "loss": 0.9189,
+      "step": 5640
+    },
+    {
+      "epoch": 1.1118662702081408,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.960777625221765e-06,
+      "loss": 0.8785,
+      "step": 5641
+    },
+    {
+      "epoch": 1.1120661652632369,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.959808190923015e-06,
+      "loss": 1.0427,
+      "step": 5642
+    },
+    {
+      "epoch": 1.112266060318333,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.958838669569793e-06,
+      "loss": 0.944,
+      "step": 5643
+    },
+    {
+      "epoch": 1.1124659553734288,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.957869061205168e-06,
+      "loss": 0.9901,
+      "step": 5644
+    },
+    {
+      "epoch": 1.112665850428525,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.956899365872207e-06,
+      "loss": 0.9979,
+      "step": 5645
+    },
+    {
+      "epoch": 1.112865745483621,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.955929583613985e-06,
+      "loss": 1.0601,
+      "step": 5646
+    },
+    {
+      "epoch": 1.1130656405387171,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.95495971447358e-06,
+      "loss": 1.0979,
+      "step": 5647
+    },
+    {
+      "epoch": 1.1132655355938132,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.953989758494075e-06,
+      "loss": 0.9665,
+      "step": 5648
+    },
+    {
+      "epoch": 1.1134654306489093,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.953019715718552e-06,
+      "loss": 0.9663,
+      "step": 5649
+    },
+    {
+      "epoch": 1.1136653257040054,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.952049586190102e-06,
+      "loss": 1.0084,
+      "step": 5650
+    },
+    {
+      "epoch": 1.1138652207591015,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.951079369951817e-06,
+      "loss": 0.9027,
+      "step": 5651
+    },
+    {
+      "epoch": 1.1140651158141976,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.950109067046797e-06,
+      "loss": 0.9388,
+      "step": 5652
+    },
+    {
+      "epoch": 1.1142650108692935,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.949138677518139e-06,
+      "loss": 1.0136,
+      "step": 5653
+    },
+    {
+      "epoch": 1.1144649059243896,
+      "grad_norm": 2.125,
+      "learning_rate": 6.948168201408949e-06,
+      "loss": 0.9745,
+      "step": 5654
+    },
+    {
+      "epoch": 1.1146648009794857,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.947197638762335e-06,
+      "loss": 0.9844,
+      "step": 5655
+    },
+    {
+      "epoch": 1.1148646960345818,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.946226989621412e-06,
+      "loss": 0.9096,
+      "step": 5656
+    },
+    {
+      "epoch": 1.115064591089678,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.945256254029292e-06,
+      "loss": 1.0328,
+      "step": 5657
+    },
+    {
+      "epoch": 1.115264486144774,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.944285432029098e-06,
+      "loss": 0.9901,
+      "step": 5658
+    },
+    {
+      "epoch": 1.1154643811998701,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.943314523663953e-06,
+      "loss": 0.9223,
+      "step": 5659
+    },
+    {
+      "epoch": 1.1156642762549662,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.942343528976984e-06,
+      "loss": 0.9988,
+      "step": 5660
+    },
+    {
+      "epoch": 1.1158641713100623,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.9413724480113224e-06,
+      "loss": 0.9521,
+      "step": 5661
+    },
+    {
+      "epoch": 1.1160640663651582,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.940401280810105e-06,
+      "loss": 0.9889,
+      "step": 5662
+    },
+    {
+      "epoch": 1.1162639614202543,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.939430027416468e-06,
+      "loss": 1.0112,
+      "step": 5663
+    },
+    {
+      "epoch": 1.1164638564753504,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.938458687873558e-06,
+      "loss": 1.0181,
+      "step": 5664
+    },
+    {
+      "epoch": 1.1166637515304465,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.93748726222452e-06,
+      "loss": 0.9091,
+      "step": 5665
+    },
+    {
+      "epoch": 1.1168636465855426,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.936515750512505e-06,
+      "loss": 0.9331,
+      "step": 5666
+    },
+    {
+      "epoch": 1.1170635416406387,
+      "grad_norm": 2.25,
+      "learning_rate": 6.935544152780666e-06,
+      "loss": 1.0873,
+      "step": 5667
+    },
+    {
+      "epoch": 1.1172634366957348,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.934572469072163e-06,
+      "loss": 0.8997,
+      "step": 5668
+    },
+    {
+      "epoch": 1.117463331750831,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.933600699430157e-06,
+      "loss": 1.0442,
+      "step": 5669
+    },
+    {
+      "epoch": 1.1176632268059268,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.932628843897816e-06,
+      "loss": 0.9619,
+      "step": 5670
+    },
+    {
+      "epoch": 1.1178631218610229,
+      "grad_norm": 2.125,
+      "learning_rate": 6.931656902518307e-06,
+      "loss": 0.9943,
+      "step": 5671
+    },
+    {
+      "epoch": 1.118063016916119,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.930684875334806e-06,
+      "loss": 0.9998,
+      "step": 5672
+    },
+    {
+      "epoch": 1.118262911971215,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.929712762390487e-06,
+      "loss": 1.0107,
+      "step": 5673
+    },
+    {
+      "epoch": 1.1184628070263112,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.928740563728533e-06,
+      "loss": 1.0266,
+      "step": 5674
+    },
+    {
+      "epoch": 1.1186627020814073,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.927768279392132e-06,
+      "loss": 0.9991,
+      "step": 5675
+    },
+    {
+      "epoch": 1.1188625971365034,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.926795909424468e-06,
+      "loss": 1.0663,
+      "step": 5676
+    },
+    {
+      "epoch": 1.1190624921915995,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.9258234538687366e-06,
+      "loss": 0.9595,
+      "step": 5677
+    },
+    {
+      "epoch": 1.1192623872466956,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.924850912768133e-06,
+      "loss": 0.9147,
+      "step": 5678
+    },
+    {
+      "epoch": 1.1194622823017915,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.923878286165856e-06,
+      "loss": 0.9804,
+      "step": 5679
+    },
+    {
+      "epoch": 1.1196621773568876,
+      "grad_norm": 2.0,
+      "learning_rate": 6.922905574105111e-06,
+      "loss": 0.8939,
+      "step": 5680
+    },
+    {
+      "epoch": 1.1198620724119837,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.921932776629107e-06,
+      "loss": 0.9842,
+      "step": 5681
+    },
+    {
+      "epoch": 1.1200619674670798,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.920959893781054e-06,
+      "loss": 1.0701,
+      "step": 5682
+    },
+    {
+      "epoch": 1.1202618625221759,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.919986925604166e-06,
+      "loss": 0.9869,
+      "step": 5683
+    },
+    {
+      "epoch": 1.120461757577272,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.919013872141667e-06,
+      "loss": 1.0938,
+      "step": 5684
+    },
+    {
+      "epoch": 1.120661652632368,
+      "grad_norm": 2.125,
+      "learning_rate": 6.918040733436774e-06,
+      "loss": 0.9659,
+      "step": 5685
+    },
+    {
+      "epoch": 1.1208615476874642,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.917067509532718e-06,
+      "loss": 0.9582,
+      "step": 5686
+    },
+    {
+      "epoch": 1.1210614427425603,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.916094200472727e-06,
+      "loss": 0.9951,
+      "step": 5687
+    },
+    {
+      "epoch": 1.1212613377976561,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.915120806300036e-06,
+      "loss": 0.9728,
+      "step": 5688
+    },
+    {
+      "epoch": 1.1214612328527522,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.914147327057885e-06,
+      "loss": 0.9721,
+      "step": 5689
+    },
+    {
+      "epoch": 1.1216611279078483,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.913173762789515e-06,
+      "loss": 0.9446,
+      "step": 5690
+    },
+    {
+      "epoch": 1.1218610229629444,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.912200113538168e-06,
+      "loss": 0.9531,
+      "step": 5691
+    },
+    {
+      "epoch": 1.1220609180180405,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.911226379347097e-06,
+      "loss": 0.9618,
+      "step": 5692
+    },
+    {
+      "epoch": 1.1222608130731366,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.910252560259555e-06,
+      "loss": 0.9827,
+      "step": 5693
+    },
+    {
+      "epoch": 1.1224607081282327,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.9092786563187975e-06,
+      "loss": 1.0449,
+      "step": 5694
+    },
+    {
+      "epoch": 1.1226606031833288,
+      "grad_norm": 2.25,
+      "learning_rate": 6.908304667568087e-06,
+      "loss": 1.0932,
+      "step": 5695
+    },
+    {
+      "epoch": 1.122860498238425,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.907330594050685e-06,
+      "loss": 0.9571,
+      "step": 5696
+    },
+    {
+      "epoch": 1.1230603932935208,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.9063564358098636e-06,
+      "loss": 1.0944,
+      "step": 5697
+    },
+    {
+      "epoch": 1.123260288348617,
+      "grad_norm": 2.25,
+      "learning_rate": 6.905382192888893e-06,
+      "loss": 1.1063,
+      "step": 5698
+    },
+    {
+      "epoch": 1.123460183403713,
+      "grad_norm": 2.25,
+      "learning_rate": 6.904407865331048e-06,
+      "loss": 0.8986,
+      "step": 5699
+    },
+    {
+      "epoch": 1.123660078458809,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.903433453179609e-06,
+      "loss": 1.0043,
+      "step": 5700
+    },
+    {
+      "epoch": 1.1238599735139052,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.90245895647786e-06,
+      "loss": 0.9294,
+      "step": 5701
+    },
+    {
+      "epoch": 1.1240598685690013,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.901484375269086e-06,
+      "loss": 0.9703,
+      "step": 5702
+    },
+    {
+      "epoch": 1.1242597636240974,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.900509709596581e-06,
+      "loss": 0.9744,
+      "step": 5703
+    },
+    {
+      "epoch": 1.1244596586791935,
+      "grad_norm": 2.375,
+      "learning_rate": 6.8995349595036365e-06,
+      "loss": 0.9381,
+      "step": 5704
+    },
+    {
+      "epoch": 1.1246595537342894,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.898560125033552e-06,
+      "loss": 0.9513,
+      "step": 5705
+    },
+    {
+      "epoch": 1.1248594487893855,
+      "grad_norm": 2.125,
+      "learning_rate": 6.897585206229631e-06,
+      "loss": 0.935,
+      "step": 5706
+    },
+    {
+      "epoch": 1.1250593438444816,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.896610203135176e-06,
+      "loss": 0.9522,
+      "step": 5707
+    },
+    {
+      "epoch": 1.1252592388995777,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.8956351157935e-06,
+      "loss": 1.0196,
+      "step": 5708
+    },
+    {
+      "epoch": 1.1254591339546738,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.894659944247914e-06,
+      "loss": 0.9577,
+      "step": 5709
+    },
+    {
+      "epoch": 1.1256590290097699,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.8936846885417344e-06,
+      "loss": 0.971,
+      "step": 5710
+    },
+    {
+      "epoch": 1.125858924064866,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.892709348718283e-06,
+      "loss": 0.9899,
+      "step": 5711
+    },
+    {
+      "epoch": 1.126058819119962,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.891733924820887e-06,
+      "loss": 0.9589,
+      "step": 5712
+    },
+    {
+      "epoch": 1.1262587141750582,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.89075841689287e-06,
+      "loss": 0.9113,
+      "step": 5713
+    },
+    {
+      "epoch": 1.1264586092301543,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.889782824977566e-06,
+      "loss": 0.9732,
+      "step": 5714
+    },
+    {
+      "epoch": 1.1266585042852502,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.8888071491183114e-06,
+      "loss": 0.941,
+      "step": 5715
+    },
+    {
+      "epoch": 1.1268583993403463,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.887831389358445e-06,
+      "loss": 0.946,
+      "step": 5716
+    },
+    {
+      "epoch": 1.1270582943954424,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.8868555457413115e-06,
+      "loss": 1.0431,
+      "step": 5717
+    },
+    {
+      "epoch": 1.1272581894505385,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.885879618310253e-06,
+      "loss": 0.9591,
+      "step": 5718
+    },
+    {
+      "epoch": 1.1274580845056346,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.884903607108624e-06,
+      "loss": 0.9041,
+      "step": 5719
+    },
+    {
+      "epoch": 1.1276579795607307,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.88392751217978e-06,
+      "loss": 1.0078,
+      "step": 5720
+    },
+    {
+      "epoch": 1.1278578746158268,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.882951333567076e-06,
+      "loss": 1.0244,
+      "step": 5721
+    },
+    {
+      "epoch": 1.1280577696709229,
+      "grad_norm": 2.125,
+      "learning_rate": 6.881975071313876e-06,
+      "loss": 0.8964,
+      "step": 5722
+    },
+    {
+      "epoch": 1.1282576647260187,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.880998725463543e-06,
+      "loss": 0.9442,
+      "step": 5723
+    },
+    {
+      "epoch": 1.1284575597811148,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.880022296059448e-06,
+      "loss": 1.0363,
+      "step": 5724
+    },
+    {
+      "epoch": 1.128657454836211,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.879045783144962e-06,
+      "loss": 0.9913,
+      "step": 5725
+    },
+    {
+      "epoch": 1.128857349891307,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.878069186763466e-06,
+      "loss": 1.0463,
+      "step": 5726
+    },
+    {
+      "epoch": 1.1290572449464031,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.877092506958334e-06,
+      "loss": 0.9575,
+      "step": 5727
+    },
+    {
+      "epoch": 1.1292571400014992,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.876115743772954e-06,
+      "loss": 0.9106,
+      "step": 5728
+    },
+    {
+      "epoch": 1.1294570350565953,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.8751388972507146e-06,
+      "loss": 0.936,
+      "step": 5729
+    },
+    {
+      "epoch": 1.1296569301116914,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.874161967435005e-06,
+      "loss": 1.0641,
+      "step": 5730
+    },
+    {
+      "epoch": 1.1298568251667875,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.87318495436922e-06,
+      "loss": 0.9322,
+      "step": 5731
+    },
+    {
+      "epoch": 1.1300567202218834,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.87220785809676e-06,
+      "loss": 0.9547,
+      "step": 5732
+    },
+    {
+      "epoch": 1.1302566152769795,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.871230678661027e-06,
+      "loss": 1.0251,
+      "step": 5733
+    },
+    {
+      "epoch": 1.1304565103320756,
+      "grad_norm": 2.125,
+      "learning_rate": 6.870253416105428e-06,
+      "loss": 0.9616,
+      "step": 5734
+    },
+    {
+      "epoch": 1.1306564053871717,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.8692760704733705e-06,
+      "loss": 0.9055,
+      "step": 5735
+    },
+    {
+      "epoch": 1.1308563004422678,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.868298641808271e-06,
+      "loss": 1.0096,
+      "step": 5736
+    },
+    {
+      "epoch": 1.131056195497364,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.867321130153545e-06,
+      "loss": 0.9167,
+      "step": 5737
+    },
+    {
+      "epoch": 1.13125609055246,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.866343535552614e-06,
+      "loss": 0.9472,
+      "step": 5738
+    },
+    {
+      "epoch": 1.1314559856075561,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.865365858048902e-06,
+      "loss": 0.9423,
+      "step": 5739
+    },
+    {
+      "epoch": 1.131655880662652,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.864388097685838e-06,
+      "loss": 1.0048,
+      "step": 5740
+    },
+    {
+      "epoch": 1.131855775717748,
+      "grad_norm": 2.125,
+      "learning_rate": 6.863410254506853e-06,
+      "loss": 0.8936,
+      "step": 5741
+    },
+    {
+      "epoch": 1.1320556707728442,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.862432328555384e-06,
+      "loss": 0.9888,
+      "step": 5742
+    },
+    {
+      "epoch": 1.1322555658279403,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.861454319874871e-06,
+      "loss": 1.0313,
+      "step": 5743
+    },
+    {
+      "epoch": 1.1324554608830364,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.860476228508755e-06,
+      "loss": 1.0116,
+      "step": 5744
+    },
+    {
+      "epoch": 1.1326553559381325,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.859498054500482e-06,
+      "loss": 1.0209,
+      "step": 5745
+    },
+    {
+      "epoch": 1.1328552509932286,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.858519797893507e-06,
+      "loss": 1.0137,
+      "step": 5746
+    },
+    {
+      "epoch": 1.1330551460483247,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.8575414587312785e-06,
+      "loss": 0.9231,
+      "step": 5747
+    },
+    {
+      "epoch": 1.1332550411034208,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.856563037057259e-06,
+      "loss": 0.9772,
+      "step": 5748
+    },
+    {
+      "epoch": 1.133454936158517,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.855584532914906e-06,
+      "loss": 0.9557,
+      "step": 5749
+    },
+    {
+      "epoch": 1.1336548312136128,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.8546059463476864e-06,
+      "loss": 0.9826,
+      "step": 5750
+    },
+    {
+      "epoch": 1.1338547262687089,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.853627277399071e-06,
+      "loss": 0.9502,
+      "step": 5751
+    },
+    {
+      "epoch": 1.134054621323805,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.852648526112529e-06,
+      "loss": 1.0073,
+      "step": 5752
+    },
+    {
+      "epoch": 1.134254516378901,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.851669692531535e-06,
+      "loss": 1.0135,
+      "step": 5753
+    },
+    {
+      "epoch": 1.1344544114339972,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.850690776699574e-06,
+      "loss": 0.986,
+      "step": 5754
+    },
+    {
+      "epoch": 1.1346543064890933,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.849711778660124e-06,
+      "loss": 1.0102,
+      "step": 5755
+    },
+    {
+      "epoch": 1.1348542015441894,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.848732698456675e-06,
+      "loss": 1.0157,
+      "step": 5756
+    },
+    {
+      "epoch": 1.1350540965992855,
+      "grad_norm": 2.125,
+      "learning_rate": 6.847753536132717e-06,
+      "loss": 0.9198,
+      "step": 5757
+    },
+    {
+      "epoch": 1.1352539916543813,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.846774291731744e-06,
+      "loss": 1.0121,
+      "step": 5758
+    },
+    {
+      "epoch": 1.1354538867094774,
+      "grad_norm": 1.953125,
+      "learning_rate": 6.845794965297254e-06,
+      "loss": 0.8884,
+      "step": 5759
+    },
+    {
+      "epoch": 1.1356537817645735,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.844815556872751e-06,
+      "loss": 0.9572,
+      "step": 5760
+    },
+    {
+      "epoch": 1.1358536768196696,
+      "grad_norm": 1.984375,
+      "learning_rate": 6.8438360665017355e-06,
+      "loss": 0.8053,
+      "step": 5761
+    },
+    {
+      "epoch": 1.1360535718747657,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.842856494227721e-06,
+      "loss": 1.0477,
+      "step": 5762
+    },
+    {
+      "epoch": 1.1362534669298618,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.841876840094216e-06,
+      "loss": 0.9417,
+      "step": 5763
+    },
+    {
+      "epoch": 1.136453361984958,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.840897104144739e-06,
+      "loss": 0.9415,
+      "step": 5764
+    },
+    {
+      "epoch": 1.136653257040054,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.839917286422811e-06,
+      "loss": 0.9455,
+      "step": 5765
+    },
+    {
+      "epoch": 1.1368531520951501,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.838937386971951e-06,
+      "loss": 1.1089,
+      "step": 5766
+    },
+    {
+      "epoch": 1.1370530471502462,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.837957405835689e-06,
+      "loss": 0.9556,
+      "step": 5767
+    },
+    {
+      "epoch": 1.1372529422053421,
+      "grad_norm": 1.984375,
+      "learning_rate": 6.836977343057558e-06,
+      "loss": 0.9827,
+      "step": 5768
+    },
+    {
+      "epoch": 1.1374528372604382,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.835997198681087e-06,
+      "loss": 0.9803,
+      "step": 5769
+    },
+    {
+      "epoch": 1.1376527323155343,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.835016972749817e-06,
+      "loss": 1.0608,
+      "step": 5770
+    },
+    {
+      "epoch": 1.1378526273706304,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.83403666530729e-06,
+      "loss": 0.9443,
+      "step": 5771
+    },
+    {
+      "epoch": 1.1380525224257265,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.8330562763970484e-06,
+      "loss": 0.9665,
+      "step": 5772
+    },
+    {
+      "epoch": 1.1382524174808226,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.832075806062644e-06,
+      "loss": 0.9547,
+      "step": 5773
+    },
+    {
+      "epoch": 1.1384523125359187,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.831095254347629e-06,
+      "loss": 1.0589,
+      "step": 5774
+    },
+    {
+      "epoch": 1.1386522075910146,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.830114621295556e-06,
+      "loss": 1.0209,
+      "step": 5775
+    },
+    {
+      "epoch": 1.1388521026461107,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.829133906949988e-06,
+      "loss": 0.9995,
+      "step": 5776
+    },
+    {
+      "epoch": 1.1390519977012068,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.8281531113544875e-06,
+      "loss": 0.9766,
+      "step": 5777
+    },
+    {
+      "epoch": 1.139251892756303,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.827172234552621e-06,
+      "loss": 0.9472,
+      "step": 5778
+    },
+    {
+      "epoch": 1.139451787811399,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.826191276587959e-06,
+      "loss": 1.0093,
+      "step": 5779
+    },
+    {
+      "epoch": 1.139651682866495,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.825210237504075e-06,
+      "loss": 0.9545,
+      "step": 5780
+    },
+    {
+      "epoch": 1.1398515779215912,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.824229117344547e-06,
+      "loss": 0.9503,
+      "step": 5781
+    },
+    {
+      "epoch": 1.1400514729766873,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.823247916152957e-06,
+      "loss": 0.988,
+      "step": 5782
+    },
+    {
+      "epoch": 1.1402513680317834,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.822266633972891e-06,
+      "loss": 0.9702,
+      "step": 5783
+    },
+    {
+      "epoch": 1.1404512630868795,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.821285270847934e-06,
+      "loss": 0.8738,
+      "step": 5784
+    },
+    {
+      "epoch": 1.1406511581419754,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.82030382682168e-06,
+      "loss": 0.8834,
+      "step": 5785
+    },
+    {
+      "epoch": 1.1408510531970715,
+      "grad_norm": 2.125,
+      "learning_rate": 6.819322301937724e-06,
+      "loss": 0.9622,
+      "step": 5786
+    },
+    {
+      "epoch": 1.1410509482521676,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.818340696239666e-06,
+      "loss": 0.9004,
+      "step": 5787
+    },
+    {
+      "epoch": 1.1412508433072637,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.817359009771109e-06,
+      "loss": 0.9325,
+      "step": 5788
+    },
+    {
+      "epoch": 1.1414507383623598,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.816377242575658e-06,
+      "loss": 0.973,
+      "step": 5789
+    },
+    {
+      "epoch": 1.1416506334174559,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.815395394696924e-06,
+      "loss": 0.934,
+      "step": 5790
+    },
+    {
+      "epoch": 1.141850528472552,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.814413466178521e-06,
+      "loss": 1.0456,
+      "step": 5791
+    },
+    {
+      "epoch": 1.142050423527648,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.813431457064064e-06,
+      "loss": 1.0741,
+      "step": 5792
+    },
+    {
+      "epoch": 1.142250318582744,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.812449367397178e-06,
+      "loss": 0.9353,
+      "step": 5793
+    },
+    {
+      "epoch": 1.14245021363784,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.811467197221483e-06,
+      "loss": 0.9993,
+      "step": 5794
+    },
+    {
+      "epoch": 1.1426501086929362,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.8104849465806086e-06,
+      "loss": 0.8419,
+      "step": 5795
+    },
+    {
+      "epoch": 1.1428500037480323,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.809502615518187e-06,
+      "loss": 0.8346,
+      "step": 5796
+    },
+    {
+      "epoch": 1.1430498988031284,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.808520204077852e-06,
+      "loss": 1.0155,
+      "step": 5797
+    },
+    {
+      "epoch": 1.1432497938582245,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.807537712303243e-06,
+      "loss": 1.03,
+      "step": 5798
+    },
+    {
+      "epoch": 1.1434496889133205,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.806555140238002e-06,
+      "loss": 1.0012,
+      "step": 5799
+    },
+    {
+      "epoch": 1.1436495839684166,
+      "grad_norm": 2.125,
+      "learning_rate": 6.805572487925774e-06,
+      "loss": 0.9254,
+      "step": 5800
+    },
+    {
+      "epoch": 1.1438494790235127,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.804589755410209e-06,
+      "loss": 1.0384,
+      "step": 5801
+    },
+    {
+      "epoch": 1.1440493740786088,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.803606942734961e-06,
+      "loss": 0.954,
+      "step": 5802
+    },
+    {
+      "epoch": 1.1442492691337047,
+      "grad_norm": 2.25,
+      "learning_rate": 6.802624049943684e-06,
+      "loss": 1.0369,
+      "step": 5803
+    },
+    {
+      "epoch": 1.1444491641888008,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.801641077080039e-06,
+      "loss": 0.9482,
+      "step": 5804
+    },
+    {
+      "epoch": 1.144649059243897,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.80065802418769e-06,
+      "loss": 1.0058,
+      "step": 5805
+    },
+    {
+      "epoch": 1.144848954298993,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.799674891310304e-06,
+      "loss": 0.9918,
+      "step": 5806
+    },
+    {
+      "epoch": 1.1450488493540891,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.798691678491552e-06,
+      "loss": 0.948,
+      "step": 5807
+    },
+    {
+      "epoch": 1.1452487444091852,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.797708385775107e-06,
+      "loss": 0.8365,
+      "step": 5808
+    },
+    {
+      "epoch": 1.1454486394642813,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.796725013204648e-06,
+      "loss": 0.9136,
+      "step": 5809
+    },
+    {
+      "epoch": 1.1456485345193772,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.795741560823856e-06,
+      "loss": 1.0892,
+      "step": 5810
+    },
+    {
+      "epoch": 1.1458484295744733,
+      "grad_norm": 2.125,
+      "learning_rate": 6.794758028676415e-06,
+      "loss": 0.9594,
+      "step": 5811
+    },
+    {
+      "epoch": 1.1460483246295694,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.793774416806014e-06,
+      "loss": 1.0519,
+      "step": 5812
+    },
+    {
+      "epoch": 1.1462482196846655,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.792790725256347e-06,
+      "loss": 0.9291,
+      "step": 5813
+    },
+    {
+      "epoch": 1.1464481147397616,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.791806954071105e-06,
+      "loss": 0.9338,
+      "step": 5814
+    },
+    {
+      "epoch": 1.1466480097948577,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.79082310329399e-06,
+      "loss": 0.9201,
+      "step": 5815
+    },
+    {
+      "epoch": 1.1468479048499538,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.789839172968705e-06,
+      "loss": 0.9484,
+      "step": 5816
+    },
+    {
+      "epoch": 1.14704779990505,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.7888551631389545e-06,
+      "loss": 0.9763,
+      "step": 5817
+    },
+    {
+      "epoch": 1.147247694960146,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.787871073848448e-06,
+      "loss": 1.0782,
+      "step": 5818
+    },
+    {
+      "epoch": 1.147447590015242,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.7868869051409e-06,
+      "loss": 0.9979,
+      "step": 5819
+    },
+    {
+      "epoch": 1.147647485070338,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.785902657060026e-06,
+      "loss": 0.9832,
+      "step": 5820
+    },
+    {
+      "epoch": 1.147847380125434,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.784918329649548e-06,
+      "loss": 0.965,
+      "step": 5821
+    },
+    {
+      "epoch": 1.1480472751805302,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.783933922953188e-06,
+      "loss": 0.9862,
+      "step": 5822
+    },
+    {
+      "epoch": 1.1482471702356263,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.782949437014672e-06,
+      "loss": 0.9793,
+      "step": 5823
+    },
+    {
+      "epoch": 1.1484470652907224,
+      "grad_norm": 2.0,
+      "learning_rate": 6.781964871877735e-06,
+      "loss": 0.9625,
+      "step": 5824
+    },
+    {
+      "epoch": 1.1486469603458185,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.780980227586107e-06,
+      "loss": 0.9612,
+      "step": 5825
+    },
+    {
+      "epoch": 1.1488468554009146,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.7799955041835276e-06,
+      "loss": 0.965,
+      "step": 5826
+    },
+    {
+      "epoch": 1.1490467504560107,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.77901070171374e-06,
+      "loss": 1.0603,
+      "step": 5827
+    },
+    {
+      "epoch": 1.1492466455111066,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.778025820220484e-06,
+      "loss": 1.0257,
+      "step": 5828
+    },
+    {
+      "epoch": 1.1494465405662027,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.777040859747512e-06,
+      "loss": 0.9683,
+      "step": 5829
+    },
+    {
+      "epoch": 1.1496464356212988,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.7760558203385765e-06,
+      "loss": 1.0206,
+      "step": 5830
+    },
+    {
+      "epoch": 1.1498463306763949,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.77507070203743e-06,
+      "loss": 0.9786,
+      "step": 5831
+    },
+    {
+      "epoch": 1.150046225731491,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.774085504887832e-06,
+      "loss": 1.0074,
+      "step": 5832
+    },
+    {
+      "epoch": 1.150246120786587,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.7731002289335455e-06,
+      "loss": 1.0272,
+      "step": 5833
+    },
+    {
+      "epoch": 1.1504460158416832,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.772114874218337e-06,
+      "loss": 0.959,
+      "step": 5834
+    },
+    {
+      "epoch": 1.1506459108967793,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.771129440785973e-06,
+      "loss": 1.0305,
+      "step": 5835
+    },
+    {
+      "epoch": 1.1508458059518754,
+      "grad_norm": 2.25,
+      "learning_rate": 6.770143928680231e-06,
+      "loss": 1.1036,
+      "step": 5836
+    },
+    {
+      "epoch": 1.1510457010069715,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.769158337944883e-06,
+      "loss": 0.983,
+      "step": 5837
+    },
+    {
+      "epoch": 1.1512455960620673,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.768172668623711e-06,
+      "loss": 1.0334,
+      "step": 5838
+    },
+    {
+      "epoch": 1.1514454911171634,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.767186920760499e-06,
+      "loss": 0.9598,
+      "step": 5839
+    },
+    {
+      "epoch": 1.1516453861722595,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.766201094399031e-06,
+      "loss": 1.0598,
+      "step": 5840
+    },
+    {
+      "epoch": 1.1518452812273556,
+      "grad_norm": 2.125,
+      "learning_rate": 6.765215189583101e-06,
+      "loss": 0.887,
+      "step": 5841
+    },
+    {
+      "epoch": 1.1520451762824517,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.764229206356498e-06,
+      "loss": 0.9367,
+      "step": 5842
+    },
+    {
+      "epoch": 1.1522450713375478,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.763243144763024e-06,
+      "loss": 1.0542,
+      "step": 5843
+    },
+    {
+      "epoch": 1.152444966392644,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.762257004846479e-06,
+      "loss": 1.0015,
+      "step": 5844
+    },
+    {
+      "epoch": 1.15264486144774,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.761270786650664e-06,
+      "loss": 0.9359,
+      "step": 5845
+    },
+    {
+      "epoch": 1.152844756502836,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.76028449021939e-06,
+      "loss": 1.0411,
+      "step": 5846
+    },
+    {
+      "epoch": 1.153044651557932,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.759298115596467e-06,
+      "loss": 0.9873,
+      "step": 5847
+    },
+    {
+      "epoch": 1.153244546613028,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.7583116628257075e-06,
+      "loss": 1.012,
+      "step": 5848
+    },
+    {
+      "epoch": 1.1534444416681242,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.757325131950934e-06,
+      "loss": 0.9291,
+      "step": 5849
+    },
+    {
+      "epoch": 1.1536443367232203,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.756338523015965e-06,
+      "loss": 1.0056,
+      "step": 5850
+    },
+    {
+      "epoch": 1.1538442317783164,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.755351836064625e-06,
+      "loss": 0.8476,
+      "step": 5851
+    },
+    {
+      "epoch": 1.1540441268334125,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.754365071140747e-06,
+      "loss": 0.9314,
+      "step": 5852
+    },
+    {
+      "epoch": 1.1542440218885086,
+      "grad_norm": 2.125,
+      "learning_rate": 6.753378228288158e-06,
+      "loss": 1.0368,
+      "step": 5853
+    },
+    {
+      "epoch": 1.1544439169436047,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.752391307550694e-06,
+      "loss": 1.0438,
+      "step": 5854
+    },
+    {
+      "epoch": 1.1546438119987006,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.751404308972198e-06,
+      "loss": 1.0567,
+      "step": 5855
+    },
+    {
+      "epoch": 1.1548437070537967,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.750417232596509e-06,
+      "loss": 0.9272,
+      "step": 5856
+    },
+    {
+      "epoch": 1.1550436021088928,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.749430078467472e-06,
+      "loss": 1.0246,
+      "step": 5857
+    },
+    {
+      "epoch": 1.1552434971639889,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.74844284662894e-06,
+      "loss": 0.9252,
+      "step": 5858
+    },
+    {
+      "epoch": 1.155443392219085,
+      "grad_norm": 2.125,
+      "learning_rate": 6.7474555371247605e-06,
+      "loss": 0.9725,
+      "step": 5859
+    },
+    {
+      "epoch": 1.155643287274181,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.746468149998796e-06,
+      "loss": 0.9515,
+      "step": 5860
+    },
+    {
+      "epoch": 1.1558431823292772,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.7454806852949015e-06,
+      "loss": 0.9205,
+      "step": 5861
+    },
+    {
+      "epoch": 1.1560430773843733,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.744493143056941e-06,
+      "loss": 0.9725,
+      "step": 5862
+    },
+    {
+      "epoch": 1.1562429724394692,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.743505523328781e-06,
+      "loss": 0.8935,
+      "step": 5863
+    },
+    {
+      "epoch": 1.1564428674945653,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.742517826154293e-06,
+      "loss": 1.0846,
+      "step": 5864
+    },
+    {
+      "epoch": 1.1566427625496614,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.741530051577347e-06,
+      "loss": 0.9303,
+      "step": 5865
+    },
+    {
+      "epoch": 1.1568426576047575,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.740542199641824e-06,
+      "loss": 0.9525,
+      "step": 5866
+    },
+    {
+      "epoch": 1.1570425526598536,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.739554270391603e-06,
+      "loss": 0.964,
+      "step": 5867
+    },
+    {
+      "epoch": 1.1572424477149497,
+      "grad_norm": 1.984375,
+      "learning_rate": 6.738566263870566e-06,
+      "loss": 0.8848,
+      "step": 5868
+    },
+    {
+      "epoch": 1.1574423427700458,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.737578180122603e-06,
+      "loss": 0.9419,
+      "step": 5869
+    },
+    {
+      "epoch": 1.1576422378251419,
+      "grad_norm": 2.125,
+      "learning_rate": 6.7365900191916e-06,
+      "loss": 0.968,
+      "step": 5870
+    },
+    {
+      "epoch": 1.157842132880238,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.735601781121454e-06,
+      "loss": 0.9856,
+      "step": 5871
+    },
+    {
+      "epoch": 1.158042027935334,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.734613465956065e-06,
+      "loss": 0.953,
+      "step": 5872
+    },
+    {
+      "epoch": 1.15824192299043,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.733625073739329e-06,
+      "loss": 0.9394,
+      "step": 5873
+    },
+    {
+      "epoch": 1.158441818045526,
+      "grad_norm": 2.25,
+      "learning_rate": 6.732636604515153e-06,
+      "loss": 0.9459,
+      "step": 5874
+    },
+    {
+      "epoch": 1.1586417131006221,
+      "grad_norm": 1.9921875,
+      "learning_rate": 6.731648058327445e-06,
+      "loss": 0.8777,
+      "step": 5875
+    },
+    {
+      "epoch": 1.1588416081557182,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.7306594352201135e-06,
+      "loss": 0.9294,
+      "step": 5876
+    },
+    {
+      "epoch": 1.1590415032108143,
+      "grad_norm": 2.125,
+      "learning_rate": 6.729670735237075e-06,
+      "loss": 1.0541,
+      "step": 5877
+    },
+    {
+      "epoch": 1.1592413982659104,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.728681958422248e-06,
+      "loss": 0.9905,
+      "step": 5878
+    },
+    {
+      "epoch": 1.1594412933210065,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.727693104819553e-06,
+      "loss": 0.9105,
+      "step": 5879
+    },
+    {
+      "epoch": 1.1596411883761026,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.726704174472913e-06,
+      "loss": 1.0022,
+      "step": 5880
+    },
+    {
+      "epoch": 1.1598410834311985,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.725715167426261e-06,
+      "loss": 1.0047,
+      "step": 5881
+    },
+    {
+      "epoch": 1.1600409784862946,
+      "grad_norm": 2.125,
+      "learning_rate": 6.724726083723523e-06,
+      "loss": 0.9738,
+      "step": 5882
+    },
+    {
+      "epoch": 1.1602408735413907,
+      "grad_norm": 2.125,
+      "learning_rate": 6.723736923408638e-06,
+      "loss": 0.9422,
+      "step": 5883
+    },
+    {
+      "epoch": 1.1604407685964868,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.722747686525543e-06,
+      "loss": 0.9893,
+      "step": 5884
+    },
+    {
+      "epoch": 1.160640663651583,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.721758373118178e-06,
+      "loss": 1.0724,
+      "step": 5885
+    },
+    {
+      "epoch": 1.160840558706679,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.720768983230492e-06,
+      "loss": 1.0144,
+      "step": 5886
+    },
+    {
+      "epoch": 1.1610404537617751,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.7197795169064305e-06,
+      "loss": 0.9422,
+      "step": 5887
+    },
+    {
+      "epoch": 1.1612403488168712,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.7187899741899465e-06,
+      "loss": 0.9876,
+      "step": 5888
+    },
+    {
+      "epoch": 1.1614402438719673,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.717800355124996e-06,
+      "loss": 0.8915,
+      "step": 5889
+    },
+    {
+      "epoch": 1.1616401389270634,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.716810659755537e-06,
+      "loss": 0.8954,
+      "step": 5890
+    },
+    {
+      "epoch": 1.1618400339821593,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.715820888125532e-06,
+      "loss": 0.9359,
+      "step": 5891
+    },
+    {
+      "epoch": 1.1620399290372554,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.714831040278946e-06,
+      "loss": 0.9886,
+      "step": 5892
+    },
+    {
+      "epoch": 1.1622398240923515,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.713841116259749e-06,
+      "loss": 0.9766,
+      "step": 5893
+    },
+    {
+      "epoch": 1.1624397191474476,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.7128511161119115e-06,
+      "loss": 0.9345,
+      "step": 5894
+    },
+    {
+      "epoch": 1.1626396142025437,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.7118610398794115e-06,
+      "loss": 0.9462,
+      "step": 5895
+    },
+    {
+      "epoch": 1.1628395092576398,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.710870887606227e-06,
+      "loss": 0.9552,
+      "step": 5896
+    },
+    {
+      "epoch": 1.163039404312736,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.70988065933634e-06,
+      "loss": 0.9285,
+      "step": 5897
+    },
+    {
+      "epoch": 1.1632392993678318,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.708890355113736e-06,
+      "loss": 1.0474,
+      "step": 5898
+    },
+    {
+      "epoch": 1.1634391944229279,
+      "grad_norm": 1.96875,
+      "learning_rate": 6.707899974982405e-06,
+      "loss": 0.8558,
+      "step": 5899
+    },
+    {
+      "epoch": 1.163639089478024,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.706909518986341e-06,
+      "loss": 0.9501,
+      "step": 5900
+    },
+    {
+      "epoch": 1.16383898453312,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.705918987169537e-06,
+      "loss": 0.9321,
+      "step": 5901
+    },
+    {
+      "epoch": 1.1640388795882162,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.704928379575993e-06,
+      "loss": 0.9636,
+      "step": 5902
+    },
+    {
+      "epoch": 1.1642387746433123,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.703937696249715e-06,
+      "loss": 1.0126,
+      "step": 5903
+    },
+    {
+      "epoch": 1.1644386696984084,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.7029469372347045e-06,
+      "loss": 1.084,
+      "step": 5904
+    },
+    {
+      "epoch": 1.1646385647535045,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.701956102574973e-06,
+      "loss": 0.9991,
+      "step": 5905
+    },
+    {
+      "epoch": 1.1648384598086006,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.700965192314536e-06,
+      "loss": 0.9644,
+      "step": 5906
+    },
+    {
+      "epoch": 1.1650383548636967,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.699974206497405e-06,
+      "loss": 1.0096,
+      "step": 5907
+    },
+    {
+      "epoch": 1.1652382499187925,
+      "grad_norm": 2.125,
+      "learning_rate": 6.6989831451676015e-06,
+      "loss": 0.9761,
+      "step": 5908
+    },
+    {
+      "epoch": 1.1654381449738886,
+      "grad_norm": 2.125,
+      "learning_rate": 6.697992008369147e-06,
+      "loss": 1.0231,
+      "step": 5909
+    },
+    {
+      "epoch": 1.1656380400289847,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.6970007961460695e-06,
+      "loss": 1.039,
+      "step": 5910
+    },
+    {
+      "epoch": 1.1658379350840808,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.6960095085423985e-06,
+      "loss": 0.9953,
+      "step": 5911
+    },
+    {
+      "epoch": 1.166037830139177,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.695018145602165e-06,
+      "loss": 1.0316,
+      "step": 5912
+    },
+    {
+      "epoch": 1.166237725194273,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.694026707369407e-06,
+      "loss": 1.0413,
+      "step": 5913
+    },
+    {
+      "epoch": 1.1664376202493691,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.693035193888164e-06,
+      "loss": 0.9899,
+      "step": 5914
+    },
+    {
+      "epoch": 1.1666375153044652,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.692043605202478e-06,
+      "loss": 0.9738,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1668374103595611,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.691051941356397e-06,
+      "loss": 0.9412,
+      "step": 5916
+    },
+    {
+      "epoch": 1.1670373054146572,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.6900602023939685e-06,
+      "loss": 1.0286,
+      "step": 5917
+    },
+    {
+      "epoch": 1.1672372004697533,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.6890683883592455e-06,
+      "loss": 0.9842,
+      "step": 5918
+    },
+    {
+      "epoch": 1.1674370955248494,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.688076499296285e-06,
+      "loss": 0.9791,
+      "step": 5919
+    },
+    {
+      "epoch": 1.1676369905799455,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.687084535249149e-06,
+      "loss": 0.9836,
+      "step": 5920
+    },
+    {
+      "epoch": 1.1678368856350416,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.686092496261896e-06,
+      "loss": 0.9675,
+      "step": 5921
+    },
+    {
+      "epoch": 1.1680367806901377,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.685100382378595e-06,
+      "loss": 0.9873,
+      "step": 5922
+    },
+    {
+      "epoch": 1.1682366757452338,
+      "grad_norm": 2.125,
+      "learning_rate": 6.684108193643317e-06,
+      "loss": 1.0282,
+      "step": 5923
+    },
+    {
+      "epoch": 1.16843657080033,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.68311593010013e-06,
+      "loss": 1.0129,
+      "step": 5924
+    },
+    {
+      "epoch": 1.168636465855426,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.682123591793114e-06,
+      "loss": 0.9194,
+      "step": 5925
+    },
+    {
+      "epoch": 1.168836360910522,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.681131178766349e-06,
+      "loss": 0.9182,
+      "step": 5926
+    },
+    {
+      "epoch": 1.169036255965618,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.680138691063914e-06,
+      "loss": 0.9553,
+      "step": 5927
+    },
+    {
+      "epoch": 1.169236151020714,
+      "grad_norm": 2.125,
+      "learning_rate": 6.679146128729901e-06,
+      "loss": 0.9834,
+      "step": 5928
+    },
+    {
+      "epoch": 1.1694360460758102,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.678153491808394e-06,
+      "loss": 1.0813,
+      "step": 5929
+    },
+    {
+      "epoch": 1.1696359411309063,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.677160780343488e-06,
+      "loss": 1.0055,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1698358361860024,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.6761679943792805e-06,
+      "loss": 0.9817,
+      "step": 5931
+    },
+    {
+      "epoch": 1.1700357312410985,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.675175133959868e-06,
+      "loss": 0.9821,
+      "step": 5932
+    },
+    {
+      "epoch": 1.1702356262961944,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.674182199129356e-06,
+      "loss": 0.8733,
+      "step": 5933
+    },
+    {
+      "epoch": 1.1704355213512905,
+      "grad_norm": 2.25,
+      "learning_rate": 6.67318918993185e-06,
+      "loss": 0.9753,
+      "step": 5934
+    },
+    {
+      "epoch": 1.1706354164063866,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.6721961064114584e-06,
+      "loss": 0.9783,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1708353114614827,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.6712029486122946e-06,
+      "loss": 0.9435,
+      "step": 5936
+    },
+    {
+      "epoch": 1.1710352065165788,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.670209716578474e-06,
+      "loss": 0.9447,
+      "step": 5937
+    },
+    {
+      "epoch": 1.1712351015716749,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.669216410354118e-06,
+      "loss": 0.9303,
+      "step": 5938
+    },
+    {
+      "epoch": 1.171434996626771,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.668223029983345e-06,
+      "loss": 0.9607,
+      "step": 5939
+    },
+    {
+      "epoch": 1.171634891681867,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.667229575510284e-06,
+      "loss": 0.9161,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1718347867369632,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.666236046979062e-06,
+      "loss": 0.9774,
+      "step": 5941
+    },
+    {
+      "epoch": 1.1720346817920593,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.665242444433815e-06,
+      "loss": 0.9084,
+      "step": 5942
+    },
+    {
+      "epoch": 1.1722345768471552,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.664248767918675e-06,
+      "loss": 0.9373,
+      "step": 5943
+    },
+    {
+      "epoch": 1.1724344719022513,
+      "grad_norm": 2.25,
+      "learning_rate": 6.663255017477783e-06,
+      "loss": 0.9009,
+      "step": 5944
+    },
+    {
+      "epoch": 1.1726343669573474,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.662261193155281e-06,
+      "loss": 1.1289,
+      "step": 5945
+    },
+    {
+      "epoch": 1.1728342620124435,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.661267294995314e-06,
+      "loss": 1.0036,
+      "step": 5946
+    },
+    {
+      "epoch": 1.1730341570675396,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.66027332304203e-06,
+      "loss": 1.0468,
+      "step": 5947
+    },
+    {
+      "epoch": 1.1732340521226357,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.659279277339584e-06,
+      "loss": 0.8799,
+      "step": 5948
+    },
+    {
+      "epoch": 1.1734339471777318,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.65828515793213e-06,
+      "loss": 0.9727,
+      "step": 5949
+    },
+    {
+      "epoch": 1.1736338422328279,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.657290964863825e-06,
+      "loss": 0.9576,
+      "step": 5950
+    },
+    {
+      "epoch": 1.1738337372879237,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.656296698178832e-06,
+      "loss": 0.8868,
+      "step": 5951
+    },
+    {
+      "epoch": 1.1740336323430198,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.655302357921318e-06,
+      "loss": 0.943,
+      "step": 5952
+    },
+    {
+      "epoch": 1.174233527398116,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.65430794413545e-06,
+      "loss": 1.0092,
+      "step": 5953
+    },
+    {
+      "epoch": 1.174433422453212,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.6533134568654e-06,
+      "loss": 0.9646,
+      "step": 5954
+    },
+    {
+      "epoch": 1.1746333175083081,
+      "grad_norm": 2.25,
+      "learning_rate": 6.652318896155342e-06,
+      "loss": 1.1088,
+      "step": 5955
+    },
+    {
+      "epoch": 1.1748332125634042,
+      "grad_norm": 2.125,
+      "learning_rate": 6.651324262049454e-06,
+      "loss": 1.0244,
+      "step": 5956
+    },
+    {
+      "epoch": 1.1750331076185003,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.650329554591921e-06,
+      "loss": 1.0009,
+      "step": 5957
+    },
+    {
+      "epoch": 1.1752330026735964,
+      "grad_norm": 2.0,
+      "learning_rate": 6.649334773826924e-06,
+      "loss": 0.9352,
+      "step": 5958
+    },
+    {
+      "epoch": 1.1754328977286925,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.648339919798654e-06,
+      "loss": 0.9745,
+      "step": 5959
+    },
+    {
+      "epoch": 1.1756327927837886,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.647344992551299e-06,
+      "loss": 1.1091,
+      "step": 5960
+    },
+    {
+      "epoch": 1.1758326878388845,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.646349992129055e-06,
+      "loss": 0.9725,
+      "step": 5961
+    },
+    {
+      "epoch": 1.1760325828939806,
+      "grad_norm": 2.375,
+      "learning_rate": 6.645354918576122e-06,
+      "loss": 0.9178,
+      "step": 5962
+    },
+    {
+      "epoch": 1.1762324779490767,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.644359771936699e-06,
+      "loss": 0.9725,
+      "step": 5963
+    },
+    {
+      "epoch": 1.1764323730041728,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.643364552254989e-06,
+      "loss": 0.8993,
+      "step": 5964
+    },
+    {
+      "epoch": 1.176632268059269,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.642369259575203e-06,
+      "loss": 0.9373,
+      "step": 5965
+    },
+    {
+      "epoch": 1.176832163114365,
+      "grad_norm": 2.25,
+      "learning_rate": 6.64137389394155e-06,
+      "loss": 1.0464,
+      "step": 5966
+    },
+    {
+      "epoch": 1.177032058169461,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.640378455398242e-06,
+      "loss": 0.9632,
+      "step": 5967
+    },
+    {
+      "epoch": 1.1772319532245572,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.6393829439895e-06,
+      "loss": 0.9081,
+      "step": 5968
+    },
+    {
+      "epoch": 1.177431848279653,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.6383873597595415e-06,
+      "loss": 1.0095,
+      "step": 5969
+    },
+    {
+      "epoch": 1.1776317433347492,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.637391702752591e-06,
+      "loss": 0.9309,
+      "step": 5970
+    },
+    {
+      "epoch": 1.1778316383898453,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.636395973012878e-06,
+      "loss": 1.0349,
+      "step": 5971
+    },
+    {
+      "epoch": 1.1780315334449414,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.635400170584629e-06,
+      "loss": 1.0354,
+      "step": 5972
+    },
+    {
+      "epoch": 1.1782314285000375,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.63440429551208e-06,
+      "loss": 0.9784,
+      "step": 5973
+    },
+    {
+      "epoch": 1.1784313235551336,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.633408347839466e-06,
+      "loss": 0.9179,
+      "step": 5974
+    },
+    {
+      "epoch": 1.1786312186102297,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.632412327611029e-06,
+      "loss": 0.9423,
+      "step": 5975
+    },
+    {
+      "epoch": 1.1788311136653258,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.63141623487101e-06,
+      "loss": 0.9934,
+      "step": 5976
+    },
+    {
+      "epoch": 1.1790310087204219,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.6304200696636545e-06,
+      "loss": 1.0466,
+      "step": 5977
+    },
+    {
+      "epoch": 1.1792309037755178,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.629423832033215e-06,
+      "loss": 0.9837,
+      "step": 5978
+    },
+    {
+      "epoch": 1.1794307988306139,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.6284275220239435e-06,
+      "loss": 0.9193,
+      "step": 5979
+    },
+    {
+      "epoch": 1.17963069388571,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.627431139680094e-06,
+      "loss": 0.9951,
+      "step": 5980
+    },
+    {
+      "epoch": 1.179830588940806,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.626434685045928e-06,
+      "loss": 0.9343,
+      "step": 5981
+    },
+    {
+      "epoch": 1.1800304839959022,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.625438158165707e-06,
+      "loss": 1.1256,
+      "step": 5982
+    },
+    {
+      "epoch": 1.1802303790509983,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.624441559083696e-06,
+      "loss": 0.9476,
+      "step": 5983
+    },
+    {
+      "epoch": 1.1804302741060944,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.623444887844166e-06,
+      "loss": 1.0104,
+      "step": 5984
+    },
+    {
+      "epoch": 1.1806301691611905,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.622448144491387e-06,
+      "loss": 0.926,
+      "step": 5985
+    },
+    {
+      "epoch": 1.1808300642162863,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.621451329069634e-06,
+      "loss": 0.9933,
+      "step": 5986
+    },
+    {
+      "epoch": 1.1810299592713824,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.6204544416231865e-06,
+      "loss": 0.988,
+      "step": 5987
+    },
+    {
+      "epoch": 1.1812298543264785,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.619457482196326e-06,
+      "loss": 0.9556,
+      "step": 5988
+    },
+    {
+      "epoch": 1.1814297493815746,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.618460450833335e-06,
+      "loss": 0.9795,
+      "step": 5989
+    },
+    {
+      "epoch": 1.1816296444366707,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.617463347578506e-06,
+      "loss": 1.0121,
+      "step": 5990
+    },
+    {
+      "epoch": 1.1818295394917668,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.6164661724761255e-06,
+      "loss": 1.0122,
+      "step": 5991
+    },
+    {
+      "epoch": 1.182029434546863,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.615468925570492e-06,
+      "loss": 1.0103,
+      "step": 5992
+    },
+    {
+      "epoch": 1.182229329601959,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.614471606905902e-06,
+      "loss": 0.9475,
+      "step": 5993
+    },
+    {
+      "epoch": 1.1824292246570551,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.6134742165266545e-06,
+      "loss": 0.9314,
+      "step": 5994
+    },
+    {
+      "epoch": 1.1826291197121512,
+      "grad_norm": 2.25,
+      "learning_rate": 6.612476754477055e-06,
+      "loss": 0.98,
+      "step": 5995
+    },
+    {
+      "epoch": 1.1828290147672471,
+      "grad_norm": 2.125,
+      "learning_rate": 6.6114792208014115e-06,
+      "loss": 0.9514,
+      "step": 5996
+    },
+    {
+      "epoch": 1.1830289098223432,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.610481615544031e-06,
+      "loss": 0.9479,
+      "step": 5997
+    },
+    {
+      "epoch": 1.1832288048774393,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.60948393874923e-06,
+      "loss": 0.9903,
+      "step": 5998
+    },
+    {
+      "epoch": 1.1834286999325354,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.608486190461324e-06,
+      "loss": 1.0254,
+      "step": 5999
+    },
+    {
+      "epoch": 1.1836285949876315,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.607488370724635e-06,
+      "loss": 0.9799,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1836285949876315,
+      "eval_loss": 0.9062411189079285,
+      "eval_runtime": 594.9136,
+      "eval_samples_per_second": 3.594,
+      "eval_steps_per_second": 3.594,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1838284900427276,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.606490479583481e-06,
+      "loss": 0.9528,
+      "step": 6001
+    },
+    {
+      "epoch": 1.1840283850978237,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.605492517082195e-06,
+      "loss": 1.0422,
+      "step": 6002
+    },
+    {
+      "epoch": 1.1842282801529198,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.604494483265101e-06,
+      "loss": 0.9536,
+      "step": 6003
+    },
+    {
+      "epoch": 1.1844281752080157,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.603496378176534e-06,
+      "loss": 0.9032,
+      "step": 6004
+    },
+    {
+      "epoch": 1.1846280702631118,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.602498201860828e-06,
+      "loss": 0.9891,
+      "step": 6005
+    },
+    {
+      "epoch": 1.1848279653182079,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.601499954362324e-06,
+      "loss": 0.9402,
+      "step": 6006
+    },
+    {
+      "epoch": 1.185027860373304,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.6005016357253624e-06,
+      "loss": 0.9325,
+      "step": 6007
+    },
+    {
+      "epoch": 1.1852277554284,
+      "grad_norm": 2.125,
+      "learning_rate": 6.5995032459942895e-06,
+      "loss": 0.923,
+      "step": 6008
+    },
+    {
+      "epoch": 1.1854276504834962,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.5985047852134535e-06,
+      "loss": 1.0531,
+      "step": 6009
+    },
+    {
+      "epoch": 1.1856275455385923,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.597506253427206e-06,
+      "loss": 0.9276,
+      "step": 6010
+    },
+    {
+      "epoch": 1.1858274405936884,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.5965076506799e-06,
+      "loss": 1.0075,
+      "step": 6011
+    },
+    {
+      "epoch": 1.1860273356487845,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.595508977015897e-06,
+      "loss": 0.9152,
+      "step": 6012
+    },
+    {
+      "epoch": 1.1862272307038806,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.594510232479553e-06,
+      "loss": 0.9343,
+      "step": 6013
+    },
+    {
+      "epoch": 1.1864271257589765,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.5935114171152345e-06,
+      "loss": 0.9068,
+      "step": 6014
+    },
+    {
+      "epoch": 1.1866270208140726,
+      "grad_norm": 2.25,
+      "learning_rate": 6.592512530967312e-06,
+      "loss": 0.9557,
+      "step": 6015
+    },
+    {
+      "epoch": 1.1868269158691687,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.591513574080152e-06,
+      "loss": 0.9594,
+      "step": 6016
+    },
+    {
+      "epoch": 1.1870268109242648,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.590514546498128e-06,
+      "loss": 0.9499,
+      "step": 6017
+    },
+    {
+      "epoch": 1.1872267059793609,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.589515448265619e-06,
+      "loss": 0.9983,
+      "step": 6018
+    },
+    {
+      "epoch": 1.187426601034457,
+      "grad_norm": 2.125,
+      "learning_rate": 6.588516279427002e-06,
+      "loss": 0.9502,
+      "step": 6019
+    },
+    {
+      "epoch": 1.187626496089553,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.587517040026662e-06,
+      "loss": 0.9559,
+      "step": 6020
+    },
+    {
+      "epoch": 1.187826391144649,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.586517730108985e-06,
+      "loss": 1.1202,
+      "step": 6021
+    },
+    {
+      "epoch": 1.188026286199745,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.58551834971836e-06,
+      "loss": 0.9888,
+      "step": 6022
+    },
+    {
+      "epoch": 1.1882261812548411,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.584518898899178e-06,
+      "loss": 1.0361,
+      "step": 6023
+    },
+    {
+      "epoch": 1.1884260763099372,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.583519377695838e-06,
+      "loss": 0.989,
+      "step": 6024
+    },
+    {
+      "epoch": 1.1886259713650333,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.582519786152735e-06,
+      "loss": 0.9821,
+      "step": 6025
+    },
+    {
+      "epoch": 1.1888258664201294,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.581520124314271e-06,
+      "loss": 0.9858,
+      "step": 6026
+    },
+    {
+      "epoch": 1.1890257614752255,
+      "grad_norm": 2.25,
+      "learning_rate": 6.580520392224854e-06,
+      "loss": 1.0612,
+      "step": 6027
+    },
+    {
+      "epoch": 1.1892256565303216,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.579520589928888e-06,
+      "loss": 0.9674,
+      "step": 6028
+    },
+    {
+      "epoch": 1.1894255515854177,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.578520717470789e-06,
+      "loss": 0.948,
+      "step": 6029
+    },
+    {
+      "epoch": 1.1896254466405138,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.577520774894967e-06,
+      "loss": 1.0566,
+      "step": 6030
+    },
+    {
+      "epoch": 1.1898253416956097,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.57652076224584e-06,
+      "loss": 1.1256,
+      "step": 6031
+    },
+    {
+      "epoch": 1.1900252367507058,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.57552067956783e-06,
+      "loss": 1.0127,
+      "step": 6032
+    },
+    {
+      "epoch": 1.190225131805802,
+      "grad_norm": 2.375,
+      "learning_rate": 6.574520526905358e-06,
+      "loss": 0.9052,
+      "step": 6033
+    },
+    {
+      "epoch": 1.190425026860898,
+      "grad_norm": 2.125,
+      "learning_rate": 6.573520304302853e-06,
+      "loss": 0.9465,
+      "step": 6034
+    },
+    {
+      "epoch": 1.1906249219159941,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.572520011804745e-06,
+      "loss": 0.9954,
+      "step": 6035
+    },
+    {
+      "epoch": 1.1908248169710902,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.571519649455464e-06,
+      "loss": 0.9809,
+      "step": 6036
+    },
+    {
+      "epoch": 1.1910247120261863,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.5705192172994505e-06,
+      "loss": 0.9703,
+      "step": 6037
+    },
+    {
+      "epoch": 1.1912246070812824,
+      "grad_norm": 2.125,
+      "learning_rate": 6.56951871538114e-06,
+      "loss": 0.9372,
+      "step": 6038
+    },
+    {
+      "epoch": 1.1914245021363783,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.568518143744977e-06,
+      "loss": 1.0186,
+      "step": 6039
+    },
+    {
+      "epoch": 1.1916243971914744,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.567517502435403e-06,
+      "loss": 0.9661,
+      "step": 6040
+    },
+    {
+      "epoch": 1.1918242922465705,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.5665167914968706e-06,
+      "loss": 0.8746,
+      "step": 6041
+    },
+    {
+      "epoch": 1.1920241873016666,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.56551601097383e-06,
+      "loss": 1.0034,
+      "step": 6042
+    },
+    {
+      "epoch": 1.1922240823567627,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.564515160910736e-06,
+      "loss": 1.093,
+      "step": 6043
+    },
+    {
+      "epoch": 1.1924239774118588,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.563514241352043e-06,
+      "loss": 1.0007,
+      "step": 6044
+    },
+    {
+      "epoch": 1.192623872466955,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.562513252342216e-06,
+      "loss": 0.9565,
+      "step": 6045
+    },
+    {
+      "epoch": 1.192823767522051,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.561512193925718e-06,
+      "loss": 0.9181,
+      "step": 6046
+    },
+    {
+      "epoch": 1.193023662577147,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.560511066147015e-06,
+      "loss": 0.9166,
+      "step": 6047
+    },
+    {
+      "epoch": 1.1932235576322432,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.559509869050575e-06,
+      "loss": 1.0238,
+      "step": 6048
+    },
+    {
+      "epoch": 1.193423452687339,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.558508602680876e-06,
+      "loss": 0.9868,
+      "step": 6049
+    },
+    {
+      "epoch": 1.1936233477424352,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.557507267082391e-06,
+      "loss": 0.9623,
+      "step": 6050
+    },
+    {
+      "epoch": 1.1938232427975313,
+      "grad_norm": 2.0,
+      "learning_rate": 6.556505862299597e-06,
+      "loss": 0.9718,
+      "step": 6051
+    },
+    {
+      "epoch": 1.1940231378526274,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.555504388376981e-06,
+      "loss": 0.9164,
+      "step": 6052
+    },
+    {
+      "epoch": 1.1942230329077235,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.554502845359026e-06,
+      "loss": 1.0381,
+      "step": 6053
+    },
+    {
+      "epoch": 1.1944229279628196,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.553501233290218e-06,
+      "loss": 1.0206,
+      "step": 6054
+    },
+    {
+      "epoch": 1.1946228230179157,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.5524995522150545e-06,
+      "loss": 1.0788,
+      "step": 6055
+    },
+    {
+      "epoch": 1.1948227180730115,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.551497802178025e-06,
+      "loss": 1.0028,
+      "step": 6056
+    },
+    {
+      "epoch": 1.1950226131281076,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.55049598322363e-06,
+      "loss": 1.0223,
+      "step": 6057
+    },
+    {
+      "epoch": 1.1952225081832037,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.549494095396368e-06,
+      "loss": 0.975,
+      "step": 6058
+    },
+    {
+      "epoch": 1.1954224032382998,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.548492138740743e-06,
+      "loss": 1.0228,
+      "step": 6059
+    },
+    {
+      "epoch": 1.195622298293396,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.547490113301265e-06,
+      "loss": 1.0139,
+      "step": 6060
+    },
+    {
+      "epoch": 1.195822193348492,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.546488019122441e-06,
+      "loss": 0.9619,
+      "step": 6061
+    },
+    {
+      "epoch": 1.1960220884035881,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.5454858562487835e-06,
+      "loss": 0.9858,
+      "step": 6062
+    },
+    {
+      "epoch": 1.1962219834586842,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.544483624724809e-06,
+      "loss": 0.9372,
+      "step": 6063
+    },
+    {
+      "epoch": 1.1964218785137803,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.543481324595037e-06,
+      "loss": 0.9888,
+      "step": 6064
+    },
+    {
+      "epoch": 1.1966217735688764,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.54247895590399e-06,
+      "loss": 0.9256,
+      "step": 6065
+    },
+    {
+      "epoch": 1.1968216686239723,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.541476518696191e-06,
+      "loss": 0.9669,
+      "step": 6066
+    },
+    {
+      "epoch": 1.1970215636790684,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.5404740130161715e-06,
+      "loss": 0.9895,
+      "step": 6067
+    },
+    {
+      "epoch": 1.1972214587341645,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.539471438908459e-06,
+      "loss": 0.8756,
+      "step": 6068
+    },
+    {
+      "epoch": 1.1974213537892606,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.5384687964175915e-06,
+      "loss": 1.0367,
+      "step": 6069
+    },
+    {
+      "epoch": 1.1976212488443567,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.537466085588104e-06,
+      "loss": 0.9823,
+      "step": 6070
+    },
+    {
+      "epoch": 1.1978211438994528,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.536463306464535e-06,
+      "loss": 0.9332,
+      "step": 6071
+    },
+    {
+      "epoch": 1.198021038954549,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.535460459091434e-06,
+      "loss": 0.9466,
+      "step": 6072
+    },
+    {
+      "epoch": 1.198220934009645,
+      "grad_norm": 2.125,
+      "learning_rate": 6.534457543513341e-06,
+      "loss": 0.9703,
+      "step": 6073
+    },
+    {
+      "epoch": 1.198420829064741,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.5334545597748075e-06,
+      "loss": 1.1124,
+      "step": 6074
+    },
+    {
+      "epoch": 1.198620724119837,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.532451507920386e-06,
+      "loss": 0.9838,
+      "step": 6075
+    },
+    {
+      "epoch": 1.198820619174933,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.531448387994634e-06,
+      "loss": 0.9515,
+      "step": 6076
+    },
+    {
+      "epoch": 1.1990205142300292,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.530445200042107e-06,
+      "loss": 1.0353,
+      "step": 6077
+    },
+    {
+      "epoch": 1.1992204092851253,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.5294419441073684e-06,
+      "loss": 1.0373,
+      "step": 6078
+    },
+    {
+      "epoch": 1.1994203043402214,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.528438620234981e-06,
+      "loss": 1.0811,
+      "step": 6079
+    },
+    {
+      "epoch": 1.1996201993953175,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.5274352284695144e-06,
+      "loss": 0.9719,
+      "step": 6080
+    },
+    {
+      "epoch": 1.1998200944504136,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.526431768855537e-06,
+      "loss": 0.8906,
+      "step": 6081
+    },
+    {
+      "epoch": 1.2000199895055097,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.5254282414376235e-06,
+      "loss": 0.9426,
+      "step": 6082
+    },
+    {
+      "epoch": 1.2002198845606058,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.524424646260351e-06,
+      "loss": 0.9419,
+      "step": 6083
+    },
+    {
+      "epoch": 1.2004197796157017,
+      "grad_norm": 2.125,
+      "learning_rate": 6.523420983368298e-06,
+      "loss": 0.9869,
+      "step": 6084
+    },
+    {
+      "epoch": 1.2006196746707978,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.522417252806048e-06,
+      "loss": 0.9829,
+      "step": 6085
+    },
+    {
+      "epoch": 1.2008195697258939,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.5214134546181865e-06,
+      "loss": 1.0244,
+      "step": 6086
+    },
+    {
+      "epoch": 1.20101946478099,
+      "grad_norm": 2.125,
+      "learning_rate": 6.520409588849301e-06,
+      "loss": 0.8331,
+      "step": 6087
+    },
+    {
+      "epoch": 1.201219359836086,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.519405655543985e-06,
+      "loss": 0.8882,
+      "step": 6088
+    },
+    {
+      "epoch": 1.2014192548911822,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.518401654746831e-06,
+      "loss": 0.9884,
+      "step": 6089
+    },
+    {
+      "epoch": 1.2016191499462783,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.517397586502439e-06,
+      "loss": 0.9536,
+      "step": 6090
+    },
+    {
+      "epoch": 1.2018190450013744,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.516393450855407e-06,
+      "loss": 0.9221,
+      "step": 6091
+    },
+    {
+      "epoch": 1.2020189400564703,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.5153892478503414e-06,
+      "loss": 0.9778,
+      "step": 6092
+    },
+    {
+      "epoch": 1.2022188351115664,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.514384977531846e-06,
+      "loss": 0.9758,
+      "step": 6093
+    },
+    {
+      "epoch": 1.2024187301666625,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.513380639944532e-06,
+      "loss": 0.9534,
+      "step": 6094
+    },
+    {
+      "epoch": 1.2026186252217586,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.512376235133011e-06,
+      "loss": 1.0497,
+      "step": 6095
+    },
+    {
+      "epoch": 1.2028185202768547,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.511371763141899e-06,
+      "loss": 1.0026,
+      "step": 6096
+    },
+    {
+      "epoch": 1.2030184153319508,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.510367224015817e-06,
+      "loss": 0.9807,
+      "step": 6097
+    },
+    {
+      "epoch": 1.2032183103870469,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.5093626177993815e-06,
+      "loss": 1.0168,
+      "step": 6098
+    },
+    {
+      "epoch": 1.203418205442143,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.508357944537221e-06,
+      "loss": 0.9322,
+      "step": 6099
+    },
+    {
+      "epoch": 1.203618100497239,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.507353204273962e-06,
+      "loss": 1.0694,
+      "step": 6100
+    },
+    {
+      "epoch": 1.203817995552335,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.506348397054233e-06,
+      "loss": 0.9876,
+      "step": 6101
+    },
+    {
+      "epoch": 1.204017890607431,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.505343522922672e-06,
+      "loss": 1.0069,
+      "step": 6102
+    },
+    {
+      "epoch": 1.2042177856625271,
+      "grad_norm": 2.125,
+      "learning_rate": 6.5043385819239095e-06,
+      "loss": 0.9277,
+      "step": 6103
+    },
+    {
+      "epoch": 1.2044176807176232,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.5033335741025885e-06,
+      "loss": 1.0108,
+      "step": 6104
+    },
+    {
+      "epoch": 1.2046175757727193,
+      "grad_norm": 2.125,
+      "learning_rate": 6.502328499503352e-06,
+      "loss": 1.0157,
+      "step": 6105
+    },
+    {
+      "epoch": 1.2048174708278154,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.5013233581708425e-06,
+      "loss": 1.0746,
+      "step": 6106
+    },
+    {
+      "epoch": 1.2050173658829115,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.500318150149711e-06,
+      "loss": 0.9329,
+      "step": 6107
+    },
+    {
+      "epoch": 1.2052172609380076,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.499312875484608e-06,
+      "loss": 1.0009,
+      "step": 6108
+    },
+    {
+      "epoch": 1.2054171559931035,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.498307534220186e-06,
+      "loss": 0.9681,
+      "step": 6109
+    },
+    {
+      "epoch": 1.2056170510481996,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.497302126401103e-06,
+      "loss": 1.043,
+      "step": 6110
+    },
+    {
+      "epoch": 1.2058169461032957,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.496296652072021e-06,
+      "loss": 0.9317,
+      "step": 6111
+    },
+    {
+      "epoch": 1.2060168411583918,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.4952911112776e-06,
+      "loss": 0.8968,
+      "step": 6112
+    },
+    {
+      "epoch": 1.206216736213488,
+      "grad_norm": 2.125,
+      "learning_rate": 6.494285504062507e-06,
+      "loss": 0.9643,
+      "step": 6113
+    },
+    {
+      "epoch": 1.206416631268584,
+      "grad_norm": 2.125,
+      "learning_rate": 6.493279830471414e-06,
+      "loss": 0.9282,
+      "step": 6114
+    },
+    {
+      "epoch": 1.20661652632368,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.492274090548989e-06,
+      "loss": 0.922,
+      "step": 6115
+    },
+    {
+      "epoch": 1.2068164213787762,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.491268284339908e-06,
+      "loss": 1.0417,
+      "step": 6116
+    },
+    {
+      "epoch": 1.2070163164338723,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.490262411888851e-06,
+      "loss": 1.0131,
+      "step": 6117
+    },
+    {
+      "epoch": 1.2072162114889684,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.489256473240493e-06,
+      "loss": 0.9763,
+      "step": 6118
+    },
+    {
+      "epoch": 1.2074161065440643,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.488250468439525e-06,
+      "loss": 0.9694,
+      "step": 6119
+    },
+    {
+      "epoch": 1.2076160015991604,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.4872443975306275e-06,
+      "loss": 1.0717,
+      "step": 6120
+    },
+    {
+      "epoch": 1.2078158966542565,
+      "grad_norm": 2.25,
+      "learning_rate": 6.486238260558495e-06,
+      "loss": 1.0287,
+      "step": 6121
+    },
+    {
+      "epoch": 1.2080157917093526,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.485232057567816e-06,
+      "loss": 0.9361,
+      "step": 6122
+    },
+    {
+      "epoch": 1.2082156867644487,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.48422578860329e-06,
+      "loss": 0.9639,
+      "step": 6123
+    },
+    {
+      "epoch": 1.2084155818195448,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.4832194537096105e-06,
+      "loss": 0.9281,
+      "step": 6124
+    },
+    {
+      "epoch": 1.2086154768746409,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.482213052931483e-06,
+      "loss": 1.0782,
+      "step": 6125
+    },
+    {
+      "epoch": 1.208815371929737,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.481206586313609e-06,
+      "loss": 0.9929,
+      "step": 6126
+    },
+    {
+      "epoch": 1.2090152669848329,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.480200053900696e-06,
+      "loss": 0.92,
+      "step": 6127
+    },
+    {
+      "epoch": 1.209215162039929,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.479193455737457e-06,
+      "loss": 0.9682,
+      "step": 6128
+    },
+    {
+      "epoch": 1.209415057095025,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.4781867918686e-06,
+      "loss": 0.9514,
+      "step": 6129
+    },
+    {
+      "epoch": 1.2096149521501212,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.477180062338845e-06,
+      "loss": 0.9545,
+      "step": 6130
+    },
+    {
+      "epoch": 1.2098148472052173,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.47617326719291e-06,
+      "loss": 1.0176,
+      "step": 6131
+    },
+    {
+      "epoch": 1.2100147422603134,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.475166406475515e-06,
+      "loss": 0.9661,
+      "step": 6132
+    },
+    {
+      "epoch": 1.2102146373154095,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.474159480231386e-06,
+      "loss": 0.9227,
+      "step": 6133
+    },
+    {
+      "epoch": 1.2104145323705056,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.47315248850525e-06,
+      "loss": 1.0038,
+      "step": 6134
+    },
+    {
+      "epoch": 1.2106144274256017,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.472145431341838e-06,
+      "loss": 0.9451,
+      "step": 6135
+    },
+    {
+      "epoch": 1.2108143224806978,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.471138308785885e-06,
+      "loss": 1.0021,
+      "step": 6136
+    },
+    {
+      "epoch": 1.2110142175357936,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.4701311208821225e-06,
+      "loss": 0.9096,
+      "step": 6137
+    },
+    {
+      "epoch": 1.2112141125908897,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.4691238676752935e-06,
+      "loss": 0.934,
+      "step": 6138
+    },
+    {
+      "epoch": 1.2114140076459858,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.468116549210142e-06,
+      "loss": 1.0087,
+      "step": 6139
+    },
+    {
+      "epoch": 1.211613902701082,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.467109165531407e-06,
+      "loss": 0.986,
+      "step": 6140
+    },
+    {
+      "epoch": 1.211813797756178,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.46610171668384e-06,
+      "loss": 1.0162,
+      "step": 6141
+    },
+    {
+      "epoch": 1.2120136928112741,
+      "grad_norm": 2.25,
+      "learning_rate": 6.465094202712192e-06,
+      "loss": 1.0288,
+      "step": 6142
+    },
+    {
+      "epoch": 1.2122135878663702,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.464086623661215e-06,
+      "loss": 0.8564,
+      "step": 6143
+    },
+    {
+      "epoch": 1.2124134829214661,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.463078979575667e-06,
+      "loss": 0.9441,
+      "step": 6144
+    },
+    {
+      "epoch": 1.2126133779765622,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.462071270500308e-06,
+      "loss": 0.9968,
+      "step": 6145
+    },
+    {
+      "epoch": 1.2128132730316583,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.461063496479899e-06,
+      "loss": 1.0823,
+      "step": 6146
+    },
+    {
+      "epoch": 1.2130131680867544,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.460055657559206e-06,
+      "loss": 0.9145,
+      "step": 6147
+    },
+    {
+      "epoch": 1.2132130631418505,
+      "grad_norm": 2.125,
+      "learning_rate": 6.459047753782994e-06,
+      "loss": 0.9598,
+      "step": 6148
+    },
+    {
+      "epoch": 1.2134129581969466,
+      "grad_norm": 2.25,
+      "learning_rate": 6.458039785196039e-06,
+      "loss": 1.0672,
+      "step": 6149
+    },
+    {
+      "epoch": 1.2136128532520427,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.457031751843113e-06,
+      "loss": 0.9857,
+      "step": 6150
+    },
+    {
+      "epoch": 1.2138127483071388,
+      "grad_norm": 2.125,
+      "learning_rate": 6.4560236537689905e-06,
+      "loss": 0.9374,
+      "step": 6151
+    },
+    {
+      "epoch": 1.214012643362235,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.455015491018452e-06,
+      "loss": 1.0154,
+      "step": 6152
+    },
+    {
+      "epoch": 1.214212538417331,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.454007263636283e-06,
+      "loss": 0.9403,
+      "step": 6153
+    },
+    {
+      "epoch": 1.2144124334724269,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.452998971667266e-06,
+      "loss": 0.9519,
+      "step": 6154
+    },
+    {
+      "epoch": 1.214612328527523,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.451990615156189e-06,
+      "loss": 0.9895,
+      "step": 6155
+    },
+    {
+      "epoch": 1.214812223582619,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.4509821941478455e-06,
+      "loss": 0.9187,
+      "step": 6156
+    },
+    {
+      "epoch": 1.2150121186377152,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.4499737086870276e-06,
+      "loss": 0.9367,
+      "step": 6157
+    },
+    {
+      "epoch": 1.2152120136928113,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.448965158818531e-06,
+      "loss": 0.9306,
+      "step": 6158
+    },
+    {
+      "epoch": 1.2154119087479074,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.447956544587158e-06,
+      "loss": 1.0404,
+      "step": 6159
+    },
+    {
+      "epoch": 1.2156118038030035,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.44694786603771e-06,
+      "loss": 0.961,
+      "step": 6160
+    },
+    {
+      "epoch": 1.2158116988580996,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.445939123214991e-06,
+      "loss": 1.0009,
+      "step": 6161
+    },
+    {
+      "epoch": 1.2160115939131955,
+      "grad_norm": 2.0,
+      "learning_rate": 6.444930316163812e-06,
+      "loss": 0.9166,
+      "step": 6162
+    },
+    {
+      "epoch": 1.2162114889682916,
+      "grad_norm": 2.125,
+      "learning_rate": 6.443921444928982e-06,
+      "loss": 0.9853,
+      "step": 6163
+    },
+    {
+      "epoch": 1.2164113840233877,
+      "grad_norm": 2.125,
+      "learning_rate": 6.442912509555316e-06,
+      "loss": 0.913,
+      "step": 6164
+    },
+    {
+      "epoch": 1.2166112790784838,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.44190351008763e-06,
+      "loss": 0.9804,
+      "step": 6165
+    },
+    {
+      "epoch": 1.2168111741335799,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.4408944465707435e-06,
+      "loss": 0.993,
+      "step": 6166
+    },
+    {
+      "epoch": 1.217011069188676,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.439885319049482e-06,
+      "loss": 0.9392,
+      "step": 6167
+    },
+    {
+      "epoch": 1.217210964243772,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.438876127568665e-06,
+      "loss": 0.9837,
+      "step": 6168
+    },
+    {
+      "epoch": 1.2174108592988682,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.437866872173127e-06,
+      "loss": 0.9335,
+      "step": 6169
+    },
+    {
+      "epoch": 1.2176107543539643,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.436857552907696e-06,
+      "loss": 0.9545,
+      "step": 6170
+    },
+    {
+      "epoch": 1.2178106494090604,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.435848169817205e-06,
+      "loss": 1.0012,
+      "step": 6171
+    },
+    {
+      "epoch": 1.2180105444641562,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.43483872294649e-06,
+      "loss": 1.0375,
+      "step": 6172
+    },
+    {
+      "epoch": 1.2182104395192523,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.433829212340394e-06,
+      "loss": 1.0023,
+      "step": 6173
+    },
+    {
+      "epoch": 1.2184103345743484,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.432819638043758e-06,
+      "loss": 0.9833,
+      "step": 6174
+    },
+    {
+      "epoch": 1.2186102296294445,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.431810000101425e-06,
+      "loss": 0.9473,
+      "step": 6175
+    },
+    {
+      "epoch": 1.2188101246845406,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.430800298558246e-06,
+      "loss": 0.9796,
+      "step": 6176
+    },
+    {
+      "epoch": 1.2190100197396367,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.429790533459071e-06,
+      "loss": 1.0393,
+      "step": 6177
+    },
+    {
+      "epoch": 1.2192099147947328,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.428780704848753e-06,
+      "loss": 1.0952,
+      "step": 6178
+    },
+    {
+      "epoch": 1.2194098098498287,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.427770812772147e-06,
+      "loss": 0.967,
+      "step": 6179
+    },
+    {
+      "epoch": 1.2196097049049248,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.426760857274115e-06,
+      "loss": 0.9787,
+      "step": 6180
+    },
+    {
+      "epoch": 1.219809599960021,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.425750838399519e-06,
+      "loss": 0.9558,
+      "step": 6181
+    },
+    {
+      "epoch": 1.220009495015117,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.4247407561932215e-06,
+      "loss": 0.9681,
+      "step": 6182
+    },
+    {
+      "epoch": 1.2202093900702131,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.423730610700092e-06,
+      "loss": 1.1079,
+      "step": 6183
+    },
+    {
+      "epoch": 1.2204092851253092,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.422720401965003e-06,
+      "loss": 0.9808,
+      "step": 6184
+    },
+    {
+      "epoch": 1.2206091801804053,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.421710130032824e-06,
+      "loss": 1.0888,
+      "step": 6185
+    },
+    {
+      "epoch": 1.2208090752355014,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.420699794948433e-06,
+      "loss": 0.9608,
+      "step": 6186
+    },
+    {
+      "epoch": 1.2210089702905975,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.419689396756709e-06,
+      "loss": 1.0215,
+      "step": 6187
+    },
+    {
+      "epoch": 1.2212088653456936,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.418678935502534e-06,
+      "loss": 0.9127,
+      "step": 6188
+    },
+    {
+      "epoch": 1.2214087604007895,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.4176684112307905e-06,
+      "loss": 0.9422,
+      "step": 6189
+    },
+    {
+      "epoch": 1.2216086554558856,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.41665782398637e-06,
+      "loss": 1.0003,
+      "step": 6190
+    },
+    {
+      "epoch": 1.2218085505109817,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.415647173814158e-06,
+      "loss": 0.9936,
+      "step": 6191
+    },
+    {
+      "epoch": 1.2220084455660778,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.414636460759052e-06,
+      "loss": 1.034,
+      "step": 6192
+    },
+    {
+      "epoch": 1.222208340621174,
+      "grad_norm": 2.0,
+      "learning_rate": 6.413625684865942e-06,
+      "loss": 0.8903,
+      "step": 6193
+    },
+    {
+      "epoch": 1.22240823567627,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.412614846179734e-06,
+      "loss": 0.9948,
+      "step": 6194
+    },
+    {
+      "epoch": 1.222608130731366,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.411603944745323e-06,
+      "loss": 1.0735,
+      "step": 6195
+    },
+    {
+      "epoch": 1.2228080257864622,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.410592980607616e-06,
+      "loss": 0.8989,
+      "step": 6196
+    },
+    {
+      "epoch": 1.223007920841558,
+      "grad_norm": 1.984375,
+      "learning_rate": 6.409581953811519e-06,
+      "loss": 0.8938,
+      "step": 6197
+    },
+    {
+      "epoch": 1.2232078158966542,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.408570864401944e-06,
+      "loss": 1.0571,
+      "step": 6198
+    },
+    {
+      "epoch": 1.2234077109517503,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.4075597124238e-06,
+      "loss": 1.0363,
+      "step": 6199
+    },
+    {
+      "epoch": 1.2236076060068464,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.4065484979220035e-06,
+      "loss": 0.938,
+      "step": 6200
+    },
+    {
+      "epoch": 1.2238075010619425,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.405537220941475e-06,
+      "loss": 0.9658,
+      "step": 6201
+    },
+    {
+      "epoch": 1.2240073961170386,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.404525881527133e-06,
+      "loss": 0.9345,
+      "step": 6202
+    },
+    {
+      "epoch": 1.2242072911721347,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.4035144797239e-06,
+      "loss": 1.0204,
+      "step": 6203
+    },
+    {
+      "epoch": 1.2244071862272308,
+      "grad_norm": 2.125,
+      "learning_rate": 6.402503015576705e-06,
+      "loss": 0.9567,
+      "step": 6204
+    },
+    {
+      "epoch": 1.2246070812823269,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.401491489130474e-06,
+      "loss": 0.9587,
+      "step": 6205
+    },
+    {
+      "epoch": 1.224806976337423,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.400479900430141e-06,
+      "loss": 1.0711,
+      "step": 6206
+    },
+    {
+      "epoch": 1.2250068713925188,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.399468249520641e-06,
+      "loss": 0.9917,
+      "step": 6207
+    },
+    {
+      "epoch": 1.225206766447615,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.398456536446912e-06,
+      "loss": 1.0016,
+      "step": 6208
+    },
+    {
+      "epoch": 1.225406661502711,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.397444761253892e-06,
+      "loss": 0.9884,
+      "step": 6209
+    },
+    {
+      "epoch": 1.2256065565578071,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.396432923986525e-06,
+      "loss": 1.0465,
+      "step": 6210
+    },
+    {
+      "epoch": 1.2258064516129032,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.3954210246897565e-06,
+      "loss": 0.8695,
+      "step": 6211
+    },
+    {
+      "epoch": 1.2260063466679993,
+      "grad_norm": 2.25,
+      "learning_rate": 6.3944090634085355e-06,
+      "loss": 1.0473,
+      "step": 6212
+    },
+    {
+      "epoch": 1.2262062417230954,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.393397040187812e-06,
+      "loss": 0.928,
+      "step": 6213
+    },
+    {
+      "epoch": 1.2264061367781913,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.392384955072541e-06,
+      "loss": 0.889,
+      "step": 6214
+    },
+    {
+      "epoch": 1.2266060318332874,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.39137280810768e-06,
+      "loss": 0.9301,
+      "step": 6215
+    },
+    {
+      "epoch": 1.2268059268883835,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.390360599338188e-06,
+      "loss": 0.947,
+      "step": 6216
+    },
+    {
+      "epoch": 1.2270058219434796,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.389348328809024e-06,
+      "loss": 1.0382,
+      "step": 6217
+    },
+    {
+      "epoch": 1.2272057169985757,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.388335996565158e-06,
+      "loss": 1.0804,
+      "step": 6218
+    },
+    {
+      "epoch": 1.2274056120536718,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.387323602651554e-06,
+      "loss": 0.894,
+      "step": 6219
+    },
+    {
+      "epoch": 1.227605507108768,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.386311147113185e-06,
+      "loss": 1.0323,
+      "step": 6220
+    },
+    {
+      "epoch": 1.227805402163864,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.385298629995021e-06,
+      "loss": 1.0044,
+      "step": 6221
+    },
+    {
+      "epoch": 1.2280052972189601,
+      "grad_norm": 2.375,
+      "learning_rate": 6.3842860513420416e-06,
+      "loss": 0.9877,
+      "step": 6222
+    },
+    {
+      "epoch": 1.2282051922740562,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.3832734111992236e-06,
+      "loss": 0.9811,
+      "step": 6223
+    },
+    {
+      "epoch": 1.228405087329152,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.382260709611547e-06,
+      "loss": 1.006,
+      "step": 6224
+    },
+    {
+      "epoch": 1.2286049823842482,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.381247946623997e-06,
+      "loss": 1.0315,
+      "step": 6225
+    },
+    {
+      "epoch": 1.2288048774393443,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.380235122281563e-06,
+      "loss": 0.9488,
+      "step": 6226
+    },
+    {
+      "epoch": 1.2290047724944404,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.379222236629231e-06,
+      "loss": 0.9502,
+      "step": 6227
+    },
+    {
+      "epoch": 1.2292046675495365,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.378209289711994e-06,
+      "loss": 0.9568,
+      "step": 6228
+    },
+    {
+      "epoch": 1.2294045626046326,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.377196281574849e-06,
+      "loss": 0.9754,
+      "step": 6229
+    },
+    {
+      "epoch": 1.2296044576597287,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.376183212262791e-06,
+      "loss": 0.9237,
+      "step": 6230
+    },
+    {
+      "epoch": 1.2298043527148248,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.375170081820823e-06,
+      "loss": 1.066,
+      "step": 6231
+    },
+    {
+      "epoch": 1.2300042477699207,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.374156890293947e-06,
+      "loss": 1.0714,
+      "step": 6232
+    },
+    {
+      "epoch": 1.2302041428250168,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.373143637727166e-06,
+      "loss": 0.9691,
+      "step": 6233
+    },
+    {
+      "epoch": 1.2304040378801129,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.372130324165493e-06,
+      "loss": 0.8767,
+      "step": 6234
+    },
+    {
+      "epoch": 1.230603932935209,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.3711169496539385e-06,
+      "loss": 0.9945,
+      "step": 6235
+    },
+    {
+      "epoch": 1.230803827990305,
+      "grad_norm": 2.125,
+      "learning_rate": 6.370103514237513e-06,
+      "loss": 0.9998,
+      "step": 6236
+    },
+    {
+      "epoch": 1.2310037230454012,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.3690900179612395e-06,
+      "loss": 0.9744,
+      "step": 6237
+    },
+    {
+      "epoch": 1.2312036181004973,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.36807646087013e-06,
+      "loss": 1.0184,
+      "step": 6238
+    },
+    {
+      "epoch": 1.2314035131555934,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.367062843009211e-06,
+      "loss": 0.9687,
+      "step": 6239
+    },
+    {
+      "epoch": 1.2316034082106895,
+      "grad_norm": 1.9921875,
+      "learning_rate": 6.366049164423508e-06,
+      "loss": 0.9406,
+      "step": 6240
+    },
+    {
+      "epoch": 1.2318033032657856,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.365035425158046e-06,
+      "loss": 0.9448,
+      "step": 6241
+    },
+    {
+      "epoch": 1.2320031983208815,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.364021625257856e-06,
+      "loss": 1.001,
+      "step": 6242
+    },
+    {
+      "epoch": 1.2322030933759776,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.363007764767972e-06,
+      "loss": 0.9919,
+      "step": 6243
+    },
+    {
+      "epoch": 1.2324029884310737,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.361993843733427e-06,
+      "loss": 1.0646,
+      "step": 6244
+    },
+    {
+      "epoch": 1.2326028834861698,
+      "grad_norm": 2.125,
+      "learning_rate": 6.360979862199262e-06,
+      "loss": 0.9798,
+      "step": 6245
+    },
+    {
+      "epoch": 1.2328027785412659,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.3599658202105175e-06,
+      "loss": 0.9565,
+      "step": 6246
+    },
+    {
+      "epoch": 1.233002673596362,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.3589517178122365e-06,
+      "loss": 0.8722,
+      "step": 6247
+    },
+    {
+      "epoch": 1.233202568651458,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.357937555049465e-06,
+      "loss": 0.9004,
+      "step": 6248
+    },
+    {
+      "epoch": 1.2334024637065542,
+      "grad_norm": 2.25,
+      "learning_rate": 6.356923331967252e-06,
+      "loss": 1.0188,
+      "step": 6249
+    },
+    {
+      "epoch": 1.23360235876165,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.355909048610649e-06,
+      "loss": 0.9297,
+      "step": 6250
+    },
+    {
+      "epoch": 1.2338022538167461,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.354894705024711e-06,
+      "loss": 0.9353,
+      "step": 6251
+    },
+    {
+      "epoch": 1.2340021488718422,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.353880301254496e-06,
+      "loss": 1.0263,
+      "step": 6252
+    },
+    {
+      "epoch": 1.2342020439269383,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.352865837345061e-06,
+      "loss": 1.0051,
+      "step": 6253
+    },
+    {
+      "epoch": 1.2344019389820344,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.351851313341473e-06,
+      "loss": 0.9922,
+      "step": 6254
+    },
+    {
+      "epoch": 1.2346018340371305,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.350836729288792e-06,
+      "loss": 0.9041,
+      "step": 6255
+    },
+    {
+      "epoch": 1.2348017290922266,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.349822085232087e-06,
+      "loss": 0.9869,
+      "step": 6256
+    },
+    {
+      "epoch": 1.2350016241473227,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.348807381216431e-06,
+      "loss": 0.9799,
+      "step": 6257
+    },
+    {
+      "epoch": 1.2352015192024188,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.347792617286893e-06,
+      "loss": 1.0542,
+      "step": 6258
+    },
+    {
+      "epoch": 1.235401414257515,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.346777793488552e-06,
+      "loss": 1.0436,
+      "step": 6259
+    },
+    {
+      "epoch": 1.2356013093126108,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.345762909866485e-06,
+      "loss": 1.0035,
+      "step": 6260
+    },
+    {
+      "epoch": 1.235801204367707,
+      "grad_norm": 2.125,
+      "learning_rate": 6.344747966465774e-06,
+      "loss": 0.9534,
+      "step": 6261
+    },
+    {
+      "epoch": 1.236001099422803,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.343732963331502e-06,
+      "loss": 0.9065,
+      "step": 6262
+    },
+    {
+      "epoch": 1.236200994477899,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.342717900508755e-06,
+      "loss": 0.9186,
+      "step": 6263
+    },
+    {
+      "epoch": 1.2364008895329952,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.341702778042622e-06,
+      "loss": 1.046,
+      "step": 6264
+    },
+    {
+      "epoch": 1.2366007845880913,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.3406875959781944e-06,
+      "loss": 0.9073,
+      "step": 6265
+    },
+    {
+      "epoch": 1.2368006796431874,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.339672354360568e-06,
+      "loss": 0.9849,
+      "step": 6266
+    },
+    {
+      "epoch": 1.2370005746982833,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.338657053234838e-06,
+      "loss": 0.9251,
+      "step": 6267
+    },
+    {
+      "epoch": 1.2372004697533794,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.337641692646106e-06,
+      "loss": 0.9606,
+      "step": 6268
+    },
+    {
+      "epoch": 1.2374003648084755,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.336626272639471e-06,
+      "loss": 0.9163,
+      "step": 6269
+    },
+    {
+      "epoch": 1.2376002598635716,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.3356107932600396e-06,
+      "loss": 0.9725,
+      "step": 6270
+    },
+    {
+      "epoch": 1.2378001549186677,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.334595254552921e-06,
+      "loss": 0.9324,
+      "step": 6271
+    },
+    {
+      "epoch": 1.2380000499737638,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.333579656563222e-06,
+      "loss": 0.9135,
+      "step": 6272
+    },
+    {
+      "epoch": 1.2381999450288599,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.332563999336059e-06,
+      "loss": 1.0486,
+      "step": 6273
+    },
+    {
+      "epoch": 1.238399840083956,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.331548282916545e-06,
+      "loss": 0.9279,
+      "step": 6274
+    },
+    {
+      "epoch": 1.238599735139052,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.330532507349798e-06,
+      "loss": 0.9609,
+      "step": 6275
+    },
+    {
+      "epoch": 1.2387996301941482,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.3295166726809396e-06,
+      "loss": 0.8932,
+      "step": 6276
+    },
+    {
+      "epoch": 1.238999525249244,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.328500778955091e-06,
+      "loss": 0.9976,
+      "step": 6277
+    },
+    {
+      "epoch": 1.2391994203043402,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.327484826217382e-06,
+      "loss": 0.9441,
+      "step": 6278
+    },
+    {
+      "epoch": 1.2393993153594363,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.326468814512937e-06,
+      "loss": 0.9518,
+      "step": 6279
+    },
+    {
+      "epoch": 1.2395992104145324,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.325452743886891e-06,
+      "loss": 0.9586,
+      "step": 6280
+    },
+    {
+      "epoch": 1.2397991054696285,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.324436614384374e-06,
+      "loss": 1.0174,
+      "step": 6281
+    },
+    {
+      "epoch": 1.2399990005247246,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.3234204260505235e-06,
+      "loss": 0.9922,
+      "step": 6282
+    },
+    {
+      "epoch": 1.2401988955798207,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.3224041789304804e-06,
+      "loss": 1.0269,
+      "step": 6283
+    },
+    {
+      "epoch": 1.2403987906349168,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.321387873069384e-06,
+      "loss": 0.9655,
+      "step": 6284
+    },
+    {
+      "epoch": 1.2405986856900126,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.3203715085123805e-06,
+      "loss": 0.9162,
+      "step": 6285
+    },
+    {
+      "epoch": 1.2407985807451087,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.319355085304615e-06,
+      "loss": 1.0348,
+      "step": 6286
+    },
+    {
+      "epoch": 1.2409984758002048,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.3183386034912364e-06,
+      "loss": 1.0138,
+      "step": 6287
+    },
+    {
+      "epoch": 1.241198370855301,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.317322063117399e-06,
+      "loss": 1.0809,
+      "step": 6288
+    },
+    {
+      "epoch": 1.241398265910397,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.316305464228256e-06,
+      "loss": 0.9187,
+      "step": 6289
+    },
+    {
+      "epoch": 1.2415981609654931,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.315288806868964e-06,
+      "loss": 0.99,
+      "step": 6290
+    },
+    {
+      "epoch": 1.2417980560205892,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.314272091084686e-06,
+      "loss": 1.0059,
+      "step": 6291
+    },
+    {
+      "epoch": 1.2419979510756853,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.31325531692058e-06,
+      "loss": 0.9138,
+      "step": 6292
+    },
+    {
+      "epoch": 1.2421978461307814,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.312238484421815e-06,
+      "loss": 0.948,
+      "step": 6293
+    },
+    {
+      "epoch": 1.2423977411858775,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.311221593633557e-06,
+      "loss": 1.025,
+      "step": 6294
+    },
+    {
+      "epoch": 1.2425976362409734,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.310204644600974e-06,
+      "loss": 1.0143,
+      "step": 6295
+    },
+    {
+      "epoch": 1.2427975312960695,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.30918763736924e-06,
+      "loss": 0.926,
+      "step": 6296
+    },
+    {
+      "epoch": 1.2429974263511656,
+      "grad_norm": 2.25,
+      "learning_rate": 6.3081705719835316e-06,
+      "loss": 1.0279,
+      "step": 6297
+    },
+    {
+      "epoch": 1.2431973214062617,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.307153448489028e-06,
+      "loss": 0.9368,
+      "step": 6298
+    },
+    {
+      "epoch": 1.2433972164613578,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.306136266930906e-06,
+      "loss": 0.9261,
+      "step": 6299
+    },
+    {
+      "epoch": 1.243597111516454,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.305119027354349e-06,
+      "loss": 0.9971,
+      "step": 6300
+    },
+    {
+      "epoch": 1.24379700657155,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.304101729804546e-06,
+      "loss": 0.9341,
+      "step": 6301
+    },
+    {
+      "epoch": 1.243996901626646,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.303084374326685e-06,
+      "loss": 1.0005,
+      "step": 6302
+    },
+    {
+      "epoch": 1.244196796681742,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.302066960965954e-06,
+      "loss": 0.8845,
+      "step": 6303
+    },
+    {
+      "epoch": 1.244396691736838,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.301049489767545e-06,
+      "loss": 1.0183,
+      "step": 6304
+    },
+    {
+      "epoch": 1.2445965867919342,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.30003196077666e-06,
+      "loss": 0.9922,
+      "step": 6305
+    },
+    {
+      "epoch": 1.2447964818470303,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.299014374038493e-06,
+      "loss": 0.9394,
+      "step": 6306
+    },
+    {
+      "epoch": 1.2449963769021264,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.297996729598247e-06,
+      "loss": 0.9718,
+      "step": 6307
+    },
+    {
+      "epoch": 1.2451962719572225,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.2969790275011245e-06,
+      "loss": 1.0739,
+      "step": 6308
+    },
+    {
+      "epoch": 1.2453961670123186,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.295961267792332e-06,
+      "loss": 0.9442,
+      "step": 6309
+    },
+    {
+      "epoch": 1.2455960620674147,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.294943450517078e-06,
+      "loss": 0.926,
+      "step": 6310
+    },
+    {
+      "epoch": 1.2457959571225108,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.293925575720575e-06,
+      "loss": 0.991,
+      "step": 6311
+    },
+    {
+      "epoch": 1.2459958521776067,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.292907643448035e-06,
+      "loss": 0.9396,
+      "step": 6312
+    },
+    {
+      "epoch": 1.2461957472327028,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.291889653744677e-06,
+      "loss": 1.037,
+      "step": 6313
+    },
+    {
+      "epoch": 1.2463956422877989,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.290871606655718e-06,
+      "loss": 0.9848,
+      "step": 6314
+    },
+    {
+      "epoch": 1.246595537342895,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.2898535022263795e-06,
+      "loss": 0.9971,
+      "step": 6315
+    },
+    {
+      "epoch": 1.246795432397991,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.288835340501886e-06,
+      "loss": 1.0538,
+      "step": 6316
+    },
+    {
+      "epoch": 1.2469953274530872,
+      "grad_norm": 2.125,
+      "learning_rate": 6.287817121527465e-06,
+      "loss": 1.141,
+      "step": 6317
+    },
+    {
+      "epoch": 1.2471952225081833,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.286798845348345e-06,
+      "loss": 1.0421,
+      "step": 6318
+    },
+    {
+      "epoch": 1.2473951175632794,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.285780512009758e-06,
+      "loss": 0.9491,
+      "step": 6319
+    },
+    {
+      "epoch": 1.2475950126183752,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.284762121556937e-06,
+      "loss": 0.9259,
+      "step": 6320
+    },
+    {
+      "epoch": 1.2477949076734713,
+      "grad_norm": 2.125,
+      "learning_rate": 6.283743674035121e-06,
+      "loss": 0.9349,
+      "step": 6321
+    },
+    {
+      "epoch": 1.2479948027285674,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.282725169489547e-06,
+      "loss": 0.8958,
+      "step": 6322
+    },
+    {
+      "epoch": 1.2481946977836635,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.281706607965459e-06,
+      "loss": 0.8789,
+      "step": 6323
+    },
+    {
+      "epoch": 1.2483945928387596,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.2806879895081006e-06,
+      "loss": 1.0401,
+      "step": 6324
+    },
+    {
+      "epoch": 1.2485944878938557,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.279669314162717e-06,
+      "loss": 0.9825,
+      "step": 6325
+    },
+    {
+      "epoch": 1.2487943829489518,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.2786505819745604e-06,
+      "loss": 0.8916,
+      "step": 6326
+    },
+    {
+      "epoch": 1.248994278004048,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.277631792988882e-06,
+      "loss": 0.985,
+      "step": 6327
+    },
+    {
+      "epoch": 1.249194173059144,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.276612947250934e-06,
+      "loss": 0.9907,
+      "step": 6328
+    },
+    {
+      "epoch": 1.2493940681142401,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.275594044805976e-06,
+      "loss": 1.0056,
+      "step": 6329
+    },
+    {
+      "epoch": 1.249593963169336,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.274575085699267e-06,
+      "loss": 0.9899,
+      "step": 6330
+    },
+    {
+      "epoch": 1.2497938582244321,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.2735560699760676e-06,
+      "loss": 0.9518,
+      "step": 6331
+    },
+    {
+      "epoch": 1.2499937532795282,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.2725369976816455e-06,
+      "loss": 1.0028,
+      "step": 6332
+    },
+    {
+      "epoch": 1.2501936483346243,
+      "grad_norm": 2.125,
+      "learning_rate": 6.271517868861266e-06,
+      "loss": 0.9737,
+      "step": 6333
+    },
+    {
+      "epoch": 1.2503935433897204,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.270498683560195e-06,
+      "loss": 0.944,
+      "step": 6334
+    },
+    {
+      "epoch": 1.2505934384448165,
+      "grad_norm": 2.25,
+      "learning_rate": 6.269479441823712e-06,
+      "loss": 0.9353,
+      "step": 6335
+    },
+    {
+      "epoch": 1.2507933334999126,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.268460143697086e-06,
+      "loss": 1.0613,
+      "step": 6336
+    },
+    {
+      "epoch": 1.2509932285550085,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.267440789225596e-06,
+      "loss": 0.9998,
+      "step": 6337
+    },
+    {
+      "epoch": 1.2511931236101046,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.266421378454524e-06,
+      "loss": 0.9825,
+      "step": 6338
+    },
+    {
+      "epoch": 1.2513930186652007,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.265401911429147e-06,
+      "loss": 0.8972,
+      "step": 6339
+    },
+    {
+      "epoch": 1.2515929137202968,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.264382388194753e-06,
+      "loss": 1.0687,
+      "step": 6340
+    },
+    {
+      "epoch": 1.251792808775393,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.263362808796627e-06,
+      "loss": 1.0331,
+      "step": 6341
+    },
+    {
+      "epoch": 1.251992703830489,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.262343173280062e-06,
+      "loss": 0.9854,
+      "step": 6342
+    },
+    {
+      "epoch": 1.252192598885585,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.261323481690347e-06,
+      "loss": 1.0489,
+      "step": 6343
+    },
+    {
+      "epoch": 1.2523924939406812,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.260303734072778e-06,
+      "loss": 0.9546,
+      "step": 6344
+    },
+    {
+      "epoch": 1.2525923889957773,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.2592839304726515e-06,
+      "loss": 1.035,
+      "step": 6345
+    },
+    {
+      "epoch": 1.2527922840508734,
+      "grad_norm": 1.9921875,
+      "learning_rate": 6.258264070935267e-06,
+      "loss": 0.933,
+      "step": 6346
+    },
+    {
+      "epoch": 1.2529921791059695,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.257244155505928e-06,
+      "loss": 0.9164,
+      "step": 6347
+    },
+    {
+      "epoch": 1.2531920741610654,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.256224184229936e-06,
+      "loss": 0.984,
+      "step": 6348
+    },
+    {
+      "epoch": 1.2533919692161615,
+      "grad_norm": 2.125,
+      "learning_rate": 6.255204157152601e-06,
+      "loss": 0.9565,
+      "step": 6349
+    },
+    {
+      "epoch": 1.2535918642712576,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.254184074319231e-06,
+      "loss": 1.0095,
+      "step": 6350
+    },
+    {
+      "epoch": 1.2537917593263537,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.253163935775139e-06,
+      "loss": 0.994,
+      "step": 6351
+    },
+    {
+      "epoch": 1.2539916543814498,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.252143741565639e-06,
+      "loss": 0.9789,
+      "step": 6352
+    },
+    {
+      "epoch": 1.2541915494365459,
+      "grad_norm": 2.125,
+      "learning_rate": 6.251123491736048e-06,
+      "loss": 0.9684,
+      "step": 6353
+    },
+    {
+      "epoch": 1.2543914444916417,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.250103186331684e-06,
+      "loss": 1.0011,
+      "step": 6354
+    },
+    {
+      "epoch": 1.2545913395467378,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.249082825397871e-06,
+      "loss": 0.9764,
+      "step": 6355
+    },
+    {
+      "epoch": 1.254791234601834,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.248062408979933e-06,
+      "loss": 0.988,
+      "step": 6356
+    },
+    {
+      "epoch": 1.25499112965693,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.247041937123194e-06,
+      "loss": 0.992,
+      "step": 6357
+    },
+    {
+      "epoch": 1.2551910247120261,
+      "grad_norm": 2.125,
+      "learning_rate": 6.246021409872987e-06,
+      "loss": 0.9474,
+      "step": 6358
+    },
+    {
+      "epoch": 1.2553909197671222,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.2450008272746395e-06,
+      "loss": 1.0386,
+      "step": 6359
+    },
+    {
+      "epoch": 1.2555908148222183,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.24398018937349e-06,
+      "loss": 0.9926,
+      "step": 6360
+    },
+    {
+      "epoch": 1.2557907098773144,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.242959496214874e-06,
+      "loss": 1.0452,
+      "step": 6361
+    },
+    {
+      "epoch": 1.2559906049324105,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.241938747844129e-06,
+      "loss": 0.9355,
+      "step": 6362
+    },
+    {
+      "epoch": 1.2561904999875066,
+      "grad_norm": 2.125,
+      "learning_rate": 6.240917944306597e-06,
+      "loss": 0.9983,
+      "step": 6363
+    },
+    {
+      "epoch": 1.2563903950426027,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.239897085647624e-06,
+      "loss": 1.0291,
+      "step": 6364
+    },
+    {
+      "epoch": 1.2565902900976986,
+      "grad_norm": 2.25,
+      "learning_rate": 6.238876171912553e-06,
+      "loss": 0.9773,
+      "step": 6365
+    },
+    {
+      "epoch": 1.2567901851527947,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.237855203146737e-06,
+      "loss": 0.9554,
+      "step": 6366
+    },
+    {
+      "epoch": 1.2569900802078908,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.2368341793955235e-06,
+      "loss": 0.9327,
+      "step": 6367
+    },
+    {
+      "epoch": 1.257189975262987,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.235813100704267e-06,
+      "loss": 0.9893,
+      "step": 6368
+    },
+    {
+      "epoch": 1.257389870318083,
+      "grad_norm": 2.125,
+      "learning_rate": 6.2347919671183265e-06,
+      "loss": 1.056,
+      "step": 6369
+    },
+    {
+      "epoch": 1.2575897653731791,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.233770778683058e-06,
+      "loss": 0.9212,
+      "step": 6370
+    },
+    {
+      "epoch": 1.2577896604282752,
+      "grad_norm": 1.9765625,
+      "learning_rate": 6.2327495354438235e-06,
+      "loss": 0.9166,
+      "step": 6371
+    },
+    {
+      "epoch": 1.257989555483371,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.231728237445987e-06,
+      "loss": 0.9113,
+      "step": 6372
+    },
+    {
+      "epoch": 1.2581894505384672,
+      "grad_norm": 2.125,
+      "learning_rate": 6.230706884734913e-06,
+      "loss": 0.9421,
+      "step": 6373
+    },
+    {
+      "epoch": 1.2583893455935633,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.2296854773559685e-06,
+      "loss": 0.9322,
+      "step": 6374
+    },
+    {
+      "epoch": 1.2585892406486594,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.22866401535453e-06,
+      "loss": 1.1273,
+      "step": 6375
+    },
+    {
+      "epoch": 1.2587891357037555,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.227642498775965e-06,
+      "loss": 1.0321,
+      "step": 6376
+    },
+    {
+      "epoch": 1.2589890307588516,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.226620927665652e-06,
+      "loss": 0.9332,
+      "step": 6377
+    },
+    {
+      "epoch": 1.2591889258139477,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.225599302068968e-06,
+      "loss": 0.9504,
+      "step": 6378
+    },
+    {
+      "epoch": 1.2593888208690438,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.224577622031294e-06,
+      "loss": 0.9156,
+      "step": 6379
+    },
+    {
+      "epoch": 1.25958871592414,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.223555887598013e-06,
+      "loss": 0.9659,
+      "step": 6380
+    },
+    {
+      "epoch": 1.259788610979236,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.22253409881451e-06,
+      "loss": 0.9739,
+      "step": 6381
+    },
+    {
+      "epoch": 1.259988506034332,
+      "grad_norm": 1.9609375,
+      "learning_rate": 6.221512255726173e-06,
+      "loss": 0.8863,
+      "step": 6382
+    },
+    {
+      "epoch": 1.260188401089428,
+      "grad_norm": 2.25,
+      "learning_rate": 6.2204903583783924e-06,
+      "loss": 0.9899,
+      "step": 6383
+    },
+    {
+      "epoch": 1.260388296144524,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.21946840681656e-06,
+      "loss": 1.0136,
+      "step": 6384
+    },
+    {
+      "epoch": 1.2605881911996202,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.218446401086071e-06,
+      "loss": 1.0317,
+      "step": 6385
+    },
+    {
+      "epoch": 1.2607880862547163,
+      "grad_norm": 2.25,
+      "learning_rate": 6.217424341232324e-06,
+      "loss": 1.0341,
+      "step": 6386
+    },
+    {
+      "epoch": 1.2609879813098124,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.216402227300717e-06,
+      "loss": 0.9551,
+      "step": 6387
+    },
+    {
+      "epoch": 1.2611878763649085,
+      "grad_norm": 2.125,
+      "learning_rate": 6.215380059336653e-06,
+      "loss": 0.9913,
+      "step": 6388
+    },
+    {
+      "epoch": 1.2613877714200046,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.214357837385538e-06,
+      "loss": 0.8355,
+      "step": 6389
+    },
+    {
+      "epoch": 1.2615876664751005,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.213335561492775e-06,
+      "loss": 0.9165,
+      "step": 6390
+    },
+    {
+      "epoch": 1.2617875615301966,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.212313231703777e-06,
+      "loss": 0.9236,
+      "step": 6391
+    },
+    {
+      "epoch": 1.2619874565852927,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.211290848063955e-06,
+      "loss": 1.0185,
+      "step": 6392
+    },
+    {
+      "epoch": 1.2621873516403888,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.210268410618723e-06,
+      "loss": 1.0065,
+      "step": 6393
+    },
+    {
+      "epoch": 1.2623872466954849,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.209245919413497e-06,
+      "loss": 0.9274,
+      "step": 6394
+    },
+    {
+      "epoch": 1.262587141750581,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.208223374493697e-06,
+      "loss": 0.9656,
+      "step": 6395
+    },
+    {
+      "epoch": 1.262787036805677,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.2072007759047426e-06,
+      "loss": 0.9707,
+      "step": 6396
+    },
+    {
+      "epoch": 1.2629869318607732,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.20617812369206e-06,
+      "loss": 0.9844,
+      "step": 6397
+    },
+    {
+      "epoch": 1.2631868269158693,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.205155417901074e-06,
+      "loss": 0.9175,
+      "step": 6398
+    },
+    {
+      "epoch": 1.2633867219709654,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.204132658577212e-06,
+      "loss": 0.9726,
+      "step": 6399
+    },
+    {
+      "epoch": 1.2635866170260615,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.203109845765907e-06,
+      "loss": 0.9825,
+      "step": 6400
+    },
+    {
+      "epoch": 1.2637865120811573,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.202086979512589e-06,
+      "loss": 0.976,
+      "step": 6401
+    },
+    {
+      "epoch": 1.2639864071362534,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.201064059862699e-06,
+      "loss": 1.0102,
+      "step": 6402
+    },
+    {
+      "epoch": 1.2641863021913495,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.20004108686167e-06,
+      "loss": 0.9396,
+      "step": 6403
+    },
+    {
+      "epoch": 1.2643861972464456,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.199018060554945e-06,
+      "loss": 1.0043,
+      "step": 6404
+    },
+    {
+      "epoch": 1.2645860923015417,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.197994980987964e-06,
+      "loss": 0.9013,
+      "step": 6405
+    },
+    {
+      "epoch": 1.2647859873566378,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.196971848206175e-06,
+      "loss": 1.0221,
+      "step": 6406
+    },
+    {
+      "epoch": 1.2649858824117337,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.195948662255024e-06,
+      "loss": 0.8652,
+      "step": 6407
+    },
+    {
+      "epoch": 1.2651857774668298,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.194925423179961e-06,
+      "loss": 0.9215,
+      "step": 6408
+    },
+    {
+      "epoch": 1.265385672521926,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.193902131026439e-06,
+      "loss": 1.0518,
+      "step": 6409
+    },
+    {
+      "epoch": 1.265585567577022,
+      "grad_norm": 2.125,
+      "learning_rate": 6.192878785839911e-06,
+      "loss": 0.9335,
+      "step": 6410
+    },
+    {
+      "epoch": 1.265785462632118,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.191855387665836e-06,
+      "loss": 1.0277,
+      "step": 6411
+    },
+    {
+      "epoch": 1.2659853576872142,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.190831936549671e-06,
+      "loss": 0.9709,
+      "step": 6412
+    },
+    {
+      "epoch": 1.2661852527423103,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.189808432536879e-06,
+      "loss": 1.0246,
+      "step": 6413
+    },
+    {
+      "epoch": 1.2663851477974064,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.188784875672923e-06,
+      "loss": 0.9244,
+      "step": 6414
+    },
+    {
+      "epoch": 1.2665850428525025,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.1877612660032706e-06,
+      "loss": 0.9935,
+      "step": 6415
+    },
+    {
+      "epoch": 1.2667849379075986,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.1867376035733885e-06,
+      "loss": 0.975,
+      "step": 6416
+    },
+    {
+      "epoch": 1.2669848329626947,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.185713888428751e-06,
+      "loss": 0.9996,
+      "step": 6417
+    },
+    {
+      "epoch": 1.2671847280177906,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.1846901206148265e-06,
+      "loss": 0.9165,
+      "step": 6418
+    },
+    {
+      "epoch": 1.2673846230728867,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.183666300177095e-06,
+      "loss": 0.92,
+      "step": 6419
+    },
+    {
+      "epoch": 1.2675845181279828,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.182642427161033e-06,
+      "loss": 0.9924,
+      "step": 6420
+    },
+    {
+      "epoch": 1.2677844131830789,
+      "grad_norm": 2.375,
+      "learning_rate": 6.181618501612119e-06,
+      "loss": 1.0589,
+      "step": 6421
+    },
+    {
+      "epoch": 1.267984308238175,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.180594523575838e-06,
+      "loss": 0.9984,
+      "step": 6422
+    },
+    {
+      "epoch": 1.268184203293271,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.179570493097675e-06,
+      "loss": 0.9574,
+      "step": 6423
+    },
+    {
+      "epoch": 1.2683840983483672,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.178546410223116e-06,
+      "loss": 0.9633,
+      "step": 6424
+    },
+    {
+      "epoch": 1.268583993403463,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.177522274997651e-06,
+      "loss": 0.9466,
+      "step": 6425
+    },
+    {
+      "epoch": 1.2687838884585592,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.176498087466775e-06,
+      "loss": 0.938,
+      "step": 6426
+    },
+    {
+      "epoch": 1.2689837835136553,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.1754738476759764e-06,
+      "loss": 1.0154,
+      "step": 6427
+    },
+    {
+      "epoch": 1.2691836785687514,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.174449555670757e-06,
+      "loss": 0.9987,
+      "step": 6428
+    },
+    {
+      "epoch": 1.2693835736238475,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.173425211496613e-06,
+      "loss": 0.9349,
+      "step": 6429
+    },
+    {
+      "epoch": 1.2695834686789436,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.1724008151990465e-06,
+      "loss": 1.0085,
+      "step": 6430
+    },
+    {
+      "epoch": 1.2697833637340397,
+      "grad_norm": 2.0,
+      "learning_rate": 6.171376366823562e-06,
+      "loss": 0.9886,
+      "step": 6431
+    },
+    {
+      "epoch": 1.2699832587891358,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.1703518664156636e-06,
+      "loss": 1.0036,
+      "step": 6432
+    },
+    {
+      "epoch": 1.2701831538442319,
+      "grad_norm": 2.25,
+      "learning_rate": 6.16932731402086e-06,
+      "loss": 0.9389,
+      "step": 6433
+    },
+    {
+      "epoch": 1.270383048899328,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.168302709684663e-06,
+      "loss": 0.9765,
+      "step": 6434
+    },
+    {
+      "epoch": 1.270582943954424,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.1672780534525825e-06,
+      "loss": 1.0638,
+      "step": 6435
+    },
+    {
+      "epoch": 1.27078283900952,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.166253345370137e-06,
+      "loss": 0.9807,
+      "step": 6436
+    },
+    {
+      "epoch": 1.270982734064616,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.165228585482842e-06,
+      "loss": 0.9761,
+      "step": 6437
+    },
+    {
+      "epoch": 1.2711826291197121,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.164203773836217e-06,
+      "loss": 0.8968,
+      "step": 6438
+    },
+    {
+      "epoch": 1.2713825241748082,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.163178910475786e-06,
+      "loss": 0.9883,
+      "step": 6439
+    },
+    {
+      "epoch": 1.2715824192299043,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.162153995447071e-06,
+      "loss": 0.9872,
+      "step": 6440
+    },
+    {
+      "epoch": 1.2717823142850004,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.1611290287956e-06,
+      "loss": 1.0828,
+      "step": 6441
+    },
+    {
+      "epoch": 1.2719822093400963,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.160104010566902e-06,
+      "loss": 1.0257,
+      "step": 6442
+    },
+    {
+      "epoch": 1.2721821043951924,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.159078940806507e-06,
+      "loss": 0.9679,
+      "step": 6443
+    },
+    {
+      "epoch": 1.2723819994502885,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.15805381955995e-06,
+      "loss": 0.9412,
+      "step": 6444
+    },
+    {
+      "epoch": 1.2725818945053846,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.157028646872765e-06,
+      "loss": 1.0561,
+      "step": 6445
+    },
+    {
+      "epoch": 1.2727817895604807,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.156003422790492e-06,
+      "loss": 0.9577,
+      "step": 6446
+    },
+    {
+      "epoch": 1.2729816846155768,
+      "grad_norm": 2.125,
+      "learning_rate": 6.15497814735867e-06,
+      "loss": 0.9266,
+      "step": 6447
+    },
+    {
+      "epoch": 1.273181579670673,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.153952820622844e-06,
+      "loss": 0.9729,
+      "step": 6448
+    },
+    {
+      "epoch": 1.273381474725769,
+      "grad_norm": 2.125,
+      "learning_rate": 6.1529274426285546e-06,
+      "loss": 0.9299,
+      "step": 6449
+    },
+    {
+      "epoch": 1.2735813697808651,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.151902013421351e-06,
+      "loss": 1.0202,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2737812648359612,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.150876533046784e-06,
+      "loss": 0.9122,
+      "step": 6451
+    },
+    {
+      "epoch": 1.2739811598910573,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.149851001550404e-06,
+      "loss": 0.8825,
+      "step": 6452
+    },
+    {
+      "epoch": 1.2741810549461532,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.1488254189777665e-06,
+      "loss": 0.9751,
+      "step": 6453
+    },
+    {
+      "epoch": 1.2743809500012493,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.147799785374425e-06,
+      "loss": 1.0231,
+      "step": 6454
+    },
+    {
+      "epoch": 1.2745808450563454,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.1467741007859405e-06,
+      "loss": 1.0558,
+      "step": 6455
+    },
+    {
+      "epoch": 1.2747807401114415,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.145748365257873e-06,
+      "loss": 0.9363,
+      "step": 6456
+    },
+    {
+      "epoch": 1.2749806351665376,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.1447225788357845e-06,
+      "loss": 0.9968,
+      "step": 6457
+    },
+    {
+      "epoch": 1.2751805302216337,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.143696741565241e-06,
+      "loss": 0.9579,
+      "step": 6458
+    },
+    {
+      "epoch": 1.2753804252767298,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.142670853491813e-06,
+      "loss": 0.9892,
+      "step": 6459
+    },
+    {
+      "epoch": 1.2755803203318257,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.141644914661066e-06,
+      "loss": 0.9936,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2757802153869218,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.140618925118574e-06,
+      "loss": 1.0118,
+      "step": 6461
+    },
+    {
+      "epoch": 1.2759801104420179,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.13959288490991e-06,
+      "loss": 0.9798,
+      "step": 6462
+    },
+    {
+      "epoch": 1.276180005497114,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.138566794080655e-06,
+      "loss": 0.8971,
+      "step": 6463
+    },
+    {
+      "epoch": 1.27637990055221,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.137540652676382e-06,
+      "loss": 0.9741,
+      "step": 6464
+    },
+    {
+      "epoch": 1.2765797956073062,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.136514460742675e-06,
+      "loss": 0.9904,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2767796906624023,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.135488218325116e-06,
+      "loss": 1.0251,
+      "step": 6466
+    },
+    {
+      "epoch": 1.2769795857174984,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.134461925469293e-06,
+      "loss": 1.0236,
+      "step": 6467
+    },
+    {
+      "epoch": 1.2771794807725945,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.133435582220794e-06,
+      "loss": 0.9359,
+      "step": 6468
+    },
+    {
+      "epoch": 1.2773793758276906,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.132409188625205e-06,
+      "loss": 1.0094,
+      "step": 6469
+    },
+    {
+      "epoch": 1.2775792708827867,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.13138274472812e-06,
+      "loss": 1.0281,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2777791659378825,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.130356250575137e-06,
+      "loss": 1.0421,
+      "step": 6471
+    },
+    {
+      "epoch": 1.2779790609929786,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.129329706211849e-06,
+      "loss": 1.0368,
+      "step": 6472
+    },
+    {
+      "epoch": 1.2781789560480747,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.1283031116838565e-06,
+      "loss": 0.9902,
+      "step": 6473
+    },
+    {
+      "epoch": 1.2783788511031708,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.127276467036759e-06,
+      "loss": 0.9434,
+      "step": 6474
+    },
+    {
+      "epoch": 1.278578746158267,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.126249772316161e-06,
+      "loss": 0.9404,
+      "step": 6475
+    },
+    {
+      "epoch": 1.278778641213363,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.1252230275676705e-06,
+      "loss": 0.9842,
+      "step": 6476
+    },
+    {
+      "epoch": 1.278978536268459,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.124196232836892e-06,
+      "loss": 1.0444,
+      "step": 6477
+    },
+    {
+      "epoch": 1.279178431323555,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.123169388169437e-06,
+      "loss": 0.9739,
+      "step": 6478
+    },
+    {
+      "epoch": 1.2793783263786511,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.122142493610918e-06,
+      "loss": 1.0631,
+      "step": 6479
+    },
+    {
+      "epoch": 1.2795782214337472,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.12111554920695e-06,
+      "loss": 0.9143,
+      "step": 6480
+    },
+    {
+      "epoch": 1.2797781164888433,
+      "grad_norm": 2.25,
+      "learning_rate": 6.120088555003148e-06,
+      "loss": 1.0243,
+      "step": 6481
+    },
+    {
+      "epoch": 1.2799780115439394,
+      "grad_norm": 2.25,
+      "learning_rate": 6.119061511045133e-06,
+      "loss": 1.0501,
+      "step": 6482
+    },
+    {
+      "epoch": 1.2801779065990355,
+      "grad_norm": 2.125,
+      "learning_rate": 6.118034417378523e-06,
+      "loss": 1.0083,
+      "step": 6483
+    },
+    {
+      "epoch": 1.2803778016541316,
+      "grad_norm": 2.125,
+      "learning_rate": 6.117007274048945e-06,
+      "loss": 0.9384,
+      "step": 6484
+    },
+    {
+      "epoch": 1.2805776967092277,
+      "grad_norm": 2.0,
+      "learning_rate": 6.115980081102024e-06,
+      "loss": 0.9705,
+      "step": 6485
+    },
+    {
+      "epoch": 1.2807775917643238,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.114952838583386e-06,
+      "loss": 1.0992,
+      "step": 6486
+    },
+    {
+      "epoch": 1.28097748681942,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.1139255465386614e-06,
+      "loss": 0.9487,
+      "step": 6487
+    },
+    {
+      "epoch": 1.2811773818745158,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.1128982050134845e-06,
+      "loss": 0.9714,
+      "step": 6488
+    },
+    {
+      "epoch": 1.281377276929612,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.111870814053487e-06,
+      "loss": 0.9044,
+      "step": 6489
+    },
+    {
+      "epoch": 1.281577171984708,
+      "grad_norm": 2.125,
+      "learning_rate": 6.110843373704307e-06,
+      "loss": 0.9707,
+      "step": 6490
+    },
+    {
+      "epoch": 1.281777067039804,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.109815884011583e-06,
+      "loss": 0.9192,
+      "step": 6491
+    },
+    {
+      "epoch": 1.2819769620949002,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.108788345020955e-06,
+      "loss": 0.8727,
+      "step": 6492
+    },
+    {
+      "epoch": 1.2821768571499963,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.10776075677807e-06,
+      "loss": 1.0429,
+      "step": 6493
+    },
+    {
+      "epoch": 1.2823767522050924,
+      "grad_norm": 2.25,
+      "learning_rate": 6.106733119328568e-06,
+      "loss": 1.0389,
+      "step": 6494
+    },
+    {
+      "epoch": 1.2825766472601883,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.1057054327181e-06,
+      "loss": 1.047,
+      "step": 6495
+    },
+    {
+      "epoch": 1.2827765423152844,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.104677696992315e-06,
+      "loss": 0.8999,
+      "step": 6496
+    },
+    {
+      "epoch": 1.2829764373703805,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.103649912196864e-06,
+      "loss": 0.9554,
+      "step": 6497
+    },
+    {
+      "epoch": 1.2831763324254766,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.102622078377401e-06,
+      "loss": 1.0413,
+      "step": 6498
+    },
+    {
+      "epoch": 1.2833762274805727,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.101594195579585e-06,
+      "loss": 1.0102,
+      "step": 6499
+    },
+    {
+      "epoch": 1.2835761225356688,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.1005662638490715e-06,
+      "loss": 0.9513,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2837760175907649,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.099538283231523e-06,
+      "loss": 1.0219,
+      "step": 6501
+    },
+    {
+      "epoch": 1.283975912645861,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.098510253772599e-06,
+      "loss": 0.9539,
+      "step": 6502
+    },
+    {
+      "epoch": 1.284175807700957,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.097482175517968e-06,
+      "loss": 0.9664,
+      "step": 6503
+    },
+    {
+      "epoch": 1.2843757027560532,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.096454048513297e-06,
+      "loss": 0.9948,
+      "step": 6504
+    },
+    {
+      "epoch": 1.2845755978111493,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.0954258728042535e-06,
+      "loss": 0.9397,
+      "step": 6505
+    },
+    {
+      "epoch": 1.2847754928662452,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.0943976484365095e-06,
+      "loss": 1.0426,
+      "step": 6506
+    },
+    {
+      "epoch": 1.2849753879213412,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.09336937545574e-06,
+      "loss": 0.9307,
+      "step": 6507
+    },
+    {
+      "epoch": 1.2851752829764373,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.092341053907618e-06,
+      "loss": 0.9363,
+      "step": 6508
+    },
+    {
+      "epoch": 1.2853751780315334,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.091312683837823e-06,
+      "loss": 0.9501,
+      "step": 6509
+    },
+    {
+      "epoch": 1.2855750730866295,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.090284265292038e-06,
+      "loss": 0.9455,
+      "step": 6510
+    },
+    {
+      "epoch": 1.2857749681417256,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.089255798315941e-06,
+      "loss": 1.0456,
+      "step": 6511
+    },
+    {
+      "epoch": 1.2859748631968217,
+      "grad_norm": 2.125,
+      "learning_rate": 6.088227282955216e-06,
+      "loss": 0.9585,
+      "step": 6512
+    },
+    {
+      "epoch": 1.2861747582519176,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.0871987192555524e-06,
+      "loss": 1.0309,
+      "step": 6513
+    },
+    {
+      "epoch": 1.2863746533070137,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.086170107262638e-06,
+      "loss": 0.9564,
+      "step": 6514
+    },
+    {
+      "epoch": 1.2865745483621098,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.085141447022162e-06,
+      "loss": 1.031,
+      "step": 6515
+    },
+    {
+      "epoch": 1.286774443417206,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.084112738579821e-06,
+      "loss": 1.0041,
+      "step": 6516
+    },
+    {
+      "epoch": 1.286974338472302,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.083083981981307e-06,
+      "loss": 1.0031,
+      "step": 6517
+    },
+    {
+      "epoch": 1.2871742335273981,
+      "grad_norm": 2.125,
+      "learning_rate": 6.0820551772723195e-06,
+      "loss": 1.0031,
+      "step": 6518
+    },
+    {
+      "epoch": 1.2873741285824942,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.081026324498554e-06,
+      "loss": 0.9412,
+      "step": 6519
+    },
+    {
+      "epoch": 1.2875740236375903,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.0799974237057155e-06,
+      "loss": 1.1451,
+      "step": 6520
+    },
+    {
+      "epoch": 1.2877739186926864,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.078968474939508e-06,
+      "loss": 1.0042,
+      "step": 6521
+    },
+    {
+      "epoch": 1.2879738137477825,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.0779394782456355e-06,
+      "loss": 0.9027,
+      "step": 6522
+    },
+    {
+      "epoch": 1.2881737088028786,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.076910433669806e-06,
+      "loss": 0.9414,
+      "step": 6523
+    },
+    {
+      "epoch": 1.2883736038579745,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.0758813412577315e-06,
+      "loss": 1.0103,
+      "step": 6524
+    },
+    {
+      "epoch": 1.2885734989130706,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.074852201055121e-06,
+      "loss": 1.0374,
+      "step": 6525
+    },
+    {
+      "epoch": 1.2887733939681667,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.0738230131076915e-06,
+      "loss": 1.066,
+      "step": 6526
+    },
+    {
+      "epoch": 1.2889732890232628,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.072793777461159e-06,
+      "loss": 1.0472,
+      "step": 6527
+    },
+    {
+      "epoch": 1.289173184078359,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.07176449416124e-06,
+      "loss": 0.9168,
+      "step": 6528
+    },
+    {
+      "epoch": 1.289373079133455,
+      "grad_norm": 2.125,
+      "learning_rate": 6.070735163253656e-06,
+      "loss": 0.9151,
+      "step": 6529
+    },
+    {
+      "epoch": 1.2895729741885509,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.069705784784133e-06,
+      "loss": 1.0055,
+      "step": 6530
+    },
+    {
+      "epoch": 1.289772869243647,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.068676358798391e-06,
+      "loss": 0.9039,
+      "step": 6531
+    },
+    {
+      "epoch": 1.289972764298743,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.067646885342161e-06,
+      "loss": 0.9601,
+      "step": 6532
+    },
+    {
+      "epoch": 1.2901726593538392,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.066617364461169e-06,
+      "loss": 0.8968,
+      "step": 6533
+    },
+    {
+      "epoch": 1.2903725544089353,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.065587796201148e-06,
+      "loss": 0.9923,
+      "step": 6534
+    },
+    {
+      "epoch": 1.2905724494640314,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.0645581806078315e-06,
+      "loss": 1.0246,
+      "step": 6535
+    },
+    {
+      "epoch": 1.2907723445191275,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.063528517726953e-06,
+      "loss": 0.9303,
+      "step": 6536
+    },
+    {
+      "epoch": 1.2909722395742236,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.0624988076042525e-06,
+      "loss": 0.9379,
+      "step": 6537
+    },
+    {
+      "epoch": 1.2911721346293197,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.061469050285469e-06,
+      "loss": 0.9697,
+      "step": 6538
+    },
+    {
+      "epoch": 1.2913720296844158,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.060439245816342e-06,
+      "loss": 0.9549,
+      "step": 6539
+    },
+    {
+      "epoch": 1.2915719247395119,
+      "grad_norm": 2.0,
+      "learning_rate": 6.059409394242618e-06,
+      "loss": 0.9163,
+      "step": 6540
+    },
+    {
+      "epoch": 1.2917718197946078,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.058379495610043e-06,
+      "loss": 1.0255,
+      "step": 6541
+    },
+    {
+      "epoch": 1.2919717148497039,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.057349549964362e-06,
+      "loss": 0.9965,
+      "step": 6542
+    },
+    {
+      "epoch": 1.2921716099048,
+      "grad_norm": 2.25,
+      "learning_rate": 6.056319557351327e-06,
+      "loss": 0.9442,
+      "step": 6543
+    },
+    {
+      "epoch": 1.292371504959896,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.055289517816691e-06,
+      "loss": 1.004,
+      "step": 6544
+    },
+    {
+      "epoch": 1.2925714000149922,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.054259431406206e-06,
+      "loss": 1.0521,
+      "step": 6545
+    },
+    {
+      "epoch": 1.2927712950700883,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.0532292981656315e-06,
+      "loss": 0.9739,
+      "step": 6546
+    },
+    {
+      "epoch": 1.2929711901251844,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.0521991181407225e-06,
+      "loss": 0.9814,
+      "step": 6547
+    },
+    {
+      "epoch": 1.2931710851802802,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.051168891377242e-06,
+      "loss": 0.9492,
+      "step": 6548
+    },
+    {
+      "epoch": 1.2933709802353763,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.050138617920951e-06,
+      "loss": 1.0261,
+      "step": 6549
+    },
+    {
+      "epoch": 1.2935708752904724,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.049108297817614e-06,
+      "loss": 1.0818,
+      "step": 6550
+    },
+    {
+      "epoch": 1.2937707703455685,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.048077931112999e-06,
+      "loss": 0.9663,
+      "step": 6551
+    },
+    {
+      "epoch": 1.2939706654006646,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.047047517852875e-06,
+      "loss": 1.0825,
+      "step": 6552
+    },
+    {
+      "epoch": 1.2941705604557607,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.046017058083011e-06,
+      "loss": 0.9238,
+      "step": 6553
+    },
+    {
+      "epoch": 1.2943704555108568,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.0449865518491825e-06,
+      "loss": 1.0634,
+      "step": 6554
+    },
+    {
+      "epoch": 1.294570350565953,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.0439559991971615e-06,
+      "loss": 0.9655,
+      "step": 6555
+    },
+    {
+      "epoch": 1.294770245621049,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.042925400172729e-06,
+      "loss": 0.971,
+      "step": 6556
+    },
+    {
+      "epoch": 1.2949701406761451,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.041894754821659e-06,
+      "loss": 1.0065,
+      "step": 6557
+    },
+    {
+      "epoch": 1.2951700357312412,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.040864063189738e-06,
+      "loss": 1.0393,
+      "step": 6558
+    },
+    {
+      "epoch": 1.295369930786337,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.039833325322744e-06,
+      "loss": 1.006,
+      "step": 6559
+    },
+    {
+      "epoch": 1.2955698258414332,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.038802541266464e-06,
+      "loss": 0.9824,
+      "step": 6560
+    },
+    {
+      "epoch": 1.2957697208965293,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.037771711066688e-06,
+      "loss": 0.8906,
+      "step": 6561
+    },
+    {
+      "epoch": 1.2959696159516254,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.036740834769202e-06,
+      "loss": 1.0133,
+      "step": 6562
+    },
+    {
+      "epoch": 1.2961695110067215,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.035709912419801e-06,
+      "loss": 0.9655,
+      "step": 6563
+    },
+    {
+      "epoch": 1.2963694060618176,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.0346789440642736e-06,
+      "loss": 0.9906,
+      "step": 6564
+    },
+    {
+      "epoch": 1.2965693011169135,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.0336479297484195e-06,
+      "loss": 1.0288,
+      "step": 6565
+    },
+    {
+      "epoch": 1.2967691961720096,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.032616869518036e-06,
+      "loss": 1.0503,
+      "step": 6566
+    },
+    {
+      "epoch": 1.2969690912271057,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.031585763418919e-06,
+      "loss": 0.9395,
+      "step": 6567
+    },
+    {
+      "epoch": 1.2971689862822018,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.030554611496873e-06,
+      "loss": 1.0274,
+      "step": 6568
+    },
+    {
+      "epoch": 1.2973688813372979,
+      "grad_norm": 2.125,
+      "learning_rate": 6.029523413797703e-06,
+      "loss": 0.9283,
+      "step": 6569
+    },
+    {
+      "epoch": 1.297568776392394,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.028492170367211e-06,
+      "loss": 0.9313,
+      "step": 6570
+    },
+    {
+      "epoch": 1.29776867144749,
+      "grad_norm": 2.25,
+      "learning_rate": 6.027460881251208e-06,
+      "loss": 1.0137,
+      "step": 6571
+    },
+    {
+      "epoch": 1.2979685665025862,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.026429546495502e-06,
+      "loss": 1.0137,
+      "step": 6572
+    },
+    {
+      "epoch": 1.2981684615576823,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.025398166145905e-06,
+      "loss": 0.9563,
+      "step": 6573
+    },
+    {
+      "epoch": 1.2983683566127784,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.024366740248231e-06,
+      "loss": 0.9904,
+      "step": 6574
+    },
+    {
+      "epoch": 1.2985682516678745,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.023335268848296e-06,
+      "loss": 0.8159,
+      "step": 6575
+    },
+    {
+      "epoch": 1.2987681467229704,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.022303751991917e-06,
+      "loss": 0.957,
+      "step": 6576
+    },
+    {
+      "epoch": 1.2989680417780665,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.021272189724916e-06,
+      "loss": 0.8763,
+      "step": 6577
+    },
+    {
+      "epoch": 1.2991679368331626,
+      "grad_norm": 2.0,
+      "learning_rate": 6.020240582093112e-06,
+      "loss": 0.8495,
+      "step": 6578
+    },
+    {
+      "epoch": 1.2993678318882587,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.019208929142329e-06,
+      "loss": 0.9718,
+      "step": 6579
+    },
+    {
+      "epoch": 1.2995677269433548,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.018177230918397e-06,
+      "loss": 0.9856,
+      "step": 6580
+    },
+    {
+      "epoch": 1.2997676219984509,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.017145487467139e-06,
+      "loss": 1.1287,
+      "step": 6581
+    },
+    {
+      "epoch": 1.299967517053547,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.016113698834388e-06,
+      "loss": 0.9454,
+      "step": 6582
+    },
+    {
+      "epoch": 1.3001674121086428,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.015081865065977e-06,
+      "loss": 1.0867,
+      "step": 6583
+    },
+    {
+      "epoch": 1.300367307163739,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.014049986207735e-06,
+      "loss": 0.9523,
+      "step": 6584
+    },
+    {
+      "epoch": 1.300567202218835,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.013018062305502e-06,
+      "loss": 1.0047,
+      "step": 6585
+    },
+    {
+      "epoch": 1.3007670972739311,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.011986093405115e-06,
+      "loss": 0.9859,
+      "step": 6586
+    },
+    {
+      "epoch": 1.3009669923290272,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.010954079552414e-06,
+      "loss": 0.9844,
+      "step": 6587
+    },
+    {
+      "epoch": 1.3011668873841233,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.009922020793241e-06,
+      "loss": 0.9852,
+      "step": 6588
+    },
+    {
+      "epoch": 1.3013667824392194,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.008889917173439e-06,
+      "loss": 1.069,
+      "step": 6589
+    },
+    {
+      "epoch": 1.3015666774943155,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.007857768738855e-06,
+      "loss": 1.019,
+      "step": 6590
+    },
+    {
+      "epoch": 1.3017665725494116,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.006825575535337e-06,
+      "loss": 0.9025,
+      "step": 6591
+    },
+    {
+      "epoch": 1.3019664676045077,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.005793337608733e-06,
+      "loss": 0.9806,
+      "step": 6592
+    },
+    {
+      "epoch": 1.3021663626596038,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.004761055004897e-06,
+      "loss": 0.999,
+      "step": 6593
+    },
+    {
+      "epoch": 1.3023662577146997,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.003728727769683e-06,
+      "loss": 0.9815,
+      "step": 6594
+    },
+    {
+      "epoch": 1.3025661527697958,
+      "grad_norm": 2.125,
+      "learning_rate": 6.002696355948945e-06,
+      "loss": 0.9643,
+      "step": 6595
+    },
+    {
+      "epoch": 1.302766047824892,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.0016639395885424e-06,
+      "loss": 0.9539,
+      "step": 6596
+    },
+    {
+      "epoch": 1.302965942879988,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.000631478734336e-06,
+      "loss": 1.0288,
+      "step": 6597
+    },
+    {
+      "epoch": 1.3031658379350841,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.999598973432184e-06,
+      "loss": 1.0162,
+      "step": 6598
+    },
+    {
+      "epoch": 1.3033657329901802,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.9985664237279515e-06,
+      "loss": 0.9597,
+      "step": 6599
+    },
+    {
+      "epoch": 1.303565628045276,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.997533829667508e-06,
+      "loss": 1.013,
+      "step": 6600
+    },
+    {
+      "epoch": 1.3037655231003722,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.996501191296717e-06,
+      "loss": 0.9786,
+      "step": 6601
+    },
+    {
+      "epoch": 1.3039654181554683,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.995468508661451e-06,
+      "loss": 0.9443,
+      "step": 6602
+    },
+    {
+      "epoch": 1.3041653132105644,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.99443578180758e-06,
+      "loss": 0.9783,
+      "step": 6603
+    },
+    {
+      "epoch": 1.3043652082656605,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.993403010780975e-06,
+      "loss": 0.9471,
+      "step": 6604
+    },
+    {
+      "epoch": 1.3045651033207566,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.992370195627516e-06,
+      "loss": 1.0114,
+      "step": 6605
+    },
+    {
+      "epoch": 1.3047649983758527,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.99133733639308e-06,
+      "loss": 1.0314,
+      "step": 6606
+    },
+    {
+      "epoch": 1.3049648934309488,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.990304433123543e-06,
+      "loss": 1.0279,
+      "step": 6607
+    },
+    {
+      "epoch": 1.305164788486045,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.98927148586479e-06,
+      "loss": 0.8717,
+      "step": 6608
+    },
+    {
+      "epoch": 1.305364683541141,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.988238494662704e-06,
+      "loss": 0.9683,
+      "step": 6609
+    },
+    {
+      "epoch": 1.305564578596237,
+      "grad_norm": 2.25,
+      "learning_rate": 5.987205459563168e-06,
+      "loss": 1.0831,
+      "step": 6610
+    },
+    {
+      "epoch": 1.305764473651333,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.986172380612073e-06,
+      "loss": 0.979,
+      "step": 6611
+    },
+    {
+      "epoch": 1.305964368706429,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.985139257855304e-06,
+      "loss": 0.9041,
+      "step": 6612
+    },
+    {
+      "epoch": 1.3061642637615252,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.984106091338756e-06,
+      "loss": 0.9586,
+      "step": 6613
+    },
+    {
+      "epoch": 1.3063641588166213,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.983072881108321e-06,
+      "loss": 0.9586,
+      "step": 6614
+    },
+    {
+      "epoch": 1.3065640538717174,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.982039627209891e-06,
+      "loss": 0.9355,
+      "step": 6615
+    },
+    {
+      "epoch": 1.3067639489268135,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.981006329689368e-06,
+      "loss": 1.0287,
+      "step": 6616
+    },
+    {
+      "epoch": 1.3069638439819096,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.979972988592648e-06,
+      "loss": 1.0191,
+      "step": 6617
+    },
+    {
+      "epoch": 1.3071637390370054,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.978939603965633e-06,
+      "loss": 1.0091,
+      "step": 6618
+    },
+    {
+      "epoch": 1.3073636340921015,
+      "grad_norm": 2.125,
+      "learning_rate": 5.977906175854224e-06,
+      "loss": 0.9668,
+      "step": 6619
+    },
+    {
+      "epoch": 1.3075635291471976,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.976872704304328e-06,
+      "loss": 0.9464,
+      "step": 6620
+    },
+    {
+      "epoch": 1.3077634242022937,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.975839189361851e-06,
+      "loss": 1.0497,
+      "step": 6621
+    },
+    {
+      "epoch": 1.3079633192573898,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.974805631072702e-06,
+      "loss": 1.0606,
+      "step": 6622
+    },
+    {
+      "epoch": 1.308163214312486,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.97377202948279e-06,
+      "loss": 1.0158,
+      "step": 6623
+    },
+    {
+      "epoch": 1.308363109367582,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.9727383846380275e-06,
+      "loss": 0.996,
+      "step": 6624
+    },
+    {
+      "epoch": 1.3085630044226781,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.971704696584332e-06,
+      "loss": 1.0327,
+      "step": 6625
+    },
+    {
+      "epoch": 1.3087628994777742,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.9706709653676156e-06,
+      "loss": 0.9993,
+      "step": 6626
+    },
+    {
+      "epoch": 1.3089627945328703,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.9696371910337995e-06,
+      "loss": 1.0207,
+      "step": 6627
+    },
+    {
+      "epoch": 1.3091626895879664,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.968603373628803e-06,
+      "loss": 0.9862,
+      "step": 6628
+    },
+    {
+      "epoch": 1.3093625846430623,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.967569513198548e-06,
+      "loss": 0.9262,
+      "step": 6629
+    },
+    {
+      "epoch": 1.3095624796981584,
+      "grad_norm": 2.125,
+      "learning_rate": 5.966535609788959e-06,
+      "loss": 1.0114,
+      "step": 6630
+    },
+    {
+      "epoch": 1.3097623747532545,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.965501663445961e-06,
+      "loss": 0.9882,
+      "step": 6631
+    },
+    {
+      "epoch": 1.3099622698083506,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.9644676742154814e-06,
+      "loss": 0.9676,
+      "step": 6632
+    },
+    {
+      "epoch": 1.3101621648634467,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.963433642143452e-06,
+      "loss": 0.954,
+      "step": 6633
+    },
+    {
+      "epoch": 1.3103620599185428,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.962399567275804e-06,
+      "loss": 0.9889,
+      "step": 6634
+    },
+    {
+      "epoch": 1.310561954973639,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.961365449658469e-06,
+      "loss": 0.9578,
+      "step": 6635
+    },
+    {
+      "epoch": 1.3107618500287348,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.960331289337383e-06,
+      "loss": 0.9138,
+      "step": 6636
+    },
+    {
+      "epoch": 1.310961745083831,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.9592970863584835e-06,
+      "loss": 1.0682,
+      "step": 6637
+    },
+    {
+      "epoch": 1.311161640138927,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.95826284076771e-06,
+      "loss": 0.9996,
+      "step": 6638
+    },
+    {
+      "epoch": 1.311361535194023,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.957228552611004e-06,
+      "loss": 1.0233,
+      "step": 6639
+    },
+    {
+      "epoch": 1.3115614302491192,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.956194221934307e-06,
+      "loss": 0.9941,
+      "step": 6640
+    },
+    {
+      "epoch": 1.3117613253042153,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.955159848783565e-06,
+      "loss": 0.9846,
+      "step": 6641
+    },
+    {
+      "epoch": 1.3119612203593114,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.954125433204726e-06,
+      "loss": 0.9759,
+      "step": 6642
+    },
+    {
+      "epoch": 1.3121611154144075,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.953090975243735e-06,
+      "loss": 0.989,
+      "step": 6643
+    },
+    {
+      "epoch": 1.3123610104695036,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.952056474946546e-06,
+      "loss": 0.9784,
+      "step": 6644
+    },
+    {
+      "epoch": 1.3125609055245997,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.9510219323591104e-06,
+      "loss": 0.9401,
+      "step": 6645
+    },
+    {
+      "epoch": 1.3127608005796958,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.94998734752738e-06,
+      "loss": 0.9958,
+      "step": 6646
+    },
+    {
+      "epoch": 1.3129606956347917,
+      "grad_norm": 2.125,
+      "learning_rate": 5.9489527204973165e-06,
+      "loss": 0.9949,
+      "step": 6647
+    },
+    {
+      "epoch": 1.3131605906898878,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.9479180513148716e-06,
+      "loss": 1.0382,
+      "step": 6648
+    },
+    {
+      "epoch": 1.3133604857449839,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.94688334002601e-06,
+      "loss": 1.0021,
+      "step": 6649
+    },
+    {
+      "epoch": 1.31356038080008,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.94584858667669e-06,
+      "loss": 0.8882,
+      "step": 6650
+    },
+    {
+      "epoch": 1.313760275855176,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.944813791312878e-06,
+      "loss": 0.9116,
+      "step": 6651
+    },
+    {
+      "epoch": 1.3139601709102722,
+      "grad_norm": 2.125,
+      "learning_rate": 5.943778953980538e-06,
+      "loss": 1.0571,
+      "step": 6652
+    },
+    {
+      "epoch": 1.314160065965368,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.942744074725638e-06,
+      "loss": 0.9611,
+      "step": 6653
+    },
+    {
+      "epoch": 1.3143599610204642,
+      "grad_norm": 2.125,
+      "learning_rate": 5.941709153594146e-06,
+      "loss": 0.9936,
+      "step": 6654
+    },
+    {
+      "epoch": 1.3145598560755603,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.940674190632034e-06,
+      "loss": 0.967,
+      "step": 6655
+    },
+    {
+      "epoch": 1.3147597511306564,
+      "grad_norm": 2.125,
+      "learning_rate": 5.939639185885276e-06,
+      "loss": 0.9805,
+      "step": 6656
+    },
+    {
+      "epoch": 1.3149596461857525,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.938604139399844e-06,
+      "loss": 0.9485,
+      "step": 6657
+    },
+    {
+      "epoch": 1.3151595412408486,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.937569051221716e-06,
+      "loss": 0.9486,
+      "step": 6658
+    },
+    {
+      "epoch": 1.3153594362959447,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.936533921396874e-06,
+      "loss": 1.0129,
+      "step": 6659
+    },
+    {
+      "epoch": 1.3155593313510408,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.935498749971293e-06,
+      "loss": 1.0116,
+      "step": 6660
+    },
+    {
+      "epoch": 1.3157592264061369,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.934463536990959e-06,
+      "loss": 0.9824,
+      "step": 6661
+    },
+    {
+      "epoch": 1.315959121461233,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.933428282501853e-06,
+      "loss": 0.9549,
+      "step": 6662
+    },
+    {
+      "epoch": 1.316159016516329,
+      "grad_norm": 2.0,
+      "learning_rate": 5.9323929865499615e-06,
+      "loss": 0.9275,
+      "step": 6663
+    },
+    {
+      "epoch": 1.316358911571425,
+      "grad_norm": 2.125,
+      "learning_rate": 5.931357649181275e-06,
+      "loss": 0.9875,
+      "step": 6664
+    },
+    {
+      "epoch": 1.316558806626521,
+      "grad_norm": 2.25,
+      "learning_rate": 5.930322270441781e-06,
+      "loss": 0.9468,
+      "step": 6665
+    },
+    {
+      "epoch": 1.3167587016816171,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.929286850377471e-06,
+      "loss": 0.9012,
+      "step": 6666
+    },
+    {
+      "epoch": 1.3169585967367132,
+      "grad_norm": 2.375,
+      "learning_rate": 5.928251389034338e-06,
+      "loss": 0.958,
+      "step": 6667
+    },
+    {
+      "epoch": 1.3171584917918093,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.927215886458377e-06,
+      "loss": 0.9074,
+      "step": 6668
+    },
+    {
+      "epoch": 1.3173583868469054,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.926180342695587e-06,
+      "loss": 0.9661,
+      "step": 6669
+    },
+    {
+      "epoch": 1.3175582819020015,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.925144757791964e-06,
+      "loss": 0.9343,
+      "step": 6670
+    },
+    {
+      "epoch": 1.3177581769570974,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.92410913179351e-06,
+      "loss": 0.9184,
+      "step": 6671
+    },
+    {
+      "epoch": 1.3179580720121935,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.923073464746225e-06,
+      "loss": 0.9682,
+      "step": 6672
+    },
+    {
+      "epoch": 1.3181579670672896,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.9220377566961175e-06,
+      "loss": 1.0067,
+      "step": 6673
+    },
+    {
+      "epoch": 1.3183578621223857,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.921002007689191e-06,
+      "loss": 0.9897,
+      "step": 6674
+    },
+    {
+      "epoch": 1.3185577571774818,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.919966217771454e-06,
+      "loss": 0.8978,
+      "step": 6675
+    },
+    {
+      "epoch": 1.318757652232578,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.9189303869889145e-06,
+      "loss": 1.0046,
+      "step": 6676
+    },
+    {
+      "epoch": 1.318957547287674,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.9178945153875856e-06,
+      "loss": 0.981,
+      "step": 6677
+    },
+    {
+      "epoch": 1.31915744234277,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.916858603013482e-06,
+      "loss": 0.9779,
+      "step": 6678
+    },
+    {
+      "epoch": 1.3193573373978662,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.915822649912616e-06,
+      "loss": 1.0136,
+      "step": 6679
+    },
+    {
+      "epoch": 1.3195572324529623,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.914786656131007e-06,
+      "loss": 0.9407,
+      "step": 6680
+    },
+    {
+      "epoch": 1.3197571275080584,
+      "grad_norm": 2.125,
+      "learning_rate": 5.913750621714671e-06,
+      "loss": 1.0132,
+      "step": 6681
+    },
+    {
+      "epoch": 1.3199570225631543,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.91271454670963e-06,
+      "loss": 0.975,
+      "step": 6682
+    },
+    {
+      "epoch": 1.3201569176182504,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.911678431161907e-06,
+      "loss": 0.9969,
+      "step": 6683
+    },
+    {
+      "epoch": 1.3203568126733465,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.910642275117525e-06,
+      "loss": 0.9714,
+      "step": 6684
+    },
+    {
+      "epoch": 1.3205567077284426,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.9096060786225095e-06,
+      "loss": 0.9532,
+      "step": 6685
+    },
+    {
+      "epoch": 1.3207566027835387,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.90856984172289e-06,
+      "loss": 0.9656,
+      "step": 6686
+    },
+    {
+      "epoch": 1.3209564978386348,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.907533564464696e-06,
+      "loss": 0.9187,
+      "step": 6687
+    },
+    {
+      "epoch": 1.3211563928937307,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.9064972468939565e-06,
+      "loss": 0.9709,
+      "step": 6688
+    },
+    {
+      "epoch": 1.3213562879488268,
+      "grad_norm": 2.125,
+      "learning_rate": 5.905460889056707e-06,
+      "loss": 0.9907,
+      "step": 6689
+    },
+    {
+      "epoch": 1.3215561830039229,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.904424490998981e-06,
+      "loss": 0.9996,
+      "step": 6690
+    },
+    {
+      "epoch": 1.321756078059019,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.903388052766817e-06,
+      "loss": 0.9861,
+      "step": 6691
+    },
+    {
+      "epoch": 1.321955973114115,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.902351574406251e-06,
+      "loss": 1.0058,
+      "step": 6692
+    },
+    {
+      "epoch": 1.3221558681692112,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.9013150559633245e-06,
+      "loss": 1.1108,
+      "step": 6693
+    },
+    {
+      "epoch": 1.3223557632243073,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.900278497484079e-06,
+      "loss": 0.9455,
+      "step": 6694
+    },
+    {
+      "epoch": 1.3225556582794034,
+      "grad_norm": 2.125,
+      "learning_rate": 5.89924189901456e-06,
+      "loss": 0.9911,
+      "step": 6695
+    },
+    {
+      "epoch": 1.3227555533344995,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.898205260600812e-06,
+      "loss": 0.8872,
+      "step": 6696
+    },
+    {
+      "epoch": 1.3229554483895956,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.897168582288881e-06,
+      "loss": 0.8304,
+      "step": 6697
+    },
+    {
+      "epoch": 1.3231553434446917,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.8961318641248185e-06,
+      "loss": 0.9498,
+      "step": 6698
+    },
+    {
+      "epoch": 1.3233552384997875,
+      "grad_norm": 2.25,
+      "learning_rate": 5.895095106154674e-06,
+      "loss": 0.9766,
+      "step": 6699
+    },
+    {
+      "epoch": 1.3235551335548836,
+      "grad_norm": 2.125,
+      "learning_rate": 5.8940583084245e-06,
+      "loss": 0.9235,
+      "step": 6700
+    },
+    {
+      "epoch": 1.3237550286099797,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.893021470980352e-06,
+      "loss": 0.9945,
+      "step": 6701
+    },
+    {
+      "epoch": 1.3239549236650758,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.891984593868285e-06,
+      "loss": 1.0195,
+      "step": 6702
+    },
+    {
+      "epoch": 1.324154818720172,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.890947677134358e-06,
+      "loss": 1.0657,
+      "step": 6703
+    },
+    {
+      "epoch": 1.324354713775268,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.889910720824631e-06,
+      "loss": 1.0126,
+      "step": 6704
+    },
+    {
+      "epoch": 1.3245546088303641,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.888873724985163e-06,
+      "loss": 0.9802,
+      "step": 6705
+    },
+    {
+      "epoch": 1.32475450388546,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.887836689662021e-06,
+      "loss": 1.0151,
+      "step": 6706
+    },
+    {
+      "epoch": 1.324954398940556,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.886799614901267e-06,
+      "loss": 1.0367,
+      "step": 6707
+    },
+    {
+      "epoch": 1.3251542939956522,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.885762500748968e-06,
+      "loss": 0.99,
+      "step": 6708
+    },
+    {
+      "epoch": 1.3253541890507483,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.8847253472511956e-06,
+      "loss": 1.0279,
+      "step": 6709
+    },
+    {
+      "epoch": 1.3255540841058444,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.883688154454016e-06,
+      "loss": 0.9555,
+      "step": 6710
+    },
+    {
+      "epoch": 1.3257539791609405,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.8826509224035046e-06,
+      "loss": 0.9638,
+      "step": 6711
+    },
+    {
+      "epoch": 1.3259538742160366,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.881613651145732e-06,
+      "loss": 0.9608,
+      "step": 6712
+    },
+    {
+      "epoch": 1.3261537692711327,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.880576340726776e-06,
+      "loss": 1.0941,
+      "step": 6713
+    },
+    {
+      "epoch": 1.3263536643262288,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.8795389911927126e-06,
+      "loss": 1.0361,
+      "step": 6714
+    },
+    {
+      "epoch": 1.326553559381325,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.8785016025896216e-06,
+      "loss": 0.9403,
+      "step": 6715
+    },
+    {
+      "epoch": 1.326753454436421,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.877464174963582e-06,
+      "loss": 0.9459,
+      "step": 6716
+    },
+    {
+      "epoch": 1.3269533494915169,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.876426708360678e-06,
+      "loss": 1.0179,
+      "step": 6717
+    },
+    {
+      "epoch": 1.327153244546613,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.875389202826994e-06,
+      "loss": 1.0085,
+      "step": 6718
+    },
+    {
+      "epoch": 1.327353139601709,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.874351658408614e-06,
+      "loss": 0.9004,
+      "step": 6719
+    },
+    {
+      "epoch": 1.3275530346568052,
+      "grad_norm": 2.125,
+      "learning_rate": 5.873314075151628e-06,
+      "loss": 1.0149,
+      "step": 6720
+    },
+    {
+      "epoch": 1.3277529297119013,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.872276453102122e-06,
+      "loss": 1.0591,
+      "step": 6721
+    },
+    {
+      "epoch": 1.3279528247669974,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.8712387923061905e-06,
+      "loss": 1.0282,
+      "step": 6722
+    },
+    {
+      "epoch": 1.3281527198220933,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.870201092809926e-06,
+      "loss": 0.917,
+      "step": 6723
+    },
+    {
+      "epoch": 1.3283526148771894,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.8691633546594206e-06,
+      "loss": 0.9753,
+      "step": 6724
+    },
+    {
+      "epoch": 1.3285525099322855,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.868125577900772e-06,
+      "loss": 0.9051,
+      "step": 6725
+    },
+    {
+      "epoch": 1.3287524049873816,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.867087762580079e-06,
+      "loss": 0.913,
+      "step": 6726
+    },
+    {
+      "epoch": 1.3289523000424777,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.866049908743439e-06,
+      "loss": 1.0433,
+      "step": 6727
+    },
+    {
+      "epoch": 1.3291521950975738,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.865012016436954e-06,
+      "loss": 1.0055,
+      "step": 6728
+    },
+    {
+      "epoch": 1.3293520901526699,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.863974085706729e-06,
+      "loss": 0.9556,
+      "step": 6729
+    },
+    {
+      "epoch": 1.329551985207766,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.862936116598866e-06,
+      "loss": 0.9934,
+      "step": 6730
+    },
+    {
+      "epoch": 1.329751880262862,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.861898109159473e-06,
+      "loss": 0.8601,
+      "step": 6731
+    },
+    {
+      "epoch": 1.3299517753179582,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.860860063434659e-06,
+      "loss": 1.0253,
+      "step": 6732
+    },
+    {
+      "epoch": 1.3301516703730543,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.859821979470532e-06,
+      "loss": 0.8942,
+      "step": 6733
+    },
+    {
+      "epoch": 1.3303515654281501,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.858783857313205e-06,
+      "loss": 1.0331,
+      "step": 6734
+    },
+    {
+      "epoch": 1.3305514604832462,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.857745697008792e-06,
+      "loss": 0.9352,
+      "step": 6735
+    },
+    {
+      "epoch": 1.3307513555383423,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.856707498603404e-06,
+      "loss": 1.0373,
+      "step": 6736
+    },
+    {
+      "epoch": 1.3309512505934384,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.855669262143163e-06,
+      "loss": 0.9652,
+      "step": 6737
+    },
+    {
+      "epoch": 1.3311511456485345,
+      "grad_norm": 2.125,
+      "learning_rate": 5.854630987674184e-06,
+      "loss": 0.8334,
+      "step": 6738
+    },
+    {
+      "epoch": 1.3313510407036306,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.853592675242587e-06,
+      "loss": 0.8961,
+      "step": 6739
+    },
+    {
+      "epoch": 1.3315509357587267,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.852554324894495e-06,
+      "loss": 0.9398,
+      "step": 6740
+    },
+    {
+      "epoch": 1.3317508308138226,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.851515936676031e-06,
+      "loss": 1.0227,
+      "step": 6741
+    },
+    {
+      "epoch": 1.3319507258689187,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.85047751063332e-06,
+      "loss": 0.9559,
+      "step": 6742
+    },
+    {
+      "epoch": 1.3321506209240148,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.84943904681249e-06,
+      "loss": 1.0473,
+      "step": 6743
+    },
+    {
+      "epoch": 1.332350515979111,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.848400545259667e-06,
+      "loss": 0.9805,
+      "step": 6744
+    },
+    {
+      "epoch": 1.332550411034207,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.847362006020982e-06,
+      "loss": 1.0385,
+      "step": 6745
+    },
+    {
+      "epoch": 1.3327503060893031,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.846323429142569e-06,
+      "loss": 0.879,
+      "step": 6746
+    },
+    {
+      "epoch": 1.3329502011443992,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.845284814670556e-06,
+      "loss": 0.9384,
+      "step": 6747
+    },
+    {
+      "epoch": 1.3331500961994953,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.844246162651085e-06,
+      "loss": 1.0013,
+      "step": 6748
+    },
+    {
+      "epoch": 1.3333499912545914,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.843207473130289e-06,
+      "loss": 1.0776,
+      "step": 6749
+    },
+    {
+      "epoch": 1.3335498863096875,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.842168746154306e-06,
+      "loss": 1.0119,
+      "step": 6750
+    },
+    {
+      "epoch": 1.3337497813647836,
+      "grad_norm": 2.125,
+      "learning_rate": 5.841129981769279e-06,
+      "loss": 0.9853,
+      "step": 6751
+    },
+    {
+      "epoch": 1.3339496764198795,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.840091180021348e-06,
+      "loss": 0.8704,
+      "step": 6752
+    },
+    {
+      "epoch": 1.3341495714749756,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.839052340956654e-06,
+      "loss": 0.9847,
+      "step": 6753
+    },
+    {
+      "epoch": 1.3343494665300717,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.838013464621347e-06,
+      "loss": 0.998,
+      "step": 6754
+    },
+    {
+      "epoch": 1.3345493615851678,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.836974551061571e-06,
+      "loss": 1.0238,
+      "step": 6755
+    },
+    {
+      "epoch": 1.334749256640264,
+      "grad_norm": 2.125,
+      "learning_rate": 5.835935600323473e-06,
+      "loss": 1.0121,
+      "step": 6756
+    },
+    {
+      "epoch": 1.33494915169536,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.834896612453209e-06,
+      "loss": 0.9981,
+      "step": 6757
+    },
+    {
+      "epoch": 1.335149046750456,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.8338575874969235e-06,
+      "loss": 0.937,
+      "step": 6758
+    },
+    {
+      "epoch": 1.335348941805552,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.832818525500773e-06,
+      "loss": 0.9186,
+      "step": 6759
+    },
+    {
+      "epoch": 1.335548836860648,
+      "grad_norm": 2.25,
+      "learning_rate": 5.831779426510914e-06,
+      "loss": 1.0156,
+      "step": 6760
+    },
+    {
+      "epoch": 1.3357487319157442,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.830740290573501e-06,
+      "loss": 0.9893,
+      "step": 6761
+    },
+    {
+      "epoch": 1.3359486269708403,
+      "grad_norm": 2.25,
+      "learning_rate": 5.829701117734692e-06,
+      "loss": 1.0315,
+      "step": 6762
+    },
+    {
+      "epoch": 1.3361485220259364,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.828661908040649e-06,
+      "loss": 0.9556,
+      "step": 6763
+    },
+    {
+      "epoch": 1.3363484170810325,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.827622661537533e-06,
+      "loss": 0.928,
+      "step": 6764
+    },
+    {
+      "epoch": 1.3365483121361286,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.826583378271506e-06,
+      "loss": 1.0471,
+      "step": 6765
+    },
+    {
+      "epoch": 1.3367482071912247,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.825544058288735e-06,
+      "loss": 1.0082,
+      "step": 6766
+    },
+    {
+      "epoch": 1.3369481022463208,
+      "grad_norm": 2.125,
+      "learning_rate": 5.824504701635383e-06,
+      "loss": 1.0759,
+      "step": 6767
+    },
+    {
+      "epoch": 1.3371479973014169,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.823465308357621e-06,
+      "loss": 0.99,
+      "step": 6768
+    },
+    {
+      "epoch": 1.337347892356513,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.822425878501618e-06,
+      "loss": 1.0142,
+      "step": 6769
+    },
+    {
+      "epoch": 1.3375477874116088,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.821386412113546e-06,
+      "loss": 0.9711,
+      "step": 6770
+    },
+    {
+      "epoch": 1.337747682466705,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.820346909239576e-06,
+      "loss": 1.06,
+      "step": 6771
+    },
+    {
+      "epoch": 1.337947577521801,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.819307369925884e-06,
+      "loss": 0.9441,
+      "step": 6772
+    },
+    {
+      "epoch": 1.3381474725768971,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.8182677942186485e-06,
+      "loss": 0.9636,
+      "step": 6773
+    },
+    {
+      "epoch": 1.3383473676319932,
+      "grad_norm": 2.125,
+      "learning_rate": 5.817228182164043e-06,
+      "loss": 0.8985,
+      "step": 6774
+    },
+    {
+      "epoch": 1.3385472626870893,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.816188533808249e-06,
+      "loss": 1.0072,
+      "step": 6775
+    },
+    {
+      "epoch": 1.3387471577421852,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.815148849197447e-06,
+      "loss": 0.993,
+      "step": 6776
+    },
+    {
+      "epoch": 1.3389470527972813,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.814109128377822e-06,
+      "loss": 0.9225,
+      "step": 6777
+    },
+    {
+      "epoch": 1.3391469478523774,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.8130693713955545e-06,
+      "loss": 0.8882,
+      "step": 6778
+    },
+    {
+      "epoch": 1.3393468429074735,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.8120295782968325e-06,
+      "loss": 0.8944,
+      "step": 6779
+    },
+    {
+      "epoch": 1.3395467379625696,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.810989749127844e-06,
+      "loss": 0.921,
+      "step": 6780
+    },
+    {
+      "epoch": 1.3397466330176657,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.809949883934778e-06,
+      "loss": 0.9186,
+      "step": 6781
+    },
+    {
+      "epoch": 1.3399465280727618,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.808909982763825e-06,
+      "loss": 0.9882,
+      "step": 6782
+    },
+    {
+      "epoch": 1.340146423127858,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.807870045661176e-06,
+      "loss": 0.9886,
+      "step": 6783
+    },
+    {
+      "epoch": 1.340346318182954,
+      "grad_norm": 2.125,
+      "learning_rate": 5.806830072673025e-06,
+      "loss": 0.9413,
+      "step": 6784
+    },
+    {
+      "epoch": 1.3405462132380501,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.80579006384557e-06,
+      "loss": 0.9557,
+      "step": 6785
+    },
+    {
+      "epoch": 1.3407461082931462,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.804750019225005e-06,
+      "loss": 1.0898,
+      "step": 6786
+    },
+    {
+      "epoch": 1.340946003348242,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.8037099388575305e-06,
+      "loss": 0.9668,
+      "step": 6787
+    },
+    {
+      "epoch": 1.3411458984033382,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.802669822789349e-06,
+      "loss": 0.9791,
+      "step": 6788
+    },
+    {
+      "epoch": 1.3413457934584343,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.801629671066657e-06,
+      "loss": 1.0095,
+      "step": 6789
+    },
+    {
+      "epoch": 1.3415456885135304,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.80058948373566e-06,
+      "loss": 1.0419,
+      "step": 6790
+    },
+    {
+      "epoch": 1.3417455835686265,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.799549260842565e-06,
+      "loss": 1.065,
+      "step": 6791
+    },
+    {
+      "epoch": 1.3419454786237226,
+      "grad_norm": 2.125,
+      "learning_rate": 5.7985090024335766e-06,
+      "loss": 0.9716,
+      "step": 6792
+    },
+    {
+      "epoch": 1.3421453736788187,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.797468708554903e-06,
+      "loss": 0.963,
+      "step": 6793
+    },
+    {
+      "epoch": 1.3423452687339146,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.796428379252756e-06,
+      "loss": 0.9844,
+      "step": 6794
+    },
+    {
+      "epoch": 1.3425451637890107,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.795388014573345e-06,
+      "loss": 0.9513,
+      "step": 6795
+    },
+    {
+      "epoch": 1.3427450588441068,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.794347614562884e-06,
+      "loss": 1.0168,
+      "step": 6796
+    },
+    {
+      "epoch": 1.3429449538992029,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.793307179267585e-06,
+      "loss": 1.0011,
+      "step": 6797
+    },
+    {
+      "epoch": 1.343144848954299,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.792266708733667e-06,
+      "loss": 0.9518,
+      "step": 6798
+    },
+    {
+      "epoch": 1.343344744009395,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.791226203007346e-06,
+      "loss": 1.0294,
+      "step": 6799
+    },
+    {
+      "epoch": 1.3435446390644912,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.790185662134842e-06,
+      "loss": 0.9154,
+      "step": 6800
+    },
+    {
+      "epoch": 1.3437445341195873,
+      "grad_norm": 2.125,
+      "learning_rate": 5.789145086162374e-06,
+      "loss": 0.9629,
+      "step": 6801
+    },
+    {
+      "epoch": 1.3439444291746834,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.788104475136168e-06,
+      "loss": 0.9986,
+      "step": 6802
+    },
+    {
+      "epoch": 1.3441443242297795,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.7870638291024436e-06,
+      "loss": 1.0083,
+      "step": 6803
+    },
+    {
+      "epoch": 1.3443442192848756,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.786023148107429e-06,
+      "loss": 1.0264,
+      "step": 6804
+    },
+    {
+      "epoch": 1.3445441143399715,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.7849824321973505e-06,
+      "loss": 0.8659,
+      "step": 6805
+    },
+    {
+      "epoch": 1.3447440093950676,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.783941681418435e-06,
+      "loss": 0.9921,
+      "step": 6806
+    },
+    {
+      "epoch": 1.3449439044501637,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.782900895816913e-06,
+      "loss": 0.9895,
+      "step": 6807
+    },
+    {
+      "epoch": 1.3451437995052598,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.781860075439019e-06,
+      "loss": 1.0123,
+      "step": 6808
+    },
+    {
+      "epoch": 1.3453436945603559,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.780819220330982e-06,
+      "loss": 0.9928,
+      "step": 6809
+    },
+    {
+      "epoch": 1.345543589615452,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.77977833053904e-06,
+      "loss": 0.9065,
+      "step": 6810
+    },
+    {
+      "epoch": 1.3457434846705478,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.778737406109427e-06,
+      "loss": 0.9993,
+      "step": 6811
+    },
+    {
+      "epoch": 1.345943379725644,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.777696447088381e-06,
+      "loss": 0.9916,
+      "step": 6812
+    },
+    {
+      "epoch": 1.34614327478074,
+      "grad_norm": 2.125,
+      "learning_rate": 5.7766554535221435e-06,
+      "loss": 0.9238,
+      "step": 6813
+    },
+    {
+      "epoch": 1.3463431698358361,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.775614425456953e-06,
+      "loss": 0.9169,
+      "step": 6814
+    },
+    {
+      "epoch": 1.3465430648909322,
+      "grad_norm": 2.125,
+      "learning_rate": 5.774573362939051e-06,
+      "loss": 0.8883,
+      "step": 6815
+    },
+    {
+      "epoch": 1.3467429599460283,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.773532266014686e-06,
+      "loss": 0.8809,
+      "step": 6816
+    },
+    {
+      "epoch": 1.3469428550011244,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.772491134730097e-06,
+      "loss": 0.9436,
+      "step": 6817
+    },
+    {
+      "epoch": 1.3471427500562205,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.771449969131536e-06,
+      "loss": 0.8461,
+      "step": 6818
+    },
+    {
+      "epoch": 1.3473426451113166,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.770408769265252e-06,
+      "loss": 1.0222,
+      "step": 6819
+    },
+    {
+      "epoch": 1.3475425401664127,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.7693675351774905e-06,
+      "loss": 0.9295,
+      "step": 6820
+    },
+    {
+      "epoch": 1.3477424352215088,
+      "grad_norm": 2.125,
+      "learning_rate": 5.768326266914505e-06,
+      "loss": 0.9204,
+      "step": 6821
+    },
+    {
+      "epoch": 1.3479423302766047,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.767284964522549e-06,
+      "loss": 1.0075,
+      "step": 6822
+    },
+    {
+      "epoch": 1.3481422253317008,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.766243628047876e-06,
+      "loss": 0.8984,
+      "step": 6823
+    },
+    {
+      "epoch": 1.348342120386797,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.765202257536744e-06,
+      "loss": 0.8898,
+      "step": 6824
+    },
+    {
+      "epoch": 1.348542015441893,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.76416085303541e-06,
+      "loss": 1.1051,
+      "step": 6825
+    },
+    {
+      "epoch": 1.348741910496989,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.7631194145901315e-06,
+      "loss": 0.9619,
+      "step": 6826
+    },
+    {
+      "epoch": 1.3489418055520852,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.76207794224717e-06,
+      "loss": 0.9741,
+      "step": 6827
+    },
+    {
+      "epoch": 1.3491417006071813,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.761036436052788e-06,
+      "loss": 1.0536,
+      "step": 6828
+    },
+    {
+      "epoch": 1.3493415956622772,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.75999489605325e-06,
+      "loss": 0.9193,
+      "step": 6829
+    },
+    {
+      "epoch": 1.3495414907173733,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.7589533222948176e-06,
+      "loss": 0.9696,
+      "step": 6830
+    },
+    {
+      "epoch": 1.3497413857724694,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.757911714823761e-06,
+      "loss": 0.9368,
+      "step": 6831
+    },
+    {
+      "epoch": 1.3499412808275655,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.756870073686347e-06,
+      "loss": 1.0816,
+      "step": 6832
+    },
+    {
+      "epoch": 1.3501411758826616,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.755828398928845e-06,
+      "loss": 1.013,
+      "step": 6833
+    },
+    {
+      "epoch": 1.3503410709377577,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.754786690597527e-06,
+      "loss": 0.931,
+      "step": 6834
+    },
+    {
+      "epoch": 1.3505409659928538,
+      "grad_norm": 2.125,
+      "learning_rate": 5.753744948738663e-06,
+      "loss": 1.0375,
+      "step": 6835
+    },
+    {
+      "epoch": 1.3507408610479499,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.75270317339853e-06,
+      "loss": 0.9461,
+      "step": 6836
+    },
+    {
+      "epoch": 1.350940756103046,
+      "grad_norm": 2.125,
+      "learning_rate": 5.751661364623403e-06,
+      "loss": 0.9873,
+      "step": 6837
+    },
+    {
+      "epoch": 1.351140651158142,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.750619522459558e-06,
+      "loss": 1.1163,
+      "step": 6838
+    },
+    {
+      "epoch": 1.3513405462132382,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.749577646953274e-06,
+      "loss": 1.0067,
+      "step": 6839
+    },
+    {
+      "epoch": 1.351540441268334,
+      "grad_norm": 1.9921875,
+      "learning_rate": 5.74853573815083e-06,
+      "loss": 0.8746,
+      "step": 6840
+    },
+    {
+      "epoch": 1.3517403363234302,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.74749379609851e-06,
+      "loss": 0.9809,
+      "step": 6841
+    },
+    {
+      "epoch": 1.3519402313785263,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.746451820842595e-06,
+      "loss": 0.9222,
+      "step": 6842
+    },
+    {
+      "epoch": 1.3521401264336224,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.7454098124293675e-06,
+      "loss": 0.9275,
+      "step": 6843
+    },
+    {
+      "epoch": 1.3523400214887185,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.744367770905119e-06,
+      "loss": 1.0583,
+      "step": 6844
+    },
+    {
+      "epoch": 1.3525399165438146,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.743325696316132e-06,
+      "loss": 0.9374,
+      "step": 6845
+    },
+    {
+      "epoch": 1.3527398115989104,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.742283588708697e-06,
+      "loss": 1.0193,
+      "step": 6846
+    },
+    {
+      "epoch": 1.3529397066540065,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.741241448129105e-06,
+      "loss": 1.0164,
+      "step": 6847
+    },
+    {
+      "epoch": 1.3531396017091026,
+      "grad_norm": 2.125,
+      "learning_rate": 5.740199274623647e-06,
+      "loss": 0.954,
+      "step": 6848
+    },
+    {
+      "epoch": 1.3533394967641987,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.7391570682386165e-06,
+      "loss": 0.9195,
+      "step": 6849
+    },
+    {
+      "epoch": 1.3535393918192948,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.738114829020307e-06,
+      "loss": 1.0024,
+      "step": 6850
+    },
+    {
+      "epoch": 1.353739286874391,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.737072557015016e-06,
+      "loss": 1.0244,
+      "step": 6851
+    },
+    {
+      "epoch": 1.353939181929487,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.73603025226904e-06,
+      "loss": 0.9202,
+      "step": 6852
+    },
+    {
+      "epoch": 1.3541390769845831,
+      "grad_norm": 2.125,
+      "learning_rate": 5.7349879148286804e-06,
+      "loss": 1.0214,
+      "step": 6853
+    },
+    {
+      "epoch": 1.3543389720396792,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.733945544740235e-06,
+      "loss": 0.8318,
+      "step": 6854
+    },
+    {
+      "epoch": 1.3545388670947753,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.732903142050008e-06,
+      "loss": 0.9818,
+      "step": 6855
+    },
+    {
+      "epoch": 1.3547387621498714,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.731860706804301e-06,
+      "loss": 0.9657,
+      "step": 6856
+    },
+    {
+      "epoch": 1.3549386572049673,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.7308182390494185e-06,
+      "loss": 1.0145,
+      "step": 6857
+    },
+    {
+      "epoch": 1.3551385522600634,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.729775738831669e-06,
+      "loss": 0.9326,
+      "step": 6858
+    },
+    {
+      "epoch": 1.3553384473151595,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.728733206197359e-06,
+      "loss": 1.0115,
+      "step": 6859
+    },
+    {
+      "epoch": 1.3555383423702556,
+      "grad_norm": 1.96875,
+      "learning_rate": 5.727690641192797e-06,
+      "loss": 0.9357,
+      "step": 6860
+    },
+    {
+      "epoch": 1.3557382374253517,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.726648043864296e-06,
+      "loss": 1.0141,
+      "step": 6861
+    },
+    {
+      "epoch": 1.3559381324804478,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.725605414258165e-06,
+      "loss": 0.988,
+      "step": 6862
+    },
+    {
+      "epoch": 1.356138027535544,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.724562752420719e-06,
+      "loss": 0.9958,
+      "step": 6863
+    },
+    {
+      "epoch": 1.3563379225906398,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.723520058398275e-06,
+      "loss": 1.0179,
+      "step": 6864
+    },
+    {
+      "epoch": 1.3565378176457359,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.7224773322371466e-06,
+      "loss": 0.935,
+      "step": 6865
+    },
+    {
+      "epoch": 1.356737712700832,
+      "grad_norm": 1.9453125,
+      "learning_rate": 5.721434573983651e-06,
+      "loss": 0.8396,
+      "step": 6866
+    },
+    {
+      "epoch": 1.356937607755928,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.720391783684109e-06,
+      "loss": 0.9771,
+      "step": 6867
+    },
+    {
+      "epoch": 1.3571375028110242,
+      "grad_norm": 2.0,
+      "learning_rate": 5.7193489613848414e-06,
+      "loss": 0.8618,
+      "step": 6868
+    },
+    {
+      "epoch": 1.3573373978661203,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.718306107132169e-06,
+      "loss": 0.9778,
+      "step": 6869
+    },
+    {
+      "epoch": 1.3575372929212164,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.717263220972418e-06,
+      "loss": 0.9241,
+      "step": 6870
+    },
+    {
+      "epoch": 1.3577371879763125,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.716220302951909e-06,
+      "loss": 1.0834,
+      "step": 6871
+    },
+    {
+      "epoch": 1.3579370830314086,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.715177353116972e-06,
+      "loss": 0.9576,
+      "step": 6872
+    },
+    {
+      "epoch": 1.3581369780865047,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.714134371513932e-06,
+      "loss": 0.9635,
+      "step": 6873
+    },
+    {
+      "epoch": 1.3583368731416008,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.71309135818912e-06,
+      "loss": 0.9623,
+      "step": 6874
+    },
+    {
+      "epoch": 1.3585367681966967,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.712048313188867e-06,
+      "loss": 0.9739,
+      "step": 6875
+    },
+    {
+      "epoch": 1.3587366632517928,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.711005236559503e-06,
+      "loss": 1.0418,
+      "step": 6876
+    },
+    {
+      "epoch": 1.3589365583068889,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.7099621283473625e-06,
+      "loss": 0.9763,
+      "step": 6877
+    },
+    {
+      "epoch": 1.359136453361985,
+      "grad_norm": 2.125,
+      "learning_rate": 5.708918988598781e-06,
+      "loss": 1.0662,
+      "step": 6878
+    },
+    {
+      "epoch": 1.359336348417081,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.707875817360092e-06,
+      "loss": 0.848,
+      "step": 6879
+    },
+    {
+      "epoch": 1.3595362434721772,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.706832614677637e-06,
+      "loss": 1.0198,
+      "step": 6880
+    },
+    {
+      "epoch": 1.3597361385272733,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.7057893805977525e-06,
+      "loss": 1.109,
+      "step": 6881
+    },
+    {
+      "epoch": 1.3599360335823691,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.704746115166777e-06,
+      "loss": 0.9673,
+      "step": 6882
+    },
+    {
+      "epoch": 1.3601359286374652,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.703702818431055e-06,
+      "loss": 1.0013,
+      "step": 6883
+    },
+    {
+      "epoch": 1.3603358236925613,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.70265949043693e-06,
+      "loss": 1.0541,
+      "step": 6884
+    },
+    {
+      "epoch": 1.3605357187476574,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.701616131230745e-06,
+      "loss": 0.9834,
+      "step": 6885
+    },
+    {
+      "epoch": 1.3607356138027535,
+      "grad_norm": 2.125,
+      "learning_rate": 5.700572740858847e-06,
+      "loss": 0.9302,
+      "step": 6886
+    },
+    {
+      "epoch": 1.3609355088578496,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.699529319367581e-06,
+      "loss": 1.0564,
+      "step": 6887
+    },
+    {
+      "epoch": 1.3611354039129457,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.698485866803298e-06,
+      "loss": 1.0996,
+      "step": 6888
+    },
+    {
+      "epoch": 1.3613352989680418,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.6974423832123494e-06,
+      "loss": 0.9551,
+      "step": 6889
+    },
+    {
+      "epoch": 1.361535194023138,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.696398868641082e-06,
+      "loss": 0.9805,
+      "step": 6890
+    },
+    {
+      "epoch": 1.361735089078234,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.695355323135852e-06,
+      "loss": 1.0134,
+      "step": 6891
+    },
+    {
+      "epoch": 1.3619349841333301,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.694311746743013e-06,
+      "loss": 0.9429,
+      "step": 6892
+    },
+    {
+      "epoch": 1.362134879188426,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.693268139508921e-06,
+      "loss": 0.8925,
+      "step": 6893
+    },
+    {
+      "epoch": 1.3623347742435221,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.6922245014799316e-06,
+      "loss": 1.0224,
+      "step": 6894
+    },
+    {
+      "epoch": 1.3625346692986182,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.6911808327024035e-06,
+      "loss": 1.0771,
+      "step": 6895
+    },
+    {
+      "epoch": 1.3627345643537143,
+      "grad_norm": 2.125,
+      "learning_rate": 5.6901371332226975e-06,
+      "loss": 0.9291,
+      "step": 6896
+    },
+    {
+      "epoch": 1.3629344594088104,
+      "grad_norm": 2.125,
+      "learning_rate": 5.689093403087173e-06,
+      "loss": 1.0084,
+      "step": 6897
+    },
+    {
+      "epoch": 1.3631343544639065,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.6880496423421936e-06,
+      "loss": 1.0037,
+      "step": 6898
+    },
+    {
+      "epoch": 1.3633342495190024,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.687005851034122e-06,
+      "loss": 1.0394,
+      "step": 6899
+    },
+    {
+      "epoch": 1.3635341445740985,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.685962029209325e-06,
+      "loss": 1.0239,
+      "step": 6900
+    },
+    {
+      "epoch": 1.3637340396291946,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.684918176914167e-06,
+      "loss": 1.044,
+      "step": 6901
+    },
+    {
+      "epoch": 1.3639339346842907,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.683874294195017e-06,
+      "loss": 0.9391,
+      "step": 6902
+    },
+    {
+      "epoch": 1.3641338297393868,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.682830381098243e-06,
+      "loss": 0.9159,
+      "step": 6903
+    },
+    {
+      "epoch": 1.364333724794483,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.681786437670217e-06,
+      "loss": 0.9494,
+      "step": 6904
+    },
+    {
+      "epoch": 1.364533619849579,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.680742463957311e-06,
+      "loss": 0.9288,
+      "step": 6905
+    },
+    {
+      "epoch": 1.364733514904675,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.679698460005897e-06,
+      "loss": 0.982,
+      "step": 6906
+    },
+    {
+      "epoch": 1.3649334099597712,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.678654425862349e-06,
+      "loss": 0.9884,
+      "step": 6907
+    },
+    {
+      "epoch": 1.3651333050148673,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.677610361573045e-06,
+      "loss": 0.978,
+      "step": 6908
+    },
+    {
+      "epoch": 1.3653332000699634,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.6765662671843615e-06,
+      "loss": 1.0478,
+      "step": 6909
+    },
+    {
+      "epoch": 1.3655330951250593,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.675522142742675e-06,
+      "loss": 0.971,
+      "step": 6910
+    },
+    {
+      "epoch": 1.3657329901801554,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.674477988294369e-06,
+      "loss": 1.0029,
+      "step": 6911
+    },
+    {
+      "epoch": 1.3659328852352515,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.6734338038858225e-06,
+      "loss": 0.9451,
+      "step": 6912
+    },
+    {
+      "epoch": 1.3661327802903476,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.672389589563417e-06,
+      "loss": 0.9869,
+      "step": 6913
+    },
+    {
+      "epoch": 1.3663326753454437,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.671345345373538e-06,
+      "loss": 1.0203,
+      "step": 6914
+    },
+    {
+      "epoch": 1.3665325704005398,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.6703010713625715e-06,
+      "loss": 1.0001,
+      "step": 6915
+    },
+    {
+      "epoch": 1.3667324654556359,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.669256767576903e-06,
+      "loss": 0.9559,
+      "step": 6916
+    },
+    {
+      "epoch": 1.3669323605107317,
+      "grad_norm": 2.25,
+      "learning_rate": 5.6682124340629195e-06,
+      "loss": 0.9698,
+      "step": 6917
+    },
+    {
+      "epoch": 1.3671322555658278,
+      "grad_norm": 2.25,
+      "learning_rate": 5.667168070867012e-06,
+      "loss": 1.0643,
+      "step": 6918
+    },
+    {
+      "epoch": 1.367332150620924,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.666123678035569e-06,
+      "loss": 1.0569,
+      "step": 6919
+    },
+    {
+      "epoch": 1.36753204567602,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.665079255614984e-06,
+      "loss": 0.9833,
+      "step": 6920
+    },
+    {
+      "epoch": 1.3677319407311161,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.664034803651649e-06,
+      "loss": 0.9298,
+      "step": 6921
+    },
+    {
+      "epoch": 1.3679318357862122,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.662990322191959e-06,
+      "loss": 1.0319,
+      "step": 6922
+    },
+    {
+      "epoch": 1.3681317308413083,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.66194581128231e-06,
+      "loss": 1.0111,
+      "step": 6923
+    },
+    {
+      "epoch": 1.3683316258964044,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.660901270969098e-06,
+      "loss": 1.0736,
+      "step": 6924
+    },
+    {
+      "epoch": 1.3685315209515005,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.659856701298724e-06,
+      "loss": 1.0559,
+      "step": 6925
+    },
+    {
+      "epoch": 1.3687314160065966,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.658812102317583e-06,
+      "loss": 0.9378,
+      "step": 6926
+    },
+    {
+      "epoch": 1.3689313110616927,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.657767474072082e-06,
+      "loss": 0.8943,
+      "step": 6927
+    },
+    {
+      "epoch": 1.3691312061167886,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.656722816608619e-06,
+      "loss": 1.0167,
+      "step": 6928
+    },
+    {
+      "epoch": 1.3693311011718847,
+      "grad_norm": 2.125,
+      "learning_rate": 5.655678129973597e-06,
+      "loss": 1.0068,
+      "step": 6929
+    },
+    {
+      "epoch": 1.3695309962269808,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.654633414213424e-06,
+      "loss": 0.9642,
+      "step": 6930
+    },
+    {
+      "epoch": 1.369730891282077,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.6535886693745044e-06,
+      "loss": 0.8641,
+      "step": 6931
+    },
+    {
+      "epoch": 1.369930786337173,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.652543895503246e-06,
+      "loss": 1.0691,
+      "step": 6932
+    },
+    {
+      "epoch": 1.3701306813922691,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.651499092646056e-06,
+      "loss": 0.9777,
+      "step": 6933
+    },
+    {
+      "epoch": 1.370330576447365,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.650454260849349e-06,
+      "loss": 0.9519,
+      "step": 6934
+    },
+    {
+      "epoch": 1.370530471502461,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.64940940015953e-06,
+      "loss": 0.8727,
+      "step": 6935
+    },
+    {
+      "epoch": 1.3707303665575572,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.648364510623016e-06,
+      "loss": 0.9795,
+      "step": 6936
+    },
+    {
+      "epoch": 1.3709302616126533,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.64731959228622e-06,
+      "loss": 0.8466,
+      "step": 6937
+    },
+    {
+      "epoch": 1.3711301566677494,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.646274645195556e-06,
+      "loss": 1.0228,
+      "step": 6938
+    },
+    {
+      "epoch": 1.3713300517228455,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.645229669397443e-06,
+      "loss": 0.9517,
+      "step": 6939
+    },
+    {
+      "epoch": 1.3715299467779416,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.644184664938296e-06,
+      "loss": 0.995,
+      "step": 6940
+    },
+    {
+      "epoch": 1.3717298418330377,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.643139631864534e-06,
+      "loss": 0.9714,
+      "step": 6941
+    },
+    {
+      "epoch": 1.3719297368881338,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.642094570222579e-06,
+      "loss": 1.0516,
+      "step": 6942
+    },
+    {
+      "epoch": 1.37212963194323,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.641049480058853e-06,
+      "loss": 1.0417,
+      "step": 6943
+    },
+    {
+      "epoch": 1.372329526998326,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.640004361419776e-06,
+      "loss": 1.0357,
+      "step": 6944
+    },
+    {
+      "epoch": 1.3725294220534219,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.638959214351775e-06,
+      "loss": 0.9773,
+      "step": 6945
+    },
+    {
+      "epoch": 1.372729317108518,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.637914038901273e-06,
+      "loss": 1.0186,
+      "step": 6946
+    },
+    {
+      "epoch": 1.372929212163614,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.6368688351146975e-06,
+      "loss": 1.0735,
+      "step": 6947
+    },
+    {
+      "epoch": 1.3731291072187102,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.635823603038476e-06,
+      "loss": 1.0468,
+      "step": 6948
+    },
+    {
+      "epoch": 1.3733290022738063,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.634778342719038e-06,
+      "loss": 1.0165,
+      "step": 6949
+    },
+    {
+      "epoch": 1.3735288973289024,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.633733054202814e-06,
+      "loss": 0.9065,
+      "step": 6950
+    },
+    {
+      "epoch": 1.3737287923839985,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.632687737536236e-06,
+      "loss": 0.8383,
+      "step": 6951
+    },
+    {
+      "epoch": 1.3739286874390944,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.631642392765736e-06,
+      "loss": 0.9439,
+      "step": 6952
+    },
+    {
+      "epoch": 1.3741285824941905,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.630597019937748e-06,
+      "loss": 1.0172,
+      "step": 6953
+    },
+    {
+      "epoch": 1.3743284775492866,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.6295516190987075e-06,
+      "loss": 0.996,
+      "step": 6954
+    },
+    {
+      "epoch": 1.3745283726043827,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.628506190295052e-06,
+      "loss": 1.0551,
+      "step": 6955
+    },
+    {
+      "epoch": 1.3747282676594788,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.627460733573219e-06,
+      "loss": 0.9798,
+      "step": 6956
+    },
+    {
+      "epoch": 1.3749281627145749,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.626415248979646e-06,
+      "loss": 1.0186,
+      "step": 6957
+    },
+    {
+      "epoch": 1.375128057769671,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.625369736560776e-06,
+      "loss": 0.9548,
+      "step": 6958
+    },
+    {
+      "epoch": 1.375327952824767,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.624324196363048e-06,
+      "loss": 0.9933,
+      "step": 6959
+    },
+    {
+      "epoch": 1.3755278478798632,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.623278628432907e-06,
+      "loss": 1.0,
+      "step": 6960
+    },
+    {
+      "epoch": 1.3757277429349593,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.622233032816795e-06,
+      "loss": 1.0621,
+      "step": 6961
+    },
+    {
+      "epoch": 1.3759276379900554,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.6211874095611585e-06,
+      "loss": 0.9978,
+      "step": 6962
+    },
+    {
+      "epoch": 1.3761275330451512,
+      "grad_norm": 1.984375,
+      "learning_rate": 5.620141758712443e-06,
+      "loss": 0.9002,
+      "step": 6963
+    },
+    {
+      "epoch": 1.3763274281002473,
+      "grad_norm": 2.125,
+      "learning_rate": 5.619096080317097e-06,
+      "loss": 0.9789,
+      "step": 6964
+    },
+    {
+      "epoch": 1.3765273231553434,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.618050374421569e-06,
+      "loss": 1.0389,
+      "step": 6965
+    },
+    {
+      "epoch": 1.3767272182104395,
+      "grad_norm": 2.125,
+      "learning_rate": 5.61700464107231e-06,
+      "loss": 0.9557,
+      "step": 6966
+    },
+    {
+      "epoch": 1.3769271132655356,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.6159588803157705e-06,
+      "loss": 0.9106,
+      "step": 6967
+    },
+    {
+      "epoch": 1.3771270083206317,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.614913092198404e-06,
+      "loss": 0.9686,
+      "step": 6968
+    },
+    {
+      "epoch": 1.3773269033757276,
+      "grad_norm": 2.125,
+      "learning_rate": 5.613867276766662e-06,
+      "loss": 0.9622,
+      "step": 6969
+    },
+    {
+      "epoch": 1.3775267984308237,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.612821434067003e-06,
+      "loss": 0.9964,
+      "step": 6970
+    },
+    {
+      "epoch": 1.3777266934859198,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.61177556414588e-06,
+      "loss": 0.9981,
+      "step": 6971
+    },
+    {
+      "epoch": 1.377926588541016,
+      "grad_norm": 2.125,
+      "learning_rate": 5.610729667049751e-06,
+      "loss": 0.9791,
+      "step": 6972
+    },
+    {
+      "epoch": 1.378126483596112,
+      "grad_norm": 1.9296875,
+      "learning_rate": 5.609683742825078e-06,
+      "loss": 0.9303,
+      "step": 6973
+    },
+    {
+      "epoch": 1.378326378651208,
+      "grad_norm": 2.25,
+      "learning_rate": 5.608637791518318e-06,
+      "loss": 1.0528,
+      "step": 6974
+    },
+    {
+      "epoch": 1.3785262737063042,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.60759181317593e-06,
+      "loss": 0.934,
+      "step": 6975
+    },
+    {
+      "epoch": 1.3787261687614003,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.60654580784438e-06,
+      "loss": 0.9334,
+      "step": 6976
+    },
+    {
+      "epoch": 1.3789260638164964,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.60549977557013e-06,
+      "loss": 0.946,
+      "step": 6977
+    },
+    {
+      "epoch": 1.3791259588715925,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.604453716399643e-06,
+      "loss": 0.9288,
+      "step": 6978
+    },
+    {
+      "epoch": 1.3793258539266886,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.60340763037939e-06,
+      "loss": 0.9337,
+      "step": 6979
+    },
+    {
+      "epoch": 1.3795257489817845,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.602361517555831e-06,
+      "loss": 0.8842,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3797256440368806,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.601315377975439e-06,
+      "loss": 1.0394,
+      "step": 6981
+    },
+    {
+      "epoch": 1.3799255390919767,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.600269211684683e-06,
+      "loss": 1.0313,
+      "step": 6982
+    },
+    {
+      "epoch": 1.3801254341470728,
+      "grad_norm": 2.25,
+      "learning_rate": 5.599223018730031e-06,
+      "loss": 1.0387,
+      "step": 6983
+    },
+    {
+      "epoch": 1.3803253292021689,
+      "grad_norm": 2.125,
+      "learning_rate": 5.5981767991579575e-06,
+      "loss": 1.0096,
+      "step": 6984
+    },
+    {
+      "epoch": 1.380525224257265,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.597130553014934e-06,
+      "loss": 1.0312,
+      "step": 6985
+    },
+    {
+      "epoch": 1.380725119312361,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.596084280347435e-06,
+      "loss": 1.0451,
+      "step": 6986
+    },
+    {
+      "epoch": 1.380925014367457,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.595037981201939e-06,
+      "loss": 0.9005,
+      "step": 6987
+    },
+    {
+      "epoch": 1.381124909422553,
+      "grad_norm": 2.125,
+      "learning_rate": 5.5939916556249155e-06,
+      "loss": 1.0645,
+      "step": 6988
+    },
+    {
+      "epoch": 1.3813248044776492,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.592945303662848e-06,
+      "loss": 0.9322,
+      "step": 6989
+    },
+    {
+      "epoch": 1.3815246995327453,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.591898925362213e-06,
+      "loss": 0.9669,
+      "step": 6990
+    },
+    {
+      "epoch": 1.3817245945878414,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.590852520769491e-06,
+      "loss": 0.993,
+      "step": 6991
+    },
+    {
+      "epoch": 1.3819244896429375,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.589806089931163e-06,
+      "loss": 1.0358,
+      "step": 6992
+    },
+    {
+      "epoch": 1.3821243846980336,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.588759632893712e-06,
+      "loss": 0.9827,
+      "step": 6993
+    },
+    {
+      "epoch": 1.3823242797531297,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.587713149703619e-06,
+      "loss": 0.9969,
+      "step": 6994
+    },
+    {
+      "epoch": 1.3825241748082258,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.586666640407373e-06,
+      "loss": 0.9349,
+      "step": 6995
+    },
+    {
+      "epoch": 1.3827240698633219,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.585620105051457e-06,
+      "loss": 1.069,
+      "step": 6996
+    },
+    {
+      "epoch": 1.382923964918418,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.584573543682358e-06,
+      "loss": 0.945,
+      "step": 6997
+    },
+    {
+      "epoch": 1.3831238599735138,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.583526956346564e-06,
+      "loss": 0.9837,
+      "step": 6998
+    },
+    {
+      "epoch": 1.38332375502861,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.582480343090566e-06,
+      "loss": 0.8857,
+      "step": 6999
+    },
+    {
+      "epoch": 1.383523650083706,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.581433703960853e-06,
+      "loss": 0.887,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3837235451388021,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.580387039003918e-06,
+      "loss": 0.9843,
+      "step": 7001
+    },
+    {
+      "epoch": 1.3839234401938982,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.579340348266251e-06,
+      "loss": 0.9901,
+      "step": 7002
+    },
+    {
+      "epoch": 1.3841233352489943,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.578293631794348e-06,
+      "loss": 0.9512,
+      "step": 7003
+    },
+    {
+      "epoch": 1.3843232303040904,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.5772468896347045e-06,
+      "loss": 0.9057,
+      "step": 7004
+    },
+    {
+      "epoch": 1.3845231253591863,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.576200121833816e-06,
+      "loss": 0.9077,
+      "step": 7005
+    },
+    {
+      "epoch": 1.3847230204142824,
+      "grad_norm": 2.125,
+      "learning_rate": 5.575153328438178e-06,
+      "loss": 1.0334,
+      "step": 7006
+    },
+    {
+      "epoch": 1.3849229154693785,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.574106509494292e-06,
+      "loss": 0.9105,
+      "step": 7007
+    },
+    {
+      "epoch": 1.3851228105244746,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.573059665048656e-06,
+      "loss": 1.0605,
+      "step": 7008
+    },
+    {
+      "epoch": 1.3853227055795707,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.57201279514777e-06,
+      "loss": 1.0571,
+      "step": 7009
+    },
+    {
+      "epoch": 1.3855226006346668,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.570965899838138e-06,
+      "loss": 0.9358,
+      "step": 7010
+    },
+    {
+      "epoch": 1.385722495689763,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.56991897916626e-06,
+      "loss": 0.9903,
+      "step": 7011
+    },
+    {
+      "epoch": 1.385922390744859,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.5688720331786425e-06,
+      "loss": 1.0151,
+      "step": 7012
+    },
+    {
+      "epoch": 1.3861222857999551,
+      "grad_norm": 2.125,
+      "learning_rate": 5.567825061921791e-06,
+      "loss": 0.9319,
+      "step": 7013
+    },
+    {
+      "epoch": 1.3863221808550512,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.56677806544221e-06,
+      "loss": 0.9802,
+      "step": 7014
+    },
+    {
+      "epoch": 1.3865220759101473,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.565731043786409e-06,
+      "loss": 0.9866,
+      "step": 7015
+    },
+    {
+      "epoch": 1.3867219709652432,
+      "grad_norm": 2.25,
+      "learning_rate": 5.564683997000893e-06,
+      "loss": 1.0267,
+      "step": 7016
+    },
+    {
+      "epoch": 1.3869218660203393,
+      "grad_norm": 2.25,
+      "learning_rate": 5.5636369251321765e-06,
+      "loss": 1.038,
+      "step": 7017
+    },
+    {
+      "epoch": 1.3871217610754354,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.5625898282267675e-06,
+      "loss": 0.9183,
+      "step": 7018
+    },
+    {
+      "epoch": 1.3873216561305315,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.561542706331178e-06,
+      "loss": 1.0668,
+      "step": 7019
+    },
+    {
+      "epoch": 1.3875215511856276,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.560495559491922e-06,
+      "loss": 1.0296,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3877214462407237,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.559448387755513e-06,
+      "loss": 0.9755,
+      "step": 7021
+    },
+    {
+      "epoch": 1.3879213412958196,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.558401191168465e-06,
+      "loss": 1.003,
+      "step": 7022
+    },
+    {
+      "epoch": 1.3881212363509157,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.557353969777297e-06,
+      "loss": 0.9762,
+      "step": 7023
+    },
+    {
+      "epoch": 1.3883211314060118,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.556306723628526e-06,
+      "loss": 0.9916,
+      "step": 7024
+    },
+    {
+      "epoch": 1.3885210264611079,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.555259452768668e-06,
+      "loss": 0.9925,
+      "step": 7025
+    },
+    {
+      "epoch": 1.388720921516204,
+      "grad_norm": 2.125,
+      "learning_rate": 5.5542121572442446e-06,
+      "loss": 0.9753,
+      "step": 7026
+    },
+    {
+      "epoch": 1.3889208165713,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.553164837101778e-06,
+      "loss": 0.971,
+      "step": 7027
+    },
+    {
+      "epoch": 1.3891207116263962,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.552117492387786e-06,
+      "loss": 0.8416,
+      "step": 7028
+    },
+    {
+      "epoch": 1.3893206066814923,
+      "grad_norm": 2.125,
+      "learning_rate": 5.551070123148795e-06,
+      "loss": 0.9155,
+      "step": 7029
+    },
+    {
+      "epoch": 1.3895205017365884,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.5500227294313295e-06,
+      "loss": 1.0296,
+      "step": 7030
+    },
+    {
+      "epoch": 1.3897203967916845,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.548975311281911e-06,
+      "loss": 1.0358,
+      "step": 7031
+    },
+    {
+      "epoch": 1.3899202918467806,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.547927868747069e-06,
+      "loss": 1.0207,
+      "step": 7032
+    },
+    {
+      "epoch": 1.3901201869018764,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.546880401873329e-06,
+      "loss": 0.8628,
+      "step": 7033
+    },
+    {
+      "epoch": 1.3903200819569725,
+      "grad_norm": 2.125,
+      "learning_rate": 5.54583291070722e-06,
+      "loss": 0.9888,
+      "step": 7034
+    },
+    {
+      "epoch": 1.3905199770120686,
+      "grad_norm": 2.0,
+      "learning_rate": 5.544785395295273e-06,
+      "loss": 0.9659,
+      "step": 7035
+    },
+    {
+      "epoch": 1.3907198720671647,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.543737855684015e-06,
+      "loss": 0.9611,
+      "step": 7036
+    },
+    {
+      "epoch": 1.3909197671222608,
+      "grad_norm": 2.25,
+      "learning_rate": 5.54269029191998e-06,
+      "loss": 0.8994,
+      "step": 7037
+    },
+    {
+      "epoch": 1.391119662177357,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.541642704049701e-06,
+      "loss": 0.9457,
+      "step": 7038
+    },
+    {
+      "epoch": 1.391319557232453,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.540595092119709e-06,
+      "loss": 0.9801,
+      "step": 7039
+    },
+    {
+      "epoch": 1.391519452287549,
+      "grad_norm": 2.125,
+      "learning_rate": 5.539547456176543e-06,
+      "loss": 0.965,
+      "step": 7040
+    },
+    {
+      "epoch": 1.391719347342645,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.538499796266735e-06,
+      "loss": 0.9907,
+      "step": 7041
+    },
+    {
+      "epoch": 1.3919192423977411,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.537452112436824e-06,
+      "loss": 0.9189,
+      "step": 7042
+    },
+    {
+      "epoch": 1.3921191374528372,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.536404404733348e-06,
+      "loss": 0.9072,
+      "step": 7043
+    },
+    {
+      "epoch": 1.3923190325079333,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.535356673202845e-06,
+      "loss": 0.9507,
+      "step": 7044
+    },
+    {
+      "epoch": 1.3925189275630294,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.534308917891856e-06,
+      "loss": 0.9909,
+      "step": 7045
+    },
+    {
+      "epoch": 1.3927188226181255,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.533261138846922e-06,
+      "loss": 0.9652,
+      "step": 7046
+    },
+    {
+      "epoch": 1.3929187176732216,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.532213336114586e-06,
+      "loss": 0.9624,
+      "step": 7047
+    },
+    {
+      "epoch": 1.3931186127283177,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.531165509741388e-06,
+      "loss": 0.9854,
+      "step": 7048
+    },
+    {
+      "epoch": 1.3933185077834138,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.5301176597738785e-06,
+      "loss": 1.0033,
+      "step": 7049
+    },
+    {
+      "epoch": 1.39351840283851,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.529069786258596e-06,
+      "loss": 0.9078,
+      "step": 7050
+    },
+    {
+      "epoch": 1.3937182978936058,
+      "grad_norm": 2.125,
+      "learning_rate": 5.5280218892420925e-06,
+      "loss": 0.9249,
+      "step": 7051
+    },
+    {
+      "epoch": 1.393918192948702,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.526973968770911e-06,
+      "loss": 1.0552,
+      "step": 7052
+    },
+    {
+      "epoch": 1.394118088003798,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.525926024891603e-06,
+      "loss": 1.0302,
+      "step": 7053
+    },
+    {
+      "epoch": 1.394317983058894,
+      "grad_norm": 2.25,
+      "learning_rate": 5.524878057650717e-06,
+      "loss": 1.002,
+      "step": 7054
+    },
+    {
+      "epoch": 1.3945178781139902,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.5238300670948044e-06,
+      "loss": 0.9046,
+      "step": 7055
+    },
+    {
+      "epoch": 1.3947177731690863,
+      "grad_norm": 2.125,
+      "learning_rate": 5.522782053270414e-06,
+      "loss": 1.0101,
+      "step": 7056
+    },
+    {
+      "epoch": 1.3949176682241822,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.521734016224103e-06,
+      "loss": 1.0308,
+      "step": 7057
+    },
+    {
+      "epoch": 1.3951175632792783,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.5206859560024215e-06,
+      "loss": 1.0799,
+      "step": 7058
+    },
+    {
+      "epoch": 1.3953174583343744,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.519637872651926e-06,
+      "loss": 0.932,
+      "step": 7059
+    },
+    {
+      "epoch": 1.3955173533894705,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.518589766219173e-06,
+      "loss": 1.0846,
+      "step": 7060
+    },
+    {
+      "epoch": 1.3957172484445666,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.517541636750715e-06,
+      "loss": 0.9484,
+      "step": 7061
+    },
+    {
+      "epoch": 1.3959171434996627,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.516493484293114e-06,
+      "loss": 1.0155,
+      "step": 7062
+    },
+    {
+      "epoch": 1.3961170385547588,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.515445308892928e-06,
+      "loss": 0.9161,
+      "step": 7063
+    },
+    {
+      "epoch": 1.3963169336098549,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.514397110596716e-06,
+      "loss": 0.9939,
+      "step": 7064
+    },
+    {
+      "epoch": 1.396516828664951,
+      "grad_norm": 2.125,
+      "learning_rate": 5.513348889451039e-06,
+      "loss": 0.903,
+      "step": 7065
+    },
+    {
+      "epoch": 1.396716723720047,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.512300645502462e-06,
+      "loss": 0.9597,
+      "step": 7066
+    },
+    {
+      "epoch": 1.3969166187751432,
+      "grad_norm": 2.125,
+      "learning_rate": 5.511252378797542e-06,
+      "loss": 0.9626,
+      "step": 7067
+    },
+    {
+      "epoch": 1.397116513830239,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.510204089382847e-06,
+      "loss": 1.0461,
+      "step": 7068
+    },
+    {
+      "epoch": 1.3973164088853351,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.509155777304941e-06,
+      "loss": 0.9781,
+      "step": 7069
+    },
+    {
+      "epoch": 1.3975163039404312,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.508107442610388e-06,
+      "loss": 1.0899,
+      "step": 7070
+    },
+    {
+      "epoch": 1.3977161989955273,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.507059085345759e-06,
+      "loss": 0.9567,
+      "step": 7071
+    },
+    {
+      "epoch": 1.3979160940506234,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.506010705557618e-06,
+      "loss": 0.9734,
+      "step": 7072
+    },
+    {
+      "epoch": 1.3981159891057195,
+      "grad_norm": 2.125,
+      "learning_rate": 5.5049623032925355e-06,
+      "loss": 0.8962,
+      "step": 7073
+    },
+    {
+      "epoch": 1.3983158841608156,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.503913878597082e-06,
+      "loss": 0.9915,
+      "step": 7074
+    },
+    {
+      "epoch": 1.3985157792159115,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.502865431517828e-06,
+      "loss": 0.966,
+      "step": 7075
+    },
+    {
+      "epoch": 1.3987156742710076,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.501816962101345e-06,
+      "loss": 0.973,
+      "step": 7076
+    },
+    {
+      "epoch": 1.3989155693261037,
+      "grad_norm": 2.25,
+      "learning_rate": 5.500768470394207e-06,
+      "loss": 1.0737,
+      "step": 7077
+    },
+    {
+      "epoch": 1.3991154643811998,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.499719956442985e-06,
+      "loss": 0.867,
+      "step": 7078
+    },
+    {
+      "epoch": 1.399315359436296,
+      "grad_norm": 2.125,
+      "learning_rate": 5.498671420294257e-06,
+      "loss": 0.9767,
+      "step": 7079
+    },
+    {
+      "epoch": 1.399515254491392,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.497622861994598e-06,
+      "loss": 1.0266,
+      "step": 7080
+    },
+    {
+      "epoch": 1.3997151495464881,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.496574281590585e-06,
+      "loss": 0.9242,
+      "step": 7081
+    },
+    {
+      "epoch": 1.3999150446015842,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.4955256791287946e-06,
+      "loss": 0.9276,
+      "step": 7082
+    },
+    {
+      "epoch": 1.4001149396566803,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.494477054655808e-06,
+      "loss": 0.9677,
+      "step": 7083
+    },
+    {
+      "epoch": 1.4003148347117764,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.493428408218202e-06,
+      "loss": 0.9379,
+      "step": 7084
+    },
+    {
+      "epoch": 1.4005147297668725,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.492379739862559e-06,
+      "loss": 0.9599,
+      "step": 7085
+    },
+    {
+      "epoch": 1.4007146248219684,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.491331049635461e-06,
+      "loss": 1.0505,
+      "step": 7086
+    },
+    {
+      "epoch": 1.4009145198770645,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.490282337583489e-06,
+      "loss": 0.9282,
+      "step": 7087
+    },
+    {
+      "epoch": 1.4011144149321606,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.489233603753228e-06,
+      "loss": 0.8507,
+      "step": 7088
+    },
+    {
+      "epoch": 1.4013143099872567,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.488184848191265e-06,
+      "loss": 0.9696,
+      "step": 7089
+    },
+    {
+      "epoch": 1.4015142050423528,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.48713607094418e-06,
+      "loss": 0.9663,
+      "step": 7090
+    },
+    {
+      "epoch": 1.401714100097449,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.486087272058566e-06,
+      "loss": 0.987,
+      "step": 7091
+    },
+    {
+      "epoch": 1.4019139951525448,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.485038451581004e-06,
+      "loss": 0.9364,
+      "step": 7092
+    },
+    {
+      "epoch": 1.4021138902076409,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.4839896095580865e-06,
+      "loss": 0.9397,
+      "step": 7093
+    },
+    {
+      "epoch": 1.402313785262737,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.4829407460364045e-06,
+      "loss": 0.9994,
+      "step": 7094
+    },
+    {
+      "epoch": 1.402513680317833,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.481891861062545e-06,
+      "loss": 0.9481,
+      "step": 7095
+    },
+    {
+      "epoch": 1.4027135753729292,
+      "grad_norm": 2.25,
+      "learning_rate": 5.480842954683099e-06,
+      "loss": 0.9541,
+      "step": 7096
+    },
+    {
+      "epoch": 1.4029134704280253,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.479794026944663e-06,
+      "loss": 1.041,
+      "step": 7097
+    },
+    {
+      "epoch": 1.4031133654831214,
+      "grad_norm": 2.25,
+      "learning_rate": 5.478745077893827e-06,
+      "loss": 1.0423,
+      "step": 7098
+    },
+    {
+      "epoch": 1.4033132605382175,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.477696107577184e-06,
+      "loss": 1.0669,
+      "step": 7099
+    },
+    {
+      "epoch": 1.4035131555933136,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.476647116041332e-06,
+      "loss": 0.9499,
+      "step": 7100
+    },
+    {
+      "epoch": 1.4037130506484097,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.4755981033328655e-06,
+      "loss": 1.0964,
+      "step": 7101
+    },
+    {
+      "epoch": 1.4039129457035058,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.474549069498381e-06,
+      "loss": 0.8984,
+      "step": 7102
+    },
+    {
+      "epoch": 1.4041128407586017,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.473500014584481e-06,
+      "loss": 0.9057,
+      "step": 7103
+    },
+    {
+      "epoch": 1.4043127358136978,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.472450938637758e-06,
+      "loss": 0.9872,
+      "step": 7104
+    },
+    {
+      "epoch": 1.4045126308687939,
+      "grad_norm": 2.125,
+      "learning_rate": 5.471401841704816e-06,
+      "loss": 0.9796,
+      "step": 7105
+    },
+    {
+      "epoch": 1.40471252592389,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.470352723832254e-06,
+      "loss": 1.0426,
+      "step": 7106
+    },
+    {
+      "epoch": 1.404912420978986,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.4693035850666734e-06,
+      "loss": 1.0271,
+      "step": 7107
+    },
+    {
+      "epoch": 1.4051123160340822,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.46825442545468e-06,
+      "loss": 0.9901,
+      "step": 7108
+    },
+    {
+      "epoch": 1.4053122110891783,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.467205245042873e-06,
+      "loss": 0.9917,
+      "step": 7109
+    },
+    {
+      "epoch": 1.4055121061442741,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.466156043877859e-06,
+      "loss": 1.0446,
+      "step": 7110
+    },
+    {
+      "epoch": 1.4057120011993702,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.465106822006244e-06,
+      "loss": 0.9722,
+      "step": 7111
+    },
+    {
+      "epoch": 1.4059118962544663,
+      "grad_norm": 2.125,
+      "learning_rate": 5.4640575794746335e-06,
+      "loss": 1.0074,
+      "step": 7112
+    },
+    {
+      "epoch": 1.4061117913095624,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.463008316329636e-06,
+      "loss": 1.0497,
+      "step": 7113
+    },
+    {
+      "epoch": 1.4063116863646585,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.461959032617857e-06,
+      "loss": 1.0138,
+      "step": 7114
+    },
+    {
+      "epoch": 1.4065115814197546,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.4609097283859084e-06,
+      "loss": 0.9636,
+      "step": 7115
+    },
+    {
+      "epoch": 1.4067114764748507,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.459860403680397e-06,
+      "loss": 0.9554,
+      "step": 7116
+    },
+    {
+      "epoch": 1.4069113715299468,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.458811058547937e-06,
+      "loss": 1.0052,
+      "step": 7117
+    },
+    {
+      "epoch": 1.407111266585043,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.457761693035139e-06,
+      "loss": 1.023,
+      "step": 7118
+    },
+    {
+      "epoch": 1.407311161640139,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.456712307188614e-06,
+      "loss": 1.0921,
+      "step": 7119
+    },
+    {
+      "epoch": 1.4075110566952351,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.4556629010549785e-06,
+      "loss": 0.9256,
+      "step": 7120
+    },
+    {
+      "epoch": 1.407710951750331,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.454613474680844e-06,
+      "loss": 0.854,
+      "step": 7121
+    },
+    {
+      "epoch": 1.407910846805427,
+      "grad_norm": 2.125,
+      "learning_rate": 5.45356402811283e-06,
+      "loss": 0.9923,
+      "step": 7122
+    },
+    {
+      "epoch": 1.4081107418605232,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.452514561397549e-06,
+      "loss": 0.9985,
+      "step": 7123
+    },
+    {
+      "epoch": 1.4083106369156193,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.451465074581619e-06,
+      "loss": 1.0087,
+      "step": 7124
+    },
+    {
+      "epoch": 1.4085105319707154,
+      "grad_norm": 2.125,
+      "learning_rate": 5.450415567711659e-06,
+      "loss": 0.9521,
+      "step": 7125
+    },
+    {
+      "epoch": 1.4087104270258115,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.449366040834287e-06,
+      "loss": 0.9508,
+      "step": 7126
+    },
+    {
+      "epoch": 1.4089103220809076,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.448316493996124e-06,
+      "loss": 1.0183,
+      "step": 7127
+    },
+    {
+      "epoch": 1.4091102171360035,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.447266927243792e-06,
+      "loss": 1.042,
+      "step": 7128
+    },
+    {
+      "epoch": 1.4093101121910996,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.446217340623909e-06,
+      "loss": 1.0149,
+      "step": 7129
+    },
+    {
+      "epoch": 1.4095100072461957,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.445167734183099e-06,
+      "loss": 0.9286,
+      "step": 7130
+    },
+    {
+      "epoch": 1.4097099023012918,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.444118107967987e-06,
+      "loss": 1.0246,
+      "step": 7131
+    },
+    {
+      "epoch": 1.4099097973563879,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.443068462025195e-06,
+      "loss": 1.0767,
+      "step": 7132
+    },
+    {
+      "epoch": 1.410109692411484,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.442018796401349e-06,
+      "loss": 0.9486,
+      "step": 7133
+    },
+    {
+      "epoch": 1.41030958746658,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.440969111143076e-06,
+      "loss": 0.9949,
+      "step": 7134
+    },
+    {
+      "epoch": 1.4105094825216762,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.439919406297002e-06,
+      "loss": 0.8934,
+      "step": 7135
+    },
+    {
+      "epoch": 1.4107093775767723,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.438869681909757e-06,
+      "loss": 0.9253,
+      "step": 7136
+    },
+    {
+      "epoch": 1.4109092726318684,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.437819938027964e-06,
+      "loss": 0.8921,
+      "step": 7137
+    },
+    {
+      "epoch": 1.4111091676869643,
+      "grad_norm": 2.125,
+      "learning_rate": 5.436770174698257e-06,
+      "loss": 1.0573,
+      "step": 7138
+    },
+    {
+      "epoch": 1.4113090627420604,
+      "grad_norm": 1.984375,
+      "learning_rate": 5.435720391967267e-06,
+      "loss": 0.9288,
+      "step": 7139
+    },
+    {
+      "epoch": 1.4115089577971565,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.434670589881623e-06,
+      "loss": 0.9548,
+      "step": 7140
+    },
+    {
+      "epoch": 1.4117088528522526,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.433620768487957e-06,
+      "loss": 0.94,
+      "step": 7141
+    },
+    {
+      "epoch": 1.4119087479073487,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.432570927832906e-06,
+      "loss": 0.9368,
+      "step": 7142
+    },
+    {
+      "epoch": 1.4121086429624448,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.431521067963097e-06,
+      "loss": 0.9632,
+      "step": 7143
+    },
+    {
+      "epoch": 1.4123085380175409,
+      "grad_norm": 2.125,
+      "learning_rate": 5.430471188925169e-06,
+      "loss": 0.9206,
+      "step": 7144
+    },
+    {
+      "epoch": 1.4125084330726367,
+      "grad_norm": 2.25,
+      "learning_rate": 5.4294212907657584e-06,
+      "loss": 0.9516,
+      "step": 7145
+    },
+    {
+      "epoch": 1.4127083281277328,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.428371373531498e-06,
+      "loss": 0.9555,
+      "step": 7146
+    },
+    {
+      "epoch": 1.412908223182829,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.427321437269027e-06,
+      "loss": 0.9503,
+      "step": 7147
+    },
+    {
+      "epoch": 1.413108118237925,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.4262714820249855e-06,
+      "loss": 0.9878,
+      "step": 7148
+    },
+    {
+      "epoch": 1.4133080132930211,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.425221507846008e-06,
+      "loss": 1.0085,
+      "step": 7149
+    },
+    {
+      "epoch": 1.4135079083481172,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.424171514778738e-06,
+      "loss": 1.0855,
+      "step": 7150
+    },
+    {
+      "epoch": 1.4137078034032133,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.423121502869814e-06,
+      "loss": 1.0379,
+      "step": 7151
+    },
+    {
+      "epoch": 1.4139076984583094,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.422071472165877e-06,
+      "loss": 0.9778,
+      "step": 7152
+    },
+    {
+      "epoch": 1.4141075935134055,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.421021422713573e-06,
+      "loss": 0.9182,
+      "step": 7153
+    },
+    {
+      "epoch": 1.4143074885685016,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.41997135455954e-06,
+      "loss": 1.0276,
+      "step": 7154
+    },
+    {
+      "epoch": 1.4145073836235977,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.418921267750425e-06,
+      "loss": 0.9951,
+      "step": 7155
+    },
+    {
+      "epoch": 1.4147072786786936,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.417871162332872e-06,
+      "loss": 0.9143,
+      "step": 7156
+    },
+    {
+      "epoch": 1.4149071737337897,
+      "grad_norm": 2.125,
+      "learning_rate": 5.416821038353526e-06,
+      "loss": 0.9185,
+      "step": 7157
+    },
+    {
+      "epoch": 1.4151070687888858,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.415770895859034e-06,
+      "loss": 0.9611,
+      "step": 7158
+    },
+    {
+      "epoch": 1.415306963843982,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.4147207348960466e-06,
+      "loss": 0.9433,
+      "step": 7159
+    },
+    {
+      "epoch": 1.415506858899078,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.413670555511204e-06,
+      "loss": 0.8962,
+      "step": 7160
+    },
+    {
+      "epoch": 1.4157067539541741,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.412620357751161e-06,
+      "loss": 0.9701,
+      "step": 7161
+    },
+    {
+      "epoch": 1.4159066490092702,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.411570141662567e-06,
+      "loss": 0.9616,
+      "step": 7162
+    },
+    {
+      "epoch": 1.416106544064366,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.41051990729207e-06,
+      "loss": 0.9964,
+      "step": 7163
+    },
+    {
+      "epoch": 1.4163064391194622,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.409469654686323e-06,
+      "loss": 0.9548,
+      "step": 7164
+    },
+    {
+      "epoch": 1.4165063341745583,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.408419383891978e-06,
+      "loss": 1.064,
+      "step": 7165
+    },
+    {
+      "epoch": 1.4167062292296544,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.407369094955685e-06,
+      "loss": 0.9294,
+      "step": 7166
+    },
+    {
+      "epoch": 1.4169061242847505,
+      "grad_norm": 2.125,
+      "learning_rate": 5.406318787924104e-06,
+      "loss": 0.9913,
+      "step": 7167
+    },
+    {
+      "epoch": 1.4171060193398466,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.4052684628438836e-06,
+      "loss": 0.9951,
+      "step": 7168
+    },
+    {
+      "epoch": 1.4173059143949427,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.404218119761682e-06,
+      "loss": 1.0213,
+      "step": 7169
+    },
+    {
+      "epoch": 1.4175058094500388,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.403167758724155e-06,
+      "loss": 0.9688,
+      "step": 7170
+    },
+    {
+      "epoch": 1.417705704505135,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.402117379777958e-06,
+      "loss": 0.982,
+      "step": 7171
+    },
+    {
+      "epoch": 1.417905599560231,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.401066982969751e-06,
+      "loss": 0.9972,
+      "step": 7172
+    },
+    {
+      "epoch": 1.418105494615327,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.400016568346192e-06,
+      "loss": 0.9878,
+      "step": 7173
+    },
+    {
+      "epoch": 1.418305389670423,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.398966135953938e-06,
+      "loss": 0.9699,
+      "step": 7174
+    },
+    {
+      "epoch": 1.418505284725519,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.397915685839652e-06,
+      "loss": 0.9125,
+      "step": 7175
+    },
+    {
+      "epoch": 1.4187051797806152,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.396865218049995e-06,
+      "loss": 0.9142,
+      "step": 7176
+    },
+    {
+      "epoch": 1.4189050748357113,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.395814732631625e-06,
+      "loss": 0.9923,
+      "step": 7177
+    },
+    {
+      "epoch": 1.4191049698908074,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.394764229631207e-06,
+      "loss": 0.976,
+      "step": 7178
+    },
+    {
+      "epoch": 1.4193048649459035,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.393713709095406e-06,
+      "loss": 0.894,
+      "step": 7179
+    },
+    {
+      "epoch": 1.4195047600009993,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.3926631710708835e-06,
+      "loss": 0.956,
+      "step": 7180
+    },
+    {
+      "epoch": 1.4197046550560954,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.3916126156043045e-06,
+      "loss": 0.8711,
+      "step": 7181
+    },
+    {
+      "epoch": 1.4199045501111915,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.3905620427423344e-06,
+      "loss": 0.9661,
+      "step": 7182
+    },
+    {
+      "epoch": 1.4201044451662876,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.38951145253164e-06,
+      "loss": 0.963,
+      "step": 7183
+    },
+    {
+      "epoch": 1.4203043402213837,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.388460845018889e-06,
+      "loss": 1.0755,
+      "step": 7184
+    },
+    {
+      "epoch": 1.4205042352764798,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.3874102202507485e-06,
+      "loss": 0.934,
+      "step": 7185
+    },
+    {
+      "epoch": 1.420704130331576,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.386359578273888e-06,
+      "loss": 0.9272,
+      "step": 7186
+    },
+    {
+      "epoch": 1.420904025386672,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.385308919134976e-06,
+      "loss": 0.9468,
+      "step": 7187
+    },
+    {
+      "epoch": 1.4211039204417681,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.384258242880682e-06,
+      "loss": 1.0303,
+      "step": 7188
+    },
+    {
+      "epoch": 1.4213038154968642,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.3832075495576794e-06,
+      "loss": 0.9291,
+      "step": 7189
+    },
+    {
+      "epoch": 1.4215037105519603,
+      "grad_norm": 2.25,
+      "learning_rate": 5.382156839212639e-06,
+      "loss": 0.9736,
+      "step": 7190
+    },
+    {
+      "epoch": 1.4217036056070562,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.381106111892231e-06,
+      "loss": 0.9767,
+      "step": 7191
+    },
+    {
+      "epoch": 1.4219035006621523,
+      "grad_norm": 2.25,
+      "learning_rate": 5.38005536764313e-06,
+      "loss": 1.0116,
+      "step": 7192
+    },
+    {
+      "epoch": 1.4221033957172484,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.3790046065120116e-06,
+      "loss": 0.9492,
+      "step": 7193
+    },
+    {
+      "epoch": 1.4223032907723445,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.377953828545548e-06,
+      "loss": 0.9302,
+      "step": 7194
+    },
+    {
+      "epoch": 1.4225031858274406,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.376903033790416e-06,
+      "loss": 1.0104,
+      "step": 7195
+    },
+    {
+      "epoch": 1.4227030808825367,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.375852222293292e-06,
+      "loss": 0.9255,
+      "step": 7196
+    },
+    {
+      "epoch": 1.4229029759376328,
+      "grad_norm": 2.125,
+      "learning_rate": 5.374801394100851e-06,
+      "loss": 0.9987,
+      "step": 7197
+    },
+    {
+      "epoch": 1.4231028709927287,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.373750549259773e-06,
+      "loss": 0.9395,
+      "step": 7198
+    },
+    {
+      "epoch": 1.4233027660478248,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.372699687816736e-06,
+      "loss": 0.8662,
+      "step": 7199
+    },
+    {
+      "epoch": 1.423502661102921,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.3716488098184175e-06,
+      "loss": 0.9063,
+      "step": 7200
+    },
+    {
+      "epoch": 1.423702556158017,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.3705979153115e-06,
+      "loss": 0.9422,
+      "step": 7201
+    },
+    {
+      "epoch": 1.423902451213113,
+      "grad_norm": 2.25,
+      "learning_rate": 5.369547004342661e-06,
+      "loss": 0.9789,
+      "step": 7202
+    },
+    {
+      "epoch": 1.4241023462682092,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.368496076958584e-06,
+      "loss": 0.9639,
+      "step": 7203
+    },
+    {
+      "epoch": 1.4243022413233053,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.367445133205952e-06,
+      "loss": 0.9358,
+      "step": 7204
+    },
+    {
+      "epoch": 1.4245021363784014,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.366394173131445e-06,
+      "loss": 0.9495,
+      "step": 7205
+    },
+    {
+      "epoch": 1.4247020314334975,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.365343196781749e-06,
+      "loss": 0.813,
+      "step": 7206
+    },
+    {
+      "epoch": 1.4249019264885936,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.364292204203548e-06,
+      "loss": 1.0009,
+      "step": 7207
+    },
+    {
+      "epoch": 1.4251018215436897,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.363241195443524e-06,
+      "loss": 1.0062,
+      "step": 7208
+    },
+    {
+      "epoch": 1.4253017165987856,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.362190170548365e-06,
+      "loss": 0.9212,
+      "step": 7209
+    },
+    {
+      "epoch": 1.4255016116538817,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.3611391295647585e-06,
+      "loss": 0.8932,
+      "step": 7210
+    },
+    {
+      "epoch": 1.4257015067089778,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.36008807253939e-06,
+      "loss": 0.9705,
+      "step": 7211
+    },
+    {
+      "epoch": 1.4259014017640739,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.359036999518948e-06,
+      "loss": 0.9085,
+      "step": 7212
+    },
+    {
+      "epoch": 1.42610129681917,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.35798591055012e-06,
+      "loss": 0.9985,
+      "step": 7213
+    },
+    {
+      "epoch": 1.426301191874266,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.356934805679597e-06,
+      "loss": 0.8962,
+      "step": 7214
+    },
+    {
+      "epoch": 1.426501086929362,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.355883684954068e-06,
+      "loss": 0.9946,
+      "step": 7215
+    },
+    {
+      "epoch": 1.426700981984458,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.354832548420222e-06,
+      "loss": 1.0097,
+      "step": 7216
+    },
+    {
+      "epoch": 1.4269008770395541,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.3537813961247546e-06,
+      "loss": 0.9663,
+      "step": 7217
+    },
+    {
+      "epoch": 1.4271007720946502,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.352730228114354e-06,
+      "loss": 0.9505,
+      "step": 7218
+    },
+    {
+      "epoch": 1.4273006671497463,
+      "grad_norm": 2.25,
+      "learning_rate": 5.351679044435714e-06,
+      "loss": 1.0214,
+      "step": 7219
+    },
+    {
+      "epoch": 1.4275005622048424,
+      "grad_norm": 2.25,
+      "learning_rate": 5.35062784513553e-06,
+      "loss": 1.0575,
+      "step": 7220
+    },
+    {
+      "epoch": 1.4277004572599385,
+      "grad_norm": 2.25,
+      "learning_rate": 5.3495766302604945e-06,
+      "loss": 0.9601,
+      "step": 7221
+    },
+    {
+      "epoch": 1.4279003523150346,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.348525399857301e-06,
+      "loss": 0.9333,
+      "step": 7222
+    },
+    {
+      "epoch": 1.4281002473701307,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.347474153972647e-06,
+      "loss": 1.025,
+      "step": 7223
+    },
+    {
+      "epoch": 1.4283001424252268,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.346422892653229e-06,
+      "loss": 0.9268,
+      "step": 7224
+    },
+    {
+      "epoch": 1.428500037480323,
+      "grad_norm": 2.25,
+      "learning_rate": 5.345371615945742e-06,
+      "loss": 0.971,
+      "step": 7225
+    },
+    {
+      "epoch": 1.4286999325354188,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.344320323896886e-06,
+      "loss": 0.9422,
+      "step": 7226
+    },
+    {
+      "epoch": 1.428899827590515,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.343269016553355e-06,
+      "loss": 0.9879,
+      "step": 7227
+    },
+    {
+      "epoch": 1.429099722645611,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.342217693961853e-06,
+      "loss": 0.9466,
+      "step": 7228
+    },
+    {
+      "epoch": 1.4292996177007071,
+      "grad_norm": 2.125,
+      "learning_rate": 5.341166356169079e-06,
+      "loss": 0.9764,
+      "step": 7229
+    },
+    {
+      "epoch": 1.4294995127558032,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.34011500322173e-06,
+      "loss": 0.9754,
+      "step": 7230
+    },
+    {
+      "epoch": 1.4296994078108993,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.339063635166508e-06,
+      "loss": 1.0708,
+      "step": 7231
+    },
+    {
+      "epoch": 1.4298993028659954,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.338012252050118e-06,
+      "loss": 1.0137,
+      "step": 7232
+    },
+    {
+      "epoch": 1.4300991979210913,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.336960853919259e-06,
+      "loss": 0.9512,
+      "step": 7233
+    },
+    {
+      "epoch": 1.4302990929761874,
+      "grad_norm": 1.9921875,
+      "learning_rate": 5.335909440820635e-06,
+      "loss": 0.861,
+      "step": 7234
+    },
+    {
+      "epoch": 1.4304989880312835,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.334858012800948e-06,
+      "loss": 1.0068,
+      "step": 7235
+    },
+    {
+      "epoch": 1.4306988830863796,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.333806569906904e-06,
+      "loss": 0.959,
+      "step": 7236
+    },
+    {
+      "epoch": 1.4308987781414757,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.3327551121852095e-06,
+      "loss": 0.8968,
+      "step": 7237
+    },
+    {
+      "epoch": 1.4310986731965718,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.331703639682568e-06,
+      "loss": 0.949,
+      "step": 7238
+    },
+    {
+      "epoch": 1.431298568251668,
+      "grad_norm": 2.125,
+      "learning_rate": 5.330652152445686e-06,
+      "loss": 0.9953,
+      "step": 7239
+    },
+    {
+      "epoch": 1.431498463306764,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.329600650521272e-06,
+      "loss": 0.9063,
+      "step": 7240
+    },
+    {
+      "epoch": 1.43169835836186,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.32854913395603e-06,
+      "loss": 0.9747,
+      "step": 7241
+    },
+    {
+      "epoch": 1.4318982534169562,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.327497602796671e-06,
+      "loss": 1.0319,
+      "step": 7242
+    },
+    {
+      "epoch": 1.4320981484720523,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.326446057089905e-06,
+      "loss": 0.9916,
+      "step": 7243
+    },
+    {
+      "epoch": 1.4322980435271482,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.325394496882439e-06,
+      "loss": 0.9394,
+      "step": 7244
+    },
+    {
+      "epoch": 1.4324979385822443,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.324342922220983e-06,
+      "loss": 0.9727,
+      "step": 7245
+    },
+    {
+      "epoch": 1.4326978336373404,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.323291333152251e-06,
+      "loss": 1.0183,
+      "step": 7246
+    },
+    {
+      "epoch": 1.4328977286924365,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.322239729722951e-06,
+      "loss": 0.9756,
+      "step": 7247
+    },
+    {
+      "epoch": 1.4330976237475326,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.321188111979797e-06,
+      "loss": 0.9,
+      "step": 7248
+    },
+    {
+      "epoch": 1.4332975188026287,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.320136479969502e-06,
+      "loss": 0.892,
+      "step": 7249
+    },
+    {
+      "epoch": 1.4334974138577246,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.319084833738779e-06,
+      "loss": 0.9366,
+      "step": 7250
+    },
+    {
+      "epoch": 1.4336973089128207,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.318033173334341e-06,
+      "loss": 0.9634,
+      "step": 7251
+    },
+    {
+      "epoch": 1.4338972039679168,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.316981498802905e-06,
+      "loss": 0.8643,
+      "step": 7252
+    },
+    {
+      "epoch": 1.4340970990230129,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.315929810191183e-06,
+      "loss": 0.977,
+      "step": 7253
+    },
+    {
+      "epoch": 1.434296994078109,
+      "grad_norm": 2.125,
+      "learning_rate": 5.3148781075458924e-06,
+      "loss": 0.9984,
+      "step": 7254
+    },
+    {
+      "epoch": 1.434496889133205,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.313826390913751e-06,
+      "loss": 0.9491,
+      "step": 7255
+    },
+    {
+      "epoch": 1.4346967841883012,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.312774660341473e-06,
+      "loss": 1.0773,
+      "step": 7256
+    },
+    {
+      "epoch": 1.4348966792433973,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.311722915875781e-06,
+      "loss": 1.0079,
+      "step": 7257
+    },
+    {
+      "epoch": 1.4350965742984934,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.310671157563387e-06,
+      "loss": 1.0116,
+      "step": 7258
+    },
+    {
+      "epoch": 1.4352964693535895,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.309619385451016e-06,
+      "loss": 1.018,
+      "step": 7259
+    },
+    {
+      "epoch": 1.4354963644086856,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.308567599585384e-06,
+      "loss": 0.965,
+      "step": 7260
+    },
+    {
+      "epoch": 1.4356962594637814,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.307515800013212e-06,
+      "loss": 1.0097,
+      "step": 7261
+    },
+    {
+      "epoch": 1.4358961545188775,
+      "grad_norm": 2.125,
+      "learning_rate": 5.3064639867812205e-06,
+      "loss": 1.0474,
+      "step": 7262
+    },
+    {
+      "epoch": 1.4360960495739736,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.305412159936133e-06,
+      "loss": 0.9374,
+      "step": 7263
+    },
+    {
+      "epoch": 1.4362959446290697,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.3043603195246684e-06,
+      "loss": 0.9425,
+      "step": 7264
+    },
+    {
+      "epoch": 1.4364958396841658,
+      "grad_norm": 2.125,
+      "learning_rate": 5.303308465593552e-06,
+      "loss": 0.9616,
+      "step": 7265
+    },
+    {
+      "epoch": 1.436695734739262,
+      "grad_norm": 2.125,
+      "learning_rate": 5.3022565981895045e-06,
+      "loss": 0.994,
+      "step": 7266
+    },
+    {
+      "epoch": 1.436895629794358,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.301204717359253e-06,
+      "loss": 1.0014,
+      "step": 7267
+    },
+    {
+      "epoch": 1.437095524849454,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.300152823149519e-06,
+      "loss": 1.0367,
+      "step": 7268
+    },
+    {
+      "epoch": 1.43729541990455,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.299100915607029e-06,
+      "loss": 0.8804,
+      "step": 7269
+    },
+    {
+      "epoch": 1.437495314959646,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.298048994778508e-06,
+      "loss": 0.8828,
+      "step": 7270
+    },
+    {
+      "epoch": 1.4376952100147422,
+      "grad_norm": 2.125,
+      "learning_rate": 5.296997060710684e-06,
+      "loss": 0.8776,
+      "step": 7271
+    },
+    {
+      "epoch": 1.4378951050698383,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.29594511345028e-06,
+      "loss": 0.98,
+      "step": 7272
+    },
+    {
+      "epoch": 1.4380950001249344,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.294893153044027e-06,
+      "loss": 1.0607,
+      "step": 7273
+    },
+    {
+      "epoch": 1.4382948951800305,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.2938411795386516e-06,
+      "loss": 0.9884,
+      "step": 7274
+    },
+    {
+      "epoch": 1.4384947902351266,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.292789192980882e-06,
+      "loss": 1.0028,
+      "step": 7275
+    },
+    {
+      "epoch": 1.4386946852902227,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.291737193417448e-06,
+      "loss": 0.9676,
+      "step": 7276
+    },
+    {
+      "epoch": 1.4388945803453188,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.290685180895078e-06,
+      "loss": 0.9076,
+      "step": 7277
+    },
+    {
+      "epoch": 1.439094475400415,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.289633155460504e-06,
+      "loss": 0.9082,
+      "step": 7278
+    },
+    {
+      "epoch": 1.4392943704555108,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.288581117160457e-06,
+      "loss": 0.9861,
+      "step": 7279
+    },
+    {
+      "epoch": 1.4394942655106069,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.287529066041666e-06,
+      "loss": 0.9693,
+      "step": 7280
+    },
+    {
+      "epoch": 1.439694160565703,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.286477002150866e-06,
+      "loss": 0.9659,
+      "step": 7281
+    },
+    {
+      "epoch": 1.439894055620799,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.285424925534788e-06,
+      "loss": 0.8919,
+      "step": 7282
+    },
+    {
+      "epoch": 1.4400939506758952,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.284372836240166e-06,
+      "loss": 0.9426,
+      "step": 7283
+    },
+    {
+      "epoch": 1.4402938457309913,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.283320734313732e-06,
+      "loss": 0.9294,
+      "step": 7284
+    },
+    {
+      "epoch": 1.4404937407860874,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.282268619802221e-06,
+      "loss": 1.0199,
+      "step": 7285
+    },
+    {
+      "epoch": 1.4406936358411833,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.281216492752368e-06,
+      "loss": 1.0224,
+      "step": 7286
+    },
+    {
+      "epoch": 1.4408935308962794,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.280164353210908e-06,
+      "loss": 0.851,
+      "step": 7287
+    },
+    {
+      "epoch": 1.4410934259513755,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.279112201224579e-06,
+      "loss": 1.0081,
+      "step": 7288
+    },
+    {
+      "epoch": 1.4412933210064716,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.278060036840115e-06,
+      "loss": 0.9771,
+      "step": 7289
+    },
+    {
+      "epoch": 1.4414932160615677,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.2770078601042526e-06,
+      "loss": 0.9269,
+      "step": 7290
+    },
+    {
+      "epoch": 1.4416931111166638,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.275955671063733e-06,
+      "loss": 0.9333,
+      "step": 7291
+    },
+    {
+      "epoch": 1.4418930061717599,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.27490346976529e-06,
+      "loss": 0.9411,
+      "step": 7292
+    },
+    {
+      "epoch": 1.442092901226856,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.273851256255664e-06,
+      "loss": 1.0253,
+      "step": 7293
+    },
+    {
+      "epoch": 1.442292796281952,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.272799030581597e-06,
+      "loss": 0.947,
+      "step": 7294
+    },
+    {
+      "epoch": 1.4424926913370482,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.271746792789824e-06,
+      "loss": 0.9798,
+      "step": 7295
+    },
+    {
+      "epoch": 1.4426925863921443,
+      "grad_norm": 2.125,
+      "learning_rate": 5.270694542927089e-06,
+      "loss": 0.947,
+      "step": 7296
+    },
+    {
+      "epoch": 1.4428924814472401,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.269642281040129e-06,
+      "loss": 0.8852,
+      "step": 7297
+    },
+    {
+      "epoch": 1.4430923765023362,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.26859000717569e-06,
+      "loss": 0.9008,
+      "step": 7298
+    },
+    {
+      "epoch": 1.4432922715574323,
+      "grad_norm": 2.125,
+      "learning_rate": 5.26753772138051e-06,
+      "loss": 1.0346,
+      "step": 7299
+    },
+    {
+      "epoch": 1.4434921666125284,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.266485423701335e-06,
+      "loss": 1.0262,
+      "step": 7300
+    },
+    {
+      "epoch": 1.4436920616676245,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.265433114184903e-06,
+      "loss": 1.0945,
+      "step": 7301
+    },
+    {
+      "epoch": 1.4438919567227206,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.264380792877963e-06,
+      "loss": 0.9568,
+      "step": 7302
+    },
+    {
+      "epoch": 1.4440918517778165,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.263328459827256e-06,
+      "loss": 0.9913,
+      "step": 7303
+    },
+    {
+      "epoch": 1.4442917468329126,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.262276115079526e-06,
+      "loss": 1.01,
+      "step": 7304
+    },
+    {
+      "epoch": 1.4444916418880087,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.26122375868152e-06,
+      "loss": 0.9188,
+      "step": 7305
+    },
+    {
+      "epoch": 1.4446915369431048,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.260171390679981e-06,
+      "loss": 1.0598,
+      "step": 7306
+    },
+    {
+      "epoch": 1.444891431998201,
+      "grad_norm": 2.25,
+      "learning_rate": 5.2591190111216574e-06,
+      "loss": 0.9469,
+      "step": 7307
+    },
+    {
+      "epoch": 1.445091327053297,
+      "grad_norm": 2.125,
+      "learning_rate": 5.258066620053295e-06,
+      "loss": 0.918,
+      "step": 7308
+    },
+    {
+      "epoch": 1.4452912221083931,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.257014217521641e-06,
+      "loss": 1.0795,
+      "step": 7309
+    },
+    {
+      "epoch": 1.4454911171634892,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.255961803573444e-06,
+      "loss": 1.0236,
+      "step": 7310
+    },
+    {
+      "epoch": 1.4456910122185853,
+      "grad_norm": 2.25,
+      "learning_rate": 5.254909378255448e-06,
+      "loss": 1.0371,
+      "step": 7311
+    },
+    {
+      "epoch": 1.4458909072736814,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.2538569416144065e-06,
+      "loss": 0.8564,
+      "step": 7312
+    },
+    {
+      "epoch": 1.4460908023287775,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.252804493697068e-06,
+      "loss": 1.0691,
+      "step": 7313
+    },
+    {
+      "epoch": 1.4462906973838734,
+      "grad_norm": 2.25,
+      "learning_rate": 5.251752034550181e-06,
+      "loss": 1.0646,
+      "step": 7314
+    },
+    {
+      "epoch": 1.4464905924389695,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.250699564220493e-06,
+      "loss": 0.9689,
+      "step": 7315
+    },
+    {
+      "epoch": 1.4466904874940656,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.24964708275476e-06,
+      "loss": 1.0675,
+      "step": 7316
+    },
+    {
+      "epoch": 1.4468903825491617,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.248594590199727e-06,
+      "loss": 0.9403,
+      "step": 7317
+    },
+    {
+      "epoch": 1.4470902776042578,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.247542086602151e-06,
+      "loss": 0.8519,
+      "step": 7318
+    },
+    {
+      "epoch": 1.447290172659354,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.2464895720087816e-06,
+      "loss": 0.908,
+      "step": 7319
+    },
+    {
+      "epoch": 1.44749006771445,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.245437046466371e-06,
+      "loss": 0.9291,
+      "step": 7320
+    },
+    {
+      "epoch": 1.4476899627695459,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.244384510021673e-06,
+      "loss": 0.9848,
+      "step": 7321
+    },
+    {
+      "epoch": 1.447889857824642,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.243331962721443e-06,
+      "loss": 0.9379,
+      "step": 7322
+    },
+    {
+      "epoch": 1.448089752879738,
+      "grad_norm": 1.90625,
+      "learning_rate": 5.242279404612431e-06,
+      "loss": 0.8042,
+      "step": 7323
+    },
+    {
+      "epoch": 1.4482896479348342,
+      "grad_norm": 2.125,
+      "learning_rate": 5.241226835741395e-06,
+      "loss": 0.943,
+      "step": 7324
+    },
+    {
+      "epoch": 1.4484895429899303,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.240174256155088e-06,
+      "loss": 0.8862,
+      "step": 7325
+    },
+    {
+      "epoch": 1.4486894380450264,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.239121665900266e-06,
+      "loss": 0.9228,
+      "step": 7326
+    },
+    {
+      "epoch": 1.4488893331001225,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.238069065023686e-06,
+      "loss": 1.0576,
+      "step": 7327
+    },
+    {
+      "epoch": 1.4490892281552186,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.237016453572103e-06,
+      "loss": 0.8776,
+      "step": 7328
+    },
+    {
+      "epoch": 1.4492891232103147,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.235963831592275e-06,
+      "loss": 1.0057,
+      "step": 7329
+    },
+    {
+      "epoch": 1.4494890182654108,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.23491119913096e-06,
+      "loss": 1.0676,
+      "step": 7330
+    },
+    {
+      "epoch": 1.4496889133205069,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.233858556234913e-06,
+      "loss": 0.9775,
+      "step": 7331
+    },
+    {
+      "epoch": 1.4498888083756027,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.232805902950894e-06,
+      "loss": 1.0474,
+      "step": 7332
+    },
+    {
+      "epoch": 1.4500887034306988,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.231753239325662e-06,
+      "loss": 0.9413,
+      "step": 7333
+    },
+    {
+      "epoch": 1.450288598485795,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.230700565405975e-06,
+      "loss": 0.9317,
+      "step": 7334
+    },
+    {
+      "epoch": 1.450488493540891,
+      "grad_norm": 2.25,
+      "learning_rate": 5.229647881238594e-06,
+      "loss": 1.0024,
+      "step": 7335
+    },
+    {
+      "epoch": 1.4506883885959871,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.228595186870279e-06,
+      "loss": 0.9927,
+      "step": 7336
+    },
+    {
+      "epoch": 1.4508882836510832,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.227542482347789e-06,
+      "loss": 0.9571,
+      "step": 7337
+    },
+    {
+      "epoch": 1.4510881787061791,
+      "grad_norm": 2.125,
+      "learning_rate": 5.226489767717887e-06,
+      "loss": 0.9475,
+      "step": 7338
+    },
+    {
+      "epoch": 1.4512880737612752,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.225437043027335e-06,
+      "loss": 1.0594,
+      "step": 7339
+    },
+    {
+      "epoch": 1.4514879688163713,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.224384308322892e-06,
+      "loss": 0.9649,
+      "step": 7340
+    },
+    {
+      "epoch": 1.4516878638714674,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.2233315636513225e-06,
+      "loss": 1.0089,
+      "step": 7341
+    },
+    {
+      "epoch": 1.4518877589265635,
+      "grad_norm": 2.125,
+      "learning_rate": 5.222278809059389e-06,
+      "loss": 1.0211,
+      "step": 7342
+    },
+    {
+      "epoch": 1.4520876539816596,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.221226044593855e-06,
+      "loss": 0.8659,
+      "step": 7343
+    },
+    {
+      "epoch": 1.4522875490367557,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.220173270301483e-06,
+      "loss": 0.9456,
+      "step": 7344
+    },
+    {
+      "epoch": 1.4524874440918518,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.219120486229038e-06,
+      "loss": 0.9313,
+      "step": 7345
+    },
+    {
+      "epoch": 1.452687339146948,
+      "grad_norm": 2.125,
+      "learning_rate": 5.2180676924232846e-06,
+      "loss": 0.9382,
+      "step": 7346
+    },
+    {
+      "epoch": 1.452887234202044,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.217014888930988e-06,
+      "loss": 0.9936,
+      "step": 7347
+    },
+    {
+      "epoch": 1.4530871292571401,
+      "grad_norm": 2.125,
+      "learning_rate": 5.215962075798912e-06,
+      "loss": 1.0392,
+      "step": 7348
+    },
+    {
+      "epoch": 1.453287024312236,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.214909253073823e-06,
+      "loss": 1.0348,
+      "step": 7349
+    },
+    {
+      "epoch": 1.453486919367332,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.21385642080249e-06,
+      "loss": 0.8826,
+      "step": 7350
+    },
+    {
+      "epoch": 1.4536868144224282,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.212803579031676e-06,
+      "loss": 0.9775,
+      "step": 7351
+    },
+    {
+      "epoch": 1.4538867094775243,
+      "grad_norm": 1.96875,
+      "learning_rate": 5.21175072780815e-06,
+      "loss": 0.8595,
+      "step": 7352
+    },
+    {
+      "epoch": 1.4540866045326204,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.2106978671786815e-06,
+      "loss": 0.9054,
+      "step": 7353
+    },
+    {
+      "epoch": 1.4542864995877165,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.209644997190033e-06,
+      "loss": 0.9559,
+      "step": 7354
+    },
+    {
+      "epoch": 1.4544863946428126,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.2085921178889785e-06,
+      "loss": 0.9998,
+      "step": 7355
+    },
+    {
+      "epoch": 1.4546862896979085,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.207539229322284e-06,
+      "loss": 1.0,
+      "step": 7356
+    },
+    {
+      "epoch": 1.4548861847530046,
+      "grad_norm": 2.125,
+      "learning_rate": 5.206486331536718e-06,
+      "loss": 1.0222,
+      "step": 7357
+    },
+    {
+      "epoch": 1.4550860798081007,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.205433424579052e-06,
+      "loss": 0.9159,
+      "step": 7358
+    },
+    {
+      "epoch": 1.4552859748631968,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.204380508496055e-06,
+      "loss": 0.9389,
+      "step": 7359
+    },
+    {
+      "epoch": 1.4554858699182929,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.203327583334499e-06,
+      "loss": 1.0703,
+      "step": 7360
+    },
+    {
+      "epoch": 1.455685764973389,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.202274649141152e-06,
+      "loss": 0.9504,
+      "step": 7361
+    },
+    {
+      "epoch": 1.455885660028485,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.2012217059627865e-06,
+      "loss": 1.0219,
+      "step": 7362
+    },
+    {
+      "epoch": 1.4560855550835812,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.200168753846174e-06,
+      "loss": 0.9764,
+      "step": 7363
+    },
+    {
+      "epoch": 1.4562854501386773,
+      "grad_norm": 2.125,
+      "learning_rate": 5.199115792838089e-06,
+      "loss": 1.0029,
+      "step": 7364
+    },
+    {
+      "epoch": 1.4564853451937734,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.1980628229852995e-06,
+      "loss": 0.9554,
+      "step": 7365
+    },
+    {
+      "epoch": 1.4566852402488695,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.197009844334581e-06,
+      "loss": 0.9349,
+      "step": 7366
+    },
+    {
+      "epoch": 1.4568851353039654,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.195956856932707e-06,
+      "loss": 0.9816,
+      "step": 7367
+    },
+    {
+      "epoch": 1.4570850303590615,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.194903860826449e-06,
+      "loss": 1.0072,
+      "step": 7368
+    },
+    {
+      "epoch": 1.4572849254141576,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.193850856062585e-06,
+      "loss": 0.9783,
+      "step": 7369
+    },
+    {
+      "epoch": 1.4574848204692537,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.192797842687883e-06,
+      "loss": 0.8432,
+      "step": 7370
+    },
+    {
+      "epoch": 1.4576847155243498,
+      "grad_norm": 2.125,
+      "learning_rate": 5.191744820749124e-06,
+      "loss": 0.8715,
+      "step": 7371
+    },
+    {
+      "epoch": 1.4578846105794458,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.1906917902930795e-06,
+      "loss": 0.9895,
+      "step": 7372
+    },
+    {
+      "epoch": 1.4580845056345417,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.189638751366527e-06,
+      "loss": 0.9933,
+      "step": 7373
+    },
+    {
+      "epoch": 1.4582844006896378,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.1885857040162406e-06,
+      "loss": 1.0211,
+      "step": 7374
+    },
+    {
+      "epoch": 1.458484295744734,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.187532648288997e-06,
+      "loss": 0.9895,
+      "step": 7375
+    },
+    {
+      "epoch": 1.45868419079983,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.186479584231575e-06,
+      "loss": 1.005,
+      "step": 7376
+    },
+    {
+      "epoch": 1.4588840858549261,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.185426511890749e-06,
+      "loss": 1.0098,
+      "step": 7377
+    },
+    {
+      "epoch": 1.4590839809100222,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.184373431313297e-06,
+      "loss": 0.9266,
+      "step": 7378
+    },
+    {
+      "epoch": 1.4592838759651183,
+      "grad_norm": 1.96875,
+      "learning_rate": 5.183320342545995e-06,
+      "loss": 0.9569,
+      "step": 7379
+    },
+    {
+      "epoch": 1.4594837710202144,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.1822672456356245e-06,
+      "loss": 0.9165,
+      "step": 7380
+    },
+    {
+      "epoch": 1.4596836660753105,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.181214140628964e-06,
+      "loss": 0.9835,
+      "step": 7381
+    },
+    {
+      "epoch": 1.4598835611304066,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.180161027572788e-06,
+      "loss": 0.9744,
+      "step": 7382
+    },
+    {
+      "epoch": 1.4600834561855027,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.179107906513879e-06,
+      "loss": 0.974,
+      "step": 7383
+    },
+    {
+      "epoch": 1.4602833512405986,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.178054777499018e-06,
+      "loss": 0.9033,
+      "step": 7384
+    },
+    {
+      "epoch": 1.4604832462956947,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.177001640574981e-06,
+      "loss": 0.9548,
+      "step": 7385
+    },
+    {
+      "epoch": 1.4606831413507908,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.1759484957885505e-06,
+      "loss": 0.9575,
+      "step": 7386
+    },
+    {
+      "epoch": 1.460883036405887,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.1748953431865064e-06,
+      "loss": 1.0753,
+      "step": 7387
+    },
+    {
+      "epoch": 1.461082931460983,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.1738421828156295e-06,
+      "loss": 0.988,
+      "step": 7388
+    },
+    {
+      "epoch": 1.461282826516079,
+      "grad_norm": 2.125,
+      "learning_rate": 5.172789014722702e-06,
+      "loss": 0.9311,
+      "step": 7389
+    },
+    {
+      "epoch": 1.4614827215711752,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.171735838954505e-06,
+      "loss": 0.9836,
+      "step": 7390
+    },
+    {
+      "epoch": 1.461682616626271,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.17068265555782e-06,
+      "loss": 0.9558,
+      "step": 7391
+    },
+    {
+      "epoch": 1.4618825116813672,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.169629464579431e-06,
+      "loss": 0.9883,
+      "step": 7392
+    },
+    {
+      "epoch": 1.4620824067364633,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.168576266066118e-06,
+      "loss": 0.9813,
+      "step": 7393
+    },
+    {
+      "epoch": 1.4622823017915594,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.1675230600646645e-06,
+      "loss": 0.898,
+      "step": 7394
+    },
+    {
+      "epoch": 1.4624821968466555,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.166469846621856e-06,
+      "loss": 0.9725,
+      "step": 7395
+    },
+    {
+      "epoch": 1.4626820919017516,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.165416625784474e-06,
+      "loss": 0.9968,
+      "step": 7396
+    },
+    {
+      "epoch": 1.4628819869568477,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.164363397599302e-06,
+      "loss": 0.9762,
+      "step": 7397
+    },
+    {
+      "epoch": 1.4630818820119438,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.163310162113128e-06,
+      "loss": 1.0231,
+      "step": 7398
+    },
+    {
+      "epoch": 1.4632817770670399,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.162256919372732e-06,
+      "loss": 1.0243,
+      "step": 7399
+    },
+    {
+      "epoch": 1.463481672122136,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.1612036694249025e-06,
+      "loss": 0.9186,
+      "step": 7400
+    },
+    {
+      "epoch": 1.463681567177232,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.160150412316421e-06,
+      "loss": 0.9296,
+      "step": 7401
+    },
+    {
+      "epoch": 1.463881462232328,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.159097148094076e-06,
+      "loss": 0.8964,
+      "step": 7402
+    },
+    {
+      "epoch": 1.464081357287424,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.158043876804655e-06,
+      "loss": 1.0496,
+      "step": 7403
+    },
+    {
+      "epoch": 1.4642812523425202,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.15699059849494e-06,
+      "loss": 0.9805,
+      "step": 7404
+    },
+    {
+      "epoch": 1.4644811473976163,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.155937313211719e-06,
+      "loss": 0.8961,
+      "step": 7405
+    },
+    {
+      "epoch": 1.4646810424527124,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.15488402100178e-06,
+      "loss": 0.9394,
+      "step": 7406
+    },
+    {
+      "epoch": 1.4648809375078085,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.15383072191191e-06,
+      "loss": 0.9571,
+      "step": 7407
+    },
+    {
+      "epoch": 1.4650808325629046,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.152777415988894e-06,
+      "loss": 0.9525,
+      "step": 7408
+    },
+    {
+      "epoch": 1.4652807276180004,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.151724103279524e-06,
+      "loss": 0.9849,
+      "step": 7409
+    },
+    {
+      "epoch": 1.4654806226730965,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.150670783830584e-06,
+      "loss": 0.9647,
+      "step": 7410
+    },
+    {
+      "epoch": 1.4656805177281926,
+      "grad_norm": 2.375,
+      "learning_rate": 5.149617457688862e-06,
+      "loss": 0.9923,
+      "step": 7411
+    },
+    {
+      "epoch": 1.4658804127832887,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.148564124901152e-06,
+      "loss": 0.9533,
+      "step": 7412
+    },
+    {
+      "epoch": 1.4660803078383848,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.147510785514237e-06,
+      "loss": 0.8886,
+      "step": 7413
+    },
+    {
+      "epoch": 1.466280202893481,
+      "grad_norm": 2.25,
+      "learning_rate": 5.14645743957491e-06,
+      "loss": 1.0243,
+      "step": 7414
+    },
+    {
+      "epoch": 1.466480097948577,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.145404087129959e-06,
+      "loss": 1.0198,
+      "step": 7415
+    },
+    {
+      "epoch": 1.4666799930036731,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.144350728226173e-06,
+      "loss": 0.9834,
+      "step": 7416
+    },
+    {
+      "epoch": 1.4668798880587692,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.143297362910344e-06,
+      "loss": 1.0275,
+      "step": 7417
+    },
+    {
+      "epoch": 1.4670797831138653,
+      "grad_norm": 5.59375,
+      "learning_rate": 5.142243991229261e-06,
+      "loss": 0.9838,
+      "step": 7418
+    },
+    {
+      "epoch": 1.4672796781689614,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.141190613229714e-06,
+      "loss": 0.9877,
+      "step": 7419
+    },
+    {
+      "epoch": 1.4674795732240573,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.1401372289584975e-06,
+      "loss": 0.9014,
+      "step": 7420
+    },
+    {
+      "epoch": 1.4676794682791534,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.139083838462398e-06,
+      "loss": 0.9056,
+      "step": 7421
+    },
+    {
+      "epoch": 1.4678793633342495,
+      "grad_norm": 2.25,
+      "learning_rate": 5.138030441788211e-06,
+      "loss": 0.9881,
+      "step": 7422
+    },
+    {
+      "epoch": 1.4680792583893456,
+      "grad_norm": 2.125,
+      "learning_rate": 5.136977038982727e-06,
+      "loss": 0.8494,
+      "step": 7423
+    },
+    {
+      "epoch": 1.4682791534444417,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.135923630092736e-06,
+      "loss": 1.0048,
+      "step": 7424
+    },
+    {
+      "epoch": 1.4684790484995378,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.134870215165032e-06,
+      "loss": 1.0214,
+      "step": 7425
+    },
+    {
+      "epoch": 1.4686789435546337,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.133816794246408e-06,
+      "loss": 0.9724,
+      "step": 7426
+    },
+    {
+      "epoch": 1.4688788386097298,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.132763367383655e-06,
+      "loss": 1.0002,
+      "step": 7427
+    },
+    {
+      "epoch": 1.4690787336648259,
+      "grad_norm": 2.125,
+      "learning_rate": 5.131709934623568e-06,
+      "loss": 0.9988,
+      "step": 7428
+    },
+    {
+      "epoch": 1.469278628719922,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.130656496012941e-06,
+      "loss": 0.8107,
+      "step": 7429
+    },
+    {
+      "epoch": 1.469478523775018,
+      "grad_norm": 2.125,
+      "learning_rate": 5.129603051598565e-06,
+      "loss": 1.0389,
+      "step": 7430
+    },
+    {
+      "epoch": 1.4696784188301142,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.128549601427237e-06,
+      "loss": 1.0429,
+      "step": 7431
+    },
+    {
+      "epoch": 1.4698783138852103,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.127496145545747e-06,
+      "loss": 1.1101,
+      "step": 7432
+    },
+    {
+      "epoch": 1.4700782089403064,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.126442684000893e-06,
+      "loss": 0.9468,
+      "step": 7433
+    },
+    {
+      "epoch": 1.4702781039954025,
+      "grad_norm": 2.125,
+      "learning_rate": 5.125389216839469e-06,
+      "loss": 0.9814,
+      "step": 7434
+    },
+    {
+      "epoch": 1.4704779990504986,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.124335744108269e-06,
+      "loss": 0.9914,
+      "step": 7435
+    },
+    {
+      "epoch": 1.4706778941055947,
+      "grad_norm": 2.25,
+      "learning_rate": 5.123282265854088e-06,
+      "loss": 0.9354,
+      "step": 7436
+    },
+    {
+      "epoch": 1.4708777891606906,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.122228782123723e-06,
+      "loss": 1.0049,
+      "step": 7437
+    },
+    {
+      "epoch": 1.4710776842157867,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.121175292963968e-06,
+      "loss": 0.9465,
+      "step": 7438
+    },
+    {
+      "epoch": 1.4712775792708828,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.120121798421619e-06,
+      "loss": 0.9718,
+      "step": 7439
+    },
+    {
+      "epoch": 1.4714774743259789,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.119068298543474e-06,
+      "loss": 0.9102,
+      "step": 7440
+    },
+    {
+      "epoch": 1.471677369381075,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.1180147933763255e-06,
+      "loss": 0.9933,
+      "step": 7441
+    },
+    {
+      "epoch": 1.471877264436171,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.116961282966975e-06,
+      "loss": 0.9802,
+      "step": 7442
+    },
+    {
+      "epoch": 1.4720771594912672,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.1159077673622145e-06,
+      "loss": 0.9478,
+      "step": 7443
+    },
+    {
+      "epoch": 1.472277054546363,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.114854246608845e-06,
+      "loss": 1.0779,
+      "step": 7444
+    },
+    {
+      "epoch": 1.4724769496014591,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.113800720753662e-06,
+      "loss": 1.0184,
+      "step": 7445
+    },
+    {
+      "epoch": 1.4726768446565552,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.112747189843462e-06,
+      "loss": 1.0157,
+      "step": 7446
+    },
+    {
+      "epoch": 1.4728767397116513,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.111693653925045e-06,
+      "loss": 1.0297,
+      "step": 7447
+    },
+    {
+      "epoch": 1.4730766347667474,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.110640113045208e-06,
+      "loss": 0.9191,
+      "step": 7448
+    },
+    {
+      "epoch": 1.4732765298218435,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.109586567250747e-06,
+      "loss": 0.9916,
+      "step": 7449
+    },
+    {
+      "epoch": 1.4734764248769396,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.108533016588464e-06,
+      "loss": 0.9522,
+      "step": 7450
+    },
+    {
+      "epoch": 1.4736763199320357,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.107479461105156e-06,
+      "loss": 0.9472,
+      "step": 7451
+    },
+    {
+      "epoch": 1.4738762149871318,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.106425900847621e-06,
+      "loss": 0.9726,
+      "step": 7452
+    },
+    {
+      "epoch": 1.474076110042228,
+      "grad_norm": 2.125,
+      "learning_rate": 5.10537233586266e-06,
+      "loss": 0.9501,
+      "step": 7453
+    },
+    {
+      "epoch": 1.474276005097324,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.104318766197072e-06,
+      "loss": 0.9535,
+      "step": 7454
+    },
+    {
+      "epoch": 1.47447590015242,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.103265191897654e-06,
+      "loss": 0.9887,
+      "step": 7455
+    },
+    {
+      "epoch": 1.474675795207516,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.102211613011207e-06,
+      "loss": 0.964,
+      "step": 7456
+    },
+    {
+      "epoch": 1.4748756902626121,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.101158029584531e-06,
+      "loss": 0.9412,
+      "step": 7457
+    },
+    {
+      "epoch": 1.4750755853177082,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.100104441664426e-06,
+      "loss": 1.0356,
+      "step": 7458
+    },
+    {
+      "epoch": 1.4752754803728043,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.099050849297694e-06,
+      "loss": 0.9289,
+      "step": 7459
+    },
+    {
+      "epoch": 1.4754753754279004,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.097997252531133e-06,
+      "loss": 0.9428,
+      "step": 7460
+    },
+    {
+      "epoch": 1.4756752704829963,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.096943651411543e-06,
+      "loss": 0.98,
+      "step": 7461
+    },
+    {
+      "epoch": 1.4758751655380924,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.095890045985729e-06,
+      "loss": 0.9712,
+      "step": 7462
+    },
+    {
+      "epoch": 1.4760750605931885,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.094836436300486e-06,
+      "loss": 0.9804,
+      "step": 7463
+    },
+    {
+      "epoch": 1.4762749556482846,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.093782822402621e-06,
+      "loss": 0.9727,
+      "step": 7464
+    },
+    {
+      "epoch": 1.4764748507033807,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.0927292043389335e-06,
+      "loss": 1.0436,
+      "step": 7465
+    },
+    {
+      "epoch": 1.4766747457584768,
+      "grad_norm": 2.125,
+      "learning_rate": 5.091675582156224e-06,
+      "loss": 1.0495,
+      "step": 7466
+    },
+    {
+      "epoch": 1.476874640813573,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.090621955901294e-06,
+      "loss": 0.9945,
+      "step": 7467
+    },
+    {
+      "epoch": 1.477074535868669,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.0895683256209485e-06,
+      "loss": 0.9012,
+      "step": 7468
+    },
+    {
+      "epoch": 1.477274430923765,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.088514691361985e-06,
+      "loss": 0.8838,
+      "step": 7469
+    },
+    {
+      "epoch": 1.4774743259788612,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.08746105317121e-06,
+      "loss": 0.9647,
+      "step": 7470
+    },
+    {
+      "epoch": 1.4776742210339573,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.086407411095424e-06,
+      "loss": 1.0347,
+      "step": 7471
+    },
+    {
+      "epoch": 1.4778741160890532,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.08535376518143e-06,
+      "loss": 0.9503,
+      "step": 7472
+    },
+    {
+      "epoch": 1.4780740111441493,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.084300115476031e-06,
+      "loss": 1.0924,
+      "step": 7473
+    },
+    {
+      "epoch": 1.4782739061992454,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.083246462026029e-06,
+      "loss": 0.9997,
+      "step": 7474
+    },
+    {
+      "epoch": 1.4784738012543415,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.082192804878229e-06,
+      "loss": 0.9579,
+      "step": 7475
+    },
+    {
+      "epoch": 1.4786736963094376,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.081139144079433e-06,
+      "loss": 1.0236,
+      "step": 7476
+    },
+    {
+      "epoch": 1.4788735913645337,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.080085479676445e-06,
+      "loss": 0.9185,
+      "step": 7477
+    },
+    {
+      "epoch": 1.4790734864196298,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.079031811716068e-06,
+      "loss": 0.9896,
+      "step": 7478
+    },
+    {
+      "epoch": 1.4792733814747256,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.077978140245108e-06,
+      "loss": 1.0319,
+      "step": 7479
+    },
+    {
+      "epoch": 1.4794732765298217,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.076924465310365e-06,
+      "loss": 0.9257,
+      "step": 7480
+    },
+    {
+      "epoch": 1.4796731715849178,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.075870786958646e-06,
+      "loss": 0.8942,
+      "step": 7481
+    },
+    {
+      "epoch": 1.479873066640014,
+      "grad_norm": 2.125,
+      "learning_rate": 5.0748171052367565e-06,
+      "loss": 0.9094,
+      "step": 7482
+    },
+    {
+      "epoch": 1.48007296169511,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.073763420191498e-06,
+      "loss": 0.9146,
+      "step": 7483
+    },
+    {
+      "epoch": 1.4802728567502061,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.072709731869677e-06,
+      "loss": 0.9358,
+      "step": 7484
+    },
+    {
+      "epoch": 1.4804727518053022,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.071656040318098e-06,
+      "loss": 0.9649,
+      "step": 7485
+    },
+    {
+      "epoch": 1.4806726468603983,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.070602345583564e-06,
+      "loss": 0.9764,
+      "step": 7486
+    },
+    {
+      "epoch": 1.4808725419154944,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.069548647712881e-06,
+      "loss": 0.9947,
+      "step": 7487
+    },
+    {
+      "epoch": 1.4810724369705905,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.0684949467528555e-06,
+      "loss": 1.0016,
+      "step": 7488
+    },
+    {
+      "epoch": 1.4812723320256866,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.06744124275029e-06,
+      "loss": 1.041,
+      "step": 7489
+    },
+    {
+      "epoch": 1.4814722270807825,
+      "grad_norm": 2.125,
+      "learning_rate": 5.066387535751995e-06,
+      "loss": 0.9154,
+      "step": 7490
+    },
+    {
+      "epoch": 1.4816721221358786,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.06533382580477e-06,
+      "loss": 1.0142,
+      "step": 7491
+    },
+    {
+      "epoch": 1.4818720171909747,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.064280112955424e-06,
+      "loss": 1.034,
+      "step": 7492
+    },
+    {
+      "epoch": 1.4820719122460708,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.063226397250763e-06,
+      "loss": 0.9688,
+      "step": 7493
+    },
+    {
+      "epoch": 1.482271807301167,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.062172678737591e-06,
+      "loss": 1.0058,
+      "step": 7494
+    },
+    {
+      "epoch": 1.482471702356263,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.061118957462716e-06,
+      "loss": 0.992,
+      "step": 7495
+    },
+    {
+      "epoch": 1.482671597411359,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.060065233472944e-06,
+      "loss": 1.0407,
+      "step": 7496
+    },
+    {
+      "epoch": 1.482871492466455,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.059011506815079e-06,
+      "loss": 1.036,
+      "step": 7497
+    },
+    {
+      "epoch": 1.483071387521551,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.0579577775359314e-06,
+      "loss": 0.8635,
+      "step": 7498
+    },
+    {
+      "epoch": 1.4832712825766472,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.0569040456823045e-06,
+      "loss": 0.9879,
+      "step": 7499
+    },
+    {
+      "epoch": 1.4834711776317433,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.055850311301006e-06,
+      "loss": 0.9769,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4834711776317433,
+      "eval_loss": 0.9009870290756226,
+      "eval_runtime": 593.6995,
+      "eval_samples_per_second": 3.601,
+      "eval_steps_per_second": 3.601,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4836710726868394,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.054796574438842e-06,
+      "loss": 1.067,
+      "step": 7501
+    },
+    {
+      "epoch": 1.4838709677419355,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.053742835142621e-06,
+      "loss": 0.8663,
+      "step": 7502
+    },
+    {
+      "epoch": 1.4840708627970316,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.052689093459148e-06,
+      "loss": 0.9994,
+      "step": 7503
+    },
+    {
+      "epoch": 1.4842707578521277,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.051635349435232e-06,
+      "loss": 1.0215,
+      "step": 7504
+    },
+    {
+      "epoch": 1.4844706529072238,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.050581603117679e-06,
+      "loss": 0.9514,
+      "step": 7505
+    },
+    {
+      "epoch": 1.48467054796232,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.049527854553296e-06,
+      "loss": 0.9356,
+      "step": 7506
+    },
+    {
+      "epoch": 1.4848704430174158,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.0484741037888915e-06,
+      "loss": 0.996,
+      "step": 7507
+    },
+    {
+      "epoch": 1.4850703380725119,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.047420350871273e-06,
+      "loss": 1.0457,
+      "step": 7508
+    },
+    {
+      "epoch": 1.485270233127608,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.046366595847246e-06,
+      "loss": 0.9795,
+      "step": 7509
+    },
+    {
+      "epoch": 1.485470128182704,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.045312838763622e-06,
+      "loss": 0.9604,
+      "step": 7510
+    },
+    {
+      "epoch": 1.4856700232378002,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.044259079667204e-06,
+      "loss": 0.9675,
+      "step": 7511
+    },
+    {
+      "epoch": 1.4858699182928963,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.043205318604804e-06,
+      "loss": 0.9365,
+      "step": 7512
+    },
+    {
+      "epoch": 1.4860698133479924,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.0421515556232285e-06,
+      "loss": 0.9545,
+      "step": 7513
+    },
+    {
+      "epoch": 1.4862697084030883,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.041097790769285e-06,
+      "loss": 1.0955,
+      "step": 7514
+    },
+    {
+      "epoch": 1.4864696034581844,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.040044024089782e-06,
+      "loss": 0.9641,
+      "step": 7515
+    },
+    {
+      "epoch": 1.4866694985132805,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.0389902556315286e-06,
+      "loss": 1.0313,
+      "step": 7516
+    },
+    {
+      "epoch": 1.4868693935683766,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.037936485441332e-06,
+      "loss": 0.9403,
+      "step": 7517
+    },
+    {
+      "epoch": 1.4870692886234727,
+      "grad_norm": 2.125,
+      "learning_rate": 5.0368827135659996e-06,
+      "loss": 1.0066,
+      "step": 7518
+    },
+    {
+      "epoch": 1.4872691836785688,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.035828940052342e-06,
+      "loss": 1.1098,
+      "step": 7519
+    },
+    {
+      "epoch": 1.4874690787336649,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.0347751649471655e-06,
+      "loss": 1.0444,
+      "step": 7520
+    },
+    {
+      "epoch": 1.487668973788761,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.033721388297281e-06,
+      "loss": 0.91,
+      "step": 7521
+    },
+    {
+      "epoch": 1.487868868843857,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.0326676101494955e-06,
+      "loss": 0.917,
+      "step": 7522
+    },
+    {
+      "epoch": 1.4880687638989532,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.031613830550618e-06,
+      "loss": 0.9712,
+      "step": 7523
+    },
+    {
+      "epoch": 1.4882686589540493,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.0305600495474586e-06,
+      "loss": 0.9923,
+      "step": 7524
+    },
+    {
+      "epoch": 1.4884685540091451,
+      "grad_norm": 2.125,
+      "learning_rate": 5.029506267186823e-06,
+      "loss": 0.9984,
+      "step": 7525
+    },
+    {
+      "epoch": 1.4886684490642412,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.028452483515524e-06,
+      "loss": 1.0691,
+      "step": 7526
+    },
+    {
+      "epoch": 1.4888683441193373,
+      "grad_norm": 2.125,
+      "learning_rate": 5.027398698580369e-06,
+      "loss": 0.9605,
+      "step": 7527
+    },
+    {
+      "epoch": 1.4890682391744334,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.026344912428166e-06,
+      "loss": 0.9946,
+      "step": 7528
+    },
+    {
+      "epoch": 1.4892681342295295,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.025291125105725e-06,
+      "loss": 0.9633,
+      "step": 7529
+    },
+    {
+      "epoch": 1.4894680292846256,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.024237336659855e-06,
+      "loss": 0.9967,
+      "step": 7530
+    },
+    {
+      "epoch": 1.4896679243397217,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.023183547137364e-06,
+      "loss": 1.0632,
+      "step": 7531
+    },
+    {
+      "epoch": 1.4898678193948176,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.022129756585063e-06,
+      "loss": 0.9735,
+      "step": 7532
+    },
+    {
+      "epoch": 1.4900677144499137,
+      "grad_norm": 2.25,
+      "learning_rate": 5.021075965049762e-06,
+      "loss": 1.0718,
+      "step": 7533
+    },
+    {
+      "epoch": 1.4902676095050098,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.0200221725782675e-06,
+      "loss": 1.0503,
+      "step": 7534
+    },
+    {
+      "epoch": 1.490467504560106,
+      "grad_norm": 1.9609375,
+      "learning_rate": 5.0189683792173905e-06,
+      "loss": 0.9196,
+      "step": 7535
+    },
+    {
+      "epoch": 1.490667399615202,
+      "grad_norm": 2.25,
+      "learning_rate": 5.01791458501394e-06,
+      "loss": 0.9358,
+      "step": 7536
+    },
+    {
+      "epoch": 1.490867294670298,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.0168607900147244e-06,
+      "loss": 0.9498,
+      "step": 7537
+    },
+    {
+      "epoch": 1.4910671897253942,
+      "grad_norm": 2.125,
+      "learning_rate": 5.015806994266556e-06,
+      "loss": 0.9967,
+      "step": 7538
+    },
+    {
+      "epoch": 1.4912670847804903,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.014753197816243e-06,
+      "loss": 0.9562,
+      "step": 7539
+    },
+    {
+      "epoch": 1.4914669798355864,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.013699400710592e-06,
+      "loss": 1.0169,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4916668748906825,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.0126456029964175e-06,
+      "loss": 1.0031,
+      "step": 7541
+    },
+    {
+      "epoch": 1.4918667699457786,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.011591804720526e-06,
+      "loss": 0.9976,
+      "step": 7542
+    },
+    {
+      "epoch": 1.4920666650008745,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.010538005929728e-06,
+      "loss": 0.9954,
+      "step": 7543
+    },
+    {
+      "epoch": 1.4922665600559706,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.009484206670831e-06,
+      "loss": 1.028,
+      "step": 7544
+    },
+    {
+      "epoch": 1.4924664551110667,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.008430406990649e-06,
+      "loss": 0.9847,
+      "step": 7545
+    },
+    {
+      "epoch": 1.4926663501661628,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.007376606935989e-06,
+      "loss": 1.0274,
+      "step": 7546
+    },
+    {
+      "epoch": 1.4928662452212589,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.0063228065536595e-06,
+      "loss": 0.9803,
+      "step": 7547
+    },
+    {
+      "epoch": 1.493066140276355,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.005269005890473e-06,
+      "loss": 0.9394,
+      "step": 7548
+    },
+    {
+      "epoch": 1.4932660353314509,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.004215204993237e-06,
+      "loss": 0.9289,
+      "step": 7549
+    },
+    {
+      "epoch": 1.493465930386547,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.003161403908762e-06,
+      "loss": 0.9714,
+      "step": 7550
+    },
+    {
+      "epoch": 1.493665825441643,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.002107602683858e-06,
+      "loss": 0.8793,
+      "step": 7551
+    },
+    {
+      "epoch": 1.4938657204967392,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.0010538013653345e-06,
+      "loss": 1.0672,
+      "step": 7552
+    },
+    {
+      "epoch": 1.4940656155518353,
+      "grad_norm": 2.28125,
+      "learning_rate": 5e-06,
+      "loss": 1.0445,
+      "step": 7553
+    },
+    {
+      "epoch": 1.4942655106069314,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.998946198634667e-06,
+      "loss": 0.9237,
+      "step": 7554
+    },
+    {
+      "epoch": 1.4944654056620275,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.997892397316144e-06,
+      "loss": 0.9672,
+      "step": 7555
+    },
+    {
+      "epoch": 1.4946653007171236,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.99683859609124e-06,
+      "loss": 1.0201,
+      "step": 7556
+    },
+    {
+      "epoch": 1.4948651957722197,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.995784795006764e-06,
+      "loss": 0.9943,
+      "step": 7557
+    },
+    {
+      "epoch": 1.4950650908273158,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.994730994109529e-06,
+      "loss": 0.9561,
+      "step": 7558
+    },
+    {
+      "epoch": 1.4952649858824119,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.993677193446341e-06,
+      "loss": 0.99,
+      "step": 7559
+    },
+    {
+      "epoch": 1.4954648809375077,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.9926233930640136e-06,
+      "loss": 0.9834,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4956647759926038,
+      "grad_norm": 2.125,
+      "learning_rate": 4.991569593009352e-06,
+      "loss": 0.9143,
+      "step": 7561
+    },
+    {
+      "epoch": 1.4958646710477,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.990515793329169e-06,
+      "loss": 1.0046,
+      "step": 7562
+    },
+    {
+      "epoch": 1.496064566102796,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.9894619940702735e-06,
+      "loss": 0.946,
+      "step": 7563
+    },
+    {
+      "epoch": 1.4962644611578921,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.988408195279475e-06,
+      "loss": 1.0447,
+      "step": 7564
+    },
+    {
+      "epoch": 1.4964643562129882,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.9873543970035825e-06,
+      "loss": 1.0406,
+      "step": 7565
+    },
+    {
+      "epoch": 1.4966642512680843,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.9863005992894085e-06,
+      "loss": 1.0905,
+      "step": 7566
+    },
+    {
+      "epoch": 1.4968641463231802,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.985246802183758e-06,
+      "loss": 1.0414,
+      "step": 7567
+    },
+    {
+      "epoch": 1.4970640413782763,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.9841930057334446e-06,
+      "loss": 0.9065,
+      "step": 7568
+    },
+    {
+      "epoch": 1.4972639364333724,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.983139209985276e-06,
+      "loss": 0.9596,
+      "step": 7569
+    },
+    {
+      "epoch": 1.4974638314884685,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.982085414986062e-06,
+      "loss": 1.0192,
+      "step": 7570
+    },
+    {
+      "epoch": 1.4976637265435646,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.981031620782612e-06,
+      "loss": 0.9386,
+      "step": 7571
+    },
+    {
+      "epoch": 1.4978636215986607,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.979977827421735e-06,
+      "loss": 0.899,
+      "step": 7572
+    },
+    {
+      "epoch": 1.4980635166537568,
+      "grad_norm": 2.25,
+      "learning_rate": 4.97892403495024e-06,
+      "loss": 1.084,
+      "step": 7573
+    },
+    {
+      "epoch": 1.498263411708853,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.977870243414938e-06,
+      "loss": 0.9711,
+      "step": 7574
+    },
+    {
+      "epoch": 1.498463306763949,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.976816452862638e-06,
+      "loss": 0.978,
+      "step": 7575
+    },
+    {
+      "epoch": 1.498663201819045,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.975762663340148e-06,
+      "loss": 1.0126,
+      "step": 7576
+    },
+    {
+      "epoch": 1.4988630968741412,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.974708874894276e-06,
+      "loss": 1.0988,
+      "step": 7577
+    },
+    {
+      "epoch": 1.499062991929237,
+      "grad_norm": 2.25,
+      "learning_rate": 4.973655087571836e-06,
+      "loss": 0.9493,
+      "step": 7578
+    },
+    {
+      "epoch": 1.4992628869843332,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.972601301419632e-06,
+      "loss": 0.9735,
+      "step": 7579
+    },
+    {
+      "epoch": 1.4994627820394293,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.971547516484477e-06,
+      "loss": 0.9846,
+      "step": 7580
+    },
+    {
+      "epoch": 1.4996626770945254,
+      "grad_norm": 2.125,
+      "learning_rate": 4.970493732813178e-06,
+      "loss": 0.9615,
+      "step": 7581
+    },
+    {
+      "epoch": 1.4998625721496215,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.969439950452543e-06,
+      "loss": 1.1397,
+      "step": 7582
+    },
+    {
+      "epoch": 1.5000624672047174,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.968386169449384e-06,
+      "loss": 0.8919,
+      "step": 7583
+    },
+    {
+      "epoch": 1.5002623622598135,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.967332389850506e-06,
+      "loss": 0.9603,
+      "step": 7584
+    },
+    {
+      "epoch": 1.5004622573149096,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.96627861170272e-06,
+      "loss": 0.8555,
+      "step": 7585
+    },
+    {
+      "epoch": 1.5006621523700057,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.965224835052836e-06,
+      "loss": 0.9833,
+      "step": 7586
+    },
+    {
+      "epoch": 1.5008620474251018,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.96417105994766e-06,
+      "loss": 0.9636,
+      "step": 7587
+    },
+    {
+      "epoch": 1.5010619424801979,
+      "grad_norm": 2.125,
+      "learning_rate": 4.963117286434003e-06,
+      "loss": 0.9915,
+      "step": 7588
+    },
+    {
+      "epoch": 1.501261837535294,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.962063514558671e-06,
+      "loss": 0.9548,
+      "step": 7589
+    },
+    {
+      "epoch": 1.50146173259039,
+      "grad_norm": 1.984375,
+      "learning_rate": 4.961009744368472e-06,
+      "loss": 0.8947,
+      "step": 7590
+    },
+    {
+      "epoch": 1.5016616276454862,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.959955975910219e-06,
+      "loss": 0.9332,
+      "step": 7591
+    },
+    {
+      "epoch": 1.5018615227005823,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.958902209230716e-06,
+      "loss": 1.0048,
+      "step": 7592
+    },
+    {
+      "epoch": 1.5020614177556784,
+      "grad_norm": 2.25,
+      "learning_rate": 4.9578484443767714e-06,
+      "loss": 0.9865,
+      "step": 7593
+    },
+    {
+      "epoch": 1.5022613128107745,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.956794681395197e-06,
+      "loss": 0.954,
+      "step": 7594
+    },
+    {
+      "epoch": 1.5024612078658706,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.955740920332796e-06,
+      "loss": 0.9588,
+      "step": 7595
+    },
+    {
+      "epoch": 1.5026611029209664,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.954687161236379e-06,
+      "loss": 0.9834,
+      "step": 7596
+    },
+    {
+      "epoch": 1.5028609979760625,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.953633404152754e-06,
+      "loss": 0.9794,
+      "step": 7597
+    },
+    {
+      "epoch": 1.5030608930311586,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.952579649128728e-06,
+      "loss": 0.9282,
+      "step": 7598
+    },
+    {
+      "epoch": 1.5032607880862547,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.951525896211109e-06,
+      "loss": 0.9748,
+      "step": 7599
+    },
+    {
+      "epoch": 1.5034606831413508,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.950472145446706e-06,
+      "loss": 0.992,
+      "step": 7600
+    },
+    {
+      "epoch": 1.5036605781964467,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.9494183968823235e-06,
+      "loss": 0.9436,
+      "step": 7601
+    },
+    {
+      "epoch": 1.5038604732515428,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.9483646505647705e-06,
+      "loss": 0.9967,
+      "step": 7602
+    },
+    {
+      "epoch": 1.504060368306639,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.947310906540854e-06,
+      "loss": 0.9897,
+      "step": 7603
+    },
+    {
+      "epoch": 1.504260263361735,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.946257164857382e-06,
+      "loss": 1.0052,
+      "step": 7604
+    },
+    {
+      "epoch": 1.5044601584168311,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.94520342556116e-06,
+      "loss": 0.9114,
+      "step": 7605
+    },
+    {
+      "epoch": 1.5046600534719272,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.944149688698997e-06,
+      "loss": 0.9928,
+      "step": 7606
+    },
+    {
+      "epoch": 1.5048599485270233,
+      "grad_norm": 2.25,
+      "learning_rate": 4.943095954317698e-06,
+      "loss": 0.91,
+      "step": 7607
+    },
+    {
+      "epoch": 1.5050598435821194,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.94204222246407e-06,
+      "loss": 0.9799,
+      "step": 7608
+    },
+    {
+      "epoch": 1.5052597386372155,
+      "grad_norm": 2.125,
+      "learning_rate": 4.940988493184922e-06,
+      "loss": 0.8856,
+      "step": 7609
+    },
+    {
+      "epoch": 1.5054596336923116,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.939934766527058e-06,
+      "loss": 1.062,
+      "step": 7610
+    },
+    {
+      "epoch": 1.5056595287474077,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.938881042537286e-06,
+      "loss": 0.9589,
+      "step": 7611
+    },
+    {
+      "epoch": 1.5058594238025038,
+      "grad_norm": 2.125,
+      "learning_rate": 4.93782732126241e-06,
+      "loss": 0.982,
+      "step": 7612
+    },
+    {
+      "epoch": 1.5060593188576,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.936773602749238e-06,
+      "loss": 1.015,
+      "step": 7613
+    },
+    {
+      "epoch": 1.5062592139126958,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.9357198870445775e-06,
+      "loss": 0.98,
+      "step": 7614
+    },
+    {
+      "epoch": 1.506459108967792,
+      "grad_norm": 2.125,
+      "learning_rate": 4.934666174195231e-06,
+      "loss": 0.9862,
+      "step": 7615
+    },
+    {
+      "epoch": 1.506659004022888,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.933612464248006e-06,
+      "loss": 1.0111,
+      "step": 7616
+    },
+    {
+      "epoch": 1.506858899077984,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.932558757249711e-06,
+      "loss": 0.9228,
+      "step": 7617
+    },
+    {
+      "epoch": 1.50705879413308,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.931505053247146e-06,
+      "loss": 1.0451,
+      "step": 7618
+    },
+    {
+      "epoch": 1.507258689188176,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.930451352287121e-06,
+      "loss": 1.0573,
+      "step": 7619
+    },
+    {
+      "epoch": 1.5074585842432722,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.929397654416438e-06,
+      "loss": 0.9944,
+      "step": 7620
+    },
+    {
+      "epoch": 1.5076584792983683,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.928343959681904e-06,
+      "loss": 0.9962,
+      "step": 7621
+    },
+    {
+      "epoch": 1.5078583743534644,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.927290268130325e-06,
+      "loss": 0.8596,
+      "step": 7622
+    },
+    {
+      "epoch": 1.5080582694085605,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.9262365798085024e-06,
+      "loss": 1.0626,
+      "step": 7623
+    },
+    {
+      "epoch": 1.5082581644636566,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.925182894763243e-06,
+      "loss": 0.9334,
+      "step": 7624
+    },
+    {
+      "epoch": 1.5084580595187527,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.924129213041353e-06,
+      "loss": 1.0244,
+      "step": 7625
+    },
+    {
+      "epoch": 1.5086579545738488,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.923075534689635e-06,
+      "loss": 1.0285,
+      "step": 7626
+    },
+    {
+      "epoch": 1.5088578496289449,
+      "grad_norm": 2.0,
+      "learning_rate": 4.922021859754894e-06,
+      "loss": 0.9209,
+      "step": 7627
+    },
+    {
+      "epoch": 1.509057744684041,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.920968188283932e-06,
+      "loss": 1.0255,
+      "step": 7628
+    },
+    {
+      "epoch": 1.509257639739137,
+      "grad_norm": 2.125,
+      "learning_rate": 4.919914520323555e-06,
+      "loss": 0.8351,
+      "step": 7629
+    },
+    {
+      "epoch": 1.5094575347942332,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.918860855920568e-06,
+      "loss": 0.9118,
+      "step": 7630
+    },
+    {
+      "epoch": 1.5096574298493293,
+      "grad_norm": 2.125,
+      "learning_rate": 4.917807195121773e-06,
+      "loss": 0.983,
+      "step": 7631
+    },
+    {
+      "epoch": 1.5098573249044251,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.916753537973973e-06,
+      "loss": 0.991,
+      "step": 7632
+    },
+    {
+      "epoch": 1.5100572199595212,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.915699884523971e-06,
+      "loss": 1.0175,
+      "step": 7633
+    },
+    {
+      "epoch": 1.5102571150146173,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.914646234818573e-06,
+      "loss": 1.0203,
+      "step": 7634
+    },
+    {
+      "epoch": 1.5104570100697134,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.913592588904578e-06,
+      "loss": 0.8762,
+      "step": 7635
+    },
+    {
+      "epoch": 1.5106569051248093,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.912538946828792e-06,
+      "loss": 0.9023,
+      "step": 7636
+    },
+    {
+      "epoch": 1.5108568001799054,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.9114853086380165e-06,
+      "loss": 0.9672,
+      "step": 7637
+    },
+    {
+      "epoch": 1.5110566952350015,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.910431674379054e-06,
+      "loss": 0.9385,
+      "step": 7638
+    },
+    {
+      "epoch": 1.5112565902900976,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.909378044098708e-06,
+      "loss": 0.8887,
+      "step": 7639
+    },
+    {
+      "epoch": 1.5114564853451937,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.908324417843779e-06,
+      "loss": 0.9445,
+      "step": 7640
+    },
+    {
+      "epoch": 1.5116563804002898,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.907270795661068e-06,
+      "loss": 1.0104,
+      "step": 7641
+    },
+    {
+      "epoch": 1.511856275455386,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.906217177597381e-06,
+      "loss": 1.0569,
+      "step": 7642
+    },
+    {
+      "epoch": 1.512056170510482,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.905163563699515e-06,
+      "loss": 1.0554,
+      "step": 7643
+    },
+    {
+      "epoch": 1.5122560655655781,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.904109954014273e-06,
+      "loss": 1.0148,
+      "step": 7644
+    },
+    {
+      "epoch": 1.5124559606206742,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.9030563485884585e-06,
+      "loss": 0.8805,
+      "step": 7645
+    },
+    {
+      "epoch": 1.5126558556757703,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.902002747468869e-06,
+      "loss": 1.0549,
+      "step": 7646
+    },
+    {
+      "epoch": 1.5128557507308664,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.900949150702308e-06,
+      "loss": 1.0335,
+      "step": 7647
+    },
+    {
+      "epoch": 1.5130556457859625,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.8998955583355755e-06,
+      "loss": 0.935,
+      "step": 7648
+    },
+    {
+      "epoch": 1.5132555408410584,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.89884197041547e-06,
+      "loss": 1.0581,
+      "step": 7649
+    },
+    {
+      "epoch": 1.5134554358961545,
+      "grad_norm": 1.8984375,
+      "learning_rate": 4.897788386988796e-06,
+      "loss": 0.8837,
+      "step": 7650
+    },
+    {
+      "epoch": 1.5136553309512506,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.896734808102348e-06,
+      "loss": 0.9951,
+      "step": 7651
+    },
+    {
+      "epoch": 1.5138552260063467,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.89568123380293e-06,
+      "loss": 0.9621,
+      "step": 7652
+    },
+    {
+      "epoch": 1.5140551210614426,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.8946276641373406e-06,
+      "loss": 0.9108,
+      "step": 7653
+    },
+    {
+      "epoch": 1.5142550161165387,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.893574099152379e-06,
+      "loss": 1.0131,
+      "step": 7654
+    },
+    {
+      "epoch": 1.5144549111716348,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.892520538894844e-06,
+      "loss": 0.9307,
+      "step": 7655
+    },
+    {
+      "epoch": 1.5146548062267309,
+      "grad_norm": 1.96875,
+      "learning_rate": 4.891466983411537e-06,
+      "loss": 0.8952,
+      "step": 7656
+    },
+    {
+      "epoch": 1.514854701281827,
+      "grad_norm": 2.5,
+      "learning_rate": 4.890413432749253e-06,
+      "loss": 1.0152,
+      "step": 7657
+    },
+    {
+      "epoch": 1.515054596336923,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.889359886954794e-06,
+      "loss": 0.9754,
+      "step": 7658
+    },
+    {
+      "epoch": 1.5152544913920192,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.888306346074956e-06,
+      "loss": 0.8848,
+      "step": 7659
+    },
+    {
+      "epoch": 1.5154543864471153,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.887252810156538e-06,
+      "loss": 1.0255,
+      "step": 7660
+    },
+    {
+      "epoch": 1.5156542815022114,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.8861992792463396e-06,
+      "loss": 0.927,
+      "step": 7661
+    },
+    {
+      "epoch": 1.5158541765573075,
+      "grad_norm": 2.125,
+      "learning_rate": 4.885145753391158e-06,
+      "loss": 1.0345,
+      "step": 7662
+    },
+    {
+      "epoch": 1.5160540716124036,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.884092232637787e-06,
+      "loss": 0.9341,
+      "step": 7663
+    },
+    {
+      "epoch": 1.5162539666674997,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.883038717033028e-06,
+      "loss": 1.0379,
+      "step": 7664
+    },
+    {
+      "epoch": 1.5164538617225958,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.881985206623676e-06,
+      "loss": 0.8957,
+      "step": 7665
+    },
+    {
+      "epoch": 1.5166537567776919,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.8809317014565286e-06,
+      "loss": 1.0101,
+      "step": 7666
+    },
+    {
+      "epoch": 1.5168536518327878,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.879878201578382e-06,
+      "loss": 0.9537,
+      "step": 7667
+    },
+    {
+      "epoch": 1.5170535468878839,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.878824707036034e-06,
+      "loss": 0.9919,
+      "step": 7668
+    },
+    {
+      "epoch": 1.51725344194298,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.877771217876279e-06,
+      "loss": 0.9453,
+      "step": 7669
+    },
+    {
+      "epoch": 1.517453336998076,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.876717734145914e-06,
+      "loss": 1.0619,
+      "step": 7670
+    },
+    {
+      "epoch": 1.517653232053172,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.875664255891733e-06,
+      "loss": 0.915,
+      "step": 7671
+    },
+    {
+      "epoch": 1.517853127108268,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.874610783160532e-06,
+      "loss": 0.9857,
+      "step": 7672
+    },
+    {
+      "epoch": 1.5180530221633641,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.873557315999108e-06,
+      "loss": 1.0185,
+      "step": 7673
+    },
+    {
+      "epoch": 1.5182529172184602,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.872503854454254e-06,
+      "loss": 0.9332,
+      "step": 7674
+    },
+    {
+      "epoch": 1.5184528122735563,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.871450398572764e-06,
+      "loss": 0.9758,
+      "step": 7675
+    },
+    {
+      "epoch": 1.5186527073286524,
+      "grad_norm": 2.25,
+      "learning_rate": 4.870396948401436e-06,
+      "loss": 0.994,
+      "step": 7676
+    },
+    {
+      "epoch": 1.5188526023837485,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.86934350398706e-06,
+      "loss": 1.0261,
+      "step": 7677
+    },
+    {
+      "epoch": 1.5190524974388446,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.868290065376433e-06,
+      "loss": 0.909,
+      "step": 7678
+    },
+    {
+      "epoch": 1.5192523924939407,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.8672366326163466e-06,
+      "loss": 0.9689,
+      "step": 7679
+    },
+    {
+      "epoch": 1.5194522875490368,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.866183205753593e-06,
+      "loss": 0.9509,
+      "step": 7680
+    },
+    {
+      "epoch": 1.519652182604133,
+      "grad_norm": 2.0,
+      "learning_rate": 4.86512978483497e-06,
+      "loss": 0.8929,
+      "step": 7681
+    },
+    {
+      "epoch": 1.519852077659229,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.864076369907265e-06,
+      "loss": 0.9983,
+      "step": 7682
+    },
+    {
+      "epoch": 1.5200519727143251,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.863022961017275e-06,
+      "loss": 1.0158,
+      "step": 7683
+    },
+    {
+      "epoch": 1.520251867769421,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.86196955821179e-06,
+      "loss": 1.0297,
+      "step": 7684
+    },
+    {
+      "epoch": 1.520451762824517,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.860916161537602e-06,
+      "loss": 1.0098,
+      "step": 7685
+    },
+    {
+      "epoch": 1.5206516578796132,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.859862771041503e-06,
+      "loss": 0.9938,
+      "step": 7686
+    },
+    {
+      "epoch": 1.5208515529347093,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.858809386770286e-06,
+      "loss": 0.9878,
+      "step": 7687
+    },
+    {
+      "epoch": 1.5210514479898054,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.85775600877074e-06,
+      "loss": 0.9156,
+      "step": 7688
+    },
+    {
+      "epoch": 1.5212513430449013,
+      "grad_norm": 2.125,
+      "learning_rate": 4.856702637089657e-06,
+      "loss": 0.9137,
+      "step": 7689
+    },
+    {
+      "epoch": 1.5214512380999974,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.8556492717738275e-06,
+      "loss": 1.0033,
+      "step": 7690
+    },
+    {
+      "epoch": 1.5216511331550935,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.854595912870041e-06,
+      "loss": 0.9716,
+      "step": 7691
+    },
+    {
+      "epoch": 1.5218510282101896,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.8535425604250915e-06,
+      "loss": 0.9526,
+      "step": 7692
+    },
+    {
+      "epoch": 1.5220509232652857,
+      "grad_norm": 2.125,
+      "learning_rate": 4.8524892144857655e-06,
+      "loss": 0.9911,
+      "step": 7693
+    },
+    {
+      "epoch": 1.5222508183203818,
+      "grad_norm": 2.25,
+      "learning_rate": 4.851435875098851e-06,
+      "loss": 0.9205,
+      "step": 7694
+    },
+    {
+      "epoch": 1.5224507133754779,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.850382542311139e-06,
+      "loss": 0.918,
+      "step": 7695
+    },
+    {
+      "epoch": 1.522650608430574,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.849329216169419e-06,
+      "loss": 1.0557,
+      "step": 7696
+    },
+    {
+      "epoch": 1.52285050348567,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.848275896720478e-06,
+      "loss": 1.0096,
+      "step": 7697
+    },
+    {
+      "epoch": 1.5230503985407662,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.847222584011107e-06,
+      "loss": 1.0347,
+      "step": 7698
+    },
+    {
+      "epoch": 1.5232502935958623,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.846169278088092e-06,
+      "loss": 0.9796,
+      "step": 7699
+    },
+    {
+      "epoch": 1.5234501886509584,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.845115978998221e-06,
+      "loss": 0.9395,
+      "step": 7700
+    },
+    {
+      "epoch": 1.5236500837060545,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.844062686788282e-06,
+      "loss": 1.0408,
+      "step": 7701
+    },
+    {
+      "epoch": 1.5238499787611504,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.843009401505062e-06,
+      "loss": 1.001,
+      "step": 7702
+    },
+    {
+      "epoch": 1.5240498738162465,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.841956123195347e-06,
+      "loss": 0.9442,
+      "step": 7703
+    },
+    {
+      "epoch": 1.5242497688713426,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.8409028519059246e-06,
+      "loss": 0.9601,
+      "step": 7704
+    },
+    {
+      "epoch": 1.5244496639264387,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.8398495876835795e-06,
+      "loss": 0.933,
+      "step": 7705
+    },
+    {
+      "epoch": 1.5246495589815345,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.838796330575099e-06,
+      "loss": 1.0609,
+      "step": 7706
+    },
+    {
+      "epoch": 1.5248494540366306,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.837743080627269e-06,
+      "loss": 0.8928,
+      "step": 7707
+    },
+    {
+      "epoch": 1.5250493490917267,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.836689837886874e-06,
+      "loss": 0.917,
+      "step": 7708
+    },
+    {
+      "epoch": 1.5252492441468228,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.835636602400699e-06,
+      "loss": 0.9897,
+      "step": 7709
+    },
+    {
+      "epoch": 1.525449139201919,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.834583374215528e-06,
+      "loss": 0.9326,
+      "step": 7710
+    },
+    {
+      "epoch": 1.525649034257015,
+      "grad_norm": 2.125,
+      "learning_rate": 4.833530153378145e-06,
+      "loss": 0.9718,
+      "step": 7711
+    },
+    {
+      "epoch": 1.5258489293121111,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.832476939935337e-06,
+      "loss": 1.0237,
+      "step": 7712
+    },
+    {
+      "epoch": 1.5260488243672072,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.831423733933884e-06,
+      "loss": 0.9722,
+      "step": 7713
+    },
+    {
+      "epoch": 1.5262487194223033,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.83037053542057e-06,
+      "loss": 0.9436,
+      "step": 7714
+    },
+    {
+      "epoch": 1.5264486144773994,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.829317344442181e-06,
+      "loss": 0.9367,
+      "step": 7715
+    },
+    {
+      "epoch": 1.5266485095324955,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.828264161045496e-06,
+      "loss": 0.9521,
+      "step": 7716
+    },
+    {
+      "epoch": 1.5268484045875916,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.827210985277299e-06,
+      "loss": 1.0105,
+      "step": 7717
+    },
+    {
+      "epoch": 1.5270482996426877,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.8261578171843705e-06,
+      "loss": 0.9382,
+      "step": 7718
+    },
+    {
+      "epoch": 1.5272481946977836,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.825104656813494e-06,
+      "loss": 0.9781,
+      "step": 7719
+    },
+    {
+      "epoch": 1.5274480897528797,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.82405150421145e-06,
+      "loss": 1.0679,
+      "step": 7720
+    },
+    {
+      "epoch": 1.5276479848079758,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.82299835942502e-06,
+      "loss": 0.9701,
+      "step": 7721
+    },
+    {
+      "epoch": 1.527847879863072,
+      "grad_norm": 2.125,
+      "learning_rate": 4.821945222500983e-06,
+      "loss": 0.9713,
+      "step": 7722
+    },
+    {
+      "epoch": 1.528047774918168,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.8208920934861206e-06,
+      "loss": 0.9887,
+      "step": 7723
+    },
+    {
+      "epoch": 1.528247669973264,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.819838972427214e-06,
+      "loss": 0.9757,
+      "step": 7724
+    },
+    {
+      "epoch": 1.52844756502836,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.818785859371039e-06,
+      "loss": 0.9716,
+      "step": 7725
+    },
+    {
+      "epoch": 1.528647460083456,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.817732754364376e-06,
+      "loss": 0.8541,
+      "step": 7726
+    },
+    {
+      "epoch": 1.5288473551385522,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.8166796574540065e-06,
+      "loss": 0.9456,
+      "step": 7727
+    },
+    {
+      "epoch": 1.5290472501936483,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.815626568686706e-06,
+      "loss": 0.9526,
+      "step": 7728
+    },
+    {
+      "epoch": 1.5292471452487444,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.8145734881092545e-06,
+      "loss": 1.0064,
+      "step": 7729
+    },
+    {
+      "epoch": 1.5294470403038405,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.813520415768427e-06,
+      "loss": 0.9533,
+      "step": 7730
+    },
+    {
+      "epoch": 1.5296469353589366,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.812467351711004e-06,
+      "loss": 1.0011,
+      "step": 7731
+    },
+    {
+      "epoch": 1.5298468304140327,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.811414295983761e-06,
+      "loss": 0.9818,
+      "step": 7732
+    },
+    {
+      "epoch": 1.5300467254691288,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.810361248633475e-06,
+      "loss": 1.009,
+      "step": 7733
+    },
+    {
+      "epoch": 1.5302466205242249,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.809308209706921e-06,
+      "loss": 0.9351,
+      "step": 7734
+    },
+    {
+      "epoch": 1.530446515579321,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.808255179250878e-06,
+      "loss": 0.9453,
+      "step": 7735
+    },
+    {
+      "epoch": 1.530646410634417,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.807202157312118e-06,
+      "loss": 0.8974,
+      "step": 7736
+    },
+    {
+      "epoch": 1.530846305689513,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.806149143937417e-06,
+      "loss": 0.9417,
+      "step": 7737
+    },
+    {
+      "epoch": 1.531046200744609,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.8050961391735516e-06,
+      "loss": 0.8945,
+      "step": 7738
+    },
+    {
+      "epoch": 1.5312460957997052,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.804043143067295e-06,
+      "loss": 1.0163,
+      "step": 7739
+    },
+    {
+      "epoch": 1.5314459908548013,
+      "grad_norm": 2.375,
+      "learning_rate": 4.802990155665421e-06,
+      "loss": 0.964,
+      "step": 7740
+    },
+    {
+      "epoch": 1.5316458859098971,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.801937177014702e-06,
+      "loss": 0.9406,
+      "step": 7741
+    },
+    {
+      "epoch": 1.5318457809649932,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.800884207161913e-06,
+      "loss": 0.9723,
+      "step": 7742
+    },
+    {
+      "epoch": 1.5320456760200893,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.799831246153827e-06,
+      "loss": 0.9697,
+      "step": 7743
+    },
+    {
+      "epoch": 1.5322455710751854,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.798778294037214e-06,
+      "loss": 0.919,
+      "step": 7744
+    },
+    {
+      "epoch": 1.5324454661302815,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.7977253508588496e-06,
+      "loss": 0.9584,
+      "step": 7745
+    },
+    {
+      "epoch": 1.5326453611853776,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.796672416665503e-06,
+      "loss": 1.0064,
+      "step": 7746
+    },
+    {
+      "epoch": 1.5328452562404737,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.795619491503946e-06,
+      "loss": 0.9836,
+      "step": 7747
+    },
+    {
+      "epoch": 1.5330451512955698,
+      "grad_norm": 2.25,
+      "learning_rate": 4.794566575420949e-06,
+      "loss": 1.0584,
+      "step": 7748
+    },
+    {
+      "epoch": 1.533245046350666,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.793513668463283e-06,
+      "loss": 0.9774,
+      "step": 7749
+    },
+    {
+      "epoch": 1.533444941405762,
+      "grad_norm": 3.0,
+      "learning_rate": 4.792460770677716e-06,
+      "loss": 0.956,
+      "step": 7750
+    },
+    {
+      "epoch": 1.5336448364608581,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.791407882111022e-06,
+      "loss": 1.0692,
+      "step": 7751
+    },
+    {
+      "epoch": 1.5338447315159542,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.790355002809967e-06,
+      "loss": 0.9051,
+      "step": 7752
+    },
+    {
+      "epoch": 1.5340446265710503,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.789302132821319e-06,
+      "loss": 1.0195,
+      "step": 7753
+    },
+    {
+      "epoch": 1.5342445216261462,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.78824927219185e-06,
+      "loss": 0.966,
+      "step": 7754
+    },
+    {
+      "epoch": 1.5344444166812423,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.787196420968326e-06,
+      "loss": 1.0108,
+      "step": 7755
+    },
+    {
+      "epoch": 1.5346443117363384,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.7861435791975124e-06,
+      "loss": 0.9161,
+      "step": 7756
+    },
+    {
+      "epoch": 1.5348442067914345,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.785090746926178e-06,
+      "loss": 0.9542,
+      "step": 7757
+    },
+    {
+      "epoch": 1.5350441018465306,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.784037924201091e-06,
+      "loss": 1.0465,
+      "step": 7758
+    },
+    {
+      "epoch": 1.5352439969016265,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.782985111069015e-06,
+      "loss": 1.057,
+      "step": 7759
+    },
+    {
+      "epoch": 1.5354438919567226,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.781932307576718e-06,
+      "loss": 0.9282,
+      "step": 7760
+    },
+    {
+      "epoch": 1.5356437870118187,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.780879513770964e-06,
+      "loss": 1.0559,
+      "step": 7761
+    },
+    {
+      "epoch": 1.5358436820669148,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.779826729698519e-06,
+      "loss": 1.0127,
+      "step": 7762
+    },
+    {
+      "epoch": 1.536043577122011,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.778773955406148e-06,
+      "loss": 1.0193,
+      "step": 7763
+    },
+    {
+      "epoch": 1.536243472177107,
+      "grad_norm": 2.125,
+      "learning_rate": 4.777721190940612e-06,
+      "loss": 0.9641,
+      "step": 7764
+    },
+    {
+      "epoch": 1.536443367232203,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.776668436348678e-06,
+      "loss": 0.9669,
+      "step": 7765
+    },
+    {
+      "epoch": 1.5366432622872992,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.77561569167711e-06,
+      "loss": 0.9652,
+      "step": 7766
+    },
+    {
+      "epoch": 1.5368431573423953,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.7745629569726664e-06,
+      "loss": 0.9537,
+      "step": 7767
+    },
+    {
+      "epoch": 1.5370430523974914,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.7735102322821145e-06,
+      "loss": 1.0082,
+      "step": 7768
+    },
+    {
+      "epoch": 1.5372429474525875,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.772457517652212e-06,
+      "loss": 0.8862,
+      "step": 7769
+    },
+    {
+      "epoch": 1.5374428425076836,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.771404813129722e-06,
+      "loss": 0.9674,
+      "step": 7770
+    },
+    {
+      "epoch": 1.5376427375627797,
+      "grad_norm": 2.25,
+      "learning_rate": 4.7703521187614075e-06,
+      "loss": 0.9746,
+      "step": 7771
+    },
+    {
+      "epoch": 1.5378426326178756,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.769299434594026e-06,
+      "loss": 0.964,
+      "step": 7772
+    },
+    {
+      "epoch": 1.5380425276729717,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.76824676067434e-06,
+      "loss": 0.9537,
+      "step": 7773
+    },
+    {
+      "epoch": 1.5382424227280678,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.767194097049108e-06,
+      "loss": 0.9013,
+      "step": 7774
+    },
+    {
+      "epoch": 1.5384423177831639,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.766141443765088e-06,
+      "loss": 0.9636,
+      "step": 7775
+    },
+    {
+      "epoch": 1.5386422128382597,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.765088800869041e-06,
+      "loss": 0.9825,
+      "step": 7776
+    },
+    {
+      "epoch": 1.5388421078933558,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.764036168407726e-06,
+      "loss": 0.9128,
+      "step": 7777
+    },
+    {
+      "epoch": 1.539042002948452,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.762983546427897e-06,
+      "loss": 0.9874,
+      "step": 7778
+    },
+    {
+      "epoch": 1.539241898003548,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.761930934976315e-06,
+      "loss": 0.9349,
+      "step": 7779
+    },
+    {
+      "epoch": 1.5394417930586441,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.760878334099734e-06,
+      "loss": 0.9253,
+      "step": 7780
+    },
+    {
+      "epoch": 1.5396416881137402,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.759825743844912e-06,
+      "loss": 0.9425,
+      "step": 7781
+    },
+    {
+      "epoch": 1.5398415831688363,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.758773164258606e-06,
+      "loss": 0.98,
+      "step": 7782
+    },
+    {
+      "epoch": 1.5400414782239324,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.75772059538757e-06,
+      "loss": 1.0028,
+      "step": 7783
+    },
+    {
+      "epoch": 1.5402413732790285,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.7566680372785575e-06,
+      "loss": 0.9386,
+      "step": 7784
+    },
+    {
+      "epoch": 1.5404412683341246,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.755615489978328e-06,
+      "loss": 1.011,
+      "step": 7785
+    },
+    {
+      "epoch": 1.5406411633892207,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.7545629535336315e-06,
+      "loss": 1.0177,
+      "step": 7786
+    },
+    {
+      "epoch": 1.5408410584443168,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.753510427991221e-06,
+      "loss": 0.9757,
+      "step": 7787
+    },
+    {
+      "epoch": 1.541040953499413,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.752457913397851e-06,
+      "loss": 0.967,
+      "step": 7788
+    },
+    {
+      "epoch": 1.541240848554509,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.751405409800275e-06,
+      "loss": 0.8561,
+      "step": 7789
+    },
+    {
+      "epoch": 1.541440743609605,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.750352917245244e-06,
+      "loss": 0.9998,
+      "step": 7790
+    },
+    {
+      "epoch": 1.541640638664701,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.74930043577951e-06,
+      "loss": 0.9589,
+      "step": 7791
+    },
+    {
+      "epoch": 1.5418405337197971,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.748247965449822e-06,
+      "loss": 0.9633,
+      "step": 7792
+    },
+    {
+      "epoch": 1.5420404287748932,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.747195506302933e-06,
+      "loss": 0.9434,
+      "step": 7793
+    },
+    {
+      "epoch": 1.542240323829989,
+      "grad_norm": 2.125,
+      "learning_rate": 4.746143058385594e-06,
+      "loss": 0.9524,
+      "step": 7794
+    },
+    {
+      "epoch": 1.5424402188850852,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.745090621744552e-06,
+      "loss": 0.9472,
+      "step": 7795
+    },
+    {
+      "epoch": 1.5426401139401813,
+      "grad_norm": 2.125,
+      "learning_rate": 4.744038196426558e-06,
+      "loss": 1.007,
+      "step": 7796
+    },
+    {
+      "epoch": 1.5428400089952774,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.742985782478361e-06,
+      "loss": 0.9061,
+      "step": 7797
+    },
+    {
+      "epoch": 1.5430399040503735,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.741933379946706e-06,
+      "loss": 1.0119,
+      "step": 7798
+    },
+    {
+      "epoch": 1.5432397991054696,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.740880988878345e-06,
+      "loss": 0.9332,
+      "step": 7799
+    },
+    {
+      "epoch": 1.5434396941605657,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.739828609320021e-06,
+      "loss": 0.8967,
+      "step": 7800
+    },
+    {
+      "epoch": 1.5436395892156618,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.738776241318482e-06,
+      "loss": 0.8662,
+      "step": 7801
+    },
+    {
+      "epoch": 1.543839484270758,
+      "grad_norm": 2.125,
+      "learning_rate": 4.737723884920476e-06,
+      "loss": 0.9848,
+      "step": 7802
+    },
+    {
+      "epoch": 1.544039379325854,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.736671540172746e-06,
+      "loss": 1.0055,
+      "step": 7803
+    },
+    {
+      "epoch": 1.54423927438095,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.735619207122038e-06,
+      "loss": 0.9009,
+      "step": 7804
+    },
+    {
+      "epoch": 1.5444391694360462,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.734566885815098e-06,
+      "loss": 1.0093,
+      "step": 7805
+    },
+    {
+      "epoch": 1.5446390644911423,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.733514576298666e-06,
+      "loss": 0.9394,
+      "step": 7806
+    },
+    {
+      "epoch": 1.5448389595462382,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.732462278619491e-06,
+      "loss": 1.0647,
+      "step": 7807
+    },
+    {
+      "epoch": 1.5450388546013343,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.731409992824311e-06,
+      "loss": 0.9244,
+      "step": 7808
+    },
+    {
+      "epoch": 1.5452387496564304,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.730357718959871e-06,
+      "loss": 0.9751,
+      "step": 7809
+    },
+    {
+      "epoch": 1.5454386447115265,
+      "grad_norm": 2.125,
+      "learning_rate": 4.729305457072913e-06,
+      "loss": 0.9699,
+      "step": 7810
+    },
+    {
+      "epoch": 1.5456385397666226,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.728253207210176e-06,
+      "loss": 0.9998,
+      "step": 7811
+    },
+    {
+      "epoch": 1.5458384348217185,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.727200969418404e-06,
+      "loss": 0.9604,
+      "step": 7812
+    },
+    {
+      "epoch": 1.5460383298768146,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.726148743744336e-06,
+      "loss": 1.0473,
+      "step": 7813
+    },
+    {
+      "epoch": 1.5462382249319107,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.72509653023471e-06,
+      "loss": 0.9781,
+      "step": 7814
+    },
+    {
+      "epoch": 1.5464381199870068,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.724044328936268e-06,
+      "loss": 0.9372,
+      "step": 7815
+    },
+    {
+      "epoch": 1.5466380150421029,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.722992139895748e-06,
+      "loss": 0.9463,
+      "step": 7816
+    },
+    {
+      "epoch": 1.546837910097199,
+      "grad_norm": 1.9921875,
+      "learning_rate": 4.7219399631598875e-06,
+      "loss": 0.888,
+      "step": 7817
+    },
+    {
+      "epoch": 1.547037805152295,
+      "grad_norm": 2.125,
+      "learning_rate": 4.720887798775424e-06,
+      "loss": 0.978,
+      "step": 7818
+    },
+    {
+      "epoch": 1.5472377002073912,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.719835646789094e-06,
+      "loss": 0.9491,
+      "step": 7819
+    },
+    {
+      "epoch": 1.5474375952624873,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.718783507247634e-06,
+      "loss": 0.9715,
+      "step": 7820
+    },
+    {
+      "epoch": 1.5476374903175834,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.717731380197782e-06,
+      "loss": 0.9565,
+      "step": 7821
+    },
+    {
+      "epoch": 1.5478373853726795,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.716679265686271e-06,
+      "loss": 0.9697,
+      "step": 7822
+    },
+    {
+      "epoch": 1.5480372804277756,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.715627163759837e-06,
+      "loss": 0.9604,
+      "step": 7823
+    },
+    {
+      "epoch": 1.5482371754828717,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.714575074465213e-06,
+      "loss": 1.0961,
+      "step": 7824
+    },
+    {
+      "epoch": 1.5484370705379675,
+      "grad_norm": 2.125,
+      "learning_rate": 4.7135229978491354e-06,
+      "loss": 1.0226,
+      "step": 7825
+    },
+    {
+      "epoch": 1.5486369655930636,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.7124709339583345e-06,
+      "loss": 0.9901,
+      "step": 7826
+    },
+    {
+      "epoch": 1.5488368606481597,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.711418882839544e-06,
+      "loss": 0.9758,
+      "step": 7827
+    },
+    {
+      "epoch": 1.5490367557032558,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.710366844539497e-06,
+      "loss": 1.0121,
+      "step": 7828
+    },
+    {
+      "epoch": 1.5492366507583517,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.7093148191049224e-06,
+      "loss": 0.8907,
+      "step": 7829
+    },
+    {
+      "epoch": 1.5494365458134478,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.708262806582554e-06,
+      "loss": 0.9393,
+      "step": 7830
+    },
+    {
+      "epoch": 1.549636440868544,
+      "grad_norm": 2.25,
+      "learning_rate": 4.70721080701912e-06,
+      "loss": 1.0503,
+      "step": 7831
+    },
+    {
+      "epoch": 1.54983633592364,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.706158820461349e-06,
+      "loss": 0.9926,
+      "step": 7832
+    },
+    {
+      "epoch": 1.550036230978736,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.705106846955975e-06,
+      "loss": 1.0444,
+      "step": 7833
+    },
+    {
+      "epoch": 1.5502361260338322,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.704054886549721e-06,
+      "loss": 0.8473,
+      "step": 7834
+    },
+    {
+      "epoch": 1.5504360210889283,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.703002939289317e-06,
+      "loss": 1.0376,
+      "step": 7835
+    },
+    {
+      "epoch": 1.5506359161440244,
+      "grad_norm": 2.125,
+      "learning_rate": 4.7019510052214924e-06,
+      "loss": 1.0081,
+      "step": 7836
+    },
+    {
+      "epoch": 1.5508358111991205,
+      "grad_norm": 2.125,
+      "learning_rate": 4.7008990843929714e-06,
+      "loss": 0.9817,
+      "step": 7837
+    },
+    {
+      "epoch": 1.5510357062542166,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.699847176850482e-06,
+      "loss": 1.0341,
+      "step": 7838
+    },
+    {
+      "epoch": 1.5512356013093127,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.698795282640748e-06,
+      "loss": 1.0323,
+      "step": 7839
+    },
+    {
+      "epoch": 1.5514354963644088,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.6977434018104955e-06,
+      "loss": 1.0385,
+      "step": 7840
+    },
+    {
+      "epoch": 1.551635391419505,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.69669153440645e-06,
+      "loss": 1.0385,
+      "step": 7841
+    },
+    {
+      "epoch": 1.5518352864746008,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.695639680475332e-06,
+      "loss": 0.8897,
+      "step": 7842
+    },
+    {
+      "epoch": 1.5520351815296969,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.694587840063868e-06,
+      "loss": 0.9869,
+      "step": 7843
+    },
+    {
+      "epoch": 1.552235076584793,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.69353601321878e-06,
+      "loss": 1.0172,
+      "step": 7844
+    },
+    {
+      "epoch": 1.552434971639889,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.692484199986789e-06,
+      "loss": 0.9036,
+      "step": 7845
+    },
+    {
+      "epoch": 1.5526348666949852,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.691432400414617e-06,
+      "loss": 0.9853,
+      "step": 7846
+    },
+    {
+      "epoch": 1.552834761750081,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.690380614548986e-06,
+      "loss": 1.0161,
+      "step": 7847
+    },
+    {
+      "epoch": 1.5530346568051772,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.6893288424366145e-06,
+      "loss": 0.9438,
+      "step": 7848
+    },
+    {
+      "epoch": 1.5532345518602733,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.688277084124222e-06,
+      "loss": 0.9814,
+      "step": 7849
+    },
+    {
+      "epoch": 1.5534344469153694,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.687225339658529e-06,
+      "loss": 0.954,
+      "step": 7850
+    },
+    {
+      "epoch": 1.5536343419704655,
+      "grad_norm": 2.25,
+      "learning_rate": 4.686173609086251e-06,
+      "loss": 1.0238,
+      "step": 7851
+    },
+    {
+      "epoch": 1.5538342370255616,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.685121892454109e-06,
+      "loss": 0.9797,
+      "step": 7852
+    },
+    {
+      "epoch": 1.5540341320806577,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.68407018980882e-06,
+      "loss": 0.9091,
+      "step": 7853
+    },
+    {
+      "epoch": 1.5542340271357538,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.683018501197098e-06,
+      "loss": 0.9587,
+      "step": 7854
+    },
+    {
+      "epoch": 1.5544339221908499,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.6819668266656594e-06,
+      "loss": 0.8859,
+      "step": 7855
+    },
+    {
+      "epoch": 1.554633817245946,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.680915166261223e-06,
+      "loss": 1.0847,
+      "step": 7856
+    },
+    {
+      "epoch": 1.554833712301042,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.679863520030499e-06,
+      "loss": 1.0824,
+      "step": 7857
+    },
+    {
+      "epoch": 1.5550336073561382,
+      "grad_norm": 2.25,
+      "learning_rate": 4.678811888020204e-06,
+      "loss": 0.9992,
+      "step": 7858
+    },
+    {
+      "epoch": 1.5552335024112343,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.6777602702770505e-06,
+      "loss": 0.9977,
+      "step": 7859
+    },
+    {
+      "epoch": 1.5554333974663301,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.67670866684775e-06,
+      "loss": 1.0028,
+      "step": 7860
+    },
+    {
+      "epoch": 1.5556332925214262,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.675657077779019e-06,
+      "loss": 0.894,
+      "step": 7861
+    },
+    {
+      "epoch": 1.5558331875765223,
+      "grad_norm": 2.125,
+      "learning_rate": 4.674605503117563e-06,
+      "loss": 1.0168,
+      "step": 7862
+    },
+    {
+      "epoch": 1.5560330826316184,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.673553942910097e-06,
+      "loss": 0.9088,
+      "step": 7863
+    },
+    {
+      "epoch": 1.5562329776867143,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.67250239720333e-06,
+      "loss": 0.8533,
+      "step": 7864
+    },
+    {
+      "epoch": 1.5564328727418104,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.6714508660439715e-06,
+      "loss": 1.0101,
+      "step": 7865
+    },
+    {
+      "epoch": 1.5566327677969065,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.67039934947873e-06,
+      "loss": 0.9974,
+      "step": 7866
+    },
+    {
+      "epoch": 1.5568326628520026,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.669347847554315e-06,
+      "loss": 0.896,
+      "step": 7867
+    },
+    {
+      "epoch": 1.5570325579070987,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.6682963603174326e-06,
+      "loss": 0.9481,
+      "step": 7868
+    },
+    {
+      "epoch": 1.5572324529621948,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.667244887814791e-06,
+      "loss": 1.0765,
+      "step": 7869
+    },
+    {
+      "epoch": 1.557432348017291,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.666193430093096e-06,
+      "loss": 1.0258,
+      "step": 7870
+    },
+    {
+      "epoch": 1.557632243072387,
+      "grad_norm": 2.125,
+      "learning_rate": 4.665141987199052e-06,
+      "loss": 0.938,
+      "step": 7871
+    },
+    {
+      "epoch": 1.5578321381274831,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.664090559179367e-06,
+      "loss": 1.0506,
+      "step": 7872
+    },
+    {
+      "epoch": 1.5580320331825792,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.663039146080742e-06,
+      "loss": 0.999,
+      "step": 7873
+    },
+    {
+      "epoch": 1.5582319282376753,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.661987747949882e-06,
+      "loss": 0.9825,
+      "step": 7874
+    },
+    {
+      "epoch": 1.5584318232927714,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.660936364833492e-06,
+      "loss": 0.8881,
+      "step": 7875
+    },
+    {
+      "epoch": 1.5586317183478675,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.659884996778271e-06,
+      "loss": 1.021,
+      "step": 7876
+    },
+    {
+      "epoch": 1.5588316134029634,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.658833643830923e-06,
+      "loss": 0.9397,
+      "step": 7877
+    },
+    {
+      "epoch": 1.5590315084580595,
+      "grad_norm": 2.125,
+      "learning_rate": 4.657782306038148e-06,
+      "loss": 0.9747,
+      "step": 7878
+    },
+    {
+      "epoch": 1.5592314035131556,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.6567309834466465e-06,
+      "loss": 1.0115,
+      "step": 7879
+    },
+    {
+      "epoch": 1.5594312985682517,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.6556796761031175e-06,
+      "loss": 0.9302,
+      "step": 7880
+    },
+    {
+      "epoch": 1.5596311936233478,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.654628384054261e-06,
+      "loss": 1.0368,
+      "step": 7881
+    },
+    {
+      "epoch": 1.5598310886784437,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.653577107346774e-06,
+      "loss": 0.9975,
+      "step": 7882
+    },
+    {
+      "epoch": 1.5600309837335398,
+      "grad_norm": 2.125,
+      "learning_rate": 4.652525846027355e-06,
+      "loss": 0.9942,
+      "step": 7883
+    },
+    {
+      "epoch": 1.5602308787886359,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.651474600142701e-06,
+      "loss": 0.8976,
+      "step": 7884
+    },
+    {
+      "epoch": 1.560430773843732,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.650423369739508e-06,
+      "loss": 0.9892,
+      "step": 7885
+    },
+    {
+      "epoch": 1.560630668898828,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.649372154864471e-06,
+      "loss": 0.9206,
+      "step": 7886
+    },
+    {
+      "epoch": 1.5608305639539242,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.648320955564287e-06,
+      "loss": 0.9585,
+      "step": 7887
+    },
+    {
+      "epoch": 1.5610304590090203,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.647269771885648e-06,
+      "loss": 0.9131,
+      "step": 7888
+    },
+    {
+      "epoch": 1.5612303540641164,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.646218603875248e-06,
+      "loss": 1.0652,
+      "step": 7889
+    },
+    {
+      "epoch": 1.5614302491192125,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.645167451579779e-06,
+      "loss": 1.0153,
+      "step": 7890
+    },
+    {
+      "epoch": 1.5616301441743086,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.644116315045933e-06,
+      "loss": 0.965,
+      "step": 7891
+    },
+    {
+      "epoch": 1.5618300392294047,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.643065194320405e-06,
+      "loss": 0.9428,
+      "step": 7892
+    },
+    {
+      "epoch": 1.5620299342845008,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.642014089449881e-06,
+      "loss": 0.9131,
+      "step": 7893
+    },
+    {
+      "epoch": 1.5622298293395969,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.6409630004810535e-06,
+      "loss": 0.9232,
+      "step": 7894
+    },
+    {
+      "epoch": 1.5624297243946927,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.639911927460612e-06,
+      "loss": 1.0809,
+      "step": 7895
+    },
+    {
+      "epoch": 1.5626296194497888,
+      "grad_norm": 2.125,
+      "learning_rate": 4.638860870435243e-06,
+      "loss": 0.9853,
+      "step": 7896
+    },
+    {
+      "epoch": 1.562829514504885,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.637809829451637e-06,
+      "loss": 0.9465,
+      "step": 7897
+    },
+    {
+      "epoch": 1.563029409559981,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.636758804556478e-06,
+      "loss": 0.8824,
+      "step": 7898
+    },
+    {
+      "epoch": 1.563229304615077,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.635707795796454e-06,
+      "loss": 1.1537,
+      "step": 7899
+    },
+    {
+      "epoch": 1.563429199670173,
+      "grad_norm": 2.25,
+      "learning_rate": 4.634656803218252e-06,
+      "loss": 0.9672,
+      "step": 7900
+    },
+    {
+      "epoch": 1.5636290947252691,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.633605826868556e-06,
+      "loss": 0.9607,
+      "step": 7901
+    },
+    {
+      "epoch": 1.5638289897803652,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.632554866794048e-06,
+      "loss": 0.951,
+      "step": 7902
+    },
+    {
+      "epoch": 1.5640288848354613,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.631503923041416e-06,
+      "loss": 0.9414,
+      "step": 7903
+    },
+    {
+      "epoch": 1.5642287798905574,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.63045299565734e-06,
+      "loss": 0.9421,
+      "step": 7904
+    },
+    {
+      "epoch": 1.5644286749456535,
+      "grad_norm": 2.125,
+      "learning_rate": 4.629402084688501e-06,
+      "loss": 0.9926,
+      "step": 7905
+    },
+    {
+      "epoch": 1.5646285700007496,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.628351190181583e-06,
+      "loss": 0.9994,
+      "step": 7906
+    },
+    {
+      "epoch": 1.5648284650558457,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.627300312183265e-06,
+      "loss": 1.1636,
+      "step": 7907
+    },
+    {
+      "epoch": 1.5650283601109418,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.6262494507402275e-06,
+      "loss": 0.9685,
+      "step": 7908
+    },
+    {
+      "epoch": 1.565228255166038,
+      "grad_norm": 2.0,
+      "learning_rate": 4.625198605899152e-06,
+      "loss": 0.9123,
+      "step": 7909
+    },
+    {
+      "epoch": 1.565428150221134,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.624147777706711e-06,
+      "loss": 0.9561,
+      "step": 7910
+    },
+    {
+      "epoch": 1.5656280452762301,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.623096966209586e-06,
+      "loss": 0.9164,
+      "step": 7911
+    },
+    {
+      "epoch": 1.5658279403313262,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.622046171454454e-06,
+      "loss": 0.9847,
+      "step": 7912
+    },
+    {
+      "epoch": 1.566027835386422,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.620995393487991e-06,
+      "loss": 0.9813,
+      "step": 7913
+    },
+    {
+      "epoch": 1.5662277304415182,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.619944632356871e-06,
+      "loss": 0.9009,
+      "step": 7914
+    },
+    {
+      "epoch": 1.5664276254966143,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.618893888107772e-06,
+      "loss": 0.9058,
+      "step": 7915
+    },
+    {
+      "epoch": 1.5666275205517104,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.617843160787364e-06,
+      "loss": 0.9579,
+      "step": 7916
+    },
+    {
+      "epoch": 1.5668274156068063,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.616792450442321e-06,
+      "loss": 0.9935,
+      "step": 7917
+    },
+    {
+      "epoch": 1.5670273106619024,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.6157417571193184e-06,
+      "loss": 0.8968,
+      "step": 7918
+    },
+    {
+      "epoch": 1.5672272057169985,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.6146910808650255e-06,
+      "loss": 0.9871,
+      "step": 7919
+    },
+    {
+      "epoch": 1.5674271007720946,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.613640421726114e-06,
+      "loss": 0.938,
+      "step": 7920
+    },
+    {
+      "epoch": 1.5676269958271907,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.612589779749252e-06,
+      "loss": 0.9615,
+      "step": 7921
+    },
+    {
+      "epoch": 1.5678268908822868,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.611539154981112e-06,
+      "loss": 0.948,
+      "step": 7922
+    },
+    {
+      "epoch": 1.5680267859373829,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.610488547468361e-06,
+      "loss": 0.984,
+      "step": 7923
+    },
+    {
+      "epoch": 1.568226680992479,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.609437957257667e-06,
+      "loss": 1.0223,
+      "step": 7924
+    },
+    {
+      "epoch": 1.568426576047575,
+      "grad_norm": 2.125,
+      "learning_rate": 4.608387384395697e-06,
+      "loss": 0.9318,
+      "step": 7925
+    },
+    {
+      "epoch": 1.5686264711026712,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.607336828929119e-06,
+      "loss": 0.9919,
+      "step": 7926
+    },
+    {
+      "epoch": 1.5688263661577673,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.606286290904595e-06,
+      "loss": 1.0524,
+      "step": 7927
+    },
+    {
+      "epoch": 1.5690262612128634,
+      "grad_norm": 2.25,
+      "learning_rate": 4.605235770368794e-06,
+      "loss": 0.9626,
+      "step": 7928
+    },
+    {
+      "epoch": 1.5692261562679595,
+      "grad_norm": 2.5,
+      "learning_rate": 4.604185267368376e-06,
+      "loss": 0.9993,
+      "step": 7929
+    },
+    {
+      "epoch": 1.5694260513230553,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.603134781950007e-06,
+      "loss": 0.9542,
+      "step": 7930
+    },
+    {
+      "epoch": 1.5696259463781514,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.6020843141603484e-06,
+      "loss": 0.952,
+      "step": 7931
+    },
+    {
+      "epoch": 1.5698258414332475,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.601033864046062e-06,
+      "loss": 0.9748,
+      "step": 7932
+    },
+    {
+      "epoch": 1.5700257364883436,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.599983431653809e-06,
+      "loss": 1.0249,
+      "step": 7933
+    },
+    {
+      "epoch": 1.5702256315434397,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.59893301703025e-06,
+      "loss": 0.9893,
+      "step": 7934
+    },
+    {
+      "epoch": 1.5704255265985356,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.597882620222043e-06,
+      "loss": 0.9827,
+      "step": 7935
+    },
+    {
+      "epoch": 1.5706254216536317,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.596832241275846e-06,
+      "loss": 1.1081,
+      "step": 7936
+    },
+    {
+      "epoch": 1.5708253167087278,
+      "grad_norm": 2.25,
+      "learning_rate": 4.595781880238319e-06,
+      "loss": 0.9389,
+      "step": 7937
+    },
+    {
+      "epoch": 1.571025211763824,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.594731537156117e-06,
+      "loss": 0.9455,
+      "step": 7938
+    },
+    {
+      "epoch": 1.57122510681892,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.593681212075898e-06,
+      "loss": 0.9311,
+      "step": 7939
+    },
+    {
+      "epoch": 1.5714250018740161,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.592630905044317e-06,
+      "loss": 0.9819,
+      "step": 7940
+    },
+    {
+      "epoch": 1.5716248969291122,
+      "grad_norm": 2.25,
+      "learning_rate": 4.591580616108026e-06,
+      "loss": 1.0248,
+      "step": 7941
+    },
+    {
+      "epoch": 1.5718247919842083,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.59053034531368e-06,
+      "loss": 0.9329,
+      "step": 7942
+    },
+    {
+      "epoch": 1.5720246870393044,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.589480092707933e-06,
+      "loss": 1.0111,
+      "step": 7943
+    },
+    {
+      "epoch": 1.5722245820944005,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.5884298583374356e-06,
+      "loss": 0.9432,
+      "step": 7944
+    },
+    {
+      "epoch": 1.5724244771494966,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.5873796422488405e-06,
+      "loss": 1.0184,
+      "step": 7945
+    },
+    {
+      "epoch": 1.5726243722045927,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.586329444488798e-06,
+      "loss": 0.9221,
+      "step": 7946
+    },
+    {
+      "epoch": 1.5728242672596888,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.585279265103957e-06,
+      "loss": 1.0654,
+      "step": 7947
+    },
+    {
+      "epoch": 1.5730241623147847,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.584229104140967e-06,
+      "loss": 1.043,
+      "step": 7948
+    },
+    {
+      "epoch": 1.5732240573698808,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.583178961646475e-06,
+      "loss": 0.8936,
+      "step": 7949
+    },
+    {
+      "epoch": 1.573423952424977,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.58212883766713e-06,
+      "loss": 0.9949,
+      "step": 7950
+    },
+    {
+      "epoch": 1.573623847480073,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.581078732249577e-06,
+      "loss": 0.9255,
+      "step": 7951
+    },
+    {
+      "epoch": 1.5738237425351689,
+      "grad_norm": 2.25,
+      "learning_rate": 4.580028645440462e-06,
+      "loss": 0.9435,
+      "step": 7952
+    },
+    {
+      "epoch": 1.574023637590265,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.5789785772864295e-06,
+      "loss": 1.0914,
+      "step": 7953
+    },
+    {
+      "epoch": 1.574223532645361,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.577928527834124e-06,
+      "loss": 1.0328,
+      "step": 7954
+    },
+    {
+      "epoch": 1.5744234277004572,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.576878497130188e-06,
+      "loss": 0.9528,
+      "step": 7955
+    },
+    {
+      "epoch": 1.5746233227555533,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.575828485221263e-06,
+      "loss": 0.9652,
+      "step": 7956
+    },
+    {
+      "epoch": 1.5748232178106494,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.574778492153993e-06,
+      "loss": 0.8972,
+      "step": 7957
+    },
+    {
+      "epoch": 1.5750231128657455,
+      "grad_norm": 2.25,
+      "learning_rate": 4.573728517975016e-06,
+      "loss": 0.9544,
+      "step": 7958
+    },
+    {
+      "epoch": 1.5752230079208416,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.5726785627309736e-06,
+      "loss": 0.9876,
+      "step": 7959
+    },
+    {
+      "epoch": 1.5754229029759377,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.571628626468503e-06,
+      "loss": 0.9727,
+      "step": 7960
+    },
+    {
+      "epoch": 1.5756227980310338,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.570578709234243e-06,
+      "loss": 1.0302,
+      "step": 7961
+    },
+    {
+      "epoch": 1.5758226930861299,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.5695288110748324e-06,
+      "loss": 0.9296,
+      "step": 7962
+    },
+    {
+      "epoch": 1.576022588141226,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.568478932036904e-06,
+      "loss": 0.9386,
+      "step": 7963
+    },
+    {
+      "epoch": 1.576222483196322,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.567429072167095e-06,
+      "loss": 0.9132,
+      "step": 7964
+    },
+    {
+      "epoch": 1.576422378251418,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.566379231512042e-06,
+      "loss": 0.9991,
+      "step": 7965
+    },
+    {
+      "epoch": 1.576622273306514,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.565329410118377e-06,
+      "loss": 0.927,
+      "step": 7966
+    },
+    {
+      "epoch": 1.5768221683616102,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.564279608032734e-06,
+      "loss": 0.9367,
+      "step": 7967
+    },
+    {
+      "epoch": 1.5770220634167063,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.563229825301743e-06,
+      "loss": 1.0996,
+      "step": 7968
+    },
+    {
+      "epoch": 1.5772219584718024,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.562180061972036e-06,
+      "loss": 0.9513,
+      "step": 7969
+    },
+    {
+      "epoch": 1.5774218535268982,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.561130318090246e-06,
+      "loss": 1.0707,
+      "step": 7970
+    },
+    {
+      "epoch": 1.5776217485819943,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.560080593703e-06,
+      "loss": 0.9936,
+      "step": 7971
+    },
+    {
+      "epoch": 1.5778216436370904,
+      "grad_norm": 2.125,
+      "learning_rate": 4.5590308888569256e-06,
+      "loss": 0.952,
+      "step": 7972
+    },
+    {
+      "epoch": 1.5780215386921865,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.557981203598652e-06,
+      "loss": 0.9375,
+      "step": 7973
+    },
+    {
+      "epoch": 1.5782214337472826,
+      "grad_norm": 2.125,
+      "learning_rate": 4.556931537974808e-06,
+      "loss": 0.9408,
+      "step": 7974
+    },
+    {
+      "epoch": 1.5784213288023787,
+      "grad_norm": 2.25,
+      "learning_rate": 4.555881892032016e-06,
+      "loss": 1.037,
+      "step": 7975
+    },
+    {
+      "epoch": 1.5786212238574748,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.554832265816903e-06,
+      "loss": 0.9225,
+      "step": 7976
+    },
+    {
+      "epoch": 1.578821118912571,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.553782659376094e-06,
+      "loss": 0.9012,
+      "step": 7977
+    },
+    {
+      "epoch": 1.579021013967667,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.552733072756211e-06,
+      "loss": 0.994,
+      "step": 7978
+    },
+    {
+      "epoch": 1.5792209090227631,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.551683506003878e-06,
+      "loss": 0.9952,
+      "step": 7979
+    },
+    {
+      "epoch": 1.5794208040778592,
+      "grad_norm": 2.25,
+      "learning_rate": 4.5506339591657145e-06,
+      "loss": 0.9756,
+      "step": 7980
+    },
+    {
+      "epoch": 1.5796206991329553,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.549584432288343e-06,
+      "loss": 0.9691,
+      "step": 7981
+    },
+    {
+      "epoch": 1.5798205941880514,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.548534925418383e-06,
+      "loss": 1.0193,
+      "step": 7982
+    },
+    {
+      "epoch": 1.5800204892431473,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.547485438602453e-06,
+      "loss": 1.0499,
+      "step": 7983
+    },
+    {
+      "epoch": 1.5802203842982434,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.546435971887172e-06,
+      "loss": 0.9644,
+      "step": 7984
+    },
+    {
+      "epoch": 1.5804202793533395,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.545386525319157e-06,
+      "loss": 0.9267,
+      "step": 7985
+    },
+    {
+      "epoch": 1.5806201744084356,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.544337098945023e-06,
+      "loss": 1.0208,
+      "step": 7986
+    },
+    {
+      "epoch": 1.5808200694635315,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.543287692811388e-06,
+      "loss": 0.903,
+      "step": 7987
+    },
+    {
+      "epoch": 1.5810199645186276,
+      "grad_norm": 2.125,
+      "learning_rate": 4.542238306964863e-06,
+      "loss": 0.9197,
+      "step": 7988
+    },
+    {
+      "epoch": 1.5812198595737237,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.5411889414520634e-06,
+      "loss": 1.0371,
+      "step": 7989
+    },
+    {
+      "epoch": 1.5814197546288198,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.540139596319604e-06,
+      "loss": 0.985,
+      "step": 7990
+    },
+    {
+      "epoch": 1.5816196496839159,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.539090271614094e-06,
+      "loss": 0.8718,
+      "step": 7991
+    },
+    {
+      "epoch": 1.581819544739012,
+      "grad_norm": 1.96875,
+      "learning_rate": 4.538040967382144e-06,
+      "loss": 0.9993,
+      "step": 7992
+    },
+    {
+      "epoch": 1.582019439794108,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.536991683670366e-06,
+      "loss": 1.0313,
+      "step": 7993
+    },
+    {
+      "epoch": 1.5822193348492042,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.5359424205253665e-06,
+      "loss": 1.0063,
+      "step": 7994
+    },
+    {
+      "epoch": 1.5824192299043003,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.534893177993756e-06,
+      "loss": 1.0112,
+      "step": 7995
+    },
+    {
+      "epoch": 1.5826191249593964,
+      "grad_norm": 2.125,
+      "learning_rate": 4.533843956122142e-06,
+      "loss": 0.9603,
+      "step": 7996
+    },
+    {
+      "epoch": 1.5828190200144925,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.532794754957128e-06,
+      "loss": 0.9179,
+      "step": 7997
+    },
+    {
+      "epoch": 1.5830189150695886,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.531745574545322e-06,
+      "loss": 0.9546,
+      "step": 7998
+    },
+    {
+      "epoch": 1.5832188101246847,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.5306964149333265e-06,
+      "loss": 0.9534,
+      "step": 7999
+    },
+    {
+      "epoch": 1.5834187051797806,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.529647276167747e-06,
+      "loss": 0.9989,
+      "step": 8000
+    },
+    {
+      "epoch": 1.5836186002348767,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.5285981582951855e-06,
+      "loss": 0.9472,
+      "step": 8001
+    },
+    {
+      "epoch": 1.5838184952899728,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.527549061362245e-06,
+      "loss": 0.98,
+      "step": 8002
+    },
+    {
+      "epoch": 1.5840183903450689,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.526499985415522e-06,
+      "loss": 1.0411,
+      "step": 8003
+    },
+    {
+      "epoch": 1.584218285400165,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.5254509305016195e-06,
+      "loss": 0.9677,
+      "step": 8004
+    },
+    {
+      "epoch": 1.5844181804552608,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.524401896667137e-06,
+      "loss": 0.952,
+      "step": 8005
+    },
+    {
+      "epoch": 1.584618075510357,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.52335288395867e-06,
+      "loss": 0.8664,
+      "step": 8006
+    },
+    {
+      "epoch": 1.584817970565453,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.522303892422818e-06,
+      "loss": 0.9897,
+      "step": 8007
+    },
+    {
+      "epoch": 1.5850178656205491,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.521254922106176e-06,
+      "loss": 0.9461,
+      "step": 8008
+    },
+    {
+      "epoch": 1.5852177606756452,
+      "grad_norm": 2.125,
+      "learning_rate": 4.520205973055339e-06,
+      "loss": 0.9887,
+      "step": 8009
+    },
+    {
+      "epoch": 1.5854176557307413,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.519157045316902e-06,
+      "loss": 0.9829,
+      "step": 8010
+    },
+    {
+      "epoch": 1.5856175507858374,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.518108138937457e-06,
+      "loss": 0.9488,
+      "step": 8011
+    },
+    {
+      "epoch": 1.5858174458409335,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.517059253963596e-06,
+      "loss": 0.8656,
+      "step": 8012
+    },
+    {
+      "epoch": 1.5860173408960296,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.516010390441914e-06,
+      "loss": 0.8928,
+      "step": 8013
+    },
+    {
+      "epoch": 1.5862172359511257,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.5149615484189965e-06,
+      "loss": 0.8735,
+      "step": 8014
+    },
+    {
+      "epoch": 1.5864171310062218,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.513912727941437e-06,
+      "loss": 0.9873,
+      "step": 8015
+    },
+    {
+      "epoch": 1.586617026061318,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.512863929055821e-06,
+      "loss": 0.9418,
+      "step": 8016
+    },
+    {
+      "epoch": 1.586816921116414,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.511815151808737e-06,
+      "loss": 0.9767,
+      "step": 8017
+    },
+    {
+      "epoch": 1.58701681617151,
+      "grad_norm": 2.125,
+      "learning_rate": 4.510766396246773e-06,
+      "loss": 1.0002,
+      "step": 8018
+    },
+    {
+      "epoch": 1.587216711226606,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.5097176624165125e-06,
+      "loss": 0.9911,
+      "step": 8019
+    },
+    {
+      "epoch": 1.5874166062817021,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.508668950364541e-06,
+      "loss": 0.8995,
+      "step": 8020
+    },
+    {
+      "epoch": 1.5876165013367982,
+      "grad_norm": 2.25,
+      "learning_rate": 4.507620260137443e-06,
+      "loss": 1.0113,
+      "step": 8021
+    },
+    {
+      "epoch": 1.587816396391894,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.5065715917818e-06,
+      "loss": 0.9905,
+      "step": 8022
+    },
+    {
+      "epoch": 1.5880162914469902,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.5055229453441935e-06,
+      "loss": 0.9579,
+      "step": 8023
+    },
+    {
+      "epoch": 1.5882161865020863,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.504474320871206e-06,
+      "loss": 1.0448,
+      "step": 8024
+    },
+    {
+      "epoch": 1.5884160815571824,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.503425718409416e-06,
+      "loss": 0.9604,
+      "step": 8025
+    },
+    {
+      "epoch": 1.5886159766122785,
+      "grad_norm": 2.125,
+      "learning_rate": 4.502377138005402e-06,
+      "loss": 1.0344,
+      "step": 8026
+    },
+    {
+      "epoch": 1.5888158716673746,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.501328579705744e-06,
+      "loss": 1.0418,
+      "step": 8027
+    },
+    {
+      "epoch": 1.5890157667224707,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.500280043557015e-06,
+      "loss": 0.9852,
+      "step": 8028
+    },
+    {
+      "epoch": 1.5892156617775668,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.499231529605795e-06,
+      "loss": 0.8487,
+      "step": 8029
+    },
+    {
+      "epoch": 1.589415556832663,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.498183037898656e-06,
+      "loss": 1.0181,
+      "step": 8030
+    },
+    {
+      "epoch": 1.589615451887759,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.497134568482172e-06,
+      "loss": 0.9443,
+      "step": 8031
+    },
+    {
+      "epoch": 1.589815346942855,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.496086121402919e-06,
+      "loss": 1.0473,
+      "step": 8032
+    },
+    {
+      "epoch": 1.5900152419979512,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.495037696707467e-06,
+      "loss": 0.9742,
+      "step": 8033
+    },
+    {
+      "epoch": 1.5902151370530473,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.493989294442385e-06,
+      "loss": 1.0032,
+      "step": 8034
+    },
+    {
+      "epoch": 1.5904150321081434,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.492940914654244e-06,
+      "loss": 0.9712,
+      "step": 8035
+    },
+    {
+      "epoch": 1.5906149271632393,
+      "grad_norm": 2.125,
+      "learning_rate": 4.491892557389614e-06,
+      "loss": 1.0108,
+      "step": 8036
+    },
+    {
+      "epoch": 1.5908148222183354,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.4908442226950625e-06,
+      "loss": 1.0197,
+      "step": 8037
+    },
+    {
+      "epoch": 1.5910147172734315,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.489795910617156e-06,
+      "loss": 0.998,
+      "step": 8038
+    },
+    {
+      "epoch": 1.5912146123285276,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.488747621202461e-06,
+      "loss": 0.8976,
+      "step": 8039
+    },
+    {
+      "epoch": 1.5914145073836234,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.487699354497541e-06,
+      "loss": 0.9769,
+      "step": 8040
+    },
+    {
+      "epoch": 1.5916144024387195,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.4866511105489625e-06,
+      "loss": 0.9799,
+      "step": 8041
+    },
+    {
+      "epoch": 1.5918142974938156,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.485602889403286e-06,
+      "loss": 0.7528,
+      "step": 8042
+    },
+    {
+      "epoch": 1.5920141925489117,
+      "grad_norm": 2.25,
+      "learning_rate": 4.484554691107073e-06,
+      "loss": 0.9563,
+      "step": 8043
+    },
+    {
+      "epoch": 1.5922140876040078,
+      "grad_norm": 2.125,
+      "learning_rate": 4.4835065157068875e-06,
+      "loss": 0.9822,
+      "step": 8044
+    },
+    {
+      "epoch": 1.592413982659104,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.4824583632492866e-06,
+      "loss": 1.0397,
+      "step": 8045
+    },
+    {
+      "epoch": 1.5926138777142,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.48141023378083e-06,
+      "loss": 1.0909,
+      "step": 8046
+    },
+    {
+      "epoch": 1.5928137727692961,
+      "grad_norm": 2.125,
+      "learning_rate": 4.480362127348075e-06,
+      "loss": 0.9063,
+      "step": 8047
+    },
+    {
+      "epoch": 1.5930136678243922,
+      "grad_norm": 2.0,
+      "learning_rate": 4.479314043997579e-06,
+      "loss": 0.9052,
+      "step": 8048
+    },
+    {
+      "epoch": 1.5932135628794883,
+      "grad_norm": 2.125,
+      "learning_rate": 4.478265983775899e-06,
+      "loss": 0.9554,
+      "step": 8049
+    },
+    {
+      "epoch": 1.5934134579345844,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.4772179467295865e-06,
+      "loss": 0.919,
+      "step": 8050
+    },
+    {
+      "epoch": 1.5936133529896805,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.476169932905197e-06,
+      "loss": 0.9487,
+      "step": 8051
+    },
+    {
+      "epoch": 1.5938132480447766,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.475121942349285e-06,
+      "loss": 1.0724,
+      "step": 8052
+    },
+    {
+      "epoch": 1.5940131430998725,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.474073975108398e-06,
+      "loss": 0.9217,
+      "step": 8053
+    },
+    {
+      "epoch": 1.5942130381549686,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.47302603122909e-06,
+      "loss": 1.0427,
+      "step": 8054
+    },
+    {
+      "epoch": 1.5944129332100647,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.47197811075791e-06,
+      "loss": 0.8782,
+      "step": 8055
+    },
+    {
+      "epoch": 1.5946128282651608,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.470930213741405e-06,
+      "loss": 0.9156,
+      "step": 8056
+    },
+    {
+      "epoch": 1.594812723320257,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.469882340226124e-06,
+      "loss": 1.0549,
+      "step": 8057
+    },
+    {
+      "epoch": 1.5950126183753528,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.468834490258612e-06,
+      "loss": 0.9348,
+      "step": 8058
+    },
+    {
+      "epoch": 1.595212513430449,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.467786663885415e-06,
+      "loss": 0.959,
+      "step": 8059
+    },
+    {
+      "epoch": 1.595412408485545,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.4667388611530785e-06,
+      "loss": 1.0165,
+      "step": 8060
+    },
+    {
+      "epoch": 1.595612303540641,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.465691082108145e-06,
+      "loss": 0.9735,
+      "step": 8061
+    },
+    {
+      "epoch": 1.5958121985957372,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.464643326797155e-06,
+      "loss": 1.0059,
+      "step": 8062
+    },
+    {
+      "epoch": 1.5960120936508333,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.463595595266653e-06,
+      "loss": 1.0127,
+      "step": 8063
+    },
+    {
+      "epoch": 1.5962119887059294,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.4625478875631785e-06,
+      "loss": 0.9852,
+      "step": 8064
+    },
+    {
+      "epoch": 1.5964118837610255,
+      "grad_norm": 2.25,
+      "learning_rate": 4.4615002037332675e-06,
+      "loss": 1.0747,
+      "step": 8065
+    },
+    {
+      "epoch": 1.5966117788161216,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.460452543823459e-06,
+      "loss": 1.091,
+      "step": 8066
+    },
+    {
+      "epoch": 1.5968116738712177,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.459404907880293e-06,
+      "loss": 0.9338,
+      "step": 8067
+    },
+    {
+      "epoch": 1.5970115689263138,
+      "grad_norm": 2.125,
+      "learning_rate": 4.458357295950302e-06,
+      "loss": 0.9632,
+      "step": 8068
+    },
+    {
+      "epoch": 1.59721146398141,
+      "grad_norm": 2.375,
+      "learning_rate": 4.457309708080022e-06,
+      "loss": 0.9838,
+      "step": 8069
+    },
+    {
+      "epoch": 1.597411359036506,
+      "grad_norm": 2.125,
+      "learning_rate": 4.456262144315987e-06,
+      "loss": 0.9952,
+      "step": 8070
+    },
+    {
+      "epoch": 1.5976112540916019,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.455214604704729e-06,
+      "loss": 0.9566,
+      "step": 8071
+    },
+    {
+      "epoch": 1.597811149146698,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.454167089292781e-06,
+      "loss": 0.9163,
+      "step": 8072
+    },
+    {
+      "epoch": 1.598011044201794,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.453119598126672e-06,
+      "loss": 0.8876,
+      "step": 8073
+    },
+    {
+      "epoch": 1.5982109392568902,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.452072131252932e-06,
+      "loss": 1.0028,
+      "step": 8074
+    },
+    {
+      "epoch": 1.598410834311986,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.45102468871809e-06,
+      "loss": 0.947,
+      "step": 8075
+    },
+    {
+      "epoch": 1.5986107293670822,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.449977270568672e-06,
+      "loss": 0.9892,
+      "step": 8076
+    },
+    {
+      "epoch": 1.5988106244221783,
+      "grad_norm": 1.984375,
+      "learning_rate": 4.4489298768512065e-06,
+      "loss": 0.9009,
+      "step": 8077
+    },
+    {
+      "epoch": 1.5990105194772744,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.4478825076122155e-06,
+      "loss": 0.9217,
+      "step": 8078
+    },
+    {
+      "epoch": 1.5992104145323705,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.446835162898224e-06,
+      "loss": 0.9853,
+      "step": 8079
+    },
+    {
+      "epoch": 1.5994103095874665,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.445787842755756e-06,
+      "loss": 1.0672,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5996102046425626,
+      "grad_norm": 2.125,
+      "learning_rate": 4.444740547231334e-06,
+      "loss": 0.9832,
+      "step": 8081
+    },
+    {
+      "epoch": 1.5998100996976587,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.443693276371476e-06,
+      "loss": 0.9773,
+      "step": 8082
+    },
+    {
+      "epoch": 1.6000099947527548,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.442646030222704e-06,
+      "loss": 0.9381,
+      "step": 8083
+    },
+    {
+      "epoch": 1.600209889807851,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.441598808831536e-06,
+      "loss": 1.1051,
+      "step": 8084
+    },
+    {
+      "epoch": 1.600409784862947,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.440551612244489e-06,
+      "loss": 0.9364,
+      "step": 8085
+    },
+    {
+      "epoch": 1.6006096799180431,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.4395044405080805e-06,
+      "loss": 1.0837,
+      "step": 8086
+    },
+    {
+      "epoch": 1.6008095749731392,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.438457293668823e-06,
+      "loss": 0.9065,
+      "step": 8087
+    },
+    {
+      "epoch": 1.6010094700282351,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.437410171773234e-06,
+      "loss": 0.9178,
+      "step": 8088
+    },
+    {
+      "epoch": 1.6012093650833312,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.436363074867824e-06,
+      "loss": 0.9614,
+      "step": 8089
+    },
+    {
+      "epoch": 1.6014092601384273,
+      "grad_norm": 2.125,
+      "learning_rate": 4.435316002999107e-06,
+      "loss": 0.9914,
+      "step": 8090
+    },
+    {
+      "epoch": 1.6016091551935234,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.434268956213593e-06,
+      "loss": 0.9617,
+      "step": 8091
+    },
+    {
+      "epoch": 1.6018090502486195,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.43322193455779e-06,
+      "loss": 1.0329,
+      "step": 8092
+    },
+    {
+      "epoch": 1.6020089453037154,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.432174938078209e-06,
+      "loss": 0.9898,
+      "step": 8093
+    },
+    {
+      "epoch": 1.6022088403588115,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.431127966821358e-06,
+      "loss": 1.0177,
+      "step": 8094
+    },
+    {
+      "epoch": 1.6024087354139076,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.4300810208337425e-06,
+      "loss": 0.9509,
+      "step": 8095
+    },
+    {
+      "epoch": 1.6026086304690037,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.4290341001618645e-06,
+      "loss": 0.9996,
+      "step": 8096
+    },
+    {
+      "epoch": 1.6028085255240998,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.4279872048522316e-06,
+      "loss": 1.0083,
+      "step": 8097
+    },
+    {
+      "epoch": 1.603008420579196,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.426940334951347e-06,
+      "loss": 0.8973,
+      "step": 8098
+    },
+    {
+      "epoch": 1.603208315634292,
+      "grad_norm": 2.125,
+      "learning_rate": 4.42589349050571e-06,
+      "loss": 0.9218,
+      "step": 8099
+    },
+    {
+      "epoch": 1.603408210689388,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.424846671561824e-06,
+      "loss": 0.9154,
+      "step": 8100
+    },
+    {
+      "epoch": 1.6036081057444842,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.423799878166187e-06,
+      "loss": 1.0048,
+      "step": 8101
+    },
+    {
+      "epoch": 1.6038080007995803,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.422753110365297e-06,
+      "loss": 1.0094,
+      "step": 8102
+    },
+    {
+      "epoch": 1.6040078958546764,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.4217063682056534e-06,
+      "loss": 1.0047,
+      "step": 8103
+    },
+    {
+      "epoch": 1.6042077909097725,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.420659651733751e-06,
+      "loss": 0.9212,
+      "step": 8104
+    },
+    {
+      "epoch": 1.6044076859648686,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.419612960996083e-06,
+      "loss": 0.935,
+      "step": 8105
+    },
+    {
+      "epoch": 1.6046075810199645,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.418566296039148e-06,
+      "loss": 0.9701,
+      "step": 8106
+    },
+    {
+      "epoch": 1.6048074760750606,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.417519656909435e-06,
+      "loss": 0.9098,
+      "step": 8107
+    },
+    {
+      "epoch": 1.6050073711301567,
+      "grad_norm": 2.125,
+      "learning_rate": 4.416473043653437e-06,
+      "loss": 0.9055,
+      "step": 8108
+    },
+    {
+      "epoch": 1.6052072661852528,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.415426456317644e-06,
+      "loss": 0.9648,
+      "step": 8109
+    },
+    {
+      "epoch": 1.6054071612403487,
+      "grad_norm": 2.0,
+      "learning_rate": 4.4143798949485445e-06,
+      "loss": 0.929,
+      "step": 8110
+    },
+    {
+      "epoch": 1.6056070562954448,
+      "grad_norm": 2.125,
+      "learning_rate": 4.4133333595926285e-06,
+      "loss": 0.9921,
+      "step": 8111
+    },
+    {
+      "epoch": 1.6058069513505409,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.4122868502963815e-06,
+      "loss": 0.8996,
+      "step": 8112
+    },
+    {
+      "epoch": 1.606006846405637,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.411240367106289e-06,
+      "loss": 1.0227,
+      "step": 8113
+    },
+    {
+      "epoch": 1.606206741460733,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.410193910068838e-06,
+      "loss": 1.0014,
+      "step": 8114
+    },
+    {
+      "epoch": 1.6064066365158292,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.40914747923051e-06,
+      "loss": 0.9274,
+      "step": 8115
+    },
+    {
+      "epoch": 1.6066065315709253,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.4081010746377875e-06,
+      "loss": 0.9208,
+      "step": 8116
+    },
+    {
+      "epoch": 1.6068064266260214,
+      "grad_norm": 2.25,
+      "learning_rate": 4.407054696337153e-06,
+      "loss": 1.0311,
+      "step": 8117
+    },
+    {
+      "epoch": 1.6070063216811175,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.4060083443750844e-06,
+      "loss": 0.9027,
+      "step": 8118
+    },
+    {
+      "epoch": 1.6072062167362136,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.404962018798064e-06,
+      "loss": 0.9151,
+      "step": 8119
+    },
+    {
+      "epoch": 1.6074061117913097,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.403915719652565e-06,
+      "loss": 0.9553,
+      "step": 8120
+    },
+    {
+      "epoch": 1.6076060068464058,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.402869446985066e-06,
+      "loss": 0.9868,
+      "step": 8121
+    },
+    {
+      "epoch": 1.6078059019015019,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.401823200842043e-06,
+      "loss": 1.0076,
+      "step": 8122
+    },
+    {
+      "epoch": 1.6080057969565977,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.400776981269969e-06,
+      "loss": 1.0536,
+      "step": 8123
+    },
+    {
+      "epoch": 1.6082056920116938,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.399730788315317e-06,
+      "loss": 0.9728,
+      "step": 8124
+    },
+    {
+      "epoch": 1.60840558706679,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.398684622024562e-06,
+      "loss": 0.9305,
+      "step": 8125
+    },
+    {
+      "epoch": 1.608605482121886,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.397638482444171e-06,
+      "loss": 1.0283,
+      "step": 8126
+    },
+    {
+      "epoch": 1.6088053771769821,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.396592369620613e-06,
+      "loss": 0.9209,
+      "step": 8127
+    },
+    {
+      "epoch": 1.609005272232078,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.395546283600359e-06,
+      "loss": 1.0647,
+      "step": 8128
+    },
+    {
+      "epoch": 1.609205167287174,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.394500224429873e-06,
+      "loss": 0.9947,
+      "step": 8129
+    },
+    {
+      "epoch": 1.6094050623422702,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.393454192155621e-06,
+      "loss": 0.9625,
+      "step": 8130
+    },
+    {
+      "epoch": 1.6096049573973663,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.392408186824072e-06,
+      "loss": 0.9379,
+      "step": 8131
+    },
+    {
+      "epoch": 1.6098048524524624,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.391362208481685e-06,
+      "loss": 1.0019,
+      "step": 8132
+    },
+    {
+      "epoch": 1.6100047475075585,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.3903162571749234e-06,
+      "loss": 0.8508,
+      "step": 8133
+    },
+    {
+      "epoch": 1.6102046425626546,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.38927033295025e-06,
+      "loss": 0.9127,
+      "step": 8134
+    },
+    {
+      "epoch": 1.6104045376177507,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.388224435854121e-06,
+      "loss": 0.9434,
+      "step": 8135
+    },
+    {
+      "epoch": 1.6106044326728468,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.3871785659329985e-06,
+      "loss": 0.9151,
+      "step": 8136
+    },
+    {
+      "epoch": 1.610804327727943,
+      "grad_norm": 2.125,
+      "learning_rate": 4.386132723233339e-06,
+      "loss": 0.994,
+      "step": 8137
+    },
+    {
+      "epoch": 1.611004222783039,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.385086907801598e-06,
+      "loss": 0.932,
+      "step": 8138
+    },
+    {
+      "epoch": 1.611204117838135,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.384041119684231e-06,
+      "loss": 1.012,
+      "step": 8139
+    },
+    {
+      "epoch": 1.6114040128932312,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.382995358927691e-06,
+      "loss": 0.9277,
+      "step": 8140
+    },
+    {
+      "epoch": 1.611603907948327,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.3819496255784314e-06,
+      "loss": 0.9755,
+      "step": 8141
+    },
+    {
+      "epoch": 1.6118038030034232,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.380903919682904e-06,
+      "loss": 1.0356,
+      "step": 8142
+    },
+    {
+      "epoch": 1.6120036980585193,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.379858241287558e-06,
+      "loss": 0.979,
+      "step": 8143
+    },
+    {
+      "epoch": 1.6122035931136154,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.378812590438843e-06,
+      "loss": 0.9252,
+      "step": 8144
+    },
+    {
+      "epoch": 1.6124034881687113,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.377766967183206e-06,
+      "loss": 0.9043,
+      "step": 8145
+    },
+    {
+      "epoch": 1.6126033832238074,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.376721371567094e-06,
+      "loss": 0.9587,
+      "step": 8146
+    },
+    {
+      "epoch": 1.6128032782789035,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.375675803636953e-06,
+      "loss": 1.0145,
+      "step": 8147
+    },
+    {
+      "epoch": 1.6130031733339996,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.374630263439225e-06,
+      "loss": 0.9653,
+      "step": 8148
+    },
+    {
+      "epoch": 1.6132030683890957,
+      "grad_norm": 2.25,
+      "learning_rate": 4.3735847510203536e-06,
+      "loss": 0.9276,
+      "step": 8149
+    },
+    {
+      "epoch": 1.6134029634441918,
+      "grad_norm": 2.125,
+      "learning_rate": 4.372539266426783e-06,
+      "loss": 0.9764,
+      "step": 8150
+    },
+    {
+      "epoch": 1.6136028584992879,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.371493809704948e-06,
+      "loss": 1.016,
+      "step": 8151
+    },
+    {
+      "epoch": 1.613802753554384,
+      "grad_norm": 2.125,
+      "learning_rate": 4.3704483809012925e-06,
+      "loss": 1.0168,
+      "step": 8152
+    },
+    {
+      "epoch": 1.61400264860948,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.369402980062253e-06,
+      "loss": 0.987,
+      "step": 8153
+    },
+    {
+      "epoch": 1.6142025436645762,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.368357607234265e-06,
+      "loss": 0.9511,
+      "step": 8154
+    },
+    {
+      "epoch": 1.6144024387196723,
+      "grad_norm": 2.25,
+      "learning_rate": 4.367312262463764e-06,
+      "loss": 0.9389,
+      "step": 8155
+    },
+    {
+      "epoch": 1.6146023337747684,
+      "grad_norm": 2.125,
+      "learning_rate": 4.366266945797187e-06,
+      "loss": 1.0277,
+      "step": 8156
+    },
+    {
+      "epoch": 1.6148022288298645,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.3652216572809645e-06,
+      "loss": 1.0851,
+      "step": 8157
+    },
+    {
+      "epoch": 1.6150021238849606,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.3641763969615255e-06,
+      "loss": 0.9183,
+      "step": 8158
+    },
+    {
+      "epoch": 1.6152020189400564,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.363131164885306e-06,
+      "loss": 0.9552,
+      "step": 8159
+    },
+    {
+      "epoch": 1.6154019139951525,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.36208596109873e-06,
+      "loss": 1.0006,
+      "step": 8160
+    },
+    {
+      "epoch": 1.6156018090502486,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.361040785648227e-06,
+      "loss": 1.0061,
+      "step": 8161
+    },
+    {
+      "epoch": 1.6158017041053447,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.359995638580226e-06,
+      "loss": 0.962,
+      "step": 8162
+    },
+    {
+      "epoch": 1.6160015991604406,
+      "grad_norm": 2.25,
+      "learning_rate": 4.358950519941149e-06,
+      "loss": 1.0112,
+      "step": 8163
+    },
+    {
+      "epoch": 1.6162014942155367,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.357905429777422e-06,
+      "loss": 0.9687,
+      "step": 8164
+    },
+    {
+      "epoch": 1.6164013892706328,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.356860368135468e-06,
+      "loss": 1.0494,
+      "step": 8165
+    },
+    {
+      "epoch": 1.616601284325729,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.3558153350617065e-06,
+      "loss": 0.9395,
+      "step": 8166
+    },
+    {
+      "epoch": 1.616801179380825,
+      "grad_norm": 2.0,
+      "learning_rate": 4.354770330602559e-06,
+      "loss": 0.9022,
+      "step": 8167
+    },
+    {
+      "epoch": 1.6170010744359211,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.353725354804445e-06,
+      "loss": 0.9848,
+      "step": 8168
+    },
+    {
+      "epoch": 1.6172009694910172,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.3526804077137816e-06,
+      "loss": 1.0458,
+      "step": 8169
+    },
+    {
+      "epoch": 1.6174008645461133,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.351635489376986e-06,
+      "loss": 0.9763,
+      "step": 8170
+    },
+    {
+      "epoch": 1.6176007596012094,
+      "grad_norm": 2.125,
+      "learning_rate": 4.350590599840472e-06,
+      "loss": 0.9329,
+      "step": 8171
+    },
+    {
+      "epoch": 1.6178006546563055,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.349545739150654e-06,
+      "loss": 0.8845,
+      "step": 8172
+    },
+    {
+      "epoch": 1.6180005497114016,
+      "grad_norm": 2.125,
+      "learning_rate": 4.3485009073539445e-06,
+      "loss": 0.9565,
+      "step": 8173
+    },
+    {
+      "epoch": 1.6182004447664977,
+      "grad_norm": 2.125,
+      "learning_rate": 4.3474561044967555e-06,
+      "loss": 1.0195,
+      "step": 8174
+    },
+    {
+      "epoch": 1.6184003398215938,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.346411330625496e-06,
+      "loss": 0.9583,
+      "step": 8175
+    },
+    {
+      "epoch": 1.6186002348766897,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.345366585786577e-06,
+      "loss": 1.0182,
+      "step": 8176
+    },
+    {
+      "epoch": 1.6188001299317858,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.344321870026404e-06,
+      "loss": 1.0077,
+      "step": 8177
+    },
+    {
+      "epoch": 1.619000024986882,
+      "grad_norm": 2.125,
+      "learning_rate": 4.343277183391384e-06,
+      "loss": 0.9333,
+      "step": 8178
+    },
+    {
+      "epoch": 1.619199920041978,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.342232525927919e-06,
+      "loss": 0.8814,
+      "step": 8179
+    },
+    {
+      "epoch": 1.619399815097074,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.341187897682416e-06,
+      "loss": 0.9667,
+      "step": 8180
+    },
+    {
+      "epoch": 1.61959971015217,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.3401432987012775e-06,
+      "loss": 1.0133,
+      "step": 8181
+    },
+    {
+      "epoch": 1.619799605207266,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.339098729030902e-06,
+      "loss": 0.913,
+      "step": 8182
+    },
+    {
+      "epoch": 1.6199995002623622,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.3380541887176904e-06,
+      "loss": 1.0063,
+      "step": 8183
+    },
+    {
+      "epoch": 1.6201993953174583,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.337009677808042e-06,
+      "loss": 0.9996,
+      "step": 8184
+    },
+    {
+      "epoch": 1.6203992903725544,
+      "grad_norm": 2.25,
+      "learning_rate": 4.335965196348352e-06,
+      "loss": 0.9258,
+      "step": 8185
+    },
+    {
+      "epoch": 1.6205991854276505,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.334920744385017e-06,
+      "loss": 0.9185,
+      "step": 8186
+    },
+    {
+      "epoch": 1.6207990804827466,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.3338763219644335e-06,
+      "loss": 1.129,
+      "step": 8187
+    },
+    {
+      "epoch": 1.6209989755378427,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.332831929132991e-06,
+      "loss": 0.9931,
+      "step": 8188
+    },
+    {
+      "epoch": 1.6211988705929388,
+      "grad_norm": 2.015625,
+      "learning_rate": 4.331787565937082e-06,
+      "loss": 0.8996,
+      "step": 8189
+    },
+    {
+      "epoch": 1.6213987656480349,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.330743232423101e-06,
+      "loss": 0.956,
+      "step": 8190
+    },
+    {
+      "epoch": 1.621598660703131,
+      "grad_norm": 2.125,
+      "learning_rate": 4.32969892863743e-06,
+      "loss": 0.9862,
+      "step": 8191
+    },
+    {
+      "epoch": 1.621798555758227,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.328654654626463e-06,
+      "loss": 1.0203,
+      "step": 8192
+    },
+    {
+      "epoch": 1.6219984508133232,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.3276104104365855e-06,
+      "loss": 0.9044,
+      "step": 8193
+    },
+    {
+      "epoch": 1.622198345868419,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.32656619611418e-06,
+      "loss": 0.9614,
+      "step": 8194
+    },
+    {
+      "epoch": 1.6223982409235151,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.3255220117056325e-06,
+      "loss": 0.9691,
+      "step": 8195
+    },
+    {
+      "epoch": 1.6225981359786112,
+      "grad_norm": 2.125,
+      "learning_rate": 4.324477857257326e-06,
+      "loss": 0.9605,
+      "step": 8196
+    },
+    {
+      "epoch": 1.6227980310337073,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.323433732815641e-06,
+      "loss": 1.0356,
+      "step": 8197
+    },
+    {
+      "epoch": 1.6229979260888032,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.322389638426957e-06,
+      "loss": 0.8856,
+      "step": 8198
+    },
+    {
+      "epoch": 1.6231978211438993,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.321345574137652e-06,
+      "loss": 0.9635,
+      "step": 8199
+    },
+    {
+      "epoch": 1.6233977161989954,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.320301539994105e-06,
+      "loss": 0.9722,
+      "step": 8200
+    },
+    {
+      "epoch": 1.6235976112540915,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.319257536042692e-06,
+      "loss": 0.9877,
+      "step": 8201
+    },
+    {
+      "epoch": 1.6237975063091876,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.318213562329784e-06,
+      "loss": 0.8835,
+      "step": 8202
+    },
+    {
+      "epoch": 1.6239974013642837,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.317169618901758e-06,
+      "loss": 0.9898,
+      "step": 8203
+    },
+    {
+      "epoch": 1.6241972964193798,
+      "grad_norm": 2.125,
+      "learning_rate": 4.316125705804986e-06,
+      "loss": 0.8772,
+      "step": 8204
+    },
+    {
+      "epoch": 1.624397191474476,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.315081823085835e-06,
+      "loss": 1.0199,
+      "step": 8205
+    },
+    {
+      "epoch": 1.624597086529572,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.3140379707906765e-06,
+      "loss": 1.0246,
+      "step": 8206
+    },
+    {
+      "epoch": 1.6247969815846681,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.312994148965879e-06,
+      "loss": 0.912,
+      "step": 8207
+    },
+    {
+      "epoch": 1.6249968766397642,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.311950357657807e-06,
+      "loss": 0.9617,
+      "step": 8208
+    },
+    {
+      "epoch": 1.6251967716948603,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.310906596912828e-06,
+      "loss": 0.9523,
+      "step": 8209
+    },
+    {
+      "epoch": 1.6253966667499564,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.309862866777303e-06,
+      "loss": 0.9298,
+      "step": 8210
+    },
+    {
+      "epoch": 1.6255965618050523,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.3088191672975965e-06,
+      "loss": 0.9174,
+      "step": 8211
+    },
+    {
+      "epoch": 1.6257964568601484,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.30777549852007e-06,
+      "loss": 1.0472,
+      "step": 8212
+    },
+    {
+      "epoch": 1.6259963519152445,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.30673186049108e-06,
+      "loss": 1.0767,
+      "step": 8213
+    },
+    {
+      "epoch": 1.6261962469703406,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.305688253256986e-06,
+      "loss": 0.9619,
+      "step": 8214
+    },
+    {
+      "epoch": 1.6263961420254367,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.304644676864149e-06,
+      "loss": 1.0446,
+      "step": 8215
+    },
+    {
+      "epoch": 1.6265960370805326,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.303601131358918e-06,
+      "loss": 1.0276,
+      "step": 8216
+    },
+    {
+      "epoch": 1.6267959321356287,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.302557616787652e-06,
+      "loss": 0.953,
+      "step": 8217
+    },
+    {
+      "epoch": 1.6269958271907248,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.3015141331967045e-06,
+      "loss": 0.9412,
+      "step": 8218
+    },
+    {
+      "epoch": 1.6271957222458209,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.300470680632421e-06,
+      "loss": 0.9793,
+      "step": 8219
+    },
+    {
+      "epoch": 1.627395617300917,
+      "grad_norm": 2.0,
+      "learning_rate": 4.299427259141155e-06,
+      "loss": 0.9653,
+      "step": 8220
+    },
+    {
+      "epoch": 1.627595512356013,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.298383868769257e-06,
+      "loss": 0.9435,
+      "step": 8221
+    },
+    {
+      "epoch": 1.6277954074111092,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.297340509563072e-06,
+      "loss": 0.9783,
+      "step": 8222
+    },
+    {
+      "epoch": 1.6279953024662053,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.296297181568946e-06,
+      "loss": 1.0285,
+      "step": 8223
+    },
+    {
+      "epoch": 1.6281951975213014,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.295253884833225e-06,
+      "loss": 0.9742,
+      "step": 8224
+    },
+    {
+      "epoch": 1.6283950925763975,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.29421061940225e-06,
+      "loss": 0.9601,
+      "step": 8225
+    },
+    {
+      "epoch": 1.6285949876314936,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.293167385322364e-06,
+      "loss": 0.9869,
+      "step": 8226
+    },
+    {
+      "epoch": 1.6287948826865897,
+      "grad_norm": 2.125,
+      "learning_rate": 4.292124182639909e-06,
+      "loss": 0.9445,
+      "step": 8227
+    },
+    {
+      "epoch": 1.6289947777416858,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.29108101140122e-06,
+      "loss": 0.9883,
+      "step": 8228
+    },
+    {
+      "epoch": 1.6291946727967817,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.290037871652639e-06,
+      "loss": 0.9173,
+      "step": 8229
+    },
+    {
+      "epoch": 1.6293945678518778,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.288994763440498e-06,
+      "loss": 1.0868,
+      "step": 8230
+    },
+    {
+      "epoch": 1.6295944629069739,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.2879516868111346e-06,
+      "loss": 1.0022,
+      "step": 8231
+    },
+    {
+      "epoch": 1.62979435796207,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.2869086418108815e-06,
+      "loss": 0.9782,
+      "step": 8232
+    },
+    {
+      "epoch": 1.6299942530171658,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.285865628486069e-06,
+      "loss": 0.9706,
+      "step": 8233
+    },
+    {
+      "epoch": 1.630194148072262,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.2848226468830295e-06,
+      "loss": 0.9746,
+      "step": 8234
+    },
+    {
+      "epoch": 1.630394043127358,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.283779697048093e-06,
+      "loss": 0.9263,
+      "step": 8235
+    },
+    {
+      "epoch": 1.6305939381824541,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.282736779027584e-06,
+      "loss": 1.057,
+      "step": 8236
+    },
+    {
+      "epoch": 1.6307938332375502,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.281693892867832e-06,
+      "loss": 0.8657,
+      "step": 8237
+    },
+    {
+      "epoch": 1.6309937282926463,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.28065103861516e-06,
+      "loss": 0.9345,
+      "step": 8238
+    },
+    {
+      "epoch": 1.6311936233477424,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.2796082163158914e-06,
+      "loss": 0.9327,
+      "step": 8239
+    },
+    {
+      "epoch": 1.6313935184028385,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.278565426016351e-06,
+      "loss": 0.8978,
+      "step": 8240
+    },
+    {
+      "epoch": 1.6315934134579346,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.277522667762855e-06,
+      "loss": 0.9943,
+      "step": 8241
+    },
+    {
+      "epoch": 1.6317933085130307,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.276479941601726e-06,
+      "loss": 0.8993,
+      "step": 8242
+    },
+    {
+      "epoch": 1.6319932035681268,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.275437247579281e-06,
+      "loss": 1.0119,
+      "step": 8243
+    },
+    {
+      "epoch": 1.632193098623223,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.274394585741835e-06,
+      "loss": 1.0501,
+      "step": 8244
+    },
+    {
+      "epoch": 1.632392993678319,
+      "grad_norm": 2.25,
+      "learning_rate": 4.273351956135704e-06,
+      "loss": 1.005,
+      "step": 8245
+    },
+    {
+      "epoch": 1.632592888733415,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.272309358807203e-06,
+      "loss": 1.0611,
+      "step": 8246
+    },
+    {
+      "epoch": 1.632792783788511,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.271266793802641e-06,
+      "loss": 1.0112,
+      "step": 8247
+    },
+    {
+      "epoch": 1.632992678843607,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.270224261168332e-06,
+      "loss": 1.0161,
+      "step": 8248
+    },
+    {
+      "epoch": 1.6331925738987032,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.269181760950584e-06,
+      "loss": 0.9708,
+      "step": 8249
+    },
+    {
+      "epoch": 1.6333924689537993,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.268139293195702e-06,
+      "loss": 0.9985,
+      "step": 8250
+    },
+    {
+      "epoch": 1.6335923640088952,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.267096857949994e-06,
+      "loss": 1.0267,
+      "step": 8251
+    },
+    {
+      "epoch": 1.6337922590639913,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.266054455259767e-06,
+      "loss": 0.9011,
+      "step": 8252
+    },
+    {
+      "epoch": 1.6339921541190874,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.265012085171322e-06,
+      "loss": 0.9273,
+      "step": 8253
+    },
+    {
+      "epoch": 1.6341920491741835,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.263969747730961e-06,
+      "loss": 0.8469,
+      "step": 8254
+    },
+    {
+      "epoch": 1.6343919442292796,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.262927442984986e-06,
+      "loss": 0.91,
+      "step": 8255
+    },
+    {
+      "epoch": 1.6345918392843757,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.261885170979695e-06,
+      "loss": 1.0107,
+      "step": 8256
+    },
+    {
+      "epoch": 1.6347917343394718,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.260842931761385e-06,
+      "loss": 0.9458,
+      "step": 8257
+    },
+    {
+      "epoch": 1.6349916293945679,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.259800725376355e-06,
+      "loss": 0.9503,
+      "step": 8258
+    },
+    {
+      "epoch": 1.635191524449664,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.258758551870896e-06,
+      "loss": 0.9429,
+      "step": 8259
+    },
+    {
+      "epoch": 1.63539141950476,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.257716411291304e-06,
+      "loss": 0.9112,
+      "step": 8260
+    },
+    {
+      "epoch": 1.6355913145598562,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.256674303683869e-06,
+      "loss": 1.0088,
+      "step": 8261
+    },
+    {
+      "epoch": 1.6357912096149523,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.255632229094882e-06,
+      "loss": 0.9294,
+      "step": 8262
+    },
+    {
+      "epoch": 1.6359911046700484,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.254590187570633e-06,
+      "loss": 0.9696,
+      "step": 8263
+    },
+    {
+      "epoch": 1.6361909997251443,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.253548179157407e-06,
+      "loss": 0.995,
+      "step": 8264
+    },
+    {
+      "epoch": 1.6363908947802404,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.252506203901491e-06,
+      "loss": 0.9733,
+      "step": 8265
+    },
+    {
+      "epoch": 1.6365907898353365,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.251464261849171e-06,
+      "loss": 0.9563,
+      "step": 8266
+    },
+    {
+      "epoch": 1.6367906848904326,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.2504223530467275e-06,
+      "loss": 1.0219,
+      "step": 8267
+    },
+    {
+      "epoch": 1.6369905799455284,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.249380477540444e-06,
+      "loss": 0.9858,
+      "step": 8268
+    },
+    {
+      "epoch": 1.6371904750006245,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.248338635376599e-06,
+      "loss": 0.9386,
+      "step": 8269
+    },
+    {
+      "epoch": 1.6373903700557206,
+      "grad_norm": 2.25,
+      "learning_rate": 4.247296826601471e-06,
+      "loss": 1.0425,
+      "step": 8270
+    },
+    {
+      "epoch": 1.6375902651108167,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.246255051261338e-06,
+      "loss": 0.9948,
+      "step": 8271
+    },
+    {
+      "epoch": 1.6377901601659128,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.245213309402475e-06,
+      "loss": 0.9659,
+      "step": 8272
+    },
+    {
+      "epoch": 1.637990055221009,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.244171601071156e-06,
+      "loss": 1.1023,
+      "step": 8273
+    },
+    {
+      "epoch": 1.638189950276105,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.243129926313654e-06,
+      "loss": 0.9661,
+      "step": 8274
+    },
+    {
+      "epoch": 1.6383898453312011,
+      "grad_norm": 2.125,
+      "learning_rate": 4.242088285176239e-06,
+      "loss": 0.9441,
+      "step": 8275
+    },
+    {
+      "epoch": 1.6385897403862972,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.241046677705183e-06,
+      "loss": 1.0038,
+      "step": 8276
+    },
+    {
+      "epoch": 1.6387896354413933,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.240005103946751e-06,
+      "loss": 0.9978,
+      "step": 8277
+    },
+    {
+      "epoch": 1.6389895304964894,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.238963563947212e-06,
+      "loss": 0.9425,
+      "step": 8278
+    },
+    {
+      "epoch": 1.6391894255515855,
+      "grad_norm": 2.25,
+      "learning_rate": 4.237922057752831e-06,
+      "loss": 1.0408,
+      "step": 8279
+    },
+    {
+      "epoch": 1.6393893206066816,
+      "grad_norm": 2.125,
+      "learning_rate": 4.236880585409872e-06,
+      "loss": 0.9364,
+      "step": 8280
+    },
+    {
+      "epoch": 1.6395892156617777,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.235839146964592e-06,
+      "loss": 0.9781,
+      "step": 8281
+    },
+    {
+      "epoch": 1.6397891107168736,
+      "grad_norm": 2.25,
+      "learning_rate": 4.234797742463258e-06,
+      "loss": 1.0187,
+      "step": 8282
+    },
+    {
+      "epoch": 1.6399890057719697,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.2337563719521254e-06,
+      "loss": 0.9255,
+      "step": 8283
+    },
+    {
+      "epoch": 1.6401889008270658,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.2327150354774536e-06,
+      "loss": 0.9822,
+      "step": 8284
+    },
+    {
+      "epoch": 1.640388795882162,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.231673733085497e-06,
+      "loss": 0.9543,
+      "step": 8285
+    },
+    {
+      "epoch": 1.6405886909372578,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.230632464822513e-06,
+      "loss": 1.0042,
+      "step": 8286
+    },
+    {
+      "epoch": 1.6407885859923539,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.229591230734751e-06,
+      "loss": 0.8565,
+      "step": 8287
+    },
+    {
+      "epoch": 1.64098848104745,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.228550030868465e-06,
+      "loss": 0.9459,
+      "step": 8288
+    },
+    {
+      "epoch": 1.641188376102546,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.227508865269904e-06,
+      "loss": 1.0064,
+      "step": 8289
+    },
+    {
+      "epoch": 1.6413882711576422,
+      "grad_norm": 2.25,
+      "learning_rate": 4.226467733985316e-06,
+      "loss": 0.999,
+      "step": 8290
+    },
+    {
+      "epoch": 1.6415881662127383,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.22542663706095e-06,
+      "loss": 0.8225,
+      "step": 8291
+    },
+    {
+      "epoch": 1.6417880612678344,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.22438557454305e-06,
+      "loss": 0.9228,
+      "step": 8292
+    },
+    {
+      "epoch": 1.6419879563229305,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.223344546477858e-06,
+      "loss": 0.9381,
+      "step": 8293
+    },
+    {
+      "epoch": 1.6421878513780266,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.2223035529116205e-06,
+      "loss": 1.005,
+      "step": 8294
+    },
+    {
+      "epoch": 1.6423877464331227,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.2212625938905745e-06,
+      "loss": 0.8917,
+      "step": 8295
+    },
+    {
+      "epoch": 1.6425876414882188,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.220221669460962e-06,
+      "loss": 1.0419,
+      "step": 8296
+    },
+    {
+      "epoch": 1.6427875365433149,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.21918077966902e-06,
+      "loss": 0.9797,
+      "step": 8297
+    },
+    {
+      "epoch": 1.642987431598411,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.218139924560983e-06,
+      "loss": 0.9871,
+      "step": 8298
+    },
+    {
+      "epoch": 1.6431873266535069,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.217099104183089e-06,
+      "loss": 0.9627,
+      "step": 8299
+    },
+    {
+      "epoch": 1.643387221708603,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.216058318581567e-06,
+      "loss": 0.9994,
+      "step": 8300
+    },
+    {
+      "epoch": 1.643587116763699,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.215017567802651e-06,
+      "loss": 0.9186,
+      "step": 8301
+    },
+    {
+      "epoch": 1.6437870118187952,
+      "grad_norm": 2.25,
+      "learning_rate": 4.213976851892573e-06,
+      "loss": 0.9584,
+      "step": 8302
+    },
+    {
+      "epoch": 1.643986906873891,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.212936170897557e-06,
+      "loss": 1.0087,
+      "step": 8303
+    },
+    {
+      "epoch": 1.6441868019289871,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.211895524863832e-06,
+      "loss": 1.0012,
+      "step": 8304
+    },
+    {
+      "epoch": 1.6443866969840832,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.210854913837625e-06,
+      "loss": 0.9215,
+      "step": 8305
+    },
+    {
+      "epoch": 1.6445865920391793,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.209814337865158e-06,
+      "loss": 0.9373,
+      "step": 8306
+    },
+    {
+      "epoch": 1.6447864870942754,
+      "grad_norm": 2.125,
+      "learning_rate": 4.2087737969926545e-06,
+      "loss": 0.8611,
+      "step": 8307
+    },
+    {
+      "epoch": 1.6449863821493715,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.2077332912663335e-06,
+      "loss": 0.8602,
+      "step": 8308
+    },
+    {
+      "epoch": 1.6451862772044676,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.206692820732415e-06,
+      "loss": 1.0522,
+      "step": 8309
+    },
+    {
+      "epoch": 1.6453861722595637,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.205652385437118e-06,
+      "loss": 0.9538,
+      "step": 8310
+    },
+    {
+      "epoch": 1.6455860673146598,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.204611985426657e-06,
+      "loss": 1.025,
+      "step": 8311
+    },
+    {
+      "epoch": 1.645785962369756,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.203571620747246e-06,
+      "loss": 1.0084,
+      "step": 8312
+    },
+    {
+      "epoch": 1.645985857424852,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.202531291445098e-06,
+      "loss": 0.9268,
+      "step": 8313
+    },
+    {
+      "epoch": 1.6461857524799481,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.201490997566426e-06,
+      "loss": 0.9529,
+      "step": 8314
+    },
+    {
+      "epoch": 1.6463856475350442,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.200450739157437e-06,
+      "loss": 1.0054,
+      "step": 8315
+    },
+    {
+      "epoch": 1.6465855425901403,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.199410516264342e-06,
+      "loss": 0.9759,
+      "step": 8316
+    },
+    {
+      "epoch": 1.6467854376452362,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.198370328933346e-06,
+      "loss": 1.0142,
+      "step": 8317
+    },
+    {
+      "epoch": 1.6469853327003323,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.197330177210654e-06,
+      "loss": 0.9481,
+      "step": 8318
+    },
+    {
+      "epoch": 1.6471852277554284,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.19629006114247e-06,
+      "loss": 0.9881,
+      "step": 8319
+    },
+    {
+      "epoch": 1.6473851228105245,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.195249980774997e-06,
+      "loss": 0.906,
+      "step": 8320
+    },
+    {
+      "epoch": 1.6475850178656204,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.194209936154431e-06,
+      "loss": 1.0061,
+      "step": 8321
+    },
+    {
+      "epoch": 1.6477849129207165,
+      "grad_norm": 2.25,
+      "learning_rate": 4.1931699273269765e-06,
+      "loss": 0.9856,
+      "step": 8322
+    },
+    {
+      "epoch": 1.6479848079758126,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.192129954338826e-06,
+      "loss": 1.0957,
+      "step": 8323
+    },
+    {
+      "epoch": 1.6481847030309087,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.191090017236177e-06,
+      "loss": 0.9819,
+      "step": 8324
+    },
+    {
+      "epoch": 1.6483845980860048,
+      "grad_norm": 2.375,
+      "learning_rate": 4.190050116065224e-06,
+      "loss": 1.0189,
+      "step": 8325
+    },
+    {
+      "epoch": 1.648584493141101,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.1890102508721565e-06,
+      "loss": 0.9359,
+      "step": 8326
+    },
+    {
+      "epoch": 1.648784388196197,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.187970421703168e-06,
+      "loss": 0.9377,
+      "step": 8327
+    },
+    {
+      "epoch": 1.648984283251293,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.186930628604447e-06,
+      "loss": 1.0538,
+      "step": 8328
+    },
+    {
+      "epoch": 1.6491841783063892,
+      "grad_norm": 2.125,
+      "learning_rate": 4.18589087162218e-06,
+      "loss": 0.9485,
+      "step": 8329
+    },
+    {
+      "epoch": 1.6493840733614853,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.184851150802554e-06,
+      "loss": 0.9153,
+      "step": 8330
+    },
+    {
+      "epoch": 1.6495839684165814,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.1838114661917525e-06,
+      "loss": 1.0567,
+      "step": 8331
+    },
+    {
+      "epoch": 1.6497838634716775,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.1827718178359586e-06,
+      "loss": 0.9752,
+      "step": 8332
+    },
+    {
+      "epoch": 1.6499837585267736,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.181732205781354e-06,
+      "loss": 0.9755,
+      "step": 8333
+    },
+    {
+      "epoch": 1.6501836535818695,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.180692630074116e-06,
+      "loss": 1.0196,
+      "step": 8334
+    },
+    {
+      "epoch": 1.6503835486369656,
+      "grad_norm": 2.125,
+      "learning_rate": 4.179653090760424e-06,
+      "loss": 0.8684,
+      "step": 8335
+    },
+    {
+      "epoch": 1.6505834436920617,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.178613587886455e-06,
+      "loss": 1.0431,
+      "step": 8336
+    },
+    {
+      "epoch": 1.6507833387471578,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.177574121498382e-06,
+      "loss": 0.9899,
+      "step": 8337
+    },
+    {
+      "epoch": 1.6509832338022539,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.17653469164238e-06,
+      "loss": 0.9906,
+      "step": 8338
+    },
+    {
+      "epoch": 1.6511831288573497,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.175495298364618e-06,
+      "loss": 0.9755,
+      "step": 8339
+    },
+    {
+      "epoch": 1.6513830239124458,
+      "grad_norm": 2.125,
+      "learning_rate": 4.174455941711266e-06,
+      "loss": 0.99,
+      "step": 8340
+    },
+    {
+      "epoch": 1.651582918967542,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.173416621728495e-06,
+      "loss": 0.9646,
+      "step": 8341
+    },
+    {
+      "epoch": 1.651782814022638,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.17237733846247e-06,
+      "loss": 0.9579,
+      "step": 8342
+    },
+    {
+      "epoch": 1.6519827090777341,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.1713380919593525e-06,
+      "loss": 0.926,
+      "step": 8343
+    },
+    {
+      "epoch": 1.6521826041328302,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.170298882265309e-06,
+      "loss": 1.0695,
+      "step": 8344
+    },
+    {
+      "epoch": 1.6523824991879263,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.169259709426502e-06,
+      "loss": 0.988,
+      "step": 8345
+    },
+    {
+      "epoch": 1.6525823942430224,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.168220573489088e-06,
+      "loss": 0.9672,
+      "step": 8346
+    },
+    {
+      "epoch": 1.6527822892981185,
+      "grad_norm": 2.25,
+      "learning_rate": 4.167181474499228e-06,
+      "loss": 0.9888,
+      "step": 8347
+    },
+    {
+      "epoch": 1.6529821843532146,
+      "grad_norm": 2.125,
+      "learning_rate": 4.166142412503078e-06,
+      "loss": 1.0004,
+      "step": 8348
+    },
+    {
+      "epoch": 1.6531820794083107,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.1651033875467935e-06,
+      "loss": 0.8566,
+      "step": 8349
+    },
+    {
+      "epoch": 1.6533819744634068,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.1640643996765275e-06,
+      "loss": 1.0011,
+      "step": 8350
+    },
+    {
+      "epoch": 1.653581869518503,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.163025448938431e-06,
+      "loss": 0.9704,
+      "step": 8351
+    },
+    {
+      "epoch": 1.6537817645735988,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.1619865353786535e-06,
+      "loss": 0.9291,
+      "step": 8352
+    },
+    {
+      "epoch": 1.653981659628695,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.160947659043347e-06,
+      "loss": 0.9484,
+      "step": 8353
+    },
+    {
+      "epoch": 1.654181554683791,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.159908819978654e-06,
+      "loss": 0.9807,
+      "step": 8354
+    },
+    {
+      "epoch": 1.6543814497388871,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.158870018230722e-06,
+      "loss": 0.8579,
+      "step": 8355
+    },
+    {
+      "epoch": 1.654581344793983,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.157831253845695e-06,
+      "loss": 0.9327,
+      "step": 8356
+    },
+    {
+      "epoch": 1.654781239849079,
+      "grad_norm": 2.25,
+      "learning_rate": 4.156792526869712e-06,
+      "loss": 0.9865,
+      "step": 8357
+    },
+    {
+      "epoch": 1.6549811349041752,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.155753837348917e-06,
+      "loss": 0.9142,
+      "step": 8358
+    },
+    {
+      "epoch": 1.6551810299592713,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.154715185329445e-06,
+      "loss": 0.9701,
+      "step": 8359
+    },
+    {
+      "epoch": 1.6553809250143674,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.153676570857434e-06,
+      "loss": 0.962,
+      "step": 8360
+    },
+    {
+      "epoch": 1.6555808200694635,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.152637993979019e-06,
+      "loss": 1.0536,
+      "step": 8361
+    },
+    {
+      "epoch": 1.6557807151245596,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.151599454740335e-06,
+      "loss": 1.0314,
+      "step": 8362
+    },
+    {
+      "epoch": 1.6559806101796557,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.150560953187511e-06,
+      "loss": 0.8813,
+      "step": 8363
+    },
+    {
+      "epoch": 1.6561805052347518,
+      "grad_norm": 2.25,
+      "learning_rate": 4.149522489366681e-06,
+      "loss": 1.0078,
+      "step": 8364
+    },
+    {
+      "epoch": 1.656380400289848,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.148484063323969e-06,
+      "loss": 1.0463,
+      "step": 8365
+    },
+    {
+      "epoch": 1.656580295344944,
+      "grad_norm": 1.984375,
+      "learning_rate": 4.147445675105506e-06,
+      "loss": 0.8978,
+      "step": 8366
+    },
+    {
+      "epoch": 1.65678019040004,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.146407324757414e-06,
+      "loss": 0.8991,
+      "step": 8367
+    },
+    {
+      "epoch": 1.6569800854551362,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.145369012325816e-06,
+      "loss": 0.9344,
+      "step": 8368
+    },
+    {
+      "epoch": 1.657179980510232,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.1443307378568385e-06,
+      "loss": 0.9609,
+      "step": 8369
+    },
+    {
+      "epoch": 1.6573798755653282,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.143292501396596e-06,
+      "loss": 1.0138,
+      "step": 8370
+    },
+    {
+      "epoch": 1.6575797706204243,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.142254302991209e-06,
+      "loss": 1.0597,
+      "step": 8371
+    },
+    {
+      "epoch": 1.6577796656755204,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.141216142686795e-06,
+      "loss": 0.9977,
+      "step": 8372
+    },
+    {
+      "epoch": 1.6579795607306165,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.14017802052947e-06,
+      "loss": 0.9488,
+      "step": 8373
+    },
+    {
+      "epoch": 1.6581794557857124,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.139139936565343e-06,
+      "loss": 0.976,
+      "step": 8374
+    },
+    {
+      "epoch": 1.6583793508408085,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.138101890840528e-06,
+      "loss": 0.915,
+      "step": 8375
+    },
+    {
+      "epoch": 1.6585792458959046,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.137063883401137e-06,
+      "loss": 0.9585,
+      "step": 8376
+    },
+    {
+      "epoch": 1.6587791409510007,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.136025914293274e-06,
+      "loss": 0.9513,
+      "step": 8377
+    },
+    {
+      "epoch": 1.6589790360060968,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.1349879835630486e-06,
+      "loss": 0.955,
+      "step": 8378
+    },
+    {
+      "epoch": 1.6591789310611929,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.133950091256564e-06,
+      "loss": 1.0102,
+      "step": 8379
+    },
+    {
+      "epoch": 1.659378826116289,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.1329122374199234e-06,
+      "loss": 0.9872,
+      "step": 8380
+    },
+    {
+      "epoch": 1.659578721171385,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.13187442209923e-06,
+      "loss": 0.9144,
+      "step": 8381
+    },
+    {
+      "epoch": 1.6597786162264812,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.130836645340581e-06,
+      "loss": 0.9445,
+      "step": 8382
+    },
+    {
+      "epoch": 1.6599785112815773,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.129798907190076e-06,
+      "loss": 0.9533,
+      "step": 8383
+    },
+    {
+      "epoch": 1.6601784063366734,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.12876120769381e-06,
+      "loss": 0.8938,
+      "step": 8384
+    },
+    {
+      "epoch": 1.6603783013917695,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.127723546897879e-06,
+      "loss": 0.9748,
+      "step": 8385
+    },
+    {
+      "epoch": 1.6605781964468656,
+      "grad_norm": 2.125,
+      "learning_rate": 4.126685924848373e-06,
+      "loss": 0.9792,
+      "step": 8386
+    },
+    {
+      "epoch": 1.6607780915019614,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.125648341591387e-06,
+      "loss": 0.9547,
+      "step": 8387
+    },
+    {
+      "epoch": 1.6609779865570575,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.124610797173008e-06,
+      "loss": 0.9843,
+      "step": 8388
+    },
+    {
+      "epoch": 1.6611778816121536,
+      "grad_norm": 2.25,
+      "learning_rate": 4.123573291639323e-06,
+      "loss": 0.9881,
+      "step": 8389
+    },
+    {
+      "epoch": 1.6613777766672497,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.12253582503642e-06,
+      "loss": 1.0074,
+      "step": 8390
+    },
+    {
+      "epoch": 1.6615776717223456,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.12149839741038e-06,
+      "loss": 0.9708,
+      "step": 8391
+    },
+    {
+      "epoch": 1.6617775667774417,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.120461008807289e-06,
+      "loss": 0.9391,
+      "step": 8392
+    },
+    {
+      "epoch": 1.6619774618325378,
+      "grad_norm": 2.125,
+      "learning_rate": 4.119423659273226e-06,
+      "loss": 0.975,
+      "step": 8393
+    },
+    {
+      "epoch": 1.662177356887634,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.1183863488542686e-06,
+      "loss": 0.8772,
+      "step": 8394
+    },
+    {
+      "epoch": 1.66237725194273,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.117349077596497e-06,
+      "loss": 1.0748,
+      "step": 8395
+    },
+    {
+      "epoch": 1.662577146997826,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.116311845545983e-06,
+      "loss": 1.0191,
+      "step": 8396
+    },
+    {
+      "epoch": 1.6627770420529222,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.115274652748806e-06,
+      "loss": 0.9597,
+      "step": 8397
+    },
+    {
+      "epoch": 1.6629769371080183,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.114237499251031e-06,
+      "loss": 0.8679,
+      "step": 8398
+    },
+    {
+      "epoch": 1.6631768321631144,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.113200385098733e-06,
+      "loss": 0.9605,
+      "step": 8399
+    },
+    {
+      "epoch": 1.6633767272182105,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.112163310337981e-06,
+      "loss": 0.8945,
+      "step": 8400
+    },
+    {
+      "epoch": 1.6635766222733066,
+      "grad_norm": 2.125,
+      "learning_rate": 4.1111262750148375e-06,
+      "loss": 0.8471,
+      "step": 8401
+    },
+    {
+      "epoch": 1.6637765173284027,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.11008927917537e-06,
+      "loss": 0.9585,
+      "step": 8402
+    },
+    {
+      "epoch": 1.6639764123834988,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.109052322865643e-06,
+      "loss": 1.0087,
+      "step": 8403
+    },
+    {
+      "epoch": 1.664176307438595,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.1080154061317175e-06,
+      "loss": 0.8538,
+      "step": 8404
+    },
+    {
+      "epoch": 1.6643762024936908,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.1069785290196505e-06,
+      "loss": 0.9302,
+      "step": 8405
+    },
+    {
+      "epoch": 1.6645760975487869,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.105941691575502e-06,
+      "loss": 1.0174,
+      "step": 8406
+    },
+    {
+      "epoch": 1.664775992603883,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.104904893845328e-06,
+      "loss": 1.0642,
+      "step": 8407
+    },
+    {
+      "epoch": 1.664975887658979,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.103868135875184e-06,
+      "loss": 0.8731,
+      "step": 8408
+    },
+    {
+      "epoch": 1.665175782714075,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.102831417711121e-06,
+      "loss": 1.0226,
+      "step": 8409
+    },
+    {
+      "epoch": 1.665375677769171,
+      "grad_norm": 2.375,
+      "learning_rate": 4.1017947393991905e-06,
+      "loss": 1.0029,
+      "step": 8410
+    },
+    {
+      "epoch": 1.6655755728242672,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.100758100985441e-06,
+      "loss": 1.0093,
+      "step": 8411
+    },
+    {
+      "epoch": 1.6657754678793633,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.099721502515922e-06,
+      "loss": 1.0087,
+      "step": 8412
+    },
+    {
+      "epoch": 1.6659753629344594,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.098684944036677e-06,
+      "loss": 1.006,
+      "step": 8413
+    },
+    {
+      "epoch": 1.6661752579895555,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.09764842559375e-06,
+      "loss": 1.0076,
+      "step": 8414
+    },
+    {
+      "epoch": 1.6663751530446516,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.096611947233185e-06,
+      "loss": 0.9425,
+      "step": 8415
+    },
+    {
+      "epoch": 1.6665750480997477,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.09557550900102e-06,
+      "loss": 0.9662,
+      "step": 8416
+    },
+    {
+      "epoch": 1.6667749431548438,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.094539110943295e-06,
+      "loss": 1.0501,
+      "step": 8417
+    },
+    {
+      "epoch": 1.6669748382099399,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.093502753106044e-06,
+      "loss": 0.8633,
+      "step": 8418
+    },
+    {
+      "epoch": 1.667174733265036,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.092466435535306e-06,
+      "loss": 0.939,
+      "step": 8419
+    },
+    {
+      "epoch": 1.667374628320132,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.091430158277112e-06,
+      "loss": 1.0204,
+      "step": 8420
+    },
+    {
+      "epoch": 1.6675745233752282,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.090393921377491e-06,
+      "loss": 0.9776,
+      "step": 8421
+    },
+    {
+      "epoch": 1.667774418430324,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.089357724882477e-06,
+      "loss": 1.0552,
+      "step": 8422
+    },
+    {
+      "epoch": 1.6679743134854201,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.088321568838095e-06,
+      "loss": 0.9877,
+      "step": 8423
+    },
+    {
+      "epoch": 1.6681742085405162,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.087285453290372e-06,
+      "loss": 0.9206,
+      "step": 8424
+    },
+    {
+      "epoch": 1.6683741035956123,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.08624937828533e-06,
+      "loss": 0.9945,
+      "step": 8425
+    },
+    {
+      "epoch": 1.6685739986507082,
+      "grad_norm": 2.125,
+      "learning_rate": 4.085213343868995e-06,
+      "loss": 0.9323,
+      "step": 8426
+    },
+    {
+      "epoch": 1.6687738937058043,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.084177350087384e-06,
+      "loss": 0.8495,
+      "step": 8427
+    },
+    {
+      "epoch": 1.6689737887609004,
+      "grad_norm": 2.125,
+      "learning_rate": 4.0831413969865195e-06,
+      "loss": 0.9174,
+      "step": 8428
+    },
+    {
+      "epoch": 1.6691736838159965,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.082105484612414e-06,
+      "loss": 0.9647,
+      "step": 8429
+    },
+    {
+      "epoch": 1.6693735788710926,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.0810696130110855e-06,
+      "loss": 0.9339,
+      "step": 8430
+    },
+    {
+      "epoch": 1.6695734739261887,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.080033782228547e-06,
+      "loss": 0.9517,
+      "step": 8431
+    },
+    {
+      "epoch": 1.6697733689812848,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.078997992310809e-06,
+      "loss": 0.9333,
+      "step": 8432
+    },
+    {
+      "epoch": 1.669973264036381,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.0779622433038825e-06,
+      "loss": 1.0035,
+      "step": 8433
+    },
+    {
+      "epoch": 1.670173159091477,
+      "grad_norm": 2.25,
+      "learning_rate": 4.076926535253775e-06,
+      "loss": 1.0483,
+      "step": 8434
+    },
+    {
+      "epoch": 1.670373054146573,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.075890868206494e-06,
+      "loss": 1.0404,
+      "step": 8435
+    },
+    {
+      "epoch": 1.6705729492016692,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.074855242208039e-06,
+      "loss": 0.94,
+      "step": 8436
+    },
+    {
+      "epoch": 1.6707728442567653,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.0738196573044155e-06,
+      "loss": 0.9711,
+      "step": 8437
+    },
+    {
+      "epoch": 1.6709727393118614,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.072784113541625e-06,
+      "loss": 0.953,
+      "step": 8438
+    },
+    {
+      "epoch": 1.6711726343669575,
+      "grad_norm": 2.125,
+      "learning_rate": 4.0717486109656636e-06,
+      "loss": 0.9248,
+      "step": 8439
+    },
+    {
+      "epoch": 1.6713725294220534,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.070713149622532e-06,
+      "loss": 0.9457,
+      "step": 8440
+    },
+    {
+      "epoch": 1.6715724244771495,
+      "grad_norm": 2.125,
+      "learning_rate": 4.06967772955822e-06,
+      "loss": 0.9245,
+      "step": 8441
+    },
+    {
+      "epoch": 1.6717723195322456,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.068642350818726e-06,
+      "loss": 0.9103,
+      "step": 8442
+    },
+    {
+      "epoch": 1.6719722145873417,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.067607013450039e-06,
+      "loss": 1.0159,
+      "step": 8443
+    },
+    {
+      "epoch": 1.6721721096424376,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.066571717498149e-06,
+      "loss": 0.9965,
+      "step": 8444
+    },
+    {
+      "epoch": 1.6723720046975337,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.065536463009043e-06,
+      "loss": 1.0675,
+      "step": 8445
+    },
+    {
+      "epoch": 1.6725718997526298,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.064501250028708e-06,
+      "loss": 0.9118,
+      "step": 8446
+    },
+    {
+      "epoch": 1.6727717948077259,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.063466078603127e-06,
+      "loss": 0.9403,
+      "step": 8447
+    },
+    {
+      "epoch": 1.672971689862822,
+      "grad_norm": 2.125,
+      "learning_rate": 4.062430948778285e-06,
+      "loss": 0.9674,
+      "step": 8448
+    },
+    {
+      "epoch": 1.673171584917918,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.061395860600157e-06,
+      "loss": 0.933,
+      "step": 8449
+    },
+    {
+      "epoch": 1.6733714799730142,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.060360814114725e-06,
+      "loss": 1.0244,
+      "step": 8450
+    },
+    {
+      "epoch": 1.6735713750281103,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.059325809367967e-06,
+      "loss": 1.1114,
+      "step": 8451
+    },
+    {
+      "epoch": 1.6737712700832064,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.058290846405856e-06,
+      "loss": 1.0581,
+      "step": 8452
+    },
+    {
+      "epoch": 1.6739711651383025,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.0572559252743635e-06,
+      "loss": 0.8494,
+      "step": 8453
+    },
+    {
+      "epoch": 1.6741710601933986,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.056221046019464e-06,
+      "loss": 0.967,
+      "step": 8454
+    },
+    {
+      "epoch": 1.6743709552484947,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.055186208687123e-06,
+      "loss": 0.9342,
+      "step": 8455
+    },
+    {
+      "epoch": 1.6745708503035908,
+      "grad_norm": 2.125,
+      "learning_rate": 4.0541514133233115e-06,
+      "loss": 0.8761,
+      "step": 8456
+    },
+    {
+      "epoch": 1.6747707453586866,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.053116659973991e-06,
+      "loss": 0.9926,
+      "step": 8457
+    },
+    {
+      "epoch": 1.6749706404137827,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.052081948685128e-06,
+      "loss": 0.9398,
+      "step": 8458
+    },
+    {
+      "epoch": 1.6751705354688788,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.051047279502686e-06,
+      "loss": 1.0457,
+      "step": 8459
+    },
+    {
+      "epoch": 1.675370430523975,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.05001265247262e-06,
+      "loss": 0.9247,
+      "step": 8460
+    },
+    {
+      "epoch": 1.675570325579071,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.04897806764089e-06,
+      "loss": 0.9881,
+      "step": 8461
+    },
+    {
+      "epoch": 1.675770220634167,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.047943525053455e-06,
+      "loss": 0.9475,
+      "step": 8462
+    },
+    {
+      "epoch": 1.675970115689263,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.046909024756265e-06,
+      "loss": 0.9269,
+      "step": 8463
+    },
+    {
+      "epoch": 1.6761700107443591,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.045874566795275e-06,
+      "loss": 1.0492,
+      "step": 8464
+    },
+    {
+      "epoch": 1.6763699057994552,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.044840151216435e-06,
+      "loss": 0.9486,
+      "step": 8465
+    },
+    {
+      "epoch": 1.6765698008545513,
+      "grad_norm": 2.125,
+      "learning_rate": 4.043805778065696e-06,
+      "loss": 0.9758,
+      "step": 8466
+    },
+    {
+      "epoch": 1.6767696959096474,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.042771447388998e-06,
+      "loss": 0.8804,
+      "step": 8467
+    },
+    {
+      "epoch": 1.6769695909647435,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.041737159232293e-06,
+      "loss": 0.9541,
+      "step": 8468
+    },
+    {
+      "epoch": 1.6771694860198396,
+      "grad_norm": 2.125,
+      "learning_rate": 4.040702913641519e-06,
+      "loss": 0.9378,
+      "step": 8469
+    },
+    {
+      "epoch": 1.6773693810749357,
+      "grad_norm": 2.125,
+      "learning_rate": 4.03966871066262e-06,
+      "loss": 0.949,
+      "step": 8470
+    },
+    {
+      "epoch": 1.6775692761300318,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.038634550341534e-06,
+      "loss": 0.984,
+      "step": 8471
+    },
+    {
+      "epoch": 1.677769171185128,
+      "grad_norm": 2.125,
+      "learning_rate": 4.037600432724199e-06,
+      "loss": 0.9495,
+      "step": 8472
+    },
+    {
+      "epoch": 1.677969066240224,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.036566357856549e-06,
+      "loss": 1.0034,
+      "step": 8473
+    },
+    {
+      "epoch": 1.6781689612953201,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.035532325784519e-06,
+      "loss": 0.912,
+      "step": 8474
+    },
+    {
+      "epoch": 1.678368856350416,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.034498336554041e-06,
+      "loss": 0.9738,
+      "step": 8475
+    },
+    {
+      "epoch": 1.678568751405512,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.033464390211043e-06,
+      "loss": 0.9138,
+      "step": 8476
+    },
+    {
+      "epoch": 1.6787686464606082,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.032430486801454e-06,
+      "loss": 1.0189,
+      "step": 8477
+    },
+    {
+      "epoch": 1.6789685415157043,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.031396626371198e-06,
+      "loss": 0.9922,
+      "step": 8478
+    },
+    {
+      "epoch": 1.6791684365708002,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.030362808966202e-06,
+      "loss": 1.0465,
+      "step": 8479
+    },
+    {
+      "epoch": 1.6793683316258963,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.029329034632386e-06,
+      "loss": 0.9951,
+      "step": 8480
+    },
+    {
+      "epoch": 1.6795682266809924,
+      "grad_norm": 2.25,
+      "learning_rate": 4.02829530341567e-06,
+      "loss": 1.0458,
+      "step": 8481
+    },
+    {
+      "epoch": 1.6797681217360885,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.027261615361973e-06,
+      "loss": 0.9252,
+      "step": 8482
+    },
+    {
+      "epoch": 1.6799680167911846,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.026227970517212e-06,
+      "loss": 0.982,
+      "step": 8483
+    },
+    {
+      "epoch": 1.6801679118462807,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.0251943689273e-06,
+      "loss": 0.9143,
+      "step": 8484
+    },
+    {
+      "epoch": 1.6803678069013768,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.02416081063815e-06,
+      "loss": 0.9427,
+      "step": 8485
+    },
+    {
+      "epoch": 1.6805677019564729,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.0231272956956725e-06,
+      "loss": 1.0324,
+      "step": 8486
+    },
+    {
+      "epoch": 1.680767597011569,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.022093824145777e-06,
+      "loss": 0.9414,
+      "step": 8487
+    },
+    {
+      "epoch": 1.680967492066665,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.021060396034369e-06,
+      "loss": 0.9265,
+      "step": 8488
+    },
+    {
+      "epoch": 1.6811673871217612,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.020027011407353e-06,
+      "loss": 0.9412,
+      "step": 8489
+    },
+    {
+      "epoch": 1.6813672821768573,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.018993670310633e-06,
+      "loss": 0.9536,
+      "step": 8490
+    },
+    {
+      "epoch": 1.6815671772319534,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.017960372790109e-06,
+      "loss": 0.9672,
+      "step": 8491
+    },
+    {
+      "epoch": 1.6817670722870492,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.01692711889168e-06,
+      "loss": 0.8579,
+      "step": 8492
+    },
+    {
+      "epoch": 1.6819669673421453,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.015893908661245e-06,
+      "loss": 0.9421,
+      "step": 8493
+    },
+    {
+      "epoch": 1.6821668623972414,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.014860742144696e-06,
+      "loss": 0.9703,
+      "step": 8494
+    },
+    {
+      "epoch": 1.6823667574523375,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.013827619387928e-06,
+      "loss": 0.9083,
+      "step": 8495
+    },
+    {
+      "epoch": 1.6825666525074336,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.012794540436832e-06,
+      "loss": 0.9675,
+      "step": 8496
+    },
+    {
+      "epoch": 1.6827665475625295,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.0117615053372986e-06,
+      "loss": 1.0955,
+      "step": 8497
+    },
+    {
+      "epoch": 1.6829664426176256,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.010728514135211e-06,
+      "loss": 0.9257,
+      "step": 8498
+    },
+    {
+      "epoch": 1.6831663376727217,
+      "grad_norm": 2.125,
+      "learning_rate": 4.009695566876459e-06,
+      "loss": 0.9684,
+      "step": 8499
+    },
+    {
+      "epoch": 1.6833662327278178,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.008662663606923e-06,
+      "loss": 0.9572,
+      "step": 8500
+    },
+    {
+      "epoch": 1.683566127782914,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.007629804372486e-06,
+      "loss": 0.9897,
+      "step": 8501
+    },
+    {
+      "epoch": 1.68376602283801,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.006596989219027e-06,
+      "loss": 0.9713,
+      "step": 8502
+    },
+    {
+      "epoch": 1.6839659178931061,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.005564218192424e-06,
+      "loss": 0.8463,
+      "step": 8503
+    },
+    {
+      "epoch": 1.6841658129482022,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.004531491338551e-06,
+      "loss": 0.9783,
+      "step": 8504
+    },
+    {
+      "epoch": 1.6843657080032983,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.0034988087032845e-06,
+      "loss": 0.993,
+      "step": 8505
+    },
+    {
+      "epoch": 1.6845656030583944,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.002466170332493e-06,
+      "loss": 0.8878,
+      "step": 8506
+    },
+    {
+      "epoch": 1.6847654981134905,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.001433576272049e-06,
+      "loss": 1.0262,
+      "step": 8507
+    },
+    {
+      "epoch": 1.6849653931685866,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.000401026567818e-06,
+      "loss": 0.9601,
+      "step": 8508
+    },
+    {
+      "epoch": 1.6851652882236827,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.999368521265667e-06,
+      "loss": 1.006,
+      "step": 8509
+    },
+    {
+      "epoch": 1.6853651832787786,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.998336060411459e-06,
+      "loss": 0.8779,
+      "step": 8510
+    },
+    {
+      "epoch": 1.6855650783338747,
+      "grad_norm": 2.25,
+      "learning_rate": 3.997303644051056e-06,
+      "loss": 1.0226,
+      "step": 8511
+    },
+    {
+      "epoch": 1.6857649733889708,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.9962712722303186e-06,
+      "loss": 0.9109,
+      "step": 8512
+    },
+    {
+      "epoch": 1.685964868444067,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.995238944995105e-06,
+      "loss": 0.9653,
+      "step": 8513
+    },
+    {
+      "epoch": 1.6861647634991628,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.994206662391269e-06,
+      "loss": 0.9835,
+      "step": 8514
+    },
+    {
+      "epoch": 1.6863646585542589,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.993174424464665e-06,
+      "loss": 0.9628,
+      "step": 8515
+    },
+    {
+      "epoch": 1.686564553609355,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.992142231261147e-06,
+      "loss": 0.9007,
+      "step": 8516
+    },
+    {
+      "epoch": 1.686764448664451,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.991110082826562e-06,
+      "loss": 0.9254,
+      "step": 8517
+    },
+    {
+      "epoch": 1.6869643437195472,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.990077979206761e-06,
+      "loss": 1.0035,
+      "step": 8518
+    },
+    {
+      "epoch": 1.6871642387746433,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.989045920447587e-06,
+      "loss": 0.945,
+      "step": 8519
+    },
+    {
+      "epoch": 1.6873641338297394,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.988013906594886e-06,
+      "loss": 0.8934,
+      "step": 8520
+    },
+    {
+      "epoch": 1.6875640288848355,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.9869819376944985e-06,
+      "loss": 0.9832,
+      "step": 8521
+    },
+    {
+      "epoch": 1.6877639239399316,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.985950013792265e-06,
+      "loss": 0.931,
+      "step": 8522
+    },
+    {
+      "epoch": 1.6879638189950277,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.984918134934024e-06,
+      "loss": 0.9826,
+      "step": 8523
+    },
+    {
+      "epoch": 1.6881637140501238,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.983886301165611e-06,
+      "loss": 0.9445,
+      "step": 8524
+    },
+    {
+      "epoch": 1.6883636091052199,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.9828545125328606e-06,
+      "loss": 0.9148,
+      "step": 8525
+    },
+    {
+      "epoch": 1.688563504160316,
+      "grad_norm": 2.25,
+      "learning_rate": 3.9818227690816045e-06,
+      "loss": 1.0161,
+      "step": 8526
+    },
+    {
+      "epoch": 1.6887633992154119,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.98079107085767e-06,
+      "loss": 0.9282,
+      "step": 8527
+    },
+    {
+      "epoch": 1.688963294270508,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.979759417906891e-06,
+      "loss": 0.9579,
+      "step": 8528
+    },
+    {
+      "epoch": 1.689163189325604,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.978727810275087e-06,
+      "loss": 0.9937,
+      "step": 8529
+    },
+    {
+      "epoch": 1.6893630843807002,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.977696248008086e-06,
+      "loss": 1.0588,
+      "step": 8530
+    },
+    {
+      "epoch": 1.6895629794357963,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.976664731151707e-06,
+      "loss": 1.0032,
+      "step": 8531
+    },
+    {
+      "epoch": 1.6897628744908921,
+      "grad_norm": 2.25,
+      "learning_rate": 3.975633259751771e-06,
+      "loss": 1.0594,
+      "step": 8532
+    },
+    {
+      "epoch": 1.6899627695459882,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.974601833854097e-06,
+      "loss": 0.9511,
+      "step": 8533
+    },
+    {
+      "epoch": 1.6901626646010843,
+      "grad_norm": 2.125,
+      "learning_rate": 3.9735704535045e-06,
+      "loss": 0.9048,
+      "step": 8534
+    },
+    {
+      "epoch": 1.6903625596561804,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.9725391187487935e-06,
+      "loss": 0.9463,
+      "step": 8535
+    },
+    {
+      "epoch": 1.6905624547112765,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.97150782963279e-06,
+      "loss": 0.9615,
+      "step": 8536
+    },
+    {
+      "epoch": 1.6907623497663726,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.9704765862022985e-06,
+      "loss": 0.9832,
+      "step": 8537
+    },
+    {
+      "epoch": 1.6909622448214687,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.969445388503128e-06,
+      "loss": 0.8743,
+      "step": 8538
+    },
+    {
+      "epoch": 1.6911621398765648,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.968414236581083e-06,
+      "loss": 0.8841,
+      "step": 8539
+    },
+    {
+      "epoch": 1.691362034931661,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.967383130481966e-06,
+      "loss": 0.9036,
+      "step": 8540
+    },
+    {
+      "epoch": 1.691561929986757,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.966352070251582e-06,
+      "loss": 0.8706,
+      "step": 8541
+    },
+    {
+      "epoch": 1.6917618250418531,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.965321055935727e-06,
+      "loss": 0.996,
+      "step": 8542
+    },
+    {
+      "epoch": 1.6919617200969492,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.9642900875802e-06,
+      "loss": 0.9702,
+      "step": 8543
+    },
+    {
+      "epoch": 1.6921616151520453,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.9632591652307985e-06,
+      "loss": 0.9192,
+      "step": 8544
+    },
+    {
+      "epoch": 1.6923615102071412,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.9622282889333135e-06,
+      "loss": 0.9064,
+      "step": 8545
+    },
+    {
+      "epoch": 1.6925614052622373,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.9611974587335375e-06,
+      "loss": 0.9418,
+      "step": 8546
+    },
+    {
+      "epoch": 1.6927613003173334,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.9601666746772586e-06,
+      "loss": 0.973,
+      "step": 8547
+    },
+    {
+      "epoch": 1.6929611953724295,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.959135936810265e-06,
+      "loss": 1.0562,
+      "step": 8548
+    },
+    {
+      "epoch": 1.6931610904275254,
+      "grad_norm": 2.25,
+      "learning_rate": 3.958105245178342e-06,
+      "loss": 0.9999,
+      "step": 8549
+    },
+    {
+      "epoch": 1.6933609854826215,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.957074599827272e-06,
+      "loss": 0.9098,
+      "step": 8550
+    },
+    {
+      "epoch": 1.6935608805377176,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.956044000802838e-06,
+      "loss": 0.9153,
+      "step": 8551
+    },
+    {
+      "epoch": 1.6937607755928137,
+      "grad_norm": 2.25,
+      "learning_rate": 3.955013448150818e-06,
+      "loss": 1.0571,
+      "step": 8552
+    },
+    {
+      "epoch": 1.6939606706479098,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.953982941916988e-06,
+      "loss": 0.9183,
+      "step": 8553
+    },
+    {
+      "epoch": 1.6941605657030059,
+      "grad_norm": 2.125,
+      "learning_rate": 3.952952482147125e-06,
+      "loss": 0.9679,
+      "step": 8554
+    },
+    {
+      "epoch": 1.694360460758102,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.9519220688870004e-06,
+      "loss": 0.9365,
+      "step": 8555
+    },
+    {
+      "epoch": 1.694560355813198,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.950891702182386e-06,
+      "loss": 0.8167,
+      "step": 8556
+    },
+    {
+      "epoch": 1.6947602508682942,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.94986138207905e-06,
+      "loss": 0.9672,
+      "step": 8557
+    },
+    {
+      "epoch": 1.6949601459233903,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.948831108622759e-06,
+      "loss": 1.0135,
+      "step": 8558
+    },
+    {
+      "epoch": 1.6951600409784864,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.94780088185928e-06,
+      "loss": 0.941,
+      "step": 8559
+    },
+    {
+      "epoch": 1.6953599360335825,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.946770701834372e-06,
+      "loss": 1.034,
+      "step": 8560
+    },
+    {
+      "epoch": 1.6955598310886786,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.945740568593796e-06,
+      "loss": 1.0687,
+      "step": 8561
+    },
+    {
+      "epoch": 1.6957597261437747,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.944710482183312e-06,
+      "loss": 0.9726,
+      "step": 8562
+    },
+    {
+      "epoch": 1.6959596211988706,
+      "grad_norm": 2.25,
+      "learning_rate": 3.943680442648675e-06,
+      "loss": 1.0225,
+      "step": 8563
+    },
+    {
+      "epoch": 1.6961595162539667,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.9426504500356415e-06,
+      "loss": 0.9074,
+      "step": 8564
+    },
+    {
+      "epoch": 1.6963594113090628,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.94162050438996e-06,
+      "loss": 0.9779,
+      "step": 8565
+    },
+    {
+      "epoch": 1.6965593063641589,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.940590605757383e-06,
+      "loss": 1.0321,
+      "step": 8566
+    },
+    {
+      "epoch": 1.6967592014192547,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.93956075418366e-06,
+      "loss": 1.026,
+      "step": 8567
+    },
+    {
+      "epoch": 1.6969590964743508,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.938530949714533e-06,
+      "loss": 1.0925,
+      "step": 8568
+    },
+    {
+      "epoch": 1.697158991529447,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.937501192395749e-06,
+      "loss": 0.975,
+      "step": 8569
+    },
+    {
+      "epoch": 1.697358886584543,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.936471482273048e-06,
+      "loss": 0.8929,
+      "step": 8570
+    },
+    {
+      "epoch": 1.6975587816396391,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.935441819392169e-06,
+      "loss": 0.9929,
+      "step": 8571
+    },
+    {
+      "epoch": 1.6977586766947352,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.934412203798853e-06,
+      "loss": 0.8692,
+      "step": 8572
+    },
+    {
+      "epoch": 1.6979585717498313,
+      "grad_norm": 2.125,
+      "learning_rate": 3.9333826355388325e-06,
+      "loss": 1.0455,
+      "step": 8573
+    },
+    {
+      "epoch": 1.6981584668049274,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.93235311465784e-06,
+      "loss": 1.0355,
+      "step": 8574
+    },
+    {
+      "epoch": 1.6983583618600235,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.93132364120161e-06,
+      "loss": 0.8788,
+      "step": 8575
+    },
+    {
+      "epoch": 1.6985582569151196,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.930294215215868e-06,
+      "loss": 1.0396,
+      "step": 8576
+    },
+    {
+      "epoch": 1.6987581519702157,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.929264836746345e-06,
+      "loss": 1.0266,
+      "step": 8577
+    },
+    {
+      "epoch": 1.6989580470253118,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.928235505838762e-06,
+      "loss": 1.0414,
+      "step": 8578
+    },
+    {
+      "epoch": 1.699157942080408,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.927206222538843e-06,
+      "loss": 0.8899,
+      "step": 8579
+    },
+    {
+      "epoch": 1.6993578371355038,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.92617698689231e-06,
+      "loss": 0.9934,
+      "step": 8580
+    },
+    {
+      "epoch": 1.6995577321906,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.92514779894488e-06,
+      "loss": 0.9783,
+      "step": 8581
+    },
+    {
+      "epoch": 1.699757627245696,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.924118658742269e-06,
+      "loss": 1.0363,
+      "step": 8582
+    },
+    {
+      "epoch": 1.6999575223007921,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.923089566330195e-06,
+      "loss": 1.0069,
+      "step": 8583
+    },
+    {
+      "epoch": 1.7001574173558882,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.922060521754365e-06,
+      "loss": 0.9617,
+      "step": 8584
+    },
+    {
+      "epoch": 1.700357312410984,
+      "grad_norm": 2.0,
+      "learning_rate": 3.921031525060493e-06,
+      "loss": 0.8853,
+      "step": 8585
+    },
+    {
+      "epoch": 1.7005572074660802,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.920002576294284e-06,
+      "loss": 0.9252,
+      "step": 8586
+    },
+    {
+      "epoch": 1.7007571025211763,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.918973675501446e-06,
+      "loss": 0.9564,
+      "step": 8587
+    },
+    {
+      "epoch": 1.7009569975762724,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.917944822727682e-06,
+      "loss": 0.9961,
+      "step": 8588
+    },
+    {
+      "epoch": 1.7011568926313685,
+      "grad_norm": 2.25,
+      "learning_rate": 3.916916018018696e-06,
+      "loss": 1.062,
+      "step": 8589
+    },
+    {
+      "epoch": 1.7013567876864646,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.915887261420181e-06,
+      "loss": 0.9978,
+      "step": 8590
+    },
+    {
+      "epoch": 1.7015566827415607,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.9148585529778385e-06,
+      "loss": 1.0687,
+      "step": 8591
+    },
+    {
+      "epoch": 1.7017565777966568,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.913829892737364e-06,
+      "loss": 0.8257,
+      "step": 8592
+    },
+    {
+      "epoch": 1.7019564728517529,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.912801280744449e-06,
+      "loss": 0.9383,
+      "step": 8593
+    },
+    {
+      "epoch": 1.702156367906849,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.911772717044786e-06,
+      "loss": 1.0214,
+      "step": 8594
+    },
+    {
+      "epoch": 1.702356262961945,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.910744201684062e-06,
+      "loss": 0.956,
+      "step": 8595
+    },
+    {
+      "epoch": 1.7025561580170412,
+      "grad_norm": 2.125,
+      "learning_rate": 3.909715734707964e-06,
+      "loss": 0.9843,
+      "step": 8596
+    },
+    {
+      "epoch": 1.7027560530721373,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.908687316162178e-06,
+      "loss": 0.9228,
+      "step": 8597
+    },
+    {
+      "epoch": 1.7029559481272332,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.907658946092383e-06,
+      "loss": 0.963,
+      "step": 8598
+    },
+    {
+      "epoch": 1.7031558431823293,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.906630624544261e-06,
+      "loss": 0.9671,
+      "step": 8599
+    },
+    {
+      "epoch": 1.7033557382374254,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.905602351563492e-06,
+      "loss": 1.0168,
+      "step": 8600
+    },
+    {
+      "epoch": 1.7035556332925215,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.904574127195747e-06,
+      "loss": 0.9687,
+      "step": 8601
+    },
+    {
+      "epoch": 1.7037555283476173,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.903545951486704e-06,
+      "loss": 1.0447,
+      "step": 8602
+    },
+    {
+      "epoch": 1.7039554234027134,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.902517824482033e-06,
+      "loss": 0.9336,
+      "step": 8603
+    },
+    {
+      "epoch": 1.7041553184578095,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.901489746227402e-06,
+      "loss": 0.898,
+      "step": 8604
+    },
+    {
+      "epoch": 1.7043552135129056,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.900461716768479e-06,
+      "loss": 0.9529,
+      "step": 8605
+    },
+    {
+      "epoch": 1.7045551085680017,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.89943373615093e-06,
+      "loss": 0.8709,
+      "step": 8606
+    },
+    {
+      "epoch": 1.7047550036230978,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.8984058044204164e-06,
+      "loss": 1.0679,
+      "step": 8607
+    },
+    {
+      "epoch": 1.704954898678194,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8973779216226e-06,
+      "loss": 0.9793,
+      "step": 8608
+    },
+    {
+      "epoch": 1.70515479373329,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.8963500878031376e-06,
+      "loss": 0.9541,
+      "step": 8609
+    },
+    {
+      "epoch": 1.7053546887883861,
+      "grad_norm": 2.125,
+      "learning_rate": 3.895322303007686e-06,
+      "loss": 0.9899,
+      "step": 8610
+    },
+    {
+      "epoch": 1.7055545838434822,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.894294567281901e-06,
+      "loss": 0.979,
+      "step": 8611
+    },
+    {
+      "epoch": 1.7057544788985783,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.893266880671433e-06,
+      "loss": 0.9598,
+      "step": 8612
+    },
+    {
+      "epoch": 1.7059543739536744,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.892239243221931e-06,
+      "loss": 1.0535,
+      "step": 8613
+    },
+    {
+      "epoch": 1.7061542690087705,
+      "grad_norm": 2.25,
+      "learning_rate": 3.891211654979045e-06,
+      "loss": 1.0797,
+      "step": 8614
+    },
+    {
+      "epoch": 1.7063541640638664,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.890184115988418e-06,
+      "loss": 1.0081,
+      "step": 8615
+    },
+    {
+      "epoch": 1.7065540591189625,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.889156626295694e-06,
+      "loss": 1.0177,
+      "step": 8616
+    },
+    {
+      "epoch": 1.7067539541740586,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.888129185946514e-06,
+      "loss": 1.0065,
+      "step": 8617
+    },
+    {
+      "epoch": 1.7069538492291547,
+      "grad_norm": 1.9921875,
+      "learning_rate": 3.887101794986516e-06,
+      "loss": 0.9566,
+      "step": 8618
+    },
+    {
+      "epoch": 1.7071537442842508,
+      "grad_norm": 2.25,
+      "learning_rate": 3.886074453461339e-06,
+      "loss": 0.9088,
+      "step": 8619
+    },
+    {
+      "epoch": 1.7073536393393467,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.885047161416618e-06,
+      "loss": 1.0895,
+      "step": 8620
+    },
+    {
+      "epoch": 1.7075535343944428,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.884019918897979e-06,
+      "loss": 0.9131,
+      "step": 8621
+    },
+    {
+      "epoch": 1.707753429449539,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.882992725951057e-06,
+      "loss": 1.0781,
+      "step": 8622
+    },
+    {
+      "epoch": 1.707953324504635,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.881965582621479e-06,
+      "loss": 0.9742,
+      "step": 8623
+    },
+    {
+      "epoch": 1.708153219559731,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.88093848895487e-06,
+      "loss": 0.9311,
+      "step": 8624
+    },
+    {
+      "epoch": 1.7083531146148272,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.879911444996854e-06,
+      "loss": 1.1179,
+      "step": 8625
+    },
+    {
+      "epoch": 1.7085530096699233,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.878884450793053e-06,
+      "loss": 0.9329,
+      "step": 8626
+    },
+    {
+      "epoch": 1.7087529047250194,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.877857506389083e-06,
+      "loss": 1.0512,
+      "step": 8627
+    },
+    {
+      "epoch": 1.7089527997801155,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.876830611830565e-06,
+      "loss": 1.0058,
+      "step": 8628
+    },
+    {
+      "epoch": 1.7091526948352116,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.87580376716311e-06,
+      "loss": 0.9564,
+      "step": 8629
+    },
+    {
+      "epoch": 1.7093525898903077,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.874776972432331e-06,
+      "loss": 0.9368,
+      "step": 8630
+    },
+    {
+      "epoch": 1.7095524849454038,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.87375022768384e-06,
+      "loss": 0.8779,
+      "step": 8631
+    },
+    {
+      "epoch": 1.7097523800005,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.872723532963242e-06,
+      "loss": 0.976,
+      "step": 8632
+    },
+    {
+      "epoch": 1.7099522750555958,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.871696888316145e-06,
+      "loss": 1.0064,
+      "step": 8633
+    },
+    {
+      "epoch": 1.7101521701106919,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.870670293788153e-06,
+      "loss": 0.9534,
+      "step": 8634
+    },
+    {
+      "epoch": 1.710352065165788,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8696437494248645e-06,
+      "loss": 1.0084,
+      "step": 8635
+    },
+    {
+      "epoch": 1.710551960220884,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.868617255271881e-06,
+      "loss": 0.9585,
+      "step": 8636
+    },
+    {
+      "epoch": 1.71075185527598,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.867590811374797e-06,
+      "loss": 1.01,
+      "step": 8637
+    },
+    {
+      "epoch": 1.710951750331076,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.866564417779208e-06,
+      "loss": 0.9748,
+      "step": 8638
+    },
+    {
+      "epoch": 1.7111516453861721,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.865538074530708e-06,
+      "loss": 0.937,
+      "step": 8639
+    },
+    {
+      "epoch": 1.7113515404412682,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.864511781674885e-06,
+      "loss": 0.9954,
+      "step": 8640
+    },
+    {
+      "epoch": 1.7115514354963643,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.863485539257326e-06,
+      "loss": 0.9562,
+      "step": 8641
+    },
+    {
+      "epoch": 1.7117513305514604,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.86245934732362e-06,
+      "loss": 1.0433,
+      "step": 8642
+    },
+    {
+      "epoch": 1.7119512256065565,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.861433205919347e-06,
+      "loss": 0.8788,
+      "step": 8643
+    },
+    {
+      "epoch": 1.7121511206616526,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.86040711509009e-06,
+      "loss": 0.9968,
+      "step": 8644
+    },
+    {
+      "epoch": 1.7123510157167487,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.859381074881427e-06,
+      "loss": 1.0523,
+      "step": 8645
+    },
+    {
+      "epoch": 1.7125509107718448,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.858355085338935e-06,
+      "loss": 0.9658,
+      "step": 8646
+    },
+    {
+      "epoch": 1.712750805826941,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.857329146508188e-06,
+      "loss": 0.907,
+      "step": 8647
+    },
+    {
+      "epoch": 1.712950700882037,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.856303258434758e-06,
+      "loss": 0.8811,
+      "step": 8648
+    },
+    {
+      "epoch": 1.7131505959371331,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.8552774211642154e-06,
+      "loss": 0.9952,
+      "step": 8649
+    },
+    {
+      "epoch": 1.713350490992229,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.854251634742128e-06,
+      "loss": 0.9703,
+      "step": 8650
+    },
+    {
+      "epoch": 1.7135503860473251,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.853225899214062e-06,
+      "loss": 0.9576,
+      "step": 8651
+    },
+    {
+      "epoch": 1.7137502811024212,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8522002146255765e-06,
+      "loss": 0.9298,
+      "step": 8652
+    },
+    {
+      "epoch": 1.7139501761575173,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.851174581022236e-06,
+      "loss": 1.0168,
+      "step": 8653
+    },
+    {
+      "epoch": 1.7141500712126134,
+      "grad_norm": 2.25,
+      "learning_rate": 3.850148998449597e-06,
+      "loss": 0.9716,
+      "step": 8654
+    },
+    {
+      "epoch": 1.7143499662677093,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.849123466953217e-06,
+      "loss": 1.003,
+      "step": 8655
+    },
+    {
+      "epoch": 1.7145498613228054,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.84809798657865e-06,
+      "loss": 0.9513,
+      "step": 8656
+    },
+    {
+      "epoch": 1.7147497563779015,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.847072557371448e-06,
+      "loss": 0.9694,
+      "step": 8657
+    },
+    {
+      "epoch": 1.7149496514329976,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.846047179377159e-06,
+      "loss": 0.9758,
+      "step": 8658
+    },
+    {
+      "epoch": 1.7151495464880937,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.8450218526413315e-06,
+      "loss": 0.9551,
+      "step": 8659
+    },
+    {
+      "epoch": 1.7153494415431898,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.843996577209509e-06,
+      "loss": 0.979,
+      "step": 8660
+    },
+    {
+      "epoch": 1.715549336598286,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.842971353127235e-06,
+      "loss": 1.0461,
+      "step": 8661
+    },
+    {
+      "epoch": 1.715749231653382,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.841946180440052e-06,
+      "loss": 1.0534,
+      "step": 8662
+    },
+    {
+      "epoch": 1.715949126708478,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.840921059193494e-06,
+      "loss": 0.9713,
+      "step": 8663
+    },
+    {
+      "epoch": 1.7161490217635742,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.839895989433099e-06,
+      "loss": 0.9854,
+      "step": 8664
+    },
+    {
+      "epoch": 1.7163489168186703,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.838870971204401e-06,
+      "loss": 1.0073,
+      "step": 8665
+    },
+    {
+      "epoch": 1.7165488118737664,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.83784600455293e-06,
+      "loss": 0.9942,
+      "step": 8666
+    },
+    {
+      "epoch": 1.7167487069288625,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.836821089524216e-06,
+      "loss": 0.9617,
+      "step": 8667
+    },
+    {
+      "epoch": 1.7169486019839584,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.835796226163784e-06,
+      "loss": 0.9522,
+      "step": 8668
+    },
+    {
+      "epoch": 1.7171484970390545,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.834771414517159e-06,
+      "loss": 0.9609,
+      "step": 8669
+    },
+    {
+      "epoch": 1.7173483920941506,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.833746654629865e-06,
+      "loss": 0.9881,
+      "step": 8670
+    },
+    {
+      "epoch": 1.7175482871492467,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.832721946547418e-06,
+      "loss": 0.917,
+      "step": 8671
+    },
+    {
+      "epoch": 1.7177481822043426,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.831697290315339e-06,
+      "loss": 1.0694,
+      "step": 8672
+    },
+    {
+      "epoch": 1.7179480772594387,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.830672685979142e-06,
+      "loss": 1.0193,
+      "step": 8673
+    },
+    {
+      "epoch": 1.7181479723145348,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.829648133584338e-06,
+      "loss": 0.895,
+      "step": 8674
+    },
+    {
+      "epoch": 1.7183478673696309,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.82862363317644e-06,
+      "loss": 0.9861,
+      "step": 8675
+    },
+    {
+      "epoch": 1.718547762424727,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.827599184800954e-06,
+      "loss": 0.9176,
+      "step": 8676
+    },
+    {
+      "epoch": 1.718747657479823,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.826574788503387e-06,
+      "loss": 0.9677,
+      "step": 8677
+    },
+    {
+      "epoch": 1.7189475525349192,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.825550444329244e-06,
+      "loss": 1.0463,
+      "step": 8678
+    },
+    {
+      "epoch": 1.7191474475900153,
+      "grad_norm": 2.25,
+      "learning_rate": 3.8245261523240235e-06,
+      "loss": 1.0227,
+      "step": 8679
+    },
+    {
+      "epoch": 1.7193473426451114,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.823501912533226e-06,
+      "loss": 0.9353,
+      "step": 8680
+    },
+    {
+      "epoch": 1.7195472377002075,
+      "grad_norm": 2.125,
+      "learning_rate": 3.822477725002348e-06,
+      "loss": 0.9909,
+      "step": 8681
+    },
+    {
+      "epoch": 1.7197471327553036,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.821453589776886e-06,
+      "loss": 0.9857,
+      "step": 8682
+    },
+    {
+      "epoch": 1.7199470278103997,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.820429506902326e-06,
+      "loss": 1.0048,
+      "step": 8683
+    },
+    {
+      "epoch": 1.7201469228654958,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.819405476424164e-06,
+      "loss": 0.9792,
+      "step": 8684
+    },
+    {
+      "epoch": 1.7203468179205919,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.818381498387883e-06,
+      "loss": 0.9946,
+      "step": 8685
+    },
+    {
+      "epoch": 1.7205467129756877,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.81735757283897e-06,
+      "loss": 0.9009,
+      "step": 8686
+    },
+    {
+      "epoch": 1.7207466080307838,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8163336998229075e-06,
+      "loss": 0.9599,
+      "step": 8687
+    },
+    {
+      "epoch": 1.72094650308588,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.815309879385176e-06,
+      "loss": 0.9272,
+      "step": 8688
+    },
+    {
+      "epoch": 1.721146398140976,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.8142861115712515e-06,
+      "loss": 0.9367,
+      "step": 8689
+    },
+    {
+      "epoch": 1.721346293196072,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.8132623964266136e-06,
+      "loss": 1.0028,
+      "step": 8690
+    },
+    {
+      "epoch": 1.721546188251168,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.8122387339967315e-06,
+      "loss": 1.0069,
+      "step": 8691
+    },
+    {
+      "epoch": 1.721746083306264,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.811215124327078e-06,
+      "loss": 1.0578,
+      "step": 8692
+    },
+    {
+      "epoch": 1.7219459783613602,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.810191567463123e-06,
+      "loss": 0.9108,
+      "step": 8693
+    },
+    {
+      "epoch": 1.7221458734164563,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8091680634503303e-06,
+      "loss": 0.9725,
+      "step": 8694
+    },
+    {
+      "epoch": 1.7223457684715524,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.808144612334165e-06,
+      "loss": 0.9575,
+      "step": 8695
+    },
+    {
+      "epoch": 1.7225456635266485,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.80712121416009e-06,
+      "loss": 1.0384,
+      "step": 8696
+    },
+    {
+      "epoch": 1.7227455585817446,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.806097868973562e-06,
+      "loss": 0.977,
+      "step": 8697
+    },
+    {
+      "epoch": 1.7229454536368407,
+      "grad_norm": 2.125,
+      "learning_rate": 3.8050745768200404e-06,
+      "loss": 0.9893,
+      "step": 8698
+    },
+    {
+      "epoch": 1.7231453486919368,
+      "grad_norm": 2.25,
+      "learning_rate": 3.8040513377449774e-06,
+      "loss": 0.9146,
+      "step": 8699
+    },
+    {
+      "epoch": 1.723345243747033,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.803028151793826e-06,
+      "loss": 1.0259,
+      "step": 8700
+    },
+    {
+      "epoch": 1.723545138802129,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.802005019012038e-06,
+      "loss": 0.9262,
+      "step": 8701
+    },
+    {
+      "epoch": 1.723745033857225,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.800981939445057e-06,
+      "loss": 0.9522,
+      "step": 8702
+    },
+    {
+      "epoch": 1.723944928912321,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.7999589131383306e-06,
+      "loss": 0.9703,
+      "step": 8703
+    },
+    {
+      "epoch": 1.724144823967417,
+      "grad_norm": 2.25,
+      "learning_rate": 3.798935940137303e-06,
+      "loss": 1.141,
+      "step": 8704
+    },
+    {
+      "epoch": 1.7243447190225132,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.7979130204874106e-06,
+      "loss": 0.9866,
+      "step": 8705
+    },
+    {
+      "epoch": 1.7245446140776093,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.7968901542340946e-06,
+      "loss": 0.9721,
+      "step": 8706
+    },
+    {
+      "epoch": 1.7247445091327054,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.795867341422789e-06,
+      "loss": 1.0209,
+      "step": 8707
+    },
+    {
+      "epoch": 1.7249444041878013,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7948445820989265e-06,
+      "loss": 0.8676,
+      "step": 8708
+    },
+    {
+      "epoch": 1.7251442992428974,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.793821876307941e-06,
+      "loss": 0.9601,
+      "step": 8709
+    },
+    {
+      "epoch": 1.7253441942979935,
+      "grad_norm": 2.125,
+      "learning_rate": 3.792799224095257e-06,
+      "loss": 0.991,
+      "step": 8710
+    },
+    {
+      "epoch": 1.7255440893530896,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.791776625506303e-06,
+      "loss": 0.983,
+      "step": 8711
+    },
+    {
+      "epoch": 1.7257439844081857,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.7907540805865035e-06,
+      "loss": 0.8535,
+      "step": 8712
+    },
+    {
+      "epoch": 1.7259438794632818,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.7897315893812796e-06,
+      "loss": 0.9902,
+      "step": 8713
+    },
+    {
+      "epoch": 1.7261437745183779,
+      "grad_norm": 2.125,
+      "learning_rate": 3.788709151936047e-06,
+      "loss": 0.9822,
+      "step": 8714
+    },
+    {
+      "epoch": 1.726343669573474,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.7876867682962244e-06,
+      "loss": 0.9171,
+      "step": 8715
+    },
+    {
+      "epoch": 1.72654356462857,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.786664438507227e-06,
+      "loss": 0.9963,
+      "step": 8716
+    },
+    {
+      "epoch": 1.7267434596836662,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.7856421626144645e-06,
+      "loss": 0.866,
+      "step": 8717
+    },
+    {
+      "epoch": 1.7269433547387623,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7846199406633493e-06,
+      "loss": 0.9655,
+      "step": 8718
+    },
+    {
+      "epoch": 1.7271432497938584,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.783597772699285e-06,
+      "loss": 0.9487,
+      "step": 8719
+    },
+    {
+      "epoch": 1.7273431448489545,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.782575658767678e-06,
+      "loss": 0.9771,
+      "step": 8720
+    },
+    {
+      "epoch": 1.7275430399040503,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.7815535989139307e-06,
+      "loss": 0.953,
+      "step": 8721
+    },
+    {
+      "epoch": 1.7277429349591464,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.7805315931834413e-06,
+      "loss": 0.9501,
+      "step": 8722
+    },
+    {
+      "epoch": 1.7279428300142425,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.779509641621609e-06,
+      "loss": 0.9922,
+      "step": 8723
+    },
+    {
+      "epoch": 1.7281427250693386,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.7784877442738283e-06,
+      "loss": 1.0136,
+      "step": 8724
+    },
+    {
+      "epoch": 1.7283426201244345,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.777465901185491e-06,
+      "loss": 0.9256,
+      "step": 8725
+    },
+    {
+      "epoch": 1.7285425151795306,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.776444112401989e-06,
+      "loss": 0.9601,
+      "step": 8726
+    },
+    {
+      "epoch": 1.7287424102346267,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.775422377968707e-06,
+      "loss": 0.9693,
+      "step": 8727
+    },
+    {
+      "epoch": 1.7289423052897228,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7744006979310333e-06,
+      "loss": 0.9509,
+      "step": 8728
+    },
+    {
+      "epoch": 1.729142200344819,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.77337907233435e-06,
+      "loss": 1.0105,
+      "step": 8729
+    },
+    {
+      "epoch": 1.729342095399915,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.7723575012240364e-06,
+      "loss": 1.0143,
+      "step": 8730
+    },
+    {
+      "epoch": 1.7295419904550111,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.771335984645471e-06,
+      "loss": 0.9855,
+      "step": 8731
+    },
+    {
+      "epoch": 1.7297418855101072,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.770314522644032e-06,
+      "loss": 0.9882,
+      "step": 8732
+    },
+    {
+      "epoch": 1.7299417805652033,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.7692931152650893e-06,
+      "loss": 1.0563,
+      "step": 8733
+    },
+    {
+      "epoch": 1.7301416756202994,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.7682717625540144e-06,
+      "loss": 1.0002,
+      "step": 8734
+    },
+    {
+      "epoch": 1.7303415706753955,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7672504645561773e-06,
+      "loss": 0.9918,
+      "step": 8735
+    },
+    {
+      "epoch": 1.7305414657304916,
+      "grad_norm": 2.125,
+      "learning_rate": 3.766229221316942e-06,
+      "loss": 0.9311,
+      "step": 8736
+    },
+    {
+      "epoch": 1.7307413607855877,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7652080328816744e-06,
+      "loss": 0.968,
+      "step": 8737
+    },
+    {
+      "epoch": 1.7309412558406836,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.7641868992957326e-06,
+      "loss": 0.9163,
+      "step": 8738
+    },
+    {
+      "epoch": 1.7311411508957797,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.763165820604477e-06,
+      "loss": 1.006,
+      "step": 8739
+    },
+    {
+      "epoch": 1.7313410459508758,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.7621447968532644e-06,
+      "loss": 0.9999,
+      "step": 8740
+    },
+    {
+      "epoch": 1.731540941005972,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7611238280874467e-06,
+      "loss": 0.9689,
+      "step": 8741
+    },
+    {
+      "epoch": 1.731740836061068,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7601029143523767e-06,
+      "loss": 0.9182,
+      "step": 8742
+    },
+    {
+      "epoch": 1.7319407311161639,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.759082055693403e-06,
+      "loss": 0.9651,
+      "step": 8743
+    },
+    {
+      "epoch": 1.73214062617126,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7580612521558737e-06,
+      "loss": 0.9859,
+      "step": 8744
+    },
+    {
+      "epoch": 1.732340521226356,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7570405037851287e-06,
+      "loss": 0.9697,
+      "step": 8745
+    },
+    {
+      "epoch": 1.7325404162814522,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.756019810626511e-06,
+      "loss": 0.9336,
+      "step": 8746
+    },
+    {
+      "epoch": 1.7327403113365483,
+      "grad_norm": 2.25,
+      "learning_rate": 3.754999172725362e-06,
+      "loss": 1.0636,
+      "step": 8747
+    },
+    {
+      "epoch": 1.7329402063916444,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.7539785901270164e-06,
+      "loss": 0.9048,
+      "step": 8748
+    },
+    {
+      "epoch": 1.7331401014467405,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.752958062876809e-06,
+      "loss": 0.9167,
+      "step": 8749
+    },
+    {
+      "epoch": 1.7333399965018366,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7519375910200706e-06,
+      "loss": 0.9092,
+      "step": 8750
+    },
+    {
+      "epoch": 1.7335398915569327,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.7509171746021305e-06,
+      "loss": 0.9155,
+      "step": 8751
+    },
+    {
+      "epoch": 1.7337397866120288,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.749896813668318e-06,
+      "loss": 1.0269,
+      "step": 8752
+    },
+    {
+      "epoch": 1.7339396816671249,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.748876508263954e-06,
+      "loss": 1.039,
+      "step": 8753
+    },
+    {
+      "epoch": 1.734139576722221,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7478562584343615e-06,
+      "loss": 0.9549,
+      "step": 8754
+    },
+    {
+      "epoch": 1.734339471777317,
+      "grad_norm": 2.25,
+      "learning_rate": 3.746836064224862e-06,
+      "loss": 1.0041,
+      "step": 8755
+    },
+    {
+      "epoch": 1.734539366832413,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.745815925680769e-06,
+      "loss": 1.0523,
+      "step": 8756
+    },
+    {
+      "epoch": 1.734739261887509,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.7447958428474006e-06,
+      "loss": 1.0006,
+      "step": 8757
+    },
+    {
+      "epoch": 1.7349391569426051,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.743775815770065e-06,
+      "loss": 1.0765,
+      "step": 8758
+    },
+    {
+      "epoch": 1.7351390519977012,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7427558444940738e-06,
+      "loss": 0.9442,
+      "step": 8759
+    },
+    {
+      "epoch": 1.7353389470527971,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.741735929064735e-06,
+      "loss": 0.9692,
+      "step": 8760
+    },
+    {
+      "epoch": 1.7355388421078932,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.74071606952735e-06,
+      "loss": 0.9094,
+      "step": 8761
+    },
+    {
+      "epoch": 1.7357387371629893,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.739696265927223e-06,
+      "loss": 0.9134,
+      "step": 8762
+    },
+    {
+      "epoch": 1.7359386322180854,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7386765183096545e-06,
+      "loss": 1.0662,
+      "step": 8763
+    },
+    {
+      "epoch": 1.7361385272731815,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.737656826719939e-06,
+      "loss": 0.9534,
+      "step": 8764
+    },
+    {
+      "epoch": 1.7363384223282776,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.7366371912033737e-06,
+      "loss": 0.9429,
+      "step": 8765
+    },
+    {
+      "epoch": 1.7365383173833737,
+      "grad_norm": 2.125,
+      "learning_rate": 3.735617611805249e-06,
+      "loss": 0.9426,
+      "step": 8766
+    },
+    {
+      "epoch": 1.7367382124384698,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7345980885708545e-06,
+      "loss": 0.9913,
+      "step": 8767
+    },
+    {
+      "epoch": 1.736938107493566,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.733578621545478e-06,
+      "loss": 0.8853,
+      "step": 8768
+    },
+    {
+      "epoch": 1.737138002548662,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7325592107744034e-06,
+      "loss": 0.9621,
+      "step": 8769
+    },
+    {
+      "epoch": 1.7373378976037581,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.7315398563029137e-06,
+      "loss": 0.9646,
+      "step": 8770
+    },
+    {
+      "epoch": 1.7375377926588542,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7305205581762895e-06,
+      "loss": 0.9537,
+      "step": 8771
+    },
+    {
+      "epoch": 1.7377376877139503,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.7295013164398042e-06,
+      "loss": 0.9745,
+      "step": 8772
+    },
+    {
+      "epoch": 1.7379375827690462,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.7284821311387356e-06,
+      "loss": 0.901,
+      "step": 8773
+    },
+    {
+      "epoch": 1.7381374778241423,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7274630023183554e-06,
+      "loss": 1.0101,
+      "step": 8774
+    },
+    {
+      "epoch": 1.7383373728792384,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.726443930023934e-06,
+      "loss": 1.0043,
+      "step": 8775
+    },
+    {
+      "epoch": 1.7385372679343345,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.7254249143007356e-06,
+      "loss": 1.0406,
+      "step": 8776
+    },
+    {
+      "epoch": 1.7387371629894306,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.724405955194027e-06,
+      "loss": 0.9588,
+      "step": 8777
+    },
+    {
+      "epoch": 1.7389370580445265,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7233870527490683e-06,
+      "loss": 0.8205,
+      "step": 8778
+    },
+    {
+      "epoch": 1.7391369530996226,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.7223682070111212e-06,
+      "loss": 0.9412,
+      "step": 8779
+    },
+    {
+      "epoch": 1.7393368481547187,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.7213494180254417e-06,
+      "loss": 1.0123,
+      "step": 8780
+    },
+    {
+      "epoch": 1.7395367432098148,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.7203306858372845e-06,
+      "loss": 0.9726,
+      "step": 8781
+    },
+    {
+      "epoch": 1.7397366382649109,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.719312010491901e-06,
+      "loss": 0.9801,
+      "step": 8782
+    },
+    {
+      "epoch": 1.739936533320007,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.7182933920345426e-06,
+      "loss": 0.9375,
+      "step": 8783
+    },
+    {
+      "epoch": 1.740136428375103,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7172748305104537e-06,
+      "loss": 0.9314,
+      "step": 8784
+    },
+    {
+      "epoch": 1.7403363234301992,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.71625632596488e-06,
+      "loss": 0.9438,
+      "step": 8785
+    },
+    {
+      "epoch": 1.7405362184852953,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.7152378784430643e-06,
+      "loss": 0.9976,
+      "step": 8786
+    },
+    {
+      "epoch": 1.7407361135403914,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.7142194879902434e-06,
+      "loss": 0.9725,
+      "step": 8787
+    },
+    {
+      "epoch": 1.7409360085954875,
+      "grad_norm": 2.0,
+      "learning_rate": 3.7132011546516568e-06,
+      "loss": 0.8509,
+      "step": 8788
+    },
+    {
+      "epoch": 1.7411359036505836,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.7121828784725357e-06,
+      "loss": 0.9757,
+      "step": 8789
+    },
+    {
+      "epoch": 1.7413357987056797,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.711164659498114e-06,
+      "loss": 0.9933,
+      "step": 8790
+    },
+    {
+      "epoch": 1.7415356937607755,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.710146497773622e-06,
+      "loss": 0.9986,
+      "step": 8791
+    },
+    {
+      "epoch": 1.7417355888158716,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7091283933442835e-06,
+      "loss": 0.9548,
+      "step": 8792
+    },
+    {
+      "epoch": 1.7419354838709677,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7081103462553234e-06,
+      "loss": 0.9078,
+      "step": 8793
+    },
+    {
+      "epoch": 1.7421353789260638,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.7070923565519655e-06,
+      "loss": 0.8792,
+      "step": 8794
+    },
+    {
+      "epoch": 1.7423352739811597,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.706074424279426e-06,
+      "loss": 1.042,
+      "step": 8795
+    },
+    {
+      "epoch": 1.7425351690362558,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.7050565494829233e-06,
+      "loss": 0.9285,
+      "step": 8796
+    },
+    {
+      "epoch": 1.742735064091352,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.704038732207669e-06,
+      "loss": 0.9947,
+      "step": 8797
+    },
+    {
+      "epoch": 1.742934959146448,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.7030209724988764e-06,
+      "loss": 0.9517,
+      "step": 8798
+    },
+    {
+      "epoch": 1.7431348542015441,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7020032704017537e-06,
+      "loss": 1.017,
+      "step": 8799
+    },
+    {
+      "epoch": 1.7433347492566402,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.7009856259615074e-06,
+      "loss": 0.9797,
+      "step": 8800
+    },
+    {
+      "epoch": 1.7435346443117363,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.6999680392233397e-06,
+      "loss": 0.9665,
+      "step": 8801
+    },
+    {
+      "epoch": 1.7437345393668324,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.698950510232454e-06,
+      "loss": 0.9752,
+      "step": 8802
+    },
+    {
+      "epoch": 1.7439344344219285,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.6979330390340472e-06,
+      "loss": 0.8478,
+      "step": 8803
+    },
+    {
+      "epoch": 1.7441343294770246,
+      "grad_norm": 2.125,
+      "learning_rate": 3.696915625673317e-06,
+      "loss": 0.899,
+      "step": 8804
+    },
+    {
+      "epoch": 1.7443342245321207,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.6958982701954536e-06,
+      "loss": 0.8745,
+      "step": 8805
+    },
+    {
+      "epoch": 1.7445341195872168,
+      "grad_norm": 2.125,
+      "learning_rate": 3.6948809726456525e-06,
+      "loss": 0.9022,
+      "step": 8806
+    },
+    {
+      "epoch": 1.744734014642313,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.6938637330690967e-06,
+      "loss": 0.9966,
+      "step": 8807
+    },
+    {
+      "epoch": 1.744933909697409,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.6928465515109756e-06,
+      "loss": 0.9684,
+      "step": 8808
+    },
+    {
+      "epoch": 1.745133804752505,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.69182942801647e-06,
+      "loss": 0.9814,
+      "step": 8809
+    },
+    {
+      "epoch": 1.745333699807601,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.6908123626307618e-06,
+      "loss": 1.0867,
+      "step": 8810
+    },
+    {
+      "epoch": 1.745533594862697,
+      "grad_norm": 2.125,
+      "learning_rate": 3.6897953553990296e-06,
+      "loss": 0.9418,
+      "step": 8811
+    },
+    {
+      "epoch": 1.7457334899177932,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.6887784063664467e-06,
+      "loss": 0.9365,
+      "step": 8812
+    },
+    {
+      "epoch": 1.745933384972889,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.6877615155781863e-06,
+      "loss": 0.9061,
+      "step": 8813
+    },
+    {
+      "epoch": 1.7461332800279852,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.6867446830794205e-06,
+      "loss": 0.9976,
+      "step": 8814
+    },
+    {
+      "epoch": 1.7463331750830813,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.6857279089153154e-06,
+      "loss": 1.0071,
+      "step": 8815
+    },
+    {
+      "epoch": 1.7465330701381774,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.6847111931310363e-06,
+      "loss": 1.0236,
+      "step": 8816
+    },
+    {
+      "epoch": 1.7467329651932735,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.683694535771745e-06,
+      "loss": 0.9984,
+      "step": 8817
+    },
+    {
+      "epoch": 1.7469328602483696,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.6826779368826015e-06,
+      "loss": 0.8575,
+      "step": 8818
+    },
+    {
+      "epoch": 1.7471327553034657,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.681661396508765e-06,
+      "loss": 0.9144,
+      "step": 8819
+    },
+    {
+      "epoch": 1.7473326503585618,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.680644914695387e-06,
+      "loss": 0.9785,
+      "step": 8820
+    },
+    {
+      "epoch": 1.7475325454136579,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.679628491487621e-06,
+      "loss": 0.9525,
+      "step": 8821
+    },
+    {
+      "epoch": 1.747732440468754,
+      "grad_norm": 2.125,
+      "learning_rate": 3.6786121269306173e-06,
+      "loss": 1.0481,
+      "step": 8822
+    },
+    {
+      "epoch": 1.74793233552385,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.677595821069521e-06,
+      "loss": 0.9572,
+      "step": 8823
+    },
+    {
+      "epoch": 1.7481322305789462,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.676579573949477e-06,
+      "loss": 0.9119,
+      "step": 8824
+    },
+    {
+      "epoch": 1.7483321256340423,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.6755633856156283e-06,
+      "loss": 1.0921,
+      "step": 8825
+    },
+    {
+      "epoch": 1.7485320206891382,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.6745472561131107e-06,
+      "loss": 0.9407,
+      "step": 8826
+    },
+    {
+      "epoch": 1.7487319157442343,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.673531185487064e-06,
+      "loss": 1.0492,
+      "step": 8827
+    },
+    {
+      "epoch": 1.7489318107993304,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.6725151737826193e-06,
+      "loss": 0.914,
+      "step": 8828
+    },
+    {
+      "epoch": 1.7491317058544265,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.6714992210449084e-06,
+      "loss": 1.0461,
+      "step": 8829
+    },
+    {
+      "epoch": 1.7493316009095226,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.670483327319062e-06,
+      "loss": 0.965,
+      "step": 8830
+    },
+    {
+      "epoch": 1.7495314959646184,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.669467492650203e-06,
+      "loss": 1.0768,
+      "step": 8831
+    },
+    {
+      "epoch": 1.7497313910197145,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.6684517170834554e-06,
+      "loss": 0.9636,
+      "step": 8832
+    },
+    {
+      "epoch": 1.7499312860748106,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.667436000663942e-06,
+      "loss": 0.9212,
+      "step": 8833
+    },
+    {
+      "epoch": 1.7501311811299067,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.666420343436777e-06,
+      "loss": 0.9206,
+      "step": 8834
+    },
+    {
+      "epoch": 1.7503310761850028,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.66540474544708e-06,
+      "loss": 0.8463,
+      "step": 8835
+    },
+    {
+      "epoch": 1.750530971240099,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.6643892067399604e-06,
+      "loss": 0.9841,
+      "step": 8836
+    },
+    {
+      "epoch": 1.750730866295195,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.663373727360532e-06,
+      "loss": 0.9349,
+      "step": 8837
+    },
+    {
+      "epoch": 1.7509307613502911,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.662358307353897e-06,
+      "loss": 0.9983,
+      "step": 8838
+    },
+    {
+      "epoch": 1.7511306564053872,
+      "grad_norm": 2.125,
+      "learning_rate": 3.6613429467651644e-06,
+      "loss": 0.8888,
+      "step": 8839
+    },
+    {
+      "epoch": 1.7513305514604833,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.6603276456394342e-06,
+      "loss": 0.8516,
+      "step": 8840
+    },
+    {
+      "epoch": 1.7515304465155794,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.659312404021807e-06,
+      "loss": 0.9168,
+      "step": 8841
+    },
+    {
+      "epoch": 1.7517303415706755,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.6582972219573808e-06,
+      "loss": 1.1102,
+      "step": 8842
+    },
+    {
+      "epoch": 1.7519302366257716,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.6572820994912474e-06,
+      "loss": 0.922,
+      "step": 8843
+    },
+    {
+      "epoch": 1.7521301316808675,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.6562670366684995e-06,
+      "loss": 0.9361,
+      "step": 8844
+    },
+    {
+      "epoch": 1.7523300267359636,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.655252033534228e-06,
+      "loss": 1.0295,
+      "step": 8845
+    },
+    {
+      "epoch": 1.7525299217910597,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.6542370901335163e-06,
+      "loss": 0.9455,
+      "step": 8846
+    },
+    {
+      "epoch": 1.7527298168461558,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.65322220651145e-06,
+      "loss": 0.9995,
+      "step": 8847
+    },
+    {
+      "epoch": 1.7529297119012517,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.652207382713109e-06,
+      "loss": 1.013,
+      "step": 8848
+    },
+    {
+      "epoch": 1.7531296069563478,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.651192618783571e-06,
+      "loss": 0.9757,
+      "step": 8849
+    },
+    {
+      "epoch": 1.7533295020114439,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.650177914767915e-06,
+      "loss": 0.9842,
+      "step": 8850
+    },
+    {
+      "epoch": 1.75352939706654,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.64916327071121e-06,
+      "loss": 0.9903,
+      "step": 8851
+    },
+    {
+      "epoch": 1.753729292121636,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.6481486866585292e-06,
+      "loss": 1.0488,
+      "step": 8852
+    },
+    {
+      "epoch": 1.7539291871767322,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.6471341626549397e-06,
+      "loss": 0.9775,
+      "step": 8853
+    },
+    {
+      "epoch": 1.7541290822318283,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.646119698745505e-06,
+      "loss": 0.943,
+      "step": 8854
+    },
+    {
+      "epoch": 1.7543289772869244,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.6451052949752902e-06,
+      "loss": 1.0549,
+      "step": 8855
+    },
+    {
+      "epoch": 1.7545288723420205,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6440909513893517e-06,
+      "loss": 0.9492,
+      "step": 8856
+    },
+    {
+      "epoch": 1.7547287673971166,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.643076668032749e-06,
+      "loss": 0.96,
+      "step": 8857
+    },
+    {
+      "epoch": 1.7549286624522127,
+      "grad_norm": 2.125,
+      "learning_rate": 3.642062444950537e-06,
+      "loss": 0.9833,
+      "step": 8858
+    },
+    {
+      "epoch": 1.7551285575073088,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.6410482821877647e-06,
+      "loss": 0.9437,
+      "step": 8859
+    },
+    {
+      "epoch": 1.7553284525624049,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.640034179789483e-06,
+      "loss": 0.9675,
+      "step": 8860
+    },
+    {
+      "epoch": 1.7555283476175008,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.6390201378007383e-06,
+      "loss": 0.9612,
+      "step": 8861
+    },
+    {
+      "epoch": 1.7557282426725969,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.638006156266573e-06,
+      "loss": 0.94,
+      "step": 8862
+    },
+    {
+      "epoch": 1.755928137727693,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.6369922352320285e-06,
+      "loss": 0.969,
+      "step": 8863
+    },
+    {
+      "epoch": 1.756128032782789,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.635978374742145e-06,
+      "loss": 0.8889,
+      "step": 8864
+    },
+    {
+      "epoch": 1.7563279278378852,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.634964574841955e-06,
+      "loss": 0.8754,
+      "step": 8865
+    },
+    {
+      "epoch": 1.756527822892981,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.6339508355764935e-06,
+      "loss": 0.9049,
+      "step": 8866
+    },
+    {
+      "epoch": 1.7567277179480771,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.632937156990789e-06,
+      "loss": 0.8882,
+      "step": 8867
+    },
+    {
+      "epoch": 1.7569276130031732,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.6319235391298724e-06,
+      "loss": 0.9551,
+      "step": 8868
+    },
+    {
+      "epoch": 1.7571275080582693,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.630909982038764e-06,
+      "loss": 1.0042,
+      "step": 8869
+    },
+    {
+      "epoch": 1.7573274031133654,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.6298964857624885e-06,
+      "loss": 1.1014,
+      "step": 8870
+    },
+    {
+      "epoch": 1.7575272981684615,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.628883050346065e-06,
+      "loss": 0.8785,
+      "step": 8871
+    },
+    {
+      "epoch": 1.7577271932235576,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.627869675834509e-06,
+      "loss": 0.992,
+      "step": 8872
+    },
+    {
+      "epoch": 1.7579270882786537,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.626856362272836e-06,
+      "loss": 0.9159,
+      "step": 8873
+    },
+    {
+      "epoch": 1.7581269833337498,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.6258431097060567e-06,
+      "loss": 1.1005,
+      "step": 8874
+    },
+    {
+      "epoch": 1.758326878388846,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.624829918179179e-06,
+      "loss": 0.9464,
+      "step": 8875
+    },
+    {
+      "epoch": 1.758526773443942,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.623816787737211e-06,
+      "loss": 0.9043,
+      "step": 8876
+    },
+    {
+      "epoch": 1.7587266684990381,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.622803718425153e-06,
+      "loss": 0.9667,
+      "step": 8877
+    },
+    {
+      "epoch": 1.7589265635541342,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.6217907102880075e-06,
+      "loss": 0.9796,
+      "step": 8878
+    },
+    {
+      "epoch": 1.7591264586092301,
+      "grad_norm": 2.25,
+      "learning_rate": 3.620777763370771e-06,
+      "loss": 0.9393,
+      "step": 8879
+    },
+    {
+      "epoch": 1.7593263536643262,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.619764877718438e-06,
+      "loss": 1.1024,
+      "step": 8880
+    },
+    {
+      "epoch": 1.7595262487194223,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.6187520533760034e-06,
+      "loss": 0.9499,
+      "step": 8881
+    },
+    {
+      "epoch": 1.7597261437745184,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.6177392903884545e-06,
+      "loss": 1.0295,
+      "step": 8882
+    },
+    {
+      "epoch": 1.7599260388296143,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.616726588800778e-06,
+      "loss": 0.9496,
+      "step": 8883
+    },
+    {
+      "epoch": 1.7601259338847104,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.61571394865796e-06,
+      "loss": 0.9193,
+      "step": 8884
+    },
+    {
+      "epoch": 1.7603258289398065,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.6147013700049793e-06,
+      "loss": 0.974,
+      "step": 8885
+    },
+    {
+      "epoch": 1.7605257239949026,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.6136888528868177e-06,
+      "loss": 0.9011,
+      "step": 8886
+    },
+    {
+      "epoch": 1.7607256190499987,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.612676397348447e-06,
+      "loss": 0.9091,
+      "step": 8887
+    },
+    {
+      "epoch": 1.7609255141050948,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.6116640034348426e-06,
+      "loss": 0.9124,
+      "step": 8888
+    },
+    {
+      "epoch": 1.761125409160191,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.610651671190977e-06,
+      "loss": 0.9292,
+      "step": 8889
+    },
+    {
+      "epoch": 1.761325304215287,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.609639400661814e-06,
+      "loss": 0.9587,
+      "step": 8890
+    },
+    {
+      "epoch": 1.761525199270383,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.60862719189232e-06,
+      "loss": 0.9473,
+      "step": 8891
+    },
+    {
+      "epoch": 1.7617250943254792,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.6076150449274595e-06,
+      "loss": 0.9856,
+      "step": 8892
+    },
+    {
+      "epoch": 1.7619249893805753,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.6066029598121883e-06,
+      "loss": 0.9389,
+      "step": 8893
+    },
+    {
+      "epoch": 1.7621248844356714,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.605590936591466e-06,
+      "loss": 0.9567,
+      "step": 8894
+    },
+    {
+      "epoch": 1.7623247794907675,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.6045789753102443e-06,
+      "loss": 0.9552,
+      "step": 8895
+    },
+    {
+      "epoch": 1.7625246745458634,
+      "grad_norm": 2.25,
+      "learning_rate": 3.6035670760134756e-06,
+      "loss": 1.0126,
+      "step": 8896
+    },
+    {
+      "epoch": 1.7627245696009595,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.6025552387461086e-06,
+      "loss": 0.9932,
+      "step": 8897
+    },
+    {
+      "epoch": 1.7629244646560556,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.6015434635530888e-06,
+      "loss": 1.0051,
+      "step": 8898
+    },
+    {
+      "epoch": 1.7631243597111517,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.6005317504793604e-06,
+      "loss": 0.9816,
+      "step": 8899
+    },
+    {
+      "epoch": 1.7633242547662478,
+      "grad_norm": 2.125,
+      "learning_rate": 3.59952009956986e-06,
+      "loss": 0.9812,
+      "step": 8900
+    },
+    {
+      "epoch": 1.7635241498213436,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5985085108695285e-06,
+      "loss": 0.975,
+      "step": 8901
+    },
+    {
+      "epoch": 1.7637240448764397,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5974969844232974e-06,
+      "loss": 0.9464,
+      "step": 8902
+    },
+    {
+      "epoch": 1.7639239399315358,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5964855202761016e-06,
+      "loss": 0.939,
+      "step": 8903
+    },
+    {
+      "epoch": 1.764123834986632,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.5954741184728698e-06,
+      "loss": 0.8953,
+      "step": 8904
+    },
+    {
+      "epoch": 1.764323730041728,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.5944627790585264e-06,
+      "loss": 1.086,
+      "step": 8905
+    },
+    {
+      "epoch": 1.7645236250968241,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.5934515020779974e-06,
+      "loss": 0.931,
+      "step": 8906
+    },
+    {
+      "epoch": 1.7647235201519202,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.592440287576201e-06,
+      "loss": 0.9439,
+      "step": 8907
+    },
+    {
+      "epoch": 1.7649234152070163,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.5914291355980573e-06,
+      "loss": 0.9707,
+      "step": 8908
+    },
+    {
+      "epoch": 1.7651233102621124,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.5904180461884818e-06,
+      "loss": 0.9281,
+      "step": 8909
+    },
+    {
+      "epoch": 1.7653232053172085,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.589407019392385e-06,
+      "loss": 0.9643,
+      "step": 8910
+    },
+    {
+      "epoch": 1.7655231003723046,
+      "grad_norm": 2.125,
+      "learning_rate": 3.5883960552546783e-06,
+      "loss": 0.9312,
+      "step": 8911
+    },
+    {
+      "epoch": 1.7657229954274007,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.587385153820269e-06,
+      "loss": 0.9281,
+      "step": 8912
+    },
+    {
+      "epoch": 1.7659228904824968,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.586374315134058e-06,
+      "loss": 1.0413,
+      "step": 8913
+    },
+    {
+      "epoch": 1.7661227855375927,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.5853635392409504e-06,
+      "loss": 0.9381,
+      "step": 8914
+    },
+    {
+      "epoch": 1.7663226805926888,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.5843528261858434e-06,
+      "loss": 0.9366,
+      "step": 8915
+    },
+    {
+      "epoch": 1.766522575647785,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.5833421760136323e-06,
+      "loss": 0.9485,
+      "step": 8916
+    },
+    {
+      "epoch": 1.766722470702881,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.582331588769211e-06,
+      "loss": 0.9681,
+      "step": 8917
+    },
+    {
+      "epoch": 1.766922365757977,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5813210644974685e-06,
+      "loss": 0.8463,
+      "step": 8918
+    },
+    {
+      "epoch": 1.767122260813073,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.5803106032432923e-06,
+      "loss": 0.9746,
+      "step": 8919
+    },
+    {
+      "epoch": 1.767322155868169,
+      "grad_norm": 2.125,
+      "learning_rate": 3.5793002050515686e-06,
+      "loss": 0.8872,
+      "step": 8920
+    },
+    {
+      "epoch": 1.7675220509232652,
+      "grad_norm": 2.125,
+      "learning_rate": 3.578289869967177e-06,
+      "loss": 0.8397,
+      "step": 8921
+    },
+    {
+      "epoch": 1.7677219459783613,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.5772795980349976e-06,
+      "loss": 0.993,
+      "step": 8922
+    },
+    {
+      "epoch": 1.7679218410334574,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.576269389299908e-06,
+      "loss": 0.9071,
+      "step": 8923
+    },
+    {
+      "epoch": 1.7681217360885535,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.5752592438067785e-06,
+      "loss": 1.0297,
+      "step": 8924
+    },
+    {
+      "epoch": 1.7683216311436496,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.5742491616004828e-06,
+      "loss": 0.9557,
+      "step": 8925
+    },
+    {
+      "epoch": 1.7685215261987457,
+      "grad_norm": 2.125,
+      "learning_rate": 3.573239142725885e-06,
+      "loss": 0.9459,
+      "step": 8926
+    },
+    {
+      "epoch": 1.7687214212538418,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.572229187227853e-06,
+      "loss": 1.0032,
+      "step": 8927
+    },
+    {
+      "epoch": 1.768921316308938,
+      "grad_norm": 2.125,
+      "learning_rate": 3.571219295151249e-06,
+      "loss": 1.0218,
+      "step": 8928
+    },
+    {
+      "epoch": 1.769121211364034,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.57020946654093e-06,
+      "loss": 1.0085,
+      "step": 8929
+    },
+    {
+      "epoch": 1.76932110641913,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.5691997014417558e-06,
+      "loss": 1.0393,
+      "step": 8930
+    },
+    {
+      "epoch": 1.7695210014742262,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.568189999898576e-06,
+      "loss": 0.9354,
+      "step": 8931
+    },
+    {
+      "epoch": 1.769720896529322,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.567180361956245e-06,
+      "loss": 0.9892,
+      "step": 8932
+    },
+    {
+      "epoch": 1.7699207915844182,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.5661707876596073e-06,
+      "loss": 0.9455,
+      "step": 8933
+    },
+    {
+      "epoch": 1.7701206866395143,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.565161277053511e-06,
+      "loss": 0.9616,
+      "step": 8934
+    },
+    {
+      "epoch": 1.7703205816946104,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.5641518301827983e-06,
+      "loss": 0.9205,
+      "step": 8935
+    },
+    {
+      "epoch": 1.7705204767497063,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.563142447092307e-06,
+      "loss": 1.0233,
+      "step": 8936
+    },
+    {
+      "epoch": 1.7707203718048024,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.5621331278268754e-06,
+      "loss": 1.0387,
+      "step": 8937
+    },
+    {
+      "epoch": 1.7709202668598985,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.5611238724313357e-06,
+      "loss": 0.9357,
+      "step": 8938
+    },
+    {
+      "epoch": 1.7711201619149946,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.5601146809505204e-06,
+      "loss": 0.8826,
+      "step": 8939
+    },
+    {
+      "epoch": 1.7713200569700907,
+      "grad_norm": 2.25,
+      "learning_rate": 3.5591055534292573e-06,
+      "loss": 1.0714,
+      "step": 8940
+    },
+    {
+      "epoch": 1.7715199520251868,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.5580964899123715e-06,
+      "loss": 0.9932,
+      "step": 8941
+    },
+    {
+      "epoch": 1.7717198470802829,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.557087490444685e-06,
+      "loss": 0.8739,
+      "step": 8942
+    },
+    {
+      "epoch": 1.771919742135379,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.55607855507102e-06,
+      "loss": 1.0024,
+      "step": 8943
+    },
+    {
+      "epoch": 1.772119637190475,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.55506968383619e-06,
+      "loss": 0.9236,
+      "step": 8944
+    },
+    {
+      "epoch": 1.7723195322455711,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.5540608767850106e-06,
+      "loss": 0.985,
+      "step": 8945
+    },
+    {
+      "epoch": 1.7725194273006672,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.5530521339622923e-06,
+      "loss": 0.985,
+      "step": 8946
+    },
+    {
+      "epoch": 1.7727193223557633,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.5520434554128437e-06,
+      "loss": 1.0439,
+      "step": 8947
+    },
+    {
+      "epoch": 1.7729192174108594,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.5510348411814705e-06,
+      "loss": 1.0106,
+      "step": 8948
+    },
+    {
+      "epoch": 1.7731191124659553,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.5500262913129745e-06,
+      "loss": 0.9443,
+      "step": 8949
+    },
+    {
+      "epoch": 1.7733190075210514,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.5490178058521553e-06,
+      "loss": 0.9458,
+      "step": 8950
+    },
+    {
+      "epoch": 1.7735189025761475,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.548009384843811e-06,
+      "loss": 0.9217,
+      "step": 8951
+    },
+    {
+      "epoch": 1.7737187976312436,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.547001028332735e-06,
+      "loss": 0.9978,
+      "step": 8952
+    },
+    {
+      "epoch": 1.7739186926863397,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.545992736363717e-06,
+      "loss": 1.0665,
+      "step": 8953
+    },
+    {
+      "epoch": 1.7741185877414356,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.544984508981548e-06,
+      "loss": 0.9647,
+      "step": 8954
+    },
+    {
+      "epoch": 1.7743184827965317,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.5439763462310107e-06,
+      "loss": 0.8975,
+      "step": 8955
+    },
+    {
+      "epoch": 1.7745183778516278,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.5429682481568894e-06,
+      "loss": 1.015,
+      "step": 8956
+    },
+    {
+      "epoch": 1.774718272906724,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.5419602148039618e-06,
+      "loss": 0.8829,
+      "step": 8957
+    },
+    {
+      "epoch": 1.77491816796182,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.5409522462170054e-06,
+      "loss": 0.9894,
+      "step": 8958
+    },
+    {
+      "epoch": 1.775118063016916,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.539944342440796e-06,
+      "loss": 1.0176,
+      "step": 8959
+    },
+    {
+      "epoch": 1.7753179580720122,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.5389365035201016e-06,
+      "loss": 0.9431,
+      "step": 8960
+    },
+    {
+      "epoch": 1.7755178531271083,
+      "grad_norm": 2.25,
+      "learning_rate": 3.537928729499694e-06,
+      "loss": 0.9652,
+      "step": 8961
+    },
+    {
+      "epoch": 1.7757177481822044,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.536921020424334e-06,
+      "loss": 0.8999,
+      "step": 8962
+    },
+    {
+      "epoch": 1.7759176432373005,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.5359133763387866e-06,
+      "loss": 1.0306,
+      "step": 8963
+    },
+    {
+      "epoch": 1.7761175382923966,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.53490579728781e-06,
+      "loss": 1.0175,
+      "step": 8964
+    },
+    {
+      "epoch": 1.7763174333474927,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.533898283316162e-06,
+      "loss": 0.9064,
+      "step": 8965
+    },
+    {
+      "epoch": 1.7765173284025888,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5328908344685952e-06,
+      "loss": 0.9434,
+      "step": 8966
+    },
+    {
+      "epoch": 1.7767172234576847,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.5318834507898607e-06,
+      "loss": 0.9599,
+      "step": 8967
+    },
+    {
+      "epoch": 1.7769171185127808,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5308761323247077e-06,
+      "loss": 0.9761,
+      "step": 8968
+    },
+    {
+      "epoch": 1.7771170135678769,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5298688791178788e-06,
+      "loss": 0.9672,
+      "step": 8969
+    },
+    {
+      "epoch": 1.777316908622973,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.528861691214117e-06,
+      "loss": 0.9342,
+      "step": 8970
+    },
+    {
+      "epoch": 1.7775168036780689,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.5278545686581633e-06,
+      "loss": 1.0538,
+      "step": 8971
+    },
+    {
+      "epoch": 1.777716698733165,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.526847511494751e-06,
+      "loss": 0.9025,
+      "step": 8972
+    },
+    {
+      "epoch": 1.777916593788261,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.5258405197686154e-06,
+      "loss": 0.9172,
+      "step": 8973
+    },
+    {
+      "epoch": 1.7781164888433572,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.524833593524487e-06,
+      "loss": 0.8952,
+      "step": 8974
+    },
+    {
+      "epoch": 1.7783163838984533,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.523826732807092e-06,
+      "loss": 0.9515,
+      "step": 8975
+    },
+    {
+      "epoch": 1.7785162789535494,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5228199376611564e-06,
+      "loss": 0.923,
+      "step": 8976
+    },
+    {
+      "epoch": 1.7787161740086455,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.521813208131401e-06,
+      "loss": 0.9622,
+      "step": 8977
+    },
+    {
+      "epoch": 1.7789160690637416,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.520806544262545e-06,
+      "loss": 1.0126,
+      "step": 8978
+    },
+    {
+      "epoch": 1.7791159641188377,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.519799946099305e-06,
+      "loss": 0.8961,
+      "step": 8979
+    },
+    {
+      "epoch": 1.7793158591739338,
+      "grad_norm": 2.125,
+      "learning_rate": 3.518793413686392e-06,
+      "loss": 1.0055,
+      "step": 8980
+    },
+    {
+      "epoch": 1.7795157542290299,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.5177869470685175e-06,
+      "loss": 0.9555,
+      "step": 8981
+    },
+    {
+      "epoch": 1.779715649284126,
+      "grad_norm": 2.25,
+      "learning_rate": 3.5167805462903903e-06,
+      "loss": 0.9801,
+      "step": 8982
+    },
+    {
+      "epoch": 1.779915544339222,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.5157742113967113e-06,
+      "loss": 0.9423,
+      "step": 8983
+    },
+    {
+      "epoch": 1.780115439394318,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5147679424321848e-06,
+      "loss": 0.9504,
+      "step": 8984
+    },
+    {
+      "epoch": 1.780315334449414,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5137617394415057e-06,
+      "loss": 1.0039,
+      "step": 8985
+    },
+    {
+      "epoch": 1.7805152295045101,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.512755602469372e-06,
+      "loss": 1.0008,
+      "step": 8986
+    },
+    {
+      "epoch": 1.7807151245596062,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5117495315604766e-06,
+      "loss": 0.9502,
+      "step": 8987
+    },
+    {
+      "epoch": 1.7809150196147023,
+      "grad_norm": 2.125,
+      "learning_rate": 3.5107435267595067e-06,
+      "loss": 0.9923,
+      "step": 8988
+    },
+    {
+      "epoch": 1.7811149146697982,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.5097375881111506e-06,
+      "loss": 1.0216,
+      "step": 8989
+    },
+    {
+      "epoch": 1.7813148097248943,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.508731715660093e-06,
+      "loss": 0.9487,
+      "step": 8990
+    },
+    {
+      "epoch": 1.7815147047799904,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.5077259094510118e-06,
+      "loss": 1.0448,
+      "step": 8991
+    },
+    {
+      "epoch": 1.7817145998350865,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.5067201695285884e-06,
+      "loss": 0.9856,
+      "step": 8992
+    },
+    {
+      "epoch": 1.7819144948901826,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.5057144959374934e-06,
+      "loss": 1.033,
+      "step": 8993
+    },
+    {
+      "epoch": 1.7821143899452787,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.5047088887224024e-06,
+      "loss": 0.9622,
+      "step": 8994
+    },
+    {
+      "epoch": 1.7823142850003748,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.5037033479279813e-06,
+      "loss": 0.8222,
+      "step": 8995
+    },
+    {
+      "epoch": 1.782514180055471,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.502697873598899e-06,
+      "loss": 0.9553,
+      "step": 8996
+    },
+    {
+      "epoch": 1.782714075110567,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.5016924657798166e-06,
+      "loss": 0.9852,
+      "step": 8997
+    },
+    {
+      "epoch": 1.782913970165663,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.5006871245153947e-06,
+      "loss": 0.9359,
+      "step": 8998
+    },
+    {
+      "epoch": 1.7831138652207592,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.499681849850291e-06,
+      "loss": 0.9947,
+      "step": 8999
+    },
+    {
+      "epoch": 1.7833137602758553,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4986766418291584e-06,
+      "loss": 1.0271,
+      "step": 9000
+    },
+    {
+      "epoch": 1.7833137602758553,
+      "eval_loss": 0.897596001625061,
+      "eval_runtime": 593.4535,
+      "eval_samples_per_second": 3.603,
+      "eval_steps_per_second": 3.603,
+      "step": 9000
+    },
+    {
+      "epoch": 1.7835136553309514,
+      "grad_norm": 2.25,
+      "learning_rate": 3.4976715004966492e-06,
+      "loss": 1.1224,
+      "step": 9001
+    },
+    {
+      "epoch": 1.7837135503860473,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4966664258974128e-06,
+      "loss": 0.8661,
+      "step": 9002
+    },
+    {
+      "epoch": 1.7839134454411434,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.4956614180760918e-06,
+      "loss": 0.8782,
+      "step": 9003
+    },
+    {
+      "epoch": 1.7841133404962395,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4946564770773305e-06,
+      "loss": 0.9052,
+      "step": 9004
+    },
+    {
+      "epoch": 1.7843132355513356,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.493651602945768e-06,
+      "loss": 0.9965,
+      "step": 9005
+    },
+    {
+      "epoch": 1.7845131306064315,
+      "grad_norm": 2.125,
+      "learning_rate": 3.49264679572604e-06,
+      "loss": 0.9936,
+      "step": 9006
+    },
+    {
+      "epoch": 1.7847130256615276,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.4916420554627806e-06,
+      "loss": 0.9557,
+      "step": 9007
+    },
+    {
+      "epoch": 1.7849129207166237,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.49063738220062e-06,
+      "loss": 0.8943,
+      "step": 9008
+    },
+    {
+      "epoch": 1.7851128157717198,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.4896327759841854e-06,
+      "loss": 0.9469,
+      "step": 9009
+    },
+    {
+      "epoch": 1.7853127108268159,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.4886282368581014e-06,
+      "loss": 1.0564,
+      "step": 9010
+    },
+    {
+      "epoch": 1.785512605881912,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4876237648669896e-06,
+      "loss": 0.9653,
+      "step": 9011
+    },
+    {
+      "epoch": 1.785712500937008,
+      "grad_norm": 2.375,
+      "learning_rate": 3.486619360055469e-06,
+      "loss": 1.0446,
+      "step": 9012
+    },
+    {
+      "epoch": 1.7859123959921042,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4856150224681557e-06,
+      "loss": 0.9555,
+      "step": 9013
+    },
+    {
+      "epoch": 1.7861122910472003,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.48461075214966e-06,
+      "loss": 0.9931,
+      "step": 9014
+    },
+    {
+      "epoch": 1.7863121861022964,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.4836065491445935e-06,
+      "loss": 1.013,
+      "step": 9015
+    },
+    {
+      "epoch": 1.7865120811573925,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.482602413497562e-06,
+      "loss": 0.9331,
+      "step": 9016
+    },
+    {
+      "epoch": 1.7867119762124886,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.481598345253169e-06,
+      "loss": 1.0055,
+      "step": 9017
+    },
+    {
+      "epoch": 1.7869118712675847,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.480594344456016e-06,
+      "loss": 1.0414,
+      "step": 9018
+    },
+    {
+      "epoch": 1.7871117663226805,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.4795904111506988e-06,
+      "loss": 1.0796,
+      "step": 9019
+    },
+    {
+      "epoch": 1.7873116613777766,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4785865453818135e-06,
+      "loss": 0.9932,
+      "step": 9020
+    },
+    {
+      "epoch": 1.7875115564328727,
+      "grad_norm": 2.125,
+      "learning_rate": 3.477582747193953e-06,
+      "loss": 1.0241,
+      "step": 9021
+    },
+    {
+      "epoch": 1.7877114514879688,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.4765790166317036e-06,
+      "loss": 0.9822,
+      "step": 9022
+    },
+    {
+      "epoch": 1.787911346543065,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.475575353739651e-06,
+      "loss": 0.9923,
+      "step": 9023
+    },
+    {
+      "epoch": 1.7881112415981608,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4745717585623778e-06,
+      "loss": 0.9738,
+      "step": 9024
+    },
+    {
+      "epoch": 1.788311136653257,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.473568231144465e-06,
+      "loss": 1.0057,
+      "step": 9025
+    },
+    {
+      "epoch": 1.788511031708353,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.4725647715304876e-06,
+      "loss": 0.9126,
+      "step": 9026
+    },
+    {
+      "epoch": 1.7887109267634491,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4715613797650216e-06,
+      "loss": 0.9076,
+      "step": 9027
+    },
+    {
+      "epoch": 1.7889108218185452,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.4705580558926337e-06,
+      "loss": 0.9161,
+      "step": 9028
+    },
+    {
+      "epoch": 1.7891107168736413,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4695547999578946e-06,
+      "loss": 0.9221,
+      "step": 9029
+    },
+    {
+      "epoch": 1.7893106119287374,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.468551612005368e-06,
+      "loss": 0.9742,
+      "step": 9030
+    },
+    {
+      "epoch": 1.7895105069838335,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4675484920796143e-06,
+      "loss": 1.0193,
+      "step": 9031
+    },
+    {
+      "epoch": 1.7897104020389296,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.4665454402251937e-06,
+      "loss": 0.9693,
+      "step": 9032
+    },
+    {
+      "epoch": 1.7899102970940257,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4655424564866615e-06,
+      "loss": 0.9386,
+      "step": 9033
+    },
+    {
+      "epoch": 1.7901101921491218,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.464539540908568e-06,
+      "loss": 1.0275,
+      "step": 9034
+    },
+    {
+      "epoch": 1.790310087204218,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.4635366935354654e-06,
+      "loss": 0.9418,
+      "step": 9035
+    },
+    {
+      "epoch": 1.790509982259314,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.462533914411898e-06,
+      "loss": 0.9292,
+      "step": 9036
+    },
+    {
+      "epoch": 1.79070987731441,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.4615312035824093e-06,
+      "loss": 0.928,
+      "step": 9037
+    },
+    {
+      "epoch": 1.790909772369506,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4605285610915417e-06,
+      "loss": 0.9658,
+      "step": 9038
+    },
+    {
+      "epoch": 1.791109667424602,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.4595259869838306e-06,
+      "loss": 0.9019,
+      "step": 9039
+    },
+    {
+      "epoch": 1.7913095624796982,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.4585234813038095e-06,
+      "loss": 1.0131,
+      "step": 9040
+    },
+    {
+      "epoch": 1.791509457534794,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.457521044096012e-06,
+      "loss": 0.9699,
+      "step": 9041
+    },
+    {
+      "epoch": 1.7917093525898902,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4565186754049647e-06,
+      "loss": 1.0297,
+      "step": 9042
+    },
+    {
+      "epoch": 1.7919092476449863,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.455516375275192e-06,
+      "loss": 0.9788,
+      "step": 9043
+    },
+    {
+      "epoch": 1.7921091427000824,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.4545141437512186e-06,
+      "loss": 0.9358,
+      "step": 9044
+    },
+    {
+      "epoch": 1.7923090377551785,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.4535119808775608e-06,
+      "loss": 1.0411,
+      "step": 9045
+    },
+    {
+      "epoch": 1.7925089328102746,
+      "grad_norm": 2.25,
+      "learning_rate": 3.4525098866987366e-06,
+      "loss": 1.0401,
+      "step": 9046
+    },
+    {
+      "epoch": 1.7927088278653707,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.451507861259257e-06,
+      "loss": 1.0765,
+      "step": 9047
+    },
+    {
+      "epoch": 1.7929087229204668,
+      "grad_norm": 2.0,
+      "learning_rate": 3.4505059046036326e-06,
+      "loss": 0.9126,
+      "step": 9048
+    },
+    {
+      "epoch": 1.7931086179755629,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.4495040167763717e-06,
+      "loss": 0.8841,
+      "step": 9049
+    },
+    {
+      "epoch": 1.793308513030659,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.4485021978219755e-06,
+      "loss": 0.9835,
+      "step": 9050
+    },
+    {
+      "epoch": 1.793508408085755,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.447500447784946e-06,
+      "loss": 1.0108,
+      "step": 9051
+    },
+    {
+      "epoch": 1.7937083031408512,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.4464987667097818e-06,
+      "loss": 1.069,
+      "step": 9052
+    },
+    {
+      "epoch": 1.7939081981959473,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4454971546409777e-06,
+      "loss": 0.9084,
+      "step": 9053
+    },
+    {
+      "epoch": 1.7941080932510434,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4444956116230216e-06,
+      "loss": 0.9625,
+      "step": 9054
+    },
+    {
+      "epoch": 1.7943079883061392,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.443494137700405e-06,
+      "loss": 0.9626,
+      "step": 9055
+    },
+    {
+      "epoch": 1.7945078833612353,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.4424927329176128e-06,
+      "loss": 1.0615,
+      "step": 9056
+    },
+    {
+      "epoch": 1.7947077784163314,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.441491397319126e-06,
+      "loss": 0.9225,
+      "step": 9057
+    },
+    {
+      "epoch": 1.7949076734714275,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.4404901309494264e-06,
+      "loss": 0.9529,
+      "step": 9058
+    },
+    {
+      "epoch": 1.7951075685265234,
+      "grad_norm": 2.25,
+      "learning_rate": 3.4394889338529874e-06,
+      "loss": 1.0643,
+      "step": 9059
+    },
+    {
+      "epoch": 1.7953074635816195,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4384878060742833e-06,
+      "loss": 0.9179,
+      "step": 9060
+    },
+    {
+      "epoch": 1.7955073586367156,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.437486747657785e-06,
+      "loss": 0.9156,
+      "step": 9061
+    },
+    {
+      "epoch": 1.7957072536918117,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4364857586479576e-06,
+      "loss": 0.9051,
+      "step": 9062
+    },
+    {
+      "epoch": 1.7959071487469078,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.435484839089266e-06,
+      "loss": 0.99,
+      "step": 9063
+    },
+    {
+      "epoch": 1.796107043802004,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.4344839890261718e-06,
+      "loss": 0.9345,
+      "step": 9064
+    },
+    {
+      "epoch": 1.7963069388571,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.43348320850313e-06,
+      "loss": 0.8536,
+      "step": 9065
+    },
+    {
+      "epoch": 1.7965068339121961,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4324824975645984e-06,
+      "loss": 0.8766,
+      "step": 9066
+    },
+    {
+      "epoch": 1.7967067289672922,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.4314818562550254e-06,
+      "loss": 1.0165,
+      "step": 9067
+    },
+    {
+      "epoch": 1.7969066240223883,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.430481284618861e-06,
+      "loss": 0.8893,
+      "step": 9068
+    },
+    {
+      "epoch": 1.7971065190774844,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.429480782700551e-06,
+      "loss": 1.0252,
+      "step": 9069
+    },
+    {
+      "epoch": 1.7973064141325805,
+      "grad_norm": 2.25,
+      "learning_rate": 3.4284803505445363e-06,
+      "loss": 1.0497,
+      "step": 9070
+    },
+    {
+      "epoch": 1.7975063091876766,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4274799881952564e-06,
+      "loss": 0.9632,
+      "step": 9071
+    },
+    {
+      "epoch": 1.7977062042427725,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.4264796956971486e-06,
+      "loss": 0.979,
+      "step": 9072
+    },
+    {
+      "epoch": 1.7979060992978686,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.425479473094644e-06,
+      "loss": 1.0513,
+      "step": 9073
+    },
+    {
+      "epoch": 1.7981059943529647,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.4244793204321727e-06,
+      "loss": 0.9315,
+      "step": 9074
+    },
+    {
+      "epoch": 1.7983058894080608,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.423479237754162e-06,
+      "loss": 0.9543,
+      "step": 9075
+    },
+    {
+      "epoch": 1.798505784463157,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.422479225105035e-06,
+      "loss": 1.0306,
+      "step": 9076
+    },
+    {
+      "epoch": 1.7987056795182528,
+      "grad_norm": 2.125,
+      "learning_rate": 3.4214792825292133e-06,
+      "loss": 0.9779,
+      "step": 9077
+    },
+    {
+      "epoch": 1.7989055745733489,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.420479410071112e-06,
+      "loss": 1.0356,
+      "step": 9078
+    },
+    {
+      "epoch": 1.799105469628445,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.419479607775147e-06,
+      "loss": 0.9675,
+      "step": 9079
+    },
+    {
+      "epoch": 1.799305364683541,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.4184798756857294e-06,
+      "loss": 0.9125,
+      "step": 9080
+    },
+    {
+      "epoch": 1.7995052597386372,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.4174802138472663e-06,
+      "loss": 1.0142,
+      "step": 9081
+    },
+    {
+      "epoch": 1.7997051547937333,
+      "grad_norm": 2.125,
+      "learning_rate": 3.416480622304163e-06,
+      "loss": 0.9753,
+      "step": 9082
+    },
+    {
+      "epoch": 1.7999050498488294,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.4154811011008224e-06,
+      "loss": 1.0182,
+      "step": 9083
+    },
+    {
+      "epoch": 1.8001049449039255,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.414481650281643e-06,
+      "loss": 0.9982,
+      "step": 9084
+    },
+    {
+      "epoch": 1.8003048399590216,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.4134822698910176e-06,
+      "loss": 0.9966,
+      "step": 9085
+    },
+    {
+      "epoch": 1.8005047350141177,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4124829599733404e-06,
+      "loss": 0.9943,
+      "step": 9086
+    },
+    {
+      "epoch": 1.8007046300692138,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.411483720573e-06,
+      "loss": 0.9798,
+      "step": 9087
+    },
+    {
+      "epoch": 1.8009045251243099,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.4104845517343837e-06,
+      "loss": 0.8583,
+      "step": 9088
+    },
+    {
+      "epoch": 1.801104420179406,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.4094854535018744e-06,
+      "loss": 1.0153,
+      "step": 9089
+    },
+    {
+      "epoch": 1.8013043152345019,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.40848642591985e-06,
+      "loss": 0.9333,
+      "step": 9090
+    },
+    {
+      "epoch": 1.801504210289598,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4074874690326897e-06,
+      "loss": 1.0667,
+      "step": 9091
+    },
+    {
+      "epoch": 1.801704105344694,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.406488582884766e-06,
+      "loss": 0.87,
+      "step": 9092
+    },
+    {
+      "epoch": 1.8019040003997902,
+      "grad_norm": 2.25,
+      "learning_rate": 3.4054897675204485e-06,
+      "loss": 1.0195,
+      "step": 9093
+    },
+    {
+      "epoch": 1.802103895454886,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.404491022984105e-06,
+      "loss": 0.9611,
+      "step": 9094
+    },
+    {
+      "epoch": 1.8023037905099821,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.403492349320101e-06,
+      "loss": 1.0533,
+      "step": 9095
+    },
+    {
+      "epoch": 1.8025036855650782,
+      "grad_norm": 2.25,
+      "learning_rate": 3.402493746572796e-06,
+      "loss": 1.0336,
+      "step": 9096
+    },
+    {
+      "epoch": 1.8027035806201743,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.401495214786548e-06,
+      "loss": 0.9208,
+      "step": 9097
+    },
+    {
+      "epoch": 1.8029034756752704,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.4004967540057117e-06,
+      "loss": 0.965,
+      "step": 9098
+    },
+    {
+      "epoch": 1.8031033707303665,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.3994983642746384e-06,
+      "loss": 0.9449,
+      "step": 9099
+    },
+    {
+      "epoch": 1.8033032657854626,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.3985000456376775e-06,
+      "loss": 0.9817,
+      "step": 9100
+    },
+    {
+      "epoch": 1.8035031608405587,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.3975017981391735e-06,
+      "loss": 0.9605,
+      "step": 9101
+    },
+    {
+      "epoch": 1.8037030558956548,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.396503621823467e-06,
+      "loss": 0.9228,
+      "step": 9102
+    },
+    {
+      "epoch": 1.803902950950751,
+      "grad_norm": 2.25,
+      "learning_rate": 3.3955055167349006e-06,
+      "loss": 1.0007,
+      "step": 9103
+    },
+    {
+      "epoch": 1.804102846005847,
+      "grad_norm": 2.25,
+      "learning_rate": 3.394507482917806e-06,
+      "loss": 0.9854,
+      "step": 9104
+    },
+    {
+      "epoch": 1.8043027410609431,
+      "grad_norm": 2.125,
+      "learning_rate": 3.393509520416519e-06,
+      "loss": 1.02,
+      "step": 9105
+    },
+    {
+      "epoch": 1.8045026361160392,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.392511629275367e-06,
+      "loss": 0.8621,
+      "step": 9106
+    },
+    {
+      "epoch": 1.804702531171135,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3915138095386758e-06,
+      "loss": 0.9195,
+      "step": 9107
+    },
+    {
+      "epoch": 1.8049024262262312,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.390516061250771e-06,
+      "loss": 0.9501,
+      "step": 9108
+    },
+    {
+      "epoch": 1.8051023212813273,
+      "grad_norm": 2.25,
+      "learning_rate": 3.38951838445597e-06,
+      "loss": 0.9412,
+      "step": 9109
+    },
+    {
+      "epoch": 1.8053022163364234,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.3885207791985898e-06,
+      "loss": 1.0443,
+      "step": 9110
+    },
+    {
+      "epoch": 1.8055021113915195,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.3875232455229455e-06,
+      "loss": 0.9549,
+      "step": 9111
+    },
+    {
+      "epoch": 1.8057020064466154,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.3865257834733454e-06,
+      "loss": 0.9883,
+      "step": 9112
+    },
+    {
+      "epoch": 1.8059019015017115,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.385528393094098e-06,
+      "loss": 0.9083,
+      "step": 9113
+    },
+    {
+      "epoch": 1.8061017965568076,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3845310744295078e-06,
+      "loss": 0.9786,
+      "step": 9114
+    },
+    {
+      "epoch": 1.8063016916119037,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.383533827523876e-06,
+      "loss": 0.9428,
+      "step": 9115
+    },
+    {
+      "epoch": 1.8065015866669998,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3825366524214965e-06,
+      "loss": 0.9264,
+      "step": 9116
+    },
+    {
+      "epoch": 1.8067014817220959,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3815395491666676e-06,
+      "loss": 0.9085,
+      "step": 9117
+    },
+    {
+      "epoch": 1.806901376777192,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3805425178036776e-06,
+      "loss": 1.0277,
+      "step": 9118
+    },
+    {
+      "epoch": 1.807101271832288,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.379545558376816e-06,
+      "loss": 0.9798,
+      "step": 9119
+    },
+    {
+      "epoch": 1.8073011668873842,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.378548670930369e-06,
+      "loss": 0.9551,
+      "step": 9120
+    },
+    {
+      "epoch": 1.8075010619424803,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.3775518555086158e-06,
+      "loss": 1.0912,
+      "step": 9121
+    },
+    {
+      "epoch": 1.8077009569975764,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.376555112155836e-06,
+      "loss": 0.9048,
+      "step": 9122
+    },
+    {
+      "epoch": 1.8079008520526725,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3755584409163058e-06,
+      "loss": 0.937,
+      "step": 9123
+    },
+    {
+      "epoch": 1.8081007471077686,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.3745618418342942e-06,
+      "loss": 0.9743,
+      "step": 9124
+    },
+    {
+      "epoch": 1.8083006421628645,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.3735653149540737e-06,
+      "loss": 0.9448,
+      "step": 9125
+    },
+    {
+      "epoch": 1.8085005372179606,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.372568860319907e-06,
+      "loss": 1.0084,
+      "step": 9126
+    },
+    {
+      "epoch": 1.8087004322730567,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3715724779760586e-06,
+      "loss": 0.9859,
+      "step": 9127
+    },
+    {
+      "epoch": 1.8089003273281528,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3705761679667865e-06,
+      "loss": 0.9299,
+      "step": 9128
+    },
+    {
+      "epoch": 1.8091002223832486,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.3695799303363463e-06,
+      "loss": 0.9784,
+      "step": 9129
+    },
+    {
+      "epoch": 1.8093001174383447,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.3685837651289922e-06,
+      "loss": 1.0141,
+      "step": 9130
+    },
+    {
+      "epoch": 1.8095000124934408,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.3675876723889735e-06,
+      "loss": 1.0483,
+      "step": 9131
+    },
+    {
+      "epoch": 1.809699907548537,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3665916521605346e-06,
+      "loss": 0.8994,
+      "step": 9132
+    },
+    {
+      "epoch": 1.809899802603633,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.3655957044879207e-06,
+      "loss": 0.9049,
+      "step": 9133
+    },
+    {
+      "epoch": 1.8100996976587291,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.364599829415372e-06,
+      "loss": 0.9387,
+      "step": 9134
+    },
+    {
+      "epoch": 1.8102995927138252,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3636040269871227e-06,
+      "loss": 0.9692,
+      "step": 9135
+    },
+    {
+      "epoch": 1.8104994877689213,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3626082972474096e-06,
+      "loss": 0.9207,
+      "step": 9136
+    },
+    {
+      "epoch": 1.8106993828240174,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3616126402404594e-06,
+      "loss": 0.9487,
+      "step": 9137
+    },
+    {
+      "epoch": 1.8108992778791135,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.360617056010501e-06,
+      "loss": 0.9734,
+      "step": 9138
+    },
+    {
+      "epoch": 1.8110991729342096,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3596215446017587e-06,
+      "loss": 1.07,
+      "step": 9139
+    },
+    {
+      "epoch": 1.8112990679893057,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.358626106058451e-06,
+      "loss": 1.0255,
+      "step": 9140
+    },
+    {
+      "epoch": 1.8114989630444018,
+      "grad_norm": 2.125,
+      "learning_rate": 3.357630740424797e-06,
+      "loss": 0.9221,
+      "step": 9141
+    },
+    {
+      "epoch": 1.8116988580994977,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.356635447745011e-06,
+      "loss": 0.9733,
+      "step": 9142
+    },
+    {
+      "epoch": 1.8118987531545938,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.3556402280633017e-06,
+      "loss": 1.0517,
+      "step": 9143
+    },
+    {
+      "epoch": 1.81209864820969,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3546450814238786e-06,
+      "loss": 1.0527,
+      "step": 9144
+    },
+    {
+      "epoch": 1.812298543264786,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3536500078709445e-06,
+      "loss": 0.9925,
+      "step": 9145
+    },
+    {
+      "epoch": 1.812498438319882,
+      "grad_norm": 2.125,
+      "learning_rate": 3.352655007448703e-06,
+      "loss": 0.9813,
+      "step": 9146
+    },
+    {
+      "epoch": 1.812698333374978,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3516600802013487e-06,
+      "loss": 1.0226,
+      "step": 9147
+    },
+    {
+      "epoch": 1.812898228430074,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.350665226173078e-06,
+      "loss": 0.9334,
+      "step": 9148
+    },
+    {
+      "epoch": 1.8130981234851702,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3496704454080807e-06,
+      "loss": 0.9609,
+      "step": 9149
+    },
+    {
+      "epoch": 1.8132980185402663,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.3486757379505465e-06,
+      "loss": 0.8925,
+      "step": 9150
+    },
+    {
+      "epoch": 1.8134979135953624,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3476811038446603e-06,
+      "loss": 1.0665,
+      "step": 9151
+    },
+    {
+      "epoch": 1.8136978086504585,
+      "grad_norm": 2.375,
+      "learning_rate": 3.3466865431346017e-06,
+      "loss": 0.9094,
+      "step": 9152
+    },
+    {
+      "epoch": 1.8138977037055546,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.345692055864551e-06,
+      "loss": 0.9556,
+      "step": 9153
+    },
+    {
+      "epoch": 1.8140975987606507,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.3446976420786835e-06,
+      "loss": 0.9073,
+      "step": 9154
+    },
+    {
+      "epoch": 1.8142974938157468,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.3437033018211682e-06,
+      "loss": 0.9369,
+      "step": 9155
+    },
+    {
+      "epoch": 1.8144973888708429,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3427090351361767e-06,
+      "loss": 0.9804,
+      "step": 9156
+    },
+    {
+      "epoch": 1.814697283925939,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.3417148420678723e-06,
+      "loss": 1.0827,
+      "step": 9157
+    },
+    {
+      "epoch": 1.814897178981035,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.3407207226604164e-06,
+      "loss": 0.8971,
+      "step": 9158
+    },
+    {
+      "epoch": 1.8150970740361312,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.339726676957971e-06,
+      "loss": 0.9962,
+      "step": 9159
+    },
+    {
+      "epoch": 1.815296969091227,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.338732705004688e-06,
+      "loss": 0.9177,
+      "step": 9160
+    },
+    {
+      "epoch": 1.8154968641463232,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.3377388068447203e-06,
+      "loss": 1.0041,
+      "step": 9161
+    },
+    {
+      "epoch": 1.8156967592014193,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3367449825222188e-06,
+      "loss": 1.0346,
+      "step": 9162
+    },
+    {
+      "epoch": 1.8158966542565154,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.3357512320813258e-06,
+      "loss": 0.9842,
+      "step": 9163
+    },
+    {
+      "epoch": 1.8160965493116112,
+      "grad_norm": 2.125,
+      "learning_rate": 3.334757555566186e-06,
+      "loss": 0.9713,
+      "step": 9164
+    },
+    {
+      "epoch": 1.8162964443667073,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.333763953020939e-06,
+      "loss": 0.9551,
+      "step": 9165
+    },
+    {
+      "epoch": 1.8164963394218034,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.3327704244897176e-06,
+      "loss": 0.9508,
+      "step": 9166
+    },
+    {
+      "epoch": 1.8166962344768995,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.331776970016657e-06,
+      "loss": 1.0225,
+      "step": 9167
+    },
+    {
+      "epoch": 1.8168961295319956,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.330783589645884e-06,
+      "loss": 0.9797,
+      "step": 9168
+    },
+    {
+      "epoch": 1.8170960245870917,
+      "grad_norm": 2.125,
+      "learning_rate": 3.329790283421526e-06,
+      "loss": 0.9614,
+      "step": 9169
+    },
+    {
+      "epoch": 1.8172959196421878,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3287970513877067e-06,
+      "loss": 1.0246,
+      "step": 9170
+    },
+    {
+      "epoch": 1.817495814697284,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.3278038935885415e-06,
+      "loss": 0.9702,
+      "step": 9171
+    },
+    {
+      "epoch": 1.81769570975238,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3268108100681494e-06,
+      "loss": 0.9394,
+      "step": 9172
+    },
+    {
+      "epoch": 1.8178956048074761,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.325817800870644e-06,
+      "loss": 0.906,
+      "step": 9173
+    },
+    {
+      "epoch": 1.8180954998625722,
+      "grad_norm": 2.25,
+      "learning_rate": 3.3248248660401317e-06,
+      "loss": 1.0423,
+      "step": 9174
+    },
+    {
+      "epoch": 1.8182953949176683,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.3238320056207208e-06,
+      "loss": 0.8253,
+      "step": 9175
+    },
+    {
+      "epoch": 1.8184952899727644,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.322839219656513e-06,
+      "loss": 1.006,
+      "step": 9176
+    },
+    {
+      "epoch": 1.8186951850278605,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.321846508191609e-06,
+      "loss": 1.0534,
+      "step": 9177
+    },
+    {
+      "epoch": 1.8188950800829564,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.320853871270102e-06,
+      "loss": 0.9099,
+      "step": 9178
+    },
+    {
+      "epoch": 1.8190949751380525,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.3198613089360875e-06,
+      "loss": 0.9262,
+      "step": 9179
+    },
+    {
+      "epoch": 1.8192948701931486,
+      "grad_norm": 2.25,
+      "learning_rate": 3.318868821233654e-06,
+      "loss": 0.9757,
+      "step": 9180
+    },
+    {
+      "epoch": 1.8194947652482447,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.317876408206887e-06,
+      "loss": 1.0132,
+      "step": 9181
+    },
+    {
+      "epoch": 1.8196946603033406,
+      "grad_norm": 2.125,
+      "learning_rate": 3.3168840698998722e-06,
+      "loss": 0.905,
+      "step": 9182
+    },
+    {
+      "epoch": 1.8198945553584367,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.315891806356686e-06,
+      "loss": 1.0229,
+      "step": 9183
+    },
+    {
+      "epoch": 1.8200944504135328,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.3148996176214054e-06,
+      "loss": 0.9095,
+      "step": 9184
+    },
+    {
+      "epoch": 1.820294345468629,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.3139075037381053e-06,
+      "loss": 0.9197,
+      "step": 9185
+    },
+    {
+      "epoch": 1.820494240523725,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.312915464750852e-06,
+      "loss": 0.9501,
+      "step": 9186
+    },
+    {
+      "epoch": 1.820694135578821,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.3119235007037155e-06,
+      "loss": 0.959,
+      "step": 9187
+    },
+    {
+      "epoch": 1.8208940306339172,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.3109316116407554e-06,
+      "loss": 0.963,
+      "step": 9188
+    },
+    {
+      "epoch": 1.8210939256890133,
+      "grad_norm": 2.125,
+      "learning_rate": 3.309939797606033e-06,
+      "loss": 0.9995,
+      "step": 9189
+    },
+    {
+      "epoch": 1.8212938207441094,
+      "grad_norm": 2.125,
+      "learning_rate": 3.308948058643605e-06,
+      "loss": 0.9072,
+      "step": 9190
+    },
+    {
+      "epoch": 1.8214937157992055,
+      "grad_norm": 2.25,
+      "learning_rate": 3.3079563947975225e-06,
+      "loss": 0.98,
+      "step": 9191
+    },
+    {
+      "epoch": 1.8216936108543016,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3069648061118366e-06,
+      "loss": 1.0131,
+      "step": 9192
+    },
+    {
+      "epoch": 1.8218935059093977,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.3059732926305943e-06,
+      "loss": 0.889,
+      "step": 9193
+    },
+    {
+      "epoch": 1.8220934009644938,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.3049818543978363e-06,
+      "loss": 0.9544,
+      "step": 9194
+    },
+    {
+      "epoch": 1.8222932960195897,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.3039904914576036e-06,
+      "loss": 0.8702,
+      "step": 9195
+    },
+    {
+      "epoch": 1.8224931910746858,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.3029992038539318e-06,
+      "loss": 0.9753,
+      "step": 9196
+    },
+    {
+      "epoch": 1.8226930861297819,
+      "grad_norm": 2.125,
+      "learning_rate": 3.302007991630854e-06,
+      "loss": 0.9726,
+      "step": 9197
+    },
+    {
+      "epoch": 1.822892981184878,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.3010168548324006e-06,
+      "loss": 1.0036,
+      "step": 9198
+    },
+    {
+      "epoch": 1.8230928762399738,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3000257935025963e-06,
+      "loss": 0.9817,
+      "step": 9199
+    },
+    {
+      "epoch": 1.82329277129507,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.299034807685465e-06,
+      "loss": 0.9199,
+      "step": 9200
+    },
+    {
+      "epoch": 1.823492666350166,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.2980438974250262e-06,
+      "loss": 1.0242,
+      "step": 9201
+    },
+    {
+      "epoch": 1.8236925614052621,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.297053062765295e-06,
+      "loss": 0.9565,
+      "step": 9202
+    },
+    {
+      "epoch": 1.8238924564603582,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.2960623037502847e-06,
+      "loss": 0.9995,
+      "step": 9203
+    },
+    {
+      "epoch": 1.8240923515154543,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.2950716204240065e-06,
+      "loss": 1.005,
+      "step": 9204
+    },
+    {
+      "epoch": 1.8242922465705504,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.2940810128304634e-06,
+      "loss": 0.9792,
+      "step": 9205
+    },
+    {
+      "epoch": 1.8244921416256465,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.293090481013661e-06,
+      "loss": 0.9746,
+      "step": 9206
+    },
+    {
+      "epoch": 1.8246920366807426,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.2921000250175948e-06,
+      "loss": 1.0629,
+      "step": 9207
+    },
+    {
+      "epoch": 1.8248919317358387,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.2911096448862666e-06,
+      "loss": 0.9234,
+      "step": 9208
+    },
+    {
+      "epoch": 1.8250918267909348,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.290119340663663e-06,
+      "loss": 1.0217,
+      "step": 9209
+    },
+    {
+      "epoch": 1.825291721846031,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2891291123937764e-06,
+      "loss": 0.9754,
+      "step": 9210
+    },
+    {
+      "epoch": 1.825491616901127,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.2881389601205906e-06,
+      "loss": 1.0121,
+      "step": 9211
+    },
+    {
+      "epoch": 1.8256915119562231,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.28714888388809e-06,
+      "loss": 0.9866,
+      "step": 9212
+    },
+    {
+      "epoch": 1.825891407011319,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.2861588837402534e-06,
+      "loss": 0.97,
+      "step": 9213
+    },
+    {
+      "epoch": 1.8260913020664151,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.2851689597210555e-06,
+      "loss": 0.9061,
+      "step": 9214
+    },
+    {
+      "epoch": 1.8262911971215112,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2841791118744704e-06,
+      "loss": 1.002,
+      "step": 9215
+    },
+    {
+      "epoch": 1.8264910921766073,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2831893402444648e-06,
+      "loss": 0.9203,
+      "step": 9216
+    },
+    {
+      "epoch": 1.8266909872317032,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.2821996448750054e-06,
+      "loss": 1.0905,
+      "step": 9217
+    },
+    {
+      "epoch": 1.8268908822867993,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2812100258100556e-06,
+      "loss": 1.0289,
+      "step": 9218
+    },
+    {
+      "epoch": 1.8270907773418954,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.280220483093571e-06,
+      "loss": 0.9328,
+      "step": 9219
+    },
+    {
+      "epoch": 1.8272906723969915,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2792310167695097e-06,
+      "loss": 0.9384,
+      "step": 9220
+    },
+    {
+      "epoch": 1.8274905674520876,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.278241626881823e-06,
+      "loss": 0.9116,
+      "step": 9221
+    },
+    {
+      "epoch": 1.8276904625071837,
+      "grad_norm": 2.25,
+      "learning_rate": 3.2772523134744592e-06,
+      "loss": 0.918,
+      "step": 9222
+    },
+    {
+      "epoch": 1.8278903575622798,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2762630765913626e-06,
+      "loss": 0.9216,
+      "step": 9223
+    },
+    {
+      "epoch": 1.828090252617376,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.275273916276478e-06,
+      "loss": 1.0595,
+      "step": 9224
+    },
+    {
+      "epoch": 1.828290147672472,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.274284832573741e-06,
+      "loss": 1.0279,
+      "step": 9225
+    },
+    {
+      "epoch": 1.828490042727568,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.273295825527088e-06,
+      "loss": 0.9786,
+      "step": 9226
+    },
+    {
+      "epoch": 1.8286899377826642,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2723068951804486e-06,
+      "loss": 1.0637,
+      "step": 9227
+    },
+    {
+      "epoch": 1.8288898328377603,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2713180415777536e-06,
+      "loss": 1.0626,
+      "step": 9228
+    },
+    {
+      "epoch": 1.8290897278928564,
+      "grad_norm": 2.125,
+      "learning_rate": 3.270329264762926e-06,
+      "loss": 0.9125,
+      "step": 9229
+    },
+    {
+      "epoch": 1.8292896229479523,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2693405647798878e-06,
+      "loss": 1.0265,
+      "step": 9230
+    },
+    {
+      "epoch": 1.8294895180030484,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.2683519416725564e-06,
+      "loss": 0.9802,
+      "step": 9231
+    },
+    {
+      "epoch": 1.8296894130581445,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.267363395484848e-06,
+      "loss": 0.9942,
+      "step": 9232
+    },
+    {
+      "epoch": 1.8298893081132406,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.2663749262606715e-06,
+      "loss": 0.965,
+      "step": 9233
+    },
+    {
+      "epoch": 1.8300892031683367,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.2653865340439366e-06,
+      "loss": 0.9769,
+      "step": 9234
+    },
+    {
+      "epoch": 1.8302890982234326,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.2643982188785457e-06,
+      "loss": 0.9969,
+      "step": 9235
+    },
+    {
+      "epoch": 1.8304889932785287,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.2634099808084004e-06,
+      "loss": 1.0316,
+      "step": 9236
+    },
+    {
+      "epoch": 1.8306888883336248,
+      "grad_norm": 2.125,
+      "learning_rate": 3.2624218198773994e-06,
+      "loss": 0.9977,
+      "step": 9237
+    },
+    {
+      "epoch": 1.8308887833887209,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.2614337361294345e-06,
+      "loss": 0.9797,
+      "step": 9238
+    },
+    {
+      "epoch": 1.831088678443817,
+      "grad_norm": 2.125,
+      "learning_rate": 3.260445729608399e-06,
+      "loss": 0.95,
+      "step": 9239
+    },
+    {
+      "epoch": 1.831288573498913,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.259457800358177e-06,
+      "loss": 1.0814,
+      "step": 9240
+    },
+    {
+      "epoch": 1.8314884685540092,
+      "grad_norm": 2.25,
+      "learning_rate": 3.2584699484226547e-06,
+      "loss": 0.9625,
+      "step": 9241
+    },
+    {
+      "epoch": 1.8316883636091053,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2574821738457096e-06,
+      "loss": 0.94,
+      "step": 9242
+    },
+    {
+      "epoch": 1.8318882586642014,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.25649447667122e-06,
+      "loss": 0.911,
+      "step": 9243
+    },
+    {
+      "epoch": 1.8320881537192975,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2555068569430614e-06,
+      "loss": 0.9125,
+      "step": 9244
+    },
+    {
+      "epoch": 1.8322880487743936,
+      "grad_norm": 2.375,
+      "learning_rate": 3.2545193147051015e-06,
+      "loss": 1.056,
+      "step": 9245
+    },
+    {
+      "epoch": 1.8324879438294897,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.253531850001207e-06,
+      "loss": 1.0154,
+      "step": 9246
+    },
+    {
+      "epoch": 1.8326878388845858,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.25254446287524e-06,
+      "loss": 1.0136,
+      "step": 9247
+    },
+    {
+      "epoch": 1.8328877339396816,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.251557153371062e-06,
+      "loss": 0.9848,
+      "step": 9248
+    },
+    {
+      "epoch": 1.8330876289947777,
+      "grad_norm": 2.125,
+      "learning_rate": 3.250569921532529e-06,
+      "loss": 0.9669,
+      "step": 9249
+    },
+    {
+      "epoch": 1.8332875240498738,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.249582767403493e-06,
+      "loss": 1.0233,
+      "step": 9250
+    },
+    {
+      "epoch": 1.83348741910497,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.2485956910278033e-06,
+      "loss": 1.081,
+      "step": 9251
+    },
+    {
+      "epoch": 1.8336873141600658,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.2476086924493067e-06,
+      "loss": 1.0206,
+      "step": 9252
+    },
+    {
+      "epoch": 1.833887209215162,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.2466217717118442e-06,
+      "loss": 0.9491,
+      "step": 9253
+    },
+    {
+      "epoch": 1.834087104270258,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2456349288592547e-06,
+      "loss": 0.947,
+      "step": 9254
+    },
+    {
+      "epoch": 1.834286999325354,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.2446481639353757e-06,
+      "loss": 0.9895,
+      "step": 9255
+    },
+    {
+      "epoch": 1.8344868943804502,
+      "grad_norm": 2.125,
+      "learning_rate": 3.2436614769840367e-06,
+      "loss": 0.9144,
+      "step": 9256
+    },
+    {
+      "epoch": 1.8346867894355463,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2426748680490684e-06,
+      "loss": 1.0338,
+      "step": 9257
+    },
+    {
+      "epoch": 1.8348866844906424,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2416883371742937e-06,
+      "loss": 0.9259,
+      "step": 9258
+    },
+    {
+      "epoch": 1.8350865795457385,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.2407018844035348e-06,
+      "loss": 0.8412,
+      "step": 9259
+    },
+    {
+      "epoch": 1.8352864746008346,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.239715509780612e-06,
+      "loss": 0.964,
+      "step": 9260
+    },
+    {
+      "epoch": 1.8354863696559307,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.238729213349337e-06,
+      "loss": 1.0716,
+      "step": 9261
+    },
+    {
+      "epoch": 1.8356862647110268,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2377429951535223e-06,
+      "loss": 1.0144,
+      "step": 9262
+    },
+    {
+      "epoch": 1.835886159766123,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.2367568552369765e-06,
+      "loss": 0.9675,
+      "step": 9263
+    },
+    {
+      "epoch": 1.836086054821219,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.2357707936435013e-06,
+      "loss": 0.9698,
+      "step": 9264
+    },
+    {
+      "epoch": 1.8362859498763149,
+      "grad_norm": 2.125,
+      "learning_rate": 3.2347848104169012e-06,
+      "loss": 0.9298,
+      "step": 9265
+    },
+    {
+      "epoch": 1.836485844931411,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.233798905600969e-06,
+      "loss": 0.8792,
+      "step": 9266
+    },
+    {
+      "epoch": 1.836685739986507,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.232813079239502e-06,
+      "loss": 0.9292,
+      "step": 9267
+    },
+    {
+      "epoch": 1.8368856350416032,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.231827331376289e-06,
+      "loss": 0.8959,
+      "step": 9268
+    },
+    {
+      "epoch": 1.8370855300966993,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2308416620551175e-06,
+      "loss": 1.0066,
+      "step": 9269
+    },
+    {
+      "epoch": 1.8372854251517952,
+      "grad_norm": 2.125,
+      "learning_rate": 3.2298560713197712e-06,
+      "loss": 1.0907,
+      "step": 9270
+    },
+    {
+      "epoch": 1.8374853202068913,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.228870559214028e-06,
+      "loss": 1.0026,
+      "step": 9271
+    },
+    {
+      "epoch": 1.8376852152619874,
+      "grad_norm": 2.125,
+      "learning_rate": 3.2278851257816655e-06,
+      "loss": 0.9451,
+      "step": 9272
+    },
+    {
+      "epoch": 1.8378851103170835,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.226899771066456e-06,
+      "loss": 0.9575,
+      "step": 9273
+    },
+    {
+      "epoch": 1.8380850053721796,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.2259144951121697e-06,
+      "loss": 0.9289,
+      "step": 9274
+    },
+    {
+      "epoch": 1.8382849004272757,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.224929297962572e-06,
+      "loss": 0.9889,
+      "step": 9275
+    },
+    {
+      "epoch": 1.8384847954823718,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2239441796614256e-06,
+      "loss": 1.069,
+      "step": 9276
+    },
+    {
+      "epoch": 1.8386846905374679,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.2229591402524894e-06,
+      "loss": 1.095,
+      "step": 9277
+    },
+    {
+      "epoch": 1.838884585592564,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2219741797795175e-06,
+      "loss": 1.061,
+      "step": 9278
+    },
+    {
+      "epoch": 1.83908448064766,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.220989298286262e-06,
+      "loss": 0.9144,
+      "step": 9279
+    },
+    {
+      "epoch": 1.8392843757027562,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.220004495816474e-06,
+      "loss": 0.9368,
+      "step": 9280
+    },
+    {
+      "epoch": 1.8394842707578523,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2190197724138943e-06,
+      "loss": 0.9001,
+      "step": 9281
+    },
+    {
+      "epoch": 1.8396841658129484,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.2180351281222665e-06,
+      "loss": 1.0191,
+      "step": 9282
+    },
+    {
+      "epoch": 1.8398840608680442,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.217050562985329e-06,
+      "loss": 0.9633,
+      "step": 9283
+    },
+    {
+      "epoch": 1.8400839559231403,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.216066077046814e-06,
+      "loss": 1.0508,
+      "step": 9284
+    },
+    {
+      "epoch": 1.8402838509782364,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.215081670350454e-06,
+      "loss": 1.0296,
+      "step": 9285
+    },
+    {
+      "epoch": 1.8404837460333325,
+      "grad_norm": 2.25,
+      "learning_rate": 3.2140973429399747e-06,
+      "loss": 1.0299,
+      "step": 9286
+    },
+    {
+      "epoch": 1.8406836410884284,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.213113094859101e-06,
+      "loss": 0.9849,
+      "step": 9287
+    },
+    {
+      "epoch": 1.8408835361435245,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.2121289261515535e-06,
+      "loss": 0.9708,
+      "step": 9288
+    },
+    {
+      "epoch": 1.8410834311986206,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.2111448368610476e-06,
+      "loss": 0.9623,
+      "step": 9289
+    },
+    {
+      "epoch": 1.8412833262537167,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.2101608270312963e-06,
+      "loss": 1.0313,
+      "step": 9290
+    },
+    {
+      "epoch": 1.8414832213088128,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.209176896706011e-06,
+      "loss": 0.9675,
+      "step": 9291
+    },
+    {
+      "epoch": 1.841683116363909,
+      "grad_norm": 2.125,
+      "learning_rate": 3.2081930459288963e-06,
+      "loss": 1.0305,
+      "step": 9292
+    },
+    {
+      "epoch": 1.841883011419005,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.2072092747436546e-06,
+      "loss": 0.9266,
+      "step": 9293
+    },
+    {
+      "epoch": 1.8420829064741011,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2062255831939863e-06,
+      "loss": 0.9272,
+      "step": 9294
+    },
+    {
+      "epoch": 1.8422828015291972,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.2052419713235854e-06,
+      "loss": 0.9379,
+      "step": 9295
+    },
+    {
+      "epoch": 1.8424826965842933,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.2042584391761457e-06,
+      "loss": 1.0046,
+      "step": 9296
+    },
+    {
+      "epoch": 1.8426825916393894,
+      "grad_norm": 2.25,
+      "learning_rate": 3.203274986795353e-06,
+      "loss": 0.9611,
+      "step": 9297
+    },
+    {
+      "epoch": 1.8428824866944855,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.2022916142248933e-06,
+      "loss": 1.0321,
+      "step": 9298
+    },
+    {
+      "epoch": 1.8430823817495816,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.2013083215084495e-06,
+      "loss": 0.9563,
+      "step": 9299
+    },
+    {
+      "epoch": 1.8432822768046777,
+      "grad_norm": 2.25,
+      "learning_rate": 3.2003251086896963e-06,
+      "loss": 1.0356,
+      "step": 9300
+    },
+    {
+      "epoch": 1.8434821718597736,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.199341975812312e-06,
+      "loss": 1.0392,
+      "step": 9301
+    },
+    {
+      "epoch": 1.8436820669148697,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.198358922919963e-06,
+      "loss": 1.0086,
+      "step": 9302
+    },
+    {
+      "epoch": 1.8438819619699658,
+      "grad_norm": 2.125,
+      "learning_rate": 3.197375950056319e-06,
+      "loss": 0.9148,
+      "step": 9303
+    },
+    {
+      "epoch": 1.8440818570250619,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1963930572650414e-06,
+      "loss": 0.8526,
+      "step": 9304
+    },
+    {
+      "epoch": 1.8442817520801578,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.1954102445897923e-06,
+      "loss": 0.9934,
+      "step": 9305
+    },
+    {
+      "epoch": 1.8444816471352539,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.194427512074228e-06,
+      "loss": 0.9847,
+      "step": 9306
+    },
+    {
+      "epoch": 1.84468154219035,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1934448597619997e-06,
+      "loss": 0.9596,
+      "step": 9307
+    },
+    {
+      "epoch": 1.844881437245446,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.192462287696759e-06,
+      "loss": 1.0517,
+      "step": 9308
+    },
+    {
+      "epoch": 1.8450813323005422,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.1914797959221495e-06,
+      "loss": 0.9667,
+      "step": 9309
+    },
+    {
+      "epoch": 1.8452812273556383,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.1904973844818144e-06,
+      "loss": 0.9874,
+      "step": 9310
+    },
+    {
+      "epoch": 1.8454811224107344,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.189515053419393e-06,
+      "loss": 0.9769,
+      "step": 9311
+    },
+    {
+      "epoch": 1.8456810174658305,
+      "grad_norm": 2.25,
+      "learning_rate": 3.188532802778518e-06,
+      "loss": 1.0035,
+      "step": 9312
+    },
+    {
+      "epoch": 1.8458809125209266,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.1875506326028226e-06,
+      "loss": 0.9881,
+      "step": 9313
+    },
+    {
+      "epoch": 1.8460808075760227,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.186568542935936e-06,
+      "loss": 1.0568,
+      "step": 9314
+    },
+    {
+      "epoch": 1.8462807026311188,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.1855865338214803e-06,
+      "loss": 0.914,
+      "step": 9315
+    },
+    {
+      "epoch": 1.8464805976862149,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1846046053030778e-06,
+      "loss": 0.9923,
+      "step": 9316
+    },
+    {
+      "epoch": 1.846680492741311,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.1836227574243434e-06,
+      "loss": 0.9697,
+      "step": 9317
+    },
+    {
+      "epoch": 1.8468803877964068,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.1826409902288925e-06,
+      "loss": 1.0062,
+      "step": 9318
+    },
+    {
+      "epoch": 1.847080282851503,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.1816593037603356e-06,
+      "loss": 0.9864,
+      "step": 9319
+    },
+    {
+      "epoch": 1.847280177906599,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.1806776980622773e-06,
+      "loss": 0.9804,
+      "step": 9320
+    },
+    {
+      "epoch": 1.8474800729616951,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.179696173178321e-06,
+      "loss": 0.985,
+      "step": 9321
+    },
+    {
+      "epoch": 1.847679968016791,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1787147291520675e-06,
+      "loss": 0.9131,
+      "step": 9322
+    },
+    {
+      "epoch": 1.8478798630718871,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1777333660271103e-06,
+      "loss": 0.8775,
+      "step": 9323
+    },
+    {
+      "epoch": 1.8480797581269832,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1767520838470433e-06,
+      "loss": 0.9141,
+      "step": 9324
+    },
+    {
+      "epoch": 1.8482796531820793,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.175770882655453e-06,
+      "loss": 0.9623,
+      "step": 9325
+    },
+    {
+      "epoch": 1.8484795482371754,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.174789762495925e-06,
+      "loss": 0.9542,
+      "step": 9326
+    },
+    {
+      "epoch": 1.8486794432922715,
+      "grad_norm": 2.25,
+      "learning_rate": 3.1738087234120426e-06,
+      "loss": 0.9142,
+      "step": 9327
+    },
+    {
+      "epoch": 1.8488793383473676,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.1728277654473793e-06,
+      "loss": 0.9389,
+      "step": 9328
+    },
+    {
+      "epoch": 1.8490792334024637,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.171846888645513e-06,
+      "loss": 0.9108,
+      "step": 9329
+    },
+    {
+      "epoch": 1.8492791284575598,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1708660930500124e-06,
+      "loss": 0.942,
+      "step": 9330
+    },
+    {
+      "epoch": 1.849479023512656,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.1698853787044447e-06,
+      "loss": 0.9554,
+      "step": 9331
+    },
+    {
+      "epoch": 1.849678918567752,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.1689047456523746e-06,
+      "loss": 0.9705,
+      "step": 9332
+    },
+    {
+      "epoch": 1.8498788136228481,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1679241939373578e-06,
+      "loss": 1.073,
+      "step": 9333
+    },
+    {
+      "epoch": 1.8500787086779442,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1669437236029536e-06,
+      "loss": 0.9328,
+      "step": 9334
+    },
+    {
+      "epoch": 1.8502786037330403,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.1659633346927133e-06,
+      "loss": 0.983,
+      "step": 9335
+    },
+    {
+      "epoch": 1.8504784987881362,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1649830272501857e-06,
+      "loss": 0.9917,
+      "step": 9336
+    },
+    {
+      "epoch": 1.8506783938432323,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.1640028013189155e-06,
+      "loss": 1.005,
+      "step": 9337
+    },
+    {
+      "epoch": 1.8508782888983284,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.163022656942445e-06,
+      "loss": 0.9806,
+      "step": 9338
+    },
+    {
+      "epoch": 1.8510781839534245,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.162042594164313e-06,
+      "loss": 0.9685,
+      "step": 9339
+    },
+    {
+      "epoch": 1.8512780790085204,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.1610626130280507e-06,
+      "loss": 0.9463,
+      "step": 9340
+    },
+    {
+      "epoch": 1.8514779740636165,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1600827135771914e-06,
+      "loss": 0.9251,
+      "step": 9341
+    },
+    {
+      "epoch": 1.8516778691187126,
+      "grad_norm": 2.25,
+      "learning_rate": 3.1591028958552627e-06,
+      "loss": 1.0171,
+      "step": 9342
+    },
+    {
+      "epoch": 1.8518777641738087,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.158123159905785e-06,
+      "loss": 0.8933,
+      "step": 9343
+    },
+    {
+      "epoch": 1.8520776592289048,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.1571435057722806e-06,
+      "loss": 0.8957,
+      "step": 9344
+    },
+    {
+      "epoch": 1.8522775542840009,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1561639334982653e-06,
+      "loss": 0.9443,
+      "step": 9345
+    },
+    {
+      "epoch": 1.852477449339097,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.1551844431272505e-06,
+      "loss": 0.9494,
+      "step": 9346
+    },
+    {
+      "epoch": 1.852677344394193,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.1542050347027465e-06,
+      "loss": 0.922,
+      "step": 9347
+    },
+    {
+      "epoch": 1.8528772394492892,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.1532257082682573e-06,
+      "loss": 0.8955,
+      "step": 9348
+    },
+    {
+      "epoch": 1.8530771345043853,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.152246463867284e-06,
+      "loss": 1.1093,
+      "step": 9349
+    },
+    {
+      "epoch": 1.8532770295594814,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.1512673015433272e-06,
+      "loss": 0.9333,
+      "step": 9350
+    },
+    {
+      "epoch": 1.8534769246145775,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.1502882213398776e-06,
+      "loss": 0.9682,
+      "step": 9351
+    },
+    {
+      "epoch": 1.8536768196696736,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.149309223300428e-06,
+      "loss": 0.9936,
+      "step": 9352
+    },
+    {
+      "epoch": 1.8538767147247694,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1483303074684663e-06,
+      "loss": 0.9197,
+      "step": 9353
+    },
+    {
+      "epoch": 1.8540766097798655,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1473514738874734e-06,
+      "loss": 1.0184,
+      "step": 9354
+    },
+    {
+      "epoch": 1.8542765048349616,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.146372722600931e-06,
+      "loss": 0.9506,
+      "step": 9355
+    },
+    {
+      "epoch": 1.8544763998900577,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.1453940536523135e-06,
+      "loss": 0.9858,
+      "step": 9356
+    },
+    {
+      "epoch": 1.8546762949451538,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.144415467085094e-06,
+      "loss": 0.998,
+      "step": 9357
+    },
+    {
+      "epoch": 1.8548761900002497,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1434369629427425e-06,
+      "loss": 0.898,
+      "step": 9358
+    },
+    {
+      "epoch": 1.8550760850553458,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.1424585412687215e-06,
+      "loss": 1.0378,
+      "step": 9359
+    },
+    {
+      "epoch": 1.855275980110442,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.141480202106494e-06,
+      "loss": 0.9777,
+      "step": 9360
+    },
+    {
+      "epoch": 1.855475875165538,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1405019454995178e-06,
+      "loss": 0.9472,
+      "step": 9361
+    },
+    {
+      "epoch": 1.8556757702206341,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.139523771491246e-06,
+      "loss": 0.9036,
+      "step": 9362
+    },
+    {
+      "epoch": 1.8558756652757302,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.138545680125132e-06,
+      "loss": 1.049,
+      "step": 9363
+    },
+    {
+      "epoch": 1.8560755603308263,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.1375676714446173e-06,
+      "loss": 0.9575,
+      "step": 9364
+    },
+    {
+      "epoch": 1.8562754553859224,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.1365897454931495e-06,
+      "loss": 0.9965,
+      "step": 9365
+    },
+    {
+      "epoch": 1.8564753504410185,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.1356119023141644e-06,
+      "loss": 0.9977,
+      "step": 9366
+    },
+    {
+      "epoch": 1.8566752454961146,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1346341419511008e-06,
+      "loss": 0.9958,
+      "step": 9367
+    },
+    {
+      "epoch": 1.8568751405512107,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.1336564644473886e-06,
+      "loss": 1.0637,
+      "step": 9368
+    },
+    {
+      "epoch": 1.8570750356063068,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1326788698464565e-06,
+      "loss": 0.9981,
+      "step": 9369
+    },
+    {
+      "epoch": 1.857274930661403,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.131701358191731e-06,
+      "loss": 0.9778,
+      "step": 9370
+    },
+    {
+      "epoch": 1.8574748257164988,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1307239295266303e-06,
+      "loss": 0.9657,
+      "step": 9371
+    },
+    {
+      "epoch": 1.857674720771595,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.129746583894573e-06,
+      "loss": 1.0535,
+      "step": 9372
+    },
+    {
+      "epoch": 1.857874615826691,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.128769321338974e-06,
+      "loss": 0.9256,
+      "step": 9373
+    },
+    {
+      "epoch": 1.858074510881787,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.127792141903241e-06,
+      "loss": 0.9787,
+      "step": 9374
+    },
+    {
+      "epoch": 1.858274405936883,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1268150456307817e-06,
+      "loss": 0.9957,
+      "step": 9375
+    },
+    {
+      "epoch": 1.858474300991979,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1258380325649973e-06,
+      "loss": 1.0009,
+      "step": 9376
+    },
+    {
+      "epoch": 1.8586741960470752,
+      "grad_norm": 2.125,
+      "learning_rate": 3.124861102749287e-06,
+      "loss": 0.9454,
+      "step": 9377
+    },
+    {
+      "epoch": 1.8588740911021713,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.123884256227047e-06,
+      "loss": 0.8927,
+      "step": 9378
+    },
+    {
+      "epoch": 1.8590739861572674,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1229074930416674e-06,
+      "loss": 0.963,
+      "step": 9379
+    },
+    {
+      "epoch": 1.8592738812123635,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1219308132365365e-06,
+      "loss": 0.969,
+      "step": 9380
+    },
+    {
+      "epoch": 1.8594737762674596,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.1209542168550393e-06,
+      "loss": 0.9602,
+      "step": 9381
+    },
+    {
+      "epoch": 1.8596736713225557,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.119977703940554e-06,
+      "loss": 1.013,
+      "step": 9382
+    },
+    {
+      "epoch": 1.8598735663776518,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.1190012745364584e-06,
+      "loss": 1.064,
+      "step": 9383
+    },
+    {
+      "epoch": 1.8600734614327479,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.118024928686126e-06,
+      "loss": 0.9555,
+      "step": 9384
+    },
+    {
+      "epoch": 1.860273356487844,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.1170486664329246e-06,
+      "loss": 1.0132,
+      "step": 9385
+    },
+    {
+      "epoch": 1.86047325154294,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1160724878202213e-06,
+      "loss": 1.0332,
+      "step": 9386
+    },
+    {
+      "epoch": 1.8606731465980362,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1150963928913756e-06,
+      "loss": 1.0393,
+      "step": 9387
+    },
+    {
+      "epoch": 1.860873041653132,
+      "grad_norm": 2.125,
+      "learning_rate": 3.114120381689747e-06,
+      "loss": 0.9029,
+      "step": 9388
+    },
+    {
+      "epoch": 1.8610729367082282,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.1131444542586906e-06,
+      "loss": 0.9409,
+      "step": 9389
+    },
+    {
+      "epoch": 1.8612728317633243,
+      "grad_norm": 2.25,
+      "learning_rate": 3.1121686106415553e-06,
+      "loss": 0.9847,
+      "step": 9390
+    },
+    {
+      "epoch": 1.8614727268184204,
+      "grad_norm": 2.125,
+      "learning_rate": 3.1111928508816885e-06,
+      "loss": 0.9834,
+      "step": 9391
+    },
+    {
+      "epoch": 1.8616726218735165,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.110217175022434e-06,
+      "loss": 0.9899,
+      "step": 9392
+    },
+    {
+      "epoch": 1.8618725169286123,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.109241583107131e-06,
+      "loss": 0.9833,
+      "step": 9393
+    },
+    {
+      "epoch": 1.8620724119837084,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.108266075179116e-06,
+      "loss": 0.9947,
+      "step": 9394
+    },
+    {
+      "epoch": 1.8622723070388045,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.107290651281718e-06,
+      "loss": 0.8924,
+      "step": 9395
+    },
+    {
+      "epoch": 1.8624722020939006,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.1063153114582677e-06,
+      "loss": 0.9985,
+      "step": 9396
+    },
+    {
+      "epoch": 1.8626720971489967,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.105340055752089e-06,
+      "loss": 0.9268,
+      "step": 9397
+    },
+    {
+      "epoch": 1.8628719922040928,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.1043648842065032e-06,
+      "loss": 1.05,
+      "step": 9398
+    },
+    {
+      "epoch": 1.863071887259189,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.1033897968648258e-06,
+      "loss": 0.9761,
+      "step": 9399
+    },
+    {
+      "epoch": 1.863271782314285,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.102414793770371e-06,
+      "loss": 0.9489,
+      "step": 9400
+    },
+    {
+      "epoch": 1.8634716773693811,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.1014398749664494e-06,
+      "loss": 0.9223,
+      "step": 9401
+    },
+    {
+      "epoch": 1.8636715724244772,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.1004650404963643e-06,
+      "loss": 0.9064,
+      "step": 9402
+    },
+    {
+      "epoch": 1.8638714674795733,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0994902904034196e-06,
+      "loss": 0.9444,
+      "step": 9403
+    },
+    {
+      "epoch": 1.8640713625346694,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0985156247309145e-06,
+      "loss": 0.8982,
+      "step": 9404
+    },
+    {
+      "epoch": 1.8642712575897655,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0975410435221415e-06,
+      "loss": 0.9376,
+      "step": 9405
+    },
+    {
+      "epoch": 1.8644711526448614,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0965665468203926e-06,
+      "loss": 1.0606,
+      "step": 9406
+    },
+    {
+      "epoch": 1.8646710476999575,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0955921346689533e-06,
+      "loss": 0.9864,
+      "step": 9407
+    },
+    {
+      "epoch": 1.8648709427550536,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.094617807111109e-06,
+      "loss": 0.9607,
+      "step": 9408
+    },
+    {
+      "epoch": 1.8650708378101497,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.093643564190138e-06,
+      "loss": 0.9449,
+      "step": 9409
+    },
+    {
+      "epoch": 1.8652707328652456,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0926694059493155e-06,
+      "loss": 0.9976,
+      "step": 9410
+    },
+    {
+      "epoch": 1.8654706279203417,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0916953324319144e-06,
+      "loss": 0.9687,
+      "step": 9411
+    },
+    {
+      "epoch": 1.8656705229754378,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0907213436812037e-06,
+      "loss": 0.9989,
+      "step": 9412
+    },
+    {
+      "epoch": 1.8658704180305339,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.0897474397404466e-06,
+      "loss": 0.9618,
+      "step": 9413
+    },
+    {
+      "epoch": 1.86607031308563,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0887736206529053e-06,
+      "loss": 1.0123,
+      "step": 9414
+    },
+    {
+      "epoch": 1.866270208140726,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0877998864618334e-06,
+      "loss": 1.0452,
+      "step": 9415
+    },
+    {
+      "epoch": 1.8664701031958222,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0868262372104873e-06,
+      "loss": 0.9167,
+      "step": 9416
+    },
+    {
+      "epoch": 1.8666699982509183,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.085852672942116e-06,
+      "loss": 1.0233,
+      "step": 9417
+    },
+    {
+      "epoch": 1.8668698933060144,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.084879193699963e-06,
+      "loss": 1.0442,
+      "step": 9418
+    },
+    {
+      "epoch": 1.8670697883611105,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.083905799527273e-06,
+      "loss": 1.0295,
+      "step": 9419
+    },
+    {
+      "epoch": 1.8672696834162066,
+      "grad_norm": 2.25,
+      "learning_rate": 3.082932490467283e-06,
+      "loss": 0.9886,
+      "step": 9420
+    },
+    {
+      "epoch": 1.8674695784713027,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0819592665632262e-06,
+      "loss": 0.9874,
+      "step": 9421
+    },
+    {
+      "epoch": 1.8676694735263988,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.080986127858334e-06,
+      "loss": 0.991,
+      "step": 9422
+    },
+    {
+      "epoch": 1.8678693685814947,
+      "grad_norm": 2.125,
+      "learning_rate": 3.080013074395834e-06,
+      "loss": 0.9268,
+      "step": 9423
+    },
+    {
+      "epoch": 1.8680692636365908,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0790401062189488e-06,
+      "loss": 0.8571,
+      "step": 9424
+    },
+    {
+      "epoch": 1.8682691586916869,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0780672233708954e-06,
+      "loss": 1.0306,
+      "step": 9425
+    },
+    {
+      "epoch": 1.868469053746783,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0770944258948913e-06,
+      "loss": 0.9527,
+      "step": 9426
+    },
+    {
+      "epoch": 1.868668948801879,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.076121713834147e-06,
+      "loss": 0.8627,
+      "step": 9427
+    },
+    {
+      "epoch": 1.868868843856975,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.07514908723187e-06,
+      "loss": 0.918,
+      "step": 9428
+    },
+    {
+      "epoch": 1.869068738912071,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.074176546131266e-06,
+      "loss": 0.9442,
+      "step": 9429
+    },
+    {
+      "epoch": 1.8692686339671671,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0732040905755333e-06,
+      "loss": 1.0042,
+      "step": 9430
+    },
+    {
+      "epoch": 1.8694685290222632,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.0722317206078692e-06,
+      "loss": 1.0081,
+      "step": 9431
+    },
+    {
+      "epoch": 1.8696684240773593,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0712594362714677e-06,
+      "loss": 0.9252,
+      "step": 9432
+    },
+    {
+      "epoch": 1.8698683191324554,
+      "grad_norm": 2.25,
+      "learning_rate": 3.0702872376095146e-06,
+      "loss": 0.8877,
+      "step": 9433
+    },
+    {
+      "epoch": 1.8700682141875515,
+      "grad_norm": 2.125,
+      "learning_rate": 3.069315124665196e-06,
+      "loss": 0.9872,
+      "step": 9434
+    },
+    {
+      "epoch": 1.8702681092426476,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.068343097481694e-06,
+      "loss": 0.9615,
+      "step": 9435
+    },
+    {
+      "epoch": 1.8704680042977437,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0673711561021857e-06,
+      "loss": 1.0107,
+      "step": 9436
+    },
+    {
+      "epoch": 1.8706678993528398,
+      "grad_norm": 2.0,
+      "learning_rate": 3.0663993005698444e-06,
+      "loss": 0.9041,
+      "step": 9437
+    },
+    {
+      "epoch": 1.870867794407936,
+      "grad_norm": 2.25,
+      "learning_rate": 3.0654275309278382e-06,
+      "loss": 1.033,
+      "step": 9438
+    },
+    {
+      "epoch": 1.871067689463032,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0644558472193355e-06,
+      "loss": 0.9913,
+      "step": 9439
+    },
+    {
+      "epoch": 1.8712675845181281,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0634842494874974e-06,
+      "loss": 0.9967,
+      "step": 9440
+    },
+    {
+      "epoch": 1.871467479573224,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0625127377754814e-06,
+      "loss": 0.9785,
+      "step": 9441
+    },
+    {
+      "epoch": 1.8716673746283201,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0615413121264427e-06,
+      "loss": 0.9842,
+      "step": 9442
+    },
+    {
+      "epoch": 1.8718672696834162,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.060569972583533e-06,
+      "loss": 0.9799,
+      "step": 9443
+    },
+    {
+      "epoch": 1.8720671647385123,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0595987191898968e-06,
+      "loss": 0.9764,
+      "step": 9444
+    },
+    {
+      "epoch": 1.8722670597936082,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.0586275519886792e-06,
+      "loss": 1.0285,
+      "step": 9445
+    },
+    {
+      "epoch": 1.8724669548487043,
+      "grad_norm": 2.0,
+      "learning_rate": 3.057656471023018e-06,
+      "loss": 0.9092,
+      "step": 9446
+    },
+    {
+      "epoch": 1.8726668499038004,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.056685476336048e-06,
+      "loss": 0.9515,
+      "step": 9447
+    },
+    {
+      "epoch": 1.8728667449588965,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0557145679709033e-06,
+      "loss": 0.9823,
+      "step": 9448
+    },
+    {
+      "epoch": 1.8730666400139926,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.0547437459707086e-06,
+      "loss": 0.9461,
+      "step": 9449
+    },
+    {
+      "epoch": 1.8732665350690887,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.053773010378589e-06,
+      "loss": 0.8853,
+      "step": 9450
+    },
+    {
+      "epoch": 1.8734664301241848,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.052802361237665e-06,
+      "loss": 1.0347,
+      "step": 9451
+    },
+    {
+      "epoch": 1.873666325179281,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0518317985910516e-06,
+      "loss": 0.9607,
+      "step": 9452
+    },
+    {
+      "epoch": 1.873866220234377,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.050861322481863e-06,
+      "loss": 0.9928,
+      "step": 9453
+    },
+    {
+      "epoch": 1.874066115289473,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0498909329532047e-06,
+      "loss": 1.031,
+      "step": 9454
+    },
+    {
+      "epoch": 1.8742660103445692,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0489206300481846e-06,
+      "loss": 0.9818,
+      "step": 9455
+    },
+    {
+      "epoch": 1.8744659053996653,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0479504138099e-06,
+      "loss": 1.0457,
+      "step": 9456
+    },
+    {
+      "epoch": 1.8746658004547614,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.046980284281451e-06,
+      "loss": 0.8744,
+      "step": 9457
+    },
+    {
+      "epoch": 1.8748656955098575,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.046010241505928e-06,
+      "loss": 0.9854,
+      "step": 9458
+    },
+    {
+      "epoch": 1.8750655905649534,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0450402855264204e-06,
+      "loss": 0.9403,
+      "step": 9459
+    },
+    {
+      "epoch": 1.8752654856200495,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0440704163860167e-06,
+      "loss": 1.02,
+      "step": 9460
+    },
+    {
+      "epoch": 1.8754653806751456,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0431006341277945e-06,
+      "loss": 0.9089,
+      "step": 9461
+    },
+    {
+      "epoch": 1.8756652757302417,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.042130938794834e-06,
+      "loss": 0.8881,
+      "step": 9462
+    },
+    {
+      "epoch": 1.8758651707853375,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.041161330430208e-06,
+      "loss": 0.8822,
+      "step": 9463
+    },
+    {
+      "epoch": 1.8760650658404336,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0401918090769868e-06,
+      "loss": 0.9422,
+      "step": 9464
+    },
+    {
+      "epoch": 1.8762649608955297,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0392223747782358e-06,
+      "loss": 0.9783,
+      "step": 9465
+    },
+    {
+      "epoch": 1.8764648559506258,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0382530275770172e-06,
+      "loss": 0.9372,
+      "step": 9466
+    },
+    {
+      "epoch": 1.876664751005722,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.037283767516389e-06,
+      "loss": 0.9253,
+      "step": 9467
+    },
+    {
+      "epoch": 1.876864646060818,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.036314594639408e-06,
+      "loss": 1.0243,
+      "step": 9468
+    },
+    {
+      "epoch": 1.8770645411159141,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0353455089891213e-06,
+      "loss": 0.9791,
+      "step": 9469
+    },
+    {
+      "epoch": 1.8772644361710102,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0343765106085778e-06,
+      "loss": 0.973,
+      "step": 9470
+    },
+    {
+      "epoch": 1.8774643312261063,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0334075995408206e-06,
+      "loss": 1.0732,
+      "step": 9471
+    },
+    {
+      "epoch": 1.8776642262812024,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.032438775828887e-06,
+      "loss": 0.931,
+      "step": 9472
+    },
+    {
+      "epoch": 1.8778641213362985,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0314700395158125e-06,
+      "loss": 1.0091,
+      "step": 9473
+    },
+    {
+      "epoch": 1.8780640163913946,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0305013906446296e-06,
+      "loss": 0.9834,
+      "step": 9474
+    },
+    {
+      "epoch": 1.8782639114464907,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.0295328292583636e-06,
+      "loss": 0.9206,
+      "step": 9475
+    },
+    {
+      "epoch": 1.8784638065015866,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.0285643554000398e-06,
+      "loss": 1.083,
+      "step": 9476
+    },
+    {
+      "epoch": 1.8786637015566827,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0275959691126763e-06,
+      "loss": 0.9827,
+      "step": 9477
+    },
+    {
+      "epoch": 1.8788635966117788,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0266276704392884e-06,
+      "loss": 0.97,
+      "step": 9478
+    },
+    {
+      "epoch": 1.879063491666875,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.0256594594228906e-06,
+      "loss": 1.0129,
+      "step": 9479
+    },
+    {
+      "epoch": 1.879263386721971,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.024691336106487e-06,
+      "loss": 0.9386,
+      "step": 9480
+    },
+    {
+      "epoch": 1.879463281777067,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0237233005330833e-06,
+      "loss": 0.9983,
+      "step": 9481
+    },
+    {
+      "epoch": 1.879663176832163,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.022755352745681e-06,
+      "loss": 0.9233,
+      "step": 9482
+    },
+    {
+      "epoch": 1.879863071887259,
+      "grad_norm": 2.125,
+      "learning_rate": 3.021787492787273e-06,
+      "loss": 1.0003,
+      "step": 9483
+    },
+    {
+      "epoch": 1.8800629669423552,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.020819720700855e-06,
+      "loss": 1.0818,
+      "step": 9484
+    },
+    {
+      "epoch": 1.8802628619974513,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.019852036529412e-06,
+      "loss": 0.9582,
+      "step": 9485
+    },
+    {
+      "epoch": 1.8804627570525474,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.018884440315932e-06,
+      "loss": 0.9328,
+      "step": 9486
+    },
+    {
+      "epoch": 1.8806626521076435,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.0179169321033926e-06,
+      "loss": 0.9897,
+      "step": 9487
+    },
+    {
+      "epoch": 1.8808625471627396,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.016949511934771e-06,
+      "loss": 0.9947,
+      "step": 9488
+    },
+    {
+      "epoch": 1.8810624422178357,
+      "grad_norm": 2.125,
+      "learning_rate": 3.015982179853041e-06,
+      "loss": 0.9682,
+      "step": 9489
+    },
+    {
+      "epoch": 1.8812623372729318,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.0150149359011694e-06,
+      "loss": 0.962,
+      "step": 9490
+    },
+    {
+      "epoch": 1.881462232328028,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0140477801221235e-06,
+      "loss": 1.0181,
+      "step": 9491
+    },
+    {
+      "epoch": 1.881662127383124,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.0130807125588625e-06,
+      "loss": 1.0336,
+      "step": 9492
+    },
+    {
+      "epoch": 1.88186202243822,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0121137332543438e-06,
+      "loss": 0.9699,
+      "step": 9493
+    },
+    {
+      "epoch": 1.882061917493316,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.0111468422515215e-06,
+      "loss": 1.0088,
+      "step": 9494
+    },
+    {
+      "epoch": 1.882261812548412,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0101800395933433e-06,
+      "loss": 0.9987,
+      "step": 9495
+    },
+    {
+      "epoch": 1.8824617076035082,
+      "grad_norm": 2.125,
+      "learning_rate": 3.0092133253227563e-06,
+      "loss": 0.9607,
+      "step": 9496
+    },
+    {
+      "epoch": 1.8826616026586043,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0082466994827e-06,
+      "loss": 0.9607,
+      "step": 9497
+    },
+    {
+      "epoch": 1.8828614977137001,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.007280162116112e-06,
+      "loss": 0.9824,
+      "step": 9498
+    },
+    {
+      "epoch": 1.8830613927687962,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.0063137132659277e-06,
+      "loss": 1.0015,
+      "step": 9499
+    },
+    {
+      "epoch": 1.8832612878238923,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.0053473529750743e-06,
+      "loss": 0.907,
+      "step": 9500
+    },
+    {
+      "epoch": 1.8834611828789884,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0043810812864783e-06,
+      "loss": 0.9253,
+      "step": 9501
+    },
+    {
+      "epoch": 1.8836610779340845,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.0034148982430624e-06,
+      "loss": 0.951,
+      "step": 9502
+    },
+    {
+      "epoch": 1.8838609729891806,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.002448803887743e-06,
+      "loss": 0.9961,
+      "step": 9503
+    },
+    {
+      "epoch": 1.8840608680442767,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.001482798263435e-06,
+      "loss": 1.0291,
+      "step": 9504
+    },
+    {
+      "epoch": 1.8842607630993728,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0005168814130463e-06,
+      "loss": 0.9918,
+      "step": 9505
+    },
+    {
+      "epoch": 1.884460658154469,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9995510533794846e-06,
+      "loss": 0.971,
+      "step": 9506
+    },
+    {
+      "epoch": 1.884660553209565,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.9985853142056527e-06,
+      "loss": 0.9807,
+      "step": 9507
+    },
+    {
+      "epoch": 1.8848604482646611,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.997619663934446e-06,
+      "loss": 0.9397,
+      "step": 9508
+    },
+    {
+      "epoch": 1.8850603433197572,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.9966541026087602e-06,
+      "loss": 0.964,
+      "step": 9509
+    },
+    {
+      "epoch": 1.8852602383748533,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.995688630271486e-06,
+      "loss": 1.0068,
+      "step": 9510
+    },
+    {
+      "epoch": 1.8854601334299492,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.994723246965508e-06,
+      "loss": 0.9264,
+      "step": 9511
+    },
+    {
+      "epoch": 1.8856600284850453,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9937579527337092e-06,
+      "loss": 0.9491,
+      "step": 9512
+    },
+    {
+      "epoch": 1.8858599235401414,
+      "grad_norm": 2.125,
+      "learning_rate": 2.992792747618969e-06,
+      "loss": 1.0221,
+      "step": 9513
+    },
+    {
+      "epoch": 1.8860598185952375,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9918276316641592e-06,
+      "loss": 0.957,
+      "step": 9514
+    },
+    {
+      "epoch": 1.8862597136503336,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9908626049121523e-06,
+      "loss": 0.936,
+      "step": 9515
+    },
+    {
+      "epoch": 1.8864596087054295,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.989897667405813e-06,
+      "loss": 0.9702,
+      "step": 9516
+    },
+    {
+      "epoch": 1.8866595037605256,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9889328191880075e-06,
+      "loss": 0.9699,
+      "step": 9517
+    },
+    {
+      "epoch": 1.8868593988156217,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.987968060301588e-06,
+      "loss": 1.0664,
+      "step": 9518
+    },
+    {
+      "epoch": 1.8870592938707178,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9870033907894146e-06,
+      "loss": 1.07,
+      "step": 9519
+    },
+    {
+      "epoch": 1.887259188925814,
+      "grad_norm": 2.125,
+      "learning_rate": 2.986038810694334e-06,
+      "loss": 0.9554,
+      "step": 9520
+    },
+    {
+      "epoch": 1.88745908398091,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.985074320059195e-06,
+      "loss": 0.9381,
+      "step": 9521
+    },
+    {
+      "epoch": 1.887658979036006,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.9841099189268397e-06,
+      "loss": 0.8263,
+      "step": 9522
+    },
+    {
+      "epoch": 1.8878588740911022,
+      "grad_norm": 2.0,
+      "learning_rate": 2.9831456073401056e-06,
+      "loss": 0.8908,
+      "step": 9523
+    },
+    {
+      "epoch": 1.8880587691461983,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.982181385341828e-06,
+      "loss": 1.0653,
+      "step": 9524
+    },
+    {
+      "epoch": 1.8882586642012944,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9812172529748395e-06,
+      "loss": 0.9158,
+      "step": 9525
+    },
+    {
+      "epoch": 1.8884585592563905,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9802532102819637e-06,
+      "loss": 1.1446,
+      "step": 9526
+    },
+    {
+      "epoch": 1.8886584543114866,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9792892573060257e-06,
+      "loss": 0.9912,
+      "step": 9527
+    },
+    {
+      "epoch": 1.8888583493665827,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9783253940898417e-06,
+      "loss": 0.9292,
+      "step": 9528
+    },
+    {
+      "epoch": 1.8890582444216786,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.977361620676228e-06,
+      "loss": 0.9676,
+      "step": 9529
+    },
+    {
+      "epoch": 1.8892581394767747,
+      "grad_norm": 2.25,
+      "learning_rate": 2.976397937107996e-06,
+      "loss": 1.0119,
+      "step": 9530
+    },
+    {
+      "epoch": 1.8894580345318708,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9754343434279504e-06,
+      "loss": 1.0663,
+      "step": 9531
+    },
+    {
+      "epoch": 1.8896579295869669,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.974470839678895e-06,
+      "loss": 0.939,
+      "step": 9532
+    },
+    {
+      "epoch": 1.8898578246420628,
+      "grad_norm": 2.5,
+      "learning_rate": 2.97350742590363e-06,
+      "loss": 0.9304,
+      "step": 9533
+    },
+    {
+      "epoch": 1.8900577196971589,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9725441021449477e-06,
+      "loss": 0.9278,
+      "step": 9534
+    },
+    {
+      "epoch": 1.890257614752255,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.9715808684456402e-06,
+      "loss": 0.952,
+      "step": 9535
+    },
+    {
+      "epoch": 1.890457509807351,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9706177248484936e-06,
+      "loss": 1.0302,
+      "step": 9536
+    },
+    {
+      "epoch": 1.8906574048624472,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9696546713962904e-06,
+      "loss": 0.9503,
+      "step": 9537
+    },
+    {
+      "epoch": 1.8908572999175433,
+      "grad_norm": 2.125,
+      "learning_rate": 2.968691708131811e-06,
+      "loss": 0.9209,
+      "step": 9538
+    },
+    {
+      "epoch": 1.8910571949726394,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.9677288350978286e-06,
+      "loss": 0.9386,
+      "step": 9539
+    },
+    {
+      "epoch": 1.8912570900277355,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9667660523371134e-06,
+      "loss": 0.9958,
+      "step": 9540
+    },
+    {
+      "epoch": 1.8914569850828316,
+      "grad_norm": 2.25,
+      "learning_rate": 2.965803359892434e-06,
+      "loss": 0.9637,
+      "step": 9541
+    },
+    {
+      "epoch": 1.8916568801379277,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.9648407578065515e-06,
+      "loss": 1.0209,
+      "step": 9542
+    },
+    {
+      "epoch": 1.8918567751930238,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9638782461222258e-06,
+      "loss": 1.0401,
+      "step": 9543
+    },
+    {
+      "epoch": 1.8920566702481199,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9629158248822105e-06,
+      "loss": 0.9794,
+      "step": 9544
+    },
+    {
+      "epoch": 1.892256565303216,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9619534941292562e-06,
+      "loss": 0.9372,
+      "step": 9545
+    },
+    {
+      "epoch": 1.8924564603583118,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9609912539061114e-06,
+      "loss": 0.9428,
+      "step": 9546
+    },
+    {
+      "epoch": 1.892656355413408,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.960029104255516e-06,
+      "loss": 0.9534,
+      "step": 9547
+    },
+    {
+      "epoch": 1.892856250468504,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.959067045220212e-06,
+      "loss": 1.0459,
+      "step": 9548
+    },
+    {
+      "epoch": 1.8930561455236001,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.95810507684293e-06,
+      "loss": 1.0112,
+      "step": 9549
+    },
+    {
+      "epoch": 1.8932560405786962,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9571431991664036e-06,
+      "loss": 0.9578,
+      "step": 9550
+    },
+    {
+      "epoch": 1.893455935633792,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9561814122333564e-06,
+      "loss": 1.0109,
+      "step": 9551
+    },
+    {
+      "epoch": 1.8936558306888882,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.955219716086514e-06,
+      "loss": 0.8693,
+      "step": 9552
+    },
+    {
+      "epoch": 1.8938557257439843,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.954258110768593e-06,
+      "loss": 0.9462,
+      "step": 9553
+    },
+    {
+      "epoch": 1.8940556207990804,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9532965963223076e-06,
+      "loss": 1.0066,
+      "step": 9554
+    },
+    {
+      "epoch": 1.8942555158541765,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.95233517279037e-06,
+      "loss": 0.9776,
+      "step": 9555
+    },
+    {
+      "epoch": 1.8944554109092726,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.951373840215484e-06,
+      "loss": 1.0024,
+      "step": 9556
+    },
+    {
+      "epoch": 1.8946553059643687,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.950412598640353e-06,
+      "loss": 0.9587,
+      "step": 9557
+    },
+    {
+      "epoch": 1.8948552010194648,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.9494514481076773e-06,
+      "loss": 1.0363,
+      "step": 9558
+    },
+    {
+      "epoch": 1.895055096074561,
+      "grad_norm": 2.25,
+      "learning_rate": 2.948490388660148e-06,
+      "loss": 1.0407,
+      "step": 9559
+    },
+    {
+      "epoch": 1.895254991129657,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9475294203404557e-06,
+      "loss": 0.9162,
+      "step": 9560
+    },
+    {
+      "epoch": 1.895454886184753,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9465685431912895e-06,
+      "loss": 0.9416,
+      "step": 9561
+    },
+    {
+      "epoch": 1.8956547812398492,
+      "grad_norm": 2.25,
+      "learning_rate": 2.945607757255328e-06,
+      "loss": 1.0465,
+      "step": 9562
+    },
+    {
+      "epoch": 1.8958546762949453,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9446470625752497e-06,
+      "loss": 0.964,
+      "step": 9563
+    },
+    {
+      "epoch": 1.8960545713500412,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9436864591937312e-06,
+      "loss": 0.9395,
+      "step": 9564
+    },
+    {
+      "epoch": 1.8962544664051373,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9427259471534396e-06,
+      "loss": 0.9681,
+      "step": 9565
+    },
+    {
+      "epoch": 1.8964543614602334,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.9417655264970424e-06,
+      "loss": 1.02,
+      "step": 9566
+    },
+    {
+      "epoch": 1.8966542565153295,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9408051972672e-06,
+      "loss": 0.9638,
+      "step": 9567
+    },
+    {
+      "epoch": 1.8968541515704254,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9398449595065705e-06,
+      "loss": 1.0314,
+      "step": 9568
+    },
+    {
+      "epoch": 1.8970540466255215,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9388848132578087e-06,
+      "loss": 0.8992,
+      "step": 9569
+    },
+    {
+      "epoch": 1.8972539416806176,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.937924758563563e-06,
+      "loss": 0.9696,
+      "step": 9570
+    },
+    {
+      "epoch": 1.8974538367357137,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9369647954664783e-06,
+      "loss": 0.9212,
+      "step": 9571
+    },
+    {
+      "epoch": 1.8976537317908098,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.9360049240091988e-06,
+      "loss": 0.9186,
+      "step": 9572
+    },
+    {
+      "epoch": 1.8978536268459059,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.9350451442343585e-06,
+      "loss": 0.9811,
+      "step": 9573
+    },
+    {
+      "epoch": 1.898053521901002,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.9340854561845945e-06,
+      "loss": 1.0047,
+      "step": 9574
+    },
+    {
+      "epoch": 1.898253416956098,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.933125859902532e-06,
+      "loss": 1.0303,
+      "step": 9575
+    },
+    {
+      "epoch": 1.8984533120111942,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9321663554307977e-06,
+      "loss": 0.9834,
+      "step": 9576
+    },
+    {
+      "epoch": 1.8986532070662903,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.931206942812015e-06,
+      "loss": 0.9697,
+      "step": 9577
+    },
+    {
+      "epoch": 1.8988531021213864,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9302476220887975e-06,
+      "loss": 1.0374,
+      "step": 9578
+    },
+    {
+      "epoch": 1.8990529971764825,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.929288393303762e-06,
+      "loss": 0.9615,
+      "step": 9579
+    },
+    {
+      "epoch": 1.8992528922315786,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.928329256499512e-06,
+      "loss": 0.9689,
+      "step": 9580
+    },
+    {
+      "epoch": 1.8994527872866747,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.9273702117186564e-06,
+      "loss": 0.9642,
+      "step": 9581
+    },
+    {
+      "epoch": 1.8996526823417705,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.926411259003794e-06,
+      "loss": 1.0004,
+      "step": 9582
+    },
+    {
+      "epoch": 1.8998525773968666,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.9254523983975224e-06,
+      "loss": 0.9807,
+      "step": 9583
+    },
+    {
+      "epoch": 1.9000524724519627,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.924493629942434e-06,
+      "loss": 0.9657,
+      "step": 9584
+    },
+    {
+      "epoch": 1.9002523675070588,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.923534953681116e-06,
+      "loss": 1.0369,
+      "step": 9585
+    },
+    {
+      "epoch": 1.9004522625621547,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.922576369656155e-06,
+      "loss": 0.8696,
+      "step": 9586
+    },
+    {
+      "epoch": 1.9006521576172508,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9216178779101276e-06,
+      "loss": 0.9599,
+      "step": 9587
+    },
+    {
+      "epoch": 1.900852052672347,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.9206594784856133e-06,
+      "loss": 0.9909,
+      "step": 9588
+    },
+    {
+      "epoch": 1.901051947727443,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9197011714251833e-06,
+      "loss": 1.0464,
+      "step": 9589
+    },
+    {
+      "epoch": 1.9012518427825391,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9187429567714044e-06,
+      "loss": 0.9008,
+      "step": 9590
+    },
+    {
+      "epoch": 1.9014517378376352,
+      "grad_norm": 2.125,
+      "learning_rate": 2.9177848345668426e-06,
+      "loss": 1.0138,
+      "step": 9591
+    },
+    {
+      "epoch": 1.9016516328927313,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9168268048540527e-06,
+      "loss": 1.0418,
+      "step": 9592
+    },
+    {
+      "epoch": 1.9018515279478274,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.9158688676755966e-06,
+      "loss": 0.9981,
+      "step": 9593
+    },
+    {
+      "epoch": 1.9020514230029235,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.914911023074023e-06,
+      "loss": 0.8566,
+      "step": 9594
+    },
+    {
+      "epoch": 1.9022513180580196,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.913953271091876e-06,
+      "loss": 0.9912,
+      "step": 9595
+    },
+    {
+      "epoch": 1.9024512131131157,
+      "grad_norm": 2.125,
+      "learning_rate": 2.912995611771705e-06,
+      "loss": 0.9226,
+      "step": 9596
+    },
+    {
+      "epoch": 1.9026511081682118,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.9120380451560456e-06,
+      "loss": 0.948,
+      "step": 9597
+    },
+    {
+      "epoch": 1.902851003223308,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.911080571287433e-06,
+      "loss": 0.9533,
+      "step": 9598
+    },
+    {
+      "epoch": 1.9030508982784038,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.9101231902083963e-06,
+      "loss": 1.0581,
+      "step": 9599
+    },
+    {
+      "epoch": 1.9032507933335,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.909165901961467e-06,
+      "loss": 1.0024,
+      "step": 9600
+    },
+    {
+      "epoch": 1.903450688388596,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.908208706589164e-06,
+      "loss": 0.9489,
+      "step": 9601
+    },
+    {
+      "epoch": 1.903650583443692,
+      "grad_norm": 2.125,
+      "learning_rate": 2.907251604134006e-06,
+      "loss": 0.9358,
+      "step": 9602
+    },
+    {
+      "epoch": 1.9038504784987882,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9062945946385092e-06,
+      "loss": 1.04,
+      "step": 9603
+    },
+    {
+      "epoch": 1.904050373553884,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.9053376781451836e-06,
+      "loss": 0.9114,
+      "step": 9604
+    },
+    {
+      "epoch": 1.9042502686089802,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.904380854696532e-06,
+      "loss": 0.9902,
+      "step": 9605
+    },
+    {
+      "epoch": 1.9044501636640763,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.9034241243350615e-06,
+      "loss": 0.9319,
+      "step": 9606
+    },
+    {
+      "epoch": 1.9046500587191724,
+      "grad_norm": 2.125,
+      "learning_rate": 2.902467487103267e-06,
+      "loss": 0.96,
+      "step": 9607
+    },
+    {
+      "epoch": 1.9048499537742685,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.901510943043641e-06,
+      "loss": 0.9625,
+      "step": 9608
+    },
+    {
+      "epoch": 1.9050498488293646,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.9005544921986774e-06,
+      "loss": 0.8741,
+      "step": 9609
+    },
+    {
+      "epoch": 1.9052497438844607,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8995981346108598e-06,
+      "loss": 1.0453,
+      "step": 9610
+    },
+    {
+      "epoch": 1.9054496389395568,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8986418703226655e-06,
+      "loss": 0.9462,
+      "step": 9611
+    },
+    {
+      "epoch": 1.9056495339946529,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8976856993765766e-06,
+      "loss": 0.9167,
+      "step": 9612
+    },
+    {
+      "epoch": 1.905849429049749,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.896729621815064e-06,
+      "loss": 1.0164,
+      "step": 9613
+    },
+    {
+      "epoch": 1.906049324104845,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8957736376805963e-06,
+      "loss": 0.9024,
+      "step": 9614
+    },
+    {
+      "epoch": 1.9062492191599412,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8948177470156404e-06,
+      "loss": 0.9204,
+      "step": 9615
+    },
+    {
+      "epoch": 1.9064491142150373,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8938619498626542e-06,
+      "loss": 0.9531,
+      "step": 9616
+    },
+    {
+      "epoch": 1.9066490092701331,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.8929062462640946e-06,
+      "loss": 1.0757,
+      "step": 9617
+    },
+    {
+      "epoch": 1.9068489043252292,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8919506362624156e-06,
+      "loss": 0.9839,
+      "step": 9618
+    },
+    {
+      "epoch": 1.9070487993803253,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8909951199000645e-06,
+      "loss": 0.8796,
+      "step": 9619
+    },
+    {
+      "epoch": 1.9072486944354214,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8900396972194834e-06,
+      "loss": 1.0045,
+      "step": 9620
+    },
+    {
+      "epoch": 1.9074485894905173,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8890843682631147e-06,
+      "loss": 0.9391,
+      "step": 9621
+    },
+    {
+      "epoch": 1.9076484845456134,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8881291330733933e-06,
+      "loss": 0.9263,
+      "step": 9622
+    },
+    {
+      "epoch": 1.9078483796007095,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8871739916927487e-06,
+      "loss": 1.0736,
+      "step": 9623
+    },
+    {
+      "epoch": 1.9080482746558056,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8862189441636113e-06,
+      "loss": 0.8903,
+      "step": 9624
+    },
+    {
+      "epoch": 1.9082481697109017,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.8852639905284026e-06,
+      "loss": 0.9666,
+      "step": 9625
+    },
+    {
+      "epoch": 1.9084480647659978,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8843091308295395e-06,
+      "loss": 0.9188,
+      "step": 9626
+    },
+    {
+      "epoch": 1.908647959821094,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.883354365109441e-06,
+      "loss": 0.917,
+      "step": 9627
+    },
+    {
+      "epoch": 1.90884785487619,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.882399693410516e-06,
+      "loss": 0.9333,
+      "step": 9628
+    },
+    {
+      "epoch": 1.9090477499312861,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8814451157751697e-06,
+      "loss": 0.9324,
+      "step": 9629
+    },
+    {
+      "epoch": 1.9092476449863822,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.880490632245803e-06,
+      "loss": 0.9385,
+      "step": 9630
+    },
+    {
+      "epoch": 1.9094475400414783,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8795362428648186e-06,
+      "loss": 0.8715,
+      "step": 9631
+    },
+    {
+      "epoch": 1.9096474350965744,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.878581947674608e-06,
+      "loss": 0.8694,
+      "step": 9632
+    },
+    {
+      "epoch": 1.9098473301516705,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8776277467175583e-06,
+      "loss": 0.8945,
+      "step": 9633
+    },
+    {
+      "epoch": 1.9100472252067664,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8766736400360595e-06,
+      "loss": 0.9825,
+      "step": 9634
+    },
+    {
+      "epoch": 1.9102471202618625,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.875719627672491e-06,
+      "loss": 1.0558,
+      "step": 9635
+    },
+    {
+      "epoch": 1.9104470153169586,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.874765709669227e-06,
+      "loss": 0.9971,
+      "step": 9636
+    },
+    {
+      "epoch": 1.9106469103720547,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8738118860686457e-06,
+      "loss": 0.9931,
+      "step": 9637
+    },
+    {
+      "epoch": 1.9108468054271508,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8728581569131137e-06,
+      "loss": 1.0429,
+      "step": 9638
+    },
+    {
+      "epoch": 1.9110467004822467,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8719045222449925e-06,
+      "loss": 0.9666,
+      "step": 9639
+    },
+    {
+      "epoch": 1.9112465955373428,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.8709509821066478e-06,
+      "loss": 0.9532,
+      "step": 9640
+    },
+    {
+      "epoch": 1.9114464905924389,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.869997536540435e-06,
+      "loss": 0.9724,
+      "step": 9641
+    },
+    {
+      "epoch": 1.911646385647535,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.8690441855887e-06,
+      "loss": 0.9422,
+      "step": 9642
+    },
+    {
+      "epoch": 1.911846280702631,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8680909292937965e-06,
+      "loss": 0.963,
+      "step": 9643
+    },
+    {
+      "epoch": 1.9120461757577272,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.867137767698066e-06,
+      "loss": 0.9424,
+      "step": 9644
+    },
+    {
+      "epoch": 1.9122460708128233,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8661847008438466e-06,
+      "loss": 0.9263,
+      "step": 9645
+    },
+    {
+      "epoch": 1.9124459658679194,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8652317287734766e-06,
+      "loss": 0.9459,
+      "step": 9646
+    },
+    {
+      "epoch": 1.9126458609230155,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8642788515292854e-06,
+      "loss": 0.9803,
+      "step": 9647
+    },
+    {
+      "epoch": 1.9128457559781116,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8633260691535973e-06,
+      "loss": 1.0479,
+      "step": 9648
+    },
+    {
+      "epoch": 1.9130456510332077,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.862373381688739e-06,
+      "loss": 0.8759,
+      "step": 9649
+    },
+    {
+      "epoch": 1.9132455460883038,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8614207891770275e-06,
+      "loss": 0.9741,
+      "step": 9650
+    },
+    {
+      "epoch": 1.9134454411433999,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.8604682916607728e-06,
+      "loss": 0.9743,
+      "step": 9651
+    },
+    {
+      "epoch": 1.9136453361984957,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.859515889182291e-06,
+      "loss": 1.0026,
+      "step": 9652
+    },
+    {
+      "epoch": 1.9138452312535918,
+      "grad_norm": 2.125,
+      "learning_rate": 2.858563581783885e-06,
+      "loss": 0.9185,
+      "step": 9653
+    },
+    {
+      "epoch": 1.914045126308688,
+      "grad_norm": 2.125,
+      "learning_rate": 2.8576113695078534e-06,
+      "loss": 0.9673,
+      "step": 9654
+    },
+    {
+      "epoch": 1.914245021363784,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.856659252396498e-06,
+      "loss": 0.9226,
+      "step": 9655
+    },
+    {
+      "epoch": 1.91444491641888,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8557072304921094e-06,
+      "loss": 1.0275,
+      "step": 9656
+    },
+    {
+      "epoch": 1.914644811473976,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.8547553038369756e-06,
+      "loss": 0.9081,
+      "step": 9657
+    },
+    {
+      "epoch": 1.9148447065290721,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.853803472473383e-06,
+      "loss": 0.988,
+      "step": 9658
+    },
+    {
+      "epoch": 1.9150446015841682,
+      "grad_norm": 2.125,
+      "learning_rate": 2.8528517364436116e-06,
+      "loss": 0.9797,
+      "step": 9659
+    },
+    {
+      "epoch": 1.9152444966392643,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.8519000957899368e-06,
+      "loss": 0.9943,
+      "step": 9660
+    },
+    {
+      "epoch": 1.9154443916943604,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.850948550554628e-06,
+      "loss": 1.0108,
+      "step": 9661
+    },
+    {
+      "epoch": 1.9156442867494565,
+      "grad_norm": 2.25,
+      "learning_rate": 2.8499971007799576e-06,
+      "loss": 0.9656,
+      "step": 9662
+    },
+    {
+      "epoch": 1.9158441818045526,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.8490457465081853e-06,
+      "loss": 0.8991,
+      "step": 9663
+    },
+    {
+      "epoch": 1.9160440768596487,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.84809448778157e-06,
+      "loss": 1.0282,
+      "step": 9664
+    },
+    {
+      "epoch": 1.9162439719147448,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.8471433246423697e-06,
+      "loss": 0.9533,
+      "step": 9665
+    },
+    {
+      "epoch": 1.916443866969841,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8461922571328327e-06,
+      "loss": 0.9021,
+      "step": 9666
+    },
+    {
+      "epoch": 1.916643762024937,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8452412852952037e-06,
+      "loss": 0.8995,
+      "step": 9667
+    },
+    {
+      "epoch": 1.9168436570800331,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.844290409171729e-06,
+      "loss": 0.9884,
+      "step": 9668
+    },
+    {
+      "epoch": 1.917043552135129,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8433396288046433e-06,
+      "loss": 0.977,
+      "step": 9669
+    },
+    {
+      "epoch": 1.917243447190225,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8423889442361797e-06,
+      "loss": 1.0235,
+      "step": 9670
+    },
+    {
+      "epoch": 1.9174433422453212,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.8414383555085708e-06,
+      "loss": 0.9775,
+      "step": 9671
+    },
+    {
+      "epoch": 1.9176432373004173,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8404878626640408e-06,
+      "loss": 0.9528,
+      "step": 9672
+    },
+    {
+      "epoch": 1.9178431323555134,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.839537465744806e-06,
+      "loss": 1.037,
+      "step": 9673
+    },
+    {
+      "epoch": 1.9180430274106093,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.8385871647930886e-06,
+      "loss": 1.0233,
+      "step": 9674
+    },
+    {
+      "epoch": 1.9182429224657054,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.837636959851098e-06,
+      "loss": 1.0528,
+      "step": 9675
+    },
+    {
+      "epoch": 1.9184428175208015,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.83668685096104e-06,
+      "loss": 0.9207,
+      "step": 9676
+    },
+    {
+      "epoch": 1.9186427125758976,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8357368381651242e-06,
+      "loss": 1.0023,
+      "step": 9677
+    },
+    {
+      "epoch": 1.9188426076309937,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.8347869215055455e-06,
+      "loss": 1.0113,
+      "step": 9678
+    },
+    {
+      "epoch": 1.9190425026860898,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8338371010244997e-06,
+      "loss": 1.0195,
+      "step": 9679
+    },
+    {
+      "epoch": 1.9192423977411859,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.83288737676418e-06,
+      "loss": 0.9366,
+      "step": 9680
+    },
+    {
+      "epoch": 1.919442292796282,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.831937748766772e-06,
+      "loss": 0.9912,
+      "step": 9681
+    },
+    {
+      "epoch": 1.919642187851378,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.830988217074455e-06,
+      "loss": 0.9882,
+      "step": 9682
+    },
+    {
+      "epoch": 1.9198420829064742,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8300387817294122e-06,
+      "loss": 1.0041,
+      "step": 9683
+    },
+    {
+      "epoch": 1.9200419779615703,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8290894427738148e-06,
+      "loss": 0.9378,
+      "step": 9684
+    },
+    {
+      "epoch": 1.9202418730166664,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.828140200249831e-06,
+      "loss": 0.9609,
+      "step": 9685
+    },
+    {
+      "epoch": 1.9204417680717625,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.827191054199629e-06,
+      "loss": 0.9683,
+      "step": 9686
+    },
+    {
+      "epoch": 1.9206416631268584,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.826242004665368e-06,
+      "loss": 1.0936,
+      "step": 9687
+    },
+    {
+      "epoch": 1.9208415581819545,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.825293051689204e-06,
+      "loss": 0.9332,
+      "step": 9688
+    },
+    {
+      "epoch": 1.9210414532370506,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8243441953132918e-06,
+      "loss": 1.0441,
+      "step": 9689
+    },
+    {
+      "epoch": 1.9212413482921467,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.8233954355797775e-06,
+      "loss": 0.9193,
+      "step": 9690
+    },
+    {
+      "epoch": 1.9214412433472425,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.8224467725308064e-06,
+      "loss": 0.9336,
+      "step": 9691
+    },
+    {
+      "epoch": 1.9216411384023386,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.821498206208515e-06,
+      "loss": 0.9434,
+      "step": 9692
+    },
+    {
+      "epoch": 1.9218410334574347,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8205497366550414e-06,
+      "loss": 1.0168,
+      "step": 9693
+    },
+    {
+      "epoch": 1.9220409285125308,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8196013639125175e-06,
+      "loss": 0.9444,
+      "step": 9694
+    },
+    {
+      "epoch": 1.922240823567627,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.818653088023065e-06,
+      "loss": 0.9618,
+      "step": 9695
+    },
+    {
+      "epoch": 1.922440718622723,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.8177049090288115e-06,
+      "loss": 0.9761,
+      "step": 9696
+    },
+    {
+      "epoch": 1.9226406136778191,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.816756826971873e-06,
+      "loss": 0.9491,
+      "step": 9697
+    },
+    {
+      "epoch": 1.9228405087329152,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.8158088418943613e-06,
+      "loss": 0.9522,
+      "step": 9698
+    },
+    {
+      "epoch": 1.9230404037880113,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.814860953838389e-06,
+      "loss": 0.9115,
+      "step": 9699
+    },
+    {
+      "epoch": 1.9232402988431074,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8139131628460605e-06,
+      "loss": 0.883,
+      "step": 9700
+    },
+    {
+      "epoch": 1.9234401938982035,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8129654689594733e-06,
+      "loss": 1.0504,
+      "step": 9701
+    },
+    {
+      "epoch": 1.9236400889532996,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8120178722207287e-06,
+      "loss": 0.9033,
+      "step": 9702
+    },
+    {
+      "epoch": 1.9238399840083957,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.811070372671918e-06,
+      "loss": 1.075,
+      "step": 9703
+    },
+    {
+      "epoch": 1.9240398790634918,
+      "grad_norm": 2.25,
+      "learning_rate": 2.810122970355124e-06,
+      "loss": 1.0194,
+      "step": 9704
+    },
+    {
+      "epoch": 1.9242397741185877,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.809175665312436e-06,
+      "loss": 0.9821,
+      "step": 9705
+    },
+    {
+      "epoch": 1.9244396691736838,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8082284575859302e-06,
+      "loss": 0.986,
+      "step": 9706
+    },
+    {
+      "epoch": 1.92463956422878,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.8072813472176807e-06,
+      "loss": 1.0015,
+      "step": 9707
+    },
+    {
+      "epoch": 1.924839459283876,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.8063343342497616e-06,
+      "loss": 0.8788,
+      "step": 9708
+    },
+    {
+      "epoch": 1.9250393543389719,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.805387418724237e-06,
+      "loss": 0.9795,
+      "step": 9709
+    },
+    {
+      "epoch": 1.925239249394068,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.804440600683167e-06,
+      "loss": 0.9401,
+      "step": 9710
+    },
+    {
+      "epoch": 1.925439144449164,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.803493880168613e-06,
+      "loss": 0.9852,
+      "step": 9711
+    },
+    {
+      "epoch": 1.9256390395042602,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.8025472572226266e-06,
+      "loss": 0.9925,
+      "step": 9712
+    },
+    {
+      "epoch": 1.9258389345593563,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.8016007318872532e-06,
+      "loss": 0.946,
+      "step": 9713
+    },
+    {
+      "epoch": 1.9260388296144524,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.800654304204543e-06,
+      "loss": 0.9504,
+      "step": 9714
+    },
+    {
+      "epoch": 1.9262387246695485,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7997079742165346e-06,
+      "loss": 0.9226,
+      "step": 9715
+    },
+    {
+      "epoch": 1.9264386197246446,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7987617419652603e-06,
+      "loss": 0.9339,
+      "step": 9716
+    },
+    {
+      "epoch": 1.9266385147797407,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.797815607492756e-06,
+      "loss": 1.0389,
+      "step": 9717
+    },
+    {
+      "epoch": 1.9268384098348368,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.7968695708410476e-06,
+      "loss": 0.9964,
+      "step": 9718
+    },
+    {
+      "epoch": 1.9270383048899329,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.7959236320521573e-06,
+      "loss": 1.0257,
+      "step": 9719
+    },
+    {
+      "epoch": 1.927238199945029,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.794977791168102e-06,
+      "loss": 0.9098,
+      "step": 9720
+    },
+    {
+      "epoch": 1.927438095000125,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.7940320482308995e-06,
+      "loss": 0.857,
+      "step": 9721
+    },
+    {
+      "epoch": 1.927637990055221,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.7930864032825582e-06,
+      "loss": 1.0006,
+      "step": 9722
+    },
+    {
+      "epoch": 1.927837885110317,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.792140856365081e-06,
+      "loss": 0.9269,
+      "step": 9723
+    },
+    {
+      "epoch": 1.9280377801654132,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.7911954075204734e-06,
+      "loss": 0.9125,
+      "step": 9724
+    },
+    {
+      "epoch": 1.9282376752205093,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7902500567907297e-06,
+      "loss": 0.8546,
+      "step": 9725
+    },
+    {
+      "epoch": 1.9284375702756054,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.7893048042178405e-06,
+      "loss": 1.0623,
+      "step": 9726
+    },
+    {
+      "epoch": 1.9286374653307012,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.788359649843797e-06,
+      "loss": 0.9814,
+      "step": 9727
+    },
+    {
+      "epoch": 1.9288373603857973,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.787414593710583e-06,
+      "loss": 1.0193,
+      "step": 9728
+    },
+    {
+      "epoch": 1.9290372554408934,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.786469635860174e-06,
+      "loss": 0.8954,
+      "step": 9729
+    },
+    {
+      "epoch": 1.9292371504959895,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.7855247763345483e-06,
+      "loss": 1.1665,
+      "step": 9730
+    },
+    {
+      "epoch": 1.9294370455510856,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7845800151756768e-06,
+      "loss": 0.9786,
+      "step": 9731
+    },
+    {
+      "epoch": 1.9296369406061817,
+      "grad_norm": 2.125,
+      "learning_rate": 2.783635352425522e-06,
+      "loss": 0.9566,
+      "step": 9732
+    },
+    {
+      "epoch": 1.9298368356612778,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.78269078812605e-06,
+      "loss": 0.8177,
+      "step": 9733
+    },
+    {
+      "epoch": 1.930036730716374,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.781746322319219e-06,
+      "loss": 1.009,
+      "step": 9734
+    },
+    {
+      "epoch": 1.93023662577147,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.7808019550469745e-06,
+      "loss": 0.9294,
+      "step": 9735
+    },
+    {
+      "epoch": 1.9304365208265661,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.779857686351273e-06,
+      "loss": 0.9833,
+      "step": 9736
+    },
+    {
+      "epoch": 1.9306364158816622,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.7789135162740555e-06,
+      "loss": 1.0077,
+      "step": 9737
+    },
+    {
+      "epoch": 1.9308363109367583,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.7779694448572605e-06,
+      "loss": 1.0207,
+      "step": 9738
+    },
+    {
+      "epoch": 1.9310362059918544,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.777025472142827e-06,
+      "loss": 0.9238,
+      "step": 9739
+    },
+    {
+      "epoch": 1.9312361010469503,
+      "grad_norm": 2.125,
+      "learning_rate": 2.7760815981726854e-06,
+      "loss": 0.9788,
+      "step": 9740
+    },
+    {
+      "epoch": 1.9314359961020464,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7751378229887586e-06,
+      "loss": 0.9759,
+      "step": 9741
+    },
+    {
+      "epoch": 1.9316358911571425,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.774194146632975e-06,
+      "loss": 0.9478,
+      "step": 9742
+    },
+    {
+      "epoch": 1.9318357862122386,
+      "grad_norm": 2.125,
+      "learning_rate": 2.773250569147249e-06,
+      "loss": 1.08,
+      "step": 9743
+    },
+    {
+      "epoch": 1.9320356812673345,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.772307090573494e-06,
+      "loss": 0.9411,
+      "step": 9744
+    },
+    {
+      "epoch": 1.9322355763224306,
+      "grad_norm": 2.25,
+      "learning_rate": 2.7713637109536207e-06,
+      "loss": 1.0289,
+      "step": 9745
+    },
+    {
+      "epoch": 1.9324354713775267,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.7704204303295348e-06,
+      "loss": 0.9375,
+      "step": 9746
+    },
+    {
+      "epoch": 1.9326353664326228,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.769477248743132e-06,
+      "loss": 1.0683,
+      "step": 9747
+    },
+    {
+      "epoch": 1.932835261487719,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.768534166236314e-06,
+      "loss": 0.8,
+      "step": 9748
+    },
+    {
+      "epoch": 1.933035156542815,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.7675911828509703e-06,
+      "loss": 0.9387,
+      "step": 9749
+    },
+    {
+      "epoch": 1.933235051597911,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.7666482986289876e-06,
+      "loss": 0.9022,
+      "step": 9750
+    },
+    {
+      "epoch": 1.9334349466530072,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.765705513612247e-06,
+      "loss": 0.9885,
+      "step": 9751
+    },
+    {
+      "epoch": 1.9336348417081033,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.7647628278426306e-06,
+      "loss": 0.9144,
+      "step": 9752
+    },
+    {
+      "epoch": 1.9338347367631994,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.7638202413620106e-06,
+      "loss": 1.0567,
+      "step": 9753
+    },
+    {
+      "epoch": 1.9340346318182955,
+      "grad_norm": 2.125,
+      "learning_rate": 2.7628777542122553e-06,
+      "loss": 0.941,
+      "step": 9754
+    },
+    {
+      "epoch": 1.9342345268733916,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.7619353664352326e-06,
+      "loss": 0.9615,
+      "step": 9755
+    },
+    {
+      "epoch": 1.9344344219284877,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.760993078072802e-06,
+      "loss": 0.9721,
+      "step": 9756
+    },
+    {
+      "epoch": 1.9346343169835836,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.760050889166818e-06,
+      "loss": 0.9166,
+      "step": 9757
+    },
+    {
+      "epoch": 1.9348342120386797,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7591087997591366e-06,
+      "loss": 0.9897,
+      "step": 9758
+    },
+    {
+      "epoch": 1.9350341070937758,
+      "grad_norm": 2.25,
+      "learning_rate": 2.7581668098916024e-06,
+      "loss": 1.0961,
+      "step": 9759
+    },
+    {
+      "epoch": 1.9352340021488719,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.7572249196060575e-06,
+      "loss": 0.9858,
+      "step": 9760
+    },
+    {
+      "epoch": 1.935433897203968,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.756283128944344e-06,
+      "loss": 1.0277,
+      "step": 9761
+    },
+    {
+      "epoch": 1.9356337922590638,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.7553414379482936e-06,
+      "loss": 1.0197,
+      "step": 9762
+    },
+    {
+      "epoch": 1.93583368731416,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.7543998466597357e-06,
+      "loss": 0.8993,
+      "step": 9763
+    },
+    {
+      "epoch": 1.936033582369256,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.753458355120498e-06,
+      "loss": 0.9828,
+      "step": 9764
+    },
+    {
+      "epoch": 1.9362334774243521,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7525169633724024e-06,
+      "loss": 0.9341,
+      "step": 9765
+    },
+    {
+      "epoch": 1.9364333724794482,
+      "grad_norm": 2.125,
+      "learning_rate": 2.7515756714572593e-06,
+      "loss": 1.0403,
+      "step": 9766
+    },
+    {
+      "epoch": 1.9366332675345443,
+      "grad_norm": 2.125,
+      "learning_rate": 2.750634479416887e-06,
+      "loss": 0.9437,
+      "step": 9767
+    },
+    {
+      "epoch": 1.9368331625896404,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.7496933872930907e-06,
+      "loss": 0.9842,
+      "step": 9768
+    },
+    {
+      "epoch": 1.9370330576447365,
+      "grad_norm": 2.25,
+      "learning_rate": 2.7487523951276716e-06,
+      "loss": 0.9598,
+      "step": 9769
+    },
+    {
+      "epoch": 1.9372329526998326,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.747811502962433e-06,
+      "loss": 1.0267,
+      "step": 9770
+    },
+    {
+      "epoch": 1.9374328477549287,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.7468707108391667e-06,
+      "loss": 0.9902,
+      "step": 9771
+    },
+    {
+      "epoch": 1.9376327428100248,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7459300187996614e-06,
+      "loss": 1.0448,
+      "step": 9772
+    },
+    {
+      "epoch": 1.937832637865121,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.7449894268857055e-06,
+      "loss": 0.9244,
+      "step": 9773
+    },
+    {
+      "epoch": 1.938032532920217,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7440489351390782e-06,
+      "loss": 1.0769,
+      "step": 9774
+    },
+    {
+      "epoch": 1.938232427975313,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.743108543601554e-06,
+      "loss": 0.9104,
+      "step": 9775
+    },
+    {
+      "epoch": 1.938432323030409,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.7421682523149097e-06,
+      "loss": 0.9787,
+      "step": 9776
+    },
+    {
+      "epoch": 1.9386322180855051,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.741228061320911e-06,
+      "loss": 0.9726,
+      "step": 9777
+    },
+    {
+      "epoch": 1.9388321131406012,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7402879706613176e-06,
+      "loss": 0.9924,
+      "step": 9778
+    },
+    {
+      "epoch": 1.939032008195697,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.7393479803778933e-06,
+      "loss": 1.009,
+      "step": 9779
+    },
+    {
+      "epoch": 1.9392319032507932,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.7384080905123912e-06,
+      "loss": 0.9448,
+      "step": 9780
+    },
+    {
+      "epoch": 1.9394317983058893,
+      "grad_norm": 2.125,
+      "learning_rate": 2.7374683011065594e-06,
+      "loss": 0.9685,
+      "step": 9781
+    },
+    {
+      "epoch": 1.9396316933609854,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.736528612202142e-06,
+      "loss": 1.0365,
+      "step": 9782
+    },
+    {
+      "epoch": 1.9398315884160815,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.7355890238408845e-06,
+      "loss": 1.0848,
+      "step": 9783
+    },
+    {
+      "epoch": 1.9400314834711776,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.73464953606452e-06,
+      "loss": 0.9731,
+      "step": 9784
+    },
+    {
+      "epoch": 1.9402313785262737,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7337101489147792e-06,
+      "loss": 1.0464,
+      "step": 9785
+    },
+    {
+      "epoch": 1.9404312735813698,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.7327708624333936e-06,
+      "loss": 1.0343,
+      "step": 9786
+    },
+    {
+      "epoch": 1.940631168636466,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.7318316766620845e-06,
+      "loss": 0.9942,
+      "step": 9787
+    },
+    {
+      "epoch": 1.940831063691562,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.7308925916425676e-06,
+      "loss": 0.9997,
+      "step": 9788
+    },
+    {
+      "epoch": 1.941030958746658,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7299536074165624e-06,
+      "loss": 0.9747,
+      "step": 9789
+    },
+    {
+      "epoch": 1.9412308538017542,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.729014724025775e-06,
+      "loss": 0.9891,
+      "step": 9790
+    },
+    {
+      "epoch": 1.9414307488568503,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7280759415119087e-06,
+      "loss": 0.961,
+      "step": 9791
+    },
+    {
+      "epoch": 1.9416306439119462,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.727137259916668e-06,
+      "loss": 1.0758,
+      "step": 9792
+    },
+    {
+      "epoch": 1.9418305389670423,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.7261986792817484e-06,
+      "loss": 1.0119,
+      "step": 9793
+    },
+    {
+      "epoch": 1.9420304340221384,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.725260199648838e-06,
+      "loss": 0.9422,
+      "step": 9794
+    },
+    {
+      "epoch": 1.9422303290772345,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7243218210596288e-06,
+      "loss": 0.9585,
+      "step": 9795
+    },
+    {
+      "epoch": 1.9424302241323306,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.7233835435558033e-06,
+      "loss": 0.9701,
+      "step": 9796
+    },
+    {
+      "epoch": 1.9426301191874265,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.722445367179034e-06,
+      "loss": 0.9562,
+      "step": 9797
+    },
+    {
+      "epoch": 1.9428300142425226,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.7215072919709996e-06,
+      "loss": 0.963,
+      "step": 9798
+    },
+    {
+      "epoch": 1.9430299092976187,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.720569317973368e-06,
+      "loss": 0.9333,
+      "step": 9799
+    },
+    {
+      "epoch": 1.9432298043527148,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.719631445227802e-06,
+      "loss": 0.9703,
+      "step": 9800
+    },
+    {
+      "epoch": 1.9434296994078109,
+      "grad_norm": 2.25,
+      "learning_rate": 2.718693673775966e-06,
+      "loss": 1.0323,
+      "step": 9801
+    },
+    {
+      "epoch": 1.943629594462907,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.7177560036595128e-06,
+      "loss": 0.9448,
+      "step": 9802
+    },
+    {
+      "epoch": 1.943829489518003,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.7168184349200926e-06,
+      "loss": 0.9027,
+      "step": 9803
+    },
+    {
+      "epoch": 1.9440293845730992,
+      "grad_norm": 2.375,
+      "learning_rate": 2.7158809675993556e-06,
+      "loss": 1.0262,
+      "step": 9804
+    },
+    {
+      "epoch": 1.9442292796281953,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.714943601738942e-06,
+      "loss": 0.9659,
+      "step": 9805
+    },
+    {
+      "epoch": 1.9444291746832914,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.714006337380487e-06,
+      "loss": 1.0365,
+      "step": 9806
+    },
+    {
+      "epoch": 1.9446290697383875,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.713069174565629e-06,
+      "loss": 1.0349,
+      "step": 9807
+    },
+    {
+      "epoch": 1.9448289647934836,
+      "grad_norm": 2.25,
+      "learning_rate": 2.712132113335994e-06,
+      "loss": 1.0042,
+      "step": 9808
+    },
+    {
+      "epoch": 1.9450288598485796,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.7111951537332058e-06,
+      "loss": 0.8987,
+      "step": 9809
+    },
+    {
+      "epoch": 1.9452287549036755,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.710258295798883e-06,
+      "loss": 0.9283,
+      "step": 9810
+    },
+    {
+      "epoch": 1.9454286499587716,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.709321539574644e-06,
+      "loss": 0.8604,
+      "step": 9811
+    },
+    {
+      "epoch": 1.9456285450138677,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.708384885102097e-06,
+      "loss": 1.0184,
+      "step": 9812
+    },
+    {
+      "epoch": 1.9458284400689638,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.7074483324228474e-06,
+      "loss": 0.8505,
+      "step": 9813
+    },
+    {
+      "epoch": 1.9460283351240597,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.7065118815785e-06,
+      "loss": 1.018,
+      "step": 9814
+    },
+    {
+      "epoch": 1.9462282301791558,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.705575532610649e-06,
+      "loss": 1.0354,
+      "step": 9815
+    },
+    {
+      "epoch": 1.946428125234252,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.704639285560886e-06,
+      "loss": 0.9605,
+      "step": 9816
+    },
+    {
+      "epoch": 1.946628020289348,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.7037031404708038e-06,
+      "loss": 1.0619,
+      "step": 9817
+    },
+    {
+      "epoch": 1.946827915344444,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.702767097381982e-06,
+      "loss": 1.083,
+      "step": 9818
+    },
+    {
+      "epoch": 1.9470278103995402,
+      "grad_norm": 2.125,
+      "learning_rate": 2.7018311563359977e-06,
+      "loss": 1.0053,
+      "step": 9819
+    },
+    {
+      "epoch": 1.9472277054546363,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.700895317374431e-06,
+      "loss": 0.9911,
+      "step": 9820
+    },
+    {
+      "epoch": 1.9474276005097324,
+      "grad_norm": 2.25,
+      "learning_rate": 2.699959580538849e-06,
+      "loss": 0.9673,
+      "step": 9821
+    },
+    {
+      "epoch": 1.9476274955648285,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.6990239458708145e-06,
+      "loss": 1.0125,
+      "step": 9822
+    },
+    {
+      "epoch": 1.9478273906199246,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.6980884134118925e-06,
+      "loss": 1.0734,
+      "step": 9823
+    },
+    {
+      "epoch": 1.9480272856750207,
+      "grad_norm": 2.125,
+      "learning_rate": 2.697152983203637e-06,
+      "loss": 0.89,
+      "step": 9824
+    },
+    {
+      "epoch": 1.9482271807301168,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.696217655287598e-06,
+      "loss": 0.8876,
+      "step": 9825
+    },
+    {
+      "epoch": 1.948427075785213,
+      "grad_norm": 2.0,
+      "learning_rate": 2.6952824297053272e-06,
+      "loss": 0.911,
+      "step": 9826
+    },
+    {
+      "epoch": 1.948626970840309,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.694347306498366e-06,
+      "loss": 0.8906,
+      "step": 9827
+    },
+    {
+      "epoch": 1.9488268658954049,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.6934122857082478e-06,
+      "loss": 0.9755,
+      "step": 9828
+    },
+    {
+      "epoch": 1.949026760950501,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.6924773673765114e-06,
+      "loss": 0.9884,
+      "step": 9829
+    },
+    {
+      "epoch": 1.949226656005597,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.6915425515446835e-06,
+      "loss": 1.0249,
+      "step": 9830
+    },
+    {
+      "epoch": 1.9494265510606932,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.6906078382542877e-06,
+      "loss": 1.0337,
+      "step": 9831
+    },
+    {
+      "epoch": 1.949626446115789,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.689673227546847e-06,
+      "loss": 1.0812,
+      "step": 9832
+    },
+    {
+      "epoch": 1.9498263411708852,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.6887387194638744e-06,
+      "loss": 1.0124,
+      "step": 9833
+    },
+    {
+      "epoch": 1.9500262362259813,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.687804314046879e-06,
+      "loss": 0.9958,
+      "step": 9834
+    },
+    {
+      "epoch": 1.9502261312810774,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.686870011337371e-06,
+      "loss": 1.0071,
+      "step": 9835
+    },
+    {
+      "epoch": 1.9504260263361735,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.6859358113768496e-06,
+      "loss": 0.8691,
+      "step": 9836
+    },
+    {
+      "epoch": 1.9506259213912696,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.6850017142068113e-06,
+      "loss": 1.082,
+      "step": 9837
+    },
+    {
+      "epoch": 1.9508258164463657,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.6840677198687515e-06,
+      "loss": 1.072,
+      "step": 9838
+    },
+    {
+      "epoch": 1.9510257115014618,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.683133828404155e-06,
+      "loss": 0.8452,
+      "step": 9839
+    },
+    {
+      "epoch": 1.9512256065565579,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.6822000398545078e-06,
+      "loss": 1.0922,
+      "step": 9840
+    },
+    {
+      "epoch": 1.951425501611654,
+      "grad_norm": 2.25,
+      "learning_rate": 2.681266354261285e-06,
+      "loss": 1.0021,
+      "step": 9841
+    },
+    {
+      "epoch": 1.95162539666675,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.6803327716659644e-06,
+      "loss": 0.9239,
+      "step": 9842
+    },
+    {
+      "epoch": 1.9518252917218462,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.6793992921100153e-06,
+      "loss": 0.9683,
+      "step": 9843
+    },
+    {
+      "epoch": 1.9520251867769423,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.678465915634899e-06,
+      "loss": 0.9021,
+      "step": 9844
+    },
+    {
+      "epoch": 1.9522250818320381,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.6775326422820813e-06,
+      "loss": 0.8585,
+      "step": 9845
+    },
+    {
+      "epoch": 1.9524249768871342,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.676599472093015e-06,
+      "loss": 0.924,
+      "step": 9846
+    },
+    {
+      "epoch": 1.9526248719422303,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.675666405109151e-06,
+      "loss": 0.8473,
+      "step": 9847
+    },
+    {
+      "epoch": 1.9528247669973264,
+      "grad_norm": 2.25,
+      "learning_rate": 2.6747334413719377e-06,
+      "loss": 1.0141,
+      "step": 9848
+    },
+    {
+      "epoch": 1.9530246620524225,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.6738005809228175e-06,
+      "loss": 0.95,
+      "step": 9849
+    },
+    {
+      "epoch": 1.9532245571075184,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.6728678238032245e-06,
+      "loss": 0.9748,
+      "step": 9850
+    },
+    {
+      "epoch": 1.9534244521626145,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.671935170054597e-06,
+      "loss": 1.0148,
+      "step": 9851
+    },
+    {
+      "epoch": 1.9536243472177106,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.6710026197183595e-06,
+      "loss": 1.0376,
+      "step": 9852
+    },
+    {
+      "epoch": 1.9538242422728067,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.670070172835936e-06,
+      "loss": 1.0051,
+      "step": 9853
+    },
+    {
+      "epoch": 1.9540241373279028,
+      "grad_norm": 2.125,
+      "learning_rate": 2.669137829448748e-06,
+      "loss": 0.9586,
+      "step": 9854
+    },
+    {
+      "epoch": 1.954224032382999,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.6682055895982085e-06,
+      "loss": 0.9388,
+      "step": 9855
+    },
+    {
+      "epoch": 1.954423927438095,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.667273453325726e-06,
+      "loss": 1.0055,
+      "step": 9856
+    },
+    {
+      "epoch": 1.954623822493191,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.6663414206727116e-06,
+      "loss": 0.9695,
+      "step": 9857
+    },
+    {
+      "epoch": 1.9548237175482872,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.66540949168056e-06,
+      "loss": 1.0922,
+      "step": 9858
+    },
+    {
+      "epoch": 1.9550236126033833,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.6644776663906674e-06,
+      "loss": 0.9678,
+      "step": 9859
+    },
+    {
+      "epoch": 1.9552235076584794,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.663545944844429e-06,
+      "loss": 1.1365,
+      "step": 9860
+    },
+    {
+      "epoch": 1.9554234027135755,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.6626143270832313e-06,
+      "loss": 0.9636,
+      "step": 9861
+    },
+    {
+      "epoch": 1.9556232977686716,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.6616828131484528e-06,
+      "loss": 0.9908,
+      "step": 9862
+    },
+    {
+      "epoch": 1.9558231928237675,
+      "grad_norm": 2.125,
+      "learning_rate": 2.6607514030814757e-06,
+      "loss": 0.9,
+      "step": 9863
+    },
+    {
+      "epoch": 1.9560230878788636,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.659820096923672e-06,
+      "loss": 0.9088,
+      "step": 9864
+    },
+    {
+      "epoch": 1.9562229829339597,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.658888894716407e-06,
+      "loss": 0.9174,
+      "step": 9865
+    },
+    {
+      "epoch": 1.9564228779890558,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.65795779650105e-06,
+      "loss": 1.1063,
+      "step": 9866
+    },
+    {
+      "epoch": 1.9566227730441517,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.657026802318957e-06,
+      "loss": 0.9581,
+      "step": 9867
+    },
+    {
+      "epoch": 1.9568226680992478,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.6560959122114815e-06,
+      "loss": 1.0477,
+      "step": 9868
+    },
+    {
+      "epoch": 1.9570225631543439,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.6551651262199773e-06,
+      "loss": 0.9407,
+      "step": 9869
+    },
+    {
+      "epoch": 1.95722245820944,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.6542344443857874e-06,
+      "loss": 0.9956,
+      "step": 9870
+    },
+    {
+      "epoch": 1.957422353264536,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.653303866750253e-06,
+      "loss": 0.9238,
+      "step": 9871
+    },
+    {
+      "epoch": 1.9576222483196322,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.652373393354709e-06,
+      "loss": 0.932,
+      "step": 9872
+    },
+    {
+      "epoch": 1.9578221433747283,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.651443024240489e-06,
+      "loss": 1.0184,
+      "step": 9873
+    },
+    {
+      "epoch": 1.9580220384298244,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.65051275944892e-06,
+      "loss": 0.9246,
+      "step": 9874
+    },
+    {
+      "epoch": 1.9582219334849205,
+      "grad_norm": 2.25,
+      "learning_rate": 2.6495825990213208e-06,
+      "loss": 0.9082,
+      "step": 9875
+    },
+    {
+      "epoch": 1.9584218285400166,
+      "grad_norm": 2.125,
+      "learning_rate": 2.6486525429990133e-06,
+      "loss": 0.9802,
+      "step": 9876
+    },
+    {
+      "epoch": 1.9586217235951127,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.647722591423309e-06,
+      "loss": 0.9473,
+      "step": 9877
+    },
+    {
+      "epoch": 1.9588216186502088,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.646792744335514e-06,
+      "loss": 0.9576,
+      "step": 9878
+    },
+    {
+      "epoch": 1.9590215137053049,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.645863001776936e-06,
+      "loss": 0.9337,
+      "step": 9879
+    },
+    {
+      "epoch": 1.9592214087604007,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.6449333637888717e-06,
+      "loss": 0.9523,
+      "step": 9880
+    },
+    {
+      "epoch": 1.9594213038154968,
+      "grad_norm": 2.125,
+      "learning_rate": 2.644003830412614e-06,
+      "loss": 0.95,
+      "step": 9881
+    },
+    {
+      "epoch": 1.959621198870593,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.643074401689457e-06,
+      "loss": 0.9245,
+      "step": 9882
+    },
+    {
+      "epoch": 1.959821093925689,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.6421450776606827e-06,
+      "loss": 0.9675,
+      "step": 9883
+    },
+    {
+      "epoch": 1.9600209889807851,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.6412158583675707e-06,
+      "loss": 1.0271,
+      "step": 9884
+    },
+    {
+      "epoch": 1.960220884035881,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.6402867438514e-06,
+      "loss": 0.9307,
+      "step": 9885
+    },
+    {
+      "epoch": 1.9604207790909771,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.63935773415344e-06,
+      "loss": 1.0636,
+      "step": 9886
+    },
+    {
+      "epoch": 1.9606206741460732,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.6384288293149572e-06,
+      "loss": 0.9815,
+      "step": 9887
+    },
+    {
+      "epoch": 1.9608205692011693,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.6375000293772144e-06,
+      "loss": 1.0368,
+      "step": 9888
+    },
+    {
+      "epoch": 1.9610204642562654,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.636571334381467e-06,
+      "loss": 0.9559,
+      "step": 9889
+    },
+    {
+      "epoch": 1.9612203593113615,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.635642744368967e-06,
+      "loss": 0.9255,
+      "step": 9890
+    },
+    {
+      "epoch": 1.9614202543664576,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.634714259380966e-06,
+      "loss": 0.9569,
+      "step": 9891
+    },
+    {
+      "epoch": 1.9616201494215537,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.6337858794587046e-06,
+      "loss": 0.8442,
+      "step": 9892
+    },
+    {
+      "epoch": 1.9618200444766498,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.63285760464342e-06,
+      "loss": 0.9336,
+      "step": 9893
+    },
+    {
+      "epoch": 1.962019939531746,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.6319294349763495e-06,
+      "loss": 0.9557,
+      "step": 9894
+    },
+    {
+      "epoch": 1.962219834586842,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.6310013704987207e-06,
+      "loss": 0.956,
+      "step": 9895
+    },
+    {
+      "epoch": 1.9624197296419381,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.6300734112517562e-06,
+      "loss": 0.9001,
+      "step": 9896
+    },
+    {
+      "epoch": 1.9626196246970342,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.6291455572766794e-06,
+      "loss": 0.8764,
+      "step": 9897
+    },
+    {
+      "epoch": 1.96281951975213,
+      "grad_norm": 2.125,
+      "learning_rate": 2.6282178086147036e-06,
+      "loss": 0.9871,
+      "step": 9898
+    },
+    {
+      "epoch": 1.9630194148072262,
+      "grad_norm": 2.25,
+      "learning_rate": 2.6272901653070397e-06,
+      "loss": 1.0325,
+      "step": 9899
+    },
+    {
+      "epoch": 1.9632193098623223,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.626362627394892e-06,
+      "loss": 0.9382,
+      "step": 9900
+    },
+    {
+      "epoch": 1.9634192049174184,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.6254351949194634e-06,
+      "loss": 1.0445,
+      "step": 9901
+    },
+    {
+      "epoch": 1.9636190999725143,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.6245078679219503e-06,
+      "loss": 0.9785,
+      "step": 9902
+    },
+    {
+      "epoch": 1.9638189950276104,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.6235806464435425e-06,
+      "loss": 1.0077,
+      "step": 9903
+    },
+    {
+      "epoch": 1.9640188900827065,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.6226535305254303e-06,
+      "loss": 0.9668,
+      "step": 9904
+    },
+    {
+      "epoch": 1.9642187851378026,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.6217265202087944e-06,
+      "loss": 0.9459,
+      "step": 9905
+    },
+    {
+      "epoch": 1.9644186801928987,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.62079961553481e-06,
+      "loss": 0.9683,
+      "step": 9906
+    },
+    {
+      "epoch": 1.9646185752479948,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.619872816544655e-06,
+      "loss": 0.909,
+      "step": 9907
+    },
+    {
+      "epoch": 1.9648184703030909,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.6189461232794956e-06,
+      "loss": 0.882,
+      "step": 9908
+    },
+    {
+      "epoch": 1.965018365358187,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.6180195357804926e-06,
+      "loss": 1.0062,
+      "step": 9909
+    },
+    {
+      "epoch": 1.965218260413283,
+      "grad_norm": 2.125,
+      "learning_rate": 2.6170930540888096e-06,
+      "loss": 0.9332,
+      "step": 9910
+    },
+    {
+      "epoch": 1.9654181554683792,
+      "grad_norm": 2.125,
+      "learning_rate": 2.6161666782455986e-06,
+      "loss": 0.9635,
+      "step": 9911
+    },
+    {
+      "epoch": 1.9656180505234753,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.615240408292007e-06,
+      "loss": 0.9407,
+      "step": 9912
+    },
+    {
+      "epoch": 1.9658179455785714,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.614314244269184e-06,
+      "loss": 1.0038,
+      "step": 9913
+    },
+    {
+      "epoch": 1.9660178406336675,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.6133881862182676e-06,
+      "loss": 0.9732,
+      "step": 9914
+    },
+    {
+      "epoch": 1.9662177356887633,
+      "grad_norm": 2.25,
+      "learning_rate": 2.612462234180391e-06,
+      "loss": 1.0319,
+      "step": 9915
+    },
+    {
+      "epoch": 1.9664176307438594,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.611536388196688e-06,
+      "loss": 0.9009,
+      "step": 9916
+    },
+    {
+      "epoch": 1.9666175257989555,
+      "grad_norm": 2.125,
+      "learning_rate": 2.610610648308285e-06,
+      "loss": 1.0454,
+      "step": 9917
+    },
+    {
+      "epoch": 1.9668174208540516,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.6096850145563014e-06,
+      "loss": 0.9378,
+      "step": 9918
+    },
+    {
+      "epoch": 1.9670173159091477,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.608759486981853e-06,
+      "loss": 0.9732,
+      "step": 9919
+    },
+    {
+      "epoch": 1.9672172109642436,
+      "grad_norm": 2.25,
+      "learning_rate": 2.6078340656260535e-06,
+      "loss": 1.0111,
+      "step": 9920
+    },
+    {
+      "epoch": 1.9674171060193397,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.606908750530008e-06,
+      "loss": 1.0066,
+      "step": 9921
+    },
+    {
+      "epoch": 1.9676170010744358,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.605983541734822e-06,
+      "loss": 0.9409,
+      "step": 9922
+    },
+    {
+      "epoch": 1.967816896129532,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.605058439281591e-06,
+      "loss": 0.9418,
+      "step": 9923
+    },
+    {
+      "epoch": 1.968016791184628,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.6041334432114064e-06,
+      "loss": 0.9253,
+      "step": 9924
+    },
+    {
+      "epoch": 1.9682166862397241,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.6032085535653605e-06,
+      "loss": 0.9287,
+      "step": 9925
+    },
+    {
+      "epoch": 1.9684165812948202,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.6022837703845346e-06,
+      "loss": 0.9976,
+      "step": 9926
+    },
+    {
+      "epoch": 1.9686164763499163,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.6013590937100054e-06,
+      "loss": 0.9212,
+      "step": 9927
+    },
+    {
+      "epoch": 1.9688163714050124,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.600434523582851e-06,
+      "loss": 0.9482,
+      "step": 9928
+    },
+    {
+      "epoch": 1.9690162664601085,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.5995100600441392e-06,
+      "loss": 0.9942,
+      "step": 9929
+    },
+    {
+      "epoch": 1.9692161615152046,
+      "grad_norm": 2.25,
+      "learning_rate": 2.598585703134934e-06,
+      "loss": 0.9624,
+      "step": 9930
+    },
+    {
+      "epoch": 1.9694160565703007,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.597661452896293e-06,
+      "loss": 0.9544,
+      "step": 9931
+    },
+    {
+      "epoch": 1.9696159516253968,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.596737309369276e-06,
+      "loss": 1.0619,
+      "step": 9932
+    },
+    {
+      "epoch": 1.9698158466804927,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.5958132725949314e-06,
+      "loss": 1.0108,
+      "step": 9933
+    },
+    {
+      "epoch": 1.9700157417355888,
+      "grad_norm": 2.25,
+      "learning_rate": 2.5948893426143018e-06,
+      "loss": 1.0122,
+      "step": 9934
+    },
+    {
+      "epoch": 1.970215636790685,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.5939655194684334e-06,
+      "loss": 0.9055,
+      "step": 9935
+    },
+    {
+      "epoch": 1.970415531845781,
+      "grad_norm": 2.125,
+      "learning_rate": 2.593041803198359e-06,
+      "loss": 1.0203,
+      "step": 9936
+    },
+    {
+      "epoch": 1.9706154269008769,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.592118193845109e-06,
+      "loss": 1.0048,
+      "step": 9937
+    },
+    {
+      "epoch": 1.970815321955973,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.5911946914497133e-06,
+      "loss": 1.0231,
+      "step": 9938
+    },
+    {
+      "epoch": 1.971015217011069,
+      "grad_norm": 2.125,
+      "learning_rate": 2.590271296053193e-06,
+      "loss": 0.9282,
+      "step": 9939
+    },
+    {
+      "epoch": 1.9712151120661652,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.5893480076965615e-06,
+      "loss": 0.9355,
+      "step": 9940
+    },
+    {
+      "epoch": 1.9714150071212613,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.588424826420836e-06,
+      "loss": 1.033,
+      "step": 9941
+    },
+    {
+      "epoch": 1.9716149021763574,
+      "grad_norm": 2.25,
+      "learning_rate": 2.5875017522670227e-06,
+      "loss": 0.9113,
+      "step": 9942
+    },
+    {
+      "epoch": 1.9718147972314535,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.5865787852761217e-06,
+      "loss": 1.0371,
+      "step": 9943
+    },
+    {
+      "epoch": 1.9720146922865496,
+      "grad_norm": 2.125,
+      "learning_rate": 2.585655925489135e-06,
+      "loss": 0.9781,
+      "step": 9944
+    },
+    {
+      "epoch": 1.9722145873416457,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.584733172947055e-06,
+      "loss": 0.9384,
+      "step": 9945
+    },
+    {
+      "epoch": 1.9724144823967418,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.5838105276908667e-06,
+      "loss": 1.029,
+      "step": 9946
+    },
+    {
+      "epoch": 1.9726143774518379,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.5828879897615587e-06,
+      "loss": 1.0109,
+      "step": 9947
+    },
+    {
+      "epoch": 1.972814272506934,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.581965559200108e-06,
+      "loss": 0.9886,
+      "step": 9948
+    },
+    {
+      "epoch": 1.97301416756203,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.58104323604749e-06,
+      "loss": 0.9924,
+      "step": 9949
+    },
+    {
+      "epoch": 1.9732140626171262,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.5801210203446718e-06,
+      "loss": 0.9618,
+      "step": 9950
+    },
+    {
+      "epoch": 1.973413957672222,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.57919891213262e-06,
+      "loss": 0.9703,
+      "step": 9951
+    },
+    {
+      "epoch": 1.9736138527273182,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.578276911452292e-06,
+      "loss": 0.9618,
+      "step": 9952
+    },
+    {
+      "epoch": 1.9738137477824143,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.5773550183446465e-06,
+      "loss": 0.9932,
+      "step": 9953
+    },
+    {
+      "epoch": 1.9740136428375104,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.5764332328506327e-06,
+      "loss": 0.9966,
+      "step": 9954
+    },
+    {
+      "epoch": 1.9742135378926062,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.5755115550111942e-06,
+      "loss": 0.986,
+      "step": 9955
+    },
+    {
+      "epoch": 1.9744134329477023,
+      "grad_norm": 2.125,
+      "learning_rate": 2.574589984867275e-06,
+      "loss": 0.9848,
+      "step": 9956
+    },
+    {
+      "epoch": 1.9746133280027984,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.5736685224598097e-06,
+      "loss": 1.038,
+      "step": 9957
+    },
+    {
+      "epoch": 1.9748132230578945,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.5727471678297277e-06,
+      "loss": 0.9692,
+      "step": 9958
+    },
+    {
+      "epoch": 1.9750131181129906,
+      "grad_norm": 2.125,
+      "learning_rate": 2.5718259210179588e-06,
+      "loss": 0.9844,
+      "step": 9959
+    },
+    {
+      "epoch": 1.9752130131680867,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.5709047820654236e-06,
+      "loss": 0.9823,
+      "step": 9960
+    },
+    {
+      "epoch": 1.9754129082231828,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.569983751013039e-06,
+      "loss": 0.9745,
+      "step": 9961
+    },
+    {
+      "epoch": 1.975612803278279,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.5690628279017136e-06,
+      "loss": 0.921,
+      "step": 9962
+    },
+    {
+      "epoch": 1.975812698333375,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.56814201277236e-06,
+      "loss": 0.9791,
+      "step": 9963
+    },
+    {
+      "epoch": 1.9760125933884711,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.567221305665879e-06,
+      "loss": 0.8148,
+      "step": 9964
+    },
+    {
+      "epoch": 1.9762124884435672,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.566300706623165e-06,
+      "loss": 0.9585,
+      "step": 9965
+    },
+    {
+      "epoch": 1.9764123834986633,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.5653802156851158e-06,
+      "loss": 0.9921,
+      "step": 9966
+    },
+    {
+      "epoch": 1.9766122785537594,
+      "grad_norm": 2.109375,
+      "learning_rate": 2.5644598328926183e-06,
+      "loss": 0.8986,
+      "step": 9967
+    },
+    {
+      "epoch": 1.9768121736088553,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.563539558286552e-06,
+      "loss": 0.9104,
+      "step": 9968
+    },
+    {
+      "epoch": 1.9770120686639514,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.5626193919078008e-06,
+      "loss": 0.9292,
+      "step": 9969
+    },
+    {
+      "epoch": 1.9772119637190475,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.561699333797236e-06,
+      "loss": 0.9468,
+      "step": 9970
+    },
+    {
+      "epoch": 1.9774118587741436,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.560779383995724e-06,
+      "loss": 1.0181,
+      "step": 9971
+    },
+    {
+      "epoch": 1.9776117538292397,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.559859542544133e-06,
+      "loss": 0.9923,
+      "step": 9972
+    },
+    {
+      "epoch": 1.9778116488843356,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.5589398094833205e-06,
+      "loss": 0.9749,
+      "step": 9973
+    },
+    {
+      "epoch": 1.9780115439394317,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.55802018485414e-06,
+      "loss": 0.9581,
+      "step": 9974
+    },
+    {
+      "epoch": 1.9782114389945278,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.557100668697443e-06,
+      "loss": 0.9778,
+      "step": 9975
+    },
+    {
+      "epoch": 1.9784113340496239,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.5561812610540736e-06,
+      "loss": 0.9616,
+      "step": 9976
+    },
+    {
+      "epoch": 1.97861122910472,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.555261961964872e-06,
+      "loss": 0.9669,
+      "step": 9977
+    },
+    {
+      "epoch": 1.978811124159816,
+      "grad_norm": 2.0,
+      "learning_rate": 2.5543427714706705e-06,
+      "loss": 0.8204,
+      "step": 9978
+    },
+    {
+      "epoch": 1.9790110192149122,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.5534236896123043e-06,
+      "loss": 1.0263,
+      "step": 9979
+    },
+    {
+      "epoch": 1.9792109142700083,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.552504716430596e-06,
+      "loss": 0.9726,
+      "step": 9980
+    },
+    {
+      "epoch": 1.9794108093251044,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.551585851966367e-06,
+      "loss": 0.9656,
+      "step": 9981
+    },
+    {
+      "epoch": 1.9796107043802005,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.550667096260433e-06,
+      "loss": 0.9558,
+      "step": 9982
+    },
+    {
+      "epoch": 1.9798105994352966,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.549748449353603e-06,
+      "loss": 0.9942,
+      "step": 9983
+    },
+    {
+      "epoch": 1.9800104944903927,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.548829911286687e-06,
+      "loss": 0.9188,
+      "step": 9984
+    },
+    {
+      "epoch": 1.9802103895454888,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.5479114821004845e-06,
+      "loss": 1.0964,
+      "step": 9985
+    },
+    {
+      "epoch": 1.9804102846005847,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.5469931618357907e-06,
+      "loss": 1.0178,
+      "step": 9986
+    },
+    {
+      "epoch": 1.9806101796556808,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.5460749505334004e-06,
+      "loss": 0.9592,
+      "step": 9987
+    },
+    {
+      "epoch": 1.9808100747107769,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.5451568482340983e-06,
+      "loss": 0.9825,
+      "step": 9988
+    },
+    {
+      "epoch": 1.981009969765873,
+      "grad_norm": 2.25,
+      "learning_rate": 2.5442388549786668e-06,
+      "loss": 0.9168,
+      "step": 9989
+    },
+    {
+      "epoch": 1.9812098648209688,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.543320970807882e-06,
+      "loss": 0.8362,
+      "step": 9990
+    },
+    {
+      "epoch": 1.981409759876065,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.5424031957625184e-06,
+      "loss": 1.0724,
+      "step": 9991
+    },
+    {
+      "epoch": 1.981609654931161,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.5414855298833423e-06,
+      "loss": 0.9505,
+      "step": 9992
+    },
+    {
+      "epoch": 1.9818095499862571,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.540567973211115e-06,
+      "loss": 1.0578,
+      "step": 9993
+    },
+    {
+      "epoch": 1.9820094450413532,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.539650525786597e-06,
+      "loss": 0.9175,
+      "step": 9994
+    },
+    {
+      "epoch": 1.9822093400964493,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.5387331876505405e-06,
+      "loss": 0.952,
+      "step": 9995
+    },
+    {
+      "epoch": 1.9824092351515454,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.5378159588436907e-06,
+      "loss": 1.0021,
+      "step": 9996
+    },
+    {
+      "epoch": 1.9826091302066415,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.536898839406795e-06,
+      "loss": 0.9979,
+      "step": 9997
+    },
+    {
+      "epoch": 1.9828090252617376,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.5359818293805893e-06,
+      "loss": 0.9755,
+      "step": 9998
+    },
+    {
+      "epoch": 1.9830089203168337,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.5350649288058065e-06,
+      "loss": 0.9423,
+      "step": 9999
+    },
+    {
+      "epoch": 1.9832088153719298,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.534148137723178e-06,
+      "loss": 0.9666,
+      "step": 10000
+    },
+    {
+      "epoch": 1.983408710427026,
+      "grad_norm": 2.046875,
+      "learning_rate": 2.5332314561734257e-06,
+      "loss": 0.823,
+      "step": 10001
+    },
+    {
+      "epoch": 1.983608605482122,
+      "grad_norm": 2.125,
+      "learning_rate": 2.532314884197267e-06,
+      "loss": 1.0064,
+      "step": 10002
+    },
+    {
+      "epoch": 1.983808500537218,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.5313984218354185e-06,
+      "loss": 1.0481,
+      "step": 10003
+    },
+    {
+      "epoch": 1.984008395592314,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.530482069128589e-06,
+      "loss": 0.9209,
+      "step": 10004
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 15006,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 5002,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.952236882409082e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}