diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10962 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.99712,
+  "eval_steps": 500,
+  "global_step": 1560,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00192,
+      "grad_norm": 3.843557834625244,
+      "learning_rate": 6.41025641025641e-08,
+      "loss": 0.5493,
+      "step": 1
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 3.8249223232269287,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 0.5359,
+      "step": 2
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 4.095667362213135,
+      "learning_rate": 1.9230769230769234e-07,
+      "loss": 0.5124,
+      "step": 3
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 3.6196255683898926,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 0.5201,
+      "step": 4
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 3.9502737522125244,
+      "learning_rate": 3.205128205128205e-07,
+      "loss": 0.5266,
+      "step": 5
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 3.8070971965789795,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 0.5315,
+      "step": 6
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 3.7366929054260254,
+      "learning_rate": 4.4871794871794876e-07,
+      "loss": 0.5782,
+      "step": 7
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 3.698350191116333,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 0.4984,
+      "step": 8
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 3.7893757820129395,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 0.5174,
+      "step": 9
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 3.5876681804656982,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 0.5407,
+      "step": 10
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 3.4170141220092773,
+      "learning_rate": 7.051282051282052e-07,
+      "loss": 0.5321,
+      "step": 11
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 3.2406179904937744,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.4483,
+      "step": 12
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 2.8420934677124023,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.5007,
+      "step": 13
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 2.869420051574707,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 0.5853,
+      "step": 14
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 2.390106439590454,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 0.3934,
+      "step": 15
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 2.772052764892578,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.4438,
+      "step": 16
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 2.3655014038085938,
+      "learning_rate": 1.0897435897435899e-06,
+      "loss": 0.4139,
+      "step": 17
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 2.1296281814575195,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 0.4922,
+      "step": 18
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 1.986196756362915,
+      "learning_rate": 1.217948717948718e-06,
+      "loss": 0.393,
+      "step": 19
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 2.0222110748291016,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.4256,
+      "step": 20
+    },
+    {
+      "epoch": 0.04032,
+      "grad_norm": 1.9421510696411133,
+      "learning_rate": 1.3461538461538462e-06,
+      "loss": 0.455,
+      "step": 21
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 1.9265691041946411,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 0.4876,
+      "step": 22
+    },
+    {
+      "epoch": 0.04416,
+      "grad_norm": 1.8183995485305786,
+      "learning_rate": 1.4743589743589745e-06,
+      "loss": 0.4725,
+      "step": 23
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 1.2524487972259521,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.4131,
+      "step": 24
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 1.238390564918518,
+      "learning_rate": 1.602564102564103e-06,
+      "loss": 0.4049,
+      "step": 25
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 1.2981340885162354,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.4275,
+      "step": 26
+    },
+    {
+      "epoch": 0.05184,
+      "grad_norm": 1.2695064544677734,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 0.4186,
+      "step": 27
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 1.1388022899627686,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 0.4181,
+      "step": 28
+    },
+    {
+      "epoch": 0.05568,
+      "grad_norm": 1.1645935773849487,
+      "learning_rate": 1.8589743589743592e-06,
+      "loss": 0.5068,
+      "step": 29
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.9908599257469177,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.4608,
+      "step": 30
+    },
+    {
+      "epoch": 0.05952,
+      "grad_norm": 0.8310354351997375,
+      "learning_rate": 1.987179487179487e-06,
+      "loss": 0.4053,
+      "step": 31
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 0.8930312395095825,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.455,
+      "step": 32
+    },
+    {
+      "epoch": 0.06336,
+      "grad_norm": 0.7913976907730103,
+      "learning_rate": 2.1153846153846155e-06,
+      "loss": 0.3993,
+      "step": 33
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 1.016400694847107,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 0.4102,
+      "step": 34
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 0.9332112073898315,
+      "learning_rate": 2.243589743589744e-06,
+      "loss": 0.4073,
+      "step": 35
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.9067287445068359,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.3914,
+      "step": 36
+    },
+    {
+      "epoch": 0.07104,
+      "grad_norm": 0.8736160397529602,
+      "learning_rate": 2.371794871794872e-06,
+      "loss": 0.4304,
+      "step": 37
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.9036316275596619,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 0.3819,
+      "step": 38
+    },
+    {
+      "epoch": 0.07488,
+      "grad_norm": 0.8593645095825195,
+      "learning_rate": 2.5e-06,
+      "loss": 0.3576,
+      "step": 39
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.8034077286720276,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.3762,
+      "step": 40
+    },
+    {
+      "epoch": 0.07872,
+      "grad_norm": 0.7795902490615845,
+      "learning_rate": 2.6282051282051286e-06,
+      "loss": 0.3585,
+      "step": 41
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.8267181515693665,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.3879,
+      "step": 42
+    },
+    {
+      "epoch": 0.08256,
+      "grad_norm": 0.7951707243919373,
+      "learning_rate": 2.756410256410257e-06,
+      "loss": 0.349,
+      "step": 43
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.8047124147415161,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 0.4085,
+      "step": 44
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.6982182860374451,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 0.3565,
+      "step": 45
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.70152348279953,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 0.3814,
+      "step": 46
+    },
+    {
+      "epoch": 0.09024,
+      "grad_norm": 0.6901643872261047,
+      "learning_rate": 3.012820512820513e-06,
+      "loss": 0.3918,
+      "step": 47
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.794810950756073,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.3505,
+      "step": 48
+    },
+    {
+      "epoch": 0.09408,
+      "grad_norm": 0.8339862823486328,
+      "learning_rate": 3.141025641025641e-06,
+      "loss": 0.4587,
+      "step": 49
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.7975308299064636,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.3878,
+      "step": 50
+    },
+    {
+      "epoch": 0.09792,
+      "grad_norm": 0.7511988282203674,
+      "learning_rate": 3.2692307692307696e-06,
+      "loss": 0.4149,
+      "step": 51
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.6813432574272156,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.3882,
+      "step": 52
+    },
+    {
+      "epoch": 0.10176,
+      "grad_norm": 0.7014743089675903,
+      "learning_rate": 3.397435897435898e-06,
+      "loss": 0.3934,
+      "step": 53
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 0.6991923451423645,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.3919,
+      "step": 54
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.6644059419631958,
+      "learning_rate": 3.5256410256410263e-06,
+      "loss": 0.3899,
+      "step": 55
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.6621732711791992,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 0.4202,
+      "step": 56
+    },
+    {
+      "epoch": 0.10944,
+      "grad_norm": 0.7253834009170532,
+      "learning_rate": 3.653846153846154e-06,
+      "loss": 0.4503,
+      "step": 57
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 0.7145866751670837,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 0.3917,
+      "step": 58
+    },
+    {
+      "epoch": 0.11328,
+      "grad_norm": 0.7308309674263,
+      "learning_rate": 3.782051282051282e-06,
+      "loss": 0.3865,
+      "step": 59
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.6961401700973511,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.4044,
+      "step": 60
+    },
+    {
+      "epoch": 0.11712,
+      "grad_norm": 0.7254306674003601,
+      "learning_rate": 3.910256410256411e-06,
+      "loss": 0.3848,
+      "step": 61
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.6466768383979797,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 0.4134,
+      "step": 62
+    },
+    {
+      "epoch": 0.12096,
+      "grad_norm": 0.7102042436599731,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 0.3736,
+      "step": 63
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.7980540990829468,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.3907,
+      "step": 64
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.687211275100708,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.4155,
+      "step": 65
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.7526315450668335,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.4125,
+      "step": 66
+    },
+    {
+      "epoch": 0.12864,
+      "grad_norm": 0.7422536015510559,
+      "learning_rate": 4.294871794871795e-06,
+      "loss": 0.3752,
+      "step": 67
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.6434139609336853,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 0.406,
+      "step": 68
+    },
+    {
+      "epoch": 0.13248,
+      "grad_norm": 0.6825657486915588,
+      "learning_rate": 4.423076923076924e-06,
+      "loss": 0.4277,
+      "step": 69
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.6319999098777771,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.3533,
+      "step": 70
+    },
+    {
+      "epoch": 0.13632,
+      "grad_norm": 0.7375587821006775,
+      "learning_rate": 4.551282051282052e-06,
+      "loss": 0.4062,
+      "step": 71
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.6587281823158264,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.3478,
+      "step": 72
+    },
+    {
+      "epoch": 0.14016,
+      "grad_norm": 0.6490545868873596,
+      "learning_rate": 4.6794871794871795e-06,
+      "loss": 0.3493,
+      "step": 73
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.6771235466003418,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 0.3708,
+      "step": 74
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.7523772120475769,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 0.4064,
+      "step": 75
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.6678730845451355,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 0.3823,
+      "step": 76
+    },
+    {
+      "epoch": 0.14784,
+      "grad_norm": 0.6507325172424316,
+      "learning_rate": 4.935897435897436e-06,
+      "loss": 0.4104,
+      "step": 77
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.6481208205223083,
+      "learning_rate": 5e-06,
+      "loss": 0.3838,
+      "step": 78
+    },
+    {
+      "epoch": 0.15168,
+      "grad_norm": 0.6811144351959229,
+      "learning_rate": 5.064102564102565e-06,
+      "loss": 0.3381,
+      "step": 79
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.7742949724197388,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.3962,
+      "step": 80
+    },
+    {
+      "epoch": 0.15552,
+      "grad_norm": 0.7335490584373474,
+      "learning_rate": 5.192307692307693e-06,
+      "loss": 0.3483,
+      "step": 81
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.6449044346809387,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 0.3535,
+      "step": 82
+    },
+    {
+      "epoch": 0.15936,
+      "grad_norm": 0.7107799053192139,
+      "learning_rate": 5.320512820512821e-06,
+      "loss": 0.3955,
+      "step": 83
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.7558763027191162,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.3937,
+      "step": 84
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.7412011027336121,
+      "learning_rate": 5.448717948717949e-06,
+      "loss": 0.3356,
+      "step": 85
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.7510801553726196,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 0.4137,
+      "step": 86
+    },
+    {
+      "epoch": 0.16704,
+      "grad_norm": 0.6146009564399719,
+      "learning_rate": 5.576923076923077e-06,
+      "loss": 0.3712,
+      "step": 87
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.7673497796058655,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 0.3757,
+      "step": 88
+    },
+    {
+      "epoch": 0.17088,
+      "grad_norm": 0.8233842849731445,
+      "learning_rate": 5.705128205128206e-06,
+      "loss": 0.3856,
+      "step": 89
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.7806938290596008,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 0.3546,
+      "step": 90
+    },
+    {
+      "epoch": 0.17472,
+      "grad_norm": 0.6763064861297607,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.3301,
+      "step": 91
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.6932587027549744,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 0.3856,
+      "step": 92
+    },
+    {
+      "epoch": 0.17856,
+      "grad_norm": 0.7267228364944458,
+      "learning_rate": 5.961538461538462e-06,
+      "loss": 0.3684,
+      "step": 93
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.6245903372764587,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 0.2943,
+      "step": 94
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.7414659261703491,
+      "learning_rate": 6.08974358974359e-06,
+      "loss": 0.3202,
+      "step": 95
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.7013316750526428,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.3953,
+      "step": 96
+    },
+    {
+      "epoch": 0.18624,
+      "grad_norm": 0.6257365942001343,
+      "learning_rate": 6.217948717948718e-06,
+      "loss": 0.3055,
+      "step": 97
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.7943385243415833,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.4025,
+      "step": 98
+    },
+    {
+      "epoch": 0.19008,
+      "grad_norm": 0.7125914692878723,
+      "learning_rate": 6.3461538461538466e-06,
+      "loss": 0.3681,
+      "step": 99
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.6895804405212402,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 0.3443,
+      "step": 100
+    },
+    {
+      "epoch": 0.19392,
+      "grad_norm": 0.61220383644104,
+      "learning_rate": 6.474358974358975e-06,
+      "loss": 0.3197,
+      "step": 101
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.6729249954223633,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.3228,
+      "step": 102
+    },
+    {
+      "epoch": 0.19776,
+      "grad_norm": 0.8359938859939575,
+      "learning_rate": 6.602564102564103e-06,
+      "loss": 0.4075,
+      "step": 103
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.7006269693374634,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3847,
+      "step": 104
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.7480886578559875,
+      "learning_rate": 6.730769230769232e-06,
+      "loss": 0.3406,
+      "step": 105
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.6175745129585266,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 0.338,
+      "step": 106
+    },
+    {
+      "epoch": 0.20544,
+      "grad_norm": 0.82331383228302,
+      "learning_rate": 6.858974358974359e-06,
+      "loss": 0.3888,
+      "step": 107
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.7225168347358704,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.3949,
+      "step": 108
+    },
+    {
+      "epoch": 0.20928,
+      "grad_norm": 0.6784982085227966,
+      "learning_rate": 6.9871794871794876e-06,
+      "loss": 0.3561,
+      "step": 109
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.640651285648346,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 0.3353,
+      "step": 110
+    },
+    {
+      "epoch": 0.21312,
+      "grad_norm": 0.671169102191925,
+      "learning_rate": 7.115384615384616e-06,
+      "loss": 0.3533,
+      "step": 111
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.7499369978904724,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.3955,
+      "step": 112
+    },
+    {
+      "epoch": 0.21696,
+      "grad_norm": 0.7039608359336853,
+      "learning_rate": 7.243589743589744e-06,
+      "loss": 0.3824,
+      "step": 113
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.7076845765113831,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.4012,
+      "step": 114
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.6565355062484741,
+      "learning_rate": 7.371794871794873e-06,
+      "loss": 0.3737,
+      "step": 115
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.8483031988143921,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 0.4697,
+      "step": 116
+    },
+    {
+      "epoch": 0.22464,
+      "grad_norm": 0.7934540510177612,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3952,
+      "step": 117
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 0.6828005313873291,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.4034,
+      "step": 118
+    },
+    {
+      "epoch": 0.22848,
+      "grad_norm": 0.606593132019043,
+      "learning_rate": 7.6282051282051286e-06,
+      "loss": 0.3438,
+      "step": 119
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.7434725761413574,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.3902,
+      "step": 120
+    },
+    {
+      "epoch": 0.23232,
+      "grad_norm": 0.6935045123100281,
+      "learning_rate": 7.756410256410258e-06,
+      "loss": 0.3516,
+      "step": 121
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.7259051203727722,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 0.4017,
+      "step": 122
+    },
+    {
+      "epoch": 0.23616,
+      "grad_norm": 0.6900324821472168,
+      "learning_rate": 7.884615384615384e-06,
+      "loss": 0.3661,
+      "step": 123
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.6937414407730103,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 0.3761,
+      "step": 124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6391271352767944,
+      "learning_rate": 8.012820512820515e-06,
+      "loss": 0.371,
+      "step": 125
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.6735659837722778,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.3582,
+      "step": 126
+    },
+    {
+      "epoch": 0.24384,
+      "grad_norm": 0.6381909847259521,
+      "learning_rate": 8.141025641025641e-06,
+      "loss": 0.345,
+      "step": 127
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.7072357535362244,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.3742,
+      "step": 128
+    },
+    {
+      "epoch": 0.24768,
+      "grad_norm": 0.6645024418830872,
+      "learning_rate": 8.26923076923077e-06,
+      "loss": 0.3421,
+      "step": 129
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.7187925577163696,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.3806,
+      "step": 130
+    },
+    {
+      "epoch": 0.25152,
+      "grad_norm": 0.6878098845481873,
+      "learning_rate": 8.397435897435898e-06,
+      "loss": 0.3748,
+      "step": 131
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.9609063267707825,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.4747,
+      "step": 132
+    },
+    {
+      "epoch": 0.25536,
+      "grad_norm": 0.7777335047721863,
+      "learning_rate": 8.525641025641026e-06,
+      "loss": 0.3586,
+      "step": 133
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.6334943175315857,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 0.4203,
+      "step": 134
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 0.9341431856155396,
+      "learning_rate": 8.653846153846155e-06,
+      "loss": 0.3596,
+      "step": 135
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.8867289423942566,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 0.51,
+      "step": 136
+    },
+    {
+      "epoch": 0.26304,
+      "grad_norm": 0.7397791147232056,
+      "learning_rate": 8.782051282051283e-06,
+      "loss": 0.3638,
+      "step": 137
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.6910991072654724,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.3991,
+      "step": 138
+    },
+    {
+      "epoch": 0.26688,
+      "grad_norm": 0.6767971515655518,
+      "learning_rate": 8.910256410256411e-06,
+      "loss": 0.3069,
+      "step": 139
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.7195672988891602,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.3706,
+      "step": 140
+    },
+    {
+      "epoch": 0.27072,
+      "grad_norm": 0.6783812046051025,
+      "learning_rate": 9.03846153846154e-06,
+      "loss": 0.3833,
+      "step": 141
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.6150461435317993,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 0.3355,
+      "step": 142
+    },
+    {
+      "epoch": 0.27456,
+      "grad_norm": 0.6504448652267456,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.3444,
+      "step": 143
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.7956386208534241,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.392,
+      "step": 144
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 1.0326496362686157,
+      "learning_rate": 9.294871794871796e-06,
+      "loss": 0.4964,
+      "step": 145
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.6709503531455994,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 0.4025,
+      "step": 146
+    },
+    {
+      "epoch": 0.28224,
+      "grad_norm": 0.7416048049926758,
+      "learning_rate": 9.423076923076923e-06,
+      "loss": 0.3976,
+      "step": 147
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.7250438332557678,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 0.3805,
+      "step": 148
+    },
+    {
+      "epoch": 0.28608,
+      "grad_norm": 0.6921824812889099,
+      "learning_rate": 9.551282051282053e-06,
+      "loss": 0.376,
+      "step": 149
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.6684398055076599,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.4057,
+      "step": 150
+    },
+    {
+      "epoch": 0.28992,
+      "grad_norm": 0.6863765120506287,
+      "learning_rate": 9.67948717948718e-06,
+      "loss": 0.384,
+      "step": 151
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 0.7271242141723633,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 0.3663,
+      "step": 152
+    },
+    {
+      "epoch": 0.29376,
+      "grad_norm": 0.7491096258163452,
+      "learning_rate": 9.807692307692308e-06,
+      "loss": 0.4624,
+      "step": 153
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.6301567554473877,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.355,
+      "step": 154
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.6892555952072144,
+      "learning_rate": 9.935897435897437e-06,
+      "loss": 0.3462,
+      "step": 155
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.673816978931427,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 156
+    },
+    {
+      "epoch": 0.30144,
+      "grad_norm": 0.6928176283836365,
+      "learning_rate": 9.999987482852865e-06,
+      "loss": 0.3357,
+      "step": 157
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.640011191368103,
+      "learning_rate": 9.99994993147413e-06,
+      "loss": 0.3705,
+      "step": 158
+    },
+    {
+      "epoch": 0.30528,
+      "grad_norm": 0.6202666759490967,
+      "learning_rate": 9.99988734605181e-06,
+      "loss": 0.3738,
+      "step": 159
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.6704613566398621,
+      "learning_rate": 9.999799726899261e-06,
+      "loss": 0.4117,
+      "step": 160
+    },
+    {
+      "epoch": 0.30912,
+      "grad_norm": 0.7389768362045288,
+      "learning_rate": 9.999687074455182e-06,
+      "loss": 0.3629,
+      "step": 161
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.734327495098114,
+      "learning_rate": 9.999549389283605e-06,
+      "loss": 0.3889,
+      "step": 162
+    },
+    {
+      "epoch": 0.31296,
+      "grad_norm": 0.7862688302993774,
+      "learning_rate": 9.999386672073902e-06,
+      "loss": 0.4319,
+      "step": 163
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.5473729968070984,
+      "learning_rate": 9.999198923640774e-06,
+      "loss": 0.3416,
+      "step": 164
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.6196820139884949,
+      "learning_rate": 9.998986144924253e-06,
+      "loss": 0.3828,
+      "step": 165
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.6770342588424683,
+      "learning_rate": 9.998748336989687e-06,
+      "loss": 0.3372,
+      "step": 166
+    },
+    {
+      "epoch": 0.32064,
+      "grad_norm": 0.6752730011940002,
+      "learning_rate": 9.998485501027752e-06,
+      "loss": 0.4141,
+      "step": 167
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.6911773681640625,
+      "learning_rate": 9.998197638354428e-06,
+      "loss": 0.4076,
+      "step": 168
+    },
+    {
+      "epoch": 0.32448,
+      "grad_norm": 0.6446501612663269,
+      "learning_rate": 9.997884750411004e-06,
+      "loss": 0.3889,
+      "step": 169
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.6675588488578796,
+      "learning_rate": 9.997546838764066e-06,
+      "loss": 0.3734,
+      "step": 170
+    },
+    {
+      "epoch": 0.32832,
+      "grad_norm": 0.6143994927406311,
+      "learning_rate": 9.997183905105487e-06,
+      "loss": 0.3454,
+      "step": 171
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.6693524122238159,
+      "learning_rate": 9.996795951252427e-06,
+      "loss": 0.3476,
+      "step": 172
+    },
+    {
+      "epoch": 0.33216,
+      "grad_norm": 0.6360855102539062,
+      "learning_rate": 9.996382979147316e-06,
+      "loss": 0.4322,
+      "step": 173
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.6526750922203064,
+      "learning_rate": 9.995944990857848e-06,
+      "loss": 0.3639,
+      "step": 174
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.6066708564758301,
+      "learning_rate": 9.995481988576968e-06,
+      "loss": 0.3756,
+      "step": 175
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.6885200142860413,
+      "learning_rate": 9.994993974622863e-06,
+      "loss": 0.3779,
+      "step": 176
+    },
+    {
+      "epoch": 0.33984,
+      "grad_norm": 0.6344889402389526,
+      "learning_rate": 9.994480951438948e-06,
+      "loss": 0.3693,
+      "step": 177
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.6831358075141907,
+      "learning_rate": 9.993942921593858e-06,
+      "loss": 0.4075,
+      "step": 178
+    },
+    {
+      "epoch": 0.34368,
+      "grad_norm": 0.6006773710250854,
+      "learning_rate": 9.993379887781436e-06,
+      "loss": 0.3784,
+      "step": 179
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.7562153339385986,
+      "learning_rate": 9.992791852820709e-06,
+      "loss": 0.4057,
+      "step": 180
+    },
+    {
+      "epoch": 0.34752,
+      "grad_norm": 0.6040899157524109,
+      "learning_rate": 9.992178819655889e-06,
+      "loss": 0.3276,
+      "step": 181
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 0.639474630355835,
+      "learning_rate": 9.991540791356342e-06,
+      "loss": 0.3374,
+      "step": 182
+    },
+    {
+      "epoch": 0.35136,
+      "grad_norm": 0.5994912981987,
+      "learning_rate": 9.990877771116588e-06,
+      "loss": 0.3753,
+      "step": 183
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.6587606072425842,
+      "learning_rate": 9.990189762256275e-06,
+      "loss": 0.3478,
+      "step": 184
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 0.5631051659584045,
+      "learning_rate": 9.98947676822017e-06,
+      "loss": 0.3078,
+      "step": 185
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.7119491100311279,
+      "learning_rate": 9.988738792578126e-06,
+      "loss": 0.3838,
+      "step": 186
+    },
+    {
+      "epoch": 0.35904,
+      "grad_norm": 0.6617155075073242,
+      "learning_rate": 9.987975839025091e-06,
+      "loss": 0.434,
+      "step": 187
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.6335970759391785,
+      "learning_rate": 9.987187911381059e-06,
+      "loss": 0.3555,
+      "step": 188
+    },
+    {
+      "epoch": 0.36288,
+      "grad_norm": 0.7126662135124207,
+      "learning_rate": 9.986375013591077e-06,
+      "loss": 0.4495,
+      "step": 189
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.6956565976142883,
+      "learning_rate": 9.985537149725207e-06,
+      "loss": 0.3805,
+      "step": 190
+    },
+    {
+      "epoch": 0.36672,
+      "grad_norm": 0.6889484524726868,
+      "learning_rate": 9.984674323978517e-06,
+      "loss": 0.4188,
+      "step": 191
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.5620262622833252,
+      "learning_rate": 9.983786540671052e-06,
+      "loss": 0.3306,
+      "step": 192
+    },
+    {
+      "epoch": 0.37056,
+      "grad_norm": 0.7724553942680359,
+      "learning_rate": 9.982873804247817e-06,
+      "loss": 0.4316,
+      "step": 193
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.6637314558029175,
+      "learning_rate": 9.981936119278758e-06,
+      "loss": 0.3696,
+      "step": 194
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 0.5833960771560669,
+      "learning_rate": 9.980973490458728e-06,
+      "loss": 0.3152,
+      "step": 195
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.6502295732498169,
+      "learning_rate": 9.979985922607476e-06,
+      "loss": 0.4064,
+      "step": 196
+    },
+    {
+      "epoch": 0.37824,
+      "grad_norm": 0.6160441040992737,
+      "learning_rate": 9.978973420669613e-06,
+      "loss": 0.3626,
+      "step": 197
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.573761522769928,
+      "learning_rate": 9.977935989714594e-06,
+      "loss": 0.3498,
+      "step": 198
+    },
+    {
+      "epoch": 0.38208,
+      "grad_norm": 0.6001765727996826,
+      "learning_rate": 9.97687363493669e-06,
+      "loss": 0.3433,
+      "step": 199
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.6546545624732971,
+      "learning_rate": 9.975786361654959e-06,
+      "loss": 0.4168,
+      "step": 200
+    },
+    {
+      "epoch": 0.38592,
+      "grad_norm": 0.6866360902786255,
+      "learning_rate": 9.974674175313228e-06,
+      "loss": 0.3813,
+      "step": 201
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.6718149185180664,
+      "learning_rate": 9.973537081480056e-06,
+      "loss": 0.383,
+      "step": 202
+    },
+    {
+      "epoch": 0.38976,
+      "grad_norm": 0.6946866512298584,
+      "learning_rate": 9.972375085848712e-06,
+      "loss": 0.4439,
+      "step": 203
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 0.5859667658805847,
+      "learning_rate": 9.971188194237141e-06,
+      "loss": 0.3816,
+      "step": 204
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.6172149181365967,
+      "learning_rate": 9.969976412587943e-06,
+      "loss": 0.3457,
+      "step": 205
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.6011773943901062,
+      "learning_rate": 9.96873974696834e-06,
+      "loss": 0.364,
+      "step": 206
+    },
+    {
+      "epoch": 0.39744,
+      "grad_norm": 0.6524738669395447,
+      "learning_rate": 9.967478203570143e-06,
+      "loss": 0.4336,
+      "step": 207
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.5141505002975464,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.3289,
+      "step": 208
+    },
+    {
+      "epoch": 0.40128,
+      "grad_norm": 0.6093929409980774,
+      "learning_rate": 9.96488050882796e-06,
+      "loss": 0.4399,
+      "step": 209
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.5254409313201904,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.3315,
+      "step": 210
+    },
+    {
+      "epoch": 0.40512,
+      "grad_norm": 0.6045888066291809,
+      "learning_rate": 9.9621833803865e-06,
+      "loss": 0.3384,
+      "step": 211
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.6646643877029419,
+      "learning_rate": 9.960797545330936e-06,
+      "loss": 0.4014,
+      "step": 212
+    },
+    {
+      "epoch": 0.40896,
+      "grad_norm": 0.6472617387771606,
+      "learning_rate": 9.95938687226226e-06,
+      "loss": 0.4723,
+      "step": 213
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.5927615165710449,
+      "learning_rate": 9.95795136824351e-06,
+      "loss": 0.4281,
+      "step": 214
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 0.561004102230072,
+      "learning_rate": 9.956491040462051e-06,
+      "loss": 0.3892,
+      "step": 215
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.6211714148521423,
+      "learning_rate": 9.955005896229543e-06,
+      "loss": 0.4637,
+      "step": 216
+    },
+    {
+      "epoch": 0.41664,
+      "grad_norm": 0.5818098783493042,
+      "learning_rate": 9.95349594298189e-06,
+      "loss": 0.3726,
+      "step": 217
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.6425796747207642,
+      "learning_rate": 9.951961188279216e-06,
+      "loss": 0.4246,
+      "step": 218
+    },
+    {
+      "epoch": 0.42048,
+      "grad_norm": 0.5746365189552307,
+      "learning_rate": 9.950401639805822e-06,
+      "loss": 0.3605,
+      "step": 219
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.5899927616119385,
+      "learning_rate": 9.948817305370145e-06,
+      "loss": 0.4032,
+      "step": 220
+    },
+    {
+      "epoch": 0.42432,
+      "grad_norm": 0.5515298247337341,
+      "learning_rate": 9.947208192904722e-06,
+      "loss": 0.3182,
+      "step": 221
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.5837955474853516,
+      "learning_rate": 9.94557431046616e-06,
+      "loss": 0.3706,
+      "step": 222
+    },
+    {
+      "epoch": 0.42816,
+      "grad_norm": 0.6058487296104431,
+      "learning_rate": 9.943915666235068e-06,
+      "loss": 0.4278,
+      "step": 223
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.5727962851524353,
+      "learning_rate": 9.942232268516051e-06,
+      "loss": 0.4167,
+      "step": 224
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 0.5703360438346863,
+      "learning_rate": 9.940524125737641e-06,
+      "loss": 0.3763,
+      "step": 225
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.5099168419837952,
+      "learning_rate": 9.938791246452267e-06,
+      "loss": 0.3436,
+      "step": 226
+    },
+    {
+      "epoch": 0.43584,
+      "grad_norm": 0.6010491251945496,
+      "learning_rate": 9.937033639336212e-06,
+      "loss": 0.3805,
+      "step": 227
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.5972442626953125,
+      "learning_rate": 9.935251313189564e-06,
+      "loss": 0.3772,
+      "step": 228
+    },
+    {
+      "epoch": 0.43968,
+      "grad_norm": 1.3225898742675781,
+      "learning_rate": 9.933444276936185e-06,
+      "loss": 0.5889,
+      "step": 229
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.5816615223884583,
+      "learning_rate": 9.931612539623643e-06,
+      "loss": 0.3584,
+      "step": 230
+    },
+    {
+      "epoch": 0.44352,
+      "grad_norm": 0.6918455958366394,
+      "learning_rate": 9.929756110423195e-06,
+      "loss": 0.3739,
+      "step": 231
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.6478462219238281,
+      "learning_rate": 9.927874998629714e-06,
+      "loss": 0.4055,
+      "step": 232
+    },
+    {
+      "epoch": 0.44736,
+      "grad_norm": 0.6581472158432007,
+      "learning_rate": 9.925969213661664e-06,
+      "loss": 0.3732,
+      "step": 233
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.6024413704872131,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.3459,
+      "step": 234
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 0.5743972659111023,
+      "learning_rate": 9.92208366249333e-06,
+      "loss": 0.369,
+      "step": 235
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.6526931524276733,
+      "learning_rate": 9.920103915747452e-06,
+      "loss": 0.3854,
+      "step": 236
+    },
+    {
+      "epoch": 0.45504,
+      "grad_norm": 0.5958468317985535,
+      "learning_rate": 9.91809953473572e-06,
+      "loss": 0.3451,
+      "step": 237
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.565719723701477,
+      "learning_rate": 9.916070529493785e-06,
+      "loss": 0.3381,
+      "step": 238
+    },
+    {
+      "epoch": 0.45888,
+      "grad_norm": 0.5998052954673767,
+      "learning_rate": 9.914016910180593e-06,
+      "loss": 0.3743,
+      "step": 239
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.609347403049469,
+      "learning_rate": 9.911938687078324e-06,
+      "loss": 0.3299,
+      "step": 240
+    },
+    {
+      "epoch": 0.46272,
+      "grad_norm": 0.7088524699211121,
+      "learning_rate": 9.90983587059235e-06,
+      "loss": 0.3504,
+      "step": 241
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 0.5974552631378174,
+      "learning_rate": 9.907708471251173e-06,
+      "loss": 0.3755,
+      "step": 242
+    },
+    {
+      "epoch": 0.46656,
+      "grad_norm": 0.6118410229682922,
+      "learning_rate": 9.905556499706384e-06,
+      "loss": 0.4054,
+      "step": 243
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.5863246917724609,
+      "learning_rate": 9.9033799667326e-06,
+      "loss": 0.3475,
+      "step": 244
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.6001723408699036,
+      "learning_rate": 9.901178883227413e-06,
+      "loss": 0.4128,
+      "step": 245
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 0.686601459980011,
+      "learning_rate": 9.89895326021134e-06,
+      "loss": 0.3957,
+      "step": 246
+    },
+    {
+      "epoch": 0.47424,
+      "grad_norm": 0.6601870656013489,
+      "learning_rate": 9.896703108827758e-06,
+      "loss": 0.4228,
+      "step": 247
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.5805381536483765,
+      "learning_rate": 9.89442844034286e-06,
+      "loss": 0.362,
+      "step": 248
+    },
+    {
+      "epoch": 0.47808,
+      "grad_norm": 0.542709231376648,
+      "learning_rate": 9.89212926614559e-06,
+      "loss": 0.3671,
+      "step": 249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5759779214859009,
+      "learning_rate": 9.889805597747588e-06,
+      "loss": 0.3974,
+      "step": 250
+    },
+    {
+      "epoch": 0.48192,
+      "grad_norm": 0.5364634394645691,
+      "learning_rate": 9.887457446783133e-06,
+      "loss": 0.3442,
+      "step": 251
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.5775154829025269,
+      "learning_rate": 9.885084825009085e-06,
+      "loss": 0.4141,
+      "step": 252
+    },
+    {
+      "epoch": 0.48576,
+      "grad_norm": 0.5697202682495117,
+      "learning_rate": 9.88268774430483e-06,
+      "loss": 0.3812,
+      "step": 253
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.6131813526153564,
+      "learning_rate": 9.88026621667221e-06,
+      "loss": 0.4309,
+      "step": 254
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.5484776496887207,
+      "learning_rate": 9.87782025423547e-06,
+      "loss": 0.3897,
+      "step": 255
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.5887479186058044,
+      "learning_rate": 9.875349869241202e-06,
+      "loss": 0.3883,
+      "step": 256
+    },
+    {
+      "epoch": 0.49344,
+      "grad_norm": 0.6489760875701904,
+      "learning_rate": 9.872855074058274e-06,
+      "loss": 0.3998,
+      "step": 257
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.6656110286712646,
+      "learning_rate": 9.870335881177774e-06,
+      "loss": 0.3887,
+      "step": 258
+    },
+    {
+      "epoch": 0.49728,
+      "grad_norm": 0.5303746461868286,
+      "learning_rate": 9.867792303212944e-06,
+      "loss": 0.336,
+      "step": 259
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.5473644137382507,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.3696,
+      "step": 260
+    },
+    {
+      "epoch": 0.50112,
+      "grad_norm": 4.5563836097717285,
+      "learning_rate": 9.862632043093667e-06,
+      "loss": 0.4792,
+      "step": 261
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.598250150680542,
+      "learning_rate": 9.860015386775915e-06,
+      "loss": 0.398,
+      "step": 262
+    },
+    {
+      "epoch": 0.50496,
+      "grad_norm": 0.5635238885879517,
+      "learning_rate": 9.857374397047093e-06,
+      "loss": 0.3413,
+      "step": 263
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.6472920179367065,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.3925,
+      "step": 264
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.5144029259681702,
+      "learning_rate": 9.852019470370254e-06,
+      "loss": 0.3816,
+      "step": 265
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.5724732279777527,
+      "learning_rate": 9.849305560233598e-06,
+      "loss": 0.3822,
+      "step": 266
+    },
+    {
+      "epoch": 0.51264,
+      "grad_norm": 0.5839284658432007,
+      "learning_rate": 9.846567370308462e-06,
+      "loss": 0.45,
+      "step": 267
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.5788620710372925,
+      "learning_rate": 9.843804914304578e-06,
+      "loss": 0.3872,
+      "step": 268
+    },
+    {
+      "epoch": 0.51648,
+      "grad_norm": 0.629429280757904,
+      "learning_rate": 9.841018206053167e-06,
+      "loss": 0.3882,
+      "step": 269
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.5664721131324768,
+      "learning_rate": 9.838207259506891e-06,
+      "loss": 0.402,
+      "step": 270
+    },
+    {
+      "epoch": 0.52032,
+      "grad_norm": 0.6412708163261414,
+      "learning_rate": 9.835372088739758e-06,
+      "loss": 0.4031,
+      "step": 271
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.5259047150611877,
+      "learning_rate": 9.83251270794707e-06,
+      "loss": 0.3165,
+      "step": 272
+    },
+    {
+      "epoch": 0.52416,
+      "grad_norm": 0.5224035978317261,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.3923,
+      "step": 273
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.5709006786346436,
+      "learning_rate": 9.826721373672235e-06,
+      "loss": 0.4312,
+      "step": 274
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.5672641396522522,
+      "learning_rate": 9.82378944918648e-06,
+      "loss": 0.4073,
+      "step": 275
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.5978093147277832,
+      "learning_rate": 9.820833372667813e-06,
+      "loss": 0.4063,
+      "step": 276
+    },
+    {
+      "epoch": 0.53184,
+      "grad_norm": 0.623744010925293,
+      "learning_rate": 9.817853158916889e-06,
+      "loss": 0.4243,
+      "step": 277
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.5311877131462097,
+      "learning_rate": 9.814848822855216e-06,
+      "loss": 0.3838,
+      "step": 278
+    },
+    {
+      "epoch": 0.53568,
+      "grad_norm": 0.5687721967697144,
+      "learning_rate": 9.811820379525085e-06,
+      "loss": 0.3591,
+      "step": 279
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.5889049172401428,
+      "learning_rate": 9.80876784408948e-06,
+      "loss": 0.3793,
+      "step": 280
+    },
+    {
+      "epoch": 0.53952,
+      "grad_norm": 0.5540534257888794,
+      "learning_rate": 9.805691231832018e-06,
+      "loss": 0.3321,
+      "step": 281
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 0.5731890797615051,
+      "learning_rate": 9.802590558156863e-06,
+      "loss": 0.3836,
+      "step": 282
+    },
+    {
+      "epoch": 0.54336,
+      "grad_norm": 0.5543009638786316,
+      "learning_rate": 9.799465838588646e-06,
+      "loss": 0.3241,
+      "step": 283
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.5468571782112122,
+      "learning_rate": 9.796317088772402e-06,
+      "loss": 0.3454,
+      "step": 284
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.6438685655593872,
+      "learning_rate": 9.793144324473473e-06,
+      "loss": 0.3702,
+      "step": 285
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.6208586096763611,
+      "learning_rate": 9.789947561577445e-06,
+      "loss": 0.4264,
+      "step": 286
+    },
+    {
+      "epoch": 0.55104,
+      "grad_norm": 0.6653385162353516,
+      "learning_rate": 9.786726816090058e-06,
+      "loss": 0.3619,
+      "step": 287
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.5749282836914062,
+      "learning_rate": 9.783482104137127e-06,
+      "loss": 0.3716,
+      "step": 288
+    },
+    {
+      "epoch": 0.55488,
+      "grad_norm": 0.5787516236305237,
+      "learning_rate": 9.78021344196447e-06,
+      "loss": 0.3619,
+      "step": 289
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.6237759590148926,
+      "learning_rate": 9.776920845937816e-06,
+      "loss": 0.3504,
+      "step": 290
+    },
+    {
+      "epoch": 0.55872,
+      "grad_norm": 0.5218870639801025,
+      "learning_rate": 9.77360433254273e-06,
+      "loss": 0.3575,
+      "step": 291
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.5251814723014832,
+      "learning_rate": 9.770263918384523e-06,
+      "loss": 0.3255,
+      "step": 292
+    },
+    {
+      "epoch": 0.56256,
+      "grad_norm": 0.5608870983123779,
+      "learning_rate": 9.766899620188181e-06,
+      "loss": 0.3696,
+      "step": 293
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.5645449757575989,
+      "learning_rate": 9.763511454798268e-06,
+      "loss": 0.3148,
+      "step": 294
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.6137270927429199,
+      "learning_rate": 9.760099439178852e-06,
+      "loss": 0.4024,
+      "step": 295
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.5873740315437317,
+      "learning_rate": 9.75666359041341e-06,
+      "loss": 0.3515,
+      "step": 296
+    },
+    {
+      "epoch": 0.57024,
+      "grad_norm": 0.6058859825134277,
+      "learning_rate": 9.753203925704756e-06,
+      "loss": 0.3752,
+      "step": 297
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.6152317523956299,
+      "learning_rate": 9.749720462374939e-06,
+      "loss": 0.3364,
+      "step": 298
+    },
+    {
+      "epoch": 0.57408,
+      "grad_norm": 0.5717517137527466,
+      "learning_rate": 9.74621321786517e-06,
+      "loss": 0.3891,
+      "step": 299
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.5540747046470642,
+      "learning_rate": 9.742682209735727e-06,
+      "loss": 0.3786,
+      "step": 300
+    },
+    {
+      "epoch": 0.57792,
+      "grad_norm": 0.5597113966941833,
+      "learning_rate": 9.739127455665872e-06,
+      "loss": 0.3938,
+      "step": 301
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.5814911127090454,
+      "learning_rate": 9.735548973453753e-06,
+      "loss": 0.3741,
+      "step": 302
+    },
+    {
+      "epoch": 0.58176,
+      "grad_norm": 0.576946496963501,
+      "learning_rate": 9.731946781016328e-06,
+      "loss": 0.3514,
+      "step": 303
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.5773646831512451,
+      "learning_rate": 9.728320896389263e-06,
+      "loss": 0.4221,
+      "step": 304
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.5906984210014343,
+      "learning_rate": 9.724671337726855e-06,
+      "loss": 0.4323,
+      "step": 305
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.5712162256240845,
+      "learning_rate": 9.720998123301924e-06,
+      "loss": 0.44,
+      "step": 306
+    },
+    {
+      "epoch": 0.58944,
+      "grad_norm": 0.5968261957168579,
+      "learning_rate": 9.717301271505739e-06,
+      "loss": 0.4329,
+      "step": 307
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.5751365423202515,
+      "learning_rate": 9.713580800847917e-06,
+      "loss": 0.3793,
+      "step": 308
+    },
+    {
+      "epoch": 0.59328,
+      "grad_norm": 0.5586972832679749,
+      "learning_rate": 9.709836729956326e-06,
+      "loss": 0.3416,
+      "step": 309
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.538479208946228,
+      "learning_rate": 9.706069077577e-06,
+      "loss": 0.3438,
+      "step": 310
+    },
+    {
+      "epoch": 0.59712,
+      "grad_norm": 0.5495683550834656,
+      "learning_rate": 9.702277862574048e-06,
+      "loss": 0.3778,
+      "step": 311
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.5954602360725403,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.4919,
+      "step": 312
+    },
+    {
+      "epoch": 0.60096,
+      "grad_norm": 0.535027265548706,
+      "learning_rate": 9.694624820743446e-06,
+      "loss": 0.3572,
+      "step": 313
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.6595924496650696,
+      "learning_rate": 9.690763032233498e-06,
+      "loss": 0.3612,
+      "step": 314
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.6070604920387268,
+      "learning_rate": 9.686877757735126e-06,
+      "loss": 0.4081,
+      "step": 315
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.5064239501953125,
+      "learning_rate": 9.682969016701357e-06,
+      "loss": 0.3175,
+      "step": 316
+    },
+    {
+      "epoch": 0.60864,
+      "grad_norm": 0.6430267691612244,
+      "learning_rate": 9.679036828702704e-06,
+      "loss": 0.3512,
+      "step": 317
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.5985868573188782,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.3535,
+      "step": 318
+    },
+    {
+      "epoch": 0.61248,
+      "grad_norm": 0.5778400301933289,
+      "learning_rate": 9.671102190679678e-06,
+      "loss": 0.3733,
+      "step": 319
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.5217545628547668,
+      "learning_rate": 9.66709978038292e-06,
+      "loss": 0.3156,
+      "step": 320
+    },
+    {
+      "epoch": 0.61632,
+      "grad_norm": 0.5111682415008545,
+      "learning_rate": 9.663074002576303e-06,
+      "loss": 0.3844,
+      "step": 321
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.5493319630622864,
+      "learning_rate": 9.659024877416328e-06,
+      "loss": 0.3659,
+      "step": 322
+    },
+    {
+      "epoch": 0.62016,
+      "grad_norm": 0.6036080121994019,
+      "learning_rate": 9.654952425176392e-06,
+      "loss": 0.3785,
+      "step": 323
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.5907517075538635,
+      "learning_rate": 9.650856666246693e-06,
+      "loss": 0.2969,
+      "step": 324
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.6222851872444153,
+      "learning_rate": 9.646737621134112e-06,
+      "loss": 0.405,
+      "step": 325
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.56783527135849,
+      "learning_rate": 9.642595310462133e-06,
+      "loss": 0.382,
+      "step": 326
+    },
+    {
+      "epoch": 0.62784,
+      "grad_norm": 0.5570917129516602,
+      "learning_rate": 9.638429754970715e-06,
+      "loss": 0.314,
+      "step": 327
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.692135214805603,
+      "learning_rate": 9.63424097551621e-06,
+      "loss": 0.3856,
+      "step": 328
+    },
+    {
+      "epoch": 0.63168,
+      "grad_norm": 0.581884503364563,
+      "learning_rate": 9.630028993071244e-06,
+      "loss": 0.4276,
+      "step": 329
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.5919851660728455,
+      "learning_rate": 9.62579382872462e-06,
+      "loss": 0.4114,
+      "step": 330
+    },
+    {
+      "epoch": 0.63552,
+      "grad_norm": 0.6581014394760132,
+      "learning_rate": 9.621535503681205e-06,
+      "loss": 0.3903,
+      "step": 331
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.6459546089172363,
+      "learning_rate": 9.617254039261835e-06,
+      "loss": 0.3729,
+      "step": 332
+    },
+    {
+      "epoch": 0.63936,
+      "grad_norm": 0.5892713069915771,
+      "learning_rate": 9.612949456903195e-06,
+      "loss": 0.4528,
+      "step": 333
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.5532912015914917,
+      "learning_rate": 9.608621778157722e-06,
+      "loss": 0.3773,
+      "step": 334
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.6035757064819336,
+      "learning_rate": 9.604271024693495e-06,
+      "loss": 0.376,
+      "step": 335
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.5886684060096741,
+      "learning_rate": 9.599897218294122e-06,
+      "loss": 0.3589,
+      "step": 336
+    },
+    {
+      "epoch": 0.64704,
+      "grad_norm": 0.6452056765556335,
+      "learning_rate": 9.595500380858632e-06,
+      "loss": 0.3944,
+      "step": 337
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.5505960583686829,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 0.3872,
+      "step": 338
+    },
+    {
+      "epoch": 0.65088,
+      "grad_norm": 0.5651102662086487,
+      "learning_rate": 9.586637701051886e-06,
+      "loss": 0.4554,
+      "step": 339
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.5067898035049438,
+      "learning_rate": 9.582171903054815e-06,
+      "loss": 0.3402,
+      "step": 340
+    },
+    {
+      "epoch": 0.65472,
+      "grad_norm": 0.6149987578392029,
+      "learning_rate": 9.577683162769781e-06,
+      "loss": 0.3641,
+      "step": 341
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.5432674884796143,
+      "learning_rate": 9.573171502671273e-06,
+      "loss": 0.3481,
+      "step": 342
+    },
+    {
+      "epoch": 0.65856,
+      "grad_norm": 0.5097435712814331,
+      "learning_rate": 9.568636945348534e-06,
+      "loss": 0.3718,
+      "step": 343
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.5743522644042969,
+      "learning_rate": 9.564079513505455e-06,
+      "loss": 0.3845,
+      "step": 344
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.5071309804916382,
+      "learning_rate": 9.55949922996045e-06,
+      "loss": 0.2939,
+      "step": 345
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.5429442524909973,
+      "learning_rate": 9.554896117646357e-06,
+      "loss": 0.3488,
+      "step": 346
+    },
+    {
+      "epoch": 0.66624,
+      "grad_norm": 0.5298962593078613,
+      "learning_rate": 9.550270199610307e-06,
+      "loss": 0.3641,
+      "step": 347
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.6157196760177612,
+      "learning_rate": 9.54562149901362e-06,
+      "loss": 0.3881,
+      "step": 348
+    },
+    {
+      "epoch": 0.67008,
+      "grad_norm": 0.5286968946456909,
+      "learning_rate": 9.54095003913168e-06,
+      "loss": 0.3822,
+      "step": 349
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.49537765979766846,
+      "learning_rate": 9.536255843353832e-06,
+      "loss": 0.3686,
+      "step": 350
+    },
+    {
+      "epoch": 0.67392,
+      "grad_norm": 0.5793171525001526,
+      "learning_rate": 9.531538935183252e-06,
+      "loss": 0.3642,
+      "step": 351
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.5152738094329834,
+      "learning_rate": 9.526799338236828e-06,
+      "loss": 0.328,
+      "step": 352
+    },
+    {
+      "epoch": 0.67776,
+      "grad_norm": 0.5933722853660583,
+      "learning_rate": 9.522037076245057e-06,
+      "loss": 0.3872,
+      "step": 353
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.5264882445335388,
+      "learning_rate": 9.517252173051912e-06,
+      "loss": 0.384,
+      "step": 354
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.5602928996086121,
+      "learning_rate": 9.512444652614728e-06,
+      "loss": 0.4314,
+      "step": 355
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.5468965768814087,
+      "learning_rate": 9.507614539004082e-06,
+      "loss": 0.3645,
+      "step": 356
+    },
+    {
+      "epoch": 0.68544,
+      "grad_norm": 0.554366409778595,
+      "learning_rate": 9.502761856403668e-06,
+      "loss": 0.3773,
+      "step": 357
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.5708198547363281,
+      "learning_rate": 9.497886629110187e-06,
+      "loss": 0.3587,
+      "step": 358
+    },
+    {
+      "epoch": 0.68928,
+      "grad_norm": 0.5647431015968323,
+      "learning_rate": 9.49298888153321e-06,
+      "loss": 0.3574,
+      "step": 359
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.53582763671875,
+      "learning_rate": 9.488068638195072e-06,
+      "loss": 0.442,
+      "step": 360
+    },
+    {
+      "epoch": 0.69312,
+      "grad_norm": 0.6588902473449707,
+      "learning_rate": 9.483125923730731e-06,
+      "loss": 0.4005,
+      "step": 361
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.5396913886070251,
+      "learning_rate": 9.478160762887668e-06,
+      "loss": 0.3446,
+      "step": 362
+    },
+    {
+      "epoch": 0.69696,
+      "grad_norm": 0.5749691128730774,
+      "learning_rate": 9.473173180525737e-06,
+      "loss": 0.3803,
+      "step": 363
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.5601921677589417,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.3744,
+      "step": 364
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.6135634779930115,
+      "learning_rate": 9.463130851245899e-06,
+      "loss": 0.3869,
+      "step": 365
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.5421634912490845,
+      "learning_rate": 9.458076154608515e-06,
+      "loss": 0.3589,
+      "step": 366
+    },
+    {
+      "epoch": 0.70464,
+      "grad_norm": 0.5679751038551331,
+      "learning_rate": 9.452999137013064e-06,
+      "loss": 0.402,
+      "step": 367
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.6569827198982239,
+      "learning_rate": 9.447899823879456e-06,
+      "loss": 0.4345,
+      "step": 368
+    },
+    {
+      "epoch": 0.70848,
+      "grad_norm": 0.505266547203064,
+      "learning_rate": 9.442778240739234e-06,
+      "loss": 0.3758,
+      "step": 369
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.5421440005302429,
+      "learning_rate": 9.437634413235437e-06,
+      "loss": 0.353,
+      "step": 370
+    },
+    {
+      "epoch": 0.71232,
+      "grad_norm": 0.5724952220916748,
+      "learning_rate": 9.43246836712249e-06,
+      "loss": 0.387,
+      "step": 371
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.5868046283721924,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.3687,
+      "step": 372
+    },
+    {
+      "epoch": 0.71616,
+      "grad_norm": 0.5674011707305908,
+      "learning_rate": 9.4220697226429e-06,
+      "loss": 0.4196,
+      "step": 373
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.5359950661659241,
+      "learning_rate": 9.416837176340807e-06,
+      "loss": 0.3736,
+      "step": 374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6915173530578613,
+      "learning_rate": 9.411582515558391e-06,
+      "loss": 0.4458,
+      "step": 375
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.579595685005188,
+      "learning_rate": 9.406305766604996e-06,
+      "loss": 0.3642,
+      "step": 376
+    },
+    {
+      "epoch": 0.72384,
+      "grad_norm": 0.5565221309661865,
+      "learning_rate": 9.401006955900555e-06,
+      "loss": 0.3311,
+      "step": 377
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.615480899810791,
+      "learning_rate": 9.395686109975475e-06,
+      "loss": 0.3451,
+      "step": 378
+    },
+    {
+      "epoch": 0.72768,
+      "grad_norm": 0.5577495098114014,
+      "learning_rate": 9.390343255470471e-06,
+      "loss": 0.4346,
+      "step": 379
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.6395245790481567,
+      "learning_rate": 9.384978419136469e-06,
+      "loss": 0.414,
+      "step": 380
+    },
+    {
+      "epoch": 0.73152,
+      "grad_norm": 0.6253337860107422,
+      "learning_rate": 9.37959162783444e-06,
+      "loss": 0.4397,
+      "step": 381
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.5820584297180176,
+      "learning_rate": 9.374182908535293e-06,
+      "loss": 0.3505,
+      "step": 382
+    },
+    {
+      "epoch": 0.73536,
+      "grad_norm": 0.6101687550544739,
+      "learning_rate": 9.368752288319722e-06,
+      "loss": 0.4072,
+      "step": 383
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.5258579254150391,
+      "learning_rate": 9.363299794378072e-06,
+      "loss": 0.3804,
+      "step": 384
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.5546537041664124,
+      "learning_rate": 9.357825454010214e-06,
+      "loss": 0.3791,
+      "step": 385
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.4797915518283844,
+      "learning_rate": 9.352329294625397e-06,
+      "loss": 0.3465,
+      "step": 386
+    },
+    {
+      "epoch": 0.74304,
+      "grad_norm": 0.5586279034614563,
+      "learning_rate": 9.346811343742115e-06,
+      "loss": 0.3209,
+      "step": 387
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.6374256610870361,
+      "learning_rate": 9.34127162898797e-06,
+      "loss": 0.4478,
+      "step": 388
+    },
+    {
+      "epoch": 0.74688,
+      "grad_norm": 0.579835832118988,
+      "learning_rate": 9.335710178099528e-06,
+      "loss": 0.3523,
+      "step": 389
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.5275377035140991,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.3315,
+      "step": 390
+    },
+    {
+      "epoch": 0.75072,
+      "grad_norm": 0.5745166540145874,
+      "learning_rate": 9.324522179410054e-06,
+      "loss": 0.3546,
+      "step": 391
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.5688664317131042,
+      "learning_rate": 9.318895687625752e-06,
+      "loss": 0.465,
+      "step": 392
+    },
+    {
+      "epoch": 0.75456,
+      "grad_norm": 0.528950035572052,
+      "learning_rate": 9.313247571740336e-06,
+      "loss": 0.3681,
+      "step": 393
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.538253903388977,
+      "learning_rate": 9.307577860033123e-06,
+      "loss": 0.3666,
+      "step": 394
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 0.5336159467697144,
+      "learning_rate": 9.301886580891563e-06,
+      "loss": 0.3998,
+      "step": 395
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.5430910587310791,
+      "learning_rate": 9.296173762811084e-06,
+      "loss": 0.372,
+      "step": 396
+    },
+    {
+      "epoch": 0.76224,
+      "grad_norm": 0.5513455271720886,
+      "learning_rate": 9.290439434394965e-06,
+      "loss": 0.3533,
+      "step": 397
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.5618960857391357,
+      "learning_rate": 9.284683624354172e-06,
+      "loss": 0.4271,
+      "step": 398
+    },
+    {
+      "epoch": 0.76608,
+      "grad_norm": 0.5100370645523071,
+      "learning_rate": 9.278906361507238e-06,
+      "loss": 0.3983,
+      "step": 399
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.60976243019104,
+      "learning_rate": 9.273107674780102e-06,
+      "loss": 0.3924,
+      "step": 400
+    },
+    {
+      "epoch": 0.76992,
+      "grad_norm": 0.5559300184249878,
+      "learning_rate": 9.26728759320597e-06,
+      "loss": 0.3588,
+      "step": 401
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.549557089805603,
+      "learning_rate": 9.261446145925167e-06,
+      "loss": 0.3982,
+      "step": 402
+    },
+    {
+      "epoch": 0.77376,
+      "grad_norm": 0.5489563941955566,
+      "learning_rate": 9.255583362184998e-06,
+      "loss": 0.3557,
+      "step": 403
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.4983119070529938,
+      "learning_rate": 9.249699271339594e-06,
+      "loss": 0.3328,
+      "step": 404
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 0.560558021068573,
+      "learning_rate": 9.243793902849764e-06,
+      "loss": 0.4001,
+      "step": 405
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.5659856796264648,
+      "learning_rate": 9.237867286282855e-06,
+      "loss": 0.349,
+      "step": 406
+    },
+    {
+      "epoch": 0.78144,
+      "grad_norm": 0.514787495136261,
+      "learning_rate": 9.231919451312603e-06,
+      "loss": 0.378,
+      "step": 407
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.5677452087402344,
+      "learning_rate": 9.225950427718974e-06,
+      "loss": 0.4029,
+      "step": 408
+    },
+    {
+      "epoch": 0.78528,
+      "grad_norm": 0.5434529781341553,
+      "learning_rate": 9.21996024538803e-06,
+      "loss": 0.413,
+      "step": 409
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.4897398054599762,
+      "learning_rate": 9.213948934311767e-06,
+      "loss": 0.3736,
+      "step": 410
+    },
+    {
+      "epoch": 0.78912,
+      "grad_norm": 0.4852549731731415,
+      "learning_rate": 9.207916524587971e-06,
+      "loss": 0.3831,
+      "step": 411
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.549034595489502,
+      "learning_rate": 9.201863046420065e-06,
+      "loss": 0.3708,
+      "step": 412
+    },
+    {
+      "epoch": 0.79296,
+      "grad_norm": 0.5679736733436584,
+      "learning_rate": 9.195788530116962e-06,
+      "loss": 0.3881,
+      "step": 413
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.5732312798500061,
+      "learning_rate": 9.189693006092907e-06,
+      "loss": 0.4388,
+      "step": 414
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 0.47129523754119873,
+      "learning_rate": 9.183576504867328e-06,
+      "loss": 0.352,
+      "step": 415
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.5004159212112427,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.3673,
+      "step": 416
+    },
+    {
+      "epoch": 0.80064,
+      "grad_norm": 0.5511000156402588,
+      "learning_rate": 9.171280693414307e-06,
+      "loss": 0.3419,
+      "step": 417
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.6543917059898376,
+      "learning_rate": 9.165101444750259e-06,
+      "loss": 0.4062,
+      "step": 418
+    },
+    {
+      "epoch": 0.80448,
+      "grad_norm": 0.5537492036819458,
+      "learning_rate": 9.158901342011163e-06,
+      "loss": 0.3893,
+      "step": 419
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.49536946415901184,
+      "learning_rate": 9.152680416240059e-06,
+      "loss": 0.3598,
+      "step": 420
+    },
+    {
+      "epoch": 0.80832,
+      "grad_norm": 0.6333961486816406,
+      "learning_rate": 9.146438698584245e-06,
+      "loss": 0.4084,
+      "step": 421
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 0.5091806054115295,
+      "learning_rate": 9.140176220295118e-06,
+      "loss": 0.3065,
+      "step": 422
+    },
+    {
+      "epoch": 0.81216,
+      "grad_norm": 0.5682697892189026,
+      "learning_rate": 9.133893012728027e-06,
+      "loss": 0.2882,
+      "step": 423
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 0.5123572945594788,
+      "learning_rate": 9.1275891073421e-06,
+      "loss": 0.3236,
+      "step": 424
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.5918817520141602,
+      "learning_rate": 9.121264535700107e-06,
+      "loss": 0.4256,
+      "step": 425
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 0.6401883959770203,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.3828,
+      "step": 426
+    },
+    {
+      "epoch": 0.81984,
+      "grad_norm": 0.48944762349128723,
+      "learning_rate": 9.10855352041618e-06,
+      "loss": 0.3162,
+      "step": 427
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 0.5616137385368347,
+      "learning_rate": 9.102167140416503e-06,
+      "loss": 0.3505,
+      "step": 428
+    },
+    {
+      "epoch": 0.82368,
+      "grad_norm": 0.46014800667762756,
+      "learning_rate": 9.09576022144496e-06,
+      "loss": 0.3222,
+      "step": 429
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.573727011680603,
+      "learning_rate": 9.089332795580085e-06,
+      "loss": 0.3844,
+      "step": 430
+    },
+    {
+      "epoch": 0.82752,
+      "grad_norm": 0.6071629524230957,
+      "learning_rate": 9.082884895003098e-06,
+      "loss": 0.3823,
+      "step": 431
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 0.5376788377761841,
+      "learning_rate": 9.076416551997721e-06,
+      "loss": 0.3846,
+      "step": 432
+    },
+    {
+      "epoch": 0.83136,
+      "grad_norm": 0.562679648399353,
+      "learning_rate": 9.06992779895004e-06,
+      "loss": 0.5272,
+      "step": 433
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 0.5987834334373474,
+      "learning_rate": 9.063418668348323e-06,
+      "loss": 0.3567,
+      "step": 434
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 0.5507369041442871,
+      "learning_rate": 9.056889192782865e-06,
+      "loss": 0.4301,
+      "step": 435
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 0.5444300770759583,
+      "learning_rate": 9.050339404945834e-06,
+      "loss": 0.364,
+      "step": 436
+    },
+    {
+      "epoch": 0.83904,
+      "grad_norm": 0.6109519004821777,
+      "learning_rate": 9.04376933763109e-06,
+      "loss": 0.4051,
+      "step": 437
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 0.5487772226333618,
+      "learning_rate": 9.037179023734036e-06,
+      "loss": 0.3661,
+      "step": 438
+    },
+    {
+      "epoch": 0.84288,
+      "grad_norm": 0.6210412383079529,
+      "learning_rate": 9.030568496251437e-06,
+      "loss": 0.4567,
+      "step": 439
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.5431700944900513,
+      "learning_rate": 9.023937788281278e-06,
+      "loss": 0.3955,
+      "step": 440
+    },
+    {
+      "epoch": 0.84672,
+      "grad_norm": 0.5094118714332581,
+      "learning_rate": 9.017286933022573e-06,
+      "loss": 0.3705,
+      "step": 441
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 0.5667728185653687,
+      "learning_rate": 9.01061596377522e-06,
+      "loss": 0.3553,
+      "step": 442
+    },
+    {
+      "epoch": 0.85056,
+      "grad_norm": 0.5509085655212402,
+      "learning_rate": 9.003924913939816e-06,
+      "loss": 0.4073,
+      "step": 443
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 0.567909836769104,
+      "learning_rate": 8.997213817017508e-06,
+      "loss": 0.3966,
+      "step": 444
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 0.5380107760429382,
+      "learning_rate": 8.990482706609805e-06,
+      "loss": 0.3839,
+      "step": 445
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 0.5862552523612976,
+      "learning_rate": 8.98373161641843e-06,
+      "loss": 0.4054,
+      "step": 446
+    },
+    {
+      "epoch": 0.85824,
+      "grad_norm": 0.5594919323921204,
+      "learning_rate": 8.97696058024514e-06,
+      "loss": 0.4418,
+      "step": 447
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 0.5139030814170837,
+      "learning_rate": 8.970169631991556e-06,
+      "loss": 0.3277,
+      "step": 448
+    },
+    {
+      "epoch": 0.86208,
+      "grad_norm": 0.5154045820236206,
+      "learning_rate": 8.963358805658998e-06,
+      "loss": 0.3804,
+      "step": 449
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.6237481832504272,
+      "learning_rate": 8.95652813534831e-06,
+      "loss": 0.4163,
+      "step": 450
+    },
+    {
+      "epoch": 0.86592,
+      "grad_norm": 0.6318241953849792,
+      "learning_rate": 8.949677655259696e-06,
+      "loss": 0.4105,
+      "step": 451
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 0.5220658779144287,
+      "learning_rate": 8.942807399692543e-06,
+      "loss": 0.3751,
+      "step": 452
+    },
+    {
+      "epoch": 0.86976,
+      "grad_norm": 0.5729755163192749,
+      "learning_rate": 8.935917403045251e-06,
+      "loss": 0.3322,
+      "step": 453
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 0.5810550451278687,
+      "learning_rate": 8.92900769981506e-06,
+      "loss": 0.3765,
+      "step": 454
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 0.5998488664627075,
+      "learning_rate": 8.92207832459788e-06,
+      "loss": 0.3494,
+      "step": 455
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 0.555694580078125,
+      "learning_rate": 8.915129312088112e-06,
+      "loss": 0.4156,
+      "step": 456
+    },
+    {
+      "epoch": 0.87744,
+      "grad_norm": 0.5433024764060974,
+      "learning_rate": 8.908160697078482e-06,
+      "loss": 0.3605,
+      "step": 457
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 0.6165239214897156,
+      "learning_rate": 8.901172514459864e-06,
+      "loss": 0.3958,
+      "step": 458
+    },
+    {
+      "epoch": 0.88128,
+      "grad_norm": 0.4919983148574829,
+      "learning_rate": 8.894164799221099e-06,
+      "loss": 0.3392,
+      "step": 459
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.5813742280006409,
+      "learning_rate": 8.88713758644883e-06,
+      "loss": 0.3889,
+      "step": 460
+    },
+    {
+      "epoch": 0.88512,
+      "grad_norm": 0.6200647354125977,
+      "learning_rate": 8.88009091132732e-06,
+      "loss": 0.4466,
+      "step": 461
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 0.5399177670478821,
+      "learning_rate": 8.873024809138272e-06,
+      "loss": 0.3807,
+      "step": 462
+    },
+    {
+      "epoch": 0.88896,
+      "grad_norm": 0.5623127818107605,
+      "learning_rate": 8.86593931526067e-06,
+      "loss": 0.3975,
+      "step": 463
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 0.5197056531906128,
+      "learning_rate": 8.858834465170576e-06,
+      "loss": 0.3533,
+      "step": 464
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 0.6290859580039978,
+      "learning_rate": 8.851710294440974e-06,
+      "loss": 0.5506,
+      "step": 465
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 0.5201876759529114,
+      "learning_rate": 8.84456683874158e-06,
+      "loss": 0.3879,
+      "step": 466
+    },
+    {
+      "epoch": 0.89664,
+      "grad_norm": 0.5704218745231628,
+      "learning_rate": 8.837404133838667e-06,
+      "loss": 0.3421,
+      "step": 467
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 0.4947088956832886,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.3449,
+      "step": 468
+    },
+    {
+      "epoch": 0.90048,
+      "grad_norm": 0.5365477204322815,
+      "learning_rate": 8.823021119969102e-06,
+      "loss": 0.336,
+      "step": 469
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.5147320032119751,
+      "learning_rate": 8.815800883016168e-06,
+      "loss": 0.3793,
+      "step": 470
+    },
+    {
+      "epoch": 0.90432,
+      "grad_norm": 0.518178403377533,
+      "learning_rate": 8.808561540886796e-06,
+      "loss": 0.344,
+      "step": 471
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 0.5249724388122559,
+      "learning_rate": 8.801303129827352e-06,
+      "loss": 0.4073,
+      "step": 472
+    },
+    {
+      "epoch": 0.90816,
+      "grad_norm": 0.6120156049728394,
+      "learning_rate": 8.794025686179677e-06,
+      "loss": 0.5028,
+      "step": 473
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 0.5620540976524353,
+      "learning_rate": 8.786729246380901e-06,
+      "loss": 0.407,
+      "step": 474
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 0.48326146602630615,
+      "learning_rate": 8.779413846963267e-06,
+      "loss": 0.3154,
+      "step": 475
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 0.4844927191734314,
+      "learning_rate": 8.772079524553951e-06,
+      "loss": 0.3318,
+      "step": 476
+    },
+    {
+      "epoch": 0.91584,
+      "grad_norm": 0.5280975103378296,
+      "learning_rate": 8.764726315874872e-06,
+      "loss": 0.3646,
+      "step": 477
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 0.513424277305603,
+      "learning_rate": 8.757354257742501e-06,
+      "loss": 0.3566,
+      "step": 478
+    },
+    {
+      "epoch": 0.91968,
+      "grad_norm": 0.5409324169158936,
+      "learning_rate": 8.749963387067697e-06,
+      "loss": 0.3879,
+      "step": 479
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.5831232666969299,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.4335,
+      "step": 480
+    },
+    {
+      "epoch": 0.92352,
+      "grad_norm": 0.5369328856468201,
+      "learning_rate": 8.735125356204982e-06,
+      "loss": 0.3796,
+      "step": 481
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 0.529913604259491,
+      "learning_rate": 8.727678270308994e-06,
+      "loss": 0.4007,
+      "step": 482
+    },
+    {
+      "epoch": 0.92736,
+      "grad_norm": 0.5112100839614868,
+      "learning_rate": 8.720212520454054e-06,
+      "loss": 0.3697,
+      "step": 483
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 0.47181791067123413,
+      "learning_rate": 8.712728144020118e-06,
+      "loss": 0.3192,
+      "step": 484
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 0.5471856594085693,
+      "learning_rate": 8.705225178480397e-06,
+      "loss": 0.3591,
+      "step": 485
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 0.5897825956344604,
+      "learning_rate": 8.697703661401187e-06,
+      "loss": 0.4018,
+      "step": 486
+    },
+    {
+      "epoch": 0.93504,
+      "grad_norm": 0.537923276424408,
+      "learning_rate": 8.69016363044166e-06,
+      "loss": 0.4189,
+      "step": 487
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 0.5146099328994751,
+      "learning_rate": 8.682605123353685e-06,
+      "loss": 0.4096,
+      "step": 488
+    },
+    {
+      "epoch": 0.93888,
+      "grad_norm": 0.5499489903450012,
+      "learning_rate": 8.675028177981643e-06,
+      "loss": 0.3816,
+      "step": 489
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.523770809173584,
+      "learning_rate": 8.66743283226223e-06,
+      "loss": 0.3716,
+      "step": 490
+    },
+    {
+      "epoch": 0.94272,
+      "grad_norm": 0.5228763818740845,
+      "learning_rate": 8.659819124224266e-06,
+      "loss": 0.3883,
+      "step": 491
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 0.48286834359169006,
+      "learning_rate": 8.652187091988516e-06,
+      "loss": 0.3341,
+      "step": 492
+    },
+    {
+      "epoch": 0.94656,
+      "grad_norm": 0.5550572276115417,
+      "learning_rate": 8.644536773767488e-06,
+      "loss": 0.3841,
+      "step": 493
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 0.5464596748352051,
+      "learning_rate": 8.636868207865244e-06,
+      "loss": 0.368,
+      "step": 494
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 0.5007494688034058,
+      "learning_rate": 8.629181432677213e-06,
+      "loss": 0.3644,
+      "step": 495
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 0.5519047379493713,
+      "learning_rate": 8.621476486689991e-06,
+      "loss": 0.3889,
+      "step": 496
+    },
+    {
+      "epoch": 0.95424,
+      "grad_norm": 0.5680062174797058,
+      "learning_rate": 8.613753408481158e-06,
+      "loss": 0.3731,
+      "step": 497
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 0.5087865591049194,
+      "learning_rate": 8.606012236719073e-06,
+      "loss": 0.3358,
+      "step": 498
+    },
+    {
+      "epoch": 0.95808,
+      "grad_norm": 0.567511260509491,
+      "learning_rate": 8.598253010162693e-06,
+      "loss": 0.414,
+      "step": 499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4895898997783661,
+      "learning_rate": 8.590475767661371e-06,
+      "loss": 0.3677,
+      "step": 500
+    },
+    {
+      "epoch": 0.96192,
+      "grad_norm": 0.5731030702590942,
+      "learning_rate": 8.58268054815466e-06,
+      "loss": 0.3405,
+      "step": 501
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 0.6070635318756104,
+      "learning_rate": 8.574867390672124e-06,
+      "loss": 0.345,
+      "step": 502
+    },
+    {
+      "epoch": 0.96576,
+      "grad_norm": 0.5701974034309387,
+      "learning_rate": 8.567036334333142e-06,
+      "loss": 0.4048,
+      "step": 503
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 0.5527924299240112,
+      "learning_rate": 8.559187418346703e-06,
+      "loss": 0.3936,
+      "step": 504
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 0.5464694499969482,
+      "learning_rate": 8.551320682011227e-06,
+      "loss": 0.3372,
+      "step": 505
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 0.5621950626373291,
+      "learning_rate": 8.543436164714351e-06,
+      "loss": 0.3637,
+      "step": 506
+    },
+    {
+      "epoch": 0.97344,
+      "grad_norm": 0.5593699812889099,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.3928,
+      "step": 507
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 0.5412471890449524,
+      "learning_rate": 8.527613945231886e-06,
+      "loss": 0.4191,
+      "step": 508
+    },
+    {
+      "epoch": 0.97728,
+      "grad_norm": 0.5321531295776367,
+      "learning_rate": 8.519676322265914e-06,
+      "loss": 0.369,
+      "step": 509
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.5176343321800232,
+      "learning_rate": 8.511721076777388e-06,
+      "loss": 0.3451,
+      "step": 510
+    },
+    {
+      "epoch": 0.98112,
+      "grad_norm": 0.610908031463623,
+      "learning_rate": 8.503748248597095e-06,
+      "loss": 0.3263,
+      "step": 511
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 0.5719373226165771,
+      "learning_rate": 8.495757877643857e-06,
+      "loss": 0.4522,
+      "step": 512
+    },
+    {
+      "epoch": 0.98496,
+      "grad_norm": 0.5898048281669617,
+      "learning_rate": 8.487750003924341e-06,
+      "loss": 0.4172,
+      "step": 513
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 0.4991302192211151,
+      "learning_rate": 8.479724667532836e-06,
+      "loss": 0.3667,
+      "step": 514
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 0.4658468961715698,
+      "learning_rate": 8.471681908651067e-06,
+      "loss": 0.2908,
+      "step": 515
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 0.6296018362045288,
+      "learning_rate": 8.463621767547998e-06,
+      "loss": 0.4177,
+      "step": 516
+    },
+    {
+      "epoch": 0.99264,
+      "grad_norm": 0.5341808199882507,
+      "learning_rate": 8.455544284579614e-06,
+      "loss": 0.3326,
+      "step": 517
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 0.576343297958374,
+      "learning_rate": 8.447449500188731e-06,
+      "loss": 0.379,
+      "step": 518
+    },
+    {
+      "epoch": 0.99648,
+      "grad_norm": 0.5497181415557861,
+      "learning_rate": 8.439337454904794e-06,
+      "loss": 0.3728,
+      "step": 519
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.5313594937324524,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.3305,
+      "step": 520
+    },
+    {
+      "epoch": 1.00128,
+      "grad_norm": 1.0214319229125977,
+      "learning_rate": 8.42306174420744e-06,
+      "loss": 0.5778,
+      "step": 521
+    },
+    {
+      "epoch": 1.0032,
+      "grad_norm": 0.4836558997631073,
+      "learning_rate": 8.414898160284208e-06,
+      "loss": 0.3022,
+      "step": 522
+    },
+    {
+      "epoch": 1.00512,
+      "grad_norm": 0.5065402388572693,
+      "learning_rate": 8.406717478447889e-06,
+      "loss": 0.2895,
+      "step": 523
+    },
+    {
+      "epoch": 1.00704,
+      "grad_norm": 0.5304985046386719,
+      "learning_rate": 8.398519739657997e-06,
+      "loss": 0.3462,
+      "step": 524
+    },
+    {
+      "epoch": 1.00896,
+      "grad_norm": 0.4894730746746063,
+      "learning_rate": 8.390304984959455e-06,
+      "loss": 0.2661,
+      "step": 525
+    },
+    {
+      "epoch": 1.01088,
+      "grad_norm": 0.528671383857727,
+      "learning_rate": 8.382073255482381e-06,
+      "loss": 0.3121,
+      "step": 526
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 0.541412353515625,
+      "learning_rate": 8.373824592441884e-06,
+      "loss": 0.3718,
+      "step": 527
+    },
+    {
+      "epoch": 1.01472,
+      "grad_norm": 0.5245410799980164,
+      "learning_rate": 8.36555903713785e-06,
+      "loss": 0.3446,
+      "step": 528
+    },
+    {
+      "epoch": 1.01664,
+      "grad_norm": 0.5120326280593872,
+      "learning_rate": 8.357276630954756e-06,
+      "loss": 0.3164,
+      "step": 529
+    },
+    {
+      "epoch": 1.01856,
+      "grad_norm": 0.5497516393661499,
+      "learning_rate": 8.348977415361435e-06,
+      "loss": 0.3402,
+      "step": 530
+    },
+    {
+      "epoch": 1.02048,
+      "grad_norm": 0.5723618268966675,
+      "learning_rate": 8.340661431910888e-06,
+      "loss": 0.304,
+      "step": 531
+    },
+    {
+      "epoch": 1.0224,
+      "grad_norm": 0.48848795890808105,
+      "learning_rate": 8.332328722240072e-06,
+      "loss": 0.2542,
+      "step": 532
+    },
+    {
+      "epoch": 1.02432,
+      "grad_norm": 0.557712972164154,
+      "learning_rate": 8.323979328069689e-06,
+      "loss": 0.4216,
+      "step": 533
+    },
+    {
+      "epoch": 1.02624,
+      "grad_norm": 0.5226695537567139,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.2848,
+      "step": 534
+    },
+    {
+      "epoch": 1.02816,
+      "grad_norm": 0.5187362432479858,
+      "learning_rate": 8.307230653530501e-06,
+      "loss": 0.3351,
+      "step": 535
+    },
+    {
+      "epoch": 1.03008,
+      "grad_norm": 0.5180396437644958,
+      "learning_rate": 8.298831457019943e-06,
+      "loss": 0.3382,
+      "step": 536
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 0.5368833541870117,
+      "learning_rate": 8.290415743725894e-06,
+      "loss": 0.3133,
+      "step": 537
+    },
+    {
+      "epoch": 1.03392,
+      "grad_norm": 0.5255162119865417,
+      "learning_rate": 8.28198355578465e-06,
+      "loss": 0.3213,
+      "step": 538
+    },
+    {
+      "epoch": 1.03584,
+      "grad_norm": 0.5252994894981384,
+      "learning_rate": 8.273534935414975e-06,
+      "loss": 0.3295,
+      "step": 539
+    },
+    {
+      "epoch": 1.03776,
+      "grad_norm": 0.5332139134407043,
+      "learning_rate": 8.265069924917925e-06,
+      "loss": 0.336,
+      "step": 540
+    },
+    {
+      "epoch": 1.03968,
+      "grad_norm": 0.4785391688346863,
+      "learning_rate": 8.256588566676616e-06,
+      "loss": 0.2827,
+      "step": 541
+    },
+    {
+      "epoch": 1.0416,
+      "grad_norm": 0.5707299113273621,
+      "learning_rate": 8.248090903156003e-06,
+      "loss": 0.3117,
+      "step": 542
+    },
+    {
+      "epoch": 1.04352,
+      "grad_norm": 0.5353734493255615,
+      "learning_rate": 8.239576976902694e-06,
+      "loss": 0.3229,
+      "step": 543
+    },
+    {
+      "epoch": 1.04544,
+      "grad_norm": 0.5370491743087769,
+      "learning_rate": 8.231046830544716e-06,
+      "loss": 0.365,
+      "step": 544
+    },
+    {
+      "epoch": 1.04736,
+      "grad_norm": 0.49072229862213135,
+      "learning_rate": 8.222500506791305e-06,
+      "loss": 0.3091,
+      "step": 545
+    },
+    {
+      "epoch": 1.04928,
+      "grad_norm": 0.5337362885475159,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.3401,
+      "step": 546
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 0.5536717176437378,
+      "learning_rate": 8.205359498339917e-06,
+      "loss": 0.34,
+      "step": 547
+    },
+    {
+      "epoch": 1.05312,
+      "grad_norm": 0.5838882923126221,
+      "learning_rate": 8.196764899464552e-06,
+      "loss": 0.3205,
+      "step": 548
+    },
+    {
+      "epoch": 1.05504,
+      "grad_norm": 0.5116017460823059,
+      "learning_rate": 8.188154294838542e-06,
+      "loss": 0.2995,
+      "step": 549
+    },
+    {
+      "epoch": 1.05696,
+      "grad_norm": 0.5031540989875793,
+      "learning_rate": 8.179527727573975e-06,
+      "loss": 0.3242,
+      "step": 550
+    },
+    {
+      "epoch": 1.05888,
+      "grad_norm": 0.5126543641090393,
+      "learning_rate": 8.170885240862854e-06,
+      "loss": 0.3118,
+      "step": 551
+    },
+    {
+      "epoch": 1.0608,
+      "grad_norm": 0.5230513215065002,
+      "learning_rate": 8.162226877976886e-06,
+      "loss": 0.33,
+      "step": 552
+    },
+    {
+      "epoch": 1.06272,
+      "grad_norm": 0.5024076700210571,
+      "learning_rate": 8.153552682267278e-06,
+      "loss": 0.3364,
+      "step": 553
+    },
+    {
+      "epoch": 1.06464,
+      "grad_norm": 0.5597369074821472,
+      "learning_rate": 8.144862697164499e-06,
+      "loss": 0.3418,
+      "step": 554
+    },
+    {
+      "epoch": 1.06656,
+      "grad_norm": 0.5870678424835205,
+      "learning_rate": 8.136156966178082e-06,
+      "loss": 0.3541,
+      "step": 555
+    },
+    {
+      "epoch": 1.06848,
+      "grad_norm": 0.5419480204582214,
+      "learning_rate": 8.127435532896388e-06,
+      "loss": 0.3884,
+      "step": 556
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 0.4972580373287201,
+      "learning_rate": 8.118698440986405e-06,
+      "loss": 0.3433,
+      "step": 557
+    },
+    {
+      "epoch": 1.07232,
+      "grad_norm": 0.5727570652961731,
+      "learning_rate": 8.10994573419352e-06,
+      "loss": 0.3673,
+      "step": 558
+    },
+    {
+      "epoch": 1.07424,
+      "grad_norm": 0.5865340232849121,
+      "learning_rate": 8.101177456341301e-06,
+      "loss": 0.37,
+      "step": 559
+    },
+    {
+      "epoch": 1.07616,
+      "grad_norm": 0.471744567155838,
+      "learning_rate": 8.092393651331275e-06,
+      "loss": 0.2805,
+      "step": 560
+    },
+    {
+      "epoch": 1.07808,
+      "grad_norm": 0.5264679193496704,
+      "learning_rate": 8.083594363142717e-06,
+      "loss": 0.3091,
+      "step": 561
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5631963610649109,
+      "learning_rate": 8.074779635832417e-06,
+      "loss": 0.3401,
+      "step": 562
+    },
+    {
+      "epoch": 1.08192,
+      "grad_norm": 0.5642712116241455,
+      "learning_rate": 8.065949513534474e-06,
+      "loss": 0.3328,
+      "step": 563
+    },
+    {
+      "epoch": 1.08384,
+      "grad_norm": 0.4875067174434662,
+      "learning_rate": 8.057104040460062e-06,
+      "loss": 0.2442,
+      "step": 564
+    },
+    {
+      "epoch": 1.08576,
+      "grad_norm": 0.5161327123641968,
+      "learning_rate": 8.048243260897216e-06,
+      "loss": 0.3174,
+      "step": 565
+    },
+    {
+      "epoch": 1.08768,
+      "grad_norm": 0.5194908976554871,
+      "learning_rate": 8.03936721921061e-06,
+      "loss": 0.3804,
+      "step": 566
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 0.5510116815567017,
+      "learning_rate": 8.030475959841333e-06,
+      "loss": 0.3032,
+      "step": 567
+    },
+    {
+      "epoch": 1.09152,
+      "grad_norm": 0.576263964176178,
+      "learning_rate": 8.021569527306663e-06,
+      "loss": 0.3598,
+      "step": 568
+    },
+    {
+      "epoch": 1.09344,
+      "grad_norm": 0.48472243547439575,
+      "learning_rate": 8.012647966199851e-06,
+      "loss": 0.2931,
+      "step": 569
+    },
+    {
+      "epoch": 1.09536,
+      "grad_norm": 0.5796002745628357,
+      "learning_rate": 8.003711321189895e-06,
+      "loss": 0.4123,
+      "step": 570
+    },
+    {
+      "epoch": 1.09728,
+      "grad_norm": 0.560088574886322,
+      "learning_rate": 7.994759637021316e-06,
+      "loss": 0.344,
+      "step": 571
+    },
+    {
+      "epoch": 1.0992,
+      "grad_norm": 0.5430134534835815,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 0.3257,
+      "step": 572
+    },
+    {
+      "epoch": 1.10112,
+      "grad_norm": 0.5343568325042725,
+      "learning_rate": 7.976811330562637e-06,
+      "loss": 0.306,
+      "step": 573
+    },
+    {
+      "epoch": 1.10304,
+      "grad_norm": 0.5186337828636169,
+      "learning_rate": 7.967814798137173e-06,
+      "loss": 0.321,
+      "step": 574
+    },
+    {
+      "epoch": 1.10496,
+      "grad_norm": 0.5277876257896423,
+      "learning_rate": 7.95880340628191e-06,
+      "loss": 0.338,
+      "step": 575
+    },
+    {
+      "epoch": 1.10688,
+      "grad_norm": 0.5278550386428833,
+      "learning_rate": 7.949777200115617e-06,
+      "loss": 0.3069,
+      "step": 576
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 0.5518340468406677,
+      "learning_rate": 7.940736224831227e-06,
+      "loss": 0.2886,
+      "step": 577
+    },
+    {
+      "epoch": 1.11072,
+      "grad_norm": 0.6374801993370056,
+      "learning_rate": 7.931680525695634e-06,
+      "loss": 0.3822,
+      "step": 578
+    },
+    {
+      "epoch": 1.11264,
+      "grad_norm": 0.4857504963874817,
+      "learning_rate": 7.922610148049445e-06,
+      "loss": 0.2834,
+      "step": 579
+    },
+    {
+      "epoch": 1.11456,
+      "grad_norm": 0.5055893063545227,
+      "learning_rate": 7.913525137306756e-06,
+      "loss": 0.3154,
+      "step": 580
+    },
+    {
+      "epoch": 1.11648,
+      "grad_norm": 0.5457763075828552,
+      "learning_rate": 7.90442553895494e-06,
+      "loss": 0.2896,
+      "step": 581
+    },
+    {
+      "epoch": 1.1184,
+      "grad_norm": 0.4903712272644043,
+      "learning_rate": 7.895311398554395e-06,
+      "loss": 0.2943,
+      "step": 582
+    },
+    {
+      "epoch": 1.12032,
+      "grad_norm": 0.5530596375465393,
+      "learning_rate": 7.886182761738339e-06,
+      "loss": 0.2997,
+      "step": 583
+    },
+    {
+      "epoch": 1.12224,
+      "grad_norm": 0.5430228114128113,
+      "learning_rate": 7.877039674212569e-06,
+      "loss": 0.331,
+      "step": 584
+    },
+    {
+      "epoch": 1.12416,
+      "grad_norm": 0.5476499795913696,
+      "learning_rate": 7.86788218175523e-06,
+      "loss": 0.3561,
+      "step": 585
+    },
+    {
+      "epoch": 1.12608,
+      "grad_norm": 0.46878013014793396,
+      "learning_rate": 7.8587103302166e-06,
+      "loss": 0.2847,
+      "step": 586
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.5620825886726379,
+      "learning_rate": 7.849524165518838e-06,
+      "loss": 0.2817,
+      "step": 587
+    },
+    {
+      "epoch": 1.12992,
+      "grad_norm": 0.5655076503753662,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.3181,
+      "step": 588
+    },
+    {
+      "epoch": 1.13184,
+      "grad_norm": 0.5563917756080627,
+      "learning_rate": 7.831109080692687e-06,
+      "loss": 0.3316,
+      "step": 589
+    },
+    {
+      "epoch": 1.13376,
+      "grad_norm": 0.5462495684623718,
+      "learning_rate": 7.821880252766025e-06,
+      "loss": 0.345,
+      "step": 590
+    },
+    {
+      "epoch": 1.13568,
+      "grad_norm": 0.5746179819107056,
+      "learning_rate": 7.812637296083235e-06,
+      "loss": 0.3229,
+      "step": 591
+    },
+    {
+      "epoch": 1.1376,
+      "grad_norm": 0.581531286239624,
+      "learning_rate": 7.803380256922495e-06,
+      "loss": 0.3141,
+      "step": 592
+    },
+    {
+      "epoch": 1.13952,
+      "grad_norm": 0.6214243173599243,
+      "learning_rate": 7.794109181632494e-06,
+      "loss": 0.3482,
+      "step": 593
+    },
+    {
+      "epoch": 1.14144,
+      "grad_norm": 0.5478833317756653,
+      "learning_rate": 7.784824116632198e-06,
+      "loss": 0.366,
+      "step": 594
+    },
+    {
+      "epoch": 1.14336,
+      "grad_norm": 0.5128333568572998,
+      "learning_rate": 7.775525108410616e-06,
+      "loss": 0.2884,
+      "step": 595
+    },
+    {
+      "epoch": 1.14528,
+      "grad_norm": 0.553466796875,
+      "learning_rate": 7.76621220352657e-06,
+      "loss": 0.3226,
+      "step": 596
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 6.04440975189209,
+      "learning_rate": 7.75688544860846e-06,
+      "loss": 0.5574,
+      "step": 597
+    },
+    {
+      "epoch": 1.14912,
+      "grad_norm": 0.6106770038604736,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 0.2983,
+      "step": 598
+    },
+    {
+      "epoch": 1.15104,
+      "grad_norm": 0.47948354482650757,
+      "learning_rate": 7.73819057553014e-06,
+      "loss": 0.276,
+      "step": 599
+    },
+    {
+      "epoch": 1.15296,
+      "grad_norm": 0.5397146940231323,
+      "learning_rate": 7.728822550972523e-06,
+      "loss": 0.3526,
+      "step": 600
+    },
+    {
+      "epoch": 1.15488,
+      "grad_norm": 0.5309719443321228,
+      "learning_rate": 7.719440863585555e-06,
+      "loss": 0.3118,
+      "step": 601
+    },
+    {
+      "epoch": 1.1568,
+      "grad_norm": 0.5489014387130737,
+      "learning_rate": 7.710045560342021e-06,
+      "loss": 0.3693,
+      "step": 602
+    },
+    {
+      "epoch": 1.15872,
+      "grad_norm": 0.5081278681755066,
+      "learning_rate": 7.700636688282878e-06,
+      "loss": 0.304,
+      "step": 603
+    },
+    {
+      "epoch": 1.16064,
+      "grad_norm": 0.6480275988578796,
+      "learning_rate": 7.69121429451702e-06,
+      "loss": 0.4215,
+      "step": 604
+    },
+    {
+      "epoch": 1.16256,
+      "grad_norm": 0.4894363284111023,
+      "learning_rate": 7.681778426221043e-06,
+      "loss": 0.2649,
+      "step": 605
+    },
+    {
+      "epoch": 1.16448,
+      "grad_norm": 0.5170885324478149,
+      "learning_rate": 7.672329130639007e-06,
+      "loss": 0.3128,
+      "step": 606
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 0.5022065043449402,
+      "learning_rate": 7.662866455082201e-06,
+      "loss": 0.2929,
+      "step": 607
+    },
+    {
+      "epoch": 1.16832,
+      "grad_norm": 0.48748090863227844,
+      "learning_rate": 7.65339044692891e-06,
+      "loss": 0.3217,
+      "step": 608
+    },
+    {
+      "epoch": 1.17024,
+      "grad_norm": 0.47743692994117737,
+      "learning_rate": 7.643901153624163e-06,
+      "loss": 0.3459,
+      "step": 609
+    },
+    {
+      "epoch": 1.17216,
+      "grad_norm": 0.5453119277954102,
+      "learning_rate": 7.634398622679517e-06,
+      "loss": 0.2998,
+      "step": 610
+    },
+    {
+      "epoch": 1.17408,
+      "grad_norm": 0.5851200222969055,
+      "learning_rate": 7.624882901672801e-06,
+      "loss": 0.3301,
+      "step": 611
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 0.5713387727737427,
+      "learning_rate": 7.615354038247889e-06,
+      "loss": 0.3708,
+      "step": 612
+    },
+    {
+      "epoch": 1.17792,
+      "grad_norm": 0.5219767689704895,
+      "learning_rate": 7.605812080114453e-06,
+      "loss": 0.309,
+      "step": 613
+    },
+    {
+      "epoch": 1.17984,
+      "grad_norm": 0.5656505823135376,
+      "learning_rate": 7.59625707504773e-06,
+      "loss": 0.3964,
+      "step": 614
+    },
+    {
+      "epoch": 1.18176,
+      "grad_norm": 0.5251127481460571,
+      "learning_rate": 7.586689070888284e-06,
+      "loss": 0.3349,
+      "step": 615
+    },
+    {
+      "epoch": 1.18368,
+      "grad_norm": 0.5733042359352112,
+      "learning_rate": 7.577108115541761e-06,
+      "loss": 0.3532,
+      "step": 616
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 0.48283764719963074,
+      "learning_rate": 7.567514256978652e-06,
+      "loss": 0.2725,
+      "step": 617
+    },
+    {
+      "epoch": 1.18752,
+      "grad_norm": 0.5394615530967712,
+      "learning_rate": 7.557907543234051e-06,
+      "loss": 0.3468,
+      "step": 618
+    },
+    {
+      "epoch": 1.18944,
+      "grad_norm": 0.47491973638534546,
+      "learning_rate": 7.54828802240742e-06,
+      "loss": 0.2847,
+      "step": 619
+    },
+    {
+      "epoch": 1.19136,
+      "grad_norm": 0.4529227018356323,
+      "learning_rate": 7.53865574266234e-06,
+      "loss": 0.2683,
+      "step": 620
+    },
+    {
+      "epoch": 1.1932800000000001,
+      "grad_norm": 0.5472639799118042,
+      "learning_rate": 7.529010752226277e-06,
+      "loss": 0.3383,
+      "step": 621
+    },
+    {
+      "epoch": 1.1952,
+      "grad_norm": 0.5126523971557617,
+      "learning_rate": 7.519353099390336e-06,
+      "loss": 0.3032,
+      "step": 622
+    },
+    {
+      "epoch": 1.19712,
+      "grad_norm": 0.5155563354492188,
+      "learning_rate": 7.509682832509024e-06,
+      "loss": 0.334,
+      "step": 623
+    },
+    {
+      "epoch": 1.19904,
+      "grad_norm": 0.5419436097145081,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3986,
+      "step": 624
+    },
+    {
+      "epoch": 1.20096,
+      "grad_norm": 0.5080655217170715,
+      "learning_rate": 7.490304650343842e-06,
+      "loss": 0.2714,
+      "step": 625
+    },
+    {
+      "epoch": 1.20288,
+      "grad_norm": 0.5573438405990601,
+      "learning_rate": 7.480596832083795e-06,
+      "loss": 0.353,
+      "step": 626
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 0.48851191997528076,
+      "learning_rate": 7.470876593825535e-06,
+      "loss": 0.3223,
+      "step": 627
+    },
+    {
+      "epoch": 1.20672,
+      "grad_norm": 0.6116130352020264,
+      "learning_rate": 7.461143984236925e-06,
+      "loss": 0.3516,
+      "step": 628
+    },
+    {
+      "epoch": 1.20864,
+      "grad_norm": 0.5926026105880737,
+      "learning_rate": 7.451399052047764e-06,
+      "loss": 0.3181,
+      "step": 629
+    },
+    {
+      "epoch": 1.21056,
+      "grad_norm": 0.5785939693450928,
+      "learning_rate": 7.441641846049557e-06,
+      "loss": 0.3145,
+      "step": 630
+    },
+    {
+      "epoch": 1.21248,
+      "grad_norm": 0.5167310237884521,
+      "learning_rate": 7.431872415095252e-06,
+      "loss": 0.3353,
+      "step": 631
+    },
+    {
+      "epoch": 1.2144,
+      "grad_norm": 0.5237314105033875,
+      "learning_rate": 7.422090808099014e-06,
+      "loss": 0.3043,
+      "step": 632
+    },
+    {
+      "epoch": 1.21632,
+      "grad_norm": 0.5447617769241333,
+      "learning_rate": 7.412297074035968e-06,
+      "loss": 0.2778,
+      "step": 633
+    },
+    {
+      "epoch": 1.21824,
+      "grad_norm": 0.675801157951355,
+      "learning_rate": 7.402491261941958e-06,
+      "loss": 0.3871,
+      "step": 634
+    },
+    {
+      "epoch": 1.22016,
+      "grad_norm": 0.5504624247550964,
+      "learning_rate": 7.3926734209133e-06,
+      "loss": 0.3361,
+      "step": 635
+    },
+    {
+      "epoch": 1.22208,
+      "grad_norm": 0.542606770992279,
+      "learning_rate": 7.382843600106539e-06,
+      "loss": 0.3023,
+      "step": 636
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 0.5649086833000183,
+      "learning_rate": 7.373001848738203e-06,
+      "loss": 0.2885,
+      "step": 637
+    },
+    {
+      "epoch": 1.22592,
+      "grad_norm": 0.5574055910110474,
+      "learning_rate": 7.363148216084548e-06,
+      "loss": 0.3332,
+      "step": 638
+    },
+    {
+      "epoch": 1.22784,
+      "grad_norm": 0.5035303831100464,
+      "learning_rate": 7.353282751481324e-06,
+      "loss": 0.3131,
+      "step": 639
+    },
+    {
+      "epoch": 1.22976,
+      "grad_norm": 0.5089981555938721,
+      "learning_rate": 7.343405504323519e-06,
+      "loss": 0.3519,
+      "step": 640
+    },
+    {
+      "epoch": 1.2316799999999999,
+      "grad_norm": 0.530159056186676,
+      "learning_rate": 7.333516524065116e-06,
+      "loss": 0.3148,
+      "step": 641
+    },
+    {
+      "epoch": 1.2336,
+      "grad_norm": 0.510208249092102,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.3043,
+      "step": 642
+    },
+    {
+      "epoch": 1.23552,
+      "grad_norm": 0.5438703298568726,
+      "learning_rate": 7.3137035623559274e-06,
+      "loss": 0.3209,
+      "step": 643
+    },
+    {
+      "epoch": 1.23744,
+      "grad_norm": 0.5391506552696228,
+      "learning_rate": 7.303779680105844e-06,
+      "loss": 0.3124,
+      "step": 644
+    },
+    {
+      "epoch": 1.23936,
+      "grad_norm": 0.5544272065162659,
+      "learning_rate": 7.2938442631560714e-06,
+      "loss": 0.3386,
+      "step": 645
+    },
+    {
+      "epoch": 1.24128,
+      "grad_norm": 0.5338462591171265,
+      "learning_rate": 7.28389736125184e-06,
+      "loss": 0.3002,
+      "step": 646
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 0.5146346688270569,
+      "learning_rate": 7.273939024195881e-06,
+      "loss": 0.3558,
+      "step": 647
+    },
+    {
+      "epoch": 1.24512,
+      "grad_norm": 0.4881197214126587,
+      "learning_rate": 7.263969301848188e-06,
+      "loss": 0.2953,
+      "step": 648
+    },
+    {
+      "epoch": 1.24704,
+      "grad_norm": 0.5438975095748901,
+      "learning_rate": 7.2539882441257484e-06,
+      "loss": 0.3654,
+      "step": 649
+    },
+    {
+      "epoch": 1.24896,
+      "grad_norm": 0.5106989741325378,
+      "learning_rate": 7.243995901002312e-06,
+      "loss": 0.3212,
+      "step": 650
+    },
+    {
+      "epoch": 1.25088,
+      "grad_norm": 0.5152694582939148,
+      "learning_rate": 7.2339923225081296e-06,
+      "loss": 0.3056,
+      "step": 651
+    },
+    {
+      "epoch": 1.2528000000000001,
+      "grad_norm": 0.6609506011009216,
+      "learning_rate": 7.223977558729707e-06,
+      "loss": 0.4505,
+      "step": 652
+    },
+    {
+      "epoch": 1.25472,
+      "grad_norm": 0.5182729959487915,
+      "learning_rate": 7.213951659809553e-06,
+      "loss": 0.3279,
+      "step": 653
+    },
+    {
+      "epoch": 1.25664,
+      "grad_norm": 0.5102232098579407,
+      "learning_rate": 7.203914675945929e-06,
+      "loss": 0.3089,
+      "step": 654
+    },
+    {
+      "epoch": 1.2585600000000001,
+      "grad_norm": 0.628922164440155,
+      "learning_rate": 7.193866657392597e-06,
+      "loss": 0.3598,
+      "step": 655
+    },
+    {
+      "epoch": 1.26048,
+      "grad_norm": 0.5390359163284302,
+      "learning_rate": 7.183807654458565e-06,
+      "loss": 0.2978,
+      "step": 656
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 0.5219982266426086,
+      "learning_rate": 7.173737717507843e-06,
+      "loss": 0.2727,
+      "step": 657
+    },
+    {
+      "epoch": 1.26432,
+      "grad_norm": 0.5432875156402588,
+      "learning_rate": 7.163656896959181e-06,
+      "loss": 0.3169,
+      "step": 658
+    },
+    {
+      "epoch": 1.26624,
+      "grad_norm": 0.5286130905151367,
+      "learning_rate": 7.153565243285829e-06,
+      "loss": 0.3218,
+      "step": 659
+    },
+    {
+      "epoch": 1.26816,
+      "grad_norm": 0.6107270121574402,
+      "learning_rate": 7.143462807015271e-06,
+      "loss": 0.3058,
+      "step": 660
+    },
+    {
+      "epoch": 1.27008,
+      "grad_norm": 0.49751877784729004,
+      "learning_rate": 7.133349638728979e-06,
+      "loss": 0.3007,
+      "step": 661
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 0.4890269339084625,
+      "learning_rate": 7.1232257890621605e-06,
+      "loss": 0.3012,
+      "step": 662
+    },
+    {
+      "epoch": 1.27392,
+      "grad_norm": 0.5718302726745605,
+      "learning_rate": 7.113091308703498e-06,
+      "loss": 0.4348,
+      "step": 663
+    },
+    {
+      "epoch": 1.27584,
+      "grad_norm": 0.5391543507575989,
+      "learning_rate": 7.102946248394908e-06,
+      "loss": 0.3198,
+      "step": 664
+    },
+    {
+      "epoch": 1.27776,
+      "grad_norm": 0.5384120345115662,
+      "learning_rate": 7.092790658931274e-06,
+      "loss": 0.3733,
+      "step": 665
+    },
+    {
+      "epoch": 1.27968,
+      "grad_norm": 0.5070114731788635,
+      "learning_rate": 7.082624591160201e-06,
+      "loss": 0.2633,
+      "step": 666
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 0.57356858253479,
+      "learning_rate": 7.072448095981751e-06,
+      "loss": 0.4372,
+      "step": 667
+    },
+    {
+      "epoch": 1.28352,
+      "grad_norm": 0.5364060401916504,
+      "learning_rate": 7.0622612243482035e-06,
+      "loss": 0.3438,
+      "step": 668
+    },
+    {
+      "epoch": 1.28544,
+      "grad_norm": 0.5661773085594177,
+      "learning_rate": 7.052064027263785e-06,
+      "loss": 0.3733,
+      "step": 669
+    },
+    {
+      "epoch": 1.28736,
+      "grad_norm": 0.5251317620277405,
+      "learning_rate": 7.041856555784421e-06,
+      "loss": 0.3243,
+      "step": 670
+    },
+    {
+      "epoch": 1.28928,
+      "grad_norm": 0.5340425372123718,
+      "learning_rate": 7.031638861017484e-06,
+      "loss": 0.3226,
+      "step": 671
+    },
+    {
+      "epoch": 1.2912,
+      "grad_norm": 0.541844367980957,
+      "learning_rate": 7.021410994121525e-06,
+      "loss": 0.3295,
+      "step": 672
+    },
+    {
+      "epoch": 1.29312,
+      "grad_norm": 0.5221070647239685,
+      "learning_rate": 7.011173006306034e-06,
+      "loss": 0.3191,
+      "step": 673
+    },
+    {
+      "epoch": 1.29504,
+      "grad_norm": 0.5229676961898804,
+      "learning_rate": 7.0009249488311685e-06,
+      "loss": 0.3233,
+      "step": 674
+    },
+    {
+      "epoch": 1.29696,
+      "grad_norm": 0.5415786504745483,
+      "learning_rate": 6.990666873007506e-06,
+      "loss": 0.3516,
+      "step": 675
+    },
+    {
+      "epoch": 1.29888,
+      "grad_norm": 0.5137751698493958,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.3515,
+      "step": 676
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 0.5411648750305176,
+      "learning_rate": 6.970120871806647e-06,
+      "loss": 0.3349,
+      "step": 677
+    },
+    {
+      "epoch": 1.3027199999999999,
+      "grad_norm": 0.5252477526664734,
+      "learning_rate": 6.959833049300376e-06,
+      "loss": 0.3384,
+      "step": 678
+    },
+    {
+      "epoch": 1.30464,
+      "grad_norm": 0.5093960165977478,
+      "learning_rate": 6.949535414186651e-06,
+      "loss": 0.3163,
+      "step": 679
+    },
+    {
+      "epoch": 1.30656,
+      "grad_norm": 0.47321853041648865,
+      "learning_rate": 6.939228018024275e-06,
+      "loss": 0.2921,
+      "step": 680
+    },
+    {
+      "epoch": 1.3084799999999999,
+      "grad_norm": 0.5221200585365295,
+      "learning_rate": 6.9289109124209276e-06,
+      "loss": 0.3138,
+      "step": 681
+    },
+    {
+      "epoch": 1.3104,
+      "grad_norm": 0.5649718642234802,
+      "learning_rate": 6.918584149032899e-06,
+      "loss": 0.3132,
+      "step": 682
+    },
+    {
+      "epoch": 1.31232,
+      "grad_norm": 0.6127787232398987,
+      "learning_rate": 6.908247779564836e-06,
+      "loss": 0.3252,
+      "step": 683
+    },
+    {
+      "epoch": 1.31424,
+      "grad_norm": 0.5303078889846802,
+      "learning_rate": 6.897901855769483e-06,
+      "loss": 0.324,
+      "step": 684
+    },
+    {
+      "epoch": 1.31616,
+      "grad_norm": 0.5830091238021851,
+      "learning_rate": 6.887546429447418e-06,
+      "loss": 0.3617,
+      "step": 685
+    },
+    {
+      "epoch": 1.31808,
+      "grad_norm": 0.5893061757087708,
+      "learning_rate": 6.8771815524468e-06,
+      "loss": 0.3667,
+      "step": 686
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5136197209358215,
+      "learning_rate": 6.8668072766631054e-06,
+      "loss": 0.2979,
+      "step": 687
+    },
+    {
+      "epoch": 1.32192,
+      "grad_norm": 0.5482615232467651,
+      "learning_rate": 6.856423654038868e-06,
+      "loss": 0.3002,
+      "step": 688
+    },
+    {
+      "epoch": 1.3238400000000001,
+      "grad_norm": 0.5251845717430115,
+      "learning_rate": 6.8460307365634225e-06,
+      "loss": 0.3028,
+      "step": 689
+    },
+    {
+      "epoch": 1.32576,
+      "grad_norm": 0.517441987991333,
+      "learning_rate": 6.8356285762726385e-06,
+      "loss": 0.31,
+      "step": 690
+    },
+    {
+      "epoch": 1.32768,
+      "grad_norm": 0.5339541435241699,
+      "learning_rate": 6.825217225248664e-06,
+      "loss": 0.3387,
+      "step": 691
+    },
+    {
+      "epoch": 1.3296000000000001,
+      "grad_norm": 0.49072977900505066,
+      "learning_rate": 6.814796735619664e-06,
+      "loss": 0.3087,
+      "step": 692
+    },
+    {
+      "epoch": 1.33152,
+      "grad_norm": 0.5255969166755676,
+      "learning_rate": 6.804367159559561e-06,
+      "loss": 0.3147,
+      "step": 693
+    },
+    {
+      "epoch": 1.33344,
+      "grad_norm": 0.5354084968566895,
+      "learning_rate": 6.793928549287767e-06,
+      "loss": 0.3834,
+      "step": 694
+    },
+    {
+      "epoch": 1.33536,
+      "grad_norm": 0.5187183618545532,
+      "learning_rate": 6.783480957068934e-06,
+      "loss": 0.3469,
+      "step": 695
+    },
+    {
+      "epoch": 1.33728,
+      "grad_norm": 0.5410125255584717,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.335,
+      "step": 696
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 0.5563452839851379,
+      "learning_rate": 6.762559036073331e-06,
+      "loss": 0.3619,
+      "step": 697
+    },
+    {
+      "epoch": 1.34112,
+      "grad_norm": 0.5452044606208801,
+      "learning_rate": 6.75208481204967e-06,
+      "loss": 0.3389,
+      "step": 698
+    },
+    {
+      "epoch": 1.34304,
+      "grad_norm": 0.5451205968856812,
+      "learning_rate": 6.74160181558465e-06,
+      "loss": 0.4036,
+      "step": 699
+    },
+    {
+      "epoch": 1.34496,
+      "grad_norm": 0.5362733006477356,
+      "learning_rate": 6.731110099165165e-06,
+      "loss": 0.3605,
+      "step": 700
+    },
+    {
+      "epoch": 1.34688,
+      "grad_norm": 0.5080693364143372,
+      "learning_rate": 6.7206097153217474e-06,
+      "loss": 0.3235,
+      "step": 701
+    },
+    {
+      "epoch": 1.3488,
+      "grad_norm": 0.54218989610672,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.3468,
+      "step": 702
+    },
+    {
+      "epoch": 1.35072,
+      "grad_norm": 0.5971646904945374,
+      "learning_rate": 6.699583155702025e-06,
+      "loss": 0.3964,
+      "step": 703
+    },
+    {
+      "epoch": 1.35264,
+      "grad_norm": 0.507707953453064,
+      "learning_rate": 6.689057085202737e-06,
+      "loss": 0.2845,
+      "step": 704
+    },
+    {
+      "epoch": 1.35456,
+      "grad_norm": 0.6143893599510193,
+      "learning_rate": 6.678522557833025e-06,
+      "loss": 0.3275,
+      "step": 705
+    },
+    {
+      "epoch": 1.35648,
+      "grad_norm": 0.5078967213630676,
+      "learning_rate": 6.667979626337783e-06,
+      "loss": 0.3129,
+      "step": 706
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 0.506790816783905,
+      "learning_rate": 6.6574283435039795e-06,
+      "loss": 0.3252,
+      "step": 707
+    },
+    {
+      "epoch": 1.36032,
+      "grad_norm": 0.5443456768989563,
+      "learning_rate": 6.646868762160399e-06,
+      "loss": 0.3535,
+      "step": 708
+    },
+    {
+      "epoch": 1.36224,
+      "grad_norm": 0.5069459676742554,
+      "learning_rate": 6.6363009351773755e-06,
+      "loss": 0.3354,
+      "step": 709
+    },
+    {
+      "epoch": 1.36416,
+      "grad_norm": 0.5225284099578857,
+      "learning_rate": 6.625724915466526e-06,
+      "loss": 0.3088,
+      "step": 710
+    },
+    {
+      "epoch": 1.36608,
+      "grad_norm": 0.5142247080802917,
+      "learning_rate": 6.615140755980491e-06,
+      "loss": 0.3708,
+      "step": 711
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 0.5313759446144104,
+      "learning_rate": 6.6045485097126585e-06,
+      "loss": 0.3031,
+      "step": 712
+    },
+    {
+      "epoch": 1.36992,
+      "grad_norm": 0.500249445438385,
+      "learning_rate": 6.593948229696915e-06,
+      "loss": 0.274,
+      "step": 713
+    },
+    {
+      "epoch": 1.37184,
+      "grad_norm": 0.5179996490478516,
+      "learning_rate": 6.583339969007364e-06,
+      "loss": 0.3371,
+      "step": 714
+    },
+    {
+      "epoch": 1.3737599999999999,
+      "grad_norm": 0.532130241394043,
+      "learning_rate": 6.572723780758069e-06,
+      "loss": 0.351,
+      "step": 715
+    },
+    {
+      "epoch": 1.37568,
+      "grad_norm": 0.5067139863967896,
+      "learning_rate": 6.562099718102788e-06,
+      "loss": 0.3136,
+      "step": 716
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 0.5194411277770996,
+      "learning_rate": 6.551467834234702e-06,
+      "loss": 0.3029,
+      "step": 717
+    },
+    {
+      "epoch": 1.37952,
+      "grad_norm": 0.5290529727935791,
+      "learning_rate": 6.540828182386154e-06,
+      "loss": 0.3566,
+      "step": 718
+    },
+    {
+      "epoch": 1.38144,
+      "grad_norm": 0.5016028881072998,
+      "learning_rate": 6.530180815828377e-06,
+      "loss": 0.3157,
+      "step": 719
+    },
+    {
+      "epoch": 1.38336,
+      "grad_norm": 0.5252622365951538,
+      "learning_rate": 6.519525787871235e-06,
+      "loss": 0.2941,
+      "step": 720
+    },
+    {
+      "epoch": 1.38528,
+      "grad_norm": 0.491657018661499,
+      "learning_rate": 6.508863151862947e-06,
+      "loss": 0.3031,
+      "step": 721
+    },
+    {
+      "epoch": 1.3872,
+      "grad_norm": 0.4873288869857788,
+      "learning_rate": 6.498192961189829e-06,
+      "loss": 0.2793,
+      "step": 722
+    },
+    {
+      "epoch": 1.3891200000000001,
+      "grad_norm": 0.46787866950035095,
+      "learning_rate": 6.487515269276015e-06,
+      "loss": 0.2721,
+      "step": 723
+    },
+    {
+      "epoch": 1.39104,
+      "grad_norm": 0.5433241128921509,
+      "learning_rate": 6.476830129583207e-06,
+      "loss": 0.3252,
+      "step": 724
+    },
+    {
+      "epoch": 1.39296,
+      "grad_norm": 0.5271470546722412,
+      "learning_rate": 6.466137595610388e-06,
+      "loss": 0.3571,
+      "step": 725
+    },
+    {
+      "epoch": 1.3948800000000001,
+      "grad_norm": 0.5099048018455505,
+      "learning_rate": 6.455437720893565e-06,
+      "loss": 0.3505,
+      "step": 726
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 0.5412172675132751,
+      "learning_rate": 6.444730559005504e-06,
+      "loss": 0.3258,
+      "step": 727
+    },
+    {
+      "epoch": 1.39872,
+      "grad_norm": 0.48186013102531433,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.2983,
+      "step": 728
+    },
+    {
+      "epoch": 1.40064,
+      "grad_norm": 0.5244131684303284,
+      "learning_rate": 6.423294588188874e-06,
+      "loss": 0.3314,
+      "step": 729
+    },
+    {
+      "epoch": 1.40256,
+      "grad_norm": 0.6155484318733215,
+      "learning_rate": 6.412565886587186e-06,
+      "loss": 0.3945,
+      "step": 730
+    },
+    {
+      "epoch": 1.40448,
+      "grad_norm": 0.5291752219200134,
+      "learning_rate": 6.401830112467479e-06,
+      "loss": 0.3243,
+      "step": 731
+    },
+    {
+      "epoch": 1.4064,
+      "grad_norm": 0.4708503484725952,
+      "learning_rate": 6.391087319582264e-06,
+      "loss": 0.2997,
+      "step": 732
+    },
+    {
+      "epoch": 1.40832,
+      "grad_norm": 0.5548900961875916,
+      "learning_rate": 6.380337561719184e-06,
+      "loss": 0.406,
+      "step": 733
+    },
+    {
+      "epoch": 1.41024,
+      "grad_norm": 0.5348383188247681,
+      "learning_rate": 6.369580892700763e-06,
+      "loss": 0.3475,
+      "step": 734
+    },
+    {
+      "epoch": 1.41216,
+      "grad_norm": 0.5502620935440063,
+      "learning_rate": 6.358817366384122e-06,
+      "loss": 0.3157,
+      "step": 735
+    },
+    {
+      "epoch": 1.41408,
+      "grad_norm": 0.5946040153503418,
+      "learning_rate": 6.34804703666072e-06,
+      "loss": 0.3255,
+      "step": 736
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 0.46333763003349304,
+      "learning_rate": 6.337269957456074e-06,
+      "loss": 0.2849,
+      "step": 737
+    },
+    {
+      "epoch": 1.41792,
+      "grad_norm": 0.5915679335594177,
+      "learning_rate": 6.326486182729504e-06,
+      "loss": 0.3539,
+      "step": 738
+    },
+    {
+      "epoch": 1.41984,
+      "grad_norm": 0.5221284031867981,
+      "learning_rate": 6.315695766473844e-06,
+      "loss": 0.3136,
+      "step": 739
+    },
+    {
+      "epoch": 1.42176,
+      "grad_norm": 0.4997546970844269,
+      "learning_rate": 6.304898762715187e-06,
+      "loss": 0.3526,
+      "step": 740
+    },
+    {
+      "epoch": 1.42368,
+      "grad_norm": 0.5197610855102539,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.3624,
+      "step": 741
+    },
+    {
+      "epoch": 1.4256,
+      "grad_norm": 0.5438770651817322,
+      "learning_rate": 6.283285208957884e-06,
+      "loss": 0.3719,
+      "step": 742
+    },
+    {
+      "epoch": 1.42752,
+      "grad_norm": 0.5013489127159119,
+      "learning_rate": 6.272468767175255e-06,
+      "loss": 0.3484,
+      "step": 743
+    },
+    {
+      "epoch": 1.42944,
+      "grad_norm": 0.49272608757019043,
+      "learning_rate": 6.261645954321109e-06,
+      "loss": 0.3261,
+      "step": 744
+    },
+    {
+      "epoch": 1.43136,
+      "grad_norm": 0.598677933216095,
+      "learning_rate": 6.2508168245837476e-06,
+      "loss": 0.4359,
+      "step": 745
+    },
+    {
+      "epoch": 1.4332799999999999,
+      "grad_norm": 0.5429832339286804,
+      "learning_rate": 6.239981432183093e-06,
+      "loss": 0.3853,
+      "step": 746
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 0.5542439818382263,
+      "learning_rate": 6.2291398313704265e-06,
+      "loss": 0.4036,
+      "step": 747
+    },
+    {
+      "epoch": 1.43712,
+      "grad_norm": 0.4820418655872345,
+      "learning_rate": 6.21829207642811e-06,
+      "loss": 0.3409,
+      "step": 748
+    },
+    {
+      "epoch": 1.4390399999999999,
+      "grad_norm": 0.45357003808021545,
+      "learning_rate": 6.207438221669325e-06,
+      "loss": 0.2595,
+      "step": 749
+    },
+    {
+      "epoch": 1.44096,
+      "grad_norm": 0.5634178519248962,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.3536,
+      "step": 750
+    },
+    {
+      "epoch": 1.44288,
+      "grad_norm": 0.45755258202552795,
+      "learning_rate": 6.185712430107489e-06,
+      "loss": 0.2903,
+      "step": 751
+    },
+    {
+      "epoch": 1.4447999999999999,
+      "grad_norm": 0.5569748282432556,
+      "learning_rate": 6.1748406020824115e-06,
+      "loss": 0.3918,
+      "step": 752
+    },
+    {
+      "epoch": 1.44672,
+      "grad_norm": 0.49455296993255615,
+      "learning_rate": 6.163962891796261e-06,
+      "loss": 0.3141,
+      "step": 753
+    },
+    {
+      "epoch": 1.44864,
+      "grad_norm": 0.5747455954551697,
+      "learning_rate": 6.153079353712201e-06,
+      "loss": 0.326,
+      "step": 754
+    },
+    {
+      "epoch": 1.45056,
+      "grad_norm": 0.5606670379638672,
+      "learning_rate": 6.142190042322569e-06,
+      "loss": 0.3525,
+      "step": 755
+    },
+    {
+      "epoch": 1.45248,
+      "grad_norm": 0.4764695167541504,
+      "learning_rate": 6.131295012148613e-06,
+      "loss": 0.2938,
+      "step": 756
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 0.5392735004425049,
+      "learning_rate": 6.120394317740205e-06,
+      "loss": 0.3203,
+      "step": 757
+    },
+    {
+      "epoch": 1.45632,
+      "grad_norm": 0.5450900793075562,
+      "learning_rate": 6.1094880136755886e-06,
+      "loss": 0.3513,
+      "step": 758
+    },
+    {
+      "epoch": 1.45824,
+      "grad_norm": 0.5558665990829468,
+      "learning_rate": 6.0985761545610865e-06,
+      "loss": 0.3284,
+      "step": 759
+    },
+    {
+      "epoch": 1.4601600000000001,
+      "grad_norm": 0.5546119213104248,
+      "learning_rate": 6.087658795030838e-06,
+      "loss": 0.3631,
+      "step": 760
+    },
+    {
+      "epoch": 1.46208,
+      "grad_norm": 0.4934506118297577,
+      "learning_rate": 6.076735989746522e-06,
+      "loss": 0.2645,
+      "step": 761
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 0.556404173374176,
+      "learning_rate": 6.065807793397081e-06,
+      "loss": 0.337,
+      "step": 762
+    },
+    {
+      "epoch": 1.4659200000000001,
+      "grad_norm": 0.5665925741195679,
+      "learning_rate": 6.0548742606984545e-06,
+      "loss": 0.3799,
+      "step": 763
+    },
+    {
+      "epoch": 1.46784,
+      "grad_norm": 0.5205016136169434,
+      "learning_rate": 6.043935446393294e-06,
+      "loss": 0.3031,
+      "step": 764
+    },
+    {
+      "epoch": 1.46976,
+      "grad_norm": 0.5917834639549255,
+      "learning_rate": 6.032991405250702e-06,
+      "loss": 0.3667,
+      "step": 765
+    },
+    {
+      "epoch": 1.47168,
+      "grad_norm": 0.553932249546051,
+      "learning_rate": 6.022042192065946e-06,
+      "loss": 0.3484,
+      "step": 766
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 0.49104544520378113,
+      "learning_rate": 6.011087861660191e-06,
+      "loss": 0.2788,
+      "step": 767
+    },
+    {
+      "epoch": 1.47552,
+      "grad_norm": 0.5559536218643188,
+      "learning_rate": 6.000128468880223e-06,
+      "loss": 0.35,
+      "step": 768
+    },
+    {
+      "epoch": 1.47744,
+      "grad_norm": 0.5878273844718933,
+      "learning_rate": 5.989164068598175e-06,
+      "loss": 0.3834,
+      "step": 769
+    },
+    {
+      "epoch": 1.47936,
+      "grad_norm": 0.5694558620452881,
+      "learning_rate": 5.978194715711254e-06,
+      "loss": 0.3529,
+      "step": 770
+    },
+    {
+      "epoch": 1.48128,
+      "grad_norm": 0.5030460357666016,
+      "learning_rate": 5.9672204651414564e-06,
+      "loss": 0.3086,
+      "step": 771
+    },
+    {
+      "epoch": 1.4832,
+      "grad_norm": 0.4853646159172058,
+      "learning_rate": 5.956241371835312e-06,
+      "loss": 0.3297,
+      "step": 772
+    },
+    {
+      "epoch": 1.48512,
+      "grad_norm": 0.4849759638309479,
+      "learning_rate": 5.945257490763588e-06,
+      "loss": 0.2822,
+      "step": 773
+    },
+    {
+      "epoch": 1.48704,
+      "grad_norm": 0.46324241161346436,
+      "learning_rate": 5.934268876921026e-06,
+      "loss": 0.259,
+      "step": 774
+    },
+    {
+      "epoch": 1.48896,
+      "grad_norm": 0.560608983039856,
+      "learning_rate": 5.923275585326064e-06,
+      "loss": 0.3438,
+      "step": 775
+    },
+    {
+      "epoch": 1.49088,
+      "grad_norm": 0.5113677978515625,
+      "learning_rate": 5.912277671020564e-06,
+      "loss": 0.2989,
+      "step": 776
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 0.5435959696769714,
+      "learning_rate": 5.90127518906953e-06,
+      "loss": 0.3074,
+      "step": 777
+    },
+    {
+      "epoch": 1.49472,
+      "grad_norm": 0.5299482941627502,
+      "learning_rate": 5.890268194560834e-06,
+      "loss": 0.3998,
+      "step": 778
+    },
+    {
+      "epoch": 1.49664,
+      "grad_norm": 0.47808772325515747,
+      "learning_rate": 5.8792567426049465e-06,
+      "loss": 0.3283,
+      "step": 779
+    },
+    {
+      "epoch": 1.49856,
+      "grad_norm": 0.4860895276069641,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.2945,
+      "step": 780
+    },
+    {
+      "epoch": 1.50048,
+      "grad_norm": 0.5218390226364136,
+      "learning_rate": 5.857220686904779e-06,
+      "loss": 0.3312,
+      "step": 781
+    },
+    {
+      "epoch": 1.5024,
+      "grad_norm": 0.5428465604782104,
+      "learning_rate": 5.846196193491919e-06,
+      "loss": 0.3569,
+      "step": 782
+    },
+    {
+      "epoch": 1.5043199999999999,
+      "grad_norm": 0.5657923817634583,
+      "learning_rate": 5.835167463294155e-06,
+      "loss": 0.3604,
+      "step": 783
+    },
+    {
+      "epoch": 1.50624,
+      "grad_norm": 0.5528233647346497,
+      "learning_rate": 5.824134551530783e-06,
+      "loss": 0.3068,
+      "step": 784
+    },
+    {
+      "epoch": 1.50816,
+      "grad_norm": 0.5969262719154358,
+      "learning_rate": 5.813097513442035e-06,
+      "loss": 0.3514,
+      "step": 785
+    },
+    {
+      "epoch": 1.5100799999999999,
+      "grad_norm": 0.5790405869483948,
+      "learning_rate": 5.8020564042888015e-06,
+      "loss": 0.4507,
+      "step": 786
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 0.4907466471195221,
+      "learning_rate": 5.79101127935236e-06,
+      "loss": 0.2509,
+      "step": 787
+    },
+    {
+      "epoch": 1.5139200000000002,
+      "grad_norm": 0.5101063251495361,
+      "learning_rate": 5.77996219393409e-06,
+      "loss": 0.3301,
+      "step": 788
+    },
+    {
+      "epoch": 1.5158399999999999,
+      "grad_norm": 0.53609699010849,
+      "learning_rate": 5.768909203355203e-06,
+      "loss": 0.3295,
+      "step": 789
+    },
+    {
+      "epoch": 1.51776,
+      "grad_norm": 0.5712200403213501,
+      "learning_rate": 5.757852362956463e-06,
+      "loss": 0.3513,
+      "step": 790
+    },
+    {
+      "epoch": 1.5196800000000001,
+      "grad_norm": 0.4859534204006195,
+      "learning_rate": 5.7467917280979105e-06,
+      "loss": 0.3136,
+      "step": 791
+    },
+    {
+      "epoch": 1.5215999999999998,
+      "grad_norm": 0.5172631740570068,
+      "learning_rate": 5.735727354158581e-06,
+      "loss": 0.3774,
+      "step": 792
+    },
+    {
+      "epoch": 1.52352,
+      "grad_norm": 0.5391514897346497,
+      "learning_rate": 5.724659296536234e-06,
+      "loss": 0.3525,
+      "step": 793
+    },
+    {
+      "epoch": 1.5254400000000001,
+      "grad_norm": 0.6238823533058167,
+      "learning_rate": 5.713587610647073e-06,
+      "loss": 0.3521,
+      "step": 794
+    },
+    {
+      "epoch": 1.52736,
+      "grad_norm": 0.6376439929008484,
+      "learning_rate": 5.7025123519254644e-06,
+      "loss": 0.3411,
+      "step": 795
+    },
+    {
+      "epoch": 1.52928,
+      "grad_norm": 0.534277081489563,
+      "learning_rate": 5.6914335758236665e-06,
+      "loss": 0.375,
+      "step": 796
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 0.5576221942901611,
+      "learning_rate": 5.680351337811547e-06,
+      "loss": 0.399,
+      "step": 797
+    },
+    {
+      "epoch": 1.53312,
+      "grad_norm": 0.5293326377868652,
+      "learning_rate": 5.669265693376309e-06,
+      "loss": 0.2972,
+      "step": 798
+    },
+    {
+      "epoch": 1.53504,
+      "grad_norm": 0.5283350944519043,
+      "learning_rate": 5.658176698022208e-06,
+      "loss": 0.2811,
+      "step": 799
+    },
+    {
+      "epoch": 1.53696,
+      "grad_norm": 0.5648627877235413,
+      "learning_rate": 5.647084407270277e-06,
+      "loss": 0.3443,
+      "step": 800
+    },
+    {
+      "epoch": 1.53888,
+      "grad_norm": 0.5358335971832275,
+      "learning_rate": 5.6359888766580555e-06,
+      "loss": 0.3104,
+      "step": 801
+    },
+    {
+      "epoch": 1.5408,
+      "grad_norm": 0.5814449191093445,
+      "learning_rate": 5.624890161739292e-06,
+      "loss": 0.381,
+      "step": 802
+    },
+    {
+      "epoch": 1.54272,
+      "grad_norm": 0.5779856443405151,
+      "learning_rate": 5.6137883180836925e-06,
+      "loss": 0.3634,
+      "step": 803
+    },
+    {
+      "epoch": 1.54464,
+      "grad_norm": 0.5728549957275391,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.3401,
+      "step": 804
+    },
+    {
+      "epoch": 1.54656,
+      "grad_norm": 0.5282721519470215,
+      "learning_rate": 5.591575466918816e-06,
+      "loss": 0.3432,
+      "step": 805
+    },
+    {
+      "epoch": 1.54848,
+      "grad_norm": 0.5740209817886353,
+      "learning_rate": 5.5804645706261515e-06,
+      "loss": 0.3558,
+      "step": 806
+    },
+    {
+      "epoch": 1.5504,
+      "grad_norm": 0.5547658205032349,
+      "learning_rate": 5.569350768029312e-06,
+      "loss": 0.3073,
+      "step": 807
+    },
+    {
+      "epoch": 1.55232,
+      "grad_norm": 0.5588148236274719,
+      "learning_rate": 5.5582341147735396e-06,
+      "loss": 0.3265,
+      "step": 808
+    },
+    {
+      "epoch": 1.55424,
+      "grad_norm": 0.5846301913261414,
+      "learning_rate": 5.5471146665183455e-06,
+      "loss": 0.4171,
+      "step": 809
+    },
+    {
+      "epoch": 1.55616,
+      "grad_norm": 0.5607483983039856,
+      "learning_rate": 5.53599247893724e-06,
+      "loss": 0.2764,
+      "step": 810
+    },
+    {
+      "epoch": 1.55808,
+      "grad_norm": 0.5195860862731934,
+      "learning_rate": 5.524867607717445e-06,
+      "loss": 0.3192,
+      "step": 811
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5422046780586243,
+      "learning_rate": 5.5137401085596224e-06,
+      "loss": 0.3355,
+      "step": 812
+    },
+    {
+      "epoch": 1.56192,
+      "grad_norm": 0.5090670585632324,
+      "learning_rate": 5.502610037177586e-06,
+      "loss": 0.2964,
+      "step": 813
+    },
+    {
+      "epoch": 1.56384,
+      "grad_norm": 0.5915701985359192,
+      "learning_rate": 5.491477449298036e-06,
+      "loss": 0.4335,
+      "step": 814
+    },
+    {
+      "epoch": 1.56576,
+      "grad_norm": 0.5539973378181458,
+      "learning_rate": 5.480342400660268e-06,
+      "loss": 0.3374,
+      "step": 815
+    },
+    {
+      "epoch": 1.56768,
+      "grad_norm": 0.5721865296363831,
+      "learning_rate": 5.469204947015897e-06,
+      "loss": 0.3415,
+      "step": 816
+    },
+    {
+      "epoch": 1.5695999999999999,
+      "grad_norm": 0.5368900299072266,
+      "learning_rate": 5.458065144128584e-06,
+      "loss": 0.2908,
+      "step": 817
+    },
+    {
+      "epoch": 1.57152,
+      "grad_norm": 0.5578900575637817,
+      "learning_rate": 5.4469230477737466e-06,
+      "loss": 0.3246,
+      "step": 818
+    },
+    {
+      "epoch": 1.57344,
+      "grad_norm": 0.4852248430252075,
+      "learning_rate": 5.435778713738292e-06,
+      "loss": 0.2689,
+      "step": 819
+    },
+    {
+      "epoch": 1.5753599999999999,
+      "grad_norm": 0.524515688419342,
+      "learning_rate": 5.424632197820325e-06,
+      "loss": 0.3091,
+      "step": 820
+    },
+    {
+      "epoch": 1.57728,
+      "grad_norm": 0.631579577922821,
+      "learning_rate": 5.413483555828879e-06,
+      "loss": 0.3848,
+      "step": 821
+    },
+    {
+      "epoch": 1.5792000000000002,
+      "grad_norm": 0.5169774889945984,
+      "learning_rate": 5.402332843583631e-06,
+      "loss": 0.3466,
+      "step": 822
+    },
+    {
+      "epoch": 1.5811199999999999,
+      "grad_norm": 0.4710410535335541,
+      "learning_rate": 5.391180116914621e-06,
+      "loss": 0.3167,
+      "step": 823
+    },
+    {
+      "epoch": 1.58304,
+      "grad_norm": 0.5386317372322083,
+      "learning_rate": 5.380025431661981e-06,
+      "loss": 0.3373,
+      "step": 824
+    },
+    {
+      "epoch": 1.5849600000000001,
+      "grad_norm": 0.4905245900154114,
+      "learning_rate": 5.368868843675642e-06,
+      "loss": 0.294,
+      "step": 825
+    },
+    {
+      "epoch": 1.5868799999999998,
+      "grad_norm": 0.5149863362312317,
+      "learning_rate": 5.3577104088150685e-06,
+      "loss": 0.329,
+      "step": 826
+    },
+    {
+      "epoch": 1.5888,
+      "grad_norm": 0.6283377408981323,
+      "learning_rate": 5.346550182948966e-06,
+      "loss": 0.5021,
+      "step": 827
+    },
+    {
+      "epoch": 1.5907200000000001,
+      "grad_norm": 0.5324111580848694,
+      "learning_rate": 5.335388221955012e-06,
+      "loss": 0.3314,
+      "step": 828
+    },
+    {
+      "epoch": 1.5926399999999998,
+      "grad_norm": 0.5217656493186951,
+      "learning_rate": 5.3242245817195705e-06,
+      "loss": 0.3284,
+      "step": 829
+    },
+    {
+      "epoch": 1.59456,
+      "grad_norm": 0.5667556524276733,
+      "learning_rate": 5.31305931813741e-06,
+      "loss": 0.4152,
+      "step": 830
+    },
+    {
+      "epoch": 1.5964800000000001,
+      "grad_norm": 0.5161786675453186,
+      "learning_rate": 5.301892487111431e-06,
+      "loss": 0.321,
+      "step": 831
+    },
+    {
+      "epoch": 1.5984,
+      "grad_norm": 0.521098792552948,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.3803,
+      "step": 832
+    },
+    {
+      "epoch": 1.60032,
+      "grad_norm": 0.4888674020767212,
+      "learning_rate": 5.279554346378572e-06,
+      "loss": 0.2717,
+      "step": 833
+    },
+    {
+      "epoch": 1.60224,
+      "grad_norm": 0.4844496548175812,
+      "learning_rate": 5.268383148515608e-06,
+      "loss": 0.3001,
+      "step": 834
+    },
+    {
+      "epoch": 1.60416,
+      "grad_norm": 0.5077923536300659,
+      "learning_rate": 5.257210606896102e-06,
+      "loss": 0.3395,
+      "step": 835
+    },
+    {
+      "epoch": 1.60608,
+      "grad_norm": 0.5035995244979858,
+      "learning_rate": 5.246036777459391e-06,
+      "loss": 0.303,
+      "step": 836
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 0.5372249484062195,
+      "learning_rate": 5.234861716151264e-06,
+      "loss": 0.3177,
+      "step": 837
+    },
+    {
+      "epoch": 1.60992,
+      "grad_norm": 0.5323225855827332,
+      "learning_rate": 5.223685478923671e-06,
+      "loss": 0.351,
+      "step": 838
+    },
+    {
+      "epoch": 1.61184,
+      "grad_norm": 0.5223349928855896,
+      "learning_rate": 5.2125081217344595e-06,
+      "loss": 0.2818,
+      "step": 839
+    },
+    {
+      "epoch": 1.61376,
+      "grad_norm": 0.484747976064682,
+      "learning_rate": 5.201329700547077e-06,
+      "loss": 0.3124,
+      "step": 840
+    },
+    {
+      "epoch": 1.61568,
+      "grad_norm": 0.5133703947067261,
+      "learning_rate": 5.190150271330298e-06,
+      "loss": 0.3322,
+      "step": 841
+    },
+    {
+      "epoch": 1.6176,
+      "grad_norm": 0.49542000889778137,
+      "learning_rate": 5.178969890057953e-06,
+      "loss": 0.2796,
+      "step": 842
+    },
+    {
+      "epoch": 1.61952,
+      "grad_norm": 0.5596998929977417,
+      "learning_rate": 5.167788612708627e-06,
+      "loss": 0.2963,
+      "step": 843
+    },
+    {
+      "epoch": 1.62144,
+      "grad_norm": 0.6021963953971863,
+      "learning_rate": 5.156606495265402e-06,
+      "loss": 0.3577,
+      "step": 844
+    },
+    {
+      "epoch": 1.62336,
+      "grad_norm": 0.501692533493042,
+      "learning_rate": 5.145423593715558e-06,
+      "loss": 0.3175,
+      "step": 845
+    },
+    {
+      "epoch": 1.62528,
+      "grad_norm": 0.5035648941993713,
+      "learning_rate": 5.1342399640503074e-06,
+      "loss": 0.3506,
+      "step": 846
+    },
+    {
+      "epoch": 1.6272,
+      "grad_norm": 0.49351170659065247,
+      "learning_rate": 5.1230556622645026e-06,
+      "loss": 0.308,
+      "step": 847
+    },
+    {
+      "epoch": 1.62912,
+      "grad_norm": 0.5333631038665771,
+      "learning_rate": 5.111870744356366e-06,
+      "loss": 0.3439,
+      "step": 848
+    },
+    {
+      "epoch": 1.63104,
+      "grad_norm": 0.509105384349823,
+      "learning_rate": 5.100685266327202e-06,
+      "loss": 0.3013,
+      "step": 849
+    },
+    {
+      "epoch": 1.63296,
+      "grad_norm": 0.9426395893096924,
+      "learning_rate": 5.0894992841811216e-06,
+      "loss": 0.4624,
+      "step": 850
+    },
+    {
+      "epoch": 1.6348799999999999,
+      "grad_norm": 0.5247693657875061,
+      "learning_rate": 5.0783128539247585e-06,
+      "loss": 0.2894,
+      "step": 851
+    },
+    {
+      "epoch": 1.6368,
+      "grad_norm": 0.4647666811943054,
+      "learning_rate": 5.067126031566988e-06,
+      "loss": 0.311,
+      "step": 852
+    },
+    {
+      "epoch": 1.63872,
+      "grad_norm": 0.4928247630596161,
+      "learning_rate": 5.055938873118653e-06,
+      "loss": 0.3161,
+      "step": 853
+    },
+    {
+      "epoch": 1.6406399999999999,
+      "grad_norm": 0.5570545792579651,
+      "learning_rate": 5.044751434592274e-06,
+      "loss": 0.3009,
+      "step": 854
+    },
+    {
+      "epoch": 1.64256,
+      "grad_norm": 0.5021554231643677,
+      "learning_rate": 5.033563772001782e-06,
+      "loss": 0.2901,
+      "step": 855
+    },
+    {
+      "epoch": 1.6444800000000002,
+      "grad_norm": 0.5140386819839478,
+      "learning_rate": 5.022375941362218e-06,
+      "loss": 0.3079,
+      "step": 856
+    },
+    {
+      "epoch": 1.6463999999999999,
+      "grad_norm": 0.4929907023906708,
+      "learning_rate": 5.011187998689474e-06,
+      "loss": 0.2768,
+      "step": 857
+    },
+    {
+      "epoch": 1.64832,
+      "grad_norm": 0.5014060139656067,
+      "learning_rate": 5e-06,
+      "loss": 0.3348,
+      "step": 858
+    },
+    {
+      "epoch": 1.6502400000000002,
+      "grad_norm": 0.5392226576805115,
+      "learning_rate": 4.988812001310528e-06,
+      "loss": 0.3785,
+      "step": 859
+    },
+    {
+      "epoch": 1.6521599999999999,
+      "grad_norm": 0.5059987902641296,
+      "learning_rate": 4.977624058637783e-06,
+      "loss": 0.3465,
+      "step": 860
+    },
+    {
+      "epoch": 1.65408,
+      "grad_norm": 0.4583433270454407,
+      "learning_rate": 4.9664362279982205e-06,
+      "loss": 0.2892,
+      "step": 861
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 0.49462541937828064,
+      "learning_rate": 4.955248565407727e-06,
+      "loss": 0.2784,
+      "step": 862
+    },
+    {
+      "epoch": 1.6579199999999998,
+      "grad_norm": 0.557277500629425,
+      "learning_rate": 4.944061126881348e-06,
+      "loss": 0.3642,
+      "step": 863
+    },
+    {
+      "epoch": 1.65984,
+      "grad_norm": 0.4494497776031494,
+      "learning_rate": 4.932873968433014e-06,
+      "loss": 0.2658,
+      "step": 864
+    },
+    {
+      "epoch": 1.6617600000000001,
+      "grad_norm": 0.5919420123100281,
+      "learning_rate": 4.921687146075244e-06,
+      "loss": 0.4344,
+      "step": 865
+    },
+    {
+      "epoch": 1.66368,
+      "grad_norm": 0.47635751962661743,
+      "learning_rate": 4.910500715818879e-06,
+      "loss": 0.3002,
+      "step": 866
+    },
+    {
+      "epoch": 1.6656,
+      "grad_norm": 0.5443571209907532,
+      "learning_rate": 4.8993147336728e-06,
+      "loss": 0.3422,
+      "step": 867
+    },
+    {
+      "epoch": 1.6675200000000001,
+      "grad_norm": 0.5035550594329834,
+      "learning_rate": 4.8881292556436355e-06,
+      "loss": 0.2509,
+      "step": 868
+    },
+    {
+      "epoch": 1.66944,
+      "grad_norm": 0.48384571075439453,
+      "learning_rate": 4.876944337735499e-06,
+      "loss": 0.2957,
+      "step": 869
+    },
+    {
+      "epoch": 1.67136,
+      "grad_norm": 0.5203081965446472,
+      "learning_rate": 4.865760035949695e-06,
+      "loss": 0.3445,
+      "step": 870
+    },
+    {
+      "epoch": 1.67328,
+      "grad_norm": 0.5037247538566589,
+      "learning_rate": 4.854576406284443e-06,
+      "loss": 0.306,
+      "step": 871
+    },
+    {
+      "epoch": 1.6752,
+      "grad_norm": 0.6113672852516174,
+      "learning_rate": 4.8433935047346e-06,
+      "loss": 0.3799,
+      "step": 872
+    },
+    {
+      "epoch": 1.67712,
+      "grad_norm": 0.4938548803329468,
+      "learning_rate": 4.832211387291374e-06,
+      "loss": 0.3326,
+      "step": 873
+    },
+    {
+      "epoch": 1.67904,
+      "grad_norm": 0.5296657085418701,
+      "learning_rate": 4.821030109942048e-06,
+      "loss": 0.3526,
+      "step": 874
+    },
+    {
+      "epoch": 1.68096,
+      "grad_norm": 0.49086233973503113,
+      "learning_rate": 4.8098497286697024e-06,
+      "loss": 0.3363,
+      "step": 875
+    },
+    {
+      "epoch": 1.68288,
+      "grad_norm": 0.5188177227973938,
+      "learning_rate": 4.798670299452926e-06,
+      "loss": 0.397,
+      "step": 876
+    },
+    {
+      "epoch": 1.6848,
+      "grad_norm": 0.4988774359226227,
+      "learning_rate": 4.787491878265542e-06,
+      "loss": 0.3138,
+      "step": 877
+    },
+    {
+      "epoch": 1.68672,
+      "grad_norm": 0.4984115958213806,
+      "learning_rate": 4.77631452107633e-06,
+      "loss": 0.3628,
+      "step": 878
+    },
+    {
+      "epoch": 1.68864,
+      "grad_norm": 0.8624169230461121,
+      "learning_rate": 4.765138283848739e-06,
+      "loss": 0.442,
+      "step": 879
+    },
+    {
+      "epoch": 1.69056,
+      "grad_norm": 0.4918670952320099,
+      "learning_rate": 4.75396322254061e-06,
+      "loss": 0.2889,
+      "step": 880
+    },
+    {
+      "epoch": 1.69248,
+      "grad_norm": 0.5083025097846985,
+      "learning_rate": 4.742789393103899e-06,
+      "loss": 0.3794,
+      "step": 881
+    },
+    {
+      "epoch": 1.6944,
+      "grad_norm": 0.4711451530456543,
+      "learning_rate": 4.731616851484392e-06,
+      "loss": 0.3062,
+      "step": 882
+    },
+    {
+      "epoch": 1.69632,
+      "grad_norm": 0.5207509994506836,
+      "learning_rate": 4.720445653621429e-06,
+      "loss": 0.3314,
+      "step": 883
+    },
+    {
+      "epoch": 1.69824,
+      "grad_norm": 0.5331737995147705,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.3325,
+      "step": 884
+    },
+    {
+      "epoch": 1.70016,
+      "grad_norm": 0.4853019416332245,
+      "learning_rate": 4.69810751288857e-06,
+      "loss": 0.309,
+      "step": 885
+    },
+    {
+      "epoch": 1.70208,
+      "grad_norm": 1.4900591373443604,
+      "learning_rate": 4.686940681862591e-06,
+      "loss": 0.4636,
+      "step": 886
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 0.47114136815071106,
+      "learning_rate": 4.675775418280432e-06,
+      "loss": 0.2645,
+      "step": 887
+    },
+    {
+      "epoch": 1.7059199999999999,
+      "grad_norm": 0.52579665184021,
+      "learning_rate": 4.664611778044988e-06,
+      "loss": 0.3138,
+      "step": 888
+    },
+    {
+      "epoch": 1.70784,
+      "grad_norm": 0.5394697785377502,
+      "learning_rate": 4.653449817051035e-06,
+      "loss": 0.3229,
+      "step": 889
+    },
+    {
+      "epoch": 1.70976,
+      "grad_norm": 0.4872913956642151,
+      "learning_rate": 4.642289591184934e-06,
+      "loss": 0.3128,
+      "step": 890
+    },
+    {
+      "epoch": 1.7116799999999999,
+      "grad_norm": 0.507679283618927,
+      "learning_rate": 4.631131156324359e-06,
+      "loss": 0.3103,
+      "step": 891
+    },
+    {
+      "epoch": 1.7136,
+      "grad_norm": 0.47934606671333313,
+      "learning_rate": 4.619974568338021e-06,
+      "loss": 0.2919,
+      "step": 892
+    },
+    {
+      "epoch": 1.7155200000000002,
+      "grad_norm": 0.5187336206436157,
+      "learning_rate": 4.60881988308538e-06,
+      "loss": 0.3744,
+      "step": 893
+    },
+    {
+      "epoch": 1.7174399999999999,
+      "grad_norm": 0.5205531120300293,
+      "learning_rate": 4.597667156416371e-06,
+      "loss": 0.3557,
+      "step": 894
+    },
+    {
+      "epoch": 1.71936,
+      "grad_norm": 0.4830275774002075,
+      "learning_rate": 4.586516444171123e-06,
+      "loss": 0.318,
+      "step": 895
+    },
+    {
+      "epoch": 1.7212800000000001,
+      "grad_norm": 0.540939450263977,
+      "learning_rate": 4.575367802179675e-06,
+      "loss": 0.3099,
+      "step": 896
+    },
+    {
+      "epoch": 1.7231999999999998,
+      "grad_norm": 0.49372294545173645,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.3284,
+      "step": 897
+    },
+    {
+      "epoch": 1.72512,
+      "grad_norm": 0.5253046751022339,
+      "learning_rate": 4.553076952226255e-06,
+      "loss": 0.365,
+      "step": 898
+    },
+    {
+      "epoch": 1.7270400000000001,
+      "grad_norm": 0.4972950220108032,
+      "learning_rate": 4.541934855871417e-06,
+      "loss": 0.3325,
+      "step": 899
+    },
+    {
+      "epoch": 1.72896,
+      "grad_norm": 0.4894527196884155,
+      "learning_rate": 4.530795052984104e-06,
+      "loss": 0.3012,
+      "step": 900
+    },
+    {
+      "epoch": 1.73088,
+      "grad_norm": 0.5010990500450134,
+      "learning_rate": 4.519657599339735e-06,
+      "loss": 0.3482,
+      "step": 901
+    },
+    {
+      "epoch": 1.7328000000000001,
+      "grad_norm": 0.5000163316726685,
+      "learning_rate": 4.508522550701965e-06,
+      "loss": 0.3146,
+      "step": 902
+    },
+    {
+      "epoch": 1.73472,
+      "grad_norm": 0.506694495677948,
+      "learning_rate": 4.497389962822416e-06,
+      "loss": 0.3063,
+      "step": 903
+    },
+    {
+      "epoch": 1.73664,
+      "grad_norm": 0.4866572618484497,
+      "learning_rate": 4.48625989144038e-06,
+      "loss": 0.3018,
+      "step": 904
+    },
+    {
+      "epoch": 1.73856,
+      "grad_norm": 0.4997923970222473,
+      "learning_rate": 4.475132392282556e-06,
+      "loss": 0.3405,
+      "step": 905
+    },
+    {
+      "epoch": 1.74048,
+      "grad_norm": 0.49642953276634216,
+      "learning_rate": 4.464007521062761e-06,
+      "loss": 0.2889,
+      "step": 906
+    },
+    {
+      "epoch": 1.7424,
+      "grad_norm": 0.5349805355072021,
+      "learning_rate": 4.452885333481657e-06,
+      "loss": 0.3819,
+      "step": 907
+    },
+    {
+      "epoch": 1.74432,
+      "grad_norm": 0.47898486256599426,
+      "learning_rate": 4.441765885226462e-06,
+      "loss": 0.2453,
+      "step": 908
+    },
+    {
+      "epoch": 1.74624,
+      "grad_norm": 0.46848607063293457,
+      "learning_rate": 4.4306492319706895e-06,
+      "loss": 0.2692,
+      "step": 909
+    },
+    {
+      "epoch": 1.74816,
+      "grad_norm": 0.5090717077255249,
+      "learning_rate": 4.4195354293738484e-06,
+      "loss": 0.3335,
+      "step": 910
+    },
+    {
+      "epoch": 1.75008,
+      "grad_norm": 0.4669017493724823,
+      "learning_rate": 4.4084245330811855e-06,
+      "loss": 0.2828,
+      "step": 911
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 0.5072060823440552,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.353,
+      "step": 912
+    },
+    {
+      "epoch": 1.75392,
+      "grad_norm": 0.5285144448280334,
+      "learning_rate": 4.386211681916309e-06,
+      "loss": 0.3773,
+      "step": 913
+    },
+    {
+      "epoch": 1.75584,
+      "grad_norm": 0.4791277348995209,
+      "learning_rate": 4.3751098382607084e-06,
+      "loss": 0.3071,
+      "step": 914
+    },
+    {
+      "epoch": 1.75776,
+      "grad_norm": 0.5386391878128052,
+      "learning_rate": 4.364011123341947e-06,
+      "loss": 0.3292,
+      "step": 915
+    },
+    {
+      "epoch": 1.75968,
+      "grad_norm": 0.5108262896537781,
+      "learning_rate": 4.352915592729723e-06,
+      "loss": 0.3116,
+      "step": 916
+    },
+    {
+      "epoch": 1.7616,
+      "grad_norm": 0.5218973755836487,
+      "learning_rate": 4.341823301977794e-06,
+      "loss": 0.3909,
+      "step": 917
+    },
+    {
+      "epoch": 1.76352,
+      "grad_norm": 0.4818607270717621,
+      "learning_rate": 4.330734306623694e-06,
+      "loss": 0.2905,
+      "step": 918
+    },
+    {
+      "epoch": 1.76544,
+      "grad_norm": 0.5633077621459961,
+      "learning_rate": 4.319648662188453e-06,
+      "loss": 0.4056,
+      "step": 919
+    },
+    {
+      "epoch": 1.76736,
+      "grad_norm": 0.5384256839752197,
+      "learning_rate": 4.308566424176336e-06,
+      "loss": 0.3148,
+      "step": 920
+    },
+    {
+      "epoch": 1.76928,
+      "grad_norm": 0.5094994306564331,
+      "learning_rate": 4.297487648074538e-06,
+      "loss": 0.3531,
+      "step": 921
+    },
+    {
+      "epoch": 1.7711999999999999,
+      "grad_norm": 0.4517587423324585,
+      "learning_rate": 4.286412389352929e-06,
+      "loss": 0.2856,
+      "step": 922
+    },
+    {
+      "epoch": 1.77312,
+      "grad_norm": 0.5219494700431824,
+      "learning_rate": 4.275340703463767e-06,
+      "loss": 0.4427,
+      "step": 923
+    },
+    {
+      "epoch": 1.77504,
+      "grad_norm": 0.5369223356246948,
+      "learning_rate": 4.264272645841419e-06,
+      "loss": 0.3449,
+      "step": 924
+    },
+    {
+      "epoch": 1.7769599999999999,
+      "grad_norm": 0.5137835144996643,
+      "learning_rate": 4.253208271902091e-06,
+      "loss": 0.2821,
+      "step": 925
+    },
+    {
+      "epoch": 1.77888,
+      "grad_norm": 0.5958774089813232,
+      "learning_rate": 4.242147637043539e-06,
+      "loss": 0.431,
+      "step": 926
+    },
+    {
+      "epoch": 1.7808000000000002,
+      "grad_norm": 0.4370100796222687,
+      "learning_rate": 4.231090796644798e-06,
+      "loss": 0.2622,
+      "step": 927
+    },
+    {
+      "epoch": 1.7827199999999999,
+      "grad_norm": 0.47108185291290283,
+      "learning_rate": 4.220037806065911e-06,
+      "loss": 0.334,
+      "step": 928
+    },
+    {
+      "epoch": 1.78464,
+      "grad_norm": 0.4901743233203888,
+      "learning_rate": 4.208988720647642e-06,
+      "loss": 0.3336,
+      "step": 929
+    },
+    {
+      "epoch": 1.7865600000000001,
+      "grad_norm": 0.5207133293151855,
+      "learning_rate": 4.1979435957111984e-06,
+      "loss": 0.3458,
+      "step": 930
+    },
+    {
+      "epoch": 1.7884799999999998,
+      "grad_norm": 0.49720364809036255,
+      "learning_rate": 4.1869024865579664e-06,
+      "loss": 0.3117,
+      "step": 931
+    },
+    {
+      "epoch": 1.7904,
+      "grad_norm": 0.4790953993797302,
+      "learning_rate": 4.175865448469219e-06,
+      "loss": 0.3201,
+      "step": 932
+    },
+    {
+      "epoch": 1.7923200000000001,
+      "grad_norm": 0.5462205410003662,
+      "learning_rate": 4.164832536705845e-06,
+      "loss": 0.3837,
+      "step": 933
+    },
+    {
+      "epoch": 1.7942399999999998,
+      "grad_norm": 0.48708873987197876,
+      "learning_rate": 4.153803806508083e-06,
+      "loss": 0.2914,
+      "step": 934
+    },
+    {
+      "epoch": 1.79616,
+      "grad_norm": 0.5103388428688049,
+      "learning_rate": 4.142779313095223e-06,
+      "loss": 0.3038,
+      "step": 935
+    },
+    {
+      "epoch": 1.7980800000000001,
+      "grad_norm": 0.4780476689338684,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.2993,
+      "step": 936
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.48637625575065613,
+      "learning_rate": 4.120743257395054e-06,
+      "loss": 0.3423,
+      "step": 937
+    },
+    {
+      "epoch": 1.80192,
+      "grad_norm": 0.50734943151474,
+      "learning_rate": 4.109731805439168e-06,
+      "loss": 0.3457,
+      "step": 938
+    },
+    {
+      "epoch": 1.80384,
+      "grad_norm": 0.5242839455604553,
+      "learning_rate": 4.098724810930472e-06,
+      "loss": 0.3042,
+      "step": 939
+    },
+    {
+      "epoch": 1.80576,
+      "grad_norm": 0.5184866786003113,
+      "learning_rate": 4.087722328979437e-06,
+      "loss": 0.3244,
+      "step": 940
+    },
+    {
+      "epoch": 1.80768,
+      "grad_norm": 0.5346887707710266,
+      "learning_rate": 4.076724414673937e-06,
+      "loss": 0.3707,
+      "step": 941
+    },
+    {
+      "epoch": 1.8096,
+      "grad_norm": 0.48652148246765137,
+      "learning_rate": 4.065731123078977e-06,
+      "loss": 0.3114,
+      "step": 942
+    },
+    {
+      "epoch": 1.81152,
+      "grad_norm": 0.49310141801834106,
+      "learning_rate": 4.054742509236416e-06,
+      "loss": 0.3154,
+      "step": 943
+    },
+    {
+      "epoch": 1.81344,
+      "grad_norm": 0.4925619959831238,
+      "learning_rate": 4.043758628164688e-06,
+      "loss": 0.3192,
+      "step": 944
+    },
+    {
+      "epoch": 1.81536,
+      "grad_norm": 0.4869747459888458,
+      "learning_rate": 4.032779534858544e-06,
+      "loss": 0.3276,
+      "step": 945
+    },
+    {
+      "epoch": 1.81728,
+      "grad_norm": 0.5204039216041565,
+      "learning_rate": 4.021805284288749e-06,
+      "loss": 0.3436,
+      "step": 946
+    },
+    {
+      "epoch": 1.8192,
+      "grad_norm": 0.5143669247627258,
+      "learning_rate": 4.0108359314018254e-06,
+      "loss": 0.3282,
+      "step": 947
+    },
+    {
+      "epoch": 1.82112,
+      "grad_norm": 0.4890744388103485,
+      "learning_rate": 3.999871531119779e-06,
+      "loss": 0.3136,
+      "step": 948
+    },
+    {
+      "epoch": 1.82304,
+      "grad_norm": 0.47771090269088745,
+      "learning_rate": 3.988912138339812e-06,
+      "loss": 0.3219,
+      "step": 949
+    },
+    {
+      "epoch": 1.82496,
+      "grad_norm": 0.5330230593681335,
+      "learning_rate": 3.977957807934055e-06,
+      "loss": 0.347,
+      "step": 950
+    },
+    {
+      "epoch": 1.82688,
+      "grad_norm": 0.5380589365959167,
+      "learning_rate": 3.9670085947493e-06,
+      "loss": 0.3797,
+      "step": 951
+    },
+    {
+      "epoch": 1.8288,
+      "grad_norm": 0.5362987518310547,
+      "learning_rate": 3.956064553606708e-06,
+      "loss": 0.3609,
+      "step": 952
+    },
+    {
+      "epoch": 1.83072,
+      "grad_norm": 0.5475422739982605,
+      "learning_rate": 3.945125739301547e-06,
+      "loss": 0.3282,
+      "step": 953
+    },
+    {
+      "epoch": 1.83264,
+      "grad_norm": 0.5517817139625549,
+      "learning_rate": 3.934192206602921e-06,
+      "loss": 0.3197,
+      "step": 954
+    },
+    {
+      "epoch": 1.83456,
+      "grad_norm": 0.47134190797805786,
+      "learning_rate": 3.923264010253479e-06,
+      "loss": 0.3046,
+      "step": 955
+    },
+    {
+      "epoch": 1.83648,
+      "grad_norm": 0.4881788194179535,
+      "learning_rate": 3.912341204969164e-06,
+      "loss": 0.3033,
+      "step": 956
+    },
+    {
+      "epoch": 1.8384,
+      "grad_norm": 0.5045731067657471,
+      "learning_rate": 3.901423845438916e-06,
+      "loss": 0.3255,
+      "step": 957
+    },
+    {
+      "epoch": 1.84032,
+      "grad_norm": 0.4945167005062103,
+      "learning_rate": 3.890511986324413e-06,
+      "loss": 0.302,
+      "step": 958
+    },
+    {
+      "epoch": 1.8422399999999999,
+      "grad_norm": 0.5825838446617126,
+      "learning_rate": 3.879605682259797e-06,
+      "loss": 0.4033,
+      "step": 959
+    },
+    {
+      "epoch": 1.84416,
+      "grad_norm": 0.5844395160675049,
+      "learning_rate": 3.86870498785139e-06,
+      "loss": 0.4172,
+      "step": 960
+    },
+    {
+      "epoch": 1.8460800000000002,
+      "grad_norm": 0.5345094203948975,
+      "learning_rate": 3.857809957677432e-06,
+      "loss": 0.3232,
+      "step": 961
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 0.49365121126174927,
+      "learning_rate": 3.8469206462878e-06,
+      "loss": 0.2926,
+      "step": 962
+    },
+    {
+      "epoch": 1.84992,
+      "grad_norm": 0.5229520201683044,
+      "learning_rate": 3.83603710820374e-06,
+      "loss": 0.373,
+      "step": 963
+    },
+    {
+      "epoch": 1.8518400000000002,
+      "grad_norm": 0.5441799163818359,
+      "learning_rate": 3.825159397917589e-06,
+      "loss": 0.3603,
+      "step": 964
+    },
+    {
+      "epoch": 1.8537599999999999,
+      "grad_norm": 0.545194685459137,
+      "learning_rate": 3.814287569892512e-06,
+      "loss": 0.3707,
+      "step": 965
+    },
+    {
+      "epoch": 1.85568,
+      "grad_norm": 0.509052574634552,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.3325,
+      "step": 966
+    },
+    {
+      "epoch": 1.8576000000000001,
+      "grad_norm": 0.5212751030921936,
+      "learning_rate": 3.7925617783306757e-06,
+      "loss": 0.3577,
+      "step": 967
+    },
+    {
+      "epoch": 1.8595199999999998,
+      "grad_norm": 0.4693075120449066,
+      "learning_rate": 3.781707923571891e-06,
+      "loss": 0.2897,
+      "step": 968
+    },
+    {
+      "epoch": 1.86144,
+      "grad_norm": 0.48696866631507874,
+      "learning_rate": 3.7708601686295756e-06,
+      "loss": 0.3903,
+      "step": 969
+    },
+    {
+      "epoch": 1.8633600000000001,
+      "grad_norm": 0.46142393350601196,
+      "learning_rate": 3.7600185678169083e-06,
+      "loss": 0.2992,
+      "step": 970
+    },
+    {
+      "epoch": 1.86528,
+      "grad_norm": 0.518894612789154,
+      "learning_rate": 3.7491831754162533e-06,
+      "loss": 0.3321,
+      "step": 971
+    },
+    {
+      "epoch": 1.8672,
+      "grad_norm": 0.5013518333435059,
+      "learning_rate": 3.7383540456788915e-06,
+      "loss": 0.3123,
+      "step": 972
+    },
+    {
+      "epoch": 1.8691200000000001,
+      "grad_norm": 0.5129708647727966,
+      "learning_rate": 3.727531232824747e-06,
+      "loss": 0.3454,
+      "step": 973
+    },
+    {
+      "epoch": 1.87104,
+      "grad_norm": 0.5593523979187012,
+      "learning_rate": 3.7167147910421165e-06,
+      "loss": 0.3901,
+      "step": 974
+    },
+    {
+      "epoch": 1.87296,
+      "grad_norm": 0.585111141204834,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.3995,
+      "step": 975
+    },
+    {
+      "epoch": 1.87488,
+      "grad_norm": 0.4892120957374573,
+      "learning_rate": 3.695101237284815e-06,
+      "loss": 0.3141,
+      "step": 976
+    },
+    {
+      "epoch": 1.8768,
+      "grad_norm": 0.5076225399971008,
+      "learning_rate": 3.6843042335261583e-06,
+      "loss": 0.3104,
+      "step": 977
+    },
+    {
+      "epoch": 1.87872,
+      "grad_norm": 0.4951135814189911,
+      "learning_rate": 3.6735138172704967e-06,
+      "loss": 0.3481,
+      "step": 978
+    },
+    {
+      "epoch": 1.88064,
+      "grad_norm": 0.5135895013809204,
+      "learning_rate": 3.662730042543926e-06,
+      "loss": 0.3486,
+      "step": 979
+    },
+    {
+      "epoch": 1.88256,
+      "grad_norm": 0.5862326622009277,
+      "learning_rate": 3.6519529633392825e-06,
+      "loss": 0.4086,
+      "step": 980
+    },
+    {
+      "epoch": 1.88448,
+      "grad_norm": 0.5124054551124573,
+      "learning_rate": 3.6411826336158785e-06,
+      "loss": 0.3064,
+      "step": 981
+    },
+    {
+      "epoch": 1.8864,
+      "grad_norm": 0.5450282096862793,
+      "learning_rate": 3.6304191072992376e-06,
+      "loss": 0.2662,
+      "step": 982
+    },
+    {
+      "epoch": 1.88832,
+      "grad_norm": 0.5234136581420898,
+      "learning_rate": 3.619662438280816e-06,
+      "loss": 0.289,
+      "step": 983
+    },
+    {
+      "epoch": 1.89024,
+      "grad_norm": 0.5349469780921936,
+      "learning_rate": 3.6089126804177373e-06,
+      "loss": 0.3375,
+      "step": 984
+    },
+    {
+      "epoch": 1.89216,
+      "grad_norm": 0.49325162172317505,
+      "learning_rate": 3.5981698875325214e-06,
+      "loss": 0.2585,
+      "step": 985
+    },
+    {
+      "epoch": 1.89408,
+      "grad_norm": 0.4329458475112915,
+      "learning_rate": 3.5874341134128156e-06,
+      "loss": 0.2449,
+      "step": 986
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 0.48916736245155334,
+      "learning_rate": 3.5767054118111266e-06,
+      "loss": 0.304,
+      "step": 987
+    },
+    {
+      "epoch": 1.89792,
+      "grad_norm": 0.5307055115699768,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.3585,
+      "step": 988
+    },
+    {
+      "epoch": 1.89984,
+      "grad_norm": 0.4914647042751312,
+      "learning_rate": 3.555269440994496e-06,
+      "loss": 0.2906,
+      "step": 989
+    },
+    {
+      "epoch": 1.90176,
+      "grad_norm": 0.49873530864715576,
+      "learning_rate": 3.544562279106436e-06,
+      "loss": 0.375,
+      "step": 990
+    },
+    {
+      "epoch": 1.90368,
+      "grad_norm": 0.48357006907463074,
+      "learning_rate": 3.5338624043896154e-06,
+      "loss": 0.3035,
+      "step": 991
+    },
+    {
+      "epoch": 1.9056,
+      "grad_norm": 0.5402414202690125,
+      "learning_rate": 3.523169870416795e-06,
+      "loss": 0.3453,
+      "step": 992
+    },
+    {
+      "epoch": 1.9075199999999999,
+      "grad_norm": 0.49305790662765503,
+      "learning_rate": 3.5124847307239863e-06,
+      "loss": 0.2936,
+      "step": 993
+    },
+    {
+      "epoch": 1.90944,
+      "grad_norm": 0.574478268623352,
+      "learning_rate": 3.501807038810174e-06,
+      "loss": 0.3802,
+      "step": 994
+    },
+    {
+      "epoch": 1.91136,
+      "grad_norm": 0.5989776253700256,
+      "learning_rate": 3.4911368481370535e-06,
+      "loss": 0.3317,
+      "step": 995
+    },
+    {
+      "epoch": 1.9132799999999999,
+      "grad_norm": 0.4545533061027527,
+      "learning_rate": 3.480474212128766e-06,
+      "loss": 0.2712,
+      "step": 996
+    },
+    {
+      "epoch": 1.9152,
+      "grad_norm": 0.5139933824539185,
+      "learning_rate": 3.469819184171623e-06,
+      "loss": 0.3247,
+      "step": 997
+    },
+    {
+      "epoch": 1.9171200000000002,
+      "grad_norm": 0.48741185665130615,
+      "learning_rate": 3.459171817613847e-06,
+      "loss": 0.3142,
+      "step": 998
+    },
+    {
+      "epoch": 1.9190399999999999,
+      "grad_norm": 0.5124534368515015,
+      "learning_rate": 3.4485321657653e-06,
+      "loss": 0.3578,
+      "step": 999
+    },
+    {
+      "epoch": 1.92096,
+      "grad_norm": 0.6058474779129028,
+      "learning_rate": 3.4379002818972122e-06,
+      "loss": 0.3741,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9228800000000001,
+      "grad_norm": 0.48686233162879944,
+      "learning_rate": 3.427276219241933e-06,
+      "loss": 0.2925,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9247999999999998,
+      "grad_norm": 0.5733014345169067,
+      "learning_rate": 3.416660030992639e-06,
+      "loss": 0.402,
+      "step": 1002
+    },
+    {
+      "epoch": 1.92672,
+      "grad_norm": 0.5262948274612427,
+      "learning_rate": 3.406051770303087e-06,
+      "loss": 0.3351,
+      "step": 1003
+    },
+    {
+      "epoch": 1.9286400000000001,
+      "grad_norm": 0.5112277269363403,
+      "learning_rate": 3.3954514902873427e-06,
+      "loss": 0.3222,
+      "step": 1004
+    },
+    {
+      "epoch": 1.93056,
+      "grad_norm": 0.49718227982521057,
+      "learning_rate": 3.3848592440195118e-06,
+      "loss": 0.3446,
+      "step": 1005
+    },
+    {
+      "epoch": 1.93248,
+      "grad_norm": 0.45362454652786255,
+      "learning_rate": 3.3742750845334748e-06,
+      "loss": 0.239,
+      "step": 1006
+    },
+    {
+      "epoch": 1.9344000000000001,
+      "grad_norm": 0.5111440420150757,
+      "learning_rate": 3.3636990648226258e-06,
+      "loss": 0.3708,
+      "step": 1007
+    },
+    {
+      "epoch": 1.93632,
+      "grad_norm": 0.5044022798538208,
+      "learning_rate": 3.3531312378396026e-06,
+      "loss": 0.3217,
+      "step": 1008
+    },
+    {
+      "epoch": 1.93824,
+      "grad_norm": 0.5394229292869568,
+      "learning_rate": 3.342571656496022e-06,
+      "loss": 0.3486,
+      "step": 1009
+    },
+    {
+      "epoch": 1.94016,
+      "grad_norm": 0.48527249693870544,
+      "learning_rate": 3.3320203736622185e-06,
+      "loss": 0.312,
+      "step": 1010
+    },
+    {
+      "epoch": 1.94208,
+      "grad_norm": 0.5112845301628113,
+      "learning_rate": 3.3214774421669777e-06,
+      "loss": 0.3368,
+      "step": 1011
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 0.572320282459259,
+      "learning_rate": 3.310942914797265e-06,
+      "loss": 0.3224,
+      "step": 1012
+    },
+    {
+      "epoch": 1.94592,
+      "grad_norm": 0.5474165081977844,
+      "learning_rate": 3.3004168442979755e-06,
+      "loss": 0.329,
+      "step": 1013
+    },
+    {
+      "epoch": 1.94784,
+      "grad_norm": 0.4958500266075134,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.3306,
+      "step": 1014
+    },
+    {
+      "epoch": 1.94976,
+      "grad_norm": 0.4885128140449524,
+      "learning_rate": 3.2793902846782534e-06,
+      "loss": 0.2881,
+      "step": 1015
+    },
+    {
+      "epoch": 1.95168,
+      "grad_norm": 0.5370290875434875,
+      "learning_rate": 3.2688899008348386e-06,
+      "loss": 0.3734,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9536,
+      "grad_norm": 0.5374308824539185,
+      "learning_rate": 3.2583981844153487e-06,
+      "loss": 0.319,
+      "step": 1017
+    },
+    {
+      "epoch": 1.95552,
+      "grad_norm": 0.4791926443576813,
+      "learning_rate": 3.2479151879503324e-06,
+      "loss": 0.3139,
+      "step": 1018
+    },
+    {
+      "epoch": 1.95744,
+      "grad_norm": 0.4811853766441345,
+      "learning_rate": 3.2374409639266695e-06,
+      "loss": 0.3214,
+      "step": 1019
+    },
+    {
+      "epoch": 1.95936,
+      "grad_norm": 0.5252090096473694,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.3168,
+      "step": 1020
+    },
+    {
+      "epoch": 1.96128,
+      "grad_norm": 0.5361338257789612,
+      "learning_rate": 3.2165190429310674e-06,
+      "loss": 0.3294,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9632,
+      "grad_norm": 0.5056690573692322,
+      "learning_rate": 3.206071450712235e-06,
+      "loss": 0.3279,
+      "step": 1022
+    },
+    {
+      "epoch": 1.96512,
+      "grad_norm": 0.5278884768486023,
+      "learning_rate": 3.1956328404404403e-06,
+      "loss": 0.3742,
+      "step": 1023
+    },
+    {
+      "epoch": 1.96704,
+      "grad_norm": 0.475342333316803,
+      "learning_rate": 3.1852032643803377e-06,
+      "loss": 0.3148,
+      "step": 1024
+    },
+    {
+      "epoch": 1.96896,
+      "grad_norm": 0.47536763548851013,
+      "learning_rate": 3.174782774751338e-06,
+      "loss": 0.2992,
+      "step": 1025
+    },
+    {
+      "epoch": 1.97088,
+      "grad_norm": 0.47321686148643494,
+      "learning_rate": 3.1643714237273628e-06,
+      "loss": 0.2885,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9727999999999999,
+      "grad_norm": 0.5139946937561035,
+      "learning_rate": 3.1539692634365788e-06,
+      "loss": 0.2855,
+      "step": 1027
+    },
+    {
+      "epoch": 1.97472,
+      "grad_norm": 0.47628602385520935,
+      "learning_rate": 3.143576345961132e-06,
+      "loss": 0.2957,
+      "step": 1028
+    },
+    {
+      "epoch": 1.97664,
+      "grad_norm": 0.4636322855949402,
+      "learning_rate": 3.1331927233368954e-06,
+      "loss": 0.3045,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9785599999999999,
+      "grad_norm": 0.5466511845588684,
+      "learning_rate": 3.1228184475532015e-06,
+      "loss": 0.2908,
+      "step": 1030
+    },
+    {
+      "epoch": 1.98048,
+      "grad_norm": 0.5866350531578064,
+      "learning_rate": 3.112453570552583e-06,
+      "loss": 0.3688,
+      "step": 1031
+    },
+    {
+      "epoch": 1.9824000000000002,
+      "grad_norm": 0.5204470157623291,
+      "learning_rate": 3.1020981442305187e-06,
+      "loss": 0.3661,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9843199999999999,
+      "grad_norm": 0.509781002998352,
+      "learning_rate": 3.091752220435166e-06,
+      "loss": 0.3282,
+      "step": 1033
+    },
+    {
+      "epoch": 1.98624,
+      "grad_norm": 0.42670103907585144,
+      "learning_rate": 3.0814158509671015e-06,
+      "loss": 0.2355,
+      "step": 1034
+    },
+    {
+      "epoch": 1.9881600000000001,
+      "grad_norm": 0.5641727447509766,
+      "learning_rate": 3.0710890875790745e-06,
+      "loss": 0.3576,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9900799999999998,
+      "grad_norm": 0.571896493434906,
+      "learning_rate": 3.0607719819757264e-06,
+      "loss": 0.3511,
+      "step": 1036
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 0.5320600271224976,
+      "learning_rate": 3.0504645858133507e-06,
+      "loss": 0.3242,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9939200000000001,
+      "grad_norm": 0.48054492473602295,
+      "learning_rate": 3.040166950699626e-06,
+      "loss": 0.2713,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9958399999999998,
+      "grad_norm": 0.5641902685165405,
+      "learning_rate": 3.029879128193356e-06,
+      "loss": 0.3373,
+      "step": 1039
+    },
+    {
+      "epoch": 1.99776,
+      "grad_norm": 0.5585317611694336,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.4263,
+      "step": 1040
+    },
+    {
+      "epoch": 2.00064,
+      "grad_norm": 0.9810692071914673,
+      "learning_rate": 3.0093331269924954e-06,
+      "loss": 0.6113,
+      "step": 1041
+    },
+    {
+      "epoch": 2.00256,
+      "grad_norm": 0.5089061856269836,
+      "learning_rate": 2.9990750511688323e-06,
+      "loss": 0.292,
+      "step": 1042
+    },
+    {
+      "epoch": 2.00448,
+      "grad_norm": 0.5237654447555542,
+      "learning_rate": 2.988826993693967e-06,
+      "loss": 0.2982,
+      "step": 1043
+    },
+    {
+      "epoch": 2.0064,
+      "grad_norm": 0.4843851327896118,
+      "learning_rate": 2.978589005878476e-06,
+      "loss": 0.2722,
+      "step": 1044
+    },
+    {
+      "epoch": 2.00832,
+      "grad_norm": 0.6039709448814392,
+      "learning_rate": 2.968361138982517e-06,
+      "loss": 0.3596,
+      "step": 1045
+    },
+    {
+      "epoch": 2.01024,
+      "grad_norm": 0.48284614086151123,
+      "learning_rate": 2.9581434442155798e-06,
+      "loss": 0.2729,
+      "step": 1046
+    },
+    {
+      "epoch": 2.01216,
+      "grad_norm": 0.4798407256603241,
+      "learning_rate": 2.947935972736217e-06,
+      "loss": 0.2764,
+      "step": 1047
+    },
+    {
+      "epoch": 2.01408,
+      "grad_norm": 0.4949873089790344,
+      "learning_rate": 2.937738775651798e-06,
+      "loss": 0.2927,
+      "step": 1048
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 0.509315550327301,
+      "learning_rate": 2.9275519040182503e-06,
+      "loss": 0.2819,
+      "step": 1049
+    },
+    {
+      "epoch": 2.01792,
+      "grad_norm": 0.5088322758674622,
+      "learning_rate": 2.917375408839803e-06,
+      "loss": 0.2797,
+      "step": 1050
+    },
+    {
+      "epoch": 2.01984,
+      "grad_norm": 0.5544795393943787,
+      "learning_rate": 2.9072093410687268e-06,
+      "loss": 0.2589,
+      "step": 1051
+    },
+    {
+      "epoch": 2.02176,
+      "grad_norm": 0.5295831561088562,
+      "learning_rate": 2.8970537516050935e-06,
+      "loss": 0.2305,
+      "step": 1052
+    },
+    {
+      "epoch": 2.02368,
+      "grad_norm": 0.543292760848999,
+      "learning_rate": 2.886908691296504e-06,
+      "loss": 0.3097,
+      "step": 1053
+    },
+    {
+      "epoch": 2.0256,
+      "grad_norm": 0.5447037220001221,
+      "learning_rate": 2.876774210937843e-06,
+      "loss": 0.3209,
+      "step": 1054
+    },
+    {
+      "epoch": 2.02752,
+      "grad_norm": 0.49107998609542847,
+      "learning_rate": 2.866650361271023e-06,
+      "loss": 0.2403,
+      "step": 1055
+    },
+    {
+      "epoch": 2.02944,
+      "grad_norm": 0.5548592805862427,
+      "learning_rate": 2.8565371929847286e-06,
+      "loss": 0.292,
+      "step": 1056
+    },
+    {
+      "epoch": 2.03136,
+      "grad_norm": 0.5970434546470642,
+      "learning_rate": 2.84643475671417e-06,
+      "loss": 0.308,
+      "step": 1057
+    },
+    {
+      "epoch": 2.03328,
+      "grad_norm": 0.5376215577125549,
+      "learning_rate": 2.836343103040819e-06,
+      "loss": 0.2446,
+      "step": 1058
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 0.5482760667800903,
+      "learning_rate": 2.8262622824921593e-06,
+      "loss": 0.307,
+      "step": 1059
+    },
+    {
+      "epoch": 2.03712,
+      "grad_norm": 0.5306827425956726,
+      "learning_rate": 2.816192345541437e-06,
+      "loss": 0.2916,
+      "step": 1060
+    },
+    {
+      "epoch": 2.03904,
+      "grad_norm": 0.5424799919128418,
+      "learning_rate": 2.8061333426074054e-06,
+      "loss": 0.3718,
+      "step": 1061
+    },
+    {
+      "epoch": 2.04096,
+      "grad_norm": 0.5038452744483948,
+      "learning_rate": 2.7960853240540703e-06,
+      "loss": 0.2913,
+      "step": 1062
+    },
+    {
+      "epoch": 2.04288,
+      "grad_norm": 0.5448192954063416,
+      "learning_rate": 2.7860483401904483e-06,
+      "loss": 0.31,
+      "step": 1063
+    },
+    {
+      "epoch": 2.0448,
+      "grad_norm": 0.4696711599826813,
+      "learning_rate": 2.776022441270295e-06,
+      "loss": 0.2149,
+      "step": 1064
+    },
+    {
+      "epoch": 2.04672,
+      "grad_norm": 0.4605139195919037,
+      "learning_rate": 2.766007677491871e-06,
+      "loss": 0.2051,
+      "step": 1065
+    },
+    {
+      "epoch": 2.04864,
+      "grad_norm": 0.48666587471961975,
+      "learning_rate": 2.7560040989976894e-06,
+      "loss": 0.2284,
+      "step": 1066
+    },
+    {
+      "epoch": 2.05056,
+      "grad_norm": 0.53001469373703,
+      "learning_rate": 2.7460117558742532e-06,
+      "loss": 0.2795,
+      "step": 1067
+    },
+    {
+      "epoch": 2.05248,
+      "grad_norm": 0.5847320556640625,
+      "learning_rate": 2.736030698151815e-06,
+      "loss": 0.3135,
+      "step": 1068
+    },
+    {
+      "epoch": 2.0544,
+      "grad_norm": 0.5066387057304382,
+      "learning_rate": 2.72606097580412e-06,
+      "loss": 0.2861,
+      "step": 1069
+    },
+    {
+      "epoch": 2.05632,
+      "grad_norm": 0.4717126190662384,
+      "learning_rate": 2.7161026387481636e-06,
+      "loss": 0.2277,
+      "step": 1070
+    },
+    {
+      "epoch": 2.05824,
+      "grad_norm": 0.5429806709289551,
+      "learning_rate": 2.7061557368439294e-06,
+      "loss": 0.3203,
+      "step": 1071
+    },
+    {
+      "epoch": 2.06016,
+      "grad_norm": 0.4770486354827881,
+      "learning_rate": 2.6962203198941587e-06,
+      "loss": 0.2398,
+      "step": 1072
+    },
+    {
+      "epoch": 2.06208,
+      "grad_norm": 0.5559279918670654,
+      "learning_rate": 2.686296437644074e-06,
+      "loss": 0.2936,
+      "step": 1073
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 0.5208328366279602,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.2434,
+      "step": 1074
+    },
+    {
+      "epoch": 2.06592,
+      "grad_norm": 0.5110089182853699,
+      "learning_rate": 2.666483475934885e-06,
+      "loss": 0.2871,
+      "step": 1075
+    },
+    {
+      "epoch": 2.06784,
+      "grad_norm": 0.49105504155158997,
+      "learning_rate": 2.656594495676482e-06,
+      "loss": 0.2733,
+      "step": 1076
+    },
+    {
+      "epoch": 2.06976,
+      "grad_norm": 0.5164961218833923,
+      "learning_rate": 2.6467172485186775e-06,
+      "loss": 0.2603,
+      "step": 1077
+    },
+    {
+      "epoch": 2.07168,
+      "grad_norm": 0.4882449209690094,
+      "learning_rate": 2.636851783915454e-06,
+      "loss": 0.2343,
+      "step": 1078
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.5346804857254028,
+      "learning_rate": 2.626998151261798e-06,
+      "loss": 0.269,
+      "step": 1079
+    },
+    {
+      "epoch": 2.07552,
+      "grad_norm": 0.5632084012031555,
+      "learning_rate": 2.6171563998934605e-06,
+      "loss": 0.2986,
+      "step": 1080
+    },
+    {
+      "epoch": 2.07744,
+      "grad_norm": 0.5827323198318481,
+      "learning_rate": 2.607326579086701e-06,
+      "loss": 0.3342,
+      "step": 1081
+    },
+    {
+      "epoch": 2.07936,
+      "grad_norm": 0.5047206878662109,
+      "learning_rate": 2.5975087380580445e-06,
+      "loss": 0.2449,
+      "step": 1082
+    },
+    {
+      "epoch": 2.08128,
+      "grad_norm": 0.480937123298645,
+      "learning_rate": 2.587702925964034e-06,
+      "loss": 0.2155,
+      "step": 1083
+    },
+    {
+      "epoch": 2.0832,
+      "grad_norm": 0.4867021143436432,
+      "learning_rate": 2.577909191900988e-06,
+      "loss": 0.2529,
+      "step": 1084
+    },
+    {
+      "epoch": 2.08512,
+      "grad_norm": 0.5163612961769104,
+      "learning_rate": 2.5681275849047482e-06,
+      "loss": 0.2653,
+      "step": 1085
+    },
+    {
+      "epoch": 2.08704,
+      "grad_norm": 0.4591393768787384,
+      "learning_rate": 2.5583581539504464e-06,
+      "loss": 0.2421,
+      "step": 1086
+    },
+    {
+      "epoch": 2.08896,
+      "grad_norm": 0.48243802785873413,
+      "learning_rate": 2.5486009479522355e-06,
+      "loss": 0.2537,
+      "step": 1087
+    },
+    {
+      "epoch": 2.09088,
+      "grad_norm": 0.5273064970970154,
+      "learning_rate": 2.5388560157630765e-06,
+      "loss": 0.2691,
+      "step": 1088
+    },
+    {
+      "epoch": 2.0928,
+      "grad_norm": 0.4881844222545624,
+      "learning_rate": 2.5291234061744655e-06,
+      "loss": 0.267,
+      "step": 1089
+    },
+    {
+      "epoch": 2.09472,
+      "grad_norm": 0.5497365593910217,
+      "learning_rate": 2.519403167916207e-06,
+      "loss": 0.3153,
+      "step": 1090
+    },
+    {
+      "epoch": 2.09664,
+      "grad_norm": 0.5377917289733887,
+      "learning_rate": 2.50969534965616e-06,
+      "loss": 0.2879,
+      "step": 1091
+    },
+    {
+      "epoch": 2.09856,
+      "grad_norm": 0.5131589770317078,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.2842,
+      "step": 1092
+    },
+    {
+      "epoch": 2.10048,
+      "grad_norm": 0.5272597670555115,
+      "learning_rate": 2.490317167490976e-06,
+      "loss": 0.2807,
+      "step": 1093
+    },
+    {
+      "epoch": 2.1024,
+      "grad_norm": 0.5617088675498962,
+      "learning_rate": 2.480646900609664e-06,
+      "loss": 0.3296,
+      "step": 1094
+    },
+    {
+      "epoch": 2.10432,
+      "grad_norm": 0.4871920645236969,
+      "learning_rate": 2.4709892477737263e-06,
+      "loss": 0.257,
+      "step": 1095
+    },
+    {
+      "epoch": 2.10624,
+      "grad_norm": 0.4916019141674042,
+      "learning_rate": 2.4613442573376625e-06,
+      "loss": 0.2445,
+      "step": 1096
+    },
+    {
+      "epoch": 2.10816,
+      "grad_norm": 0.5486736297607422,
+      "learning_rate": 2.4517119775925824e-06,
+      "loss": 0.2901,
+      "step": 1097
+    },
+    {
+      "epoch": 2.11008,
+      "grad_norm": 0.5128265619277954,
+      "learning_rate": 2.4420924567659508e-06,
+      "loss": 0.2832,
+      "step": 1098
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.5287452340126038,
+      "learning_rate": 2.4324857430213504e-06,
+      "loss": 0.2665,
+      "step": 1099
+    },
+    {
+      "epoch": 2.11392,
+      "grad_norm": 0.49356284737586975,
+      "learning_rate": 2.422891884458241e-06,
+      "loss": 0.2753,
+      "step": 1100
+    },
+    {
+      "epoch": 2.11584,
+      "grad_norm": 0.5281261205673218,
+      "learning_rate": 2.4133109291117156e-06,
+      "loss": 0.2949,
+      "step": 1101
+    },
+    {
+      "epoch": 2.11776,
+      "grad_norm": 0.5272106528282166,
+      "learning_rate": 2.4037429249522702e-06,
+      "loss": 0.2848,
+      "step": 1102
+    },
+    {
+      "epoch": 2.11968,
+      "grad_norm": 0.5452120304107666,
+      "learning_rate": 2.394187919885548e-06,
+      "loss": 0.3272,
+      "step": 1103
+    },
+    {
+      "epoch": 2.1216,
+      "grad_norm": 0.5306313037872314,
+      "learning_rate": 2.384645961752113e-06,
+      "loss": 0.2829,
+      "step": 1104
+    },
+    {
+      "epoch": 2.12352,
+      "grad_norm": 0.5973362326622009,
+      "learning_rate": 2.3751170983272e-06,
+      "loss": 0.3881,
+      "step": 1105
+    },
+    {
+      "epoch": 2.12544,
+      "grad_norm": 0.5198131203651428,
+      "learning_rate": 2.3656013773204843e-06,
+      "loss": 0.2631,
+      "step": 1106
+    },
+    {
+      "epoch": 2.12736,
+      "grad_norm": 0.5155800580978394,
+      "learning_rate": 2.3560988463758366e-06,
+      "loss": 0.3125,
+      "step": 1107
+    },
+    {
+      "epoch": 2.12928,
+      "grad_norm": 0.5644299387931824,
+      "learning_rate": 2.346609553071093e-06,
+      "loss": 0.3214,
+      "step": 1108
+    },
+    {
+      "epoch": 2.1312,
+      "grad_norm": 0.5407156944274902,
+      "learning_rate": 2.3371335449178006e-06,
+      "loss": 0.2575,
+      "step": 1109
+    },
+    {
+      "epoch": 2.13312,
+      "grad_norm": 0.534443199634552,
+      "learning_rate": 2.3276708693609947e-06,
+      "loss": 0.2452,
+      "step": 1110
+    },
+    {
+      "epoch": 2.13504,
+      "grad_norm": 0.5076003074645996,
+      "learning_rate": 2.3182215737789593e-06,
+      "loss": 0.2387,
+      "step": 1111
+    },
+    {
+      "epoch": 2.13696,
+      "grad_norm": 0.5253971815109253,
+      "learning_rate": 2.308785705482982e-06,
+      "loss": 0.2937,
+      "step": 1112
+    },
+    {
+      "epoch": 2.13888,
+      "grad_norm": 0.5186853408813477,
+      "learning_rate": 2.2993633117171243e-06,
+      "loss": 0.2991,
+      "step": 1113
+    },
+    {
+      "epoch": 2.1408,
+      "grad_norm": 0.5558339357376099,
+      "learning_rate": 2.289954439657981e-06,
+      "loss": 0.3091,
+      "step": 1114
+    },
+    {
+      "epoch": 2.14272,
+      "grad_norm": 0.5345395803451538,
+      "learning_rate": 2.2805591364144446e-06,
+      "loss": 0.2954,
+      "step": 1115
+    },
+    {
+      "epoch": 2.14464,
+      "grad_norm": 0.4834584593772888,
+      "learning_rate": 2.2711774490274767e-06,
+      "loss": 0.2441,
+      "step": 1116
+    },
+    {
+      "epoch": 2.14656,
+      "grad_norm": 0.5163441300392151,
+      "learning_rate": 2.2618094244698614e-06,
+      "loss": 0.2508,
+      "step": 1117
+    },
+    {
+      "epoch": 2.14848,
+      "grad_norm": 0.5603058338165283,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 0.3042,
+      "step": 1118
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.5417174100875854,
+      "learning_rate": 2.243114551391542e-06,
+      "loss": 0.3124,
+      "step": 1119
+    },
+    {
+      "epoch": 2.15232,
+      "grad_norm": 0.5522037744522095,
+      "learning_rate": 2.2337877964734324e-06,
+      "loss": 0.3128,
+      "step": 1120
+    },
+    {
+      "epoch": 2.15424,
+      "grad_norm": 0.5324065685272217,
+      "learning_rate": 2.224474891589386e-06,
+      "loss": 0.2875,
+      "step": 1121
+    },
+    {
+      "epoch": 2.15616,
+      "grad_norm": 0.5070877075195312,
+      "learning_rate": 2.2151758833678044e-06,
+      "loss": 0.265,
+      "step": 1122
+    },
+    {
+      "epoch": 2.15808,
+      "grad_norm": 0.598729133605957,
+      "learning_rate": 2.205890818367508e-06,
+      "loss": 0.2975,
+      "step": 1123
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.496507465839386,
+      "learning_rate": 2.1966197430775056e-06,
+      "loss": 0.2429,
+      "step": 1124
+    },
+    {
+      "epoch": 2.16192,
+      "grad_norm": 0.5321954488754272,
+      "learning_rate": 2.187362703916766e-06,
+      "loss": 0.2986,
+      "step": 1125
+    },
+    {
+      "epoch": 2.16384,
+      "grad_norm": 0.4799526035785675,
+      "learning_rate": 2.178119747233976e-06,
+      "loss": 0.237,
+      "step": 1126
+    },
+    {
+      "epoch": 2.16576,
+      "grad_norm": 0.5974825620651245,
+      "learning_rate": 2.168890919307315e-06,
+      "loss": 0.3313,
+      "step": 1127
+    },
+    {
+      "epoch": 2.16768,
+      "grad_norm": 0.5782894492149353,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.2724,
+      "step": 1128
+    },
+    {
+      "epoch": 2.1696,
+      "grad_norm": 0.5559664964675903,
+      "learning_rate": 2.1504758344811615e-06,
+      "loss": 0.2844,
+      "step": 1129
+    },
+    {
+      "epoch": 2.17152,
+      "grad_norm": 0.5257934927940369,
+      "learning_rate": 2.141289669783401e-06,
+      "loss": 0.2583,
+      "step": 1130
+    },
+    {
+      "epoch": 2.17344,
+      "grad_norm": 0.5752183794975281,
+      "learning_rate": 2.132117818244771e-06,
+      "loss": 0.3677,
+      "step": 1131
+    },
+    {
+      "epoch": 2.17536,
+      "grad_norm": 0.5480060577392578,
+      "learning_rate": 2.122960325787432e-06,
+      "loss": 0.2846,
+      "step": 1132
+    },
+    {
+      "epoch": 2.17728,
+      "grad_norm": 0.5211611986160278,
+      "learning_rate": 2.1138172382616612e-06,
+      "loss": 0.2487,
+      "step": 1133
+    },
+    {
+      "epoch": 2.1792,
+      "grad_norm": 0.5457812547683716,
+      "learning_rate": 2.104688601445606e-06,
+      "loss": 0.2932,
+      "step": 1134
+    },
+    {
+      "epoch": 2.18112,
+      "grad_norm": 0.5075328946113586,
+      "learning_rate": 2.0955744610450616e-06,
+      "loss": 0.236,
+      "step": 1135
+    },
+    {
+      "epoch": 2.18304,
+      "grad_norm": 0.5714215040206909,
+      "learning_rate": 2.086474862693244e-06,
+      "loss": 0.3177,
+      "step": 1136
+    },
+    {
+      "epoch": 2.1849600000000002,
+      "grad_norm": 0.5517022609710693,
+      "learning_rate": 2.077389851950557e-06,
+      "loss": 0.2541,
+      "step": 1137
+    },
+    {
+      "epoch": 2.18688,
+      "grad_norm": 0.5161548256874084,
+      "learning_rate": 2.068319474304365e-06,
+      "loss": 0.316,
+      "step": 1138
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.5095581412315369,
+      "learning_rate": 2.059263775168773e-06,
+      "loss": 0.2821,
+      "step": 1139
+    },
+    {
+      "epoch": 2.19072,
+      "grad_norm": 0.547309935092926,
+      "learning_rate": 2.050222799884387e-06,
+      "loss": 0.3012,
+      "step": 1140
+    },
+    {
+      "epoch": 2.19264,
+      "grad_norm": 0.5637189745903015,
+      "learning_rate": 2.0411965937180907e-06,
+      "loss": 0.3268,
+      "step": 1141
+    },
+    {
+      "epoch": 2.19456,
+      "grad_norm": 0.48354989290237427,
+      "learning_rate": 2.0321852018628278e-06,
+      "loss": 0.2227,
+      "step": 1142
+    },
+    {
+      "epoch": 2.19648,
+      "grad_norm": 0.5226564407348633,
+      "learning_rate": 2.0231886694373653e-06,
+      "loss": 0.3259,
+      "step": 1143
+    },
+    {
+      "epoch": 2.1984,
+      "grad_norm": 0.4894440472126007,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.2324,
+      "step": 1144
+    },
+    {
+      "epoch": 2.20032,
+      "grad_norm": 0.5285873413085938,
+      "learning_rate": 2.005240362978686e-06,
+      "loss": 0.2593,
+      "step": 1145
+    },
+    {
+      "epoch": 2.20224,
+      "grad_norm": 0.5923100709915161,
+      "learning_rate": 1.996288678810105e-06,
+      "loss": 0.3179,
+      "step": 1146
+    },
+    {
+      "epoch": 2.20416,
+      "grad_norm": 0.48019418120384216,
+      "learning_rate": 1.98735203380015e-06,
+      "loss": 0.2501,
+      "step": 1147
+    },
+    {
+      "epoch": 2.20608,
+      "grad_norm": 0.4968745708465576,
+      "learning_rate": 1.9784304726933384e-06,
+      "loss": 0.3037,
+      "step": 1148
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 0.5125188231468201,
+      "learning_rate": 1.9695240401586687e-06,
+      "loss": 0.2821,
+      "step": 1149
+    },
+    {
+      "epoch": 2.20992,
+      "grad_norm": 0.5009004473686218,
+      "learning_rate": 1.9606327807893905e-06,
+      "loss": 0.2633,
+      "step": 1150
+    },
+    {
+      "epoch": 2.21184,
+      "grad_norm": 0.5285592079162598,
+      "learning_rate": 1.9517567391027853e-06,
+      "loss": 0.315,
+      "step": 1151
+    },
+    {
+      "epoch": 2.21376,
+      "grad_norm": 0.5298715829849243,
+      "learning_rate": 1.942895959539939e-06,
+      "loss": 0.2805,
+      "step": 1152
+    },
+    {
+      "epoch": 2.21568,
+      "grad_norm": 0.5409754514694214,
+      "learning_rate": 1.934050486465529e-06,
+      "loss": 0.322,
+      "step": 1153
+    },
+    {
+      "epoch": 2.2176,
+      "grad_norm": 0.5477048754692078,
+      "learning_rate": 1.9252203641675854e-06,
+      "loss": 0.2354,
+      "step": 1154
+    },
+    {
+      "epoch": 2.21952,
+      "grad_norm": 0.5203359127044678,
+      "learning_rate": 1.9164056368572847e-06,
+      "loss": 0.224,
+      "step": 1155
+    },
+    {
+      "epoch": 2.22144,
+      "grad_norm": 0.5266945362091064,
+      "learning_rate": 1.9076063486687256e-06,
+      "loss": 0.2508,
+      "step": 1156
+    },
+    {
+      "epoch": 2.22336,
+      "grad_norm": 0.544944703578949,
+      "learning_rate": 1.8988225436587005e-06,
+      "loss": 0.2899,
+      "step": 1157
+    },
+    {
+      "epoch": 2.22528,
+      "grad_norm": 0.5391412377357483,
+      "learning_rate": 1.8900542658064807e-06,
+      "loss": 0.3677,
+      "step": 1158
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.5213735103607178,
+      "learning_rate": 1.8813015590135963e-06,
+      "loss": 0.2632,
+      "step": 1159
+    },
+    {
+      "epoch": 2.22912,
+      "grad_norm": 0.5155541300773621,
+      "learning_rate": 1.8725644671036125e-06,
+      "loss": 0.2177,
+      "step": 1160
+    },
+    {
+      "epoch": 2.23104,
+      "grad_norm": 0.5886062383651733,
+      "learning_rate": 1.8638430338219199e-06,
+      "loss": 0.3509,
+      "step": 1161
+    },
+    {
+      "epoch": 2.23296,
+      "grad_norm": 0.5050294995307922,
+      "learning_rate": 1.8551373028355013e-06,
+      "loss": 0.2756,
+      "step": 1162
+    },
+    {
+      "epoch": 2.23488,
+      "grad_norm": 0.5493663549423218,
+      "learning_rate": 1.846447317732723e-06,
+      "loss": 0.2805,
+      "step": 1163
+    },
+    {
+      "epoch": 2.2368,
+      "grad_norm": 0.573564350605011,
+      "learning_rate": 1.8377731220231144e-06,
+      "loss": 0.3016,
+      "step": 1164
+    },
+    {
+      "epoch": 2.23872,
+      "grad_norm": 0.5422305464744568,
+      "learning_rate": 1.8291147591371482e-06,
+      "loss": 0.274,
+      "step": 1165
+    },
+    {
+      "epoch": 2.24064,
+      "grad_norm": 0.5040318965911865,
+      "learning_rate": 1.8204722724260266e-06,
+      "loss": 0.2734,
+      "step": 1166
+    },
+    {
+      "epoch": 2.24256,
+      "grad_norm": 0.5195218920707703,
+      "learning_rate": 1.8118457051614591e-06,
+      "loss": 0.2816,
+      "step": 1167
+    },
+    {
+      "epoch": 2.24448,
+      "grad_norm": 0.5320647358894348,
+      "learning_rate": 1.803235100535452e-06,
+      "loss": 0.2622,
+      "step": 1168
+    },
+    {
+      "epoch": 2.2464,
+      "grad_norm": 0.5619749426841736,
+      "learning_rate": 1.7946405016600843e-06,
+      "loss": 0.2683,
+      "step": 1169
+    },
+    {
+      "epoch": 2.24832,
+      "grad_norm": 0.5170308947563171,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.2477,
+      "step": 1170
+    },
+    {
+      "epoch": 2.25024,
+      "grad_norm": 0.5504286289215088,
+      "learning_rate": 1.7774994932086976e-06,
+      "loss": 0.2759,
+      "step": 1171
+    },
+    {
+      "epoch": 2.25216,
+      "grad_norm": 0.5571184158325195,
+      "learning_rate": 1.7689531694552863e-06,
+      "loss": 0.3073,
+      "step": 1172
+    },
+    {
+      "epoch": 2.25408,
+      "grad_norm": 0.5318464636802673,
+      "learning_rate": 1.7604230230973068e-06,
+      "loss": 0.2975,
+      "step": 1173
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.48411232233047485,
+      "learning_rate": 1.7519090968439966e-06,
+      "loss": 0.2688,
+      "step": 1174
+    },
+    {
+      "epoch": 2.25792,
+      "grad_norm": 0.5787471532821655,
+      "learning_rate": 1.7434114333233853e-06,
+      "loss": 0.3319,
+      "step": 1175
+    },
+    {
+      "epoch": 2.25984,
+      "grad_norm": 0.5259526968002319,
+      "learning_rate": 1.7349300750820758e-06,
+      "loss": 0.2462,
+      "step": 1176
+    },
+    {
+      "epoch": 2.2617599999999998,
+      "grad_norm": 0.5181608200073242,
+      "learning_rate": 1.7264650645850256e-06,
+      "loss": 0.2655,
+      "step": 1177
+    },
+    {
+      "epoch": 2.26368,
+      "grad_norm": 0.522636353969574,
+      "learning_rate": 1.7180164442153529e-06,
+      "loss": 0.2895,
+      "step": 1178
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.4949401021003723,
+      "learning_rate": 1.709584256274106e-06,
+      "loss": 0.2397,
+      "step": 1179
+    },
+    {
+      "epoch": 2.26752,
+      "grad_norm": 0.547947108745575,
+      "learning_rate": 1.7011685429800596e-06,
+      "loss": 0.2746,
+      "step": 1180
+    },
+    {
+      "epoch": 2.26944,
+      "grad_norm": 0.5172644257545471,
+      "learning_rate": 1.6927693464695022e-06,
+      "loss": 0.2764,
+      "step": 1181
+    },
+    {
+      "epoch": 2.27136,
+      "grad_norm": 0.5107510685920715,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.2413,
+      "step": 1182
+    },
+    {
+      "epoch": 2.27328,
+      "grad_norm": 0.5045897364616394,
+      "learning_rate": 1.6760206719303107e-06,
+      "loss": 0.2616,
+      "step": 1183
+    },
+    {
+      "epoch": 2.2752,
+      "grad_norm": 0.5340535044670105,
+      "learning_rate": 1.6676712777599275e-06,
+      "loss": 0.2777,
+      "step": 1184
+    },
+    {
+      "epoch": 2.27712,
+      "grad_norm": 1.0574698448181152,
+      "learning_rate": 1.6593385680891139e-06,
+      "loss": 0.3968,
+      "step": 1185
+    },
+    {
+      "epoch": 2.27904,
+      "grad_norm": 0.5148729681968689,
+      "learning_rate": 1.6510225846385668e-06,
+      "loss": 0.2556,
+      "step": 1186
+    },
+    {
+      "epoch": 2.28096,
+      "grad_norm": 0.5266612768173218,
+      "learning_rate": 1.6427233690452455e-06,
+      "loss": 0.2766,
+      "step": 1187
+    },
+    {
+      "epoch": 2.28288,
+      "grad_norm": 0.4794199466705322,
+      "learning_rate": 1.6344409628621482e-06,
+      "loss": 0.2335,
+      "step": 1188
+    },
+    {
+      "epoch": 2.2848,
+      "grad_norm": 0.5271382927894592,
+      "learning_rate": 1.6261754075581187e-06,
+      "loss": 0.3039,
+      "step": 1189
+    },
+    {
+      "epoch": 2.28672,
+      "grad_norm": 0.5311862230300903,
+      "learning_rate": 1.6179267445176206e-06,
+      "loss": 0.2806,
+      "step": 1190
+    },
+    {
+      "epoch": 2.28864,
+      "grad_norm": 0.5555956959724426,
+      "learning_rate": 1.6096950150405454e-06,
+      "loss": 0.3126,
+      "step": 1191
+    },
+    {
+      "epoch": 2.29056,
+      "grad_norm": 0.5290018320083618,
+      "learning_rate": 1.6014802603420044e-06,
+      "loss": 0.2562,
+      "step": 1192
+    },
+    {
+      "epoch": 2.29248,
+      "grad_norm": 0.49365177750587463,
+      "learning_rate": 1.593282521552113e-06,
+      "loss": 0.2328,
+      "step": 1193
+    },
+    {
+      "epoch": 2.2944,
+      "grad_norm": 0.5647149085998535,
+      "learning_rate": 1.5851018397157918e-06,
+      "loss": 0.2998,
+      "step": 1194
+    },
+    {
+      "epoch": 2.29632,
+      "grad_norm": 0.5351371765136719,
+      "learning_rate": 1.5769382557925612e-06,
+      "loss": 0.2547,
+      "step": 1195
+    },
+    {
+      "epoch": 2.29824,
+      "grad_norm": 0.5579474568367004,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.2526,
+      "step": 1196
+    },
+    {
+      "epoch": 2.30016,
+      "grad_norm": 0.5138181447982788,
+      "learning_rate": 1.5606625450952062e-06,
+      "loss": 0.2328,
+      "step": 1197
+    },
+    {
+      "epoch": 2.30208,
+      "grad_norm": 0.4934709072113037,
+      "learning_rate": 1.5525504998112717e-06,
+      "loss": 0.2404,
+      "step": 1198
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.5571762323379517,
+      "learning_rate": 1.5444557154203892e-06,
+      "loss": 0.2838,
+      "step": 1199
+    },
+    {
+      "epoch": 2.30592,
+      "grad_norm": 0.4902985394001007,
+      "learning_rate": 1.5363782324520033e-06,
+      "loss": 0.2412,
+      "step": 1200
+    },
+    {
+      "epoch": 2.30784,
+      "grad_norm": 0.47329410910606384,
+      "learning_rate": 1.5283180913489326e-06,
+      "loss": 0.2266,
+      "step": 1201
+    },
+    {
+      "epoch": 2.30976,
+      "grad_norm": 0.5809300541877747,
+      "learning_rate": 1.520275332467166e-06,
+      "loss": 0.3494,
+      "step": 1202
+    },
+    {
+      "epoch": 2.31168,
+      "grad_norm": 0.5269177556037903,
+      "learning_rate": 1.5122499960756604e-06,
+      "loss": 0.2942,
+      "step": 1203
+    },
+    {
+      "epoch": 2.3136,
+      "grad_norm": 0.5624419450759888,
+      "learning_rate": 1.504242122356143e-06,
+      "loss": 0.2776,
+      "step": 1204
+    },
+    {
+      "epoch": 2.3155200000000002,
+      "grad_norm": 0.5004557967185974,
+      "learning_rate": 1.4962517514029069e-06,
+      "loss": 0.265,
+      "step": 1205
+    },
+    {
+      "epoch": 2.31744,
+      "grad_norm": 0.4810078740119934,
+      "learning_rate": 1.4882789232226124e-06,
+      "loss": 0.2342,
+      "step": 1206
+    },
+    {
+      "epoch": 2.31936,
+      "grad_norm": 0.5699446201324463,
+      "learning_rate": 1.4803236777340856e-06,
+      "loss": 0.3356,
+      "step": 1207
+    },
+    {
+      "epoch": 2.32128,
+      "grad_norm": 0.49257543683052063,
+      "learning_rate": 1.4723860547681163e-06,
+      "loss": 0.2502,
+      "step": 1208
+    },
+    {
+      "epoch": 2.3232,
+      "grad_norm": 0.5249313116073608,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.2918,
+      "step": 1209
+    },
+    {
+      "epoch": 2.32512,
+      "grad_norm": 0.5423191785812378,
+      "learning_rate": 1.4565638352856504e-06,
+      "loss": 0.282,
+      "step": 1210
+    },
+    {
+      "epoch": 2.32704,
+      "grad_norm": 0.5252957940101624,
+      "learning_rate": 1.4486793179887738e-06,
+      "loss": 0.2432,
+      "step": 1211
+    },
+    {
+      "epoch": 2.32896,
+      "grad_norm": 0.5464475750923157,
+      "learning_rate": 1.4408125816532981e-06,
+      "loss": 0.2842,
+      "step": 1212
+    },
+    {
+      "epoch": 2.33088,
+      "grad_norm": 0.5821936130523682,
+      "learning_rate": 1.4329636656668617e-06,
+      "loss": 0.3484,
+      "step": 1213
+    },
+    {
+      "epoch": 2.3327999999999998,
+      "grad_norm": 0.5434349775314331,
+      "learning_rate": 1.4251326093278773e-06,
+      "loss": 0.3122,
+      "step": 1214
+    },
+    {
+      "epoch": 2.33472,
+      "grad_norm": 0.5610452890396118,
+      "learning_rate": 1.4173194518453415e-06,
+      "loss": 0.3511,
+      "step": 1215
+    },
+    {
+      "epoch": 2.33664,
+      "grad_norm": 0.5676388740539551,
+      "learning_rate": 1.4095242323386305e-06,
+      "loss": 0.2674,
+      "step": 1216
+    },
+    {
+      "epoch": 2.33856,
+      "grad_norm": 0.5156210660934448,
+      "learning_rate": 1.4017469898373077e-06,
+      "loss": 0.2417,
+      "step": 1217
+    },
+    {
+      "epoch": 2.34048,
+      "grad_norm": 0.5702019333839417,
+      "learning_rate": 1.3939877632809279e-06,
+      "loss": 0.3422,
+      "step": 1218
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.5302916169166565,
+      "learning_rate": 1.3862465915188427e-06,
+      "loss": 0.2915,
+      "step": 1219
+    },
+    {
+      "epoch": 2.34432,
+      "grad_norm": 0.524100124835968,
+      "learning_rate": 1.3785235133100088e-06,
+      "loss": 0.2567,
+      "step": 1220
+    },
+    {
+      "epoch": 2.34624,
+      "grad_norm": 0.5496824383735657,
+      "learning_rate": 1.3708185673227896e-06,
+      "loss": 0.3192,
+      "step": 1221
+    },
+    {
+      "epoch": 2.34816,
+      "grad_norm": 0.5403966903686523,
+      "learning_rate": 1.3631317921347564e-06,
+      "loss": 0.2993,
+      "step": 1222
+    },
+    {
+      "epoch": 2.35008,
+      "grad_norm": 0.5469822883605957,
+      "learning_rate": 1.3554632262325129e-06,
+      "loss": 0.2804,
+      "step": 1223
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 0.5396835207939148,
+      "learning_rate": 1.347812908011485e-06,
+      "loss": 0.2972,
+      "step": 1224
+    },
+    {
+      "epoch": 2.35392,
+      "grad_norm": 0.589891791343689,
+      "learning_rate": 1.340180875775735e-06,
+      "loss": 0.2825,
+      "step": 1225
+    },
+    {
+      "epoch": 2.35584,
+      "grad_norm": 0.5493536591529846,
+      "learning_rate": 1.3325671677377727e-06,
+      "loss": 0.3004,
+      "step": 1226
+    },
+    {
+      "epoch": 2.35776,
+      "grad_norm": 0.5246999859809875,
+      "learning_rate": 1.3249718220183583e-06,
+      "loss": 0.2935,
+      "step": 1227
+    },
+    {
+      "epoch": 2.35968,
+      "grad_norm": 0.5652483105659485,
+      "learning_rate": 1.3173948766463146e-06,
+      "loss": 0.3328,
+      "step": 1228
+    },
+    {
+      "epoch": 2.3616,
+      "grad_norm": 0.5446557402610779,
+      "learning_rate": 1.309836369558341e-06,
+      "loss": 0.253,
+      "step": 1229
+    },
+    {
+      "epoch": 2.36352,
+      "grad_norm": 0.5502628087997437,
+      "learning_rate": 1.3022963385988153e-06,
+      "loss": 0.3301,
+      "step": 1230
+    },
+    {
+      "epoch": 2.36544,
+      "grad_norm": 0.47435635328292847,
+      "learning_rate": 1.2947748215196038e-06,
+      "loss": 0.2497,
+      "step": 1231
+    },
+    {
+      "epoch": 2.36736,
+      "grad_norm": 0.5310259461402893,
+      "learning_rate": 1.2872718559798852e-06,
+      "loss": 0.3,
+      "step": 1232
+    },
+    {
+      "epoch": 2.36928,
+      "grad_norm": 0.5287906527519226,
+      "learning_rate": 1.2797874795459464e-06,
+      "loss": 0.2983,
+      "step": 1233
+    },
+    {
+      "epoch": 2.3712,
+      "grad_norm": 0.5118058919906616,
+      "learning_rate": 1.2723217296910078e-06,
+      "loss": 0.2687,
+      "step": 1234
+    },
+    {
+      "epoch": 2.37312,
+      "grad_norm": 0.5420043468475342,
+      "learning_rate": 1.264874643795021e-06,
+      "loss": 0.253,
+      "step": 1235
+    },
+    {
+      "epoch": 2.37504,
+      "grad_norm": 0.5595210790634155,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.2965,
+      "step": 1236
+    },
+    {
+      "epoch": 2.37696,
+      "grad_norm": 0.5088095664978027,
+      "learning_rate": 1.2500366129323039e-06,
+      "loss": 0.261,
+      "step": 1237
+    },
+    {
+      "epoch": 2.37888,
+      "grad_norm": 0.5904267430305481,
+      "learning_rate": 1.2426457422575e-06,
+      "loss": 0.333,
+      "step": 1238
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.5221114754676819,
+      "learning_rate": 1.2352736841251306e-06,
+      "loss": 0.2533,
+      "step": 1239
+    },
+    {
+      "epoch": 2.38272,
+      "grad_norm": 0.5373600125312805,
+      "learning_rate": 1.2279204754460494e-06,
+      "loss": 0.3,
+      "step": 1240
+    },
+    {
+      "epoch": 2.38464,
+      "grad_norm": 0.5690154433250427,
+      "learning_rate": 1.2205861530367342e-06,
+      "loss": 0.3129,
+      "step": 1241
+    },
+    {
+      "epoch": 2.3865600000000002,
+      "grad_norm": 0.4852672219276428,
+      "learning_rate": 1.2132707536191008e-06,
+      "loss": 0.2413,
+      "step": 1242
+    },
+    {
+      "epoch": 2.38848,
+      "grad_norm": 0.5786780714988708,
+      "learning_rate": 1.2059743138203256e-06,
+      "loss": 0.332,
+      "step": 1243
+    },
+    {
+      "epoch": 2.3904,
+      "grad_norm": 0.5776085257530212,
+      "learning_rate": 1.1986968701726492e-06,
+      "loss": 0.3274,
+      "step": 1244
+    },
+    {
+      "epoch": 2.39232,
+      "grad_norm": 0.5448642373085022,
+      "learning_rate": 1.1914384591132045e-06,
+      "loss": 0.294,
+      "step": 1245
+    },
+    {
+      "epoch": 2.39424,
+      "grad_norm": 0.505533754825592,
+      "learning_rate": 1.184199116983834e-06,
+      "loss": 0.2792,
+      "step": 1246
+    },
+    {
+      "epoch": 2.39616,
+      "grad_norm": 0.537585973739624,
+      "learning_rate": 1.1769788800309001e-06,
+      "loss": 0.279,
+      "step": 1247
+    },
+    {
+      "epoch": 2.39808,
+      "grad_norm": 0.5108551979064941,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.2372,
+      "step": 1248
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.5080101490020752,
+      "learning_rate": 1.1625958661613345e-06,
+      "loss": 0.273,
+      "step": 1249
+    },
+    {
+      "epoch": 2.40192,
+      "grad_norm": 0.5584346055984497,
+      "learning_rate": 1.1554331612584218e-06,
+      "loss": 0.3326,
+      "step": 1250
+    },
+    {
+      "epoch": 2.4038399999999998,
+      "grad_norm": 0.5501477122306824,
+      "learning_rate": 1.1482897055590275e-06,
+      "loss": 0.2955,
+      "step": 1251
+    },
+    {
+      "epoch": 2.40576,
+      "grad_norm": 0.5100276470184326,
+      "learning_rate": 1.141165534829425e-06,
+      "loss": 0.2364,
+      "step": 1252
+    },
+    {
+      "epoch": 2.40768,
+      "grad_norm": 0.5747215151786804,
+      "learning_rate": 1.134060684739331e-06,
+      "loss": 0.336,
+      "step": 1253
+    },
+    {
+      "epoch": 2.4096,
+      "grad_norm": 0.5547710657119751,
+      "learning_rate": 1.1269751908617277e-06,
+      "loss": 0.3122,
+      "step": 1254
+    },
+    {
+      "epoch": 2.41152,
+      "grad_norm": 0.5013325214385986,
+      "learning_rate": 1.119909088672682e-06,
+      "loss": 0.2332,
+      "step": 1255
+    },
+    {
+      "epoch": 2.41344,
+      "grad_norm": 0.5268529057502747,
+      "learning_rate": 1.1128624135511712e-06,
+      "loss": 0.2737,
+      "step": 1256
+    },
+    {
+      "epoch": 2.41536,
+      "grad_norm": 0.5017078518867493,
+      "learning_rate": 1.105835200778902e-06,
+      "loss": 0.2658,
+      "step": 1257
+    },
+    {
+      "epoch": 2.41728,
+      "grad_norm": 0.5137251019477844,
+      "learning_rate": 1.0988274855401377e-06,
+      "loss": 0.2831,
+      "step": 1258
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.5609878301620483,
+      "learning_rate": 1.091839302921518e-06,
+      "loss": 0.2951,
+      "step": 1259
+    },
+    {
+      "epoch": 2.42112,
+      "grad_norm": 0.5148438811302185,
+      "learning_rate": 1.0848706879118893e-06,
+      "loss": 0.2689,
+      "step": 1260
+    },
+    {
+      "epoch": 2.42304,
+      "grad_norm": 0.5275246500968933,
+      "learning_rate": 1.0779216754021215e-06,
+      "loss": 0.261,
+      "step": 1261
+    },
+    {
+      "epoch": 2.42496,
+      "grad_norm": 0.5768818855285645,
+      "learning_rate": 1.070992300184941e-06,
+      "loss": 0.3116,
+      "step": 1262
+    },
+    {
+      "epoch": 2.42688,
+      "grad_norm": 0.520849883556366,
+      "learning_rate": 1.0640825969547498e-06,
+      "loss": 0.2565,
+      "step": 1263
+    },
+    {
+      "epoch": 2.4288,
+      "grad_norm": 0.5152717232704163,
+      "learning_rate": 1.057192600307456e-06,
+      "loss": 0.2492,
+      "step": 1264
+    },
+    {
+      "epoch": 2.43072,
+      "grad_norm": 0.514140248298645,
+      "learning_rate": 1.0503223447403033e-06,
+      "loss": 0.2323,
+      "step": 1265
+    },
+    {
+      "epoch": 2.43264,
+      "grad_norm": 0.525887131690979,
+      "learning_rate": 1.0434718646516917e-06,
+      "loss": 0.2783,
+      "step": 1266
+    },
+    {
+      "epoch": 2.43456,
+      "grad_norm": 0.5362196564674377,
+      "learning_rate": 1.0366411943410033e-06,
+      "loss": 0.3023,
+      "step": 1267
+    },
+    {
+      "epoch": 2.43648,
+      "grad_norm": 0.49444660544395447,
+      "learning_rate": 1.0298303680084448e-06,
+      "loss": 0.2371,
+      "step": 1268
+    },
+    {
+      "epoch": 2.4384,
+      "grad_norm": 0.5334476232528687,
+      "learning_rate": 1.0230394197548605e-06,
+      "loss": 0.2902,
+      "step": 1269
+    },
+    {
+      "epoch": 2.44032,
+      "grad_norm": 0.5205954909324646,
+      "learning_rate": 1.0162683835815706e-06,
+      "loss": 0.2608,
+      "step": 1270
+    },
+    {
+      "epoch": 2.44224,
+      "grad_norm": 0.551737904548645,
+      "learning_rate": 1.009517293390197e-06,
+      "loss": 0.3059,
+      "step": 1271
+    },
+    {
+      "epoch": 2.44416,
+      "grad_norm": 0.5070509314537048,
+      "learning_rate": 1.0027861829824953e-06,
+      "loss": 0.2304,
+      "step": 1272
+    },
+    {
+      "epoch": 2.44608,
+      "grad_norm": 0.5290709137916565,
+      "learning_rate": 9.960750860601842e-07,
+      "loss": 0.275,
+      "step": 1273
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 0.5123082995414734,
+      "learning_rate": 9.893840362247809e-07,
+      "loss": 0.2464,
+      "step": 1274
+    },
+    {
+      "epoch": 2.44992,
+      "grad_norm": 0.5409252643585205,
+      "learning_rate": 9.82713066977427e-07,
+      "loss": 0.3267,
+      "step": 1275
+    },
+    {
+      "epoch": 2.45184,
+      "grad_norm": 0.5274913907051086,
+      "learning_rate": 9.760622117187234e-07,
+      "loss": 0.283,
+      "step": 1276
+    },
+    {
+      "epoch": 2.45376,
+      "grad_norm": 0.5742137432098389,
+      "learning_rate": 9.694315037485635e-07,
+      "loss": 0.3158,
+      "step": 1277
+    },
+    {
+      "epoch": 2.45568,
+      "grad_norm": 0.557791531085968,
+      "learning_rate": 9.628209762659658e-07,
+      "loss": 0.2781,
+      "step": 1278
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.506764829158783,
+      "learning_rate": 9.562306623689111e-07,
+      "loss": 0.2642,
+      "step": 1279
+    },
+    {
+      "epoch": 2.45952,
+      "grad_norm": 0.5207616686820984,
+      "learning_rate": 9.496605950541676e-07,
+      "loss": 0.2859,
+      "step": 1280
+    },
+    {
+      "epoch": 2.46144,
+      "grad_norm": 0.5234065055847168,
+      "learning_rate": 9.431108072171346e-07,
+      "loss": 0.2982,
+      "step": 1281
+    },
+    {
+      "epoch": 2.4633599999999998,
+      "grad_norm": 0.4819236099720001,
+      "learning_rate": 9.365813316516787e-07,
+      "loss": 0.2341,
+      "step": 1282
+    },
+    {
+      "epoch": 2.46528,
+      "grad_norm": 0.5364812612533569,
+      "learning_rate": 9.300722010499608e-07,
+      "loss": 0.2598,
+      "step": 1283
+    },
+    {
+      "epoch": 2.4672,
+      "grad_norm": 0.4972287714481354,
+      "learning_rate": 9.235834480022788e-07,
+      "loss": 0.2796,
+      "step": 1284
+    },
+    {
+      "epoch": 2.46912,
+      "grad_norm": 0.4950721263885498,
+      "learning_rate": 9.17115104996903e-07,
+      "loss": 0.2913,
+      "step": 1285
+    },
+    {
+      "epoch": 2.47104,
+      "grad_norm": 0.5206244587898254,
+      "learning_rate": 9.10667204419915e-07,
+      "loss": 0.2709,
+      "step": 1286
+    },
+    {
+      "epoch": 2.47296,
+      "grad_norm": 0.55411696434021,
+      "learning_rate": 9.042397785550405e-07,
+      "loss": 0.3023,
+      "step": 1287
+    },
+    {
+      "epoch": 2.47488,
+      "grad_norm": 0.4549674093723297,
+      "learning_rate": 8.978328595834984e-07,
+      "loss": 0.2053,
+      "step": 1288
+    },
+    {
+      "epoch": 2.4768,
+      "grad_norm": 0.509245753288269,
+      "learning_rate": 8.91446479583823e-07,
+      "loss": 0.2834,
+      "step": 1289
+    },
+    {
+      "epoch": 2.47872,
+      "grad_norm": 0.5446407794952393,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.3241,
+      "step": 1290
+    },
+    {
+      "epoch": 2.48064,
+      "grad_norm": 0.5171636343002319,
+      "learning_rate": 8.787354642998936e-07,
+      "loss": 0.281,
+      "step": 1291
+    },
+    {
+      "epoch": 2.48256,
+      "grad_norm": 0.5980998277664185,
+      "learning_rate": 8.724108926579e-07,
+      "loss": 0.3467,
+      "step": 1292
+    },
+    {
+      "epoch": 2.48448,
+      "grad_norm": 0.5404484868049622,
+      "learning_rate": 8.661069872719746e-07,
+      "loss": 0.2908,
+      "step": 1293
+    },
+    {
+      "epoch": 2.4864,
+      "grad_norm": 0.5078650116920471,
+      "learning_rate": 8.598237797048825e-07,
+      "loss": 0.2373,
+      "step": 1294
+    },
+    {
+      "epoch": 2.48832,
+      "grad_norm": 0.5432282090187073,
+      "learning_rate": 8.535613014157556e-07,
+      "loss": 0.2674,
+      "step": 1295
+    },
+    {
+      "epoch": 2.49024,
+      "grad_norm": 0.5542681813240051,
+      "learning_rate": 8.473195837599419e-07,
+      "loss": 0.2795,
+      "step": 1296
+    },
+    {
+      "epoch": 2.49216,
+      "grad_norm": 0.5242416858673096,
+      "learning_rate": 8.410986579888381e-07,
+      "loss": 0.249,
+      "step": 1297
+    },
+    {
+      "epoch": 2.49408,
+      "grad_norm": 0.5228531956672668,
+      "learning_rate": 8.348985552497424e-07,
+      "loss": 0.2598,
+      "step": 1298
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.5574326515197754,
+      "learning_rate": 8.287193065856936e-07,
+      "loss": 0.3093,
+      "step": 1299
+    },
+    {
+      "epoch": 2.49792,
+      "grad_norm": 0.5175071358680725,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.2588,
+      "step": 1300
+    },
+    {
+      "epoch": 2.49984,
+      "grad_norm": 0.5172858238220215,
+      "learning_rate": 8.164234951326727e-07,
+      "loss": 0.2647,
+      "step": 1301
+    },
+    {
+      "epoch": 2.50176,
+      "grad_norm": 0.5424558520317078,
+      "learning_rate": 8.103069939070945e-07,
+      "loss": 0.3114,
+      "step": 1302
+    },
+    {
+      "epoch": 2.50368,
+      "grad_norm": 0.5101452469825745,
+      "learning_rate": 8.042114698830394e-07,
+      "loss": 0.2519,
+      "step": 1303
+    },
+    {
+      "epoch": 2.5056000000000003,
+      "grad_norm": 0.6045923829078674,
+      "learning_rate": 7.981369535799354e-07,
+      "loss": 0.3774,
+      "step": 1304
+    },
+    {
+      "epoch": 2.50752,
+      "grad_norm": 0.5323867797851562,
+      "learning_rate": 7.920834754120305e-07,
+      "loss": 0.2571,
+      "step": 1305
+    },
+    {
+      "epoch": 2.50944,
+      "grad_norm": 0.5404714345932007,
+      "learning_rate": 7.860510656882342e-07,
+      "loss": 0.2785,
+      "step": 1306
+    },
+    {
+      "epoch": 2.51136,
+      "grad_norm": 0.5398850440979004,
+      "learning_rate": 7.800397546119709e-07,
+      "loss": 0.2802,
+      "step": 1307
+    },
+    {
+      "epoch": 2.51328,
+      "grad_norm": 0.5384976267814636,
+      "learning_rate": 7.740495722810271e-07,
+      "loss": 0.2644,
+      "step": 1308
+    },
+    {
+      "epoch": 2.5152,
+      "grad_norm": 0.5045672655105591,
+      "learning_rate": 7.680805486873977e-07,
+      "loss": 0.2316,
+      "step": 1309
+    },
+    {
+      "epoch": 2.5171200000000002,
+      "grad_norm": 0.5286737084388733,
+      "learning_rate": 7.621327137171447e-07,
+      "loss": 0.2545,
+      "step": 1310
+    },
+    {
+      "epoch": 2.51904,
+      "grad_norm": 0.5331920385360718,
+      "learning_rate": 7.562060971502383e-07,
+      "loss": 0.295,
+      "step": 1311
+    },
+    {
+      "epoch": 2.52096,
+      "grad_norm": 0.5143847465515137,
+      "learning_rate": 7.50300728660407e-07,
+      "loss": 0.2849,
+      "step": 1312
+    },
+    {
+      "epoch": 2.52288,
+      "grad_norm": 0.5072587132453918,
+      "learning_rate": 7.444166378150014e-07,
+      "loss": 0.2467,
+      "step": 1313
+    },
+    {
+      "epoch": 2.5248,
+      "grad_norm": 0.47314944863319397,
+      "learning_rate": 7.385538540748327e-07,
+      "loss": 0.2366,
+      "step": 1314
+    },
+    {
+      "epoch": 2.52672,
+      "grad_norm": 0.5704914927482605,
+      "learning_rate": 7.327124067940311e-07,
+      "loss": 0.315,
+      "step": 1315
+    },
+    {
+      "epoch": 2.52864,
+      "grad_norm": 0.5133654475212097,
+      "learning_rate": 7.26892325219899e-07,
+      "loss": 0.2683,
+      "step": 1316
+    },
+    {
+      "epoch": 2.53056,
+      "grad_norm": 0.5594615936279297,
+      "learning_rate": 7.210936384927631e-07,
+      "loss": 0.3586,
+      "step": 1317
+    },
+    {
+      "epoch": 2.53248,
+      "grad_norm": 0.5097527503967285,
+      "learning_rate": 7.153163756458287e-07,
+      "loss": 0.2533,
+      "step": 1318
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.5137017965316772,
+      "learning_rate": 7.09560565605037e-07,
+      "loss": 0.2736,
+      "step": 1319
+    },
+    {
+      "epoch": 2.53632,
+      "grad_norm": 0.5312138199806213,
+      "learning_rate": 7.03826237188916e-07,
+      "loss": 0.295,
+      "step": 1320
+    },
+    {
+      "epoch": 2.53824,
+      "grad_norm": 0.6559857130050659,
+      "learning_rate": 6.981134191084388e-07,
+      "loss": 0.3476,
+      "step": 1321
+    },
+    {
+      "epoch": 2.54016,
+      "grad_norm": 0.5578992962837219,
+      "learning_rate": 6.924221399668785e-07,
+      "loss": 0.3141,
+      "step": 1322
+    },
+    {
+      "epoch": 2.54208,
+      "grad_norm": 0.5448708534240723,
+      "learning_rate": 6.867524282596655e-07,
+      "loss": 0.2562,
+      "step": 1323
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.5386422276496887,
+      "learning_rate": 6.811043123742494e-07,
+      "loss": 0.2846,
+      "step": 1324
+    },
+    {
+      "epoch": 2.5459199999999997,
+      "grad_norm": 0.5408101081848145,
+      "learning_rate": 6.754778205899465e-07,
+      "loss": 0.2938,
+      "step": 1325
+    },
+    {
+      "epoch": 2.54784,
+      "grad_norm": 0.5798414349555969,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.3057,
+      "step": 1326
+    },
+    {
+      "epoch": 2.54976,
+      "grad_norm": 0.5611661672592163,
+      "learning_rate": 6.642898219004723e-07,
+      "loss": 0.3181,
+      "step": 1327
+    },
+    {
+      "epoch": 2.55168,
+      "grad_norm": 0.5378629565238953,
+      "learning_rate": 6.587283710120324e-07,
+      "loss": 0.3251,
+      "step": 1328
+    },
+    {
+      "epoch": 2.5536,
+      "grad_norm": 0.5189611315727234,
+      "learning_rate": 6.531886562578859e-07,
+      "loss": 0.3166,
+      "step": 1329
+    },
+    {
+      "epoch": 2.55552,
+      "grad_norm": 0.5893354415893555,
+      "learning_rate": 6.47670705374604e-07,
+      "loss": 0.3166,
+      "step": 1330
+    },
+    {
+      "epoch": 2.55744,
+      "grad_norm": 0.4974201023578644,
+      "learning_rate": 6.421745459897871e-07,
+      "loss": 0.2625,
+      "step": 1331
+    },
+    {
+      "epoch": 2.55936,
+      "grad_norm": 0.5556225180625916,
+      "learning_rate": 6.367002056219285e-07,
+      "loss": 0.3331,
+      "step": 1332
+    },
+    {
+      "epoch": 2.56128,
+      "grad_norm": 0.5511367917060852,
+      "learning_rate": 6.312477116802807e-07,
+      "loss": 0.2919,
+      "step": 1333
+    },
+    {
+      "epoch": 2.5632,
+      "grad_norm": 0.5125117301940918,
+      "learning_rate": 6.258170914647077e-07,
+      "loss": 0.2486,
+      "step": 1334
+    },
+    {
+      "epoch": 2.56512,
+      "grad_norm": 0.5501939654350281,
+      "learning_rate": 6.204083721655607e-07,
+      "loss": 0.3543,
+      "step": 1335
+    },
+    {
+      "epoch": 2.56704,
+      "grad_norm": 0.5546405911445618,
+      "learning_rate": 6.150215808635334e-07,
+      "loss": 0.299,
+      "step": 1336
+    },
+    {
+      "epoch": 2.56896,
+      "grad_norm": 0.543989896774292,
+      "learning_rate": 6.096567445295298e-07,
+      "loss": 0.2406,
+      "step": 1337
+    },
+    {
+      "epoch": 2.57088,
+      "grad_norm": 0.5333052277565002,
+      "learning_rate": 6.043138900245277e-07,
+      "loss": 0.2663,
+      "step": 1338
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.5557252764701843,
+      "learning_rate": 5.989930440994451e-07,
+      "loss": 0.3188,
+      "step": 1339
+    },
+    {
+      "epoch": 2.57472,
+      "grad_norm": 0.5499855875968933,
+      "learning_rate": 5.936942333950063e-07,
+      "loss": 0.3269,
+      "step": 1340
+    },
+    {
+      "epoch": 2.5766400000000003,
+      "grad_norm": 0.5173619985580444,
+      "learning_rate": 5.884174844416102e-07,
+      "loss": 0.2649,
+      "step": 1341
+    },
+    {
+      "epoch": 2.57856,
+      "grad_norm": 0.5802469849586487,
+      "learning_rate": 5.831628236591929e-07,
+      "loss": 0.2957,
+      "step": 1342
+    },
+    {
+      "epoch": 2.58048,
+      "grad_norm": 0.548316478729248,
+      "learning_rate": 5.779302773570994e-07,
+      "loss": 0.284,
+      "step": 1343
+    },
+    {
+      "epoch": 2.5824,
+      "grad_norm": 0.523848295211792,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.2587,
+      "step": 1344
+    },
+    {
+      "epoch": 2.58432,
+      "grad_norm": 0.47859248518943787,
+      "learning_rate": 5.675316328775126e-07,
+      "loss": 0.1909,
+      "step": 1345
+    },
+    {
+      "epoch": 2.58624,
+      "grad_norm": 0.5654980540275574,
+      "learning_rate": 5.623655867645628e-07,
+      "loss": 0.3127,
+      "step": 1346
+    },
+    {
+      "epoch": 2.5881600000000002,
+      "grad_norm": 0.5302689075469971,
+      "learning_rate": 5.572217592607687e-07,
+      "loss": 0.2703,
+      "step": 1347
+    },
+    {
+      "epoch": 2.59008,
+      "grad_norm": 0.5951128005981445,
+      "learning_rate": 5.521001761205441e-07,
+      "loss": 0.3452,
+      "step": 1348
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 0.5373216867446899,
+      "learning_rate": 5.470008629869367e-07,
+      "loss": 0.2658,
+      "step": 1349
+    },
+    {
+      "epoch": 2.59392,
+      "grad_norm": 0.53899085521698,
+      "learning_rate": 5.41923845391486e-07,
+      "loss": 0.2687,
+      "step": 1350
+    },
+    {
+      "epoch": 2.59584,
+      "grad_norm": 0.5003790855407715,
+      "learning_rate": 5.368691487541027e-07,
+      "loss": 0.2468,
+      "step": 1351
+    },
+    {
+      "epoch": 2.59776,
+      "grad_norm": 0.5328875184059143,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.2675,
+      "step": 1352
+    },
+    {
+      "epoch": 2.59968,
+      "grad_norm": 0.47491270303726196,
+      "learning_rate": 5.268268194742638e-07,
+      "loss": 0.23,
+      "step": 1353
+    },
+    {
+      "epoch": 2.6016,
+      "grad_norm": 0.5315960049629211,
+      "learning_rate": 5.218392371123326e-07,
+      "loss": 0.2912,
+      "step": 1354
+    },
+    {
+      "epoch": 2.60352,
+      "grad_norm": 0.5693609118461609,
+      "learning_rate": 5.168740762692681e-07,
+      "loss": 0.3021,
+      "step": 1355
+    },
+    {
+      "epoch": 2.6054399999999998,
+      "grad_norm": 0.5275635719299316,
+      "learning_rate": 5.119313618049309e-07,
+      "loss": 0.2744,
+      "step": 1356
+    },
+    {
+      "epoch": 2.60736,
+      "grad_norm": 0.5745298862457275,
+      "learning_rate": 5.070111184667908e-07,
+      "loss": 0.2931,
+      "step": 1357
+    },
+    {
+      "epoch": 2.60928,
+      "grad_norm": 0.5557462573051453,
+      "learning_rate": 5.021133708898146e-07,
+      "loss": 0.277,
+      "step": 1358
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.5096452236175537,
+      "learning_rate": 4.972381435963331e-07,
+      "loss": 0.2272,
+      "step": 1359
+    },
+    {
+      "epoch": 2.61312,
+      "grad_norm": 0.5025963187217712,
+      "learning_rate": 4.9238546099592e-07,
+      "loss": 0.2286,
+      "step": 1360
+    },
+    {
+      "epoch": 2.61504,
+      "grad_norm": 0.5286290645599365,
+      "learning_rate": 4.875553473852735e-07,
+      "loss": 0.2512,
+      "step": 1361
+    },
+    {
+      "epoch": 2.6169599999999997,
+      "grad_norm": 0.5427853465080261,
+      "learning_rate": 4.827478269480895e-07,
+      "loss": 0.2912,
+      "step": 1362
+    },
+    {
+      "epoch": 2.61888,
+      "grad_norm": 0.5329648852348328,
+      "learning_rate": 4.779629237549438e-07,
+      "loss": 0.3203,
+      "step": 1363
+    },
+    {
+      "epoch": 2.6208,
+      "grad_norm": 0.5288243293762207,
+      "learning_rate": 4.732006617631729e-07,
+      "loss": 0.2687,
+      "step": 1364
+    },
+    {
+      "epoch": 2.62272,
+      "grad_norm": 0.5017338395118713,
+      "learning_rate": 4.6846106481675035e-07,
+      "loss": 0.2648,
+      "step": 1365
+    },
+    {
+      "epoch": 2.62464,
+      "grad_norm": 0.5405363440513611,
+      "learning_rate": 4.637441566461681e-07,
+      "loss": 0.2545,
+      "step": 1366
+    },
+    {
+      "epoch": 2.62656,
+      "grad_norm": 0.5540486574172974,
+      "learning_rate": 4.590499608683202e-07,
+      "loss": 0.2814,
+      "step": 1367
+    },
+    {
+      "epoch": 2.62848,
+      "grad_norm": 0.5053295493125916,
+      "learning_rate": 4.54378500986381e-07,
+      "loss": 0.2753,
+      "step": 1368
+    },
+    {
+      "epoch": 2.6304,
+      "grad_norm": 0.5184414386749268,
+      "learning_rate": 4.497298003896944e-07,
+      "loss": 0.248,
+      "step": 1369
+    },
+    {
+      "epoch": 2.63232,
+      "grad_norm": 0.5374047756195068,
+      "learning_rate": 4.451038823536441e-07,
+      "loss": 0.3257,
+      "step": 1370
+    },
+    {
+      "epoch": 2.63424,
+      "grad_norm": 0.516406238079071,
+      "learning_rate": 4.405007700395497e-07,
+      "loss": 0.2536,
+      "step": 1371
+    },
+    {
+      "epoch": 2.63616,
+      "grad_norm": 0.5297756791114807,
+      "learning_rate": 4.35920486494546e-07,
+      "loss": 0.2851,
+      "step": 1372
+    },
+    {
+      "epoch": 2.63808,
+      "grad_norm": 0.5142978429794312,
+      "learning_rate": 4.313630546514663e-07,
+      "loss": 0.2494,
+      "step": 1373
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.5551294088363647,
+      "learning_rate": 4.268284973287273e-07,
+      "loss": 0.276,
+      "step": 1374
+    },
+    {
+      "epoch": 2.64192,
+      "grad_norm": 0.5049953460693359,
+      "learning_rate": 4.223168372302189e-07,
+      "loss": 0.2775,
+      "step": 1375
+    },
+    {
+      "epoch": 2.64384,
+      "grad_norm": 0.5217283964157104,
+      "learning_rate": 4.1782809694518533e-07,
+      "loss": 0.2716,
+      "step": 1376
+    },
+    {
+      "epoch": 2.64576,
+      "grad_norm": 0.5295341610908508,
+      "learning_rate": 4.1336229894811454e-07,
+      "loss": 0.2599,
+      "step": 1377
+    },
+    {
+      "epoch": 2.6476800000000003,
+      "grad_norm": 0.5834762454032898,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 0.3451,
+      "step": 1378
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.45285993814468384,
+      "learning_rate": 4.044996191413686e-07,
+      "loss": 0.1924,
+      "step": 1379
+    },
+    {
+      "epoch": 2.65152,
+      "grad_norm": 0.5599640607833862,
+      "learning_rate": 4.001027817058789e-07,
+      "loss": 0.3083,
+      "step": 1380
+    },
+    {
+      "epoch": 2.65344,
+      "grad_norm": 0.56117844581604,
+      "learning_rate": 3.957289753065052e-07,
+      "loss": 0.3127,
+      "step": 1381
+    },
+    {
+      "epoch": 2.65536,
+      "grad_norm": 0.570381760597229,
+      "learning_rate": 3.9137822184227845e-07,
+      "loss": 0.3358,
+      "step": 1382
+    },
+    {
+      "epoch": 2.65728,
+      "grad_norm": 0.5211979746818542,
+      "learning_rate": 3.870505430968069e-07,
+      "loss": 0.2688,
+      "step": 1383
+    },
+    {
+      "epoch": 2.6592000000000002,
+      "grad_norm": 0.5594597458839417,
+      "learning_rate": 3.8274596073816784e-07,
+      "loss": 0.3128,
+      "step": 1384
+    },
+    {
+      "epoch": 2.66112,
+      "grad_norm": 0.5592339038848877,
+      "learning_rate": 3.7846449631879664e-07,
+      "loss": 0.2808,
+      "step": 1385
+    },
+    {
+      "epoch": 2.66304,
+      "grad_norm": 0.4964348077774048,
+      "learning_rate": 3.742061712753825e-07,
+      "loss": 0.259,
+      "step": 1386
+    },
+    {
+      "epoch": 2.6649599999999998,
+      "grad_norm": 0.4814138114452362,
+      "learning_rate": 3.699710069287571e-07,
+      "loss": 0.2064,
+      "step": 1387
+    },
+    {
+      "epoch": 2.66688,
+      "grad_norm": 0.5795491933822632,
+      "learning_rate": 3.657590244837911e-07,
+      "loss": 0.2884,
+      "step": 1388
+    },
+    {
+      "epoch": 2.6688,
+      "grad_norm": 0.5160412788391113,
+      "learning_rate": 3.615702450292857e-07,
+      "loss": 0.2585,
+      "step": 1389
+    },
+    {
+      "epoch": 2.67072,
+      "grad_norm": 0.5075104236602783,
+      "learning_rate": 3.5740468953786854e-07,
+      "loss": 0.2576,
+      "step": 1390
+    },
+    {
+      "epoch": 2.67264,
+      "grad_norm": 0.57828289270401,
+      "learning_rate": 3.5326237886588734e-07,
+      "loss": 0.3278,
+      "step": 1391
+    },
+    {
+      "epoch": 2.67456,
+      "grad_norm": 0.5309354662895203,
+      "learning_rate": 3.49143333753309e-07,
+      "loss": 0.2874,
+      "step": 1392
+    },
+    {
+      "epoch": 2.6764799999999997,
+      "grad_norm": 0.5397410988807678,
+      "learning_rate": 3.4504757482360817e-07,
+      "loss": 0.2618,
+      "step": 1393
+    },
+    {
+      "epoch": 2.6784,
+      "grad_norm": 0.5536224842071533,
+      "learning_rate": 3.4097512258367385e-07,
+      "loss": 0.2769,
+      "step": 1394
+    },
+    {
+      "epoch": 2.68032,
+      "grad_norm": 0.5112101435661316,
+      "learning_rate": 3.369259974236988e-07,
+      "loss": 0.2524,
+      "step": 1395
+    },
+    {
+      "epoch": 2.68224,
+      "grad_norm": 0.5118189454078674,
+      "learning_rate": 3.3290021961708163e-07,
+      "loss": 0.289,
+      "step": 1396
+    },
+    {
+      "epoch": 2.68416,
+      "grad_norm": 0.5124532580375671,
+      "learning_rate": 3.288978093203227e-07,
+      "loss": 0.2423,
+      "step": 1397
+    },
+    {
+      "epoch": 2.68608,
+      "grad_norm": 0.5080134272575378,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.2353,
+      "step": 1398
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.5464827418327332,
+      "learning_rate": 3.209631712972966e-07,
+      "loss": 0.3018,
+      "step": 1399
+    },
+    {
+      "epoch": 2.68992,
+      "grad_norm": 0.5050497055053711,
+      "learning_rate": 3.1703098329864237e-07,
+      "loss": 0.2773,
+      "step": 1400
+    },
+    {
+      "epoch": 2.69184,
+      "grad_norm": 0.4953087866306305,
+      "learning_rate": 3.131222422648744e-07,
+      "loss": 0.214,
+      "step": 1401
+    },
+    {
+      "epoch": 2.69376,
+      "grad_norm": 0.611307680606842,
+      "learning_rate": 3.0923696776650414e-07,
+      "loss": 0.2778,
+      "step": 1402
+    },
+    {
+      "epoch": 2.69568,
+      "grad_norm": 0.5462144613265991,
+      "learning_rate": 3.0537517925655567e-07,
+      "loss": 0.2766,
+      "step": 1403
+    },
+    {
+      "epoch": 2.6976,
+      "grad_norm": 0.5388204455375671,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.2576,
+      "step": 1404
+    },
+    {
+      "epoch": 2.69952,
+      "grad_norm": 0.5153400301933289,
+      "learning_rate": 2.9772213742595367e-07,
+      "loss": 0.2431,
+      "step": 1405
+    },
+    {
+      "epoch": 2.70144,
+      "grad_norm": 0.5605869293212891,
+      "learning_rate": 2.9393092242300026e-07,
+      "loss": 0.3015,
+      "step": 1406
+    },
+    {
+      "epoch": 2.70336,
+      "grad_norm": 0.5621346235275269,
+      "learning_rate": 2.901632700436757e-07,
+      "loss": 0.3077,
+      "step": 1407
+    },
+    {
+      "epoch": 2.70528,
+      "grad_norm": 0.5033831596374512,
+      "learning_rate": 2.864191991520848e-07,
+      "loss": 0.2455,
+      "step": 1408
+    },
+    {
+      "epoch": 2.7072000000000003,
+      "grad_norm": 0.48300206661224365,
+      "learning_rate": 2.8269872849426114e-07,
+      "loss": 0.2373,
+      "step": 1409
+    },
+    {
+      "epoch": 2.70912,
+      "grad_norm": 0.5201036930084229,
+      "learning_rate": 2.790018766980773e-07,
+      "loss": 0.2575,
+      "step": 1410
+    },
+    {
+      "epoch": 2.71104,
+      "grad_norm": 0.5643584728240967,
+      "learning_rate": 2.7532866227314714e-07,
+      "loss": 0.3004,
+      "step": 1411
+    },
+    {
+      "epoch": 2.71296,
+      "grad_norm": 0.5160210728645325,
+      "learning_rate": 2.71679103610738e-07,
+      "loss": 0.2289,
+      "step": 1412
+    },
+    {
+      "epoch": 2.71488,
+      "grad_norm": 0.5517868399620056,
+      "learning_rate": 2.6805321898367323e-07,
+      "loss": 0.2671,
+      "step": 1413
+    },
+    {
+      "epoch": 2.7168,
+      "grad_norm": 0.5592690110206604,
+      "learning_rate": 2.64451026546248e-07,
+      "loss": 0.2917,
+      "step": 1414
+    },
+    {
+      "epoch": 2.7187200000000002,
+      "grad_norm": 0.5464155673980713,
+      "learning_rate": 2.6087254433412924e-07,
+      "loss": 0.2611,
+      "step": 1415
+    },
+    {
+      "epoch": 2.72064,
+      "grad_norm": 0.517124593257904,
+      "learning_rate": 2.573177902642726e-07,
+      "loss": 0.2764,
+      "step": 1416
+    },
+    {
+      "epoch": 2.72256,
+      "grad_norm": 0.5529623627662659,
+      "learning_rate": 2.5378678213483057e-07,
+      "loss": 0.2961,
+      "step": 1417
+    },
+    {
+      "epoch": 2.72448,
+      "grad_norm": 0.556036651134491,
+      "learning_rate": 2.502795376250622e-07,
+      "loss": 0.2282,
+      "step": 1418
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.5442557334899902,
+      "learning_rate": 2.467960742952463e-07,
+      "loss": 0.2649,
+      "step": 1419
+    },
+    {
+      "epoch": 2.72832,
+      "grad_norm": 0.49297502636909485,
+      "learning_rate": 2.4333640958659144e-07,
+      "loss": 0.2018,
+      "step": 1420
+    },
+    {
+      "epoch": 2.7302400000000002,
+      "grad_norm": 0.5577881336212158,
+      "learning_rate": 2.399005608211502e-07,
+      "loss": 0.3163,
+      "step": 1421
+    },
+    {
+      "epoch": 2.73216,
+      "grad_norm": 0.517428994178772,
+      "learning_rate": 2.3648854520173237e-07,
+      "loss": 0.2742,
+      "step": 1422
+    },
+    {
+      "epoch": 2.73408,
+      "grad_norm": 0.5350176095962524,
+      "learning_rate": 2.3310037981182088e-07,
+      "loss": 0.327,
+      "step": 1423
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 0.5728635787963867,
+      "learning_rate": 2.2973608161547755e-07,
+      "loss": 0.3192,
+      "step": 1424
+    },
+    {
+      "epoch": 2.73792,
+      "grad_norm": 0.5502520799636841,
+      "learning_rate": 2.2639566745727203e-07,
+      "loss": 0.2951,
+      "step": 1425
+    },
+    {
+      "epoch": 2.73984,
+      "grad_norm": 0.4843711853027344,
+      "learning_rate": 2.2307915406218517e-07,
+      "loss": 0.2122,
+      "step": 1426
+    },
+    {
+      "epoch": 2.74176,
+      "grad_norm": 0.49999791383743286,
+      "learning_rate": 2.1978655803553128e-07,
+      "loss": 0.219,
+      "step": 1427
+    },
+    {
+      "epoch": 2.74368,
+      "grad_norm": 0.5275260210037231,
+      "learning_rate": 2.1651789586287442e-07,
+      "loss": 0.2885,
+      "step": 1428
+    },
+    {
+      "epoch": 2.7456,
+      "grad_norm": 0.5150967240333557,
+      "learning_rate": 2.1327318390994445e-07,
+      "loss": 0.2798,
+      "step": 1429
+    },
+    {
+      "epoch": 2.7475199999999997,
+      "grad_norm": 0.5025932788848877,
+      "learning_rate": 2.1005243842255552e-07,
+      "loss": 0.2474,
+      "step": 1430
+    },
+    {
+      "epoch": 2.74944,
+      "grad_norm": 0.5199427008628845,
+      "learning_rate": 2.068556755265272e-07,
+      "loss": 0.243,
+      "step": 1431
+    },
+    {
+      "epoch": 2.75136,
+      "grad_norm": 1.9392038583755493,
+      "learning_rate": 2.0368291122759898e-07,
+      "loss": 0.5163,
+      "step": 1432
+    },
+    {
+      "epoch": 2.75328,
+      "grad_norm": 0.48683902621269226,
+      "learning_rate": 2.005341614113543e-07,
+      "loss": 0.2123,
+      "step": 1433
+    },
+    {
+      "epoch": 2.7552,
+      "grad_norm": 0.5586715340614319,
+      "learning_rate": 1.9740944184313882e-07,
+      "loss": 0.2685,
+      "step": 1434
+    },
+    {
+      "epoch": 2.75712,
+      "grad_norm": 0.558089017868042,
+      "learning_rate": 1.9430876816798228e-07,
+      "loss": 0.2777,
+      "step": 1435
+    },
+    {
+      "epoch": 2.75904,
+      "grad_norm": 0.5511989593505859,
+      "learning_rate": 1.9123215591052014e-07,
+      "loss": 0.2983,
+      "step": 1436
+    },
+    {
+      "epoch": 2.76096,
+      "grad_norm": 0.5335795879364014,
+      "learning_rate": 1.88179620474917e-07,
+      "loss": 0.2619,
+      "step": 1437
+    },
+    {
+      "epoch": 2.76288,
+      "grad_norm": 0.5755378007888794,
+      "learning_rate": 1.8515117714478447e-07,
+      "loss": 0.2938,
+      "step": 1438
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.5381505489349365,
+      "learning_rate": 1.8214684108311286e-07,
+      "loss": 0.275,
+      "step": 1439
+    },
+    {
+      "epoch": 2.76672,
+      "grad_norm": 0.5264760255813599,
+      "learning_rate": 1.7916662733218848e-07,
+      "loss": 0.2587,
+      "step": 1440
+    },
+    {
+      "epoch": 2.76864,
+      "grad_norm": 0.5486288666725159,
+      "learning_rate": 1.762105508135198e-07,
+      "loss": 0.3161,
+      "step": 1441
+    },
+    {
+      "epoch": 2.77056,
+      "grad_norm": 0.49116531014442444,
+      "learning_rate": 1.7327862632776638e-07,
+      "loss": 0.2452,
+      "step": 1442
+    },
+    {
+      "epoch": 2.77248,
+      "grad_norm": 0.5521030426025391,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.2754,
+      "step": 1443
+    },
+    {
+      "epoch": 2.7744,
+      "grad_norm": 0.5492038726806641,
+      "learning_rate": 1.6748729205293024e-07,
+      "loss": 0.2891,
+      "step": 1444
+    },
+    {
+      "epoch": 2.77632,
+      "grad_norm": 0.503020703792572,
+      "learning_rate": 1.646279112602417e-07,
+      "loss": 0.2218,
+      "step": 1445
+    },
+    {
+      "epoch": 2.7782400000000003,
+      "grad_norm": 0.5517632365226746,
+      "learning_rate": 1.6179274049310966e-07,
+      "loss": 0.2349,
+      "step": 1446
+    },
+    {
+      "epoch": 2.78016,
+      "grad_norm": 0.5212957859039307,
+      "learning_rate": 1.5898179394683244e-07,
+      "loss": 0.2746,
+      "step": 1447
+    },
+    {
+      "epoch": 2.78208,
+      "grad_norm": 0.47663435339927673,
+      "learning_rate": 1.5619508569542363e-07,
+      "loss": 0.2274,
+      "step": 1448
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 0.5489935278892517,
+      "learning_rate": 1.5343262969153781e-07,
+      "loss": 0.3223,
+      "step": 1449
+    },
+    {
+      "epoch": 2.78592,
+      "grad_norm": 0.5325907468795776,
+      "learning_rate": 1.5069443976640287e-07,
+      "loss": 0.302,
+      "step": 1450
+    },
+    {
+      "epoch": 2.78784,
+      "grad_norm": 0.5111280679702759,
+      "learning_rate": 1.4798052962974874e-07,
+      "loss": 0.2661,
+      "step": 1451
+    },
+    {
+      "epoch": 2.7897600000000002,
+      "grad_norm": 0.5166310667991638,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.2915,
+      "step": 1452
+    },
+    {
+      "epoch": 2.79168,
+      "grad_norm": 0.5559850931167603,
+      "learning_rate": 1.4262560295290884e-07,
+      "loss": 0.287,
+      "step": 1453
+    },
+    {
+      "epoch": 2.7936,
+      "grad_norm": 0.5579703450202942,
+      "learning_rate": 1.3998461322408563e-07,
+      "loss": 0.3283,
+      "step": 1454
+    },
+    {
+      "epoch": 2.79552,
+      "grad_norm": 0.5109795331954956,
+      "learning_rate": 1.3736795690633353e-07,
+      "loss": 0.2792,
+      "step": 1455
+    },
+    {
+      "epoch": 2.79744,
+      "grad_norm": 0.5248427987098694,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.2703,
+      "step": 1456
+    },
+    {
+      "epoch": 2.79936,
+      "grad_norm": 0.5487046837806702,
+      "learning_rate": 1.3220769678705724e-07,
+      "loss": 0.2748,
+      "step": 1457
+    },
+    {
+      "epoch": 2.80128,
+      "grad_norm": 0.5361766815185547,
+      "learning_rate": 1.2966411882222695e-07,
+      "loss": 0.2723,
+      "step": 1458
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.5451945066452026,
+      "learning_rate": 1.271449259417268e-07,
+      "loss": 0.2974,
+      "step": 1459
+    },
+    {
+      "epoch": 2.80512,
+      "grad_norm": 0.4694383442401886,
+      "learning_rate": 1.2465013075879884e-07,
+      "loss": 0.2235,
+      "step": 1460
+    },
+    {
+      "epoch": 2.8070399999999998,
+      "grad_norm": 0.5395259857177734,
+      "learning_rate": 1.2217974576453072e-07,
+      "loss": 0.2839,
+      "step": 1461
+    },
+    {
+      "epoch": 2.80896,
+      "grad_norm": 0.5818226933479309,
+      "learning_rate": 1.1973378332779229e-07,
+      "loss": 0.3077,
+      "step": 1462
+    },
+    {
+      "epoch": 2.81088,
+      "grad_norm": 0.5247942805290222,
+      "learning_rate": 1.1731225569517113e-07,
+      "loss": 0.2808,
+      "step": 1463
+    },
+    {
+      "epoch": 2.8128,
+      "grad_norm": 0.5117696523666382,
+      "learning_rate": 1.1491517499091498e-07,
+      "loss": 0.264,
+      "step": 1464
+    },
+    {
+      "epoch": 2.81472,
+      "grad_norm": 0.5266653895378113,
+      "learning_rate": 1.1254255321686836e-07,
+      "loss": 0.3011,
+      "step": 1465
+    },
+    {
+      "epoch": 2.81664,
+      "grad_norm": 0.4915022552013397,
+      "learning_rate": 1.1019440225241317e-07,
+      "loss": 0.2493,
+      "step": 1466
+    },
+    {
+      "epoch": 2.8185599999999997,
+      "grad_norm": 0.5022535920143127,
+      "learning_rate": 1.078707338544105e-07,
+      "loss": 0.2716,
+      "step": 1467
+    },
+    {
+      "epoch": 2.82048,
+      "grad_norm": 0.5364386439323425,
+      "learning_rate": 1.055715596571405e-07,
+      "loss": 0.2946,
+      "step": 1468
+    },
+    {
+      "epoch": 2.8224,
+      "grad_norm": 0.549530029296875,
+      "learning_rate": 1.0329689117224262e-07,
+      "loss": 0.2976,
+      "step": 1469
+    },
+    {
+      "epoch": 2.82432,
+      "grad_norm": 0.5654832124710083,
+      "learning_rate": 1.0104673978866164e-07,
+      "loss": 0.3099,
+      "step": 1470
+    },
+    {
+      "epoch": 2.82624,
+      "grad_norm": 0.5266042947769165,
+      "learning_rate": 9.882111677258777e-08,
+      "loss": 0.227,
+      "step": 1471
+    },
+    {
+      "epoch": 2.82816,
+      "grad_norm": 0.541802704334259,
+      "learning_rate": 9.662003326740166e-08,
+      "loss": 0.2759,
+      "step": 1472
+    },
+    {
+      "epoch": 2.83008,
+      "grad_norm": 0.5359547734260559,
+      "learning_rate": 9.444350029361671e-08,
+      "loss": 0.2853,
+      "step": 1473
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.48531007766723633,
+      "learning_rate": 9.22915287488274e-08,
+      "loss": 0.2232,
+      "step": 1474
+    },
+    {
+      "epoch": 2.83392,
+      "grad_norm": 0.5301419496536255,
+      "learning_rate": 9.016412940765107e-08,
+      "loss": 0.2313,
+      "step": 1475
+    },
+    {
+      "epoch": 2.83584,
+      "grad_norm": 0.5007804036140442,
+      "learning_rate": 8.80613129216762e-08,
+      "loss": 0.2342,
+      "step": 1476
+    },
+    {
+      "epoch": 2.83776,
+      "grad_norm": 0.517315685749054,
+      "learning_rate": 8.598308981940751e-08,
+      "loss": 0.2697,
+      "step": 1477
+    },
+    {
+      "epoch": 2.83968,
+      "grad_norm": 0.5181477665901184,
+      "learning_rate": 8.392947050621603e-08,
+      "loss": 0.254,
+      "step": 1478
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.5087987184524536,
+      "learning_rate": 8.190046526428241e-08,
+      "loss": 0.2459,
+      "step": 1479
+    },
+    {
+      "epoch": 2.84352,
+      "grad_norm": 0.5246613025665283,
+      "learning_rate": 7.989608425254924e-08,
+      "loss": 0.2472,
+      "step": 1480
+    },
+    {
+      "epoch": 2.84544,
+      "grad_norm": 0.5786083340644836,
+      "learning_rate": 7.791633750667105e-08,
+      "loss": 0.2974,
+      "step": 1481
+    },
+    {
+      "epoch": 2.84736,
+      "grad_norm": 0.5533187985420227,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.2744,
+      "step": 1482
+    },
+    {
+      "epoch": 2.8492800000000003,
+      "grad_norm": 0.5378976464271545,
+      "learning_rate": 7.403078633833716e-08,
+      "loss": 0.2833,
+      "step": 1483
+    },
+    {
+      "epoch": 2.8512,
+      "grad_norm": 0.48698195815086365,
+      "learning_rate": 7.212500137028789e-08,
+      "loss": 0.2438,
+      "step": 1484
+    },
+    {
+      "epoch": 2.85312,
+      "grad_norm": 0.5447365641593933,
+      "learning_rate": 7.024388957680705e-08,
+      "loss": 0.2705,
+      "step": 1485
+    },
+    {
+      "epoch": 2.85504,
+      "grad_norm": 0.5075899362564087,
+      "learning_rate": 6.838746037635735e-08,
+      "loss": 0.262,
+      "step": 1486
+    },
+    {
+      "epoch": 2.85696,
+      "grad_norm": 0.5052075386047363,
+      "learning_rate": 6.655572306381696e-08,
+      "loss": 0.2844,
+      "step": 1487
+    },
+    {
+      "epoch": 2.85888,
+      "grad_norm": 0.5378866195678711,
+      "learning_rate": 6.474868681043578e-08,
+      "loss": 0.3133,
+      "step": 1488
+    },
+    {
+      "epoch": 2.8608000000000002,
+      "grad_norm": 0.5053765177726746,
+      "learning_rate": 6.29663606637898e-08,
+      "loss": 0.26,
+      "step": 1489
+    },
+    {
+      "epoch": 2.86272,
+      "grad_norm": 0.4990273714065552,
+      "learning_rate": 6.120875354773459e-08,
+      "loss": 0.2775,
+      "step": 1490
+    },
+    {
+      "epoch": 2.86464,
+      "grad_norm": 0.561270534992218,
+      "learning_rate": 5.947587426236079e-08,
+      "loss": 0.3022,
+      "step": 1491
+    },
+    {
+      "epoch": 2.8665599999999998,
+      "grad_norm": 0.5420502424240112,
+      "learning_rate": 5.776773148394976e-08,
+      "loss": 0.2721,
+      "step": 1492
+    },
+    {
+      "epoch": 2.86848,
+      "grad_norm": 0.5910217761993408,
+      "learning_rate": 5.608433376493194e-08,
+      "loss": 0.299,
+      "step": 1493
+    },
+    {
+      "epoch": 2.8704,
+      "grad_norm": 0.5273252129554749,
+      "learning_rate": 5.4425689533841864e-08,
+      "loss": 0.2863,
+      "step": 1494
+    },
+    {
+      "epoch": 2.87232,
+      "grad_norm": 0.5191953182220459,
+      "learning_rate": 5.279180709527765e-08,
+      "loss": 0.2507,
+      "step": 1495
+    },
+    {
+      "epoch": 2.87424,
+      "grad_norm": 0.5528905391693115,
+      "learning_rate": 5.1182694629857145e-08,
+      "loss": 0.2747,
+      "step": 1496
+    },
+    {
+      "epoch": 2.87616,
+      "grad_norm": 0.5159099102020264,
+      "learning_rate": 4.959836019417963e-08,
+      "loss": 0.2692,
+      "step": 1497
+    },
+    {
+      "epoch": 2.8780799999999997,
+      "grad_norm": 0.5562239289283752,
+      "learning_rate": 4.803881172078473e-08,
+      "loss": 0.3109,
+      "step": 1498
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4771888256072998,
+      "learning_rate": 4.650405701811078e-08,
+      "loss": 0.2306,
+      "step": 1499
+    },
+    {
+      "epoch": 2.88192,
+      "grad_norm": 0.5593058466911316,
+      "learning_rate": 4.499410377045765e-08,
+      "loss": 0.2662,
+      "step": 1500
+    },
+    {
+      "epoch": 2.88384,
+      "grad_norm": 0.5127353072166443,
+      "learning_rate": 4.350895953794898e-08,
+      "loss": 0.2511,
+      "step": 1501
+    },
+    {
+      "epoch": 2.88576,
+      "grad_norm": 0.5308008193969727,
+      "learning_rate": 4.2048631756492206e-08,
+      "loss": 0.2739,
+      "step": 1502
+    },
+    {
+      "epoch": 2.88768,
+      "grad_norm": 0.5492320656776428,
+      "learning_rate": 4.0613127737741396e-08,
+      "loss": 0.308,
+      "step": 1503
+    },
+    {
+      "epoch": 2.8895999999999997,
+      "grad_norm": 0.5101546049118042,
+      "learning_rate": 3.9202454669063915e-08,
+      "loss": 0.2731,
+      "step": 1504
+    },
+    {
+      "epoch": 2.89152,
+      "grad_norm": 0.5620161294937134,
+      "learning_rate": 3.781661961349992e-08,
+      "loss": 0.2376,
+      "step": 1505
+    },
+    {
+      "epoch": 2.89344,
+      "grad_norm": 0.4927712380886078,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.2435,
+      "step": 1506
+    },
+    {
+      "epoch": 2.89536,
+      "grad_norm": 0.5900700092315674,
+      "learning_rate": 3.5119491172039836e-08,
+      "loss": 0.3185,
+      "step": 1507
+    },
+    {
+      "epoch": 2.89728,
+      "grad_norm": 0.5799671411514282,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.3245,
+      "step": 1508
+    },
+    {
+      "epoch": 2.8992,
+      "grad_norm": 0.5340503454208374,
+      "learning_rate": 3.252179642985909e-08,
+      "loss": 0.2901,
+      "step": 1509
+    },
+    {
+      "epoch": 2.90112,
+      "grad_norm": 0.5165550112724304,
+      "learning_rate": 3.1260253031660247e-08,
+      "loss": 0.2528,
+      "step": 1510
+    },
+    {
+      "epoch": 2.90304,
+      "grad_norm": 0.5425903797149658,
+      "learning_rate": 3.002358741205691e-08,
+      "loss": 0.3238,
+      "step": 1511
+    },
+    {
+      "epoch": 2.90496,
+      "grad_norm": 0.5243244171142578,
+      "learning_rate": 2.8811805762860578e-08,
+      "loss": 0.2829,
+      "step": 1512
+    },
+    {
+      "epoch": 2.90688,
+      "grad_norm": 0.4840536415576935,
+      "learning_rate": 2.762491415128965e-08,
+      "loss": 0.2684,
+      "step": 1513
+    },
+    {
+      "epoch": 2.9088000000000003,
+      "grad_norm": 0.5116623640060425,
+      "learning_rate": 2.6462918519944425e-08,
+      "loss": 0.2579,
+      "step": 1514
+    },
+    {
+      "epoch": 2.91072,
+      "grad_norm": 0.5284073352813721,
+      "learning_rate": 2.5325824686772138e-08,
+      "loss": 0.2696,
+      "step": 1515
+    },
+    {
+      "epoch": 2.91264,
+      "grad_norm": 0.5507498383522034,
+      "learning_rate": 2.4213638345040868e-08,
+      "loss": 0.3055,
+      "step": 1516
+    },
+    {
+      "epoch": 2.91456,
+      "grad_norm": 0.5134804248809814,
+      "learning_rate": 2.3126365063311228e-08,
+      "loss": 0.2837,
+      "step": 1517
+    },
+    {
+      "epoch": 2.91648,
+      "grad_norm": 0.5148767828941345,
+      "learning_rate": 2.206401028540639e-08,
+      "loss": 0.2794,
+      "step": 1518
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.5791639089584351,
+      "learning_rate": 2.1026579330387655e-08,
+      "loss": 0.3157,
+      "step": 1519
+    },
+    {
+      "epoch": 2.9203200000000002,
+      "grad_norm": 0.5207650661468506,
+      "learning_rate": 2.0014077392525035e-08,
+      "loss": 0.2842,
+      "step": 1520
+    },
+    {
+      "epoch": 2.92224,
+      "grad_norm": 0.5276824831962585,
+      "learning_rate": 1.9026509541272276e-08,
+      "loss": 0.2829,
+      "step": 1521
+    },
+    {
+      "epoch": 2.92416,
+      "grad_norm": 0.5397416353225708,
+      "learning_rate": 1.8063880721242986e-08,
+      "loss": 0.2676,
+      "step": 1522
+    },
+    {
+      "epoch": 2.92608,
+      "grad_norm": 0.5123454928398132,
+      "learning_rate": 1.712619575218344e-08,
+      "loss": 0.2925,
+      "step": 1523
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 2.336402177810669,
+      "learning_rate": 1.6213459328950355e-08,
+      "loss": 0.4248,
+      "step": 1524
+    },
+    {
+      "epoch": 2.92992,
+      "grad_norm": 0.532516360282898,
+      "learning_rate": 1.5325676021484826e-08,
+      "loss": 0.2648,
+      "step": 1525
+    },
+    {
+      "epoch": 2.9318400000000002,
+      "grad_norm": 0.5240092277526855,
+      "learning_rate": 1.4462850274794548e-08,
+      "loss": 0.2786,
+      "step": 1526
+    },
+    {
+      "epoch": 2.93376,
+      "grad_norm": 0.5047268867492676,
+      "learning_rate": 1.3624986408924956e-08,
+      "loss": 0.2454,
+      "step": 1527
+    },
+    {
+      "epoch": 2.93568,
+      "grad_norm": 0.5109764337539673,
+      "learning_rate": 1.2812088618942009e-08,
+      "loss": 0.258,
+      "step": 1528
+    },
+    {
+      "epoch": 2.9375999999999998,
+      "grad_norm": 0.5353485345840454,
+      "learning_rate": 1.2024160974911103e-08,
+      "loss": 0.2839,
+      "step": 1529
+    },
+    {
+      "epoch": 2.93952,
+      "grad_norm": 0.5175056457519531,
+      "learning_rate": 1.1261207421874309e-08,
+      "loss": 0.2847,
+      "step": 1530
+    },
+    {
+      "epoch": 2.94144,
+      "grad_norm": 0.5646983981132507,
+      "learning_rate": 1.0523231779832055e-08,
+      "loss": 0.3153,
+      "step": 1531
+    },
+    {
+      "epoch": 2.94336,
+      "grad_norm": 0.573734700679779,
+      "learning_rate": 9.810237743724805e-09,
+      "loss": 0.3068,
+      "step": 1532
+    },
+    {
+      "epoch": 2.94528,
+      "grad_norm": 0.5229532718658447,
+      "learning_rate": 9.12222888341252e-09,
+      "loss": 0.2929,
+      "step": 1533
+    },
+    {
+      "epoch": 2.9472,
+      "grad_norm": 0.5395614504814148,
+      "learning_rate": 8.459208643659122e-09,
+      "loss": 0.3148,
+      "step": 1534
+    },
+    {
+      "epoch": 2.9491199999999997,
+      "grad_norm": 0.5242195129394531,
+      "learning_rate": 7.8211803441125e-09,
+      "loss": 0.286,
+      "step": 1535
+    },
+    {
+      "epoch": 2.95104,
+      "grad_norm": 0.5721161961555481,
+      "learning_rate": 7.2081471792911914e-09,
+      "loss": 0.3343,
+      "step": 1536
+    },
+    {
+      "epoch": 2.95296,
+      "grad_norm": 0.5378016829490662,
+      "learning_rate": 6.6201122185649555e-09,
+      "loss": 0.2675,
+      "step": 1537
+    },
+    {
+      "epoch": 2.95488,
+      "grad_norm": 0.5330981016159058,
+      "learning_rate": 6.057078406142003e-09,
+      "loss": 0.309,
+      "step": 1538
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.5302271246910095,
+      "learning_rate": 5.519048561053453e-09,
+      "loss": 0.2788,
+      "step": 1539
+    },
+    {
+      "epoch": 2.95872,
+      "grad_norm": 0.5372453331947327,
+      "learning_rate": 5.006025377138901e-09,
+      "loss": 0.2719,
+      "step": 1540
+    },
+    {
+      "epoch": 2.96064,
+      "grad_norm": 0.5358155369758606,
+      "learning_rate": 4.518011423032542e-09,
+      "loss": 0.2698,
+      "step": 1541
+    },
+    {
+      "epoch": 2.96256,
+      "grad_norm": 0.557539165019989,
+      "learning_rate": 4.055009142152066e-09,
+      "loss": 0.3262,
+      "step": 1542
+    },
+    {
+      "epoch": 2.96448,
+      "grad_norm": 0.5364840626716614,
+      "learning_rate": 3.6170208526836724e-09,
+      "loss": 0.2993,
+      "step": 1543
+    },
+    {
+      "epoch": 2.9664,
+      "grad_norm": 0.5672816038131714,
+      "learning_rate": 3.204048747573185e-09,
+      "loss": 0.3519,
+      "step": 1544
+    },
+    {
+      "epoch": 2.96832,
+      "grad_norm": 0.5199512243270874,
+      "learning_rate": 2.816094894513843e-09,
+      "loss": 0.2523,
+      "step": 1545
+    },
+    {
+      "epoch": 2.97024,
+      "grad_norm": 0.5317375063896179,
+      "learning_rate": 2.4531612359363077e-09,
+      "loss": 0.2475,
+      "step": 1546
+    },
+    {
+      "epoch": 2.97216,
+      "grad_norm": 0.4999130368232727,
+      "learning_rate": 2.1152495889970035e-09,
+      "loss": 0.2126,
+      "step": 1547
+    },
+    {
+      "epoch": 2.97408,
+      "grad_norm": 0.5565233826637268,
+      "learning_rate": 1.8023616455731253e-09,
+      "loss": 0.2994,
+      "step": 1548
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 0.5300474762916565,
+      "learning_rate": 1.514498972249312e-09,
+      "loss": 0.2751,
+      "step": 1549
+    },
+    {
+      "epoch": 2.97792,
+      "grad_norm": 0.5208839178085327,
+      "learning_rate": 1.2516630103137638e-09,
+      "loss": 0.2556,
+      "step": 1550
+    },
+    {
+      "epoch": 2.9798400000000003,
+      "grad_norm": 0.5471765995025635,
+      "learning_rate": 1.0138550757493592e-09,
+      "loss": 0.2431,
+      "step": 1551
+    },
+    {
+      "epoch": 2.98176,
+      "grad_norm": 0.5384496450424194,
+      "learning_rate": 8.010763592264381e-10,
+      "loss": 0.2813,
+      "step": 1552
+    },
+    {
+      "epoch": 2.98368,
+      "grad_norm": 0.5266402363777161,
+      "learning_rate": 6.133279260983616e-10,
+      "loss": 0.2696,
+      "step": 1553
+    },
+    {
+      "epoch": 2.9856,
+      "grad_norm": 0.4907943308353424,
+      "learning_rate": 4.506107163948503e-10,
+      "loss": 0.2419,
+      "step": 1554
+    },
+    {
+      "epoch": 2.98752,
+      "grad_norm": 0.5272263884544373,
+      "learning_rate": 3.12925544818099e-10,
+      "loss": 0.297,
+      "step": 1555
+    },
+    {
+      "epoch": 2.98944,
+      "grad_norm": 0.5567149519920349,
+      "learning_rate": 2.0027310073833516e-10,
+      "loss": 0.2985,
+      "step": 1556
+    },
+    {
+      "epoch": 2.9913600000000002,
+      "grad_norm": 0.547590970993042,
+      "learning_rate": 1.1265394818993358e-10,
+      "loss": 0.2916,
+      "step": 1557
+    },
+    {
+      "epoch": 2.99328,
+      "grad_norm": 0.4852241277694702,
+      "learning_rate": 5.0068525870305974e-11,
+      "loss": 0.2621,
+      "step": 1558
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.5260079503059387,
+      "learning_rate": 1.251714713546015e-11,
+      "loss": 0.2989,
+      "step": 1559
+    },
+    {
+      "epoch": 2.99712,
+      "grad_norm": 0.5325884222984314,
+      "learning_rate": 0.0,
+      "loss": 0.2841,
+      "step": 1560
+    },
+    {
+      "epoch": 2.99712,
+      "step": 1560,
+      "total_flos": 228476960497664.0,
+      "train_loss": 0.3325708398452172,
+      "train_runtime": 14325.8133,
+      "train_samples_per_second": 10.471,
+      "train_steps_per_second": 0.109
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1560,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 228476960497664.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}