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": 4.794714450836182,
+      "learning_rate": 6.41025641025641e-08,
+      "loss": 0.5572,
+      "step": 1
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 4.548453330993652,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 0.5289,
+      "step": 2
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 4.554799556732178,
+      "learning_rate": 1.9230769230769234e-07,
+      "loss": 0.5279,
+      "step": 3
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 4.350290775299072,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 0.5273,
+      "step": 4
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 4.668323516845703,
+      "learning_rate": 3.205128205128205e-07,
+      "loss": 0.5837,
+      "step": 5
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 4.25593376159668,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 0.5365,
+      "step": 6
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 4.435504913330078,
+      "learning_rate": 4.4871794871794876e-07,
+      "loss": 0.5539,
+      "step": 7
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 4.584949970245361,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 0.5765,
+      "step": 8
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 4.665410041809082,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 0.534,
+      "step": 9
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 3.739230155944824,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 0.5364,
+      "step": 10
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 3.928215742111206,
+      "learning_rate": 7.051282051282052e-07,
+      "loss": 0.5264,
+      "step": 11
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 3.7986998558044434,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.5429,
+      "step": 12
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 2.96742844581604,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.5747,
+      "step": 13
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 2.881573438644409,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 0.5471,
+      "step": 14
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 2.5934195518493652,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 0.5479,
+      "step": 15
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 2.7261855602264404,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.4923,
+      "step": 16
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 2.330930709838867,
+      "learning_rate": 1.0897435897435899e-06,
+      "loss": 0.5848,
+      "step": 17
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 1.980789065361023,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 0.4853,
+      "step": 18
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 2.1258254051208496,
+      "learning_rate": 1.217948717948718e-06,
+      "loss": 0.506,
+      "step": 19
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 1.8639053106307983,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.4573,
+      "step": 20
+    },
+    {
+      "epoch": 0.04032,
+      "grad_norm": 1.6065678596496582,
+      "learning_rate": 1.3461538461538462e-06,
+      "loss": 0.4433,
+      "step": 21
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 1.7313145399093628,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 0.5144,
+      "step": 22
+    },
+    {
+      "epoch": 0.04416,
+      "grad_norm": 1.4784044027328491,
+      "learning_rate": 1.4743589743589745e-06,
+      "loss": 0.4692,
+      "step": 23
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 1.474663257598877,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.4937,
+      "step": 24
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 1.7213897705078125,
+      "learning_rate": 1.602564102564103e-06,
+      "loss": 0.5155,
+      "step": 25
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 1.6058335304260254,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.4463,
+      "step": 26
+    },
+    {
+      "epoch": 0.05184,
+      "grad_norm": 1.4729127883911133,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 0.5221,
+      "step": 27
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 1.304975986480713,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 0.4906,
+      "step": 28
+    },
+    {
+      "epoch": 0.05568,
+      "grad_norm": 1.208889126777649,
+      "learning_rate": 1.8589743589743592e-06,
+      "loss": 0.4469,
+      "step": 29
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 1.2039344310760498,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.5211,
+      "step": 30
+    },
+    {
+      "epoch": 0.05952,
+      "grad_norm": 0.9915562272071838,
+      "learning_rate": 1.987179487179487e-06,
+      "loss": 0.5018,
+      "step": 31
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 0.9129545092582703,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.4418,
+      "step": 32
+    },
+    {
+      "epoch": 0.06336,
+      "grad_norm": 0.9248676300048828,
+      "learning_rate": 2.1153846153846155e-06,
+      "loss": 0.4894,
+      "step": 33
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 1.0229066610336304,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 0.5429,
+      "step": 34
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 1.056071162223816,
+      "learning_rate": 2.243589743589744e-06,
+      "loss": 0.4699,
+      "step": 35
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.9281153678894043,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.3992,
+      "step": 36
+    },
+    {
+      "epoch": 0.07104,
+      "grad_norm": 0.9566601514816284,
+      "learning_rate": 2.371794871794872e-06,
+      "loss": 0.4824,
+      "step": 37
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.9954881072044373,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 0.5959,
+      "step": 38
+    },
+    {
+      "epoch": 0.07488,
+      "grad_norm": 0.917946457862854,
+      "learning_rate": 2.5e-06,
+      "loss": 0.4385,
+      "step": 39
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.8246254324913025,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.4143,
+      "step": 40
+    },
+    {
+      "epoch": 0.07872,
+      "grad_norm": 0.8681561946868896,
+      "learning_rate": 2.6282051282051286e-06,
+      "loss": 0.4518,
+      "step": 41
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.7804803848266602,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.4778,
+      "step": 42
+    },
+    {
+      "epoch": 0.08256,
+      "grad_norm": 0.7801416516304016,
+      "learning_rate": 2.756410256410257e-06,
+      "loss": 0.5147,
+      "step": 43
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.8374272584915161,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 0.4261,
+      "step": 44
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.8440502285957336,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 0.4642,
+      "step": 45
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.7083030343055725,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 0.3911,
+      "step": 46
+    },
+    {
+      "epoch": 0.09024,
+      "grad_norm": 0.8209000825881958,
+      "learning_rate": 3.012820512820513e-06,
+      "loss": 0.4815,
+      "step": 47
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.8322147130966187,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.5364,
+      "step": 48
+    },
+    {
+      "epoch": 0.09408,
+      "grad_norm": 0.7259360551834106,
+      "learning_rate": 3.141025641025641e-06,
+      "loss": 0.4375,
+      "step": 49
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.7483527660369873,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.4116,
+      "step": 50
+    },
+    {
+      "epoch": 0.09792,
+      "grad_norm": 0.6951253414154053,
+      "learning_rate": 3.2692307692307696e-06,
+      "loss": 0.4373,
+      "step": 51
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.8435599207878113,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.4604,
+      "step": 52
+    },
+    {
+      "epoch": 0.10176,
+      "grad_norm": 0.6284298896789551,
+      "learning_rate": 3.397435897435898e-06,
+      "loss": 0.3838,
+      "step": 53
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 0.6833722591400146,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.4403,
+      "step": 54
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.6884530186653137,
+      "learning_rate": 3.5256410256410263e-06,
+      "loss": 0.4289,
+      "step": 55
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.69256591796875,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 0.46,
+      "step": 56
+    },
+    {
+      "epoch": 0.10944,
+      "grad_norm": 0.6705825328826904,
+      "learning_rate": 3.653846153846154e-06,
+      "loss": 0.4314,
+      "step": 57
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 0.6995909214019775,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 0.4822,
+      "step": 58
+    },
+    {
+      "epoch": 0.11328,
+      "grad_norm": 0.6814873218536377,
+      "learning_rate": 3.782051282051282e-06,
+      "loss": 0.4168,
+      "step": 59
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.7259885668754578,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.4753,
+      "step": 60
+    },
+    {
+      "epoch": 0.11712,
+      "grad_norm": 0.6892727017402649,
+      "learning_rate": 3.910256410256411e-06,
+      "loss": 0.4129,
+      "step": 61
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.7582497000694275,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 0.4572,
+      "step": 62
+    },
+    {
+      "epoch": 0.12096,
+      "grad_norm": 0.780265212059021,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 0.4214,
+      "step": 63
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.6739459037780762,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.4117,
+      "step": 64
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.6292528510093689,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.4484,
+      "step": 65
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.6516684889793396,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.4726,
+      "step": 66
+    },
+    {
+      "epoch": 0.12864,
+      "grad_norm": 0.680350661277771,
+      "learning_rate": 4.294871794871795e-06,
+      "loss": 0.4938,
+      "step": 67
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.7237058877944946,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 0.4749,
+      "step": 68
+    },
+    {
+      "epoch": 0.13248,
+      "grad_norm": 0.73295658826828,
+      "learning_rate": 4.423076923076924e-06,
+      "loss": 0.4643,
+      "step": 69
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.6375187039375305,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.4392,
+      "step": 70
+    },
+    {
+      "epoch": 0.13632,
+      "grad_norm": 0.7278439998626709,
+      "learning_rate": 4.551282051282052e-06,
+      "loss": 0.4884,
+      "step": 71
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.739063024520874,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.4662,
+      "step": 72
+    },
+    {
+      "epoch": 0.14016,
+      "grad_norm": 0.6500280499458313,
+      "learning_rate": 4.6794871794871795e-06,
+      "loss": 0.4318,
+      "step": 73
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.7568433284759521,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 0.4708,
+      "step": 74
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.7707805037498474,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 0.3847,
+      "step": 75
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.6852964162826538,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 0.4196,
+      "step": 76
+    },
+    {
+      "epoch": 0.14784,
+      "grad_norm": 0.726940393447876,
+      "learning_rate": 4.935897435897436e-06,
+      "loss": 0.4267,
+      "step": 77
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.6975735425949097,
+      "learning_rate": 5e-06,
+      "loss": 0.4958,
+      "step": 78
+    },
+    {
+      "epoch": 0.15168,
+      "grad_norm": 0.7175726294517517,
+      "learning_rate": 5.064102564102565e-06,
+      "loss": 0.4589,
+      "step": 79
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.690256655216217,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.4189,
+      "step": 80
+    },
+    {
+      "epoch": 0.15552,
+      "grad_norm": 0.6484289169311523,
+      "learning_rate": 5.192307692307693e-06,
+      "loss": 0.3955,
+      "step": 81
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.7284436821937561,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 0.4585,
+      "step": 82
+    },
+    {
+      "epoch": 0.15936,
+      "grad_norm": 0.7118803858757019,
+      "learning_rate": 5.320512820512821e-06,
+      "loss": 0.4456,
+      "step": 83
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.7958849668502808,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.425,
+      "step": 84
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.6842144131660461,
+      "learning_rate": 5.448717948717949e-06,
+      "loss": 0.384,
+      "step": 85
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.6901838779449463,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 0.4534,
+      "step": 86
+    },
+    {
+      "epoch": 0.16704,
+      "grad_norm": 0.6572390198707581,
+      "learning_rate": 5.576923076923077e-06,
+      "loss": 0.4093,
+      "step": 87
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.7032268047332764,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 0.4149,
+      "step": 88
+    },
+    {
+      "epoch": 0.17088,
+      "grad_norm": 0.8220327496528625,
+      "learning_rate": 5.705128205128206e-06,
+      "loss": 0.4735,
+      "step": 89
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.909488320350647,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 0.4418,
+      "step": 90
+    },
+    {
+      "epoch": 0.17472,
+      "grad_norm": 0.7101952433586121,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.472,
+      "step": 91
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.8168191909790039,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 0.4661,
+      "step": 92
+    },
+    {
+      "epoch": 0.17856,
+      "grad_norm": 0.7897205352783203,
+      "learning_rate": 5.961538461538462e-06,
+      "loss": 0.41,
+      "step": 93
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.7651470303535461,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 0.4122,
+      "step": 94
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.6624057292938232,
+      "learning_rate": 6.08974358974359e-06,
+      "loss": 0.4467,
+      "step": 95
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.8730647563934326,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.4507,
+      "step": 96
+    },
+    {
+      "epoch": 0.18624,
+      "grad_norm": 1.1064871549606323,
+      "learning_rate": 6.217948717948718e-06,
+      "loss": 0.4492,
+      "step": 97
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.620094895362854,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.4097,
+      "step": 98
+    },
+    {
+      "epoch": 0.19008,
+      "grad_norm": 0.6669363975524902,
+      "learning_rate": 6.3461538461538466e-06,
+      "loss": 0.4264,
+      "step": 99
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.7385616898536682,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 0.398,
+      "step": 100
+    },
+    {
+      "epoch": 0.19392,
+      "grad_norm": 0.9896393418312073,
+      "learning_rate": 6.474358974358975e-06,
+      "loss": 0.494,
+      "step": 101
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.7829315662384033,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.4465,
+      "step": 102
+    },
+    {
+      "epoch": 0.19776,
+      "grad_norm": 0.6535574197769165,
+      "learning_rate": 6.602564102564103e-06,
+      "loss": 0.3951,
+      "step": 103
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.6937007308006287,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.414,
+      "step": 104
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.8500528335571289,
+      "learning_rate": 6.730769230769232e-06,
+      "loss": 0.4909,
+      "step": 105
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.8101775050163269,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 0.459,
+      "step": 106
+    },
+    {
+      "epoch": 0.20544,
+      "grad_norm": 0.6628045439720154,
+      "learning_rate": 6.858974358974359e-06,
+      "loss": 0.3955,
+      "step": 107
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.6649342775344849,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.4492,
+      "step": 108
+    },
+    {
+      "epoch": 0.20928,
+      "grad_norm": 0.7124112844467163,
+      "learning_rate": 6.9871794871794876e-06,
+      "loss": 0.3934,
+      "step": 109
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.8023260831832886,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 0.473,
+      "step": 110
+    },
+    {
+      "epoch": 0.21312,
+      "grad_norm": 0.7043531537055969,
+      "learning_rate": 7.115384615384616e-06,
+      "loss": 0.425,
+      "step": 111
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.6678853631019592,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.4305,
+      "step": 112
+    },
+    {
+      "epoch": 0.21696,
+      "grad_norm": 0.7686051726341248,
+      "learning_rate": 7.243589743589744e-06,
+      "loss": 0.4382,
+      "step": 113
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.8249655365943909,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.4338,
+      "step": 114
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.8552616238594055,
+      "learning_rate": 7.371794871794873e-06,
+      "loss": 0.4371,
+      "step": 115
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.6678420305252075,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 0.4315,
+      "step": 116
+    },
+    {
+      "epoch": 0.22464,
+      "grad_norm": 0.7653200030326843,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.4254,
+      "step": 117
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 0.8432084918022156,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.4101,
+      "step": 118
+    },
+    {
+      "epoch": 0.22848,
+      "grad_norm": 0.7131649255752563,
+      "learning_rate": 7.6282051282051286e-06,
+      "loss": 0.4324,
+      "step": 119
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.7141339182853699,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.4195,
+      "step": 120
+    },
+    {
+      "epoch": 0.23232,
+      "grad_norm": 0.6580355763435364,
+      "learning_rate": 7.756410256410258e-06,
+      "loss": 0.4326,
+      "step": 121
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.7327094078063965,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 0.4524,
+      "step": 122
+    },
+    {
+      "epoch": 0.23616,
+      "grad_norm": 0.7456380724906921,
+      "learning_rate": 7.884615384615384e-06,
+      "loss": 0.4548,
+      "step": 123
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.8268933892250061,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 0.4222,
+      "step": 124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7346514463424683,
+      "learning_rate": 8.012820512820515e-06,
+      "loss": 0.4601,
+      "step": 125
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.673126220703125,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.4118,
+      "step": 126
+    },
+    {
+      "epoch": 0.24384,
+      "grad_norm": 0.8967100977897644,
+      "learning_rate": 8.141025641025641e-06,
+      "loss": 0.4359,
+      "step": 127
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.8696978688240051,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.391,
+      "step": 128
+    },
+    {
+      "epoch": 0.24768,
+      "grad_norm": 0.6482024192810059,
+      "learning_rate": 8.26923076923077e-06,
+      "loss": 0.4203,
+      "step": 129
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.7619929313659668,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.4814,
+      "step": 130
+    },
+    {
+      "epoch": 0.25152,
+      "grad_norm": 0.8412722945213318,
+      "learning_rate": 8.397435897435898e-06,
+      "loss": 0.3987,
+      "step": 131
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.8173087239265442,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.435,
+      "step": 132
+    },
+    {
+      "epoch": 0.25536,
+      "grad_norm": 0.6569345593452454,
+      "learning_rate": 8.525641025641026e-06,
+      "loss": 0.4308,
+      "step": 133
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.7619485855102539,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 0.3929,
+      "step": 134
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 1.025268316268921,
+      "learning_rate": 8.653846153846155e-06,
+      "loss": 0.4622,
+      "step": 135
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.901294469833374,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 0.4536,
+      "step": 136
+    },
+    {
+      "epoch": 0.26304,
+      "grad_norm": 0.777611255645752,
+      "learning_rate": 8.782051282051283e-06,
+      "loss": 0.4844,
+      "step": 137
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.9970526099205017,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.4652,
+      "step": 138
+    },
+    {
+      "epoch": 0.26688,
+      "grad_norm": 0.7287512421607971,
+      "learning_rate": 8.910256410256411e-06,
+      "loss": 0.4126,
+      "step": 139
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.7563086748123169,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.4678,
+      "step": 140
+    },
+    {
+      "epoch": 0.27072,
+      "grad_norm": 0.7214654684066772,
+      "learning_rate": 9.03846153846154e-06,
+      "loss": 0.4665,
+      "step": 141
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.9660215377807617,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 0.4936,
+      "step": 142
+    },
+    {
+      "epoch": 0.27456,
+      "grad_norm": 0.7381386756896973,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.4179,
+      "step": 143
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.8013956546783447,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.4581,
+      "step": 144
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.7392879128456116,
+      "learning_rate": 9.294871794871796e-06,
+      "loss": 0.422,
+      "step": 145
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.6284559965133667,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 0.4316,
+      "step": 146
+    },
+    {
+      "epoch": 0.28224,
+      "grad_norm": 0.7184411883354187,
+      "learning_rate": 9.423076923076923e-06,
+      "loss": 0.4438,
+      "step": 147
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.7426531314849854,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 0.4487,
+      "step": 148
+    },
+    {
+      "epoch": 0.28608,
+      "grad_norm": 0.7444310784339905,
+      "learning_rate": 9.551282051282053e-06,
+      "loss": 0.4733,
+      "step": 149
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.6773303151130676,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.4069,
+      "step": 150
+    },
+    {
+      "epoch": 0.28992,
+      "grad_norm": 0.6402796506881714,
+      "learning_rate": 9.67948717948718e-06,
+      "loss": 0.4233,
+      "step": 151
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 0.7720417380332947,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 0.4242,
+      "step": 152
+    },
+    {
+      "epoch": 0.29376,
+      "grad_norm": 0.7832808494567871,
+      "learning_rate": 9.807692307692308e-06,
+      "loss": 0.4129,
+      "step": 153
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.6694121360778809,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.412,
+      "step": 154
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.6468183398246765,
+      "learning_rate": 9.935897435897437e-06,
+      "loss": 0.4127,
+      "step": 155
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.6488353610038757,
+      "learning_rate": 1e-05,
+      "loss": 0.4603,
+      "step": 156
+    },
+    {
+      "epoch": 0.30144,
+      "grad_norm": 0.6530375480651855,
+      "learning_rate": 9.999987482852865e-06,
+      "loss": 0.3888,
+      "step": 157
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.6056983470916748,
+      "learning_rate": 9.99994993147413e-06,
+      "loss": 0.4075,
+      "step": 158
+    },
+    {
+      "epoch": 0.30528,
+      "grad_norm": 0.7294408082962036,
+      "learning_rate": 9.99988734605181e-06,
+      "loss": 0.466,
+      "step": 159
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.6635957956314087,
+      "learning_rate": 9.999799726899261e-06,
+      "loss": 0.4384,
+      "step": 160
+    },
+    {
+      "epoch": 0.30912,
+      "grad_norm": 0.6894209384918213,
+      "learning_rate": 9.999687074455182e-06,
+      "loss": 0.4938,
+      "step": 161
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.7321233749389648,
+      "learning_rate": 9.999549389283605e-06,
+      "loss": 0.4634,
+      "step": 162
+    },
+    {
+      "epoch": 0.31296,
+      "grad_norm": 0.6581103205680847,
+      "learning_rate": 9.999386672073902e-06,
+      "loss": 0.4564,
+      "step": 163
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.7082487344741821,
+      "learning_rate": 9.999198923640774e-06,
+      "loss": 0.4423,
+      "step": 164
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.6569457650184631,
+      "learning_rate": 9.998986144924253e-06,
+      "loss": 0.4199,
+      "step": 165
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.6754735708236694,
+      "learning_rate": 9.998748336989687e-06,
+      "loss": 0.4647,
+      "step": 166
+    },
+    {
+      "epoch": 0.32064,
+      "grad_norm": 0.7099205255508423,
+      "learning_rate": 9.998485501027752e-06,
+      "loss": 0.4512,
+      "step": 167
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.6508955955505371,
+      "learning_rate": 9.998197638354428e-06,
+      "loss": 0.4003,
+      "step": 168
+    },
+    {
+      "epoch": 0.32448,
+      "grad_norm": 0.8200316429138184,
+      "learning_rate": 9.997884750411004e-06,
+      "loss": 0.4731,
+      "step": 169
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.7495537996292114,
+      "learning_rate": 9.997546838764066e-06,
+      "loss": 0.4161,
+      "step": 170
+    },
+    {
+      "epoch": 0.32832,
+      "grad_norm": 0.6588367223739624,
+      "learning_rate": 9.997183905105487e-06,
+      "loss": 0.485,
+      "step": 171
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.7129039764404297,
+      "learning_rate": 9.996795951252427e-06,
+      "loss": 0.476,
+      "step": 172
+    },
+    {
+      "epoch": 0.33216,
+      "grad_norm": 0.7017580270767212,
+      "learning_rate": 9.996382979147316e-06,
+      "loss": 0.4879,
+      "step": 173
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.6430830955505371,
+      "learning_rate": 9.995944990857848e-06,
+      "loss": 0.4107,
+      "step": 174
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.7420734763145447,
+      "learning_rate": 9.995481988576968e-06,
+      "loss": 0.4112,
+      "step": 175
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.6747193336486816,
+      "learning_rate": 9.994993974622863e-06,
+      "loss": 0.4599,
+      "step": 176
+    },
+    {
+      "epoch": 0.33984,
+      "grad_norm": 0.7100633382797241,
+      "learning_rate": 9.994480951438948e-06,
+      "loss": 0.4171,
+      "step": 177
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.8811770677566528,
+      "learning_rate": 9.993942921593858e-06,
+      "loss": 0.4258,
+      "step": 178
+    },
+    {
+      "epoch": 0.34368,
+      "grad_norm": 0.6834702491760254,
+      "learning_rate": 9.993379887781436e-06,
+      "loss": 0.4173,
+      "step": 179
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.7509815096855164,
+      "learning_rate": 9.992791852820709e-06,
+      "loss": 0.4462,
+      "step": 180
+    },
+    {
+      "epoch": 0.34752,
+      "grad_norm": 0.8992944955825806,
+      "learning_rate": 9.992178819655889e-06,
+      "loss": 0.4775,
+      "step": 181
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 0.671084463596344,
+      "learning_rate": 9.991540791356342e-06,
+      "loss": 0.4299,
+      "step": 182
+    },
+    {
+      "epoch": 0.35136,
+      "grad_norm": 0.6163865327835083,
+      "learning_rate": 9.990877771116588e-06,
+      "loss": 0.407,
+      "step": 183
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.7145616412162781,
+      "learning_rate": 9.990189762256275e-06,
+      "loss": 0.4024,
+      "step": 184
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 0.7685941457748413,
+      "learning_rate": 9.98947676822017e-06,
+      "loss": 0.4636,
+      "step": 185
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.7406034469604492,
+      "learning_rate": 9.988738792578126e-06,
+      "loss": 0.4403,
+      "step": 186
+    },
+    {
+      "epoch": 0.35904,
+      "grad_norm": 0.6860826015472412,
+      "learning_rate": 9.987975839025091e-06,
+      "loss": 0.4432,
+      "step": 187
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.7041403651237488,
+      "learning_rate": 9.987187911381059e-06,
+      "loss": 0.4213,
+      "step": 188
+    },
+    {
+      "epoch": 0.36288,
+      "grad_norm": 0.6703921556472778,
+      "learning_rate": 9.986375013591077e-06,
+      "loss": 0.4049,
+      "step": 189
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.7673277258872986,
+      "learning_rate": 9.985537149725207e-06,
+      "loss": 0.4543,
+      "step": 190
+    },
+    {
+      "epoch": 0.36672,
+      "grad_norm": 0.6445493698120117,
+      "learning_rate": 9.984674323978517e-06,
+      "loss": 0.4421,
+      "step": 191
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.7062714099884033,
+      "learning_rate": 9.983786540671052e-06,
+      "loss": 0.4792,
+      "step": 192
+    },
+    {
+      "epoch": 0.37056,
+      "grad_norm": 0.9028751850128174,
+      "learning_rate": 9.982873804247817e-06,
+      "loss": 0.5189,
+      "step": 193
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.6526753902435303,
+      "learning_rate": 9.981936119278758e-06,
+      "loss": 0.4676,
+      "step": 194
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 0.6702433824539185,
+      "learning_rate": 9.980973490458728e-06,
+      "loss": 0.4327,
+      "step": 195
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.6686440110206604,
+      "learning_rate": 9.979985922607476e-06,
+      "loss": 0.4439,
+      "step": 196
+    },
+    {
+      "epoch": 0.37824,
+      "grad_norm": 0.7222212553024292,
+      "learning_rate": 9.978973420669613e-06,
+      "loss": 0.4263,
+      "step": 197
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.5915130972862244,
+      "learning_rate": 9.977935989714594e-06,
+      "loss": 0.4229,
+      "step": 198
+    },
+    {
+      "epoch": 0.38208,
+      "grad_norm": 0.6441517472267151,
+      "learning_rate": 9.97687363493669e-06,
+      "loss": 0.438,
+      "step": 199
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.725648045539856,
+      "learning_rate": 9.975786361654959e-06,
+      "loss": 0.4721,
+      "step": 200
+    },
+    {
+      "epoch": 0.38592,
+      "grad_norm": 0.6560202240943909,
+      "learning_rate": 9.974674175313228e-06,
+      "loss": 0.3761,
+      "step": 201
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.6972739100456238,
+      "learning_rate": 9.973537081480056e-06,
+      "loss": 0.4646,
+      "step": 202
+    },
+    {
+      "epoch": 0.38976,
+      "grad_norm": 0.5617237091064453,
+      "learning_rate": 9.972375085848712e-06,
+      "loss": 0.4114,
+      "step": 203
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 0.6500780582427979,
+      "learning_rate": 9.971188194237141e-06,
+      "loss": 0.4276,
+      "step": 204
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.6391369104385376,
+      "learning_rate": 9.969976412587943e-06,
+      "loss": 0.4606,
+      "step": 205
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.6811553835868835,
+      "learning_rate": 9.96873974696834e-06,
+      "loss": 0.4288,
+      "step": 206
+    },
+    {
+      "epoch": 0.39744,
+      "grad_norm": 0.6640064716339111,
+      "learning_rate": 9.967478203570143e-06,
+      "loss": 0.4155,
+      "step": 207
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.6844574809074402,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.4409,
+      "step": 208
+    },
+    {
+      "epoch": 0.40128,
+      "grad_norm": 0.6361032724380493,
+      "learning_rate": 9.96488050882796e-06,
+      "loss": 0.4491,
+      "step": 209
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.5900434851646423,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.4437,
+      "step": 210
+    },
+    {
+      "epoch": 0.40512,
+      "grad_norm": 0.687255859375,
+      "learning_rate": 9.9621833803865e-06,
+      "loss": 0.4203,
+      "step": 211
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.5974467396736145,
+      "learning_rate": 9.960797545330936e-06,
+      "loss": 0.4267,
+      "step": 212
+    },
+    {
+      "epoch": 0.40896,
+      "grad_norm": 0.6312135457992554,
+      "learning_rate": 9.95938687226226e-06,
+      "loss": 0.4462,
+      "step": 213
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.630776047706604,
+      "learning_rate": 9.95795136824351e-06,
+      "loss": 0.4024,
+      "step": 214
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 0.6300383806228638,
+      "learning_rate": 9.956491040462051e-06,
+      "loss": 0.4057,
+      "step": 215
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.6183411478996277,
+      "learning_rate": 9.955005896229543e-06,
+      "loss": 0.3925,
+      "step": 216
+    },
+    {
+      "epoch": 0.41664,
+      "grad_norm": 0.61863774061203,
+      "learning_rate": 9.95349594298189e-06,
+      "loss": 0.4484,
+      "step": 217
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.5878365635871887,
+      "learning_rate": 9.951961188279216e-06,
+      "loss": 0.4229,
+      "step": 218
+    },
+    {
+      "epoch": 0.42048,
+      "grad_norm": 0.6332812309265137,
+      "learning_rate": 9.950401639805822e-06,
+      "loss": 0.4414,
+      "step": 219
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.6726149320602417,
+      "learning_rate": 9.948817305370145e-06,
+      "loss": 0.4223,
+      "step": 220
+    },
+    {
+      "epoch": 0.42432,
+      "grad_norm": 0.6683290004730225,
+      "learning_rate": 9.947208192904722e-06,
+      "loss": 0.4361,
+      "step": 221
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.624133288860321,
+      "learning_rate": 9.94557431046616e-06,
+      "loss": 0.4223,
+      "step": 222
+    },
+    {
+      "epoch": 0.42816,
+      "grad_norm": 0.6811025142669678,
+      "learning_rate": 9.943915666235068e-06,
+      "loss": 0.4534,
+      "step": 223
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.6343306303024292,
+      "learning_rate": 9.942232268516051e-06,
+      "loss": 0.4455,
+      "step": 224
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 0.6387416124343872,
+      "learning_rate": 9.940524125737641e-06,
+      "loss": 0.4532,
+      "step": 225
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.6800745725631714,
+      "learning_rate": 9.938791246452267e-06,
+      "loss": 0.4158,
+      "step": 226
+    },
+    {
+      "epoch": 0.43584,
+      "grad_norm": 0.5845776200294495,
+      "learning_rate": 9.937033639336212e-06,
+      "loss": 0.375,
+      "step": 227
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.6098273396492004,
+      "learning_rate": 9.935251313189564e-06,
+      "loss": 0.4385,
+      "step": 228
+    },
+    {
+      "epoch": 0.43968,
+      "grad_norm": 0.7030534148216248,
+      "learning_rate": 9.933444276936185e-06,
+      "loss": 0.4351,
+      "step": 229
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.7210600972175598,
+      "learning_rate": 9.931612539623643e-06,
+      "loss": 0.432,
+      "step": 230
+    },
+    {
+      "epoch": 0.44352,
+      "grad_norm": 0.6231402158737183,
+      "learning_rate": 9.929756110423195e-06,
+      "loss": 0.397,
+      "step": 231
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.6284440755844116,
+      "learning_rate": 9.927874998629714e-06,
+      "loss": 0.4186,
+      "step": 232
+    },
+    {
+      "epoch": 0.44736,
+      "grad_norm": 0.6185586452484131,
+      "learning_rate": 9.925969213661664e-06,
+      "loss": 0.3991,
+      "step": 233
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.6638928651809692,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.4219,
+      "step": 234
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 0.7105096578598022,
+      "learning_rate": 9.92208366249333e-06,
+      "loss": 0.4497,
+      "step": 235
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.6381037831306458,
+      "learning_rate": 9.920103915747452e-06,
+      "loss": 0.4179,
+      "step": 236
+    },
+    {
+      "epoch": 0.45504,
+      "grad_norm": 0.5987352132797241,
+      "learning_rate": 9.91809953473572e-06,
+      "loss": 0.3952,
+      "step": 237
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.6437040567398071,
+      "learning_rate": 9.916070529493785e-06,
+      "loss": 0.4973,
+      "step": 238
+    },
+    {
+      "epoch": 0.45888,
+      "grad_norm": 0.6221977472305298,
+      "learning_rate": 9.914016910180593e-06,
+      "loss": 0.4405,
+      "step": 239
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.6501063704490662,
+      "learning_rate": 9.911938687078324e-06,
+      "loss": 0.4634,
+      "step": 240
+    },
+    {
+      "epoch": 0.46272,
+      "grad_norm": 0.6384264826774597,
+      "learning_rate": 9.90983587059235e-06,
+      "loss": 0.4436,
+      "step": 241
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 0.6768873333930969,
+      "learning_rate": 9.907708471251173e-06,
+      "loss": 0.4155,
+      "step": 242
+    },
+    {
+      "epoch": 0.46656,
+      "grad_norm": 0.6218531131744385,
+      "learning_rate": 9.905556499706384e-06,
+      "loss": 0.4123,
+      "step": 243
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.5687350630760193,
+      "learning_rate": 9.9033799667326e-06,
+      "loss": 0.4096,
+      "step": 244
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.6120277643203735,
+      "learning_rate": 9.901178883227413e-06,
+      "loss": 0.4207,
+      "step": 245
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 0.6349164843559265,
+      "learning_rate": 9.89895326021134e-06,
+      "loss": 0.4391,
+      "step": 246
+    },
+    {
+      "epoch": 0.47424,
+      "grad_norm": 0.6295879483222961,
+      "learning_rate": 9.896703108827758e-06,
+      "loss": 0.4619,
+      "step": 247
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.6572540998458862,
+      "learning_rate": 9.89442844034286e-06,
+      "loss": 0.4269,
+      "step": 248
+    },
+    {
+      "epoch": 0.47808,
+      "grad_norm": 0.5945138931274414,
+      "learning_rate": 9.89212926614559e-06,
+      "loss": 0.4383,
+      "step": 249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6150192022323608,
+      "learning_rate": 9.889805597747588e-06,
+      "loss": 0.3992,
+      "step": 250
+    },
+    {
+      "epoch": 0.48192,
+      "grad_norm": 0.6160690188407898,
+      "learning_rate": 9.887457446783133e-06,
+      "loss": 0.4265,
+      "step": 251
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.5769174098968506,
+      "learning_rate": 9.885084825009085e-06,
+      "loss": 0.4445,
+      "step": 252
+    },
+    {
+      "epoch": 0.48576,
+      "grad_norm": 0.6055333614349365,
+      "learning_rate": 9.88268774430483e-06,
+      "loss": 0.4388,
+      "step": 253
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.6275864243507385,
+      "learning_rate": 9.88026621667221e-06,
+      "loss": 0.4272,
+      "step": 254
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.6018068790435791,
+      "learning_rate": 9.87782025423547e-06,
+      "loss": 0.409,
+      "step": 255
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.5812795758247375,
+      "learning_rate": 9.875349869241202e-06,
+      "loss": 0.398,
+      "step": 256
+    },
+    {
+      "epoch": 0.49344,
+      "grad_norm": 0.579362690448761,
+      "learning_rate": 9.872855074058274e-06,
+      "loss": 0.4159,
+      "step": 257
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.6873140335083008,
+      "learning_rate": 9.870335881177774e-06,
+      "loss": 0.4524,
+      "step": 258
+    },
+    {
+      "epoch": 0.49728,
+      "grad_norm": 0.6087861657142639,
+      "learning_rate": 9.867792303212944e-06,
+      "loss": 0.4035,
+      "step": 259
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.6202896237373352,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.4074,
+      "step": 260
+    },
+    {
+      "epoch": 0.50112,
+      "grad_norm": 0.5939845442771912,
+      "learning_rate": 9.862632043093667e-06,
+      "loss": 0.395,
+      "step": 261
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.6610350608825684,
+      "learning_rate": 9.860015386775915e-06,
+      "loss": 0.4267,
+      "step": 262
+    },
+    {
+      "epoch": 0.50496,
+      "grad_norm": 0.6722172498703003,
+      "learning_rate": 9.857374397047093e-06,
+      "loss": 0.4425,
+      "step": 263
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.6368960738182068,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.3928,
+      "step": 264
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.6006765365600586,
+      "learning_rate": 9.852019470370254e-06,
+      "loss": 0.4043,
+      "step": 265
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.6915842890739441,
+      "learning_rate": 9.849305560233598e-06,
+      "loss": 0.4742,
+      "step": 266
+    },
+    {
+      "epoch": 0.51264,
+      "grad_norm": 0.6464575529098511,
+      "learning_rate": 9.846567370308462e-06,
+      "loss": 0.4735,
+      "step": 267
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.6143791675567627,
+      "learning_rate": 9.843804914304578e-06,
+      "loss": 0.4138,
+      "step": 268
+    },
+    {
+      "epoch": 0.51648,
+      "grad_norm": 0.6347455978393555,
+      "learning_rate": 9.841018206053167e-06,
+      "loss": 0.4569,
+      "step": 269
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.6528032422065735,
+      "learning_rate": 9.838207259506891e-06,
+      "loss": 0.4427,
+      "step": 270
+    },
+    {
+      "epoch": 0.52032,
+      "grad_norm": 0.6099432706832886,
+      "learning_rate": 9.835372088739758e-06,
+      "loss": 0.4502,
+      "step": 271
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.6289542317390442,
+      "learning_rate": 9.83251270794707e-06,
+      "loss": 0.4039,
+      "step": 272
+    },
+    {
+      "epoch": 0.52416,
+      "grad_norm": 0.6122071146965027,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.412,
+      "step": 273
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.5679907202720642,
+      "learning_rate": 9.826721373672235e-06,
+      "loss": 0.3802,
+      "step": 274
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.5918080806732178,
+      "learning_rate": 9.82378944918648e-06,
+      "loss": 0.4347,
+      "step": 275
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.5497361421585083,
+      "learning_rate": 9.820833372667813e-06,
+      "loss": 0.4382,
+      "step": 276
+    },
+    {
+      "epoch": 0.53184,
+      "grad_norm": 0.6946800947189331,
+      "learning_rate": 9.817853158916889e-06,
+      "loss": 0.5166,
+      "step": 277
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.6019163131713867,
+      "learning_rate": 9.814848822855216e-06,
+      "loss": 0.4029,
+      "step": 278
+    },
+    {
+      "epoch": 0.53568,
+      "grad_norm": 0.6074923276901245,
+      "learning_rate": 9.811820379525085e-06,
+      "loss": 0.4684,
+      "step": 279
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.6609743237495422,
+      "learning_rate": 9.80876784408948e-06,
+      "loss": 0.4236,
+      "step": 280
+    },
+    {
+      "epoch": 0.53952,
+      "grad_norm": 0.5728402137756348,
+      "learning_rate": 9.805691231832018e-06,
+      "loss": 0.443,
+      "step": 281
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 0.5943120718002319,
+      "learning_rate": 9.802590558156863e-06,
+      "loss": 0.4433,
+      "step": 282
+    },
+    {
+      "epoch": 0.54336,
+      "grad_norm": 0.5700099468231201,
+      "learning_rate": 9.799465838588646e-06,
+      "loss": 0.4127,
+      "step": 283
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.6051260828971863,
+      "learning_rate": 9.796317088772402e-06,
+      "loss": 0.4639,
+      "step": 284
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.6286738514900208,
+      "learning_rate": 9.793144324473473e-06,
+      "loss": 0.4209,
+      "step": 285
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.5681666731834412,
+      "learning_rate": 9.789947561577445e-06,
+      "loss": 0.3722,
+      "step": 286
+    },
+    {
+      "epoch": 0.55104,
+      "grad_norm": 0.5979229807853699,
+      "learning_rate": 9.786726816090058e-06,
+      "loss": 0.4756,
+      "step": 287
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.6027213335037231,
+      "learning_rate": 9.783482104137127e-06,
+      "loss": 0.4411,
+      "step": 288
+    },
+    {
+      "epoch": 0.55488,
+      "grad_norm": 0.6172551512718201,
+      "learning_rate": 9.78021344196447e-06,
+      "loss": 0.4524,
+      "step": 289
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.5813352465629578,
+      "learning_rate": 9.776920845937816e-06,
+      "loss": 0.3884,
+      "step": 290
+    },
+    {
+      "epoch": 0.55872,
+      "grad_norm": 0.5678222179412842,
+      "learning_rate": 9.77360433254273e-06,
+      "loss": 0.4155,
+      "step": 291
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.6574405431747437,
+      "learning_rate": 9.770263918384523e-06,
+      "loss": 0.4773,
+      "step": 292
+    },
+    {
+      "epoch": 0.56256,
+      "grad_norm": 0.5848900675773621,
+      "learning_rate": 9.766899620188181e-06,
+      "loss": 0.4535,
+      "step": 293
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.6124178171157837,
+      "learning_rate": 9.763511454798268e-06,
+      "loss": 0.4747,
+      "step": 294
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.5285205841064453,
+      "learning_rate": 9.760099439178852e-06,
+      "loss": 0.3537,
+      "step": 295
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.5999187231063843,
+      "learning_rate": 9.75666359041341e-06,
+      "loss": 0.4371,
+      "step": 296
+    },
+    {
+      "epoch": 0.57024,
+      "grad_norm": 0.5562451481819153,
+      "learning_rate": 9.753203925704756e-06,
+      "loss": 0.4199,
+      "step": 297
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.5493106245994568,
+      "learning_rate": 9.749720462374939e-06,
+      "loss": 0.3849,
+      "step": 298
+    },
+    {
+      "epoch": 0.57408,
+      "grad_norm": 0.5920756459236145,
+      "learning_rate": 9.74621321786517e-06,
+      "loss": 0.3978,
+      "step": 299
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.5868974328041077,
+      "learning_rate": 9.742682209735727e-06,
+      "loss": 0.4195,
+      "step": 300
+    },
+    {
+      "epoch": 0.57792,
+      "grad_norm": 0.5794445872306824,
+      "learning_rate": 9.739127455665872e-06,
+      "loss": 0.4021,
+      "step": 301
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.5720506310462952,
+      "learning_rate": 9.735548973453753e-06,
+      "loss": 0.4117,
+      "step": 302
+    },
+    {
+      "epoch": 0.58176,
+      "grad_norm": 0.6064119935035706,
+      "learning_rate": 9.731946781016328e-06,
+      "loss": 0.4126,
+      "step": 303
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.620589554309845,
+      "learning_rate": 9.728320896389263e-06,
+      "loss": 0.4525,
+      "step": 304
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.5793087482452393,
+      "learning_rate": 9.724671337726855e-06,
+      "loss": 0.4321,
+      "step": 305
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.5964763760566711,
+      "learning_rate": 9.720998123301924e-06,
+      "loss": 0.426,
+      "step": 306
+    },
+    {
+      "epoch": 0.58944,
+      "grad_norm": 0.6096430420875549,
+      "learning_rate": 9.717301271505739e-06,
+      "loss": 0.4212,
+      "step": 307
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.5234261751174927,
+      "learning_rate": 9.713580800847917e-06,
+      "loss": 0.391,
+      "step": 308
+    },
+    {
+      "epoch": 0.59328,
+      "grad_norm": 0.5809370279312134,
+      "learning_rate": 9.709836729956326e-06,
+      "loss": 0.445,
+      "step": 309
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.5623607039451599,
+      "learning_rate": 9.706069077577e-06,
+      "loss": 0.4148,
+      "step": 310
+    },
+    {
+      "epoch": 0.59712,
+      "grad_norm": 0.5807476043701172,
+      "learning_rate": 9.702277862574048e-06,
+      "loss": 0.4191,
+      "step": 311
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.622277021408081,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.4427,
+      "step": 312
+    },
+    {
+      "epoch": 0.60096,
+      "grad_norm": 0.6130335330963135,
+      "learning_rate": 9.694624820743446e-06,
+      "loss": 0.4362,
+      "step": 313
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.5820890069007874,
+      "learning_rate": 9.690763032233498e-06,
+      "loss": 0.4592,
+      "step": 314
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.5634021162986755,
+      "learning_rate": 9.686877757735126e-06,
+      "loss": 0.4113,
+      "step": 315
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.5894038081169128,
+      "learning_rate": 9.682969016701357e-06,
+      "loss": 0.4355,
+      "step": 316
+    },
+    {
+      "epoch": 0.60864,
+      "grad_norm": 0.6619282364845276,
+      "learning_rate": 9.679036828702704e-06,
+      "loss": 0.4613,
+      "step": 317
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.6012680530548096,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.4147,
+      "step": 318
+    },
+    {
+      "epoch": 0.61248,
+      "grad_norm": 0.5383720993995667,
+      "learning_rate": 9.671102190679678e-06,
+      "loss": 0.3618,
+      "step": 319
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.5907011032104492,
+      "learning_rate": 9.66709978038292e-06,
+      "loss": 0.4541,
+      "step": 320
+    },
+    {
+      "epoch": 0.61632,
+      "grad_norm": 0.5301948189735413,
+      "learning_rate": 9.663074002576303e-06,
+      "loss": 0.4175,
+      "step": 321
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.6011309027671814,
+      "learning_rate": 9.659024877416328e-06,
+      "loss": 0.4153,
+      "step": 322
+    },
+    {
+      "epoch": 0.62016,
+      "grad_norm": 0.6379223465919495,
+      "learning_rate": 9.654952425176392e-06,
+      "loss": 0.425,
+      "step": 323
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.5878227949142456,
+      "learning_rate": 9.650856666246693e-06,
+      "loss": 0.4245,
+      "step": 324
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.5655598044395447,
+      "learning_rate": 9.646737621134112e-06,
+      "loss": 0.3999,
+      "step": 325
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.609563410282135,
+      "learning_rate": 9.642595310462133e-06,
+      "loss": 0.3977,
+      "step": 326
+    },
+    {
+      "epoch": 0.62784,
+      "grad_norm": 0.5848413109779358,
+      "learning_rate": 9.638429754970715e-06,
+      "loss": 0.3945,
+      "step": 327
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.5481098294258118,
+      "learning_rate": 9.63424097551621e-06,
+      "loss": 0.426,
+      "step": 328
+    },
+    {
+      "epoch": 0.63168,
+      "grad_norm": 0.5994529724121094,
+      "learning_rate": 9.630028993071244e-06,
+      "loss": 0.4887,
+      "step": 329
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.5384455323219299,
+      "learning_rate": 9.62579382872462e-06,
+      "loss": 0.3897,
+      "step": 330
+    },
+    {
+      "epoch": 0.63552,
+      "grad_norm": 0.5639839172363281,
+      "learning_rate": 9.621535503681205e-06,
+      "loss": 0.4105,
+      "step": 331
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.5973184108734131,
+      "learning_rate": 9.617254039261835e-06,
+      "loss": 0.4521,
+      "step": 332
+    },
+    {
+      "epoch": 0.63936,
+      "grad_norm": 0.5906018018722534,
+      "learning_rate": 9.612949456903195e-06,
+      "loss": 0.4512,
+      "step": 333
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.6174817681312561,
+      "learning_rate": 9.608621778157722e-06,
+      "loss": 0.4601,
+      "step": 334
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.5569174885749817,
+      "learning_rate": 9.604271024693495e-06,
+      "loss": 0.3926,
+      "step": 335
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.5581832528114319,
+      "learning_rate": 9.599897218294122e-06,
+      "loss": 0.4407,
+      "step": 336
+    },
+    {
+      "epoch": 0.64704,
+      "grad_norm": 0.6117115616798401,
+      "learning_rate": 9.595500380858632e-06,
+      "loss": 0.4793,
+      "step": 337
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.6087374091148376,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 0.451,
+      "step": 338
+    },
+    {
+      "epoch": 0.65088,
+      "grad_norm": 0.6251009106636047,
+      "learning_rate": 9.586637701051886e-06,
+      "loss": 0.3753,
+      "step": 339
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.5525182485580444,
+      "learning_rate": 9.582171903054815e-06,
+      "loss": 0.4089,
+      "step": 340
+    },
+    {
+      "epoch": 0.65472,
+      "grad_norm": 0.6613112688064575,
+      "learning_rate": 9.577683162769781e-06,
+      "loss": 0.4657,
+      "step": 341
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.5987522602081299,
+      "learning_rate": 9.573171502671273e-06,
+      "loss": 0.4392,
+      "step": 342
+    },
+    {
+      "epoch": 0.65856,
+      "grad_norm": 0.647398054599762,
+      "learning_rate": 9.568636945348534e-06,
+      "loss": 0.5387,
+      "step": 343
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.5918810963630676,
+      "learning_rate": 9.564079513505455e-06,
+      "loss": 0.4251,
+      "step": 344
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.5995827913284302,
+      "learning_rate": 9.55949922996045e-06,
+      "loss": 0.4066,
+      "step": 345
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.5888463854789734,
+      "learning_rate": 9.554896117646357e-06,
+      "loss": 0.4409,
+      "step": 346
+    },
+    {
+      "epoch": 0.66624,
+      "grad_norm": 0.5619899034500122,
+      "learning_rate": 9.550270199610307e-06,
+      "loss": 0.4193,
+      "step": 347
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.6418078541755676,
+      "learning_rate": 9.54562149901362e-06,
+      "loss": 0.4664,
+      "step": 348
+    },
+    {
+      "epoch": 0.67008,
+      "grad_norm": 0.6333869695663452,
+      "learning_rate": 9.54095003913168e-06,
+      "loss": 0.4392,
+      "step": 349
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.5406534671783447,
+      "learning_rate": 9.536255843353832e-06,
+      "loss": 0.4016,
+      "step": 350
+    },
+    {
+      "epoch": 0.67392,
+      "grad_norm": 0.59455806016922,
+      "learning_rate": 9.531538935183252e-06,
+      "loss": 0.4574,
+      "step": 351
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.5838527083396912,
+      "learning_rate": 9.526799338236828e-06,
+      "loss": 0.4236,
+      "step": 352
+    },
+    {
+      "epoch": 0.67776,
+      "grad_norm": 0.6499077081680298,
+      "learning_rate": 9.522037076245057e-06,
+      "loss": 0.4752,
+      "step": 353
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.5874121785163879,
+      "learning_rate": 9.517252173051912e-06,
+      "loss": 0.4171,
+      "step": 354
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.6133200526237488,
+      "learning_rate": 9.512444652614728e-06,
+      "loss": 0.4427,
+      "step": 355
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.5895383954048157,
+      "learning_rate": 9.507614539004082e-06,
+      "loss": 0.4339,
+      "step": 356
+    },
+    {
+      "epoch": 0.68544,
+      "grad_norm": 0.604487955570221,
+      "learning_rate": 9.502761856403668e-06,
+      "loss": 0.4111,
+      "step": 357
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.6303328275680542,
+      "learning_rate": 9.497886629110187e-06,
+      "loss": 0.4324,
+      "step": 358
+    },
+    {
+      "epoch": 0.68928,
+      "grad_norm": 0.6146121025085449,
+      "learning_rate": 9.49298888153321e-06,
+      "loss": 0.4098,
+      "step": 359
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.6525639891624451,
+      "learning_rate": 9.488068638195072e-06,
+      "loss": 0.4705,
+      "step": 360
+    },
+    {
+      "epoch": 0.69312,
+      "grad_norm": 0.5884914398193359,
+      "learning_rate": 9.483125923730731e-06,
+      "loss": 0.3988,
+      "step": 361
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.5391362309455872,
+      "learning_rate": 9.478160762887668e-06,
+      "loss": 0.4215,
+      "step": 362
+    },
+    {
+      "epoch": 0.69696,
+      "grad_norm": 0.6375428438186646,
+      "learning_rate": 9.473173180525737e-06,
+      "loss": 0.4543,
+      "step": 363
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.7459134459495544,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.4958,
+      "step": 364
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.5543957948684692,
+      "learning_rate": 9.463130851245899e-06,
+      "loss": 0.4124,
+      "step": 365
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.6711146235466003,
+      "learning_rate": 9.458076154608515e-06,
+      "loss": 0.4432,
+      "step": 366
+    },
+    {
+      "epoch": 0.70464,
+      "grad_norm": 0.6898747682571411,
+      "learning_rate": 9.452999137013064e-06,
+      "loss": 0.4302,
+      "step": 367
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.6382907629013062,
+      "learning_rate": 9.447899823879456e-06,
+      "loss": 0.4229,
+      "step": 368
+    },
+    {
+      "epoch": 0.70848,
+      "grad_norm": 0.5684764385223389,
+      "learning_rate": 9.442778240739234e-06,
+      "loss": 0.4112,
+      "step": 369
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.5965732336044312,
+      "learning_rate": 9.437634413235437e-06,
+      "loss": 0.4221,
+      "step": 370
+    },
+    {
+      "epoch": 0.71232,
+      "grad_norm": 0.6396543979644775,
+      "learning_rate": 9.43246836712249e-06,
+      "loss": 0.4621,
+      "step": 371
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.5788828134536743,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.47,
+      "step": 372
+    },
+    {
+      "epoch": 0.71616,
+      "grad_norm": 0.5893690586090088,
+      "learning_rate": 9.4220697226429e-06,
+      "loss": 0.4209,
+      "step": 373
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.5851166248321533,
+      "learning_rate": 9.416837176340807e-06,
+      "loss": 0.4736,
+      "step": 374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.700449526309967,
+      "learning_rate": 9.411582515558391e-06,
+      "loss": 0.4429,
+      "step": 375
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.634381890296936,
+      "learning_rate": 9.406305766604996e-06,
+      "loss": 0.4101,
+      "step": 376
+    },
+    {
+      "epoch": 0.72384,
+      "grad_norm": 0.6350017189979553,
+      "learning_rate": 9.401006955900555e-06,
+      "loss": 0.5164,
+      "step": 377
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.6132837533950806,
+      "learning_rate": 9.395686109975475e-06,
+      "loss": 0.4049,
+      "step": 378
+    },
+    {
+      "epoch": 0.72768,
+      "grad_norm": 0.7326400876045227,
+      "learning_rate": 9.390343255470471e-06,
+      "loss": 0.4473,
+      "step": 379
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.7300869226455688,
+      "learning_rate": 9.384978419136469e-06,
+      "loss": 0.457,
+      "step": 380
+    },
+    {
+      "epoch": 0.73152,
+      "grad_norm": 0.5627624988555908,
+      "learning_rate": 9.37959162783444e-06,
+      "loss": 0.4173,
+      "step": 381
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.5816049575805664,
+      "learning_rate": 9.374182908535293e-06,
+      "loss": 0.4075,
+      "step": 382
+    },
+    {
+      "epoch": 0.73536,
+      "grad_norm": 0.6696531176567078,
+      "learning_rate": 9.368752288319722e-06,
+      "loss": 0.4489,
+      "step": 383
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.5709719657897949,
+      "learning_rate": 9.363299794378072e-06,
+      "loss": 0.4153,
+      "step": 384
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.5871385335922241,
+      "learning_rate": 9.357825454010214e-06,
+      "loss": 0.4297,
+      "step": 385
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.6235693693161011,
+      "learning_rate": 9.352329294625397e-06,
+      "loss": 0.4268,
+      "step": 386
+    },
+    {
+      "epoch": 0.74304,
+      "grad_norm": 0.5997891426086426,
+      "learning_rate": 9.346811343742115e-06,
+      "loss": 0.4004,
+      "step": 387
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.5980295538902283,
+      "learning_rate": 9.34127162898797e-06,
+      "loss": 0.41,
+      "step": 388
+    },
+    {
+      "epoch": 0.74688,
+      "grad_norm": 0.5848520994186401,
+      "learning_rate": 9.335710178099528e-06,
+      "loss": 0.4029,
+      "step": 389
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.6449246406555176,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.428,
+      "step": 390
+    },
+    {
+      "epoch": 0.75072,
+      "grad_norm": 0.5723674297332764,
+      "learning_rate": 9.324522179410054e-06,
+      "loss": 0.4093,
+      "step": 391
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.6265943050384521,
+      "learning_rate": 9.318895687625752e-06,
+      "loss": 0.5252,
+      "step": 392
+    },
+    {
+      "epoch": 0.75456,
+      "grad_norm": 0.5903204679489136,
+      "learning_rate": 9.313247571740336e-06,
+      "loss": 0.3662,
+      "step": 393
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.5904682278633118,
+      "learning_rate": 9.307577860033123e-06,
+      "loss": 0.4642,
+      "step": 394
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 0.5902845859527588,
+      "learning_rate": 9.301886580891563e-06,
+      "loss": 0.4624,
+      "step": 395
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.521574079990387,
+      "learning_rate": 9.296173762811084e-06,
+      "loss": 0.4103,
+      "step": 396
+    },
+    {
+      "epoch": 0.76224,
+      "grad_norm": 0.63703453540802,
+      "learning_rate": 9.290439434394965e-06,
+      "loss": 0.428,
+      "step": 397
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.6102745532989502,
+      "learning_rate": 9.284683624354172e-06,
+      "loss": 0.4683,
+      "step": 398
+    },
+    {
+      "epoch": 0.76608,
+      "grad_norm": 0.6004009246826172,
+      "learning_rate": 9.278906361507238e-06,
+      "loss": 0.4261,
+      "step": 399
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.5680023431777954,
+      "learning_rate": 9.273107674780102e-06,
+      "loss": 0.3898,
+      "step": 400
+    },
+    {
+      "epoch": 0.76992,
+      "grad_norm": 0.5901056528091431,
+      "learning_rate": 9.26728759320597e-06,
+      "loss": 0.4552,
+      "step": 401
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.5861181616783142,
+      "learning_rate": 9.261446145925167e-06,
+      "loss": 0.4619,
+      "step": 402
+    },
+    {
+      "epoch": 0.77376,
+      "grad_norm": 0.5946195721626282,
+      "learning_rate": 9.255583362184998e-06,
+      "loss": 0.4523,
+      "step": 403
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.5602298378944397,
+      "learning_rate": 9.249699271339594e-06,
+      "loss": 0.4485,
+      "step": 404
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 0.5812543034553528,
+      "learning_rate": 9.243793902849764e-06,
+      "loss": 0.4654,
+      "step": 405
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.6400187015533447,
+      "learning_rate": 9.237867286282855e-06,
+      "loss": 0.4725,
+      "step": 406
+    },
+    {
+      "epoch": 0.78144,
+      "grad_norm": 0.5182756781578064,
+      "learning_rate": 9.231919451312603e-06,
+      "loss": 0.4331,
+      "step": 407
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.5420542359352112,
+      "learning_rate": 9.225950427718974e-06,
+      "loss": 0.4187,
+      "step": 408
+    },
+    {
+      "epoch": 0.78528,
+      "grad_norm": 0.575011670589447,
+      "learning_rate": 9.21996024538803e-06,
+      "loss": 0.3977,
+      "step": 409
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.5888107419013977,
+      "learning_rate": 9.213948934311767e-06,
+      "loss": 0.4624,
+      "step": 410
+    },
+    {
+      "epoch": 0.78912,
+      "grad_norm": 0.5791131258010864,
+      "learning_rate": 9.207916524587971e-06,
+      "loss": 0.4255,
+      "step": 411
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.5643811225891113,
+      "learning_rate": 9.201863046420065e-06,
+      "loss": 0.4356,
+      "step": 412
+    },
+    {
+      "epoch": 0.79296,
+      "grad_norm": 0.49760690331459045,
+      "learning_rate": 9.195788530116962e-06,
+      "loss": 0.3719,
+      "step": 413
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.5795424580574036,
+      "learning_rate": 9.189693006092907e-06,
+      "loss": 0.4399,
+      "step": 414
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 0.5537222623825073,
+      "learning_rate": 9.183576504867328e-06,
+      "loss": 0.4331,
+      "step": 415
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.528248131275177,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.3826,
+      "step": 416
+    },
+    {
+      "epoch": 0.80064,
+      "grad_norm": 0.541611909866333,
+      "learning_rate": 9.171280693414307e-06,
+      "loss": 0.4224,
+      "step": 417
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.5405563116073608,
+      "learning_rate": 9.165101444750259e-06,
+      "loss": 0.4304,
+      "step": 418
+    },
+    {
+      "epoch": 0.80448,
+      "grad_norm": 0.5537245869636536,
+      "learning_rate": 9.158901342011163e-06,
+      "loss": 0.4189,
+      "step": 419
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.5463759899139404,
+      "learning_rate": 9.152680416240059e-06,
+      "loss": 0.4083,
+      "step": 420
+    },
+    {
+      "epoch": 0.80832,
+      "grad_norm": 0.6896761655807495,
+      "learning_rate": 9.146438698584245e-06,
+      "loss": 0.4842,
+      "step": 421
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 0.6586357355117798,
+      "learning_rate": 9.140176220295118e-06,
+      "loss": 0.4278,
+      "step": 422
+    },
+    {
+      "epoch": 0.81216,
+      "grad_norm": 0.5526915192604065,
+      "learning_rate": 9.133893012728027e-06,
+      "loss": 0.3785,
+      "step": 423
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 0.6459162831306458,
+      "learning_rate": 9.1275891073421e-06,
+      "loss": 0.4189,
+      "step": 424
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.6584529280662537,
+      "learning_rate": 9.121264535700107e-06,
+      "loss": 0.4406,
+      "step": 425
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 0.5623626708984375,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.392,
+      "step": 426
+    },
+    {
+      "epoch": 0.81984,
+      "grad_norm": 0.5952051877975464,
+      "learning_rate": 9.10855352041618e-06,
+      "loss": 0.4722,
+      "step": 427
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 0.5571552515029907,
+      "learning_rate": 9.102167140416503e-06,
+      "loss": 0.4392,
+      "step": 428
+    },
+    {
+      "epoch": 0.82368,
+      "grad_norm": 0.595314085483551,
+      "learning_rate": 9.09576022144496e-06,
+      "loss": 0.4396,
+      "step": 429
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.6492066979408264,
+      "learning_rate": 9.089332795580085e-06,
+      "loss": 0.4403,
+      "step": 430
+    },
+    {
+      "epoch": 0.82752,
+      "grad_norm": 0.566978394985199,
+      "learning_rate": 9.082884895003098e-06,
+      "loss": 0.4289,
+      "step": 431
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 0.5322848558425903,
+      "learning_rate": 9.076416551997721e-06,
+      "loss": 0.3977,
+      "step": 432
+    },
+    {
+      "epoch": 0.83136,
+      "grad_norm": 0.5986286401748657,
+      "learning_rate": 9.06992779895004e-06,
+      "loss": 0.4326,
+      "step": 433
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 0.5888142585754395,
+      "learning_rate": 9.063418668348323e-06,
+      "loss": 0.3905,
+      "step": 434
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 0.5863139033317566,
+      "learning_rate": 9.056889192782865e-06,
+      "loss": 0.3788,
+      "step": 435
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 0.568784236907959,
+      "learning_rate": 9.050339404945834e-06,
+      "loss": 0.3838,
+      "step": 436
+    },
+    {
+      "epoch": 0.83904,
+      "grad_norm": 0.8033542037010193,
+      "learning_rate": 9.04376933763109e-06,
+      "loss": 0.4938,
+      "step": 437
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 0.6318104267120361,
+      "learning_rate": 9.037179023734036e-06,
+      "loss": 0.4305,
+      "step": 438
+    },
+    {
+      "epoch": 0.84288,
+      "grad_norm": 0.5430580377578735,
+      "learning_rate": 9.030568496251437e-06,
+      "loss": 0.3917,
+      "step": 439
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.5487372279167175,
+      "learning_rate": 9.023937788281278e-06,
+      "loss": 0.4044,
+      "step": 440
+    },
+    {
+      "epoch": 0.84672,
+      "grad_norm": 0.7298694252967834,
+      "learning_rate": 9.017286933022573e-06,
+      "loss": 0.5494,
+      "step": 441
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 0.7551047205924988,
+      "learning_rate": 9.01061596377522e-06,
+      "loss": 0.4642,
+      "step": 442
+    },
+    {
+      "epoch": 0.85056,
+      "grad_norm": 0.5360585451126099,
+      "learning_rate": 9.003924913939816e-06,
+      "loss": 0.4576,
+      "step": 443
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 0.671170711517334,
+      "learning_rate": 8.997213817017508e-06,
+      "loss": 0.4261,
+      "step": 444
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 0.6138496398925781,
+      "learning_rate": 8.990482706609805e-06,
+      "loss": 0.4213,
+      "step": 445
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 0.6199293732643127,
+      "learning_rate": 8.98373161641843e-06,
+      "loss": 0.4497,
+      "step": 446
+    },
+    {
+      "epoch": 0.85824,
+      "grad_norm": 0.5127760171890259,
+      "learning_rate": 8.97696058024514e-06,
+      "loss": 0.3875,
+      "step": 447
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 0.5768069624900818,
+      "learning_rate": 8.970169631991556e-06,
+      "loss": 0.4275,
+      "step": 448
+    },
+    {
+      "epoch": 0.86208,
+      "grad_norm": 0.5681172013282776,
+      "learning_rate": 8.963358805658998e-06,
+      "loss": 0.4574,
+      "step": 449
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.5542353391647339,
+      "learning_rate": 8.95652813534831e-06,
+      "loss": 0.4075,
+      "step": 450
+    },
+    {
+      "epoch": 0.86592,
+      "grad_norm": 0.56357741355896,
+      "learning_rate": 8.949677655259696e-06,
+      "loss": 0.4269,
+      "step": 451
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 0.5429962277412415,
+      "learning_rate": 8.942807399692543e-06,
+      "loss": 0.3894,
+      "step": 452
+    },
+    {
+      "epoch": 0.86976,
+      "grad_norm": 0.5428287386894226,
+      "learning_rate": 8.935917403045251e-06,
+      "loss": 0.4325,
+      "step": 453
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 0.5382682085037231,
+      "learning_rate": 8.92900769981506e-06,
+      "loss": 0.4247,
+      "step": 454
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 0.5389488339424133,
+      "learning_rate": 8.92207832459788e-06,
+      "loss": 0.3997,
+      "step": 455
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 0.5954126119613647,
+      "learning_rate": 8.915129312088112e-06,
+      "loss": 0.4077,
+      "step": 456
+    },
+    {
+      "epoch": 0.87744,
+      "grad_norm": 0.5518102049827576,
+      "learning_rate": 8.908160697078482e-06,
+      "loss": 0.4369,
+      "step": 457
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 0.6229602694511414,
+      "learning_rate": 8.901172514459864e-06,
+      "loss": 0.4735,
+      "step": 458
+    },
+    {
+      "epoch": 0.88128,
+      "grad_norm": 0.5334439277648926,
+      "learning_rate": 8.894164799221099e-06,
+      "loss": 0.4227,
+      "step": 459
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.5671420097351074,
+      "learning_rate": 8.88713758644883e-06,
+      "loss": 0.436,
+      "step": 460
+    },
+    {
+      "epoch": 0.88512,
+      "grad_norm": 0.5621917843818665,
+      "learning_rate": 8.88009091132732e-06,
+      "loss": 0.4559,
+      "step": 461
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 0.5582394599914551,
+      "learning_rate": 8.873024809138272e-06,
+      "loss": 0.4195,
+      "step": 462
+    },
+    {
+      "epoch": 0.88896,
+      "grad_norm": 0.5513332486152649,
+      "learning_rate": 8.86593931526067e-06,
+      "loss": 0.4371,
+      "step": 463
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 0.48494118452072144,
+      "learning_rate": 8.858834465170576e-06,
+      "loss": 0.3799,
+      "step": 464
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 0.5130149722099304,
+      "learning_rate": 8.851710294440974e-06,
+      "loss": 0.387,
+      "step": 465
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 0.611595630645752,
+      "learning_rate": 8.84456683874158e-06,
+      "loss": 0.4747,
+      "step": 466
+    },
+    {
+      "epoch": 0.89664,
+      "grad_norm": 0.5914707183837891,
+      "learning_rate": 8.837404133838667e-06,
+      "loss": 0.4508,
+      "step": 467
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 0.5432965159416199,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.4016,
+      "step": 468
+    },
+    {
+      "epoch": 0.90048,
+      "grad_norm": 0.5468518137931824,
+      "learning_rate": 8.823021119969102e-06,
+      "loss": 0.4149,
+      "step": 469
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.5804511308670044,
+      "learning_rate": 8.815800883016168e-06,
+      "loss": 0.4257,
+      "step": 470
+    },
+    {
+      "epoch": 0.90432,
+      "grad_norm": 0.6590668559074402,
+      "learning_rate": 8.808561540886796e-06,
+      "loss": 0.4441,
+      "step": 471
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 0.5548150539398193,
+      "learning_rate": 8.801303129827352e-06,
+      "loss": 0.3839,
+      "step": 472
+    },
+    {
+      "epoch": 0.90816,
+      "grad_norm": 0.5706216096878052,
+      "learning_rate": 8.794025686179677e-06,
+      "loss": 0.4647,
+      "step": 473
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 0.5289030075073242,
+      "learning_rate": 8.786729246380901e-06,
+      "loss": 0.3645,
+      "step": 474
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 0.6097282767295837,
+      "learning_rate": 8.779413846963267e-06,
+      "loss": 0.438,
+      "step": 475
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 0.5199111104011536,
+      "learning_rate": 8.772079524553951e-06,
+      "loss": 0.4026,
+      "step": 476
+    },
+    {
+      "epoch": 0.91584,
+      "grad_norm": 0.5859712362289429,
+      "learning_rate": 8.764726315874872e-06,
+      "loss": 0.4767,
+      "step": 477
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 0.5710155963897705,
+      "learning_rate": 8.757354257742501e-06,
+      "loss": 0.4556,
+      "step": 478
+    },
+    {
+      "epoch": 0.91968,
+      "grad_norm": 0.584568977355957,
+      "learning_rate": 8.749963387067697e-06,
+      "loss": 0.4386,
+      "step": 479
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.590584397315979,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.4557,
+      "step": 480
+    },
+    {
+      "epoch": 0.92352,
+      "grad_norm": 0.578332781791687,
+      "learning_rate": 8.735125356204982e-06,
+      "loss": 0.3876,
+      "step": 481
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 0.5473009347915649,
+      "learning_rate": 8.727678270308994e-06,
+      "loss": 0.4304,
+      "step": 482
+    },
+    {
+      "epoch": 0.92736,
+      "grad_norm": 0.6294142007827759,
+      "learning_rate": 8.720212520454054e-06,
+      "loss": 0.4221,
+      "step": 483
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 0.5512722134590149,
+      "learning_rate": 8.712728144020118e-06,
+      "loss": 0.3782,
+      "step": 484
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 0.5106618404388428,
+      "learning_rate": 8.705225178480397e-06,
+      "loss": 0.3912,
+      "step": 485
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 0.5381516814231873,
+      "learning_rate": 8.697703661401187e-06,
+      "loss": 0.4165,
+      "step": 486
+    },
+    {
+      "epoch": 0.93504,
+      "grad_norm": 0.5492082834243774,
+      "learning_rate": 8.69016363044166e-06,
+      "loss": 0.4538,
+      "step": 487
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 0.5437434315681458,
+      "learning_rate": 8.682605123353685e-06,
+      "loss": 0.36,
+      "step": 488
+    },
+    {
+      "epoch": 0.93888,
+      "grad_norm": 0.6525633931159973,
+      "learning_rate": 8.675028177981643e-06,
+      "loss": 0.4132,
+      "step": 489
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.5652430057525635,
+      "learning_rate": 8.66743283226223e-06,
+      "loss": 0.4679,
+      "step": 490
+    },
+    {
+      "epoch": 0.94272,
+      "grad_norm": 0.5668798685073853,
+      "learning_rate": 8.659819124224266e-06,
+      "loss": 0.4436,
+      "step": 491
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 0.5731828212738037,
+      "learning_rate": 8.652187091988516e-06,
+      "loss": 0.4414,
+      "step": 492
+    },
+    {
+      "epoch": 0.94656,
+      "grad_norm": 0.5267658233642578,
+      "learning_rate": 8.644536773767488e-06,
+      "loss": 0.3962,
+      "step": 493
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 0.526160717010498,
+      "learning_rate": 8.636868207865244e-06,
+      "loss": 0.4123,
+      "step": 494
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 0.5530787706375122,
+      "learning_rate": 8.629181432677213e-06,
+      "loss": 0.4315,
+      "step": 495
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 0.5535915493965149,
+      "learning_rate": 8.621476486689991e-06,
+      "loss": 0.4081,
+      "step": 496
+    },
+    {
+      "epoch": 0.95424,
+      "grad_norm": 0.5536392331123352,
+      "learning_rate": 8.613753408481158e-06,
+      "loss": 0.4526,
+      "step": 497
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 0.5924801826477051,
+      "learning_rate": 8.606012236719073e-06,
+      "loss": 0.4607,
+      "step": 498
+    },
+    {
+      "epoch": 0.95808,
+      "grad_norm": 0.5603639483451843,
+      "learning_rate": 8.598253010162693e-06,
+      "loss": 0.4305,
+      "step": 499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5551688075065613,
+      "learning_rate": 8.590475767661371e-06,
+      "loss": 0.4179,
+      "step": 500
+    },
+    {
+      "epoch": 0.96192,
+      "grad_norm": 0.6050774455070496,
+      "learning_rate": 8.58268054815466e-06,
+      "loss": 0.4733,
+      "step": 501
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 0.56812584400177,
+      "learning_rate": 8.574867390672124e-06,
+      "loss": 0.4617,
+      "step": 502
+    },
+    {
+      "epoch": 0.96576,
+      "grad_norm": 0.5135716199874878,
+      "learning_rate": 8.567036334333142e-06,
+      "loss": 0.3682,
+      "step": 503
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 0.54925936460495,
+      "learning_rate": 8.559187418346703e-06,
+      "loss": 0.4606,
+      "step": 504
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 0.49934929609298706,
+      "learning_rate": 8.551320682011227e-06,
+      "loss": 0.4085,
+      "step": 505
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 0.5625557899475098,
+      "learning_rate": 8.543436164714351e-06,
+      "loss": 0.4525,
+      "step": 506
+    },
+    {
+      "epoch": 0.97344,
+      "grad_norm": 0.5004458427429199,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.405,
+      "step": 507
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 0.5780734419822693,
+      "learning_rate": 8.527613945231886e-06,
+      "loss": 0.4327,
+      "step": 508
+    },
+    {
+      "epoch": 0.97728,
+      "grad_norm": 0.5655471086502075,
+      "learning_rate": 8.519676322265914e-06,
+      "loss": 0.4343,
+      "step": 509
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.5845369696617126,
+      "learning_rate": 8.511721076777388e-06,
+      "loss": 0.4477,
+      "step": 510
+    },
+    {
+      "epoch": 0.98112,
+      "grad_norm": 0.5894936323165894,
+      "learning_rate": 8.503748248597095e-06,
+      "loss": 0.4903,
+      "step": 511
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 0.5438454151153564,
+      "learning_rate": 8.495757877643857e-06,
+      "loss": 0.3845,
+      "step": 512
+    },
+    {
+      "epoch": 0.98496,
+      "grad_norm": 0.5755746364593506,
+      "learning_rate": 8.487750003924341e-06,
+      "loss": 0.3887,
+      "step": 513
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 0.5982465147972107,
+      "learning_rate": 8.479724667532836e-06,
+      "loss": 0.4191,
+      "step": 514
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 0.5450482964515686,
+      "learning_rate": 8.471681908651067e-06,
+      "loss": 0.4346,
+      "step": 515
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 0.5907667875289917,
+      "learning_rate": 8.463621767547998e-06,
+      "loss": 0.4163,
+      "step": 516
+    },
+    {
+      "epoch": 0.99264,
+      "grad_norm": 0.6019169688224792,
+      "learning_rate": 8.455544284579614e-06,
+      "loss": 0.4192,
+      "step": 517
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 0.5811045169830322,
+      "learning_rate": 8.447449500188731e-06,
+      "loss": 0.4444,
+      "step": 518
+    },
+    {
+      "epoch": 0.99648,
+      "grad_norm": 0.5790154337882996,
+      "learning_rate": 8.439337454904794e-06,
+      "loss": 0.4445,
+      "step": 519
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.6267328262329102,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.46,
+      "step": 520
+    },
+    {
+      "epoch": 1.00128,
+      "grad_norm": 1.100299596786499,
+      "learning_rate": 8.42306174420744e-06,
+      "loss": 0.7123,
+      "step": 521
+    },
+    {
+      "epoch": 1.0032,
+      "grad_norm": 0.5067075490951538,
+      "learning_rate": 8.414898160284208e-06,
+      "loss": 0.3578,
+      "step": 522
+    },
+    {
+      "epoch": 1.00512,
+      "grad_norm": 0.5103561282157898,
+      "learning_rate": 8.406717478447889e-06,
+      "loss": 0.3232,
+      "step": 523
+    },
+    {
+      "epoch": 1.00704,
+      "grad_norm": 0.5728906989097595,
+      "learning_rate": 8.398519739657997e-06,
+      "loss": 0.4039,
+      "step": 524
+    },
+    {
+      "epoch": 1.00896,
+      "grad_norm": 0.5371456146240234,
+      "learning_rate": 8.390304984959455e-06,
+      "loss": 0.3954,
+      "step": 525
+    },
+    {
+      "epoch": 1.01088,
+      "grad_norm": 0.5132201910018921,
+      "learning_rate": 8.382073255482381e-06,
+      "loss": 0.3458,
+      "step": 526
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 0.5323919057846069,
+      "learning_rate": 8.373824592441884e-06,
+      "loss": 0.3484,
+      "step": 527
+    },
+    {
+      "epoch": 1.01472,
+      "grad_norm": 0.5784719586372375,
+      "learning_rate": 8.36555903713785e-06,
+      "loss": 0.3519,
+      "step": 528
+    },
+    {
+      "epoch": 1.01664,
+      "grad_norm": 0.5782860517501831,
+      "learning_rate": 8.357276630954756e-06,
+      "loss": 0.3919,
+      "step": 529
+    },
+    {
+      "epoch": 1.01856,
+      "grad_norm": 0.5787063837051392,
+      "learning_rate": 8.348977415361435e-06,
+      "loss": 0.3811,
+      "step": 530
+    },
+    {
+      "epoch": 1.02048,
+      "grad_norm": 0.5830230116844177,
+      "learning_rate": 8.340661431910888e-06,
+      "loss": 0.3333,
+      "step": 531
+    },
+    {
+      "epoch": 1.0224,
+      "grad_norm": 0.5619094371795654,
+      "learning_rate": 8.332328722240072e-06,
+      "loss": 0.3468,
+      "step": 532
+    },
+    {
+      "epoch": 1.02432,
+      "grad_norm": 0.6257872581481934,
+      "learning_rate": 8.323979328069689e-06,
+      "loss": 0.4171,
+      "step": 533
+    },
+    {
+      "epoch": 1.02624,
+      "grad_norm": 0.5823887586593628,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.3618,
+      "step": 534
+    },
+    {
+      "epoch": 1.02816,
+      "grad_norm": 0.6039524078369141,
+      "learning_rate": 8.307230653530501e-06,
+      "loss": 0.3584,
+      "step": 535
+    },
+    {
+      "epoch": 1.03008,
+      "grad_norm": 0.5869462490081787,
+      "learning_rate": 8.298831457019943e-06,
+      "loss": 0.3064,
+      "step": 536
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 0.5875105261802673,
+      "learning_rate": 8.290415743725894e-06,
+      "loss": 0.3839,
+      "step": 537
+    },
+    {
+      "epoch": 1.03392,
+      "grad_norm": 0.5947169661521912,
+      "learning_rate": 8.28198355578465e-06,
+      "loss": 0.3511,
+      "step": 538
+    },
+    {
+      "epoch": 1.03584,
+      "grad_norm": 0.5495089888572693,
+      "learning_rate": 8.273534935414975e-06,
+      "loss": 0.3539,
+      "step": 539
+    },
+    {
+      "epoch": 1.03776,
+      "grad_norm": 0.6429109573364258,
+      "learning_rate": 8.265069924917925e-06,
+      "loss": 0.3654,
+      "step": 540
+    },
+    {
+      "epoch": 1.03968,
+      "grad_norm": 0.582603931427002,
+      "learning_rate": 8.256588566676616e-06,
+      "loss": 0.3391,
+      "step": 541
+    },
+    {
+      "epoch": 1.0416,
+      "grad_norm": 0.5939660668373108,
+      "learning_rate": 8.248090903156003e-06,
+      "loss": 0.368,
+      "step": 542
+    },
+    {
+      "epoch": 1.04352,
+      "grad_norm": 0.5589175224304199,
+      "learning_rate": 8.239576976902694e-06,
+      "loss": 0.3614,
+      "step": 543
+    },
+    {
+      "epoch": 1.04544,
+      "grad_norm": 0.5333201885223389,
+      "learning_rate": 8.231046830544716e-06,
+      "loss": 0.3414,
+      "step": 544
+    },
+    {
+      "epoch": 1.04736,
+      "grad_norm": 0.5665721893310547,
+      "learning_rate": 8.222500506791305e-06,
+      "loss": 0.3665,
+      "step": 545
+    },
+    {
+      "epoch": 1.04928,
+      "grad_norm": 0.5572689175605774,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.3258,
+      "step": 546
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 0.5759371519088745,
+      "learning_rate": 8.205359498339917e-06,
+      "loss": 0.3542,
+      "step": 547
+    },
+    {
+      "epoch": 1.05312,
+      "grad_norm": 0.5441709160804749,
+      "learning_rate": 8.196764899464552e-06,
+      "loss": 0.3171,
+      "step": 548
+    },
+    {
+      "epoch": 1.05504,
+      "grad_norm": 0.575310468673706,
+      "learning_rate": 8.188154294838542e-06,
+      "loss": 0.3595,
+      "step": 549
+    },
+    {
+      "epoch": 1.05696,
+      "grad_norm": 0.6133948564529419,
+      "learning_rate": 8.179527727573975e-06,
+      "loss": 0.3707,
+      "step": 550
+    },
+    {
+      "epoch": 1.05888,
+      "grad_norm": 0.5399485230445862,
+      "learning_rate": 8.170885240862854e-06,
+      "loss": 0.3291,
+      "step": 551
+    },
+    {
+      "epoch": 1.0608,
+      "grad_norm": 0.6897387504577637,
+      "learning_rate": 8.162226877976886e-06,
+      "loss": 0.3804,
+      "step": 552
+    },
+    {
+      "epoch": 1.06272,
+      "grad_norm": 0.6381528973579407,
+      "learning_rate": 8.153552682267278e-06,
+      "loss": 0.442,
+      "step": 553
+    },
+    {
+      "epoch": 1.06464,
+      "grad_norm": 0.6141320466995239,
+      "learning_rate": 8.144862697164499e-06,
+      "loss": 0.3776,
+      "step": 554
+    },
+    {
+      "epoch": 1.06656,
+      "grad_norm": 0.5737519860267639,
+      "learning_rate": 8.136156966178082e-06,
+      "loss": 0.3997,
+      "step": 555
+    },
+    {
+      "epoch": 1.06848,
+      "grad_norm": 0.5241888761520386,
+      "learning_rate": 8.127435532896388e-06,
+      "loss": 0.3581,
+      "step": 556
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 0.5723682641983032,
+      "learning_rate": 8.118698440986405e-06,
+      "loss": 0.3668,
+      "step": 557
+    },
+    {
+      "epoch": 1.07232,
+      "grad_norm": 0.6237109899520874,
+      "learning_rate": 8.10994573419352e-06,
+      "loss": 0.3949,
+      "step": 558
+    },
+    {
+      "epoch": 1.07424,
+      "grad_norm": 0.5409807562828064,
+      "learning_rate": 8.101177456341301e-06,
+      "loss": 0.3461,
+      "step": 559
+    },
+    {
+      "epoch": 1.07616,
+      "grad_norm": 0.587260901927948,
+      "learning_rate": 8.092393651331275e-06,
+      "loss": 0.3957,
+      "step": 560
+    },
+    {
+      "epoch": 1.07808,
+      "grad_norm": 0.559928834438324,
+      "learning_rate": 8.083594363142717e-06,
+      "loss": 0.352,
+      "step": 561
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5672323107719421,
+      "learning_rate": 8.074779635832417e-06,
+      "loss": 0.4027,
+      "step": 562
+    },
+    {
+      "epoch": 1.08192,
+      "grad_norm": 0.5665669441223145,
+      "learning_rate": 8.065949513534474e-06,
+      "loss": 0.3713,
+      "step": 563
+    },
+    {
+      "epoch": 1.08384,
+      "grad_norm": 0.5343528985977173,
+      "learning_rate": 8.057104040460062e-06,
+      "loss": 0.3648,
+      "step": 564
+    },
+    {
+      "epoch": 1.08576,
+      "grad_norm": 0.5320156812667847,
+      "learning_rate": 8.048243260897216e-06,
+      "loss": 0.3431,
+      "step": 565
+    },
+    {
+      "epoch": 1.08768,
+      "grad_norm": 0.5923605561256409,
+      "learning_rate": 8.03936721921061e-06,
+      "loss": 0.4191,
+      "step": 566
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 0.505659282207489,
+      "learning_rate": 8.030475959841333e-06,
+      "loss": 0.3386,
+      "step": 567
+    },
+    {
+      "epoch": 1.09152,
+      "grad_norm": 0.5640355944633484,
+      "learning_rate": 8.021569527306663e-06,
+      "loss": 0.4094,
+      "step": 568
+    },
+    {
+      "epoch": 1.09344,
+      "grad_norm": 0.5868680477142334,
+      "learning_rate": 8.012647966199851e-06,
+      "loss": 0.3909,
+      "step": 569
+    },
+    {
+      "epoch": 1.09536,
+      "grad_norm": 0.5564755201339722,
+      "learning_rate": 8.003711321189895e-06,
+      "loss": 0.3617,
+      "step": 570
+    },
+    {
+      "epoch": 1.09728,
+      "grad_norm": 0.5832654237747192,
+      "learning_rate": 7.994759637021316e-06,
+      "loss": 0.3651,
+      "step": 571
+    },
+    {
+      "epoch": 1.0992,
+      "grad_norm": 0.5585725903511047,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 0.3528,
+      "step": 572
+    },
+    {
+      "epoch": 1.10112,
+      "grad_norm": 0.48755979537963867,
+      "learning_rate": 7.976811330562637e-06,
+      "loss": 0.3179,
+      "step": 573
+    },
+    {
+      "epoch": 1.10304,
+      "grad_norm": 0.5658108592033386,
+      "learning_rate": 7.967814798137173e-06,
+      "loss": 0.3782,
+      "step": 574
+    },
+    {
+      "epoch": 1.10496,
+      "grad_norm": 0.5519687533378601,
+      "learning_rate": 7.95880340628191e-06,
+      "loss": 0.3447,
+      "step": 575
+    },
+    {
+      "epoch": 1.10688,
+      "grad_norm": 0.5374343395233154,
+      "learning_rate": 7.949777200115617e-06,
+      "loss": 0.3618,
+      "step": 576
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 0.5478836297988892,
+      "learning_rate": 7.940736224831227e-06,
+      "loss": 0.3714,
+      "step": 577
+    },
+    {
+      "epoch": 1.11072,
+      "grad_norm": 0.5985942482948303,
+      "learning_rate": 7.931680525695634e-06,
+      "loss": 0.4108,
+      "step": 578
+    },
+    {
+      "epoch": 1.11264,
+      "grad_norm": 0.5715672969818115,
+      "learning_rate": 7.922610148049445e-06,
+      "loss": 0.4026,
+      "step": 579
+    },
+    {
+      "epoch": 1.11456,
+      "grad_norm": 0.5735968947410583,
+      "learning_rate": 7.913525137306756e-06,
+      "loss": 0.3546,
+      "step": 580
+    },
+    {
+      "epoch": 1.11648,
+      "grad_norm": 0.5736159682273865,
+      "learning_rate": 7.90442553895494e-06,
+      "loss": 0.3976,
+      "step": 581
+    },
+    {
+      "epoch": 1.1184,
+      "grad_norm": 0.5330926775932312,
+      "learning_rate": 7.895311398554395e-06,
+      "loss": 0.3447,
+      "step": 582
+    },
+    {
+      "epoch": 1.12032,
+      "grad_norm": 0.6086499691009521,
+      "learning_rate": 7.886182761738339e-06,
+      "loss": 0.4126,
+      "step": 583
+    },
+    {
+      "epoch": 1.12224,
+      "grad_norm": 0.5583930015563965,
+      "learning_rate": 7.877039674212569e-06,
+      "loss": 0.3635,
+      "step": 584
+    },
+    {
+      "epoch": 1.12416,
+      "grad_norm": 0.568538248538971,
+      "learning_rate": 7.86788218175523e-06,
+      "loss": 0.3838,
+      "step": 585
+    },
+    {
+      "epoch": 1.12608,
+      "grad_norm": 0.5865702629089355,
+      "learning_rate": 7.8587103302166e-06,
+      "loss": 0.394,
+      "step": 586
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.5577906966209412,
+      "learning_rate": 7.849524165518838e-06,
+      "loss": 0.3968,
+      "step": 587
+    },
+    {
+      "epoch": 1.12992,
+      "grad_norm": 0.5969308614730835,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.3869,
+      "step": 588
+    },
+    {
+      "epoch": 1.13184,
+      "grad_norm": 0.6029490828514099,
+      "learning_rate": 7.831109080692687e-06,
+      "loss": 0.3896,
+      "step": 589
+    },
+    {
+      "epoch": 1.13376,
+      "grad_norm": 0.5661328434944153,
+      "learning_rate": 7.821880252766025e-06,
+      "loss": 0.3558,
+      "step": 590
+    },
+    {
+      "epoch": 1.13568,
+      "grad_norm": 0.5487807989120483,
+      "learning_rate": 7.812637296083235e-06,
+      "loss": 0.3831,
+      "step": 591
+    },
+    {
+      "epoch": 1.1376,
+      "grad_norm": 0.5510612726211548,
+      "learning_rate": 7.803380256922495e-06,
+      "loss": 0.3423,
+      "step": 592
+    },
+    {
+      "epoch": 1.13952,
+      "grad_norm": 0.59043288230896,
+      "learning_rate": 7.794109181632494e-06,
+      "loss": 0.3627,
+      "step": 593
+    },
+    {
+      "epoch": 1.14144,
+      "grad_norm": 0.5743687748908997,
+      "learning_rate": 7.784824116632198e-06,
+      "loss": 0.3273,
+      "step": 594
+    },
+    {
+      "epoch": 1.14336,
+      "grad_norm": 0.5270614624023438,
+      "learning_rate": 7.775525108410616e-06,
+      "loss": 0.3335,
+      "step": 595
+    },
+    {
+      "epoch": 1.14528,
+      "grad_norm": 0.5700528621673584,
+      "learning_rate": 7.76621220352657e-06,
+      "loss": 0.3845,
+      "step": 596
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 0.5575633645057678,
+      "learning_rate": 7.75688544860846e-06,
+      "loss": 0.361,
+      "step": 597
+    },
+    {
+      "epoch": 1.14912,
+      "grad_norm": 0.5090420842170715,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 0.3679,
+      "step": 598
+    },
+    {
+      "epoch": 1.15104,
+      "grad_norm": 0.5497364401817322,
+      "learning_rate": 7.73819057553014e-06,
+      "loss": 0.3787,
+      "step": 599
+    },
+    {
+      "epoch": 1.15296,
+      "grad_norm": 0.5635077357292175,
+      "learning_rate": 7.728822550972523e-06,
+      "loss": 0.3791,
+      "step": 600
+    },
+    {
+      "epoch": 1.15488,
+      "grad_norm": 0.5452158451080322,
+      "learning_rate": 7.719440863585555e-06,
+      "loss": 0.3957,
+      "step": 601
+    },
+    {
+      "epoch": 1.1568,
+      "grad_norm": 0.5858789086341858,
+      "learning_rate": 7.710045560342021e-06,
+      "loss": 0.4012,
+      "step": 602
+    },
+    {
+      "epoch": 1.15872,
+      "grad_norm": 0.5860247611999512,
+      "learning_rate": 7.700636688282878e-06,
+      "loss": 0.3562,
+      "step": 603
+    },
+    {
+      "epoch": 1.16064,
+      "grad_norm": 0.5257615447044373,
+      "learning_rate": 7.69121429451702e-06,
+      "loss": 0.3525,
+      "step": 604
+    },
+    {
+      "epoch": 1.16256,
+      "grad_norm": 0.5720390677452087,
+      "learning_rate": 7.681778426221043e-06,
+      "loss": 0.3726,
+      "step": 605
+    },
+    {
+      "epoch": 1.16448,
+      "grad_norm": 0.5522862076759338,
+      "learning_rate": 7.672329130639007e-06,
+      "loss": 0.3681,
+      "step": 606
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 0.6177027821540833,
+      "learning_rate": 7.662866455082201e-06,
+      "loss": 0.4221,
+      "step": 607
+    },
+    {
+      "epoch": 1.16832,
+      "grad_norm": 0.5884987711906433,
+      "learning_rate": 7.65339044692891e-06,
+      "loss": 0.4061,
+      "step": 608
+    },
+    {
+      "epoch": 1.17024,
+      "grad_norm": 0.6003137230873108,
+      "learning_rate": 7.643901153624163e-06,
+      "loss": 0.4125,
+      "step": 609
+    },
+    {
+      "epoch": 1.17216,
+      "grad_norm": 0.5626486539840698,
+      "learning_rate": 7.634398622679517e-06,
+      "loss": 0.361,
+      "step": 610
+    },
+    {
+      "epoch": 1.17408,
+      "grad_norm": 0.5053431987762451,
+      "learning_rate": 7.624882901672801e-06,
+      "loss": 0.3134,
+      "step": 611
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 0.5010710954666138,
+      "learning_rate": 7.615354038247889e-06,
+      "loss": 0.281,
+      "step": 612
+    },
+    {
+      "epoch": 1.17792,
+      "grad_norm": 0.5858680009841919,
+      "learning_rate": 7.605812080114453e-06,
+      "loss": 0.3755,
+      "step": 613
+    },
+    {
+      "epoch": 1.17984,
+      "grad_norm": 0.5540910363197327,
+      "learning_rate": 7.59625707504773e-06,
+      "loss": 0.3475,
+      "step": 614
+    },
+    {
+      "epoch": 1.18176,
+      "grad_norm": 0.5937021374702454,
+      "learning_rate": 7.586689070888284e-06,
+      "loss": 0.4207,
+      "step": 615
+    },
+    {
+      "epoch": 1.18368,
+      "grad_norm": 0.6152872443199158,
+      "learning_rate": 7.577108115541761e-06,
+      "loss": 0.3953,
+      "step": 616
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 0.5496403574943542,
+      "learning_rate": 7.567514256978652e-06,
+      "loss": 0.3432,
+      "step": 617
+    },
+    {
+      "epoch": 1.18752,
+      "grad_norm": 0.5806377530097961,
+      "learning_rate": 7.557907543234051e-06,
+      "loss": 0.3835,
+      "step": 618
+    },
+    {
+      "epoch": 1.18944,
+      "grad_norm": 0.5831543207168579,
+      "learning_rate": 7.54828802240742e-06,
+      "loss": 0.3702,
+      "step": 619
+    },
+    {
+      "epoch": 1.19136,
+      "grad_norm": 0.6438750624656677,
+      "learning_rate": 7.53865574266234e-06,
+      "loss": 0.385,
+      "step": 620
+    },
+    {
+      "epoch": 1.1932800000000001,
+      "grad_norm": 0.5649431943893433,
+      "learning_rate": 7.529010752226277e-06,
+      "loss": 0.3866,
+      "step": 621
+    },
+    {
+      "epoch": 1.1952,
+      "grad_norm": 0.6866464018821716,
+      "learning_rate": 7.519353099390336e-06,
+      "loss": 0.4149,
+      "step": 622
+    },
+    {
+      "epoch": 1.19712,
+      "grad_norm": 0.5543093085289001,
+      "learning_rate": 7.509682832509024e-06,
+      "loss": 0.3563,
+      "step": 623
+    },
+    {
+      "epoch": 1.19904,
+      "grad_norm": 0.542264461517334,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3474,
+      "step": 624
+    },
+    {
+      "epoch": 1.20096,
+      "grad_norm": 0.5008151531219482,
+      "learning_rate": 7.490304650343842e-06,
+      "loss": 0.3108,
+      "step": 625
+    },
+    {
+      "epoch": 1.20288,
+      "grad_norm": 0.6101180911064148,
+      "learning_rate": 7.480596832083795e-06,
+      "loss": 0.402,
+      "step": 626
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 0.5587049126625061,
+      "learning_rate": 7.470876593825535e-06,
+      "loss": 0.3484,
+      "step": 627
+    },
+    {
+      "epoch": 1.20672,
+      "grad_norm": 0.5417701601982117,
+      "learning_rate": 7.461143984236925e-06,
+      "loss": 0.3689,
+      "step": 628
+    },
+    {
+      "epoch": 1.20864,
+      "grad_norm": 0.5674274563789368,
+      "learning_rate": 7.451399052047764e-06,
+      "loss": 0.3366,
+      "step": 629
+    },
+    {
+      "epoch": 1.21056,
+      "grad_norm": 0.5525009632110596,
+      "learning_rate": 7.441641846049557e-06,
+      "loss": 0.3299,
+      "step": 630
+    },
+    {
+      "epoch": 1.21248,
+      "grad_norm": 0.5465267300605774,
+      "learning_rate": 7.431872415095252e-06,
+      "loss": 0.363,
+      "step": 631
+    },
+    {
+      "epoch": 1.2144,
+      "grad_norm": 0.5430969595909119,
+      "learning_rate": 7.422090808099014e-06,
+      "loss": 0.3549,
+      "step": 632
+    },
+    {
+      "epoch": 1.21632,
+      "grad_norm": 0.5617967844009399,
+      "learning_rate": 7.412297074035968e-06,
+      "loss": 0.3871,
+      "step": 633
+    },
+    {
+      "epoch": 1.21824,
+      "grad_norm": 0.535500705242157,
+      "learning_rate": 7.402491261941958e-06,
+      "loss": 0.37,
+      "step": 634
+    },
+    {
+      "epoch": 1.22016,
+      "grad_norm": 0.5272918939590454,
+      "learning_rate": 7.3926734209133e-06,
+      "loss": 0.3438,
+      "step": 635
+    },
+    {
+      "epoch": 1.22208,
+      "grad_norm": 0.559246838092804,
+      "learning_rate": 7.382843600106539e-06,
+      "loss": 0.3956,
+      "step": 636
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 0.5945393443107605,
+      "learning_rate": 7.373001848738203e-06,
+      "loss": 0.4068,
+      "step": 637
+    },
+    {
+      "epoch": 1.22592,
+      "grad_norm": 0.5096088647842407,
+      "learning_rate": 7.363148216084548e-06,
+      "loss": 0.3575,
+      "step": 638
+    },
+    {
+      "epoch": 1.22784,
+      "grad_norm": 0.615565299987793,
+      "learning_rate": 7.353282751481324e-06,
+      "loss": 0.4184,
+      "step": 639
+    },
+    {
+      "epoch": 1.22976,
+      "grad_norm": 0.5571386218070984,
+      "learning_rate": 7.343405504323519e-06,
+      "loss": 0.3495,
+      "step": 640
+    },
+    {
+      "epoch": 1.2316799999999999,
+      "grad_norm": 0.5487669110298157,
+      "learning_rate": 7.333516524065116e-06,
+      "loss": 0.4232,
+      "step": 641
+    },
+    {
+      "epoch": 1.2336,
+      "grad_norm": 0.5621005892753601,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.3406,
+      "step": 642
+    },
+    {
+      "epoch": 1.23552,
+      "grad_norm": 0.5904276371002197,
+      "learning_rate": 7.3137035623559274e-06,
+      "loss": 0.3672,
+      "step": 643
+    },
+    {
+      "epoch": 1.23744,
+      "grad_norm": 0.5038149356842041,
+      "learning_rate": 7.303779680105844e-06,
+      "loss": 0.3419,
+      "step": 644
+    },
+    {
+      "epoch": 1.23936,
+      "grad_norm": 0.5708621740341187,
+      "learning_rate": 7.2938442631560714e-06,
+      "loss": 0.3757,
+      "step": 645
+    },
+    {
+      "epoch": 1.24128,
+      "grad_norm": 0.5482596158981323,
+      "learning_rate": 7.28389736125184e-06,
+      "loss": 0.3537,
+      "step": 646
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 0.5996248126029968,
+      "learning_rate": 7.273939024195881e-06,
+      "loss": 0.3547,
+      "step": 647
+    },
+    {
+      "epoch": 1.24512,
+      "grad_norm": 0.6012654900550842,
+      "learning_rate": 7.263969301848188e-06,
+      "loss": 0.398,
+      "step": 648
+    },
+    {
+      "epoch": 1.24704,
+      "grad_norm": 0.5477873682975769,
+      "learning_rate": 7.2539882441257484e-06,
+      "loss": 0.3666,
+      "step": 649
+    },
+    {
+      "epoch": 1.24896,
+      "grad_norm": 0.5989547371864319,
+      "learning_rate": 7.243995901002312e-06,
+      "loss": 0.3786,
+      "step": 650
+    },
+    {
+      "epoch": 1.25088,
+      "grad_norm": 0.6050532460212708,
+      "learning_rate": 7.2339923225081296e-06,
+      "loss": 0.3789,
+      "step": 651
+    },
+    {
+      "epoch": 1.2528000000000001,
+      "grad_norm": 0.5515075922012329,
+      "learning_rate": 7.223977558729707e-06,
+      "loss": 0.3695,
+      "step": 652
+    },
+    {
+      "epoch": 1.25472,
+      "grad_norm": 0.4945664405822754,
+      "learning_rate": 7.213951659809553e-06,
+      "loss": 0.3496,
+      "step": 653
+    },
+    {
+      "epoch": 1.25664,
+      "grad_norm": 0.5737729668617249,
+      "learning_rate": 7.203914675945929e-06,
+      "loss": 0.3806,
+      "step": 654
+    },
+    {
+      "epoch": 1.2585600000000001,
+      "grad_norm": 0.5845323801040649,
+      "learning_rate": 7.193866657392597e-06,
+      "loss": 0.3841,
+      "step": 655
+    },
+    {
+      "epoch": 1.26048,
+      "grad_norm": 0.5330585837364197,
+      "learning_rate": 7.183807654458565e-06,
+      "loss": 0.3508,
+      "step": 656
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 0.5641207695007324,
+      "learning_rate": 7.173737717507843e-06,
+      "loss": 0.3637,
+      "step": 657
+    },
+    {
+      "epoch": 1.26432,
+      "grad_norm": 0.557246744632721,
+      "learning_rate": 7.163656896959181e-06,
+      "loss": 0.384,
+      "step": 658
+    },
+    {
+      "epoch": 1.26624,
+      "grad_norm": 0.5714312791824341,
+      "learning_rate": 7.153565243285829e-06,
+      "loss": 0.3677,
+      "step": 659
+    },
+    {
+      "epoch": 1.26816,
+      "grad_norm": 0.5794839262962341,
+      "learning_rate": 7.143462807015271e-06,
+      "loss": 0.3867,
+      "step": 660
+    },
+    {
+      "epoch": 1.27008,
+      "grad_norm": 0.5434252619743347,
+      "learning_rate": 7.133349638728979e-06,
+      "loss": 0.342,
+      "step": 661
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 0.5557631850242615,
+      "learning_rate": 7.1232257890621605e-06,
+      "loss": 0.3937,
+      "step": 662
+    },
+    {
+      "epoch": 1.27392,
+      "grad_norm": 0.5878427028656006,
+      "learning_rate": 7.113091308703498e-06,
+      "loss": 0.4081,
+      "step": 663
+    },
+    {
+      "epoch": 1.27584,
+      "grad_norm": 0.5944617390632629,
+      "learning_rate": 7.102946248394908e-06,
+      "loss": 0.3764,
+      "step": 664
+    },
+    {
+      "epoch": 1.27776,
+      "grad_norm": 0.5725439190864563,
+      "learning_rate": 7.092790658931274e-06,
+      "loss": 0.365,
+      "step": 665
+    },
+    {
+      "epoch": 1.27968,
+      "grad_norm": 0.5907331705093384,
+      "learning_rate": 7.082624591160201e-06,
+      "loss": 0.3435,
+      "step": 666
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 0.5718739628791809,
+      "learning_rate": 7.072448095981751e-06,
+      "loss": 0.3474,
+      "step": 667
+    },
+    {
+      "epoch": 1.28352,
+      "grad_norm": 0.5484753251075745,
+      "learning_rate": 7.0622612243482035e-06,
+      "loss": 0.3681,
+      "step": 668
+    },
+    {
+      "epoch": 1.28544,
+      "grad_norm": 0.5785349607467651,
+      "learning_rate": 7.052064027263785e-06,
+      "loss": 0.3605,
+      "step": 669
+    },
+    {
+      "epoch": 1.28736,
+      "grad_norm": 0.5944696664810181,
+      "learning_rate": 7.041856555784421e-06,
+      "loss": 0.3779,
+      "step": 670
+    },
+    {
+      "epoch": 1.28928,
+      "grad_norm": 0.5899400115013123,
+      "learning_rate": 7.031638861017484e-06,
+      "loss": 0.4245,
+      "step": 671
+    },
+    {
+      "epoch": 1.2912,
+      "grad_norm": 0.5744240283966064,
+      "learning_rate": 7.021410994121525e-06,
+      "loss": 0.3846,
+      "step": 672
+    },
+    {
+      "epoch": 1.29312,
+      "grad_norm": 0.5999711751937866,
+      "learning_rate": 7.011173006306034e-06,
+      "loss": 0.3634,
+      "step": 673
+    },
+    {
+      "epoch": 1.29504,
+      "grad_norm": 0.5760245323181152,
+      "learning_rate": 7.0009249488311685e-06,
+      "loss": 0.3873,
+      "step": 674
+    },
+    {
+      "epoch": 1.29696,
+      "grad_norm": 0.6277198195457458,
+      "learning_rate": 6.990666873007506e-06,
+      "loss": 0.4289,
+      "step": 675
+    },
+    {
+      "epoch": 1.29888,
+      "grad_norm": 0.6124325394630432,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.4424,
+      "step": 676
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 0.5946184396743774,
+      "learning_rate": 6.970120871806647e-06,
+      "loss": 0.3763,
+      "step": 677
+    },
+    {
+      "epoch": 1.3027199999999999,
+      "grad_norm": 0.5659326314926147,
+      "learning_rate": 6.959833049300376e-06,
+      "loss": 0.3468,
+      "step": 678
+    },
+    {
+      "epoch": 1.30464,
+      "grad_norm": 0.5912370085716248,
+      "learning_rate": 6.949535414186651e-06,
+      "loss": 0.3897,
+      "step": 679
+    },
+    {
+      "epoch": 1.30656,
+      "grad_norm": 0.6038317680358887,
+      "learning_rate": 6.939228018024275e-06,
+      "loss": 0.4312,
+      "step": 680
+    },
+    {
+      "epoch": 1.3084799999999999,
+      "grad_norm": 0.5715245008468628,
+      "learning_rate": 6.9289109124209276e-06,
+      "loss": 0.3671,
+      "step": 681
+    },
+    {
+      "epoch": 1.3104,
+      "grad_norm": 0.5397374033927917,
+      "learning_rate": 6.918584149032899e-06,
+      "loss": 0.326,
+      "step": 682
+    },
+    {
+      "epoch": 1.31232,
+      "grad_norm": 0.5480877757072449,
+      "learning_rate": 6.908247779564836e-06,
+      "loss": 0.3771,
+      "step": 683
+    },
+    {
+      "epoch": 1.31424,
+      "grad_norm": 0.5419209599494934,
+      "learning_rate": 6.897901855769483e-06,
+      "loss": 0.3335,
+      "step": 684
+    },
+    {
+      "epoch": 1.31616,
+      "grad_norm": 0.6512284874916077,
+      "learning_rate": 6.887546429447418e-06,
+      "loss": 0.4097,
+      "step": 685
+    },
+    {
+      "epoch": 1.31808,
+      "grad_norm": 0.6101759672164917,
+      "learning_rate": 6.8771815524468e-06,
+      "loss": 0.3515,
+      "step": 686
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5833544731140137,
+      "learning_rate": 6.8668072766631054e-06,
+      "loss": 0.3863,
+      "step": 687
+    },
+    {
+      "epoch": 1.32192,
+      "grad_norm": 0.5798306465148926,
+      "learning_rate": 6.856423654038868e-06,
+      "loss": 0.3714,
+      "step": 688
+    },
+    {
+      "epoch": 1.3238400000000001,
+      "grad_norm": 0.6920048594474792,
+      "learning_rate": 6.8460307365634225e-06,
+      "loss": 0.3858,
+      "step": 689
+    },
+    {
+      "epoch": 1.32576,
+      "grad_norm": 0.6107429265975952,
+      "learning_rate": 6.8356285762726385e-06,
+      "loss": 0.3391,
+      "step": 690
+    },
+    {
+      "epoch": 1.32768,
+      "grad_norm": 0.5508682727813721,
+      "learning_rate": 6.825217225248664e-06,
+      "loss": 0.3467,
+      "step": 691
+    },
+    {
+      "epoch": 1.3296000000000001,
+      "grad_norm": 0.5802949070930481,
+      "learning_rate": 6.814796735619664e-06,
+      "loss": 0.3901,
+      "step": 692
+    },
+    {
+      "epoch": 1.33152,
+      "grad_norm": 0.6609660387039185,
+      "learning_rate": 6.804367159559561e-06,
+      "loss": 0.3696,
+      "step": 693
+    },
+    {
+      "epoch": 1.33344,
+      "grad_norm": 0.7042303085327148,
+      "learning_rate": 6.793928549287767e-06,
+      "loss": 0.4217,
+      "step": 694
+    },
+    {
+      "epoch": 1.33536,
+      "grad_norm": 0.5676562786102295,
+      "learning_rate": 6.783480957068934e-06,
+      "loss": 0.4187,
+      "step": 695
+    },
+    {
+      "epoch": 1.33728,
+      "grad_norm": 0.5554609894752502,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.3769,
+      "step": 696
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 0.6118761897087097,
+      "learning_rate": 6.762559036073331e-06,
+      "loss": 0.3933,
+      "step": 697
+    },
+    {
+      "epoch": 1.34112,
+      "grad_norm": 0.547964334487915,
+      "learning_rate": 6.75208481204967e-06,
+      "loss": 0.3214,
+      "step": 698
+    },
+    {
+      "epoch": 1.34304,
+      "grad_norm": 0.5246742367744446,
+      "learning_rate": 6.74160181558465e-06,
+      "loss": 0.3149,
+      "step": 699
+    },
+    {
+      "epoch": 1.34496,
+      "grad_norm": 0.5433777570724487,
+      "learning_rate": 6.731110099165165e-06,
+      "loss": 0.3526,
+      "step": 700
+    },
+    {
+      "epoch": 1.34688,
+      "grad_norm": 0.5239468812942505,
+      "learning_rate": 6.7206097153217474e-06,
+      "loss": 0.3283,
+      "step": 701
+    },
+    {
+      "epoch": 1.3488,
+      "grad_norm": 0.5656136870384216,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.3355,
+      "step": 702
+    },
+    {
+      "epoch": 1.35072,
+      "grad_norm": 0.5359403491020203,
+      "learning_rate": 6.699583155702025e-06,
+      "loss": 0.3637,
+      "step": 703
+    },
+    {
+      "epoch": 1.35264,
+      "grad_norm": 0.5552080273628235,
+      "learning_rate": 6.689057085202737e-06,
+      "loss": 0.3438,
+      "step": 704
+    },
+    {
+      "epoch": 1.35456,
+      "grad_norm": 0.519082248210907,
+      "learning_rate": 6.678522557833025e-06,
+      "loss": 0.3209,
+      "step": 705
+    },
+    {
+      "epoch": 1.35648,
+      "grad_norm": 0.5399109721183777,
+      "learning_rate": 6.667979626337783e-06,
+      "loss": 0.3586,
+      "step": 706
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 0.5169100761413574,
+      "learning_rate": 6.6574283435039795e-06,
+      "loss": 0.3398,
+      "step": 707
+    },
+    {
+      "epoch": 1.36032,
+      "grad_norm": 0.5389205813407898,
+      "learning_rate": 6.646868762160399e-06,
+      "loss": 0.3483,
+      "step": 708
+    },
+    {
+      "epoch": 1.36224,
+      "grad_norm": 0.5997045040130615,
+      "learning_rate": 6.6363009351773755e-06,
+      "loss": 0.3602,
+      "step": 709
+    },
+    {
+      "epoch": 1.36416,
+      "grad_norm": 0.561926543712616,
+      "learning_rate": 6.625724915466526e-06,
+      "loss": 0.3739,
+      "step": 710
+    },
+    {
+      "epoch": 1.36608,
+      "grad_norm": 0.5001537799835205,
+      "learning_rate": 6.615140755980491e-06,
+      "loss": 0.331,
+      "step": 711
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 0.5765603184700012,
+      "learning_rate": 6.6045485097126585e-06,
+      "loss": 0.3753,
+      "step": 712
+    },
+    {
+      "epoch": 1.36992,
+      "grad_norm": 0.5498204231262207,
+      "learning_rate": 6.593948229696915e-06,
+      "loss": 0.4058,
+      "step": 713
+    },
+    {
+      "epoch": 1.37184,
+      "grad_norm": 0.5711533427238464,
+      "learning_rate": 6.583339969007364e-06,
+      "loss": 0.3865,
+      "step": 714
+    },
+    {
+      "epoch": 1.3737599999999999,
+      "grad_norm": 0.5368412733078003,
+      "learning_rate": 6.572723780758069e-06,
+      "loss": 0.3475,
+      "step": 715
+    },
+    {
+      "epoch": 1.37568,
+      "grad_norm": 0.5234925746917725,
+      "learning_rate": 6.562099718102788e-06,
+      "loss": 0.3298,
+      "step": 716
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 0.5610650777816772,
+      "learning_rate": 6.551467834234702e-06,
+      "loss": 0.3555,
+      "step": 717
+    },
+    {
+      "epoch": 1.37952,
+      "grad_norm": 0.5173355937004089,
+      "learning_rate": 6.540828182386154e-06,
+      "loss": 0.322,
+      "step": 718
+    },
+    {
+      "epoch": 1.38144,
+      "grad_norm": 0.5313178896903992,
+      "learning_rate": 6.530180815828377e-06,
+      "loss": 0.3659,
+      "step": 719
+    },
+    {
+      "epoch": 1.38336,
+      "grad_norm": 0.5848520994186401,
+      "learning_rate": 6.519525787871235e-06,
+      "loss": 0.3862,
+      "step": 720
+    },
+    {
+      "epoch": 1.38528,
+      "grad_norm": 0.5176764130592346,
+      "learning_rate": 6.508863151862947e-06,
+      "loss": 0.353,
+      "step": 721
+    },
+    {
+      "epoch": 1.3872,
+      "grad_norm": 0.6958459615707397,
+      "learning_rate": 6.498192961189829e-06,
+      "loss": 0.4059,
+      "step": 722
+    },
+    {
+      "epoch": 1.3891200000000001,
+      "grad_norm": 0.5808752179145813,
+      "learning_rate": 6.487515269276015e-06,
+      "loss": 0.3804,
+      "step": 723
+    },
+    {
+      "epoch": 1.39104,
+      "grad_norm": 0.5837891101837158,
+      "learning_rate": 6.476830129583207e-06,
+      "loss": 0.3845,
+      "step": 724
+    },
+    {
+      "epoch": 1.39296,
+      "grad_norm": 0.5993189811706543,
+      "learning_rate": 6.466137595610388e-06,
+      "loss": 0.3709,
+      "step": 725
+    },
+    {
+      "epoch": 1.3948800000000001,
+      "grad_norm": 0.6214917898178101,
+      "learning_rate": 6.455437720893565e-06,
+      "loss": 0.3397,
+      "step": 726
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 0.5897639393806458,
+      "learning_rate": 6.444730559005504e-06,
+      "loss": 0.3434,
+      "step": 727
+    },
+    {
+      "epoch": 1.39872,
+      "grad_norm": 0.6529344320297241,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.4053,
+      "step": 728
+    },
+    {
+      "epoch": 1.40064,
+      "grad_norm": 0.5655744075775146,
+      "learning_rate": 6.423294588188874e-06,
+      "loss": 0.362,
+      "step": 729
+    },
+    {
+      "epoch": 1.40256,
+      "grad_norm": 0.6439656019210815,
+      "learning_rate": 6.412565886587186e-06,
+      "loss": 0.3605,
+      "step": 730
+    },
+    {
+      "epoch": 1.40448,
+      "grad_norm": 0.6847288012504578,
+      "learning_rate": 6.401830112467479e-06,
+      "loss": 0.4112,
+      "step": 731
+    },
+    {
+      "epoch": 1.4064,
+      "grad_norm": 0.6503743529319763,
+      "learning_rate": 6.391087319582264e-06,
+      "loss": 0.3858,
+      "step": 732
+    },
+    {
+      "epoch": 1.40832,
+      "grad_norm": 0.5181986093521118,
+      "learning_rate": 6.380337561719184e-06,
+      "loss": 0.3455,
+      "step": 733
+    },
+    {
+      "epoch": 1.41024,
+      "grad_norm": 0.5788087844848633,
+      "learning_rate": 6.369580892700763e-06,
+      "loss": 0.3647,
+      "step": 734
+    },
+    {
+      "epoch": 1.41216,
+      "grad_norm": 0.5603296160697937,
+      "learning_rate": 6.358817366384122e-06,
+      "loss": 0.363,
+      "step": 735
+    },
+    {
+      "epoch": 1.41408,
+      "grad_norm": 0.6358045339584351,
+      "learning_rate": 6.34804703666072e-06,
+      "loss": 0.3769,
+      "step": 736
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 0.5851025581359863,
+      "learning_rate": 6.337269957456074e-06,
+      "loss": 0.3593,
+      "step": 737
+    },
+    {
+      "epoch": 1.41792,
+      "grad_norm": 0.569039523601532,
+      "learning_rate": 6.326486182729504e-06,
+      "loss": 0.404,
+      "step": 738
+    },
+    {
+      "epoch": 1.41984,
+      "grad_norm": 0.5932394862174988,
+      "learning_rate": 6.315695766473844e-06,
+      "loss": 0.3885,
+      "step": 739
+    },
+    {
+      "epoch": 1.42176,
+      "grad_norm": 0.6062045097351074,
+      "learning_rate": 6.304898762715187e-06,
+      "loss": 0.3868,
+      "step": 740
+    },
+    {
+      "epoch": 1.42368,
+      "grad_norm": 0.5724964737892151,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.3931,
+      "step": 741
+    },
+    {
+      "epoch": 1.4256,
+      "grad_norm": 0.5028746128082275,
+      "learning_rate": 6.283285208957884e-06,
+      "loss": 0.3298,
+      "step": 742
+    },
+    {
+      "epoch": 1.42752,
+      "grad_norm": 0.5761634707450867,
+      "learning_rate": 6.272468767175255e-06,
+      "loss": 0.3967,
+      "step": 743
+    },
+    {
+      "epoch": 1.42944,
+      "grad_norm": 0.5469388365745544,
+      "learning_rate": 6.261645954321109e-06,
+      "loss": 0.3673,
+      "step": 744
+    },
+    {
+      "epoch": 1.43136,
+      "grad_norm": 0.5997186899185181,
+      "learning_rate": 6.2508168245837476e-06,
+      "loss": 0.4111,
+      "step": 745
+    },
+    {
+      "epoch": 1.4332799999999999,
+      "grad_norm": 0.5492907762527466,
+      "learning_rate": 6.239981432183093e-06,
+      "loss": 0.3734,
+      "step": 746
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 0.50548255443573,
+      "learning_rate": 6.2291398313704265e-06,
+      "loss": 0.3356,
+      "step": 747
+    },
+    {
+      "epoch": 1.43712,
+      "grad_norm": 0.5548710823059082,
+      "learning_rate": 6.21829207642811e-06,
+      "loss": 0.3419,
+      "step": 748
+    },
+    {
+      "epoch": 1.4390399999999999,
+      "grad_norm": 0.5736958384513855,
+      "learning_rate": 6.207438221669325e-06,
+      "loss": 0.3574,
+      "step": 749
+    },
+    {
+      "epoch": 1.44096,
+      "grad_norm": 0.5194618701934814,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.3582,
+      "step": 750
+    },
+    {
+      "epoch": 1.44288,
+      "grad_norm": 0.5740842223167419,
+      "learning_rate": 6.185712430107489e-06,
+      "loss": 0.3645,
+      "step": 751
+    },
+    {
+      "epoch": 1.4447999999999999,
+      "grad_norm": 0.5583177804946899,
+      "learning_rate": 6.1748406020824115e-06,
+      "loss": 0.35,
+      "step": 752
+    },
+    {
+      "epoch": 1.44672,
+      "grad_norm": 0.5404021739959717,
+      "learning_rate": 6.163962891796261e-06,
+      "loss": 0.3504,
+      "step": 753
+    },
+    {
+      "epoch": 1.44864,
+      "grad_norm": 0.5572770833969116,
+      "learning_rate": 6.153079353712201e-06,
+      "loss": 0.3594,
+      "step": 754
+    },
+    {
+      "epoch": 1.45056,
+      "grad_norm": 0.5232715010643005,
+      "learning_rate": 6.142190042322569e-06,
+      "loss": 0.3741,
+      "step": 755
+    },
+    {
+      "epoch": 1.45248,
+      "grad_norm": 0.5563902258872986,
+      "learning_rate": 6.131295012148613e-06,
+      "loss": 0.4005,
+      "step": 756
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 0.5317257642745972,
+      "learning_rate": 6.120394317740205e-06,
+      "loss": 0.3891,
+      "step": 757
+    },
+    {
+      "epoch": 1.45632,
+      "grad_norm": 0.5544328093528748,
+      "learning_rate": 6.1094880136755886e-06,
+      "loss": 0.3583,
+      "step": 758
+    },
+    {
+      "epoch": 1.45824,
+      "grad_norm": 0.5211480259895325,
+      "learning_rate": 6.0985761545610865e-06,
+      "loss": 0.3366,
+      "step": 759
+    },
+    {
+      "epoch": 1.4601600000000001,
+      "grad_norm": 0.5635758638381958,
+      "learning_rate": 6.087658795030838e-06,
+      "loss": 0.392,
+      "step": 760
+    },
+    {
+      "epoch": 1.46208,
+      "grad_norm": 0.5777500867843628,
+      "learning_rate": 6.076735989746522e-06,
+      "loss": 0.3818,
+      "step": 761
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 0.5805518627166748,
+      "learning_rate": 6.065807793397081e-06,
+      "loss": 0.3874,
+      "step": 762
+    },
+    {
+      "epoch": 1.4659200000000001,
+      "grad_norm": 0.5196983814239502,
+      "learning_rate": 6.0548742606984545e-06,
+      "loss": 0.3289,
+      "step": 763
+    },
+    {
+      "epoch": 1.46784,
+      "grad_norm": 0.5974670052528381,
+      "learning_rate": 6.043935446393294e-06,
+      "loss": 0.3879,
+      "step": 764
+    },
+    {
+      "epoch": 1.46976,
+      "grad_norm": 0.5634557604789734,
+      "learning_rate": 6.032991405250702e-06,
+      "loss": 0.403,
+      "step": 765
+    },
+    {
+      "epoch": 1.47168,
+      "grad_norm": 0.5320404171943665,
+      "learning_rate": 6.022042192065946e-06,
+      "loss": 0.3455,
+      "step": 766
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 0.5829054117202759,
+      "learning_rate": 6.011087861660191e-06,
+      "loss": 0.3609,
+      "step": 767
+    },
+    {
+      "epoch": 1.47552,
+      "grad_norm": 0.6140908598899841,
+      "learning_rate": 6.000128468880223e-06,
+      "loss": 0.3931,
+      "step": 768
+    },
+    {
+      "epoch": 1.47744,
+      "grad_norm": 0.5233393907546997,
+      "learning_rate": 5.989164068598175e-06,
+      "loss": 0.3464,
+      "step": 769
+    },
+    {
+      "epoch": 1.47936,
+      "grad_norm": 0.5652267336845398,
+      "learning_rate": 5.978194715711254e-06,
+      "loss": 0.3436,
+      "step": 770
+    },
+    {
+      "epoch": 1.48128,
+      "grad_norm": 0.559798002243042,
+      "learning_rate": 5.9672204651414564e-06,
+      "loss": 0.3639,
+      "step": 771
+    },
+    {
+      "epoch": 1.4832,
+      "grad_norm": 0.5387717485427856,
+      "learning_rate": 5.956241371835312e-06,
+      "loss": 0.3482,
+      "step": 772
+    },
+    {
+      "epoch": 1.48512,
+      "grad_norm": 0.5308160781860352,
+      "learning_rate": 5.945257490763588e-06,
+      "loss": 0.3599,
+      "step": 773
+    },
+    {
+      "epoch": 1.48704,
+      "grad_norm": 0.548055112361908,
+      "learning_rate": 5.934268876921026e-06,
+      "loss": 0.3309,
+      "step": 774
+    },
+    {
+      "epoch": 1.48896,
+      "grad_norm": 0.576321542263031,
+      "learning_rate": 5.923275585326064e-06,
+      "loss": 0.373,
+      "step": 775
+    },
+    {
+      "epoch": 1.49088,
+      "grad_norm": 0.5497064590454102,
+      "learning_rate": 5.912277671020564e-06,
+      "loss": 0.3578,
+      "step": 776
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 0.5383490324020386,
+      "learning_rate": 5.90127518906953e-06,
+      "loss": 0.3646,
+      "step": 777
+    },
+    {
+      "epoch": 1.49472,
+      "grad_norm": 0.5955314636230469,
+      "learning_rate": 5.890268194560834e-06,
+      "loss": 0.3739,
+      "step": 778
+    },
+    {
+      "epoch": 1.49664,
+      "grad_norm": 0.5158449411392212,
+      "learning_rate": 5.8792567426049465e-06,
+      "loss": 0.3383,
+      "step": 779
+    },
+    {
+      "epoch": 1.49856,
+      "grad_norm": 0.6251073479652405,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.3526,
+      "step": 780
+    },
+    {
+      "epoch": 1.50048,
+      "grad_norm": 0.5680933594703674,
+      "learning_rate": 5.857220686904779e-06,
+      "loss": 0.3781,
+      "step": 781
+    },
+    {
+      "epoch": 1.5024,
+      "grad_norm": 0.5707352161407471,
+      "learning_rate": 5.846196193491919e-06,
+      "loss": 0.3529,
+      "step": 782
+    },
+    {
+      "epoch": 1.5043199999999999,
+      "grad_norm": 0.5336598753929138,
+      "learning_rate": 5.835167463294155e-06,
+      "loss": 0.3459,
+      "step": 783
+    },
+    {
+      "epoch": 1.50624,
+      "grad_norm": 0.5436135530471802,
+      "learning_rate": 5.824134551530783e-06,
+      "loss": 0.3501,
+      "step": 784
+    },
+    {
+      "epoch": 1.50816,
+      "grad_norm": 0.5132091641426086,
+      "learning_rate": 5.813097513442035e-06,
+      "loss": 0.3575,
+      "step": 785
+    },
+    {
+      "epoch": 1.5100799999999999,
+      "grad_norm": 0.5584456324577332,
+      "learning_rate": 5.8020564042888015e-06,
+      "loss": 0.3819,
+      "step": 786
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 0.5445287227630615,
+      "learning_rate": 5.79101127935236e-06,
+      "loss": 0.3718,
+      "step": 787
+    },
+    {
+      "epoch": 1.5139200000000002,
+      "grad_norm": 0.541552722454071,
+      "learning_rate": 5.77996219393409e-06,
+      "loss": 0.3701,
+      "step": 788
+    },
+    {
+      "epoch": 1.5158399999999999,
+      "grad_norm": 0.5952560901641846,
+      "learning_rate": 5.768909203355203e-06,
+      "loss": 0.3813,
+      "step": 789
+    },
+    {
+      "epoch": 1.51776,
+      "grad_norm": 0.5288509130477905,
+      "learning_rate": 5.757852362956463e-06,
+      "loss": 0.3592,
+      "step": 790
+    },
+    {
+      "epoch": 1.5196800000000001,
+      "grad_norm": 0.5748488306999207,
+      "learning_rate": 5.7467917280979105e-06,
+      "loss": 0.347,
+      "step": 791
+    },
+    {
+      "epoch": 1.5215999999999998,
+      "grad_norm": 0.5155866146087646,
+      "learning_rate": 5.735727354158581e-06,
+      "loss": 0.3234,
+      "step": 792
+    },
+    {
+      "epoch": 1.52352,
+      "grad_norm": 0.5519395470619202,
+      "learning_rate": 5.724659296536234e-06,
+      "loss": 0.3436,
+      "step": 793
+    },
+    {
+      "epoch": 1.5254400000000001,
+      "grad_norm": 0.6205373406410217,
+      "learning_rate": 5.713587610647073e-06,
+      "loss": 0.3855,
+      "step": 794
+    },
+    {
+      "epoch": 1.52736,
+      "grad_norm": 0.571591317653656,
+      "learning_rate": 5.7025123519254644e-06,
+      "loss": 0.3978,
+      "step": 795
+    },
+    {
+      "epoch": 1.52928,
+      "grad_norm": 0.5796583294868469,
+      "learning_rate": 5.6914335758236665e-06,
+      "loss": 0.3707,
+      "step": 796
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 0.5975468754768372,
+      "learning_rate": 5.680351337811547e-06,
+      "loss": 0.3816,
+      "step": 797
+    },
+    {
+      "epoch": 1.53312,
+      "grad_norm": 0.53078693151474,
+      "learning_rate": 5.669265693376309e-06,
+      "loss": 0.3608,
+      "step": 798
+    },
+    {
+      "epoch": 1.53504,
+      "grad_norm": 0.5230347514152527,
+      "learning_rate": 5.658176698022208e-06,
+      "loss": 0.3453,
+      "step": 799
+    },
+    {
+      "epoch": 1.53696,
+      "grad_norm": 0.5867753028869629,
+      "learning_rate": 5.647084407270277e-06,
+      "loss": 0.3872,
+      "step": 800
+    },
+    {
+      "epoch": 1.53888,
+      "grad_norm": 0.632377564907074,
+      "learning_rate": 5.6359888766580555e-06,
+      "loss": 0.3845,
+      "step": 801
+    },
+    {
+      "epoch": 1.5408,
+      "grad_norm": 0.5891371369361877,
+      "learning_rate": 5.624890161739292e-06,
+      "loss": 0.399,
+      "step": 802
+    },
+    {
+      "epoch": 1.54272,
+      "grad_norm": 0.5270923376083374,
+      "learning_rate": 5.6137883180836925e-06,
+      "loss": 0.3011,
+      "step": 803
+    },
+    {
+      "epoch": 1.54464,
+      "grad_norm": 0.5440530776977539,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.3539,
+      "step": 804
+    },
+    {
+      "epoch": 1.54656,
+      "grad_norm": 0.5289129614830017,
+      "learning_rate": 5.591575466918816e-06,
+      "loss": 0.336,
+      "step": 805
+    },
+    {
+      "epoch": 1.54848,
+      "grad_norm": 0.5319774746894836,
+      "learning_rate": 5.5804645706261515e-06,
+      "loss": 0.3509,
+      "step": 806
+    },
+    {
+      "epoch": 1.5504,
+      "grad_norm": 0.5817285180091858,
+      "learning_rate": 5.569350768029312e-06,
+      "loss": 0.3855,
+      "step": 807
+    },
+    {
+      "epoch": 1.55232,
+      "grad_norm": 0.5126243233680725,
+      "learning_rate": 5.5582341147735396e-06,
+      "loss": 0.3469,
+      "step": 808
+    },
+    {
+      "epoch": 1.55424,
+      "grad_norm": 0.5536636710166931,
+      "learning_rate": 5.5471146665183455e-06,
+      "loss": 0.3704,
+      "step": 809
+    },
+    {
+      "epoch": 1.55616,
+      "grad_norm": 0.5727599859237671,
+      "learning_rate": 5.53599247893724e-06,
+      "loss": 0.399,
+      "step": 810
+    },
+    {
+      "epoch": 1.55808,
+      "grad_norm": 0.5912884473800659,
+      "learning_rate": 5.524867607717445e-06,
+      "loss": 0.3451,
+      "step": 811
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5964463949203491,
+      "learning_rate": 5.5137401085596224e-06,
+      "loss": 0.3826,
+      "step": 812
+    },
+    {
+      "epoch": 1.56192,
+      "grad_norm": 0.547386109828949,
+      "learning_rate": 5.502610037177586e-06,
+      "loss": 0.3823,
+      "step": 813
+    },
+    {
+      "epoch": 1.56384,
+      "grad_norm": 0.534135103225708,
+      "learning_rate": 5.491477449298036e-06,
+      "loss": 0.3504,
+      "step": 814
+    },
+    {
+      "epoch": 1.56576,
+      "grad_norm": 0.6092579364776611,
+      "learning_rate": 5.480342400660268e-06,
+      "loss": 0.4181,
+      "step": 815
+    },
+    {
+      "epoch": 1.56768,
+      "grad_norm": 0.5891258716583252,
+      "learning_rate": 5.469204947015897e-06,
+      "loss": 0.36,
+      "step": 816
+    },
+    {
+      "epoch": 1.5695999999999999,
+      "grad_norm": 0.5518959760665894,
+      "learning_rate": 5.458065144128584e-06,
+      "loss": 0.3357,
+      "step": 817
+    },
+    {
+      "epoch": 1.57152,
+      "grad_norm": 0.5938230156898499,
+      "learning_rate": 5.4469230477737466e-06,
+      "loss": 0.3976,
+      "step": 818
+    },
+    {
+      "epoch": 1.57344,
+      "grad_norm": 0.5118221640586853,
+      "learning_rate": 5.435778713738292e-06,
+      "loss": 0.334,
+      "step": 819
+    },
+    {
+      "epoch": 1.5753599999999999,
+      "grad_norm": 4.7820658683776855,
+      "learning_rate": 5.424632197820325e-06,
+      "loss": 0.5372,
+      "step": 820
+    },
+    {
+      "epoch": 1.57728,
+      "grad_norm": 0.5630938410758972,
+      "learning_rate": 5.413483555828879e-06,
+      "loss": 0.3874,
+      "step": 821
+    },
+    {
+      "epoch": 1.5792000000000002,
+      "grad_norm": 0.5636446475982666,
+      "learning_rate": 5.402332843583631e-06,
+      "loss": 0.3517,
+      "step": 822
+    },
+    {
+      "epoch": 1.5811199999999999,
+      "grad_norm": 0.47537457942962646,
+      "learning_rate": 5.391180116914621e-06,
+      "loss": 0.3061,
+      "step": 823
+    },
+    {
+      "epoch": 1.58304,
+      "grad_norm": 0.5182272791862488,
+      "learning_rate": 5.380025431661981e-06,
+      "loss": 0.3446,
+      "step": 824
+    },
+    {
+      "epoch": 1.5849600000000001,
+      "grad_norm": 0.6314375996589661,
+      "learning_rate": 5.368868843675642e-06,
+      "loss": 0.3632,
+      "step": 825
+    },
+    {
+      "epoch": 1.5868799999999998,
+      "grad_norm": 0.5926188826560974,
+      "learning_rate": 5.3577104088150685e-06,
+      "loss": 0.4132,
+      "step": 826
+    },
+    {
+      "epoch": 1.5888,
+      "grad_norm": 0.5382211208343506,
+      "learning_rate": 5.346550182948966e-06,
+      "loss": 0.3243,
+      "step": 827
+    },
+    {
+      "epoch": 1.5907200000000001,
+      "grad_norm": 0.5299359560012817,
+      "learning_rate": 5.335388221955012e-06,
+      "loss": 0.3412,
+      "step": 828
+    },
+    {
+      "epoch": 1.5926399999999998,
+      "grad_norm": 0.5448858141899109,
+      "learning_rate": 5.3242245817195705e-06,
+      "loss": 0.3508,
+      "step": 829
+    },
+    {
+      "epoch": 1.59456,
+      "grad_norm": 0.5473446249961853,
+      "learning_rate": 5.31305931813741e-06,
+      "loss": 0.3759,
+      "step": 830
+    },
+    {
+      "epoch": 1.5964800000000001,
+      "grad_norm": 0.5329676866531372,
+      "learning_rate": 5.301892487111431e-06,
+      "loss": 0.3714,
+      "step": 831
+    },
+    {
+      "epoch": 1.5984,
+      "grad_norm": 0.5705289840698242,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.4021,
+      "step": 832
+    },
+    {
+      "epoch": 1.60032,
+      "grad_norm": 0.5319229960441589,
+      "learning_rate": 5.279554346378572e-06,
+      "loss": 0.3666,
+      "step": 833
+    },
+    {
+      "epoch": 1.60224,
+      "grad_norm": 0.5316363573074341,
+      "learning_rate": 5.268383148515608e-06,
+      "loss": 0.3707,
+      "step": 834
+    },
+    {
+      "epoch": 1.60416,
+      "grad_norm": 0.590303361415863,
+      "learning_rate": 5.257210606896102e-06,
+      "loss": 0.4312,
+      "step": 835
+    },
+    {
+      "epoch": 1.60608,
+      "grad_norm": 0.4883015751838684,
+      "learning_rate": 5.246036777459391e-06,
+      "loss": 0.3006,
+      "step": 836
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 0.5569459199905396,
+      "learning_rate": 5.234861716151264e-06,
+      "loss": 0.3835,
+      "step": 837
+    },
+    {
+      "epoch": 1.60992,
+      "grad_norm": 0.5687525272369385,
+      "learning_rate": 5.223685478923671e-06,
+      "loss": 0.3909,
+      "step": 838
+    },
+    {
+      "epoch": 1.61184,
+      "grad_norm": 0.5588273406028748,
+      "learning_rate": 5.2125081217344595e-06,
+      "loss": 0.343,
+      "step": 839
+    },
+    {
+      "epoch": 1.61376,
+      "grad_norm": 0.5663592219352722,
+      "learning_rate": 5.201329700547077e-06,
+      "loss": 0.4031,
+      "step": 840
+    },
+    {
+      "epoch": 1.61568,
+      "grad_norm": 0.5880895853042603,
+      "learning_rate": 5.190150271330298e-06,
+      "loss": 0.391,
+      "step": 841
+    },
+    {
+      "epoch": 1.6176,
+      "grad_norm": 0.6419731974601746,
+      "learning_rate": 5.178969890057953e-06,
+      "loss": 0.4096,
+      "step": 842
+    },
+    {
+      "epoch": 1.61952,
+      "grad_norm": 0.5403954982757568,
+      "learning_rate": 5.167788612708627e-06,
+      "loss": 0.3445,
+      "step": 843
+    },
+    {
+      "epoch": 1.62144,
+      "grad_norm": 0.5250660181045532,
+      "learning_rate": 5.156606495265402e-06,
+      "loss": 0.3236,
+      "step": 844
+    },
+    {
+      "epoch": 1.62336,
+      "grad_norm": 0.582247257232666,
+      "learning_rate": 5.145423593715558e-06,
+      "loss": 0.4054,
+      "step": 845
+    },
+    {
+      "epoch": 1.62528,
+      "grad_norm": 0.561072051525116,
+      "learning_rate": 5.1342399640503074e-06,
+      "loss": 0.3768,
+      "step": 846
+    },
+    {
+      "epoch": 1.6272,
+      "grad_norm": 0.5809715986251831,
+      "learning_rate": 5.1230556622645026e-06,
+      "loss": 0.3578,
+      "step": 847
+    },
+    {
+      "epoch": 1.62912,
+      "grad_norm": 0.5471744537353516,
+      "learning_rate": 5.111870744356366e-06,
+      "loss": 0.3654,
+      "step": 848
+    },
+    {
+      "epoch": 1.63104,
+      "grad_norm": 0.5228375792503357,
+      "learning_rate": 5.100685266327202e-06,
+      "loss": 0.3503,
+      "step": 849
+    },
+    {
+      "epoch": 1.63296,
+      "grad_norm": 0.5058294534683228,
+      "learning_rate": 5.0894992841811216e-06,
+      "loss": 0.335,
+      "step": 850
+    },
+    {
+      "epoch": 1.6348799999999999,
+      "grad_norm": 0.5544894933700562,
+      "learning_rate": 5.0783128539247585e-06,
+      "loss": 0.4017,
+      "step": 851
+    },
+    {
+      "epoch": 1.6368,
+      "grad_norm": 0.5488715767860413,
+      "learning_rate": 5.067126031566988e-06,
+      "loss": 0.3539,
+      "step": 852
+    },
+    {
+      "epoch": 1.63872,
+      "grad_norm": 0.5239600539207458,
+      "learning_rate": 5.055938873118653e-06,
+      "loss": 0.3736,
+      "step": 853
+    },
+    {
+      "epoch": 1.6406399999999999,
+      "grad_norm": 0.5390827059745789,
+      "learning_rate": 5.044751434592274e-06,
+      "loss": 0.3615,
+      "step": 854
+    },
+    {
+      "epoch": 1.64256,
+      "grad_norm": 0.47422555088996887,
+      "learning_rate": 5.033563772001782e-06,
+      "loss": 0.3079,
+      "step": 855
+    },
+    {
+      "epoch": 1.6444800000000002,
+      "grad_norm": 0.5674540996551514,
+      "learning_rate": 5.022375941362218e-06,
+      "loss": 0.3778,
+      "step": 856
+    },
+    {
+      "epoch": 1.6463999999999999,
+      "grad_norm": 0.5452783703804016,
+      "learning_rate": 5.011187998689474e-06,
+      "loss": 0.3285,
+      "step": 857
+    },
+    {
+      "epoch": 1.64832,
+      "grad_norm": 0.5746825337409973,
+      "learning_rate": 5e-06,
+      "loss": 0.3951,
+      "step": 858
+    },
+    {
+      "epoch": 1.6502400000000002,
+      "grad_norm": 0.521946132183075,
+      "learning_rate": 4.988812001310528e-06,
+      "loss": 0.3859,
+      "step": 859
+    },
+    {
+      "epoch": 1.6521599999999999,
+      "grad_norm": 0.5596005320549011,
+      "learning_rate": 4.977624058637783e-06,
+      "loss": 0.391,
+      "step": 860
+    },
+    {
+      "epoch": 1.65408,
+      "grad_norm": 0.5569255352020264,
+      "learning_rate": 4.9664362279982205e-06,
+      "loss": 0.3602,
+      "step": 861
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 0.5943171381950378,
+      "learning_rate": 4.955248565407727e-06,
+      "loss": 0.4087,
+      "step": 862
+    },
+    {
+      "epoch": 1.6579199999999998,
+      "grad_norm": 0.5397094488143921,
+      "learning_rate": 4.944061126881348e-06,
+      "loss": 0.3269,
+      "step": 863
+    },
+    {
+      "epoch": 1.65984,
+      "grad_norm": 0.5474894046783447,
+      "learning_rate": 4.932873968433014e-06,
+      "loss": 0.3509,
+      "step": 864
+    },
+    {
+      "epoch": 1.6617600000000001,
+      "grad_norm": 0.5501736998558044,
+      "learning_rate": 4.921687146075244e-06,
+      "loss": 0.3923,
+      "step": 865
+    },
+    {
+      "epoch": 1.66368,
+      "grad_norm": 0.5606262683868408,
+      "learning_rate": 4.910500715818879e-06,
+      "loss": 0.3566,
+      "step": 866
+    },
+    {
+      "epoch": 1.6656,
+      "grad_norm": 0.530535876750946,
+      "learning_rate": 4.8993147336728e-06,
+      "loss": 0.3172,
+      "step": 867
+    },
+    {
+      "epoch": 1.6675200000000001,
+      "grad_norm": 0.5624592900276184,
+      "learning_rate": 4.8881292556436355e-06,
+      "loss": 0.3246,
+      "step": 868
+    },
+    {
+      "epoch": 1.66944,
+      "grad_norm": 0.5291157960891724,
+      "learning_rate": 4.876944337735499e-06,
+      "loss": 0.3902,
+      "step": 869
+    },
+    {
+      "epoch": 1.67136,
+      "grad_norm": 0.6082698702812195,
+      "learning_rate": 4.865760035949695e-06,
+      "loss": 0.4108,
+      "step": 870
+    },
+    {
+      "epoch": 1.67328,
+      "grad_norm": 0.6572030782699585,
+      "learning_rate": 4.854576406284443e-06,
+      "loss": 0.3745,
+      "step": 871
+    },
+    {
+      "epoch": 1.6752,
+      "grad_norm": 0.5564181208610535,
+      "learning_rate": 4.8433935047346e-06,
+      "loss": 0.3617,
+      "step": 872
+    },
+    {
+      "epoch": 1.67712,
+      "grad_norm": 0.5101797580718994,
+      "learning_rate": 4.832211387291374e-06,
+      "loss": 0.3251,
+      "step": 873
+    },
+    {
+      "epoch": 1.67904,
+      "grad_norm": 0.5082737803459167,
+      "learning_rate": 4.821030109942048e-06,
+      "loss": 0.3304,
+      "step": 874
+    },
+    {
+      "epoch": 1.68096,
+      "grad_norm": 0.5568008422851562,
+      "learning_rate": 4.8098497286697024e-06,
+      "loss": 0.3657,
+      "step": 875
+    },
+    {
+      "epoch": 1.68288,
+      "grad_norm": 0.539011538028717,
+      "learning_rate": 4.798670299452926e-06,
+      "loss": 0.3741,
+      "step": 876
+    },
+    {
+      "epoch": 1.6848,
+      "grad_norm": 0.5386427640914917,
+      "learning_rate": 4.787491878265542e-06,
+      "loss": 0.3893,
+      "step": 877
+    },
+    {
+      "epoch": 1.68672,
+      "grad_norm": 0.5311138033866882,
+      "learning_rate": 4.77631452107633e-06,
+      "loss": 0.3439,
+      "step": 878
+    },
+    {
+      "epoch": 1.68864,
+      "grad_norm": 0.5659523606300354,
+      "learning_rate": 4.765138283848739e-06,
+      "loss": 0.4049,
+      "step": 879
+    },
+    {
+      "epoch": 1.69056,
+      "grad_norm": 0.5437709093093872,
+      "learning_rate": 4.75396322254061e-06,
+      "loss": 0.3346,
+      "step": 880
+    },
+    {
+      "epoch": 1.69248,
+      "grad_norm": 0.5165103077888489,
+      "learning_rate": 4.742789393103899e-06,
+      "loss": 0.3337,
+      "step": 881
+    },
+    {
+      "epoch": 1.6944,
+      "grad_norm": 0.5523893237113953,
+      "learning_rate": 4.731616851484392e-06,
+      "loss": 0.338,
+      "step": 882
+    },
+    {
+      "epoch": 1.69632,
+      "grad_norm": 0.5683523416519165,
+      "learning_rate": 4.720445653621429e-06,
+      "loss": 0.3502,
+      "step": 883
+    },
+    {
+      "epoch": 1.69824,
+      "grad_norm": 0.5193224549293518,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.3275,
+      "step": 884
+    },
+    {
+      "epoch": 1.70016,
+      "grad_norm": 0.578707754611969,
+      "learning_rate": 4.69810751288857e-06,
+      "loss": 0.4053,
+      "step": 885
+    },
+    {
+      "epoch": 1.70208,
+      "grad_norm": 0.5384418964385986,
+      "learning_rate": 4.686940681862591e-06,
+      "loss": 0.3601,
+      "step": 886
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 0.5045821070671082,
+      "learning_rate": 4.675775418280432e-06,
+      "loss": 0.3539,
+      "step": 887
+    },
+    {
+      "epoch": 1.7059199999999999,
+      "grad_norm": 0.5180543661117554,
+      "learning_rate": 4.664611778044988e-06,
+      "loss": 0.3463,
+      "step": 888
+    },
+    {
+      "epoch": 1.70784,
+      "grad_norm": 0.5512346625328064,
+      "learning_rate": 4.653449817051035e-06,
+      "loss": 0.396,
+      "step": 889
+    },
+    {
+      "epoch": 1.70976,
+      "grad_norm": 0.5784295797348022,
+      "learning_rate": 4.642289591184934e-06,
+      "loss": 0.3735,
+      "step": 890
+    },
+    {
+      "epoch": 1.7116799999999999,
+      "grad_norm": 0.5158311128616333,
+      "learning_rate": 4.631131156324359e-06,
+      "loss": 0.3312,
+      "step": 891
+    },
+    {
+      "epoch": 1.7136,
+      "grad_norm": 0.5407631397247314,
+      "learning_rate": 4.619974568338021e-06,
+      "loss": 0.3902,
+      "step": 892
+    },
+    {
+      "epoch": 1.7155200000000002,
+      "grad_norm": 0.5775696039199829,
+      "learning_rate": 4.60881988308538e-06,
+      "loss": 0.4204,
+      "step": 893
+    },
+    {
+      "epoch": 1.7174399999999999,
+      "grad_norm": 0.5379625558853149,
+      "learning_rate": 4.597667156416371e-06,
+      "loss": 0.3813,
+      "step": 894
+    },
+    {
+      "epoch": 1.71936,
+      "grad_norm": 0.5556676387786865,
+      "learning_rate": 4.586516444171123e-06,
+      "loss": 0.3732,
+      "step": 895
+    },
+    {
+      "epoch": 1.7212800000000001,
+      "grad_norm": 0.5742902159690857,
+      "learning_rate": 4.575367802179675e-06,
+      "loss": 0.3558,
+      "step": 896
+    },
+    {
+      "epoch": 1.7231999999999998,
+      "grad_norm": 0.582093358039856,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.4151,
+      "step": 897
+    },
+    {
+      "epoch": 1.72512,
+      "grad_norm": 0.5696203112602234,
+      "learning_rate": 4.553076952226255e-06,
+      "loss": 0.3747,
+      "step": 898
+    },
+    {
+      "epoch": 1.7270400000000001,
+      "grad_norm": 0.584201455116272,
+      "learning_rate": 4.541934855871417e-06,
+      "loss": 0.391,
+      "step": 899
+    },
+    {
+      "epoch": 1.72896,
+      "grad_norm": 0.5832385420799255,
+      "learning_rate": 4.530795052984104e-06,
+      "loss": 0.3622,
+      "step": 900
+    },
+    {
+      "epoch": 1.73088,
+      "grad_norm": 0.5924862623214722,
+      "learning_rate": 4.519657599339735e-06,
+      "loss": 0.4282,
+      "step": 901
+    },
+    {
+      "epoch": 1.7328000000000001,
+      "grad_norm": 0.5580012202262878,
+      "learning_rate": 4.508522550701965e-06,
+      "loss": 0.3675,
+      "step": 902
+    },
+    {
+      "epoch": 1.73472,
+      "grad_norm": 0.5788021087646484,
+      "learning_rate": 4.497389962822416e-06,
+      "loss": 0.4349,
+      "step": 903
+    },
+    {
+      "epoch": 1.73664,
+      "grad_norm": 0.5625857710838318,
+      "learning_rate": 4.48625989144038e-06,
+      "loss": 0.3873,
+      "step": 904
+    },
+    {
+      "epoch": 1.73856,
+      "grad_norm": 0.612031102180481,
+      "learning_rate": 4.475132392282556e-06,
+      "loss": 0.3826,
+      "step": 905
+    },
+    {
+      "epoch": 1.74048,
+      "grad_norm": 0.5689667463302612,
+      "learning_rate": 4.464007521062761e-06,
+      "loss": 0.3558,
+      "step": 906
+    },
+    {
+      "epoch": 1.7424,
+      "grad_norm": 0.5157765746116638,
+      "learning_rate": 4.452885333481657e-06,
+      "loss": 0.3294,
+      "step": 907
+    },
+    {
+      "epoch": 1.74432,
+      "grad_norm": 0.5565506815910339,
+      "learning_rate": 4.441765885226462e-06,
+      "loss": 0.3833,
+      "step": 908
+    },
+    {
+      "epoch": 1.74624,
+      "grad_norm": 0.5499540567398071,
+      "learning_rate": 4.4306492319706895e-06,
+      "loss": 0.3687,
+      "step": 909
+    },
+    {
+      "epoch": 1.74816,
+      "grad_norm": 0.5661587715148926,
+      "learning_rate": 4.4195354293738484e-06,
+      "loss": 0.3495,
+      "step": 910
+    },
+    {
+      "epoch": 1.75008,
+      "grad_norm": 0.586073637008667,
+      "learning_rate": 4.4084245330811855e-06,
+      "loss": 0.3749,
+      "step": 911
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 0.5580382943153381,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.3443,
+      "step": 912
+    },
+    {
+      "epoch": 1.75392,
+      "grad_norm": 0.5540907979011536,
+      "learning_rate": 4.386211681916309e-06,
+      "loss": 0.3686,
+      "step": 913
+    },
+    {
+      "epoch": 1.75584,
+      "grad_norm": 0.5244554877281189,
+      "learning_rate": 4.3751098382607084e-06,
+      "loss": 0.3297,
+      "step": 914
+    },
+    {
+      "epoch": 1.75776,
+      "grad_norm": 0.5991559624671936,
+      "learning_rate": 4.364011123341947e-06,
+      "loss": 0.3935,
+      "step": 915
+    },
+    {
+      "epoch": 1.75968,
+      "grad_norm": 0.5043613314628601,
+      "learning_rate": 4.352915592729723e-06,
+      "loss": 0.3208,
+      "step": 916
+    },
+    {
+      "epoch": 1.7616,
+      "grad_norm": 0.6101287007331848,
+      "learning_rate": 4.341823301977794e-06,
+      "loss": 0.4001,
+      "step": 917
+    },
+    {
+      "epoch": 1.76352,
+      "grad_norm": 0.5094191431999207,
+      "learning_rate": 4.330734306623694e-06,
+      "loss": 0.338,
+      "step": 918
+    },
+    {
+      "epoch": 1.76544,
+      "grad_norm": 0.5617073178291321,
+      "learning_rate": 4.319648662188453e-06,
+      "loss": 0.3898,
+      "step": 919
+    },
+    {
+      "epoch": 1.76736,
+      "grad_norm": 0.5536214709281921,
+      "learning_rate": 4.308566424176336e-06,
+      "loss": 0.35,
+      "step": 920
+    },
+    {
+      "epoch": 1.76928,
+      "grad_norm": 0.5714005827903748,
+      "learning_rate": 4.297487648074538e-06,
+      "loss": 0.3944,
+      "step": 921
+    },
+    {
+      "epoch": 1.7711999999999999,
+      "grad_norm": 0.5356638431549072,
+      "learning_rate": 4.286412389352929e-06,
+      "loss": 0.3734,
+      "step": 922
+    },
+    {
+      "epoch": 1.77312,
+      "grad_norm": 0.5743429064750671,
+      "learning_rate": 4.275340703463767e-06,
+      "loss": 0.374,
+      "step": 923
+    },
+    {
+      "epoch": 1.77504,
+      "grad_norm": 0.6010902523994446,
+      "learning_rate": 4.264272645841419e-06,
+      "loss": 0.4203,
+      "step": 924
+    },
+    {
+      "epoch": 1.7769599999999999,
+      "grad_norm": 0.5743195414543152,
+      "learning_rate": 4.253208271902091e-06,
+      "loss": 0.3346,
+      "step": 925
+    },
+    {
+      "epoch": 1.77888,
+      "grad_norm": 0.5518834590911865,
+      "learning_rate": 4.242147637043539e-06,
+      "loss": 0.3619,
+      "step": 926
+    },
+    {
+      "epoch": 1.7808000000000002,
+      "grad_norm": 0.5507184863090515,
+      "learning_rate": 4.231090796644798e-06,
+      "loss": 0.373,
+      "step": 927
+    },
+    {
+      "epoch": 1.7827199999999999,
+      "grad_norm": 0.5593322515487671,
+      "learning_rate": 4.220037806065911e-06,
+      "loss": 0.3461,
+      "step": 928
+    },
+    {
+      "epoch": 1.78464,
+      "grad_norm": 0.5120484828948975,
+      "learning_rate": 4.208988720647642e-06,
+      "loss": 0.3509,
+      "step": 929
+    },
+    {
+      "epoch": 1.7865600000000001,
+      "grad_norm": 0.5518879294395447,
+      "learning_rate": 4.1979435957111984e-06,
+      "loss": 0.3407,
+      "step": 930
+    },
+    {
+      "epoch": 1.7884799999999998,
+      "grad_norm": 0.5580224990844727,
+      "learning_rate": 4.1869024865579664e-06,
+      "loss": 0.4045,
+      "step": 931
+    },
+    {
+      "epoch": 1.7904,
+      "grad_norm": 0.5249459743499756,
+      "learning_rate": 4.175865448469219e-06,
+      "loss": 0.297,
+      "step": 932
+    },
+    {
+      "epoch": 1.7923200000000001,
+      "grad_norm": 0.5686970353126526,
+      "learning_rate": 4.164832536705845e-06,
+      "loss": 0.2825,
+      "step": 933
+    },
+    {
+      "epoch": 1.7942399999999998,
+      "grad_norm": 0.6355201005935669,
+      "learning_rate": 4.153803806508083e-06,
+      "loss": 0.3718,
+      "step": 934
+    },
+    {
+      "epoch": 1.79616,
+      "grad_norm": 0.5277770161628723,
+      "learning_rate": 4.142779313095223e-06,
+      "loss": 0.3602,
+      "step": 935
+    },
+    {
+      "epoch": 1.7980800000000001,
+      "grad_norm": 0.5232786536216736,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.3713,
+      "step": 936
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.55926114320755,
+      "learning_rate": 4.120743257395054e-06,
+      "loss": 0.3971,
+      "step": 937
+    },
+    {
+      "epoch": 1.80192,
+      "grad_norm": 0.6262421011924744,
+      "learning_rate": 4.109731805439168e-06,
+      "loss": 0.3409,
+      "step": 938
+    },
+    {
+      "epoch": 1.80384,
+      "grad_norm": 0.5595870018005371,
+      "learning_rate": 4.098724810930472e-06,
+      "loss": 0.4066,
+      "step": 939
+    },
+    {
+      "epoch": 1.80576,
+      "grad_norm": 0.6054223775863647,
+      "learning_rate": 4.087722328979437e-06,
+      "loss": 0.4063,
+      "step": 940
+    },
+    {
+      "epoch": 1.80768,
+      "grad_norm": 0.5683716535568237,
+      "learning_rate": 4.076724414673937e-06,
+      "loss": 0.3514,
+      "step": 941
+    },
+    {
+      "epoch": 1.8096,
+      "grad_norm": 0.5242344737052917,
+      "learning_rate": 4.065731123078977e-06,
+      "loss": 0.3334,
+      "step": 942
+    },
+    {
+      "epoch": 1.81152,
+      "grad_norm": 0.5685484409332275,
+      "learning_rate": 4.054742509236416e-06,
+      "loss": 0.3887,
+      "step": 943
+    },
+    {
+      "epoch": 1.81344,
+      "grad_norm": 0.5299404263496399,
+      "learning_rate": 4.043758628164688e-06,
+      "loss": 0.3519,
+      "step": 944
+    },
+    {
+      "epoch": 1.81536,
+      "grad_norm": 0.529046356678009,
+      "learning_rate": 4.032779534858544e-06,
+      "loss": 0.3568,
+      "step": 945
+    },
+    {
+      "epoch": 1.81728,
+      "grad_norm": 0.5307803750038147,
+      "learning_rate": 4.021805284288749e-06,
+      "loss": 0.3648,
+      "step": 946
+    },
+    {
+      "epoch": 1.8192,
+      "grad_norm": 0.5602819919586182,
+      "learning_rate": 4.0108359314018254e-06,
+      "loss": 0.3795,
+      "step": 947
+    },
+    {
+      "epoch": 1.82112,
+      "grad_norm": 0.6019018292427063,
+      "learning_rate": 3.999871531119779e-06,
+      "loss": 0.4077,
+      "step": 948
+    },
+    {
+      "epoch": 1.82304,
+      "grad_norm": 0.5332850217819214,
+      "learning_rate": 3.988912138339812e-06,
+      "loss": 0.3455,
+      "step": 949
+    },
+    {
+      "epoch": 1.82496,
+      "grad_norm": 0.5987565517425537,
+      "learning_rate": 3.977957807934055e-06,
+      "loss": 0.3971,
+      "step": 950
+    },
+    {
+      "epoch": 1.82688,
+      "grad_norm": 0.5839775204658508,
+      "learning_rate": 3.9670085947493e-06,
+      "loss": 0.3364,
+      "step": 951
+    },
+    {
+      "epoch": 1.8288,
+      "grad_norm": 0.6140446066856384,
+      "learning_rate": 3.956064553606708e-06,
+      "loss": 0.3881,
+      "step": 952
+    },
+    {
+      "epoch": 1.83072,
+      "grad_norm": 0.5003345012664795,
+      "learning_rate": 3.945125739301547e-06,
+      "loss": 0.3204,
+      "step": 953
+    },
+    {
+      "epoch": 1.83264,
+      "grad_norm": 0.541694164276123,
+      "learning_rate": 3.934192206602921e-06,
+      "loss": 0.399,
+      "step": 954
+    },
+    {
+      "epoch": 1.83456,
+      "grad_norm": 0.5641539692878723,
+      "learning_rate": 3.923264010253479e-06,
+      "loss": 0.3789,
+      "step": 955
+    },
+    {
+      "epoch": 1.83648,
+      "grad_norm": 0.5361934900283813,
+      "learning_rate": 3.912341204969164e-06,
+      "loss": 0.3339,
+      "step": 956
+    },
+    {
+      "epoch": 1.8384,
+      "grad_norm": 0.5608660578727722,
+      "learning_rate": 3.901423845438916e-06,
+      "loss": 0.373,
+      "step": 957
+    },
+    {
+      "epoch": 1.84032,
+      "grad_norm": 0.6265275478363037,
+      "learning_rate": 3.890511986324413e-06,
+      "loss": 0.3986,
+      "step": 958
+    },
+    {
+      "epoch": 1.8422399999999999,
+      "grad_norm": 0.6064483523368835,
+      "learning_rate": 3.879605682259797e-06,
+      "loss": 0.3571,
+      "step": 959
+    },
+    {
+      "epoch": 1.84416,
+      "grad_norm": 0.5702192187309265,
+      "learning_rate": 3.86870498785139e-06,
+      "loss": 0.3579,
+      "step": 960
+    },
+    {
+      "epoch": 1.8460800000000002,
+      "grad_norm": 0.5513043999671936,
+      "learning_rate": 3.857809957677432e-06,
+      "loss": 0.3818,
+      "step": 961
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 0.5345434546470642,
+      "learning_rate": 3.8469206462878e-06,
+      "loss": 0.3556,
+      "step": 962
+    },
+    {
+      "epoch": 1.84992,
+      "grad_norm": 0.5314241051673889,
+      "learning_rate": 3.83603710820374e-06,
+      "loss": 0.3762,
+      "step": 963
+    },
+    {
+      "epoch": 1.8518400000000002,
+      "grad_norm": 0.5409510731697083,
+      "learning_rate": 3.825159397917589e-06,
+      "loss": 0.3365,
+      "step": 964
+    },
+    {
+      "epoch": 1.8537599999999999,
+      "grad_norm": 0.5638481974601746,
+      "learning_rate": 3.814287569892512e-06,
+      "loss": 0.3677,
+      "step": 965
+    },
+    {
+      "epoch": 1.85568,
+      "grad_norm": 0.5408279895782471,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.3515,
+      "step": 966
+    },
+    {
+      "epoch": 1.8576000000000001,
+      "grad_norm": 0.5382869839668274,
+      "learning_rate": 3.7925617783306757e-06,
+      "loss": 0.3496,
+      "step": 967
+    },
+    {
+      "epoch": 1.8595199999999998,
+      "grad_norm": 0.5887214541435242,
+      "learning_rate": 3.781707923571891e-06,
+      "loss": 0.4202,
+      "step": 968
+    },
+    {
+      "epoch": 1.86144,
+      "grad_norm": 0.530247151851654,
+      "learning_rate": 3.7708601686295756e-06,
+      "loss": 0.3736,
+      "step": 969
+    },
+    {
+      "epoch": 1.8633600000000001,
+      "grad_norm": 0.5137759447097778,
+      "learning_rate": 3.7600185678169083e-06,
+      "loss": 0.3535,
+      "step": 970
+    },
+    {
+      "epoch": 1.86528,
+      "grad_norm": 0.5769482851028442,
+      "learning_rate": 3.7491831754162533e-06,
+      "loss": 0.3985,
+      "step": 971
+    },
+    {
+      "epoch": 1.8672,
+      "grad_norm": 0.5656195282936096,
+      "learning_rate": 3.7383540456788915e-06,
+      "loss": 0.3603,
+      "step": 972
+    },
+    {
+      "epoch": 1.8691200000000001,
+      "grad_norm": 0.6007447838783264,
+      "learning_rate": 3.727531232824747e-06,
+      "loss": 0.3899,
+      "step": 973
+    },
+    {
+      "epoch": 1.87104,
+      "grad_norm": 0.564092755317688,
+      "learning_rate": 3.7167147910421165e-06,
+      "loss": 0.3856,
+      "step": 974
+    },
+    {
+      "epoch": 1.87296,
+      "grad_norm": 0.559190034866333,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.3887,
+      "step": 975
+    },
+    {
+      "epoch": 1.87488,
+      "grad_norm": 0.5306927561759949,
+      "learning_rate": 3.695101237284815e-06,
+      "loss": 0.3415,
+      "step": 976
+    },
+    {
+      "epoch": 1.8768,
+      "grad_norm": 0.5489330887794495,
+      "learning_rate": 3.6843042335261583e-06,
+      "loss": 0.3856,
+      "step": 977
+    },
+    {
+      "epoch": 1.87872,
+      "grad_norm": 0.5632824897766113,
+      "learning_rate": 3.6735138172704967e-06,
+      "loss": 0.3667,
+      "step": 978
+    },
+    {
+      "epoch": 1.88064,
+      "grad_norm": 0.5681033134460449,
+      "learning_rate": 3.662730042543926e-06,
+      "loss": 0.3936,
+      "step": 979
+    },
+    {
+      "epoch": 1.88256,
+      "grad_norm": 0.5619406700134277,
+      "learning_rate": 3.6519529633392825e-06,
+      "loss": 0.3426,
+      "step": 980
+    },
+    {
+      "epoch": 1.88448,
+      "grad_norm": 0.5577789545059204,
+      "learning_rate": 3.6411826336158785e-06,
+      "loss": 0.3559,
+      "step": 981
+    },
+    {
+      "epoch": 1.8864,
+      "grad_norm": 0.5660808682441711,
+      "learning_rate": 3.6304191072992376e-06,
+      "loss": 0.4005,
+      "step": 982
+    },
+    {
+      "epoch": 1.88832,
+      "grad_norm": 0.5848060250282288,
+      "learning_rate": 3.619662438280816e-06,
+      "loss": 0.3208,
+      "step": 983
+    },
+    {
+      "epoch": 1.89024,
+      "grad_norm": 0.5017204284667969,
+      "learning_rate": 3.6089126804177373e-06,
+      "loss": 0.3377,
+      "step": 984
+    },
+    {
+      "epoch": 1.89216,
+      "grad_norm": 0.5458231568336487,
+      "learning_rate": 3.5981698875325214e-06,
+      "loss": 0.3698,
+      "step": 985
+    },
+    {
+      "epoch": 1.89408,
+      "grad_norm": 0.5595934987068176,
+      "learning_rate": 3.5874341134128156e-06,
+      "loss": 0.3487,
+      "step": 986
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 0.5343905687332153,
+      "learning_rate": 3.5767054118111266e-06,
+      "loss": 0.3651,
+      "step": 987
+    },
+    {
+      "epoch": 1.89792,
+      "grad_norm": 0.6023030281066895,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.4365,
+      "step": 988
+    },
+    {
+      "epoch": 1.89984,
+      "grad_norm": 0.5929659605026245,
+      "learning_rate": 3.555269440994496e-06,
+      "loss": 0.3673,
+      "step": 989
+    },
+    {
+      "epoch": 1.90176,
+      "grad_norm": 0.5221308469772339,
+      "learning_rate": 3.544562279106436e-06,
+      "loss": 0.3396,
+      "step": 990
+    },
+    {
+      "epoch": 1.90368,
+      "grad_norm": 0.5567911863327026,
+      "learning_rate": 3.5338624043896154e-06,
+      "loss": 0.3945,
+      "step": 991
+    },
+    {
+      "epoch": 1.9056,
+      "grad_norm": 0.5417887568473816,
+      "learning_rate": 3.523169870416795e-06,
+      "loss": 0.3223,
+      "step": 992
+    },
+    {
+      "epoch": 1.9075199999999999,
+      "grad_norm": 0.561205267906189,
+      "learning_rate": 3.5124847307239863e-06,
+      "loss": 0.3693,
+      "step": 993
+    },
+    {
+      "epoch": 1.90944,
+      "grad_norm": 0.5760982036590576,
+      "learning_rate": 3.501807038810174e-06,
+      "loss": 0.3764,
+      "step": 994
+    },
+    {
+      "epoch": 1.91136,
+      "grad_norm": 0.6661338210105896,
+      "learning_rate": 3.4911368481370535e-06,
+      "loss": 0.4907,
+      "step": 995
+    },
+    {
+      "epoch": 1.9132799999999999,
+      "grad_norm": 0.523639440536499,
+      "learning_rate": 3.480474212128766e-06,
+      "loss": 0.3294,
+      "step": 996
+    },
+    {
+      "epoch": 1.9152,
+      "grad_norm": 0.5651407241821289,
+      "learning_rate": 3.469819184171623e-06,
+      "loss": 0.3478,
+      "step": 997
+    },
+    {
+      "epoch": 1.9171200000000002,
+      "grad_norm": 0.6051636338233948,
+      "learning_rate": 3.459171817613847e-06,
+      "loss": 0.3917,
+      "step": 998
+    },
+    {
+      "epoch": 1.9190399999999999,
+      "grad_norm": 0.6386885046958923,
+      "learning_rate": 3.4485321657653e-06,
+      "loss": 0.4514,
+      "step": 999
+    },
+    {
+      "epoch": 1.92096,
+      "grad_norm": 0.5221706628799438,
+      "learning_rate": 3.4379002818972122e-06,
+      "loss": 0.3378,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9228800000000001,
+      "grad_norm": 0.5450828671455383,
+      "learning_rate": 3.427276219241933e-06,
+      "loss": 0.4057,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9247999999999998,
+      "grad_norm": 0.5941938161849976,
+      "learning_rate": 3.416660030992639e-06,
+      "loss": 0.3835,
+      "step": 1002
+    },
+    {
+      "epoch": 1.92672,
+      "grad_norm": 0.5343064069747925,
+      "learning_rate": 3.406051770303087e-06,
+      "loss": 0.3556,
+      "step": 1003
+    },
+    {
+      "epoch": 1.9286400000000001,
+      "grad_norm": 0.6187920570373535,
+      "learning_rate": 3.3954514902873427e-06,
+      "loss": 0.4467,
+      "step": 1004
+    },
+    {
+      "epoch": 1.93056,
+      "grad_norm": 0.5299493074417114,
+      "learning_rate": 3.3848592440195118e-06,
+      "loss": 0.3762,
+      "step": 1005
+    },
+    {
+      "epoch": 1.93248,
+      "grad_norm": 0.5261619091033936,
+      "learning_rate": 3.3742750845334748e-06,
+      "loss": 0.3593,
+      "step": 1006
+    },
+    {
+      "epoch": 1.9344000000000001,
+      "grad_norm": 0.5474284887313843,
+      "learning_rate": 3.3636990648226258e-06,
+      "loss": 0.3842,
+      "step": 1007
+    },
+    {
+      "epoch": 1.93632,
+      "grad_norm": 0.5688561201095581,
+      "learning_rate": 3.3531312378396026e-06,
+      "loss": 0.4023,
+      "step": 1008
+    },
+    {
+      "epoch": 1.93824,
+      "grad_norm": 0.5590647459030151,
+      "learning_rate": 3.342571656496022e-06,
+      "loss": 0.374,
+      "step": 1009
+    },
+    {
+      "epoch": 1.94016,
+      "grad_norm": 0.561633825302124,
+      "learning_rate": 3.3320203736622185e-06,
+      "loss": 0.3897,
+      "step": 1010
+    },
+    {
+      "epoch": 1.94208,
+      "grad_norm": 0.58525151014328,
+      "learning_rate": 3.3214774421669777e-06,
+      "loss": 0.4162,
+      "step": 1011
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 0.507087230682373,
+      "learning_rate": 3.310942914797265e-06,
+      "loss": 0.3386,
+      "step": 1012
+    },
+    {
+      "epoch": 1.94592,
+      "grad_norm": 0.5507297515869141,
+      "learning_rate": 3.3004168442979755e-06,
+      "loss": 0.3908,
+      "step": 1013
+    },
+    {
+      "epoch": 1.94784,
+      "grad_norm": 0.5400334596633911,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.3477,
+      "step": 1014
+    },
+    {
+      "epoch": 1.94976,
+      "grad_norm": 0.5403082370758057,
+      "learning_rate": 3.2793902846782534e-06,
+      "loss": 0.3602,
+      "step": 1015
+    },
+    {
+      "epoch": 1.95168,
+      "grad_norm": 0.5268360376358032,
+      "learning_rate": 3.2688899008348386e-06,
+      "loss": 0.3441,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9536,
+      "grad_norm": 0.5958834886550903,
+      "learning_rate": 3.2583981844153487e-06,
+      "loss": 0.3952,
+      "step": 1017
+    },
+    {
+      "epoch": 1.95552,
+      "grad_norm": 0.5213354825973511,
+      "learning_rate": 3.2479151879503324e-06,
+      "loss": 0.3479,
+      "step": 1018
+    },
+    {
+      "epoch": 1.95744,
+      "grad_norm": 0.5577940940856934,
+      "learning_rate": 3.2374409639266695e-06,
+      "loss": 0.3529,
+      "step": 1019
+    },
+    {
+      "epoch": 1.95936,
+      "grad_norm": 0.5322228074073792,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.3316,
+      "step": 1020
+    },
+    {
+      "epoch": 1.96128,
+      "grad_norm": 0.5604656934738159,
+      "learning_rate": 3.2165190429310674e-06,
+      "loss": 0.3734,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9632,
+      "grad_norm": 0.5235439538955688,
+      "learning_rate": 3.206071450712235e-06,
+      "loss": 0.3689,
+      "step": 1022
+    },
+    {
+      "epoch": 1.96512,
+      "grad_norm": 0.5378146171569824,
+      "learning_rate": 3.1956328404404403e-06,
+      "loss": 0.3885,
+      "step": 1023
+    },
+    {
+      "epoch": 1.96704,
+      "grad_norm": 0.575285792350769,
+      "learning_rate": 3.1852032643803377e-06,
+      "loss": 0.3481,
+      "step": 1024
+    },
+    {
+      "epoch": 1.96896,
+      "grad_norm": 0.5725017786026001,
+      "learning_rate": 3.174782774751338e-06,
+      "loss": 0.4221,
+      "step": 1025
+    },
+    {
+      "epoch": 1.97088,
+      "grad_norm": 0.577306866645813,
+      "learning_rate": 3.1643714237273628e-06,
+      "loss": 0.3515,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9727999999999999,
+      "grad_norm": 0.5049177408218384,
+      "learning_rate": 3.1539692634365788e-06,
+      "loss": 0.3422,
+      "step": 1027
+    },
+    {
+      "epoch": 1.97472,
+      "grad_norm": 0.5968877077102661,
+      "learning_rate": 3.143576345961132e-06,
+      "loss": 0.3651,
+      "step": 1028
+    },
+    {
+      "epoch": 1.97664,
+      "grad_norm": 0.5615143775939941,
+      "learning_rate": 3.1331927233368954e-06,
+      "loss": 0.3493,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9785599999999999,
+      "grad_norm": 0.5772520303726196,
+      "learning_rate": 3.1228184475532015e-06,
+      "loss": 0.3815,
+      "step": 1030
+    },
+    {
+      "epoch": 1.98048,
+      "grad_norm": 0.5537616014480591,
+      "learning_rate": 3.112453570552583e-06,
+      "loss": 0.376,
+      "step": 1031
+    },
+    {
+      "epoch": 1.9824000000000002,
+      "grad_norm": 0.5214160084724426,
+      "learning_rate": 3.1020981442305187e-06,
+      "loss": 0.3128,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9843199999999999,
+      "grad_norm": 0.5592936277389526,
+      "learning_rate": 3.091752220435166e-06,
+      "loss": 0.3753,
+      "step": 1033
+    },
+    {
+      "epoch": 1.98624,
+      "grad_norm": 0.5977579951286316,
+      "learning_rate": 3.0814158509671015e-06,
+      "loss": 0.3696,
+      "step": 1034
+    },
+    {
+      "epoch": 1.9881600000000001,
+      "grad_norm": 0.5269472599029541,
+      "learning_rate": 3.0710890875790745e-06,
+      "loss": 0.3576,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9900799999999998,
+      "grad_norm": 0.586826503276825,
+      "learning_rate": 3.0607719819757264e-06,
+      "loss": 0.381,
+      "step": 1036
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 0.548525333404541,
+      "learning_rate": 3.0504645858133507e-06,
+      "loss": 0.3488,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9939200000000001,
+      "grad_norm": 0.5786440968513489,
+      "learning_rate": 3.040166950699626e-06,
+      "loss": 0.3858,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9958399999999998,
+      "grad_norm": 0.5861607789993286,
+      "learning_rate": 3.029879128193356e-06,
+      "loss": 0.3843,
+      "step": 1039
+    },
+    {
+      "epoch": 1.99776,
+      "grad_norm": 0.5301262736320496,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.3602,
+      "step": 1040
+    },
+    {
+      "epoch": 2.00064,
+      "grad_norm": 1.0699694156646729,
+      "learning_rate": 3.0093331269924954e-06,
+      "loss": 0.5307,
+      "step": 1041
+    },
+    {
+      "epoch": 2.00256,
+      "grad_norm": 0.5843510031700134,
+      "learning_rate": 2.9990750511688323e-06,
+      "loss": 0.2989,
+      "step": 1042
+    },
+    {
+      "epoch": 2.00448,
+      "grad_norm": 0.5885567665100098,
+      "learning_rate": 2.988826993693967e-06,
+      "loss": 0.4073,
+      "step": 1043
+    },
+    {
+      "epoch": 2.0064,
+      "grad_norm": 0.5879434943199158,
+      "learning_rate": 2.978589005878476e-06,
+      "loss": 0.3277,
+      "step": 1044
+    },
+    {
+      "epoch": 2.00832,
+      "grad_norm": 0.5412276983261108,
+      "learning_rate": 2.968361138982517e-06,
+      "loss": 0.3406,
+      "step": 1045
+    },
+    {
+      "epoch": 2.01024,
+      "grad_norm": 0.5053864121437073,
+      "learning_rate": 2.9581434442155798e-06,
+      "loss": 0.2773,
+      "step": 1046
+    },
+    {
+      "epoch": 2.01216,
+      "grad_norm": 0.5611642599105835,
+      "learning_rate": 2.947935972736217e-06,
+      "loss": 0.3021,
+      "step": 1047
+    },
+    {
+      "epoch": 2.01408,
+      "grad_norm": 0.549683690071106,
+      "learning_rate": 2.937738775651798e-06,
+      "loss": 0.3025,
+      "step": 1048
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 0.5979468822479248,
+      "learning_rate": 2.9275519040182503e-06,
+      "loss": 0.3206,
+      "step": 1049
+    },
+    {
+      "epoch": 2.01792,
+      "grad_norm": 0.5925368666648865,
+      "learning_rate": 2.917375408839803e-06,
+      "loss": 0.3204,
+      "step": 1050
+    },
+    {
+      "epoch": 2.01984,
+      "grad_norm": 0.5685588717460632,
+      "learning_rate": 2.9072093410687268e-06,
+      "loss": 0.2987,
+      "step": 1051
+    },
+    {
+      "epoch": 2.02176,
+      "grad_norm": 0.5929874181747437,
+      "learning_rate": 2.8970537516050935e-06,
+      "loss": 0.3404,
+      "step": 1052
+    },
+    {
+      "epoch": 2.02368,
+      "grad_norm": 0.5990848541259766,
+      "learning_rate": 2.886908691296504e-06,
+      "loss": 0.3201,
+      "step": 1053
+    },
+    {
+      "epoch": 2.0256,
+      "grad_norm": 0.6347576379776001,
+      "learning_rate": 2.876774210937843e-06,
+      "loss": 0.3118,
+      "step": 1054
+    },
+    {
+      "epoch": 2.02752,
+      "grad_norm": 0.5757642388343811,
+      "learning_rate": 2.866650361271023e-06,
+      "loss": 0.2951,
+      "step": 1055
+    },
+    {
+      "epoch": 2.02944,
+      "grad_norm": 0.628146767616272,
+      "learning_rate": 2.8565371929847286e-06,
+      "loss": 0.3099,
+      "step": 1056
+    },
+    {
+      "epoch": 2.03136,
+      "grad_norm": 0.5957508087158203,
+      "learning_rate": 2.84643475671417e-06,
+      "loss": 0.2827,
+      "step": 1057
+    },
+    {
+      "epoch": 2.03328,
+      "grad_norm": 0.530911386013031,
+      "learning_rate": 2.836343103040819e-06,
+      "loss": 0.2691,
+      "step": 1058
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 0.5892682671546936,
+      "learning_rate": 2.8262622824921593e-06,
+      "loss": 0.3144,
+      "step": 1059
+    },
+    {
+      "epoch": 2.03712,
+      "grad_norm": 0.6083415746688843,
+      "learning_rate": 2.816192345541437e-06,
+      "loss": 0.305,
+      "step": 1060
+    },
+    {
+      "epoch": 2.03904,
+      "grad_norm": 0.5969704389572144,
+      "learning_rate": 2.8061333426074054e-06,
+      "loss": 0.2854,
+      "step": 1061
+    },
+    {
+      "epoch": 2.04096,
+      "grad_norm": 0.6448900103569031,
+      "learning_rate": 2.7960853240540703e-06,
+      "loss": 0.3117,
+      "step": 1062
+    },
+    {
+      "epoch": 2.04288,
+      "grad_norm": 0.6071145534515381,
+      "learning_rate": 2.7860483401904483e-06,
+      "loss": 0.3129,
+      "step": 1063
+    },
+    {
+      "epoch": 2.0448,
+      "grad_norm": 0.6165956854820251,
+      "learning_rate": 2.776022441270295e-06,
+      "loss": 0.3664,
+      "step": 1064
+    },
+    {
+      "epoch": 2.04672,
+      "grad_norm": 0.5551939010620117,
+      "learning_rate": 2.766007677491871e-06,
+      "loss": 0.3056,
+      "step": 1065
+    },
+    {
+      "epoch": 2.04864,
+      "grad_norm": 0.6051166653633118,
+      "learning_rate": 2.7560040989976894e-06,
+      "loss": 0.3166,
+      "step": 1066
+    },
+    {
+      "epoch": 2.05056,
+      "grad_norm": 0.5648452639579773,
+      "learning_rate": 2.7460117558742532e-06,
+      "loss": 0.2875,
+      "step": 1067
+    },
+    {
+      "epoch": 2.05248,
+      "grad_norm": 0.5626317262649536,
+      "learning_rate": 2.736030698151815e-06,
+      "loss": 0.2987,
+      "step": 1068
+    },
+    {
+      "epoch": 2.0544,
+      "grad_norm": 0.6308444738388062,
+      "learning_rate": 2.72606097580412e-06,
+      "loss": 0.2995,
+      "step": 1069
+    },
+    {
+      "epoch": 2.05632,
+      "grad_norm": 0.5499259233474731,
+      "learning_rate": 2.7161026387481636e-06,
+      "loss": 0.2989,
+      "step": 1070
+    },
+    {
+      "epoch": 2.05824,
+      "grad_norm": 0.5939814448356628,
+      "learning_rate": 2.7061557368439294e-06,
+      "loss": 0.3352,
+      "step": 1071
+    },
+    {
+      "epoch": 2.06016,
+      "grad_norm": 0.5373671054840088,
+      "learning_rate": 2.6962203198941587e-06,
+      "loss": 0.2818,
+      "step": 1072
+    },
+    {
+      "epoch": 2.06208,
+      "grad_norm": 0.5360935926437378,
+      "learning_rate": 2.686296437644074e-06,
+      "loss": 0.307,
+      "step": 1073
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 0.547386884689331,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.3131,
+      "step": 1074
+    },
+    {
+      "epoch": 2.06592,
+      "grad_norm": 0.5415403842926025,
+      "learning_rate": 2.666483475934885e-06,
+      "loss": 0.2678,
+      "step": 1075
+    },
+    {
+      "epoch": 2.06784,
+      "grad_norm": 0.5443956851959229,
+      "learning_rate": 2.656594495676482e-06,
+      "loss": 0.2677,
+      "step": 1076
+    },
+    {
+      "epoch": 2.06976,
+      "grad_norm": 0.5876948237419128,
+      "learning_rate": 2.6467172485186775e-06,
+      "loss": 0.3059,
+      "step": 1077
+    },
+    {
+      "epoch": 2.07168,
+      "grad_norm": 0.5025241374969482,
+      "learning_rate": 2.636851783915454e-06,
+      "loss": 0.2773,
+      "step": 1078
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.5622445344924927,
+      "learning_rate": 2.626998151261798e-06,
+      "loss": 0.3347,
+      "step": 1079
+    },
+    {
+      "epoch": 2.07552,
+      "grad_norm": 0.584555983543396,
+      "learning_rate": 2.6171563998934605e-06,
+      "loss": 0.3284,
+      "step": 1080
+    },
+    {
+      "epoch": 2.07744,
+      "grad_norm": 0.5684719681739807,
+      "learning_rate": 2.607326579086701e-06,
+      "loss": 0.3396,
+      "step": 1081
+    },
+    {
+      "epoch": 2.07936,
+      "grad_norm": 0.5275523662567139,
+      "learning_rate": 2.5975087380580445e-06,
+      "loss": 0.2656,
+      "step": 1082
+    },
+    {
+      "epoch": 2.08128,
+      "grad_norm": 0.5579650402069092,
+      "learning_rate": 2.587702925964034e-06,
+      "loss": 0.3165,
+      "step": 1083
+    },
+    {
+      "epoch": 2.0832,
+      "grad_norm": 0.5256741642951965,
+      "learning_rate": 2.577909191900988e-06,
+      "loss": 0.2647,
+      "step": 1084
+    },
+    {
+      "epoch": 2.08512,
+      "grad_norm": 0.567886233329773,
+      "learning_rate": 2.5681275849047482e-06,
+      "loss": 0.3134,
+      "step": 1085
+    },
+    {
+      "epoch": 2.08704,
+      "grad_norm": 0.5466117858886719,
+      "learning_rate": 2.5583581539504464e-06,
+      "loss": 0.2788,
+      "step": 1086
+    },
+    {
+      "epoch": 2.08896,
+      "grad_norm": 0.5347439646720886,
+      "learning_rate": 2.5486009479522355e-06,
+      "loss": 0.2848,
+      "step": 1087
+    },
+    {
+      "epoch": 2.09088,
+      "grad_norm": 0.5583649277687073,
+      "learning_rate": 2.5388560157630765e-06,
+      "loss": 0.2895,
+      "step": 1088
+    },
+    {
+      "epoch": 2.0928,
+      "grad_norm": 0.6265268325805664,
+      "learning_rate": 2.5291234061744655e-06,
+      "loss": 0.3953,
+      "step": 1089
+    },
+    {
+      "epoch": 2.09472,
+      "grad_norm": 0.5785489082336426,
+      "learning_rate": 2.519403167916207e-06,
+      "loss": 0.319,
+      "step": 1090
+    },
+    {
+      "epoch": 2.09664,
+      "grad_norm": 0.5578694939613342,
+      "learning_rate": 2.50969534965616e-06,
+      "loss": 0.2964,
+      "step": 1091
+    },
+    {
+      "epoch": 2.09856,
+      "grad_norm": 0.5523391366004944,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.301,
+      "step": 1092
+    },
+    {
+      "epoch": 2.10048,
+      "grad_norm": 0.5072388052940369,
+      "learning_rate": 2.490317167490976e-06,
+      "loss": 0.2635,
+      "step": 1093
+    },
+    {
+      "epoch": 2.1024,
+      "grad_norm": 0.5640236139297485,
+      "learning_rate": 2.480646900609664e-06,
+      "loss": 0.2895,
+      "step": 1094
+    },
+    {
+      "epoch": 2.10432,
+      "grad_norm": 0.5470246076583862,
+      "learning_rate": 2.4709892477737263e-06,
+      "loss": 0.272,
+      "step": 1095
+    },
+    {
+      "epoch": 2.10624,
+      "grad_norm": 0.5707826614379883,
+      "learning_rate": 2.4613442573376625e-06,
+      "loss": 0.2935,
+      "step": 1096
+    },
+    {
+      "epoch": 2.10816,
+      "grad_norm": 0.6262672543525696,
+      "learning_rate": 2.4517119775925824e-06,
+      "loss": 0.3624,
+      "step": 1097
+    },
+    {
+      "epoch": 2.11008,
+      "grad_norm": 0.5375233888626099,
+      "learning_rate": 2.4420924567659508e-06,
+      "loss": 0.2352,
+      "step": 1098
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.6190659403800964,
+      "learning_rate": 2.4324857430213504e-06,
+      "loss": 0.3429,
+      "step": 1099
+    },
+    {
+      "epoch": 2.11392,
+      "grad_norm": 0.570245087146759,
+      "learning_rate": 2.422891884458241e-06,
+      "loss": 0.3384,
+      "step": 1100
+    },
+    {
+      "epoch": 2.11584,
+      "grad_norm": 0.5666949152946472,
+      "learning_rate": 2.4133109291117156e-06,
+      "loss": 0.3312,
+      "step": 1101
+    },
+    {
+      "epoch": 2.11776,
+      "grad_norm": 0.5367165803909302,
+      "learning_rate": 2.4037429249522702e-06,
+      "loss": 0.3288,
+      "step": 1102
+    },
+    {
+      "epoch": 2.11968,
+      "grad_norm": 0.5449112057685852,
+      "learning_rate": 2.394187919885548e-06,
+      "loss": 0.3009,
+      "step": 1103
+    },
+    {
+      "epoch": 2.1216,
+      "grad_norm": 0.5667607188224792,
+      "learning_rate": 2.384645961752113e-06,
+      "loss": 0.3301,
+      "step": 1104
+    },
+    {
+      "epoch": 2.12352,
+      "grad_norm": 0.5558924674987793,
+      "learning_rate": 2.3751170983272e-06,
+      "loss": 0.3332,
+      "step": 1105
+    },
+    {
+      "epoch": 2.12544,
+      "grad_norm": 0.541942298412323,
+      "learning_rate": 2.3656013773204843e-06,
+      "loss": 0.301,
+      "step": 1106
+    },
+    {
+      "epoch": 2.12736,
+      "grad_norm": 0.5786070823669434,
+      "learning_rate": 2.3560988463758366e-06,
+      "loss": 0.3362,
+      "step": 1107
+    },
+    {
+      "epoch": 2.12928,
+      "grad_norm": 0.591190755367279,
+      "learning_rate": 2.346609553071093e-06,
+      "loss": 0.3174,
+      "step": 1108
+    },
+    {
+      "epoch": 2.1312,
+      "grad_norm": 0.5622370839118958,
+      "learning_rate": 2.3371335449178006e-06,
+      "loss": 0.308,
+      "step": 1109
+    },
+    {
+      "epoch": 2.13312,
+      "grad_norm": 0.560448944568634,
+      "learning_rate": 2.3276708693609947e-06,
+      "loss": 0.3106,
+      "step": 1110
+    },
+    {
+      "epoch": 2.13504,
+      "grad_norm": 0.5368326306343079,
+      "learning_rate": 2.3182215737789593e-06,
+      "loss": 0.2654,
+      "step": 1111
+    },
+    {
+      "epoch": 2.13696,
+      "grad_norm": 0.5856943726539612,
+      "learning_rate": 2.308785705482982e-06,
+      "loss": 0.2983,
+      "step": 1112
+    },
+    {
+      "epoch": 2.13888,
+      "grad_norm": 0.5470758676528931,
+      "learning_rate": 2.2993633117171243e-06,
+      "loss": 0.3052,
+      "step": 1113
+    },
+    {
+      "epoch": 2.1408,
+      "grad_norm": 0.5670667886734009,
+      "learning_rate": 2.289954439657981e-06,
+      "loss": 0.2841,
+      "step": 1114
+    },
+    {
+      "epoch": 2.14272,
+      "grad_norm": 0.5341038107872009,
+      "learning_rate": 2.2805591364144446e-06,
+      "loss": 0.2971,
+      "step": 1115
+    },
+    {
+      "epoch": 2.14464,
+      "grad_norm": 0.5586668252944946,
+      "learning_rate": 2.2711774490274767e-06,
+      "loss": 0.2741,
+      "step": 1116
+    },
+    {
+      "epoch": 2.14656,
+      "grad_norm": 0.5637398958206177,
+      "learning_rate": 2.2618094244698614e-06,
+      "loss": 0.3038,
+      "step": 1117
+    },
+    {
+      "epoch": 2.14848,
+      "grad_norm": 0.5505224466323853,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 0.2791,
+      "step": 1118
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.5483053922653198,
+      "learning_rate": 2.243114551391542e-06,
+      "loss": 0.303,
+      "step": 1119
+    },
+    {
+      "epoch": 2.15232,
+      "grad_norm": 0.5604549646377563,
+      "learning_rate": 2.2337877964734324e-06,
+      "loss": 0.2734,
+      "step": 1120
+    },
+    {
+      "epoch": 2.15424,
+      "grad_norm": 0.5676697492599487,
+      "learning_rate": 2.224474891589386e-06,
+      "loss": 0.2874,
+      "step": 1121
+    },
+    {
+      "epoch": 2.15616,
+      "grad_norm": 0.5403993725776672,
+      "learning_rate": 2.2151758833678044e-06,
+      "loss": 0.2827,
+      "step": 1122
+    },
+    {
+      "epoch": 2.15808,
+      "grad_norm": 0.5599153637886047,
+      "learning_rate": 2.205890818367508e-06,
+      "loss": 0.2937,
+      "step": 1123
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.5756741762161255,
+      "learning_rate": 2.1966197430775056e-06,
+      "loss": 0.3003,
+      "step": 1124
+    },
+    {
+      "epoch": 2.16192,
+      "grad_norm": 0.6014906167984009,
+      "learning_rate": 2.187362703916766e-06,
+      "loss": 0.314,
+      "step": 1125
+    },
+    {
+      "epoch": 2.16384,
+      "grad_norm": 0.554108738899231,
+      "learning_rate": 2.178119747233976e-06,
+      "loss": 0.3037,
+      "step": 1126
+    },
+    {
+      "epoch": 2.16576,
+      "grad_norm": 0.5447950959205627,
+      "learning_rate": 2.168890919307315e-06,
+      "loss": 0.2694,
+      "step": 1127
+    },
+    {
+      "epoch": 2.16768,
+      "grad_norm": 0.5252242684364319,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.2592,
+      "step": 1128
+    },
+    {
+      "epoch": 2.1696,
+      "grad_norm": 0.5808660984039307,
+      "learning_rate": 2.1504758344811615e-06,
+      "loss": 0.3366,
+      "step": 1129
+    },
+    {
+      "epoch": 2.17152,
+      "grad_norm": 0.5725222229957581,
+      "learning_rate": 2.141289669783401e-06,
+      "loss": 0.3124,
+      "step": 1130
+    },
+    {
+      "epoch": 2.17344,
+      "grad_norm": 0.5473577380180359,
+      "learning_rate": 2.132117818244771e-06,
+      "loss": 0.2737,
+      "step": 1131
+    },
+    {
+      "epoch": 2.17536,
+      "grad_norm": 0.5586757659912109,
+      "learning_rate": 2.122960325787432e-06,
+      "loss": 0.3268,
+      "step": 1132
+    },
+    {
+      "epoch": 2.17728,
+      "grad_norm": 0.5499918460845947,
+      "learning_rate": 2.1138172382616612e-06,
+      "loss": 0.3109,
+      "step": 1133
+    },
+    {
+      "epoch": 2.1792,
+      "grad_norm": 0.5620728135108948,
+      "learning_rate": 2.104688601445606e-06,
+      "loss": 0.3129,
+      "step": 1134
+    },
+    {
+      "epoch": 2.18112,
+      "grad_norm": 0.6615064740180969,
+      "learning_rate": 2.0955744610450616e-06,
+      "loss": 0.3599,
+      "step": 1135
+    },
+    {
+      "epoch": 2.18304,
+      "grad_norm": 0.5560691356658936,
+      "learning_rate": 2.086474862693244e-06,
+      "loss": 0.3385,
+      "step": 1136
+    },
+    {
+      "epoch": 2.1849600000000002,
+      "grad_norm": 0.5569999814033508,
+      "learning_rate": 2.077389851950557e-06,
+      "loss": 0.2885,
+      "step": 1137
+    },
+    {
+      "epoch": 2.18688,
+      "grad_norm": 0.532078742980957,
+      "learning_rate": 2.068319474304365e-06,
+      "loss": 0.2873,
+      "step": 1138
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.593733012676239,
+      "learning_rate": 2.059263775168773e-06,
+      "loss": 0.3387,
+      "step": 1139
+    },
+    {
+      "epoch": 2.19072,
+      "grad_norm": 0.5717750787734985,
+      "learning_rate": 2.050222799884387e-06,
+      "loss": 0.2915,
+      "step": 1140
+    },
+    {
+      "epoch": 2.19264,
+      "grad_norm": 0.5853467583656311,
+      "learning_rate": 2.0411965937180907e-06,
+      "loss": 0.3437,
+      "step": 1141
+    },
+    {
+      "epoch": 2.19456,
+      "grad_norm": 0.6035574078559875,
+      "learning_rate": 2.0321852018628278e-06,
+      "loss": 0.3228,
+      "step": 1142
+    },
+    {
+      "epoch": 2.19648,
+      "grad_norm": 0.5406397581100464,
+      "learning_rate": 2.0231886694373653e-06,
+      "loss": 0.2491,
+      "step": 1143
+    },
+    {
+      "epoch": 2.1984,
+      "grad_norm": 0.5605498552322388,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.3129,
+      "step": 1144
+    },
+    {
+      "epoch": 2.20032,
+      "grad_norm": 0.519410252571106,
+      "learning_rate": 2.005240362978686e-06,
+      "loss": 0.2591,
+      "step": 1145
+    },
+    {
+      "epoch": 2.20224,
+      "grad_norm": 0.5592628717422485,
+      "learning_rate": 1.996288678810105e-06,
+      "loss": 0.3206,
+      "step": 1146
+    },
+    {
+      "epoch": 2.20416,
+      "grad_norm": 0.5862335562705994,
+      "learning_rate": 1.98735203380015e-06,
+      "loss": 0.2902,
+      "step": 1147
+    },
+    {
+      "epoch": 2.20608,
+      "grad_norm": 0.5700810551643372,
+      "learning_rate": 1.9784304726933384e-06,
+      "loss": 0.3203,
+      "step": 1148
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 0.5940973162651062,
+      "learning_rate": 1.9695240401586687e-06,
+      "loss": 0.3488,
+      "step": 1149
+    },
+    {
+      "epoch": 2.20992,
+      "grad_norm": 0.6147710084915161,
+      "learning_rate": 1.9606327807893905e-06,
+      "loss": 0.3516,
+      "step": 1150
+    },
+    {
+      "epoch": 2.21184,
+      "grad_norm": 0.5687532424926758,
+      "learning_rate": 1.9517567391027853e-06,
+      "loss": 0.3133,
+      "step": 1151
+    },
+    {
+      "epoch": 2.21376,
+      "grad_norm": 0.5686992406845093,
+      "learning_rate": 1.942895959539939e-06,
+      "loss": 0.298,
+      "step": 1152
+    },
+    {
+      "epoch": 2.21568,
+      "grad_norm": 0.5280912518501282,
+      "learning_rate": 1.934050486465529e-06,
+      "loss": 0.3018,
+      "step": 1153
+    },
+    {
+      "epoch": 2.2176,
+      "grad_norm": 0.5686296224594116,
+      "learning_rate": 1.9252203641675854e-06,
+      "loss": 0.2729,
+      "step": 1154
+    },
+    {
+      "epoch": 2.21952,
+      "grad_norm": 0.5466453433036804,
+      "learning_rate": 1.9164056368572847e-06,
+      "loss": 0.2803,
+      "step": 1155
+    },
+    {
+      "epoch": 2.22144,
+      "grad_norm": 0.6525296568870544,
+      "learning_rate": 1.9076063486687256e-06,
+      "loss": 0.3347,
+      "step": 1156
+    },
+    {
+      "epoch": 2.22336,
+      "grad_norm": 0.5216550230979919,
+      "learning_rate": 1.8988225436587005e-06,
+      "loss": 0.2691,
+      "step": 1157
+    },
+    {
+      "epoch": 2.22528,
+      "grad_norm": 0.5835052132606506,
+      "learning_rate": 1.8900542658064807e-06,
+      "loss": 0.3581,
+      "step": 1158
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.5276742577552795,
+      "learning_rate": 1.8813015590135963e-06,
+      "loss": 0.2692,
+      "step": 1159
+    },
+    {
+      "epoch": 2.22912,
+      "grad_norm": 0.6019517183303833,
+      "learning_rate": 1.8725644671036125e-06,
+      "loss": 0.3244,
+      "step": 1160
+    },
+    {
+      "epoch": 2.23104,
+      "grad_norm": 0.5910352468490601,
+      "learning_rate": 1.8638430338219199e-06,
+      "loss": 0.3394,
+      "step": 1161
+    },
+    {
+      "epoch": 2.23296,
+      "grad_norm": 0.5284815430641174,
+      "learning_rate": 1.8551373028355013e-06,
+      "loss": 0.279,
+      "step": 1162
+    },
+    {
+      "epoch": 2.23488,
+      "grad_norm": 0.5661656856536865,
+      "learning_rate": 1.846447317732723e-06,
+      "loss": 0.2875,
+      "step": 1163
+    },
+    {
+      "epoch": 2.2368,
+      "grad_norm": 0.5936484336853027,
+      "learning_rate": 1.8377731220231144e-06,
+      "loss": 0.3471,
+      "step": 1164
+    },
+    {
+      "epoch": 2.23872,
+      "grad_norm": 0.5249510407447815,
+      "learning_rate": 1.8291147591371482e-06,
+      "loss": 0.2981,
+      "step": 1165
+    },
+    {
+      "epoch": 2.24064,
+      "grad_norm": 0.6210512518882751,
+      "learning_rate": 1.8204722724260266e-06,
+      "loss": 0.3654,
+      "step": 1166
+    },
+    {
+      "epoch": 2.24256,
+      "grad_norm": 0.5676133632659912,
+      "learning_rate": 1.8118457051614591e-06,
+      "loss": 0.3042,
+      "step": 1167
+    },
+    {
+      "epoch": 2.24448,
+      "grad_norm": 0.561954140663147,
+      "learning_rate": 1.803235100535452e-06,
+      "loss": 0.2837,
+      "step": 1168
+    },
+    {
+      "epoch": 2.2464,
+      "grad_norm": 0.5541426539421082,
+      "learning_rate": 1.7946405016600843e-06,
+      "loss": 0.3168,
+      "step": 1169
+    },
+    {
+      "epoch": 2.24832,
+      "grad_norm": 0.5643410086631775,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.3094,
+      "step": 1170
+    },
+    {
+      "epoch": 2.25024,
+      "grad_norm": 0.5392529964447021,
+      "learning_rate": 1.7774994932086976e-06,
+      "loss": 0.2978,
+      "step": 1171
+    },
+    {
+      "epoch": 2.25216,
+      "grad_norm": 0.54856938123703,
+      "learning_rate": 1.7689531694552863e-06,
+      "loss": 0.2867,
+      "step": 1172
+    },
+    {
+      "epoch": 2.25408,
+      "grad_norm": 0.5615019798278809,
+      "learning_rate": 1.7604230230973068e-06,
+      "loss": 0.3405,
+      "step": 1173
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.5297316908836365,
+      "learning_rate": 1.7519090968439966e-06,
+      "loss": 0.2877,
+      "step": 1174
+    },
+    {
+      "epoch": 2.25792,
+      "grad_norm": 0.5461461544036865,
+      "learning_rate": 1.7434114333233853e-06,
+      "loss": 0.3004,
+      "step": 1175
+    },
+    {
+      "epoch": 2.25984,
+      "grad_norm": 0.5500417351722717,
+      "learning_rate": 1.7349300750820758e-06,
+      "loss": 0.2897,
+      "step": 1176
+    },
+    {
+      "epoch": 2.2617599999999998,
+      "grad_norm": 0.6003739833831787,
+      "learning_rate": 1.7264650645850256e-06,
+      "loss": 0.3605,
+      "step": 1177
+    },
+    {
+      "epoch": 2.26368,
+      "grad_norm": 0.5664680600166321,
+      "learning_rate": 1.7180164442153529e-06,
+      "loss": 0.3175,
+      "step": 1178
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.5552890300750732,
+      "learning_rate": 1.709584256274106e-06,
+      "loss": 0.3292,
+      "step": 1179
+    },
+    {
+      "epoch": 2.26752,
+      "grad_norm": 0.5299292206764221,
+      "learning_rate": 1.7011685429800596e-06,
+      "loss": 0.2767,
+      "step": 1180
+    },
+    {
+      "epoch": 2.26944,
+      "grad_norm": 0.4991382658481598,
+      "learning_rate": 1.6927693464695022e-06,
+      "loss": 0.2821,
+      "step": 1181
+    },
+    {
+      "epoch": 2.27136,
+      "grad_norm": 0.5633355379104614,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.2759,
+      "step": 1182
+    },
+    {
+      "epoch": 2.27328,
+      "grad_norm": 0.5878853797912598,
+      "learning_rate": 1.6760206719303107e-06,
+      "loss": 0.3583,
+      "step": 1183
+    },
+    {
+      "epoch": 2.2752,
+      "grad_norm": 0.5704030990600586,
+      "learning_rate": 1.6676712777599275e-06,
+      "loss": 0.3306,
+      "step": 1184
+    },
+    {
+      "epoch": 2.27712,
+      "grad_norm": 0.5675718188285828,
+      "learning_rate": 1.6593385680891139e-06,
+      "loss": 0.3369,
+      "step": 1185
+    },
+    {
+      "epoch": 2.27904,
+      "grad_norm": 0.5157074928283691,
+      "learning_rate": 1.6510225846385668e-06,
+      "loss": 0.2608,
+      "step": 1186
+    },
+    {
+      "epoch": 2.28096,
+      "grad_norm": 0.5530833005905151,
+      "learning_rate": 1.6427233690452455e-06,
+      "loss": 0.3333,
+      "step": 1187
+    },
+    {
+      "epoch": 2.28288,
+      "grad_norm": 0.5693645477294922,
+      "learning_rate": 1.6344409628621482e-06,
+      "loss": 0.3184,
+      "step": 1188
+    },
+    {
+      "epoch": 2.2848,
+      "grad_norm": 0.49818363785743713,
+      "learning_rate": 1.6261754075581187e-06,
+      "loss": 0.249,
+      "step": 1189
+    },
+    {
+      "epoch": 2.28672,
+      "grad_norm": 0.5964416861534119,
+      "learning_rate": 1.6179267445176206e-06,
+      "loss": 0.3054,
+      "step": 1190
+    },
+    {
+      "epoch": 2.28864,
+      "grad_norm": 0.5627337098121643,
+      "learning_rate": 1.6096950150405454e-06,
+      "loss": 0.3037,
+      "step": 1191
+    },
+    {
+      "epoch": 2.29056,
+      "grad_norm": 0.5663759112358093,
+      "learning_rate": 1.6014802603420044e-06,
+      "loss": 0.3291,
+      "step": 1192
+    },
+    {
+      "epoch": 2.29248,
+      "grad_norm": 0.597136914730072,
+      "learning_rate": 1.593282521552113e-06,
+      "loss": 0.3486,
+      "step": 1193
+    },
+    {
+      "epoch": 2.2944,
+      "grad_norm": 0.5951170325279236,
+      "learning_rate": 1.5851018397157918e-06,
+      "loss": 0.3337,
+      "step": 1194
+    },
+    {
+      "epoch": 2.29632,
+      "grad_norm": 0.5457406640052795,
+      "learning_rate": 1.5769382557925612e-06,
+      "loss": 0.2923,
+      "step": 1195
+    },
+    {
+      "epoch": 2.29824,
+      "grad_norm": 0.5755424499511719,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.3415,
+      "step": 1196
+    },
+    {
+      "epoch": 2.30016,
+      "grad_norm": 0.5980664491653442,
+      "learning_rate": 1.5606625450952062e-06,
+      "loss": 0.2973,
+      "step": 1197
+    },
+    {
+      "epoch": 2.30208,
+      "grad_norm": 0.6022819876670837,
+      "learning_rate": 1.5525504998112717e-06,
+      "loss": 0.3344,
+      "step": 1198
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.57796311378479,
+      "learning_rate": 1.5444557154203892e-06,
+      "loss": 0.3033,
+      "step": 1199
+    },
+    {
+      "epoch": 2.30592,
+      "grad_norm": 0.598685085773468,
+      "learning_rate": 1.5363782324520033e-06,
+      "loss": 0.3442,
+      "step": 1200
+    },
+    {
+      "epoch": 2.30784,
+      "grad_norm": 0.5483145713806152,
+      "learning_rate": 1.5283180913489326e-06,
+      "loss": 0.3017,
+      "step": 1201
+    },
+    {
+      "epoch": 2.30976,
+      "grad_norm": 0.5511487126350403,
+      "learning_rate": 1.520275332467166e-06,
+      "loss": 0.2883,
+      "step": 1202
+    },
+    {
+      "epoch": 2.31168,
+      "grad_norm": 0.5519542098045349,
+      "learning_rate": 1.5122499960756604e-06,
+      "loss": 0.2986,
+      "step": 1203
+    },
+    {
+      "epoch": 2.3136,
+      "grad_norm": 0.566226065158844,
+      "learning_rate": 1.504242122356143e-06,
+      "loss": 0.2869,
+      "step": 1204
+    },
+    {
+      "epoch": 2.3155200000000002,
+      "grad_norm": 0.5418808460235596,
+      "learning_rate": 1.4962517514029069e-06,
+      "loss": 0.2822,
+      "step": 1205
+    },
+    {
+      "epoch": 2.31744,
+      "grad_norm": 0.5311360359191895,
+      "learning_rate": 1.4882789232226124e-06,
+      "loss": 0.2492,
+      "step": 1206
+    },
+    {
+      "epoch": 2.31936,
+      "grad_norm": 0.576831042766571,
+      "learning_rate": 1.4803236777340856e-06,
+      "loss": 0.3251,
+      "step": 1207
+    },
+    {
+      "epoch": 2.32128,
+      "grad_norm": 0.5796308517456055,
+      "learning_rate": 1.4723860547681163e-06,
+      "loss": 0.365,
+      "step": 1208
+    },
+    {
+      "epoch": 2.3232,
+      "grad_norm": 0.5780469179153442,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.3083,
+      "step": 1209
+    },
+    {
+      "epoch": 2.32512,
+      "grad_norm": 0.53952956199646,
+      "learning_rate": 1.4565638352856504e-06,
+      "loss": 0.2963,
+      "step": 1210
+    },
+    {
+      "epoch": 2.32704,
+      "grad_norm": 0.5655388832092285,
+      "learning_rate": 1.4486793179887738e-06,
+      "loss": 0.33,
+      "step": 1211
+    },
+    {
+      "epoch": 2.32896,
+      "grad_norm": 0.5814327597618103,
+      "learning_rate": 1.4408125816532981e-06,
+      "loss": 0.3156,
+      "step": 1212
+    },
+    {
+      "epoch": 2.33088,
+      "grad_norm": 0.5972678661346436,
+      "learning_rate": 1.4329636656668617e-06,
+      "loss": 0.3417,
+      "step": 1213
+    },
+    {
+      "epoch": 2.3327999999999998,
+      "grad_norm": 0.5857495665550232,
+      "learning_rate": 1.4251326093278773e-06,
+      "loss": 0.3218,
+      "step": 1214
+    },
+    {
+      "epoch": 2.33472,
+      "grad_norm": 0.5907280445098877,
+      "learning_rate": 1.4173194518453415e-06,
+      "loss": 0.3389,
+      "step": 1215
+    },
+    {
+      "epoch": 2.33664,
+      "grad_norm": 0.5681069493293762,
+      "learning_rate": 1.4095242323386305e-06,
+      "loss": 0.3073,
+      "step": 1216
+    },
+    {
+      "epoch": 2.33856,
+      "grad_norm": 0.5561332702636719,
+      "learning_rate": 1.4017469898373077e-06,
+      "loss": 0.2742,
+      "step": 1217
+    },
+    {
+      "epoch": 2.34048,
+      "grad_norm": 0.568260908126831,
+      "learning_rate": 1.3939877632809279e-06,
+      "loss": 0.29,
+      "step": 1218
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.564420223236084,
+      "learning_rate": 1.3862465915188427e-06,
+      "loss": 0.3178,
+      "step": 1219
+    },
+    {
+      "epoch": 2.34432,
+      "grad_norm": 0.5590744614601135,
+      "learning_rate": 1.3785235133100088e-06,
+      "loss": 0.3353,
+      "step": 1220
+    },
+    {
+      "epoch": 2.34624,
+      "grad_norm": 0.5926837921142578,
+      "learning_rate": 1.3708185673227896e-06,
+      "loss": 0.342,
+      "step": 1221
+    },
+    {
+      "epoch": 2.34816,
+      "grad_norm": 0.5307691693305969,
+      "learning_rate": 1.3631317921347564e-06,
+      "loss": 0.29,
+      "step": 1222
+    },
+    {
+      "epoch": 2.35008,
+      "grad_norm": 0.5744847655296326,
+      "learning_rate": 1.3554632262325129e-06,
+      "loss": 0.306,
+      "step": 1223
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 0.5999277830123901,
+      "learning_rate": 1.347812908011485e-06,
+      "loss": 0.2718,
+      "step": 1224
+    },
+    {
+      "epoch": 2.35392,
+      "grad_norm": 0.5368554592132568,
+      "learning_rate": 1.340180875775735e-06,
+      "loss": 0.2917,
+      "step": 1225
+    },
+    {
+      "epoch": 2.35584,
+      "grad_norm": 0.5271530747413635,
+      "learning_rate": 1.3325671677377727e-06,
+      "loss": 0.2615,
+      "step": 1226
+    },
+    {
+      "epoch": 2.35776,
+      "grad_norm": 0.5575255751609802,
+      "learning_rate": 1.3249718220183583e-06,
+      "loss": 0.2769,
+      "step": 1227
+    },
+    {
+      "epoch": 2.35968,
+      "grad_norm": 0.5836583375930786,
+      "learning_rate": 1.3173948766463146e-06,
+      "loss": 0.3179,
+      "step": 1228
+    },
+    {
+      "epoch": 2.3616,
+      "grad_norm": 0.5757988691329956,
+      "learning_rate": 1.309836369558341e-06,
+      "loss": 0.2981,
+      "step": 1229
+    },
+    {
+      "epoch": 2.36352,
+      "grad_norm": 0.5630492568016052,
+      "learning_rate": 1.3022963385988153e-06,
+      "loss": 0.2905,
+      "step": 1230
+    },
+    {
+      "epoch": 2.36544,
+      "grad_norm": 0.5522061586380005,
+      "learning_rate": 1.2947748215196038e-06,
+      "loss": 0.278,
+      "step": 1231
+    },
+    {
+      "epoch": 2.36736,
+      "grad_norm": 0.5176620483398438,
+      "learning_rate": 1.2872718559798852e-06,
+      "loss": 0.2652,
+      "step": 1232
+    },
+    {
+      "epoch": 2.36928,
+      "grad_norm": 0.5579131841659546,
+      "learning_rate": 1.2797874795459464e-06,
+      "loss": 0.3055,
+      "step": 1233
+    },
+    {
+      "epoch": 2.3712,
+      "grad_norm": 0.5547510981559753,
+      "learning_rate": 1.2723217296910078e-06,
+      "loss": 0.3107,
+      "step": 1234
+    },
+    {
+      "epoch": 2.37312,
+      "grad_norm": 0.6101371645927429,
+      "learning_rate": 1.264874643795021e-06,
+      "loss": 0.3437,
+      "step": 1235
+    },
+    {
+      "epoch": 2.37504,
+      "grad_norm": 0.5506192445755005,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.2667,
+      "step": 1236
+    },
+    {
+      "epoch": 2.37696,
+      "grad_norm": 0.5524115562438965,
+      "learning_rate": 1.2500366129323039e-06,
+      "loss": 0.3026,
+      "step": 1237
+    },
+    {
+      "epoch": 2.37888,
+      "grad_norm": 0.60694420337677,
+      "learning_rate": 1.2426457422575e-06,
+      "loss": 0.3045,
+      "step": 1238
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.5660543441772461,
+      "learning_rate": 1.2352736841251306e-06,
+      "loss": 0.3062,
+      "step": 1239
+    },
+    {
+      "epoch": 2.38272,
+      "grad_norm": 0.5220528841018677,
+      "learning_rate": 1.2279204754460494e-06,
+      "loss": 0.2931,
+      "step": 1240
+    },
+    {
+      "epoch": 2.38464,
+      "grad_norm": 0.5629206895828247,
+      "learning_rate": 1.2205861530367342e-06,
+      "loss": 0.3265,
+      "step": 1241
+    },
+    {
+      "epoch": 2.3865600000000002,
+      "grad_norm": 0.5631869435310364,
+      "learning_rate": 1.2132707536191008e-06,
+      "loss": 0.2883,
+      "step": 1242
+    },
+    {
+      "epoch": 2.38848,
+      "grad_norm": 0.613427996635437,
+      "learning_rate": 1.2059743138203256e-06,
+      "loss": 0.2855,
+      "step": 1243
+    },
+    {
+      "epoch": 2.3904,
+      "grad_norm": 0.6336319446563721,
+      "learning_rate": 1.1986968701726492e-06,
+      "loss": 0.3225,
+      "step": 1244
+    },
+    {
+      "epoch": 2.39232,
+      "grad_norm": 0.5840649008750916,
+      "learning_rate": 1.1914384591132045e-06,
+      "loss": 0.3331,
+      "step": 1245
+    },
+    {
+      "epoch": 2.39424,
+      "grad_norm": 0.562660813331604,
+      "learning_rate": 1.184199116983834e-06,
+      "loss": 0.3102,
+      "step": 1246
+    },
+    {
+      "epoch": 2.39616,
+      "grad_norm": 0.5466341972351074,
+      "learning_rate": 1.1769788800309001e-06,
+      "loss": 0.2743,
+      "step": 1247
+    },
+    {
+      "epoch": 2.39808,
+      "grad_norm": 0.5944238901138306,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.3689,
+      "step": 1248
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.5709063410758972,
+      "learning_rate": 1.1625958661613345e-06,
+      "loss": 0.2983,
+      "step": 1249
+    },
+    {
+      "epoch": 2.40192,
+      "grad_norm": 0.5517197251319885,
+      "learning_rate": 1.1554331612584218e-06,
+      "loss": 0.3011,
+      "step": 1250
+    },
+    {
+      "epoch": 2.4038399999999998,
+      "grad_norm": 0.5444364547729492,
+      "learning_rate": 1.1482897055590275e-06,
+      "loss": 0.2651,
+      "step": 1251
+    },
+    {
+      "epoch": 2.40576,
+      "grad_norm": 0.52352374792099,
+      "learning_rate": 1.141165534829425e-06,
+      "loss": 0.2871,
+      "step": 1252
+    },
+    {
+      "epoch": 2.40768,
+      "grad_norm": 0.5225202441215515,
+      "learning_rate": 1.134060684739331e-06,
+      "loss": 0.2838,
+      "step": 1253
+    },
+    {
+      "epoch": 2.4096,
+      "grad_norm": 0.5135831236839294,
+      "learning_rate": 1.1269751908617277e-06,
+      "loss": 0.2898,
+      "step": 1254
+    },
+    {
+      "epoch": 2.41152,
+      "grad_norm": 0.5847892165184021,
+      "learning_rate": 1.119909088672682e-06,
+      "loss": 0.2934,
+      "step": 1255
+    },
+    {
+      "epoch": 2.41344,
+      "grad_norm": 0.591886579990387,
+      "learning_rate": 1.1128624135511712e-06,
+      "loss": 0.3038,
+      "step": 1256
+    },
+    {
+      "epoch": 2.41536,
+      "grad_norm": 0.5736411213874817,
+      "learning_rate": 1.105835200778902e-06,
+      "loss": 0.3015,
+      "step": 1257
+    },
+    {
+      "epoch": 2.41728,
+      "grad_norm": 0.5322389006614685,
+      "learning_rate": 1.0988274855401377e-06,
+      "loss": 0.2764,
+      "step": 1258
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.54026859998703,
+      "learning_rate": 1.091839302921518e-06,
+      "loss": 0.2604,
+      "step": 1259
+    },
+    {
+      "epoch": 2.42112,
+      "grad_norm": 0.5842394828796387,
+      "learning_rate": 1.0848706879118893e-06,
+      "loss": 0.3151,
+      "step": 1260
+    },
+    {
+      "epoch": 2.42304,
+      "grad_norm": 0.5830526351928711,
+      "learning_rate": 1.0779216754021215e-06,
+      "loss": 0.2921,
+      "step": 1261
+    },
+    {
+      "epoch": 2.42496,
+      "grad_norm": 0.5301140546798706,
+      "learning_rate": 1.070992300184941e-06,
+      "loss": 0.2815,
+      "step": 1262
+    },
+    {
+      "epoch": 2.42688,
+      "grad_norm": 0.5471583604812622,
+      "learning_rate": 1.0640825969547498e-06,
+      "loss": 0.3205,
+      "step": 1263
+    },
+    {
+      "epoch": 2.4288,
+      "grad_norm": 0.6082338094711304,
+      "learning_rate": 1.057192600307456e-06,
+      "loss": 0.2917,
+      "step": 1264
+    },
+    {
+      "epoch": 2.43072,
+      "grad_norm": 0.5065282583236694,
+      "learning_rate": 1.0503223447403033e-06,
+      "loss": 0.2649,
+      "step": 1265
+    },
+    {
+      "epoch": 2.43264,
+      "grad_norm": 0.5389804840087891,
+      "learning_rate": 1.0434718646516917e-06,
+      "loss": 0.2748,
+      "step": 1266
+    },
+    {
+      "epoch": 2.43456,
+      "grad_norm": 0.5444163680076599,
+      "learning_rate": 1.0366411943410033e-06,
+      "loss": 0.2953,
+      "step": 1267
+    },
+    {
+      "epoch": 2.43648,
+      "grad_norm": 0.5636538863182068,
+      "learning_rate": 1.0298303680084448e-06,
+      "loss": 0.2866,
+      "step": 1268
+    },
+    {
+      "epoch": 2.4384,
+      "grad_norm": 0.5241268873214722,
+      "learning_rate": 1.0230394197548605e-06,
+      "loss": 0.2714,
+      "step": 1269
+    },
+    {
+      "epoch": 2.44032,
+      "grad_norm": 0.5515684485435486,
+      "learning_rate": 1.0162683835815706e-06,
+      "loss": 0.3063,
+      "step": 1270
+    },
+    {
+      "epoch": 2.44224,
+      "grad_norm": 0.5493499040603638,
+      "learning_rate": 1.009517293390197e-06,
+      "loss": 0.3097,
+      "step": 1271
+    },
+    {
+      "epoch": 2.44416,
+      "grad_norm": 0.5881457328796387,
+      "learning_rate": 1.0027861829824953e-06,
+      "loss": 0.3584,
+      "step": 1272
+    },
+    {
+      "epoch": 2.44608,
+      "grad_norm": 0.573180079460144,
+      "learning_rate": 9.960750860601842e-07,
+      "loss": 0.3059,
+      "step": 1273
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 0.5450122952461243,
+      "learning_rate": 9.893840362247809e-07,
+      "loss": 0.2587,
+      "step": 1274
+    },
+    {
+      "epoch": 2.44992,
+      "grad_norm": 0.5546673536300659,
+      "learning_rate": 9.82713066977427e-07,
+      "loss": 0.2928,
+      "step": 1275
+    },
+    {
+      "epoch": 2.45184,
+      "grad_norm": 0.6369803547859192,
+      "learning_rate": 9.760622117187234e-07,
+      "loss": 0.357,
+      "step": 1276
+    },
+    {
+      "epoch": 2.45376,
+      "grad_norm": 0.5507545471191406,
+      "learning_rate": 9.694315037485635e-07,
+      "loss": 0.2934,
+      "step": 1277
+    },
+    {
+      "epoch": 2.45568,
+      "grad_norm": 0.5355435609817505,
+      "learning_rate": 9.628209762659658e-07,
+      "loss": 0.3026,
+      "step": 1278
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.5143349766731262,
+      "learning_rate": 9.562306623689111e-07,
+      "loss": 0.2862,
+      "step": 1279
+    },
+    {
+      "epoch": 2.45952,
+      "grad_norm": 0.5520386099815369,
+      "learning_rate": 9.496605950541676e-07,
+      "loss": 0.3306,
+      "step": 1280
+    },
+    {
+      "epoch": 2.46144,
+      "grad_norm": 0.5393406748771667,
+      "learning_rate": 9.431108072171346e-07,
+      "loss": 0.3069,
+      "step": 1281
+    },
+    {
+      "epoch": 2.4633599999999998,
+      "grad_norm": 0.5416886210441589,
+      "learning_rate": 9.365813316516787e-07,
+      "loss": 0.3056,
+      "step": 1282
+    },
+    {
+      "epoch": 2.46528,
+      "grad_norm": 0.5515227913856506,
+      "learning_rate": 9.300722010499608e-07,
+      "loss": 0.2974,
+      "step": 1283
+    },
+    {
+      "epoch": 2.4672,
+      "grad_norm": 0.5814557075500488,
+      "learning_rate": 9.235834480022788e-07,
+      "loss": 0.3014,
+      "step": 1284
+    },
+    {
+      "epoch": 2.46912,
+      "grad_norm": 0.5457692742347717,
+      "learning_rate": 9.17115104996903e-07,
+      "loss": 0.3207,
+      "step": 1285
+    },
+    {
+      "epoch": 2.47104,
+      "grad_norm": 0.5245400667190552,
+      "learning_rate": 9.10667204419915e-07,
+      "loss": 0.2763,
+      "step": 1286
+    },
+    {
+      "epoch": 2.47296,
+      "grad_norm": 0.5794905424118042,
+      "learning_rate": 9.042397785550405e-07,
+      "loss": 0.309,
+      "step": 1287
+    },
+    {
+      "epoch": 2.47488,
+      "grad_norm": 0.595958948135376,
+      "learning_rate": 8.978328595834984e-07,
+      "loss": 0.307,
+      "step": 1288
+    },
+    {
+      "epoch": 2.4768,
+      "grad_norm": 0.6133590936660767,
+      "learning_rate": 8.91446479583823e-07,
+      "loss": 0.3156,
+      "step": 1289
+    },
+    {
+      "epoch": 2.47872,
+      "grad_norm": 0.5469968914985657,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.2899,
+      "step": 1290
+    },
+    {
+      "epoch": 2.48064,
+      "grad_norm": 0.5700621008872986,
+      "learning_rate": 8.787354642998936e-07,
+      "loss": 0.2974,
+      "step": 1291
+    },
+    {
+      "epoch": 2.48256,
+      "grad_norm": 0.5310498476028442,
+      "learning_rate": 8.724108926579e-07,
+      "loss": 0.2734,
+      "step": 1292
+    },
+    {
+      "epoch": 2.48448,
+      "grad_norm": 0.5221157073974609,
+      "learning_rate": 8.661069872719746e-07,
+      "loss": 0.2966,
+      "step": 1293
+    },
+    {
+      "epoch": 2.4864,
+      "grad_norm": 0.5556731224060059,
+      "learning_rate": 8.598237797048825e-07,
+      "loss": 0.3333,
+      "step": 1294
+    },
+    {
+      "epoch": 2.48832,
+      "grad_norm": 0.5634422898292542,
+      "learning_rate": 8.535613014157556e-07,
+      "loss": 0.2937,
+      "step": 1295
+    },
+    {
+      "epoch": 2.49024,
+      "grad_norm": 0.5687916278839111,
+      "learning_rate": 8.473195837599419e-07,
+      "loss": 0.321,
+      "step": 1296
+    },
+    {
+      "epoch": 2.49216,
+      "grad_norm": 0.5855703353881836,
+      "learning_rate": 8.410986579888381e-07,
+      "loss": 0.2951,
+      "step": 1297
+    },
+    {
+      "epoch": 2.49408,
+      "grad_norm": 0.5657078623771667,
+      "learning_rate": 8.348985552497424e-07,
+      "loss": 0.299,
+      "step": 1298
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.5629979968070984,
+      "learning_rate": 8.287193065856936e-07,
+      "loss": 0.3089,
+      "step": 1299
+    },
+    {
+      "epoch": 2.49792,
+      "grad_norm": 0.5918774604797363,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.3437,
+      "step": 1300
+    },
+    {
+      "epoch": 2.49984,
+      "grad_norm": 0.5529206991195679,
+      "learning_rate": 8.164234951326727e-07,
+      "loss": 0.2929,
+      "step": 1301
+    },
+    {
+      "epoch": 2.50176,
+      "grad_norm": 0.5409730672836304,
+      "learning_rate": 8.103069939070945e-07,
+      "loss": 0.297,
+      "step": 1302
+    },
+    {
+      "epoch": 2.50368,
+      "grad_norm": 0.5309858322143555,
+      "learning_rate": 8.042114698830394e-07,
+      "loss": 0.2628,
+      "step": 1303
+    },
+    {
+      "epoch": 2.5056000000000003,
+      "grad_norm": 0.5169317126274109,
+      "learning_rate": 7.981369535799354e-07,
+      "loss": 0.2853,
+      "step": 1304
+    },
+    {
+      "epoch": 2.50752,
+      "grad_norm": 0.5260899662971497,
+      "learning_rate": 7.920834754120305e-07,
+      "loss": 0.272,
+      "step": 1305
+    },
+    {
+      "epoch": 2.50944,
+      "grad_norm": 0.545064389705658,
+      "learning_rate": 7.860510656882342e-07,
+      "loss": 0.3035,
+      "step": 1306
+    },
+    {
+      "epoch": 2.51136,
+      "grad_norm": 0.5196812748908997,
+      "learning_rate": 7.800397546119709e-07,
+      "loss": 0.2648,
+      "step": 1307
+    },
+    {
+      "epoch": 2.51328,
+      "grad_norm": 0.5809112787246704,
+      "learning_rate": 7.740495722810271e-07,
+      "loss": 0.3005,
+      "step": 1308
+    },
+    {
+      "epoch": 2.5152,
+      "grad_norm": 0.5616962313652039,
+      "learning_rate": 7.680805486873977e-07,
+      "loss": 0.3149,
+      "step": 1309
+    },
+    {
+      "epoch": 2.5171200000000002,
+      "grad_norm": 0.5568909049034119,
+      "learning_rate": 7.621327137171447e-07,
+      "loss": 0.3028,
+      "step": 1310
+    },
+    {
+      "epoch": 2.51904,
+      "grad_norm": 0.582168698310852,
+      "learning_rate": 7.562060971502383e-07,
+      "loss": 0.306,
+      "step": 1311
+    },
+    {
+      "epoch": 2.52096,
+      "grad_norm": 0.6206565499305725,
+      "learning_rate": 7.50300728660407e-07,
+      "loss": 0.3553,
+      "step": 1312
+    },
+    {
+      "epoch": 2.52288,
+      "grad_norm": 0.5818870663642883,
+      "learning_rate": 7.444166378150014e-07,
+      "loss": 0.3395,
+      "step": 1313
+    },
+    {
+      "epoch": 2.5248,
+      "grad_norm": 0.5885866284370422,
+      "learning_rate": 7.385538540748327e-07,
+      "loss": 0.3283,
+      "step": 1314
+    },
+    {
+      "epoch": 2.52672,
+      "grad_norm": 0.5720000267028809,
+      "learning_rate": 7.327124067940311e-07,
+      "loss": 0.2899,
+      "step": 1315
+    },
+    {
+      "epoch": 2.52864,
+      "grad_norm": 0.5514774918556213,
+      "learning_rate": 7.26892325219899e-07,
+      "loss": 0.2933,
+      "step": 1316
+    },
+    {
+      "epoch": 2.53056,
+      "grad_norm": 0.5508315563201904,
+      "learning_rate": 7.210936384927631e-07,
+      "loss": 0.3152,
+      "step": 1317
+    },
+    {
+      "epoch": 2.53248,
+      "grad_norm": 0.5942166447639465,
+      "learning_rate": 7.153163756458287e-07,
+      "loss": 0.3218,
+      "step": 1318
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.54239821434021,
+      "learning_rate": 7.09560565605037e-07,
+      "loss": 0.2996,
+      "step": 1319
+    },
+    {
+      "epoch": 2.53632,
+      "grad_norm": 0.5864148139953613,
+      "learning_rate": 7.03826237188916e-07,
+      "loss": 0.3259,
+      "step": 1320
+    },
+    {
+      "epoch": 2.53824,
+      "grad_norm": 0.5506216883659363,
+      "learning_rate": 6.981134191084388e-07,
+      "loss": 0.2823,
+      "step": 1321
+    },
+    {
+      "epoch": 2.54016,
+      "grad_norm": 0.5380348563194275,
+      "learning_rate": 6.924221399668785e-07,
+      "loss": 0.2729,
+      "step": 1322
+    },
+    {
+      "epoch": 2.54208,
+      "grad_norm": 0.562001645565033,
+      "learning_rate": 6.867524282596655e-07,
+      "loss": 0.3061,
+      "step": 1323
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.5345205664634705,
+      "learning_rate": 6.811043123742494e-07,
+      "loss": 0.2974,
+      "step": 1324
+    },
+    {
+      "epoch": 2.5459199999999997,
+      "grad_norm": 0.5839608311653137,
+      "learning_rate": 6.754778205899465e-07,
+      "loss": 0.3288,
+      "step": 1325
+    },
+    {
+      "epoch": 2.54784,
+      "grad_norm": 0.5835815072059631,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.3491,
+      "step": 1326
+    },
+    {
+      "epoch": 2.54976,
+      "grad_norm": 0.5368748903274536,
+      "learning_rate": 6.642898219004723e-07,
+      "loss": 0.303,
+      "step": 1327
+    },
+    {
+      "epoch": 2.55168,
+      "grad_norm": 0.5371608138084412,
+      "learning_rate": 6.587283710120324e-07,
+      "loss": 0.3057,
+      "step": 1328
+    },
+    {
+      "epoch": 2.5536,
+      "grad_norm": 0.5403703451156616,
+      "learning_rate": 6.531886562578859e-07,
+      "loss": 0.2867,
+      "step": 1329
+    },
+    {
+      "epoch": 2.55552,
+      "grad_norm": 0.5724359154701233,
+      "learning_rate": 6.47670705374604e-07,
+      "loss": 0.3143,
+      "step": 1330
+    },
+    {
+      "epoch": 2.55744,
+      "grad_norm": 0.5384829640388489,
+      "learning_rate": 6.421745459897871e-07,
+      "loss": 0.2636,
+      "step": 1331
+    },
+    {
+      "epoch": 2.55936,
+      "grad_norm": 0.5424496531486511,
+      "learning_rate": 6.367002056219285e-07,
+      "loss": 0.2722,
+      "step": 1332
+    },
+    {
+      "epoch": 2.56128,
+      "grad_norm": 0.6021274328231812,
+      "learning_rate": 6.312477116802807e-07,
+      "loss": 0.3291,
+      "step": 1333
+    },
+    {
+      "epoch": 2.5632,
+      "grad_norm": 0.5530632138252258,
+      "learning_rate": 6.258170914647077e-07,
+      "loss": 0.3121,
+      "step": 1334
+    },
+    {
+      "epoch": 2.56512,
+      "grad_norm": 0.5618712902069092,
+      "learning_rate": 6.204083721655607e-07,
+      "loss": 0.3453,
+      "step": 1335
+    },
+    {
+      "epoch": 2.56704,
+      "grad_norm": 0.5514795184135437,
+      "learning_rate": 6.150215808635334e-07,
+      "loss": 0.2736,
+      "step": 1336
+    },
+    {
+      "epoch": 2.56896,
+      "grad_norm": 0.5560357570648193,
+      "learning_rate": 6.096567445295298e-07,
+      "loss": 0.3054,
+      "step": 1337
+    },
+    {
+      "epoch": 2.57088,
+      "grad_norm": 0.5912596583366394,
+      "learning_rate": 6.043138900245277e-07,
+      "loss": 0.3397,
+      "step": 1338
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.556064248085022,
+      "learning_rate": 5.989930440994451e-07,
+      "loss": 0.282,
+      "step": 1339
+    },
+    {
+      "epoch": 2.57472,
+      "grad_norm": 0.5416197180747986,
+      "learning_rate": 5.936942333950063e-07,
+      "loss": 0.2952,
+      "step": 1340
+    },
+    {
+      "epoch": 2.5766400000000003,
+      "grad_norm": 0.5501132011413574,
+      "learning_rate": 5.884174844416102e-07,
+      "loss": 0.2909,
+      "step": 1341
+    },
+    {
+      "epoch": 2.57856,
+      "grad_norm": 0.5210188031196594,
+      "learning_rate": 5.831628236591929e-07,
+      "loss": 0.2718,
+      "step": 1342
+    },
+    {
+      "epoch": 2.58048,
+      "grad_norm": 0.536419689655304,
+      "learning_rate": 5.779302773570994e-07,
+      "loss": 0.2935,
+      "step": 1343
+    },
+    {
+      "epoch": 2.5824,
+      "grad_norm": 0.5744022130966187,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.3049,
+      "step": 1344
+    },
+    {
+      "epoch": 2.58432,
+      "grad_norm": 0.5372785329818726,
+      "learning_rate": 5.675316328775126e-07,
+      "loss": 0.2787,
+      "step": 1345
+    },
+    {
+      "epoch": 2.58624,
+      "grad_norm": 0.5580140948295593,
+      "learning_rate": 5.623655867645628e-07,
+      "loss": 0.2974,
+      "step": 1346
+    },
+    {
+      "epoch": 2.5881600000000002,
+      "grad_norm": 0.5411311984062195,
+      "learning_rate": 5.572217592607687e-07,
+      "loss": 0.3141,
+      "step": 1347
+    },
+    {
+      "epoch": 2.59008,
+      "grad_norm": 0.5491883158683777,
+      "learning_rate": 5.521001761205441e-07,
+      "loss": 0.2976,
+      "step": 1348
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 0.5661610960960388,
+      "learning_rate": 5.470008629869367e-07,
+      "loss": 0.2997,
+      "step": 1349
+    },
+    {
+      "epoch": 2.59392,
+      "grad_norm": 0.8253101706504822,
+      "learning_rate": 5.41923845391486e-07,
+      "loss": 0.4359,
+      "step": 1350
+    },
+    {
+      "epoch": 2.59584,
+      "grad_norm": 0.5413922667503357,
+      "learning_rate": 5.368691487541027e-07,
+      "loss": 0.3122,
+      "step": 1351
+    },
+    {
+      "epoch": 2.59776,
+      "grad_norm": 0.5520449280738831,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.2959,
+      "step": 1352
+    },
+    {
+      "epoch": 2.59968,
+      "grad_norm": 0.5748773813247681,
+      "learning_rate": 5.268268194742638e-07,
+      "loss": 0.3317,
+      "step": 1353
+    },
+    {
+      "epoch": 2.6016,
+      "grad_norm": 0.5702695250511169,
+      "learning_rate": 5.218392371123326e-07,
+      "loss": 0.3192,
+      "step": 1354
+    },
+    {
+      "epoch": 2.60352,
+      "grad_norm": 0.5375118851661682,
+      "learning_rate": 5.168740762692681e-07,
+      "loss": 0.29,
+      "step": 1355
+    },
+    {
+      "epoch": 2.6054399999999998,
+      "grad_norm": 0.5678666830062866,
+      "learning_rate": 5.119313618049309e-07,
+      "loss": 0.3283,
+      "step": 1356
+    },
+    {
+      "epoch": 2.60736,
+      "grad_norm": 0.5683631896972656,
+      "learning_rate": 5.070111184667908e-07,
+      "loss": 0.3067,
+      "step": 1357
+    },
+    {
+      "epoch": 2.60928,
+      "grad_norm": 0.5689677000045776,
+      "learning_rate": 5.021133708898146e-07,
+      "loss": 0.307,
+      "step": 1358
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.5744563341140747,
+      "learning_rate": 4.972381435963331e-07,
+      "loss": 0.3036,
+      "step": 1359
+    },
+    {
+      "epoch": 2.61312,
+      "grad_norm": 0.5442294478416443,
+      "learning_rate": 4.9238546099592e-07,
+      "loss": 0.2972,
+      "step": 1360
+    },
+    {
+      "epoch": 2.61504,
+      "grad_norm": 0.5406044125556946,
+      "learning_rate": 4.875553473852735e-07,
+      "loss": 0.283,
+      "step": 1361
+    },
+    {
+      "epoch": 2.6169599999999997,
+      "grad_norm": 0.558731198310852,
+      "learning_rate": 4.827478269480895e-07,
+      "loss": 0.3452,
+      "step": 1362
+    },
+    {
+      "epoch": 2.61888,
+      "grad_norm": 0.547626793384552,
+      "learning_rate": 4.779629237549438e-07,
+      "loss": 0.3312,
+      "step": 1363
+    },
+    {
+      "epoch": 2.6208,
+      "grad_norm": 0.5578754544258118,
+      "learning_rate": 4.732006617631729e-07,
+      "loss": 0.3007,
+      "step": 1364
+    },
+    {
+      "epoch": 2.62272,
+      "grad_norm": 0.5665492415428162,
+      "learning_rate": 4.6846106481675035e-07,
+      "loss": 0.2978,
+      "step": 1365
+    },
+    {
+      "epoch": 2.62464,
+      "grad_norm": 0.56627357006073,
+      "learning_rate": 4.637441566461681e-07,
+      "loss": 0.3239,
+      "step": 1366
+    },
+    {
+      "epoch": 2.62656,
+      "grad_norm": 0.5490041971206665,
+      "learning_rate": 4.590499608683202e-07,
+      "loss": 0.2833,
+      "step": 1367
+    },
+    {
+      "epoch": 2.62848,
+      "grad_norm": 0.5790018439292908,
+      "learning_rate": 4.54378500986381e-07,
+      "loss": 0.3046,
+      "step": 1368
+    },
+    {
+      "epoch": 2.6304,
+      "grad_norm": 0.5299497246742249,
+      "learning_rate": 4.497298003896944e-07,
+      "loss": 0.3003,
+      "step": 1369
+    },
+    {
+      "epoch": 2.63232,
+      "grad_norm": 0.5806832909584045,
+      "learning_rate": 4.451038823536441e-07,
+      "loss": 0.3241,
+      "step": 1370
+    },
+    {
+      "epoch": 2.63424,
+      "grad_norm": 0.5430147051811218,
+      "learning_rate": 4.405007700395497e-07,
+      "loss": 0.2829,
+      "step": 1371
+    },
+    {
+      "epoch": 2.63616,
+      "grad_norm": 0.5127598643302917,
+      "learning_rate": 4.35920486494546e-07,
+      "loss": 0.2514,
+      "step": 1372
+    },
+    {
+      "epoch": 2.63808,
+      "grad_norm": 0.5535872578620911,
+      "learning_rate": 4.313630546514663e-07,
+      "loss": 0.3038,
+      "step": 1373
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.5273986458778381,
+      "learning_rate": 4.268284973287273e-07,
+      "loss": 0.29,
+      "step": 1374
+    },
+    {
+      "epoch": 2.64192,
+      "grad_norm": 0.5864711403846741,
+      "learning_rate": 4.223168372302189e-07,
+      "loss": 0.3191,
+      "step": 1375
+    },
+    {
+      "epoch": 2.64384,
+      "grad_norm": 0.5816910266876221,
+      "learning_rate": 4.1782809694518533e-07,
+      "loss": 0.3214,
+      "step": 1376
+    },
+    {
+      "epoch": 2.64576,
+      "grad_norm": 0.5795580148696899,
+      "learning_rate": 4.1336229894811454e-07,
+      "loss": 0.3264,
+      "step": 1377
+    },
+    {
+      "epoch": 2.6476800000000003,
+      "grad_norm": 0.6075361371040344,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 0.345,
+      "step": 1378
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.5767363905906677,
+      "learning_rate": 4.044996191413686e-07,
+      "loss": 0.336,
+      "step": 1379
+    },
+    {
+      "epoch": 2.65152,
+      "grad_norm": 0.5655356049537659,
+      "learning_rate": 4.001027817058789e-07,
+      "loss": 0.2847,
+      "step": 1380
+    },
+    {
+      "epoch": 2.65344,
+      "grad_norm": 0.5491154789924622,
+      "learning_rate": 3.957289753065052e-07,
+      "loss": 0.3016,
+      "step": 1381
+    },
+    {
+      "epoch": 2.65536,
+      "grad_norm": 0.5363990068435669,
+      "learning_rate": 3.9137822184227845e-07,
+      "loss": 0.2559,
+      "step": 1382
+    },
+    {
+      "epoch": 2.65728,
+      "grad_norm": 0.5405572056770325,
+      "learning_rate": 3.870505430968069e-07,
+      "loss": 0.3342,
+      "step": 1383
+    },
+    {
+      "epoch": 2.6592000000000002,
+      "grad_norm": 0.5687881112098694,
+      "learning_rate": 3.8274596073816784e-07,
+      "loss": 0.3244,
+      "step": 1384
+    },
+    {
+      "epoch": 2.66112,
+      "grad_norm": 0.6006290912628174,
+      "learning_rate": 3.7846449631879664e-07,
+      "loss": 0.3874,
+      "step": 1385
+    },
+    {
+      "epoch": 2.66304,
+      "grad_norm": 0.560851514339447,
+      "learning_rate": 3.742061712753825e-07,
+      "loss": 0.3218,
+      "step": 1386
+    },
+    {
+      "epoch": 2.6649599999999998,
+      "grad_norm": 0.5569449663162231,
+      "learning_rate": 3.699710069287571e-07,
+      "loss": 0.2996,
+      "step": 1387
+    },
+    {
+      "epoch": 2.66688,
+      "grad_norm": 0.5382348895072937,
+      "learning_rate": 3.657590244837911e-07,
+      "loss": 0.2759,
+      "step": 1388
+    },
+    {
+      "epoch": 2.6688,
+      "grad_norm": 0.5894468426704407,
+      "learning_rate": 3.615702450292857e-07,
+      "loss": 0.3137,
+      "step": 1389
+    },
+    {
+      "epoch": 2.67072,
+      "grad_norm": 0.5606824159622192,
+      "learning_rate": 3.5740468953786854e-07,
+      "loss": 0.2963,
+      "step": 1390
+    },
+    {
+      "epoch": 2.67264,
+      "grad_norm": 0.567693829536438,
+      "learning_rate": 3.5326237886588734e-07,
+      "loss": 0.3421,
+      "step": 1391
+    },
+    {
+      "epoch": 2.67456,
+      "grad_norm": 0.5793265700340271,
+      "learning_rate": 3.49143333753309e-07,
+      "loss": 0.3478,
+      "step": 1392
+    },
+    {
+      "epoch": 2.6764799999999997,
+      "grad_norm": 0.5931764841079712,
+      "learning_rate": 3.4504757482360817e-07,
+      "loss": 0.3421,
+      "step": 1393
+    },
+    {
+      "epoch": 2.6784,
+      "grad_norm": 0.5348607897758484,
+      "learning_rate": 3.4097512258367385e-07,
+      "loss": 0.2681,
+      "step": 1394
+    },
+    {
+      "epoch": 2.68032,
+      "grad_norm": 0.5695550441741943,
+      "learning_rate": 3.369259974236988e-07,
+      "loss": 0.2997,
+      "step": 1395
+    },
+    {
+      "epoch": 2.68224,
+      "grad_norm": 0.5733218789100647,
+      "learning_rate": 3.3290021961708163e-07,
+      "loss": 0.3255,
+      "step": 1396
+    },
+    {
+      "epoch": 2.68416,
+      "grad_norm": 0.5123270153999329,
+      "learning_rate": 3.288978093203227e-07,
+      "loss": 0.2519,
+      "step": 1397
+    },
+    {
+      "epoch": 2.68608,
+      "grad_norm": 0.6017299294471741,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.3225,
+      "step": 1398
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.5499048829078674,
+      "learning_rate": 3.209631712972966e-07,
+      "loss": 0.2935,
+      "step": 1399
+    },
+    {
+      "epoch": 2.68992,
+      "grad_norm": 0.5104941725730896,
+      "learning_rate": 3.1703098329864237e-07,
+      "loss": 0.2432,
+      "step": 1400
+    },
+    {
+      "epoch": 2.69184,
+      "grad_norm": 0.6025027632713318,
+      "learning_rate": 3.131222422648744e-07,
+      "loss": 0.3357,
+      "step": 1401
+    },
+    {
+      "epoch": 2.69376,
+      "grad_norm": 0.5640344619750977,
+      "learning_rate": 3.0923696776650414e-07,
+      "loss": 0.2895,
+      "step": 1402
+    },
+    {
+      "epoch": 2.69568,
+      "grad_norm": 0.5519618988037109,
+      "learning_rate": 3.0537517925655567e-07,
+      "loss": 0.3145,
+      "step": 1403
+    },
+    {
+      "epoch": 2.6976,
+      "grad_norm": 0.5527986884117126,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.2924,
+      "step": 1404
+    },
+    {
+      "epoch": 2.69952,
+      "grad_norm": 0.5133557915687561,
+      "learning_rate": 2.9772213742595367e-07,
+      "loss": 0.2715,
+      "step": 1405
+    },
+    {
+      "epoch": 2.70144,
+      "grad_norm": 0.5926811099052429,
+      "learning_rate": 2.9393092242300026e-07,
+      "loss": 0.3594,
+      "step": 1406
+    },
+    {
+      "epoch": 2.70336,
+      "grad_norm": 0.55426424741745,
+      "learning_rate": 2.901632700436757e-07,
+      "loss": 0.2984,
+      "step": 1407
+    },
+    {
+      "epoch": 2.70528,
+      "grad_norm": 0.5845705270767212,
+      "learning_rate": 2.864191991520848e-07,
+      "loss": 0.3226,
+      "step": 1408
+    },
+    {
+      "epoch": 2.7072000000000003,
+      "grad_norm": 0.5657607316970825,
+      "learning_rate": 2.8269872849426114e-07,
+      "loss": 0.2864,
+      "step": 1409
+    },
+    {
+      "epoch": 2.70912,
+      "grad_norm": 0.5616539716720581,
+      "learning_rate": 2.790018766980773e-07,
+      "loss": 0.2974,
+      "step": 1410
+    },
+    {
+      "epoch": 2.71104,
+      "grad_norm": 0.5201040506362915,
+      "learning_rate": 2.7532866227314714e-07,
+      "loss": 0.2509,
+      "step": 1411
+    },
+    {
+      "epoch": 2.71296,
+      "grad_norm": 0.5703771114349365,
+      "learning_rate": 2.71679103610738e-07,
+      "loss": 0.3106,
+      "step": 1412
+    },
+    {
+      "epoch": 2.71488,
+      "grad_norm": 0.5862473845481873,
+      "learning_rate": 2.6805321898367323e-07,
+      "loss": 0.2982,
+      "step": 1413
+    },
+    {
+      "epoch": 2.7168,
+      "grad_norm": 0.49550262093544006,
+      "learning_rate": 2.64451026546248e-07,
+      "loss": 0.2528,
+      "step": 1414
+    },
+    {
+      "epoch": 2.7187200000000002,
+      "grad_norm": 0.5279282331466675,
+      "learning_rate": 2.6087254433412924e-07,
+      "loss": 0.32,
+      "step": 1415
+    },
+    {
+      "epoch": 2.72064,
+      "grad_norm": 0.5335554480552673,
+      "learning_rate": 2.573177902642726e-07,
+      "loss": 0.2753,
+      "step": 1416
+    },
+    {
+      "epoch": 2.72256,
+      "grad_norm": 0.5452460050582886,
+      "learning_rate": 2.5378678213483057e-07,
+      "loss": 0.3014,
+      "step": 1417
+    },
+    {
+      "epoch": 2.72448,
+      "grad_norm": 0.5915035605430603,
+      "learning_rate": 2.502795376250622e-07,
+      "loss": 0.3623,
+      "step": 1418
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.5688404440879822,
+      "learning_rate": 2.467960742952463e-07,
+      "loss": 0.3546,
+      "step": 1419
+    },
+    {
+      "epoch": 2.72832,
+      "grad_norm": 0.5816324353218079,
+      "learning_rate": 2.4333640958659144e-07,
+      "loss": 0.302,
+      "step": 1420
+    },
+    {
+      "epoch": 2.7302400000000002,
+      "grad_norm": 0.5857483148574829,
+      "learning_rate": 2.399005608211502e-07,
+      "loss": 0.2811,
+      "step": 1421
+    },
+    {
+      "epoch": 2.73216,
+      "grad_norm": 0.5573190450668335,
+      "learning_rate": 2.3648854520173237e-07,
+      "loss": 0.28,
+      "step": 1422
+    },
+    {
+      "epoch": 2.73408,
+      "grad_norm": 0.6168860197067261,
+      "learning_rate": 2.3310037981182088e-07,
+      "loss": 0.3208,
+      "step": 1423
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 0.5611668229103088,
+      "learning_rate": 2.2973608161547755e-07,
+      "loss": 0.3221,
+      "step": 1424
+    },
+    {
+      "epoch": 2.73792,
+      "grad_norm": 0.537710428237915,
+      "learning_rate": 2.2639566745727203e-07,
+      "loss": 0.2768,
+      "step": 1425
+    },
+    {
+      "epoch": 2.73984,
+      "grad_norm": 0.5652465224266052,
+      "learning_rate": 2.2307915406218517e-07,
+      "loss": 0.2947,
+      "step": 1426
+    },
+    {
+      "epoch": 2.74176,
+      "grad_norm": 0.5713388919830322,
+      "learning_rate": 2.1978655803553128e-07,
+      "loss": 0.2954,
+      "step": 1427
+    },
+    {
+      "epoch": 2.74368,
+      "grad_norm": 0.5514291524887085,
+      "learning_rate": 2.1651789586287442e-07,
+      "loss": 0.2762,
+      "step": 1428
+    },
+    {
+      "epoch": 2.7456,
+      "grad_norm": 0.5554014444351196,
+      "learning_rate": 2.1327318390994445e-07,
+      "loss": 0.3158,
+      "step": 1429
+    },
+    {
+      "epoch": 2.7475199999999997,
+      "grad_norm": 0.5611034631729126,
+      "learning_rate": 2.1005243842255552e-07,
+      "loss": 0.3316,
+      "step": 1430
+    },
+    {
+      "epoch": 2.74944,
+      "grad_norm": 0.5305187106132507,
+      "learning_rate": 2.068556755265272e-07,
+      "loss": 0.3038,
+      "step": 1431
+    },
+    {
+      "epoch": 2.75136,
+      "grad_norm": 0.5081883668899536,
+      "learning_rate": 2.0368291122759898e-07,
+      "loss": 0.2625,
+      "step": 1432
+    },
+    {
+      "epoch": 2.75328,
+      "grad_norm": 0.558817446231842,
+      "learning_rate": 2.005341614113543e-07,
+      "loss": 0.3291,
+      "step": 1433
+    },
+    {
+      "epoch": 2.7552,
+      "grad_norm": 0.5524738430976868,
+      "learning_rate": 1.9740944184313882e-07,
+      "loss": 0.289,
+      "step": 1434
+    },
+    {
+      "epoch": 2.75712,
+      "grad_norm": 0.5792908668518066,
+      "learning_rate": 1.9430876816798228e-07,
+      "loss": 0.3198,
+      "step": 1435
+    },
+    {
+      "epoch": 2.75904,
+      "grad_norm": 0.5416079759597778,
+      "learning_rate": 1.9123215591052014e-07,
+      "loss": 0.3105,
+      "step": 1436
+    },
+    {
+      "epoch": 2.76096,
+      "grad_norm": 0.5259556174278259,
+      "learning_rate": 1.88179620474917e-07,
+      "loss": 0.2771,
+      "step": 1437
+    },
+    {
+      "epoch": 2.76288,
+      "grad_norm": 0.5737584233283997,
+      "learning_rate": 1.8515117714478447e-07,
+      "loss": 0.3013,
+      "step": 1438
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.5395472645759583,
+      "learning_rate": 1.8214684108311286e-07,
+      "loss": 0.2846,
+      "step": 1439
+    },
+    {
+      "epoch": 2.76672,
+      "grad_norm": 0.5394297242164612,
+      "learning_rate": 1.7916662733218848e-07,
+      "loss": 0.3198,
+      "step": 1440
+    },
+    {
+      "epoch": 2.76864,
+      "grad_norm": 0.5442410111427307,
+      "learning_rate": 1.762105508135198e-07,
+      "loss": 0.2909,
+      "step": 1441
+    },
+    {
+      "epoch": 2.77056,
+      "grad_norm": 0.5793753862380981,
+      "learning_rate": 1.7327862632776638e-07,
+      "loss": 0.2918,
+      "step": 1442
+    },
+    {
+      "epoch": 2.77248,
+      "grad_norm": 0.5948330163955688,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.3296,
+      "step": 1443
+    },
+    {
+      "epoch": 2.7744,
+      "grad_norm": 0.5751848220825195,
+      "learning_rate": 1.6748729205293024e-07,
+      "loss": 0.3495,
+      "step": 1444
+    },
+    {
+      "epoch": 2.77632,
+      "grad_norm": 0.5466296076774597,
+      "learning_rate": 1.646279112602417e-07,
+      "loss": 0.3077,
+      "step": 1445
+    },
+    {
+      "epoch": 2.7782400000000003,
+      "grad_norm": 0.5667700171470642,
+      "learning_rate": 1.6179274049310966e-07,
+      "loss": 0.3313,
+      "step": 1446
+    },
+    {
+      "epoch": 2.78016,
+      "grad_norm": 0.5523081421852112,
+      "learning_rate": 1.5898179394683244e-07,
+      "loss": 0.3286,
+      "step": 1447
+    },
+    {
+      "epoch": 2.78208,
+      "grad_norm": 0.5246440172195435,
+      "learning_rate": 1.5619508569542363e-07,
+      "loss": 0.2743,
+      "step": 1448
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 0.5458906292915344,
+      "learning_rate": 1.5343262969153781e-07,
+      "loss": 0.2855,
+      "step": 1449
+    },
+    {
+      "epoch": 2.78592,
+      "grad_norm": 0.5353618860244751,
+      "learning_rate": 1.5069443976640287e-07,
+      "loss": 0.29,
+      "step": 1450
+    },
+    {
+      "epoch": 2.78784,
+      "grad_norm": 0.5422212481498718,
+      "learning_rate": 1.4798052962974874e-07,
+      "loss": 0.2988,
+      "step": 1451
+    },
+    {
+      "epoch": 2.7897600000000002,
+      "grad_norm": 0.566254734992981,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.2868,
+      "step": 1452
+    },
+    {
+      "epoch": 2.79168,
+      "grad_norm": 0.5234791040420532,
+      "learning_rate": 1.4262560295290884e-07,
+      "loss": 0.27,
+      "step": 1453
+    },
+    {
+      "epoch": 2.7936,
+      "grad_norm": 0.5908464193344116,
+      "learning_rate": 1.3998461322408563e-07,
+      "loss": 0.3296,
+      "step": 1454
+    },
+    {
+      "epoch": 2.79552,
+      "grad_norm": 0.5589529275894165,
+      "learning_rate": 1.3736795690633353e-07,
+      "loss": 0.2973,
+      "step": 1455
+    },
+    {
+      "epoch": 2.79744,
+      "grad_norm": 0.5492869019508362,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.2949,
+      "step": 1456
+    },
+    {
+      "epoch": 2.79936,
+      "grad_norm": 0.5032598972320557,
+      "learning_rate": 1.3220769678705724e-07,
+      "loss": 0.2265,
+      "step": 1457
+    },
+    {
+      "epoch": 2.80128,
+      "grad_norm": 0.5773717164993286,
+      "learning_rate": 1.2966411882222695e-07,
+      "loss": 0.3126,
+      "step": 1458
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.5409242510795593,
+      "learning_rate": 1.271449259417268e-07,
+      "loss": 0.2674,
+      "step": 1459
+    },
+    {
+      "epoch": 2.80512,
+      "grad_norm": 0.5595743060112,
+      "learning_rate": 1.2465013075879884e-07,
+      "loss": 0.2973,
+      "step": 1460
+    },
+    {
+      "epoch": 2.8070399999999998,
+      "grad_norm": 0.5802177786827087,
+      "learning_rate": 1.2217974576453072e-07,
+      "loss": 0.3245,
+      "step": 1461
+    },
+    {
+      "epoch": 2.80896,
+      "grad_norm": 0.544562816619873,
+      "learning_rate": 1.1973378332779229e-07,
+      "loss": 0.2925,
+      "step": 1462
+    },
+    {
+      "epoch": 2.81088,
+      "grad_norm": 0.5287115573883057,
+      "learning_rate": 1.1731225569517113e-07,
+      "loss": 0.2765,
+      "step": 1463
+    },
+    {
+      "epoch": 2.8128,
+      "grad_norm": 0.5500369668006897,
+      "learning_rate": 1.1491517499091498e-07,
+      "loss": 0.2925,
+      "step": 1464
+    },
+    {
+      "epoch": 2.81472,
+      "grad_norm": 0.5520563721656799,
+      "learning_rate": 1.1254255321686836e-07,
+      "loss": 0.2904,
+      "step": 1465
+    },
+    {
+      "epoch": 2.81664,
+      "grad_norm": 0.5520586967468262,
+      "learning_rate": 1.1019440225241317e-07,
+      "loss": 0.2846,
+      "step": 1466
+    },
+    {
+      "epoch": 2.8185599999999997,
+      "grad_norm": 0.5287692546844482,
+      "learning_rate": 1.078707338544105e-07,
+      "loss": 0.2967,
+      "step": 1467
+    },
+    {
+      "epoch": 2.82048,
+      "grad_norm": 0.4905540943145752,
+      "learning_rate": 1.055715596571405e-07,
+      "loss": 0.2594,
+      "step": 1468
+    },
+    {
+      "epoch": 2.8224,
+      "grad_norm": 0.5458372235298157,
+      "learning_rate": 1.0329689117224262e-07,
+      "loss": 0.2689,
+      "step": 1469
+    },
+    {
+      "epoch": 2.82432,
+      "grad_norm": 0.5809189677238464,
+      "learning_rate": 1.0104673978866164e-07,
+      "loss": 0.3084,
+      "step": 1470
+    },
+    {
+      "epoch": 2.82624,
+      "grad_norm": 0.5623963475227356,
+      "learning_rate": 9.882111677258777e-08,
+      "loss": 0.3295,
+      "step": 1471
+    },
+    {
+      "epoch": 2.82816,
+      "grad_norm": 0.5905405282974243,
+      "learning_rate": 9.662003326740166e-08,
+      "loss": 0.325,
+      "step": 1472
+    },
+    {
+      "epoch": 2.83008,
+      "grad_norm": 0.5417467355728149,
+      "learning_rate": 9.444350029361671e-08,
+      "loss": 0.2823,
+      "step": 1473
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.5448617339134216,
+      "learning_rate": 9.22915287488274e-08,
+      "loss": 0.3106,
+      "step": 1474
+    },
+    {
+      "epoch": 2.83392,
+      "grad_norm": 0.5453553199768066,
+      "learning_rate": 9.016412940765107e-08,
+      "loss": 0.2892,
+      "step": 1475
+    },
+    {
+      "epoch": 2.83584,
+      "grad_norm": 0.6035391092300415,
+      "learning_rate": 8.80613129216762e-08,
+      "loss": 0.3492,
+      "step": 1476
+    },
+    {
+      "epoch": 2.83776,
+      "grad_norm": 0.540359616279602,
+      "learning_rate": 8.598308981940751e-08,
+      "loss": 0.2931,
+      "step": 1477
+    },
+    {
+      "epoch": 2.83968,
+      "grad_norm": 0.6145979762077332,
+      "learning_rate": 8.392947050621603e-08,
+      "loss": 0.3372,
+      "step": 1478
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.5423272252082825,
+      "learning_rate": 8.190046526428241e-08,
+      "loss": 0.2524,
+      "step": 1479
+    },
+    {
+      "epoch": 2.84352,
+      "grad_norm": 0.5985803604125977,
+      "learning_rate": 7.989608425254924e-08,
+      "loss": 0.331,
+      "step": 1480
+    },
+    {
+      "epoch": 2.84544,
+      "grad_norm": 0.5351382493972778,
+      "learning_rate": 7.791633750667105e-08,
+      "loss": 0.3061,
+      "step": 1481
+    },
+    {
+      "epoch": 2.84736,
+      "grad_norm": 0.5449694991111755,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.2747,
+      "step": 1482
+    },
+    {
+      "epoch": 2.8492800000000003,
+      "grad_norm": 0.5747963786125183,
+      "learning_rate": 7.403078633833716e-08,
+      "loss": 0.3059,
+      "step": 1483
+    },
+    {
+      "epoch": 2.8512,
+      "grad_norm": 0.5565093755722046,
+      "learning_rate": 7.212500137028789e-08,
+      "loss": 0.285,
+      "step": 1484
+    },
+    {
+      "epoch": 2.85312,
+      "grad_norm": 0.5490142703056335,
+      "learning_rate": 7.024388957680705e-08,
+      "loss": 0.2823,
+      "step": 1485
+    },
+    {
+      "epoch": 2.85504,
+      "grad_norm": 0.5610407590866089,
+      "learning_rate": 6.838746037635735e-08,
+      "loss": 0.2809,
+      "step": 1486
+    },
+    {
+      "epoch": 2.85696,
+      "grad_norm": 0.5866698026657104,
+      "learning_rate": 6.655572306381696e-08,
+      "loss": 0.3125,
+      "step": 1487
+    },
+    {
+      "epoch": 2.85888,
+      "grad_norm": 0.5672308802604675,
+      "learning_rate": 6.474868681043578e-08,
+      "loss": 0.3126,
+      "step": 1488
+    },
+    {
+      "epoch": 2.8608000000000002,
+      "grad_norm": 0.5824835300445557,
+      "learning_rate": 6.29663606637898e-08,
+      "loss": 0.3239,
+      "step": 1489
+    },
+    {
+      "epoch": 2.86272,
+      "grad_norm": 0.5627408027648926,
+      "learning_rate": 6.120875354773459e-08,
+      "loss": 0.3208,
+      "step": 1490
+    },
+    {
+      "epoch": 2.86464,
+      "grad_norm": 0.5675732493400574,
+      "learning_rate": 5.947587426236079e-08,
+      "loss": 0.2856,
+      "step": 1491
+    },
+    {
+      "epoch": 2.8665599999999998,
+      "grad_norm": 0.6041935682296753,
+      "learning_rate": 5.776773148394976e-08,
+      "loss": 0.3268,
+      "step": 1492
+    },
+    {
+      "epoch": 2.86848,
+      "grad_norm": 0.5871978402137756,
+      "learning_rate": 5.608433376493194e-08,
+      "loss": 0.2714,
+      "step": 1493
+    },
+    {
+      "epoch": 2.8704,
+      "grad_norm": 0.5642929673194885,
+      "learning_rate": 5.4425689533841864e-08,
+      "loss": 0.2847,
+      "step": 1494
+    },
+    {
+      "epoch": 2.87232,
+      "grad_norm": 0.6233981251716614,
+      "learning_rate": 5.279180709527765e-08,
+      "loss": 0.3584,
+      "step": 1495
+    },
+    {
+      "epoch": 2.87424,
+      "grad_norm": 0.5557087063789368,
+      "learning_rate": 5.1182694629857145e-08,
+      "loss": 0.3004,
+      "step": 1496
+    },
+    {
+      "epoch": 2.87616,
+      "grad_norm": 0.5885350704193115,
+      "learning_rate": 4.959836019417963e-08,
+      "loss": 0.3413,
+      "step": 1497
+    },
+    {
+      "epoch": 2.8780799999999997,
+      "grad_norm": 0.594765305519104,
+      "learning_rate": 4.803881172078473e-08,
+      "loss": 0.3121,
+      "step": 1498
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.550088107585907,
+      "learning_rate": 4.650405701811078e-08,
+      "loss": 0.2841,
+      "step": 1499
+    },
+    {
+      "epoch": 2.88192,
+      "grad_norm": 0.5379555821418762,
+      "learning_rate": 4.499410377045765e-08,
+      "loss": 0.2925,
+      "step": 1500
+    },
+    {
+      "epoch": 2.88384,
+      "grad_norm": 0.5393018126487732,
+      "learning_rate": 4.350895953794898e-08,
+      "loss": 0.3025,
+      "step": 1501
+    },
+    {
+      "epoch": 2.88576,
+      "grad_norm": 0.5759963989257812,
+      "learning_rate": 4.2048631756492206e-08,
+      "loss": 0.2979,
+      "step": 1502
+    },
+    {
+      "epoch": 2.88768,
+      "grad_norm": 0.5605481266975403,
+      "learning_rate": 4.0613127737741396e-08,
+      "loss": 0.2814,
+      "step": 1503
+    },
+    {
+      "epoch": 2.8895999999999997,
+      "grad_norm": 0.5624799132347107,
+      "learning_rate": 3.9202454669063915e-08,
+      "loss": 0.3094,
+      "step": 1504
+    },
+    {
+      "epoch": 2.89152,
+      "grad_norm": 0.6108046174049377,
+      "learning_rate": 3.781661961349992e-08,
+      "loss": 0.3335,
+      "step": 1505
+    },
+    {
+      "epoch": 2.89344,
+      "grad_norm": 0.5063559412956238,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.2639,
+      "step": 1506
+    },
+    {
+      "epoch": 2.89536,
+      "grad_norm": 0.6236696839332581,
+      "learning_rate": 3.5119491172039836e-08,
+      "loss": 0.3292,
+      "step": 1507
+    },
+    {
+      "epoch": 2.89728,
+      "grad_norm": 0.5660619139671326,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.3442,
+      "step": 1508
+    },
+    {
+      "epoch": 2.8992,
+      "grad_norm": 0.5539608597755432,
+      "learning_rate": 3.252179642985909e-08,
+      "loss": 0.2992,
+      "step": 1509
+    },
+    {
+      "epoch": 2.90112,
+      "grad_norm": 0.5649862289428711,
+      "learning_rate": 3.1260253031660247e-08,
+      "loss": 0.3084,
+      "step": 1510
+    },
+    {
+      "epoch": 2.90304,
+      "grad_norm": 0.5749421119689941,
+      "learning_rate": 3.002358741205691e-08,
+      "loss": 0.2846,
+      "step": 1511
+    },
+    {
+      "epoch": 2.90496,
+      "grad_norm": 0.5437770485877991,
+      "learning_rate": 2.8811805762860578e-08,
+      "loss": 0.2855,
+      "step": 1512
+    },
+    {
+      "epoch": 2.90688,
+      "grad_norm": 0.6007202863693237,
+      "learning_rate": 2.762491415128965e-08,
+      "loss": 0.3384,
+      "step": 1513
+    },
+    {
+      "epoch": 2.9088000000000003,
+      "grad_norm": 0.546088695526123,
+      "learning_rate": 2.6462918519944425e-08,
+      "loss": 0.2787,
+      "step": 1514
+    },
+    {
+      "epoch": 2.91072,
+      "grad_norm": 0.5877922773361206,
+      "learning_rate": 2.5325824686772138e-08,
+      "loss": 0.3025,
+      "step": 1515
+    },
+    {
+      "epoch": 2.91264,
+      "grad_norm": 0.565007746219635,
+      "learning_rate": 2.4213638345040868e-08,
+      "loss": 0.3325,
+      "step": 1516
+    },
+    {
+      "epoch": 2.91456,
+      "grad_norm": 0.5694823861122131,
+      "learning_rate": 2.3126365063311228e-08,
+      "loss": 0.307,
+      "step": 1517
+    },
+    {
+      "epoch": 2.91648,
+      "grad_norm": 0.5272535681724548,
+      "learning_rate": 2.206401028540639e-08,
+      "loss": 0.2741,
+      "step": 1518
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.5387914776802063,
+      "learning_rate": 2.1026579330387655e-08,
+      "loss": 0.2629,
+      "step": 1519
+    },
+    {
+      "epoch": 2.9203200000000002,
+      "grad_norm": 0.558050811290741,
+      "learning_rate": 2.0014077392525035e-08,
+      "loss": 0.2984,
+      "step": 1520
+    },
+    {
+      "epoch": 2.92224,
+      "grad_norm": 0.5727524757385254,
+      "learning_rate": 1.9026509541272276e-08,
+      "loss": 0.3438,
+      "step": 1521
+    },
+    {
+      "epoch": 2.92416,
+      "grad_norm": 0.5330016016960144,
+      "learning_rate": 1.8063880721242986e-08,
+      "loss": 0.2944,
+      "step": 1522
+    },
+    {
+      "epoch": 2.92608,
+      "grad_norm": 0.595766007900238,
+      "learning_rate": 1.712619575218344e-08,
+      "loss": 0.3396,
+      "step": 1523
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 0.5626906752586365,
+      "learning_rate": 1.6213459328950355e-08,
+      "loss": 0.3065,
+      "step": 1524
+    },
+    {
+      "epoch": 2.92992,
+      "grad_norm": 0.5716311931610107,
+      "learning_rate": 1.5325676021484826e-08,
+      "loss": 0.2791,
+      "step": 1525
+    },
+    {
+      "epoch": 2.9318400000000002,
+      "grad_norm": 0.5433459877967834,
+      "learning_rate": 1.4462850274794548e-08,
+      "loss": 0.2781,
+      "step": 1526
+    },
+    {
+      "epoch": 2.93376,
+      "grad_norm": 0.5663525462150574,
+      "learning_rate": 1.3624986408924956e-08,
+      "loss": 0.3523,
+      "step": 1527
+    },
+    {
+      "epoch": 2.93568,
+      "grad_norm": 0.5889537334442139,
+      "learning_rate": 1.2812088618942009e-08,
+      "loss": 0.3014,
+      "step": 1528
+    },
+    {
+      "epoch": 2.9375999999999998,
+      "grad_norm": 0.5145683288574219,
+      "learning_rate": 1.2024160974911103e-08,
+      "loss": 0.2709,
+      "step": 1529
+    },
+    {
+      "epoch": 2.93952,
+      "grad_norm": 0.5143959522247314,
+      "learning_rate": 1.1261207421874309e-08,
+      "loss": 0.2825,
+      "step": 1530
+    },
+    {
+      "epoch": 2.94144,
+      "grad_norm": 0.5216909050941467,
+      "learning_rate": 1.0523231779832055e-08,
+      "loss": 0.2921,
+      "step": 1531
+    },
+    {
+      "epoch": 2.94336,
+      "grad_norm": 0.5247679948806763,
+      "learning_rate": 9.810237743724805e-09,
+      "loss": 0.2987,
+      "step": 1532
+    },
+    {
+      "epoch": 2.94528,
+      "grad_norm": 0.5490642786026001,
+      "learning_rate": 9.12222888341252e-09,
+      "loss": 0.293,
+      "step": 1533
+    },
+    {
+      "epoch": 2.9472,
+      "grad_norm": 0.543799877166748,
+      "learning_rate": 8.459208643659122e-09,
+      "loss": 0.3116,
+      "step": 1534
+    },
+    {
+      "epoch": 2.9491199999999997,
+      "grad_norm": 0.5274407863616943,
+      "learning_rate": 7.8211803441125e-09,
+      "loss": 0.2874,
+      "step": 1535
+    },
+    {
+      "epoch": 2.95104,
+      "grad_norm": 0.5699525475502014,
+      "learning_rate": 7.2081471792911914e-09,
+      "loss": 0.3317,
+      "step": 1536
+    },
+    {
+      "epoch": 2.95296,
+      "grad_norm": 0.5978836417198181,
+      "learning_rate": 6.6201122185649555e-09,
+      "loss": 0.3144,
+      "step": 1537
+    },
+    {
+      "epoch": 2.95488,
+      "grad_norm": 0.544627845287323,
+      "learning_rate": 6.057078406142003e-09,
+      "loss": 0.2797,
+      "step": 1538
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.5838849544525146,
+      "learning_rate": 5.519048561053453e-09,
+      "loss": 0.3329,
+      "step": 1539
+    },
+    {
+      "epoch": 2.95872,
+      "grad_norm": 0.5277259945869446,
+      "learning_rate": 5.006025377138901e-09,
+      "loss": 0.2879,
+      "step": 1540
+    },
+    {
+      "epoch": 2.96064,
+      "grad_norm": 0.5589214563369751,
+      "learning_rate": 4.518011423032542e-09,
+      "loss": 0.291,
+      "step": 1541
+    },
+    {
+      "epoch": 2.96256,
+      "grad_norm": 0.5628105401992798,
+      "learning_rate": 4.055009142152066e-09,
+      "loss": 0.2998,
+      "step": 1542
+    },
+    {
+      "epoch": 2.96448,
+      "grad_norm": 0.553152859210968,
+      "learning_rate": 3.6170208526836724e-09,
+      "loss": 0.299,
+      "step": 1543
+    },
+    {
+      "epoch": 2.9664,
+      "grad_norm": 0.5817992687225342,
+      "learning_rate": 3.204048747573185e-09,
+      "loss": 0.3575,
+      "step": 1544
+    },
+    {
+      "epoch": 2.96832,
+      "grad_norm": 0.5697669982910156,
+      "learning_rate": 2.816094894513843e-09,
+      "loss": 0.3402,
+      "step": 1545
+    },
+    {
+      "epoch": 2.97024,
+      "grad_norm": 0.5430675148963928,
+      "learning_rate": 2.4531612359363077e-09,
+      "loss": 0.328,
+      "step": 1546
+    },
+    {
+      "epoch": 2.97216,
+      "grad_norm": 0.5338860750198364,
+      "learning_rate": 2.1152495889970035e-09,
+      "loss": 0.2769,
+      "step": 1547
+    },
+    {
+      "epoch": 2.97408,
+      "grad_norm": 0.601998507976532,
+      "learning_rate": 1.8023616455731253e-09,
+      "loss": 0.3718,
+      "step": 1548
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 0.548792839050293,
+      "learning_rate": 1.514498972249312e-09,
+      "loss": 0.3133,
+      "step": 1549
+    },
+    {
+      "epoch": 2.97792,
+      "grad_norm": 0.5115510821342468,
+      "learning_rate": 1.2516630103137638e-09,
+      "loss": 0.2673,
+      "step": 1550
+    },
+    {
+      "epoch": 2.9798400000000003,
+      "grad_norm": 0.5228269100189209,
+      "learning_rate": 1.0138550757493592e-09,
+      "loss": 0.2658,
+      "step": 1551
+    },
+    {
+      "epoch": 2.98176,
+      "grad_norm": 0.528329610824585,
+      "learning_rate": 8.010763592264381e-10,
+      "loss": 0.2851,
+      "step": 1552
+    },
+    {
+      "epoch": 2.98368,
+      "grad_norm": 0.5231673121452332,
+      "learning_rate": 6.133279260983616e-10,
+      "loss": 0.2909,
+      "step": 1553
+    },
+    {
+      "epoch": 2.9856,
+      "grad_norm": 0.5630878210067749,
+      "learning_rate": 4.506107163948503e-10,
+      "loss": 0.3072,
+      "step": 1554
+    },
+    {
+      "epoch": 2.98752,
+      "grad_norm": 0.567456066608429,
+      "learning_rate": 3.12925544818099e-10,
+      "loss": 0.3041,
+      "step": 1555
+    },
+    {
+      "epoch": 2.98944,
+      "grad_norm": 0.561774492263794,
+      "learning_rate": 2.0027310073833516e-10,
+      "loss": 0.2994,
+      "step": 1556
+    },
+    {
+      "epoch": 2.9913600000000002,
+      "grad_norm": 0.603220522403717,
+      "learning_rate": 1.1265394818993358e-10,
+      "loss": 0.3189,
+      "step": 1557
+    },
+    {
+      "epoch": 2.99328,
+      "grad_norm": 0.5047410130500793,
+      "learning_rate": 5.0068525870305974e-11,
+      "loss": 0.2327,
+      "step": 1558
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.5660232901573181,
+      "learning_rate": 1.251714713546015e-11,
+      "loss": 0.3071,
+      "step": 1559
+    },
+    {
+      "epoch": 2.99712,
+      "grad_norm": 0.5263440012931824,
+      "learning_rate": 0.0,
+      "loss": 0.269,
+      "step": 1560
+    },
+    {
+      "epoch": 2.99712,
+      "step": 1560,
+      "total_flos": 266431779422208.0,
+      "train_loss": 0.3710105074999424,
+      "train_runtime": 14130.5903,
+      "train_samples_per_second": 10.615,
+      "train_steps_per_second": 0.11
+    }
+  ],
+  "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": 266431779422208.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}