diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.988354430379747,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 7.527076096692714,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.8709,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 7.582557022662077,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 0.8699,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 7.548924645509905,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.8666,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 7.475113724687576,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.8798,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.799123121443767,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.8387,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 6.840080193130509,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.8262,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.772394552773234,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.8124,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 5.42927643225118,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.8177,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.089622766499983,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.7283,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.7730088988367307,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.7166,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.493275555470038,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.7297,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.7975598425918866,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.6967,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.237090021488338,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.6986,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.768238859996337,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.668,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.4502973994294686,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.6539,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.985220841347346,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.6151,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.933891056774459,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.5994,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.5461672685000853,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.5908,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.9610024623194584,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.5876,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.4364041618541956,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.5867,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.3604057336269975,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.5403,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.653946624518525,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.5459,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.568193018564551,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.5285,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.093292271502812,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.5338,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.8865066957578928,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.5287,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1068027817557777,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.5234,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.0061193950681666,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.5054,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.765847602909696,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.5109,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.7508812150202899,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.5238,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8422316175454332,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.4838,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8578103836590061,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.5105,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.62093158227521,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.4684,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.6585077932793563,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.4917,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.6777758592301273,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.4769,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.6607938304510559,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.4871,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.6386670538193097,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.4662,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.6882048608556005,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.4793,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6765501235557523,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.4725,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6005280015594142,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.4795,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7044051971355302,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.4848,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.6206163680767645,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.4582,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.7155260403101985,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.4758,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.5890870219089316,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.4623,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.5318663624871932,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.4497,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6865034736164682,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.473,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6500633827768643,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.4335,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7364356120346403,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.4301,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.6027138245761108,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.4535,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.9143834103821317,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.4545,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.6528957107740687,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.4557,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.6518918652064657,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.4542,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.605749814703612,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.4403,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.66014669808256,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.4668,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.5654821985935837,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.4553,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.6442533640984571,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.4542,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.6795914229228068,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.4369,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.680339419901832,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.4418,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8120598211005785,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.4497,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.822101277259329,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.4415,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.7774153453968174,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.4316,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.8614901104233991,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.4454,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.727285235570131,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.431,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.9629685845106366,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.4388,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.7632686871368123,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.4333,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.6817139821636937,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.4449,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.671907760484157,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.4414,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.8290239228884142,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.4301,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.6435508452443077,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.4438,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.6060124162832746,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.4225,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.7260662162622552,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.4267,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.5687902965065843,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.428,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.7584088514560654,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.4436,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.7472179705722637,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.4403,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.9734664873352147,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.4394,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.7078965890640506,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.4354,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.8876965176119289,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.4375,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.5998573457105476,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.4172,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.84993333547805,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.4384,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5701257185691007,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.4173,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.6832792573803107,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.447,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.5927880750221343,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.4174,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.717181776613709,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4385,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.6326789750912956,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.4294,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.6993613030274495,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.4441,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.7077242762572704,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.4216,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.6700669909867093,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.4123,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.7419645477976812,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.4228,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.6815647955268159,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.4272,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.6773724857288099,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.4081,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.7347627347395281,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.4174,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.8821797460366304,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.4291,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.7202841056131238,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.4294,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.6281601519534314,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.441,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.6182705557947554,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.4068,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.7150871400634707,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.4399,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.6991847968343469,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.4289,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.6862408086562818,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.4397,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.5717402380425879,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.4388,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.8128381363312509,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.425,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.9195311093380673,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.4317,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.8451079894251896,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.4385,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.9144350112642171,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.4392,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.571374595250725,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.4215,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.8374408977620348,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.4295,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.7818231699250037,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.4235,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.6399015701579468,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.4208,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.8254837587671889,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.4113,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.9898895910518165,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.4327,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.7188174917831045,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.4456,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.6756219040108241,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.4155,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.8495873072977147,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.441,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.6583160773422243,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.4228,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.6772888866051994,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.4285,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.9604207534076106,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.4422,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.8586834066657907,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.4256,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.6069667408902724,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.4125,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.869178195235486,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.4547,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.7212427839968107,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.4191,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.6867614038042704,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.4049,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.4985081887366083,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.426,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.6403896999241384,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.4244,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.6398595809058498,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.4207,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.649381495703123,
+      "learning_rate": 4e-05,
+      "loss": 0.4113,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.5161111508094524,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.4372,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.5693724227658651,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.4347,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.6842328073994829,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.4365,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.6419144468340265,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.4231,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.5709212301804785,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.3873,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.6339853686969806,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.4243,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.814575145226862,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.4269,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.6841703112712746,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.4372,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.715446908251466,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.4216,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.7260779384113824,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.4245,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.7806446348407472,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.4366,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.6017726151061273,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.4451,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.6268315385372626,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.4235,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.6745209129524968,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.4154,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.5884005397107197,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.4259,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.5690625351942581,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.4166,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.7313755040207544,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.4247,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.6492787566658088,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.4065,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.7019923762081841,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.4142,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.846522694181363,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.4296,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.5096679164815313,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.426,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.7646132514179469,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.4079,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.5963494420819231,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.4221,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.6388030399712905,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.4153,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.546379328657648,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.4079,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.7337439986080107,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.4238,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.5815299510641235,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.4259,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.5889078158489808,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.4262,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.5715805872495453,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.4335,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.5763386002789992,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.4154,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.5465789585943186,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.412,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.5421666020268171,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.4147,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.5398874445939679,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.4105,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.44473888246714466,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.3887,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5487038497937811,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.42,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.5387175022978156,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.4278,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.48735790758915126,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.4211,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.5966990605658485,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.4131,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.5813667802276059,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.4234,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.49355597649530103,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.404,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.558862241104456,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.439,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.5121653951887409,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.4004,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.5187166886128561,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.4207,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.5290927171643314,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.4186,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.4776012327010876,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.4076,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.641281433871652,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.4297,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.6212381736411501,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.4003,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.49340667195965004,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.4163,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.7019363093129196,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.4095,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.5425315451433032,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.4204,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.6143277328819339,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.425,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.5633549392219812,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.4038,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.6208317516724524,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.3977,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.46914618467249797,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.4266,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.5181049282861401,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.4097,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.4756166932390534,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.3967,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.48282050748664196,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.424,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.4997662000463143,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.4169,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.5719904871712101,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.4099,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.483329811756118,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.4033,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.5801080217353537,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.4064,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.45511757377460316,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.4054,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.557026371020462,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.4031,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.5211448088839995,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.4213,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.5386685711265724,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.4355,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.4643902489760019,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.4105,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.5677220684083739,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.3944,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.4969845445765455,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.4137,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.504170399397714,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.3873,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.559098940494722,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.4109,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6420832017051156,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.4078,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.6002197123502346,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.4101,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.4996484181327684,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.4112,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.5400270401093356,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.4036,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6280206081283952,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.4002,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.4073086159554425,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.4151,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.641956821249613,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.3975,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.6986378918941479,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.4223,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.4213758124141269,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.3972,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.67421787477896,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.4171,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.4162955618335279,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.3978,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.5870919374486856,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.4009,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.5407274879917621,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.4242,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.5316230729519162,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.4051,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.6892710025330142,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.3983,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.5238709689279634,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.4078,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.762100228708086,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.4181,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.7456534318529437,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.4031,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5890185074145464,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.403,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.7672853413933207,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.3903,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.5414988002460092,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.4069,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.9091859406824638,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.4186,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6449351132411659,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.4203,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.6572249398536855,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.415,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.7629622382377921,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.4049,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.43912463028033966,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.4093,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.6996349846244798,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.4096,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.4909557540208431,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.4225,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.576749926035123,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.4,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 1.2461345283568779,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.4507,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.5724054897298048,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.4149,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.5737145826045262,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.4001,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.4965806082249067,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.4135,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.5113515431189364,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.4095,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.44996375523857113,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.3941,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.5181508742761869,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.4111,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.4880063966422817,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.3991,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.5162968459040623,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.3852,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.440392394837802,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.4085,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.5210249607464091,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.392,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.4520389544237369,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.4197,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.5612541899124449,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.4179,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.4903846297764731,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.3934,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5190433767344278,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.4041,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.4688099341183763,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.4143,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.5715237285484925,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.3912,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.425080634042541,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.4109,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.6341899019224486,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.3888,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.46265447035510154,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.3995,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.49898428346784324,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.3839,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.490471044947893,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.3979,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.5128164074973695,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.4094,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.4845307047794919,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.4024,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.5571478368975784,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.3885,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.3986477999127251,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.325,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.5448672585984008,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.3405,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.44640675766192145,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.3313,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.44372145366299875,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.3265,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.49114518146038905,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.348,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.4721631791220245,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.3343,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.5268888999227905,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.3332,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.4580320233264645,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.3468,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.5159281054288687,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.3426,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.42855719676506077,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.338,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.4348525288109394,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.3276,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.4350268840995788,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.3387,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.5592273500168142,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.3462,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.4146487247944348,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.3132,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.5764010357885209,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.3251,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.48319859423580214,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.3178,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.4529010182657376,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.3218,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.5153907425833623,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.342,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.4167461681697551,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.3169,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.6326417532343238,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.3521,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.4755653961691437,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.3376,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.4481129346940663,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.3159,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.5947325024765462,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.3232,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.4434157448762071,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.309,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.6195291044028429,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.317,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.5109533276675696,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.3419,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.48567854564687984,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.3209,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.5724995655384286,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.3269,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.4290796781774386,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.3149,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.5142758106289028,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.327,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.47909758019829535,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.3275,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.4723510140056827,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.3239,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.5168755292872745,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.3408,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.5071412172771096,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.3218,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.4457247418690259,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.3498,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.5032757562516043,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.3341,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.4416686009282585,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.3231,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.41476990650080825,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.3364,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.4492298092691964,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.3247,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.39584589954792615,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.3156,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.42302826855661035,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.3384,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.48236357434491867,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.3331,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.3794644808215738,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.334,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.45890778220711625,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.3261,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.3686131428818877,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.3318,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.4537725368048026,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.3412,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.4558284624441498,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.3238,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.4133506402783624,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.3243,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.43315175484317525,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.3334,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.4322470544422062,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.3376,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.4817558636959325,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.3396,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.4370625873674432,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.3385,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.40467452387620645,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.3411,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.42152157991545086,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.3182,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.43085821167194843,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.351,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.42032402451159395,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.334,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.4747149560174886,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.3392,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.41108633176923626,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.3466,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.4922756944929064,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.353,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.4186174834772363,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.3279,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.39238193872670374,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.307,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.5035528423183314,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.353,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.38195353263418536,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.3391,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.4757706622522796,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.3202,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.44613717899162575,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.3542,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.5050751819265551,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.3307,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.4882512377779982,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.3497,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.47022354397045957,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.3331,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.4931711228858973,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.3469,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.3924992817227343,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.3419,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.48187831736264614,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.3372,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.41225774507855384,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.3509,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.4489360776809179,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.3309,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.4360509245392919,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.3342,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.4260933103213882,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.3322,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.43198915116205633,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.3228,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.5236211184779326,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.3407,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.39512031475672377,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.3218,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.4859915935784157,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.3417,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.4001223798562388,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.3514,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.5052903692485984,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.3011,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.39762522898564606,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.3084,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.5084496262196139,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.3221,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.46647915662832107,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.3526,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.46862760502402023,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.332,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.436377053220926,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.3499,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.40322569266150876,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.3419,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.41708847166048607,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.3233,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.4232942024674408,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.3515,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.41689826609406755,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.3495,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.3977517596962275,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.345,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.43354752564391213,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.3324,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.4366643492077072,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.3499,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.3971196709589341,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.3177,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.38144002928244614,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.3483,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.4593815266875286,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.3364,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.3928496485823181,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.3336,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.4214527187281214,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.3454,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.41970343068349075,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.3584,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.3832102830663753,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.3325,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.4002737600748524,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.3245,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.43215549555335336,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.3472,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.41786409315192585,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.3595,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.4556085007413393,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.3409,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.4240259252528791,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.3297,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.46794405054227456,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.3595,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.40651271304845515,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.3248,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.4992758656394574,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.3307,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.4669841846239505,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.3227,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.463927772524752,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.3363,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.41290786998999146,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.338,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.43235210645276884,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.3487,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.42899729249863744,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.3303,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.4047121117815278,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.3365,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.3803099468615304,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.3357,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.4302654310767862,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.3356,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.41358293239623617,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.3302,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.40007383498258653,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.3305,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.40903744622597427,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.3391,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.38329288970575187,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.3452,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.4200839291648381,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.3062,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.4987086200013767,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.3303,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.41577475691028026,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.3267,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.4412737664860633,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.3279,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.4971445842486023,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.3495,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.411581445535149,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.3259,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.4710415196449669,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.3424,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.42665091385099607,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.3319,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.4699259706349483,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.3447,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.3681728534214568,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.3317,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.4499961659558405,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.3518,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.3860777692634264,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.3365,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.44264978610455985,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.334,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.3699370500965797,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.3273,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.3942873880538376,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.3295,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.4028013625353162,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.3476,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.42555891421464426,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.335,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.4157994868050664,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.3433,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.40871862463646025,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.3225,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.40149229139236925,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.3049,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.4205219181134412,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.3301,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.4118468927701489,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.3382,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.4268761853335801,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.3321,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.4374375054680576,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.3242,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.3827812097108136,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.3207,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.4561222354470972,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.3348,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.4028234615926401,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.3208,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.43799589360934793,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.3395,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.4481206644719914,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.3276,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.42343071608802013,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.3348,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 1.1284544347366896,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.357,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.44936089082741826,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.3335,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.3565715089182343,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.3362,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.41449655863156365,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.3437,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.3663988135931478,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.3223,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.43937618518819715,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.3381,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.3814336433829766,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.3321,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.38223380350483377,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.3276,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.38533820793212553,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.326,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.4097106124012283,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.3391,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.3718478141489936,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.3367,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.44995692850672603,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.3434,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.3883389576093801,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.344,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.3874831878183108,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.3275,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.443218575787233,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.3253,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.3563972203038915,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.3267,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.41089560808696496,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.3167,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.4117186052720832,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.3361,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.4323812738168094,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.3155,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.4327357639425891,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.3234,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.39331705076435086,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.3352,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.44559330903784783,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.3433,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.4061387124378416,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.3194,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.38455068486401767,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.3038,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.4350902806295156,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.3294,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.3937866880767571,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.3279,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.3699595847100245,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.3651,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.3974602105690045,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.3393,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.3981124746450097,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.337,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.37174343675741334,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.3299,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.45153310974066013,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.3499,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.41446776312384803,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.3333,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.4074107222112874,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.3469,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.42062820413001273,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.3419,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.421213025021969,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.3386,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.4025474100909486,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.3255,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.4362302828184045,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.3329,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.4720109489526487,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.3258,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.3910701488708621,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.3295,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.42136600492742393,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.3297,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.4029664989290793,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.3511,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.40941778696490644,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.333,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.43943752229163435,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.318,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.43685045714629245,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.3299,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.4035878569537456,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.3453,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.37939236421584893,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.3282,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.41750292538797396,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.3299,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.384229632959281,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.3303,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.4341166049342145,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.3329,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.3961023722954063,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.3416,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.415054250827641,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.3411,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.4046476692886668,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.3385,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.45279067697574277,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.3397,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.47053757554695597,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.3242,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.4237399739042954,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.3414,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.45244923080781846,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.3541,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.3993970917010382,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.3323,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.46486424433463736,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.3283,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.4113107746789965,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.3248,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.41128469845778015,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.3314,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.4185480585421337,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.331,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.410316246818797,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.3287,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.40069771821482075,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.3381,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.41286389930128325,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.3176,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.3879703054504979,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.3164,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.3788458569214131,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.3369,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.45324429598370786,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.3572,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.38823373225523256,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.3484,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.4200638669012839,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.338,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.42285674048463745,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.3337,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.41656015389647705,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.3399,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.38675188436028274,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.334,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.38539437762495665,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.3225,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.36579671412533066,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.342,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.415382382573576,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.3442,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.3734824828299513,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.3341,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.3814795522735639,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.3507,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.34536778784675126,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.3298,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.3972776278187328,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.3226,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.41139043469004927,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.3346,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.3900554631154005,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.3476,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.4378785678332928,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.3043,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.3924466020011781,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.3384,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.3817041761477797,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.3159,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.4397346608467035,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.3229,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.3557741236583033,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.3464,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.4198998002264315,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.3264,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.399510865783311,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.3455,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.4002507587911729,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.3346,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.36399234556492055,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.3206,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.37377326478965983,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.3344,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.362806562902706,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.3264,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.4007226874825636,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.3244,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.39083863065696584,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.3431,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.4307861249042579,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3246,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.44558356298579105,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.3343,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 0.5733482034441348,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.2565,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.397567601801192,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.2562,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 0.9359429039044225,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.2629,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.4744561763359087,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.2351,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.4866684439826479,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.2523,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.4423942970706683,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.2341,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.40617186448769804,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.2468,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.46266341752200124,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.241,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.439501339189632,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.2437,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.41601233288166684,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.2419,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.43560466164053135,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.2196,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.4648383366697023,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.2262,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.4107514350132585,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.2403,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.4410456416034379,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.2359,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.4313383551444157,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.2418,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.39755083889570053,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.249,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.42905834792198844,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.2305,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.3695175390254489,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.2397,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.3779585056893355,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.2347,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.4193878941364583,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.2453,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.37895618254809327,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.2627,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.36948182980933225,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.2495,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.3958498042938425,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2321,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.38190502280846994,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.2217,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.36182278439665166,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.2488,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.4364455656162927,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.2387,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.3811929804489559,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.2472,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.37388404632421335,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.2253,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.43761356144051194,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.228,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.42090078632940464,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.2377,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.364965996869903,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.2379,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.42099756170658365,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.2426,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.4052485965876613,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.2404,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.3705649656399807,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.24,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.3935497696837462,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.238,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.3847596148076446,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.2556,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.3913402502574812,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.2298,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.37912115720037975,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.2573,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.4070855937102483,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.2131,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.3560199085405656,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.2367,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.416788178503481,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.2322,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.4029416244639351,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.2591,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.38228348910186155,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.2415,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.3998766454320136,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.2417,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.3569012045310857,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.2317,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.3719070614040113,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.2315,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.32724742089043657,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.2454,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.40946188471026207,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.2562,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.3439086974686623,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.2413,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.36951294918751953,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.2331,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.372651790454717,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.2438,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.4289609149016661,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.2507,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.34411825729981316,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.2376,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.4196683080867956,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.2389,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.3888022489946864,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.2394,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.34540712866108325,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.2221,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.38862274861298873,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.2612,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.35882762863923034,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.2468,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.3623274730027531,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.2336,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.37269026110711445,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.2164,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.37060052519284165,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.2544,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.37523543736578363,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.2379,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.34264167925812405,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.2425,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.3471928120827856,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.2356,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.3964951445570785,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.2391,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.3790313940907886,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.2518,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.3562855266764715,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.23,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.3929473705216026,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.2421,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.3906934393487318,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.2496,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.3542367698827356,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.2632,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.39383163822656697,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.2471,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.41076269722886855,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.2312,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.3557105965954298,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.2318,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.35870245706352655,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.2478,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.38193053094941337,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.2402,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.38152983941508073,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.2422,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.3880485851093977,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.2345,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.3615486377050177,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.2467,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.3972862456065456,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.2333,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.37757520094601116,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.2344,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.34782904706403134,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.25,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.38158884644469776,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.2436,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.35967688608304543,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.256,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.3663113950597336,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.2325,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.33090654850660645,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.2448,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.33073256387197564,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.2371,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.3344244380138724,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.2629,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.35481637561141,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.2427,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.3630667687395913,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.2511,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.3594232952387801,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.2517,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.33330142691821896,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.2323,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.36353295356716336,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.243,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.34489688844457167,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.2443,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.385186699816653,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.2415,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.3500985080585077,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.2335,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.364454905125375,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.2532,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.3597009035900142,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.2458,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.363945576299989,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.2253,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.37414394692092645,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.2241,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.3738178131726136,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.2501,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.35162881434347093,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.2468,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.35415604438676984,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.2457,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.36045789630740227,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.2434,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.37147132550252554,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.25,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.3185679559773856,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.229,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.37243601872178145,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.2724,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.40459636238703456,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.2333,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.3341720558874203,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.2561,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.3611698750689612,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.2396,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.3547048616499007,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.2259,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.3545218398554794,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.2367,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.3546964105034664,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.245,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.39617935209425265,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.232,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.3578710347335086,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.2607,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.4040733334600705,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.2626,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.3752427458635307,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.2319,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.3564071329311155,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.2449,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.4051415539904142,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.2482,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.34199001256860834,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.263,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.3469816060965267,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.2296,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.3364086424135848,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.24,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.350429452860175,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2353,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.3499259754958485,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.2404,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.3619129431289598,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.2508,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.36011012558996996,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.2614,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.40656925747428807,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.252,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.3456963779719087,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.2426,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.37412594798183507,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.2559,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.32713676677023235,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.2518,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.3235618009018623,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.2468,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.3737292664795582,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.2395,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.34360418406617627,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.2519,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.333001622522385,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.2525,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.3458124698769787,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.231,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.3502530322285603,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.2561,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.3340030895018751,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.2581,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.3455788140342786,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.2555,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.3491805488567238,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.2435,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.3206603944479966,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.2304,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.33840882565333813,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.2338,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.3368687512782083,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.2425,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.33380078244850286,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.2269,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.33328751263481493,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2424,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.31434082924265067,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.2519,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.33828794688698116,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.238,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.35859575289725404,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.2196,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.33640002900189314,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.2475,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.33353860697560667,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.2338,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.399560792508508,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.2511,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.3430868206357577,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.2362,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.34832690792860665,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.2328,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.4116031920296893,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.2494,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.3397743966684084,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.2378,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.4006106318672933,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.2476,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.36514775462943083,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.2341,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.3568950850145061,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.2319,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.38495055270380696,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.2458,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.4057508574739205,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.2284,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.32184984891193624,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.258,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.37332089368324906,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.2335,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.3528038634143796,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.2413,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.32445076980456367,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.2269,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.3483448305010462,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.2374,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.328787719088629,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.246,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.3892106238165899,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.235,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.30818928053299105,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.2318,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.3258458798509179,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.219,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.3530020540626817,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.2769,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.3352573589764864,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.254,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.33365730344537103,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.2141,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.35249242355022514,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.2585,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.347629375774613,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.2422,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.34499124215646176,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.2502,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.36349258364679116,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.2411,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 1.5987660950042564,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.2543,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.39427707488733443,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.2361,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.36436972975314663,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.2268,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.3674405054081652,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.2528,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 1.1796021622448674,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.2409,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.4001380882478764,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.2478,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.34889595409133967,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.2386,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.3349519788026442,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.2478,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.3634482415078469,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.2514,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.34448058215312777,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.2414,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.34256167554924155,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.2648,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.4042007722571759,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.2345,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.333929296945041,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.243,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.34183180468504215,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.2372,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.36863307350037827,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.2636,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.3239872670324215,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.2464,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.36682861274460493,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.2598,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.3226813488513336,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.2609,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.3422982314189772,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.2379,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.34812190928032627,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.2288,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.3385617243339986,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.2368,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.32006184408040456,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.2211,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.34219372800679587,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.2183,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.33720552099714024,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.2297,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.33581832545604573,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.2435,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.3216121444173283,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.2559,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.33222730152329877,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.2422,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.3383619470080389,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.2617,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.3358668570179573,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.2313,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.3364813057655903,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.2629,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.3190925270443395,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.249,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.3629564148847516,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.235,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.32618731593779815,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.2295,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.36660208360847046,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.2405,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.3561935175693026,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.2577,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.3259586934769585,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.2355,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.37949777458718525,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.2459,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.3305130271448903,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.2273,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.30354994774905036,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.2543,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.33965350903337294,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.2373,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.33731763511855706,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.2347,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.3521326951881875,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.2404,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.3242086519360183,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.2433,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.34833656985823575,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.2466,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.3469086490355028,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.2465,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.3226213797714079,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.2315,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.34863142723445,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.2303,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.35027680247571585,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.2406,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.33475241900895253,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.2306,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.3386645563940142,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.241,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.36836600108335865,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.2386,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.3330626224227087,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.2465,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.35864907029972154,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.2401,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.3491112557069058,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.2406,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.346660120937888,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.2286,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.3361763983870647,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.2461,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.36775384224026597,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.2404,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.34242283059015277,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.2368,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.31616775817822773,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.2312,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.343582778670758,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.2433,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.34409595470341325,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.2359,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.3170511515650247,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.2414,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.3112921398765536,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.2547,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.3372159737515423,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.2285,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.35240574594020113,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.2423,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.33531064514720216,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.2335,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.3348816653892991,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.2123,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.31391461095362205,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.2391,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.34835679102064654,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.2347,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.31683115921905225,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.2404,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.3509725472784912,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2426,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.3217638559415799,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.2467,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.3082805060461396,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.2043,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.3737795545652187,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.1575,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.31771118149738,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1708,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.434586217432909,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1746,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.4688691694007289,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.1668,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.3520165176809714,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1526,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.391988432446571,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1651,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.3776768478182973,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.171,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.3295567255106475,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1692,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.38808648438295473,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.1737,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.3602628152155221,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1838,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.3532115234601053,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1825,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.3433340715165645,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1558,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.30275031278247205,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1754,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.37448396988105564,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1521,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.3127590154123285,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1548,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.3305293047250143,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1477,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.33622743952419404,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1483,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.30988159973038343,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1729,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.3378340628737843,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1698,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.33078087027855624,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.186,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.302683321057565,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1857,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.3129418230041575,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1731,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.3058014236367372,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1616,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.2882368778341601,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1891,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.31073737997814666,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1697,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.32211082143732533,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1691,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.30453746947166394,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1714,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.3084719460107315,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1666,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.31182247539744445,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1822,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.2942124504854903,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1477,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.2927028656862569,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1615,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.3038908341127498,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1711,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.29932591600071595,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.1727,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.3065957344463919,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1547,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.31625160088588655,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1681,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.3173097805257988,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1639,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.30809740375772887,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.152,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.3147980913360742,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1691,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.32058506409147225,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1721,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.3242542835821205,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1565,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.31744727066577827,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1504,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.3060431606595756,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1615,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.3054696329591124,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1673,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.31622078918441715,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1815,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.29673417422175163,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1533,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.29232195188911686,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1678,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.3069410153710998,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.1563,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.30647240354958805,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1611,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.30901662736086144,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1652,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.3042886949112334,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1567,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.2955884750830877,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.1714,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.3093542911260947,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1786,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.2992105069118941,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1647,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.2729368204361087,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1691,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.32016568487306796,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1753,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.30774743983013797,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.18,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.2905622437860254,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1693,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.29587466200535933,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1455,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.29699848970079623,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1493,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.29007927962021085,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1786,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.30142758102401024,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1708,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.3050495745883441,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1646,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.3096421231541482,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.1588,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.3087220848166549,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.143,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.2888684750254325,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1822,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.3085551841832258,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1914,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.29285859085212884,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1718,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.30752143017398703,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1717,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.3048999250477642,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.156,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.30801429949605913,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1705,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.3080078395755238,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1494,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.3090211078254331,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1605,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.2877787497684695,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.168,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.31433556607232815,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1742,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.28445958042433817,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1609,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.2969461084410041,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1744,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.3092337873389138,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1649,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.30036286791877614,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1725,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.2995944688766452,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1603,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.2961974017854544,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1501,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.31049938852476855,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1593,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.3102058016037818,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1665,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.3045792748497218,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1702,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.3182151044953695,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1683,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.2938406130940762,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1509,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.30886520552933955,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1714,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.3124775039881817,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1509,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.3003074117143575,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.1722,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.31021154250093497,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1695,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.30786479542149486,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1751,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.2874577168284571,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1694,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.30437990000825654,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1702,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.2884010415712962,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1765,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.30200132472808866,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1441,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.2941163944106803,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.175,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.29352027539916353,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1706,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.29906564838134175,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.1752,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.28813754298643945,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1672,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.29860105099941836,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.152,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.2967689409826107,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1517,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.30664403001100154,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1607,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.3122820379571278,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1635,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.28950574939327006,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1618,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.3093798644616152,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1665,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.29573880588657414,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1834,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.3009169249879688,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1628,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.2929716703566274,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1823,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.2860485779400074,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.168,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.2984759812740656,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1781,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.29887035696218284,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1599,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.29045005414570974,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1665,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.29538489873159435,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1591,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.2945953102493551,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.171,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.2944507372396471,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.153,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.2775363829439252,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.154,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.3033641307162131,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1653,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.2909440638386303,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1591,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.2760433732163866,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1979,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.2886356873896863,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1655,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.2761041785586394,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1765,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.30282491177094173,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1608,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.28138249709754426,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1677,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.28125221604274864,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1613,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.28573644579340635,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1623,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.2902543199849343,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.17,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.30058530177500326,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.1649,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.27417586448025655,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1609,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.2748756517804535,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1581,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.30624621146899866,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1438,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.28726744451915487,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1589,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.2726011045991505,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1777,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.2833229262207905,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1486,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.29331679933775745,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1435,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.3056807348522248,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1844,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.29105158028110856,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1517,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.3138443124040847,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.1613,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.28784035322193696,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1577,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.29417357958830187,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1667,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.28606969721267006,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.177,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.2988818010429278,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1648,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.30206161643356283,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1775,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.30739862416349684,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1685,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.2814003200194986,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1556,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.28892072394050217,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.1656,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.2794474766982607,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1505,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.2929181204896763,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1726,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.27516445116826577,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1763,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.2929829004402186,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1492,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.27738094268449026,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1694,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.28847722417029814,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1717,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.29152625661498394,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1577,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.28191708291642,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1485,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.29402989288123926,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1528,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.26683069237487966,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1867,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.2817158925928116,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1741,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.27861772921050154,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1744,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.2775441220378693,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1739,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.272894278932011,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1701,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.2835381951351384,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.1609,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.2872722884021677,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.155,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.27850344330120963,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1527,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.2793989332322497,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1595,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.2795516566273682,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1585,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.28172876960268717,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.17,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.2749313134807833,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1484,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.26698139391873815,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1974,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.2734796558751321,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1442,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.28499408472797866,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1416,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.26959683616912145,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1602,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.2761606719518361,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1591,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.2709508938885257,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1586,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.27441991953518335,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1808,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.2945580811777501,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1447,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.27457926602202676,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1544,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.2806424476649439,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1601,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.2774476509868807,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1555,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.2724381910399199,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1654,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.28411788167051416,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1545,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.2694850307075435,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.157,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.27152383670754393,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1747,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.30124282277950126,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1686,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.2827878306408377,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1453,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.28593036940093286,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1652,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.2903772956546489,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1501,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.2807733685162814,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1564,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.27483294455974594,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1527,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.2670725753788277,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1842,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.2814451417912293,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1673,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.2612003687609873,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1685,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.2738240554045644,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1669,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.27224382214732096,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.166,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.27241457496804516,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1709,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.2658805011127306,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1541,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.2730435875287565,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.166,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.2667958412761635,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1409,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.2860261550703979,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1583,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.28720583657513565,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1648,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.2813770243881661,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1721,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.28150073177089696,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1405,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.27214158372970393,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1624,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.2807279701239281,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1543,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.2793881121252571,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1583,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.2705978814109788,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1616,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.2937869349745439,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1606,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.2704549116452371,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1524,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.2613989998807937,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.174,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.27920039703420113,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1394,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.2874311068445236,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1535,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.272936370077748,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1656,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.3012771175878465,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1401,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.2675514182437548,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1692,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.2687746974088081,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.1541,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.2732196761026443,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1549,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.2727785077684225,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1561,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.29689398678489554,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1614,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.8892252904299419,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.169,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.2704453494523041,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1711,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.2880841178483063,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1809,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.26789038631338563,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1731,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.26753063548381645,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1629,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.27352249941870926,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1604,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.26458238131989437,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1634,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.27167828453677706,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1632,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.274284851649774,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1651,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.2844011875813177,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1736,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.27343099119341363,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1547,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.26424404507662125,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1744,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.28033538271303654,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1611,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.2711549968177438,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1655,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.2619393885810456,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1556,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.25639498346781137,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1685,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.2748971957356019,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1694,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.2731023597198101,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1659,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.27542222580993714,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1574,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.263637466897985,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1721,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.27519092919705357,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1604,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.27723352604436546,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1564,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.27404867116456383,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1519,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.2734427063940602,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1612,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.27218513947342027,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1664,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.26913446788769013,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1566,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.2689854064989129,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.159,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.27153101813365343,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1588,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.275942858644361,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1503,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.28222832760388994,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1817,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.2565981017913237,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1629,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.3473502143553144,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.1045,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.2847589435203046,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1164,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.2522635749139119,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1116,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.23232386251283554,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1046,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.24552761272962143,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1218,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.290291911295569,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1075,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.33819602505022783,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1309,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.3125642197284788,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1225,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.29175874424408027,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1453,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.24023284311881246,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.1546,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.2628536733220237,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.0925,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.2756715556902154,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1317,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.2805452051127995,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1056,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.2347106815478855,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.105,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.24480172321432192,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1234,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.2356115554358658,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.0982,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.237294795147572,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1149,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.24462593381995795,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1399,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.24460415303539967,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.0937,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.25214193281039693,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1247,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.238711283980518,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1201,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.2357654735684518,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1338,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.2331858812305692,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.127,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.22141367786124694,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.11,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.23267817223291423,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1212,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.23255822092567366,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1105,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.2298111639336183,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1312,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.23015226018401555,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1212,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.22313931645510032,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1147,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.22477278654009547,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1025,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.23829357750318908,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1158,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.2257172940763584,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1456,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.22484459832767034,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.0939,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.21585107845811863,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1145,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.22127918722563253,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.1037,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.23105230958848127,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1174,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.22422631502467194,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1118,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.22459532988135877,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1302,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.24512625162360552,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.136,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.21825123886793982,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1003,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.22868793354146055,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.121,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.22195945400803277,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1245,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.21799457081094145,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1167,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.22227047204423017,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1131,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.2268993201794593,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1421,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.22659580391340142,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1268,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.2252113885979066,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1316,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.2270676413973957,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1178,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.21994061105626989,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1517,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.2072549885058109,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1155,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.22556930747873577,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1195,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.21102536604294073,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1177,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.2135584553727959,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1238,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.23300453683638317,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1256,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.22781514984232037,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.119,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.22007376239215756,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1006,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.2205600897168326,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1217,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.21393037853578006,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1269,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.21768298484190873,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1052,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.22708371573403616,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1064,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.22157476535520534,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1056,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.2151149824158524,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1345,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.22031618423821578,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1395,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.22909080117412542,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.112,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.22966761544643688,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.0992,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.21744894331993586,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.123,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.22887179173845745,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.128,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.21471847949518974,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1231,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.2174304761739108,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1209,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.21709676953593907,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.1018,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.2136120751647193,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1208,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.22509157709124322,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1072,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.21507353735719606,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.11,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.22276227462842857,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.115,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.23180463311826868,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1196,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.23211077133987865,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1163,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.2216888779565285,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1151,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.23810349690296354,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1139,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.2198833533419359,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1258,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.22193516040637537,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1156,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.22880906445596574,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1166,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.23294415296697787,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1258,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.22850966357061311,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.096,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.22023341849084602,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1152,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.23232681943869682,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.111,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.21178940588629053,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1196,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.221564290808231,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1087,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.21902912684926074,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1248,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.25175813509331657,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1239,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.23040119672853135,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1079,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.2247163312475984,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1023,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.22138001876258537,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1115,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.21057003470273866,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1165,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.2193130859374176,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.1108,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.2253925878881806,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1151,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.22128042390949668,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1223,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.2156955657698815,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1303,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.2239462603620744,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1384,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.21816026253110307,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1171,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.22041117093943458,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1404,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.22857702802582142,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1128,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.21596329560534971,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1387,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.21351990903993914,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1151,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.23262062286169372,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1164,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.22804932283836077,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1321,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.21412374998295122,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1467,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.22176238107055424,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1225,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.21203804845148935,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1075,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.22330279221542865,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1275,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.2302759463044603,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1286,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.2180779408787249,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1061,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.22532925905220016,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1166,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.21315200957616082,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1134,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.2343586536915463,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1169,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.21637977066568675,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.132,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.23880598391550936,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1369,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.21735156546907713,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1134,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.2132540516253041,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1127,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.2072715788984184,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1208,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.21363740970439465,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1004,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.22342923383906194,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.109,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.21936403210992594,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.123,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.2112017279668416,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1204,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.2146944239583926,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.109,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.24624647707131517,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.113,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.20738181697730268,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1082,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.22160677125019182,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1077,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.22216779579604726,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.128,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.22180236064895636,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1333,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.21436724548080588,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1154,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.21904491114869115,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1347,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.2364273600050615,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1282,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.2180852209089089,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1079,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.2241288712118843,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1337,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.22039859437924791,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.13,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.2193213741835611,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.11,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.22438658681201037,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.0992,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.2087026533578337,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1286,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.22290799306913994,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1074,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.21848504192466967,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.1233,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.20941466969501743,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1296,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.21980446574234055,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1155,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.21711528889684079,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.135,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.2228958844366628,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1385,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.2317817507328144,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1337,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.2334387147679599,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1154,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.22615234542557275,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1159,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.23005952174879798,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1201,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.22205605231621892,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.113,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.21859778645989245,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1132,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.21370344914191552,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1022,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.22002438836204724,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1197,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.22976120947747095,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1229,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.21983494099773382,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1212,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.2093462255602368,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1143,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.2154449948018785,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1092,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.21803153824308047,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1213,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.2125802255296246,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.121,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.2091889960131764,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1063,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.21371861838813783,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1491,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.22717333973408607,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1129,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.220532709496273,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1124,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.2168229990695833,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1334,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.22302382529706255,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1237,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.2198069904663175,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1088,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.22236571577441008,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1171,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.21260209412457975,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1219,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.21105960986783445,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1128,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.21266427086331854,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1191,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.23153715733149888,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.0915,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.21322419599718692,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1116,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.21896176096023776,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.134,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.2133660877306697,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1012,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.22459553139505473,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.1114,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.21894716293139943,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1165,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.21909736285564566,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1178,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.22454105164746338,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.1229,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.21316434946754576,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1129,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.21365564875310417,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.139,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.21424576429331502,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1076,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.2247894264237759,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1298,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.2172240032907659,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.0979,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.21943534467202344,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1234,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.21716981311814962,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1044,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.22445666966639224,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1384,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.21158205565064483,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1516,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.22327556913109742,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1242,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.22278529346607376,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.1254,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.2062585620628424,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1207,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.22120987186221122,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1417,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.22284089883304656,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1179,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.22186077324080355,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1241,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.223931962320012,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1216,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.21627530179544543,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1184,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.21826345565819477,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1074,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.21898563135441987,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1204,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.21887917975307997,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.0946,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.2259259601827037,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1176,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.2174066668386943,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.125,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.22489768498345486,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1422,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.2124456140067316,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1317,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.2216916173344711,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1341,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.21475253888857054,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.127,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.21243022039126347,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1305,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.2190154548855226,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1382,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.22105731444831006,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1256,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.23000873828474933,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1243,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.2050153919664028,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1149,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.21479302619512544,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1207,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.21902709604293064,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1076,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.2216363305778955,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1052,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.2154958879752841,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1116,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.21430951571474174,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1225,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.20981391569003618,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.133,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.2147318299631988,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1253,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.2239015596429572,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.114,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.21640983991778806,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1128,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.23708056415065065,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.111,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.21636378719146834,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1091,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.2093939748799308,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1283,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.21217111284584697,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1112,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.22211323168726482,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1171,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.21673687104069222,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1213,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.2150852004120508,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1101,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.2143671308732413,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1159,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.22115654034754437,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1232,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.22598429991035546,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1129,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.21604448063651122,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.1011,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.22156774943157423,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1137,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.21848894412882977,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1122,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.21026692405228647,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1407,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.2081102092451164,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1192,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.21167995308882276,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1057,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.23272244741057188,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.121,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.21982652187089946,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1338,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.7269271309817587,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1334,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2237742498551278,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.0988,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.21175592605488835,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1181,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.22794513496848332,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1152,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.21676263767988727,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1276,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.21534565866530123,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1152,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.21308882860160636,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1333,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.2107374578869834,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1313,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.22793837813782036,
+      "learning_rate": 0.0,
+      "loss": 0.1338,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 2.0620707338356e+18,
+      "train_loss": 0.2627544824609427,
+      "train_runtime": 84719.4901,
+      "train_samples_per_second": 1.865,
+      "train_steps_per_second": 0.015
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "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": 2.0620707338356e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}