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": 4.996455696202531,
+  "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": 6.465711266413276,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.0365,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.607219071018812,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.0548,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.620207299036213,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.0462,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.556096980852716,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.0635,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.131320896344567,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.0462,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.874883366691952,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.0301,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.787446417041451,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.0052,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.591115579918033,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.9961,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.7961445003166316,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.966,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.6157172400130855,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9674,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.4732770848135748,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.9543,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.9493826066198388,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9418,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.247558013487754,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.9238,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.137778561649873,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.9301,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.804186384886837,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9239,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.8158131117774903,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.8735,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.6003201911914426,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.8774,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.081933577749037,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.8482,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.5640562094921773,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8313,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.2552065904804715,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.8,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.2318538226306897,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8039,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.3550559132797642,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.7872,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.1504573479797011,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.7848,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.1527133673520173,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.7577,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.9278841250092421,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.7587,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.8706637344201955,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.7579,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.8447502722313452,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.7419,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.7499414297783786,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.7383,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.6809807486528612,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.7414,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.7130805913159811,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7056,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8267746257729714,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7333,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.6541902930406758,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.7092,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.6039349584547471,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.7103,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.6150496647489553,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.7114,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.693493887162491,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.6959,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.6360786240397739,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.69,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.5678314106363934,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.715,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6877682960906321,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.6916,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6186823288462386,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.6897,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7864885274540604,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.6903,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.6754301278042151,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6803,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.6802842694887583,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.6731,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.6541082601672414,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.6885,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.6846014460939492,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.6814,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7042185912164343,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.6607,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6109220928956943,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.6713,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.8256193309358218,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.6736,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.6930046873845234,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.6569,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.6663562910525779,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.6537,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7984724924054366,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.6633,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.6096841642140087,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.6594,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7707678415298873,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.6662,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.745163470021058,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.665,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.5808723802763294,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.653,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.771566559700636,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.661,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.6033007545166813,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.6584,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.7360025875122352,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.6411,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.6272721064118025,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.6266,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7040058976752948,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.644,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.8182387707638039,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6257,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.7751970935581758,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6481,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7366370923184147,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.6422,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.5917514420203566,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6443,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.8199447704718442,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.6294,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.8210335590916825,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.6465,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.8152559748684239,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.6239,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.6310799039922009,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.6362,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.85576962878205,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.6394,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.029761364638883,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.6361,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8266675133955985,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6256,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.9029190747104121,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.6256,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.9003183143425921,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.6309,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.0614135154877307,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.6371,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.8085603112999976,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.6301,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.7282470227307456,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.6289,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.567127536382807,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.628,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.7645577773426411,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.6344,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.8830312450938312,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.6258,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3920496921011205,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.6075,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.741709539690981,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.621,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.057963441368092,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6184,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.7316366742228677,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6298,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9157014828111576,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.6143,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 2.564946189084993,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.635,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 2.0481034678833594,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.63,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.785002109462161,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.615,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.9536019463540357,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.624,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.4636833791459531,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6206,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.3845967704462083,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.6239,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.4332999481796096,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6048,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.0829079410409814,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.6063,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.4550869424312227,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.6226,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.8101899688687995,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6078,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.4714831065744018,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.6159,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.7309774021124482,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6107,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.5010042317113936,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.6145,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.8486597818107885,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.6198,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.7008802817049966,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.5989,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.1065896842978427,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.6098,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.8664415302876374,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.6319,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.6262441714287155,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.6329,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.4604619747553487,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.6245,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.3776491693495405,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.6128,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.3975702607663594,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.6053,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.158159472492283,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6068,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.3047732040404871,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.6043,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.2226501607322409,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.63,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.276638776043817,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.6202,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.8974680253616613,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6026,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.5308073185809752,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.6139,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.0659205460728776,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.6175,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.5594159418368452,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.6017,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.2964398098179089,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6087,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.253442468272072,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.595,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.1230603633090004,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.5991,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.4033558293618118,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.6236,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.0974040901224473,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6145,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.3333684620127388,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6097,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.9746188624723459,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.61,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.5484357385688845,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.6054,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.1362808929625952,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.6071,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.5199827175756533,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6026,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.4167742094618387,
+      "learning_rate": 4e-05,
+      "loss": 0.6098,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.1935807957198288,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.6143,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.1179302769705879,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.6112,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.3668212911933715,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.606,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.0149920656609863,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.6066,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.0936464460570539,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.6018,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.9081902829905248,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.5946,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.3477099986300594,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.6059,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.1621472792644807,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.602,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.8936541859041591,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.6136,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.165414165208242,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.602,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.0898178586550542,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.6166,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.9173737322102044,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.6045,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.8788866266895952,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.5976,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.048615924811541,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.5945,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.9541683300527272,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.6025,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.8215325949379723,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.5962,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.6767148616984162,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.9174134149801051,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6218,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.9188682501975907,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.5949,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.7277914166028492,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.591,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.8563885222597342,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.593,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.1898227549815557,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.5959,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.7425739867757253,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.6109,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.9561985836464357,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.5989,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.8020436501577503,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.5857,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.8395224988464036,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.5884,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.8728186600825415,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.5958,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.7852348795595929,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.5962,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.7425533780296052,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6051,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.660340623233992,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.6045,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.7199966411354108,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6065,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.7839210883347717,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.5954,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.992622903014384,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.6041,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 1.451658046528815,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.5993,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6846645700746701,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.5759,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.5610138611561186,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.5988,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.7375476441072767,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.5978,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.4992655574051332,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.589,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.7869338391406451,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.5909,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.5508040206681721,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.6098,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.8910658778467502,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.5929,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.8002597765804884,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.5996,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 1.4593426790851913,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.5886,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.1868605831272738,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.5991,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 1.4317425032501625,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.6121,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.8066178678020944,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6047,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.1948584526212094,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.5974,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.7691763316374775,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.5965,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.9496579581656424,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.5793,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.8984994629639325,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.5835,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.8381471755484526,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.5887,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.7740566717536295,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.5922,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.6597720114552952,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.5784,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.7606952725922187,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.5882,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.631178964693463,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.5956,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.6544875786672019,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.573,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7115742006382975,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.6045,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.8255260220533697,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.5904,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.8004251206951006,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.591,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.4876283323456751,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.5923,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.6978275350100326,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.5851,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.5919946343910951,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.5977,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.6156282063740146,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.5801,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.5248910398040175,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.5891,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.7028176524224351,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.5797,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.6178753787807173,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.5956,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.5111013850244029,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.5761,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.5942105496596377,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.5882,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.5082435167455861,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.5829,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.6229325295359238,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.589,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6093576351478974,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.5823,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.5200094084332099,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.5949,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.5344520632646615,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.579,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.5842176727050034,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.5795,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.5077444756837567,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.5783,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.5540671728497173,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.5705,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.5035242601242992,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.5859,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.581313086495215,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.5845,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.47618206620889003,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.5727,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.738036758447841,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.5981,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.5156101976453624,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.5991,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.5794175700485699,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.5923,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6109206984399284,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.5808,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.4664645091240642,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.5798,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.496648224899303,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.5633,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.5081574328276751,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.5736,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.43899126466725025,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.5976,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.5432315661908386,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.5973,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5252768406344567,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.5709,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.5003712374597047,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.5912,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.4850322158134018,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.5671,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.4386700067332097,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.5805,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.46692473535054924,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.5782,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.6034346559474051,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.5777,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.5762824548116016,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.563,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.457266766413901,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.5755,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.660329213849558,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.5789,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.5826210948067808,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.5643,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.5634043480405045,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.5593,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.5064756218160257,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.5797,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.5017288693507665,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.5697,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6143628903880458,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.5711,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.5541161587743463,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.5717,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.6999580422782304,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.5819,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.6828984917443417,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.5621,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.43675679854591026,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.5703,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.5791554948887311,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.6012,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.478694898441376,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.5628,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.45212809894076306,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.5549,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.5240416136895056,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.5566,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.5500068862125738,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.5877,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.5916344409104826,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.566,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.5572217634882011,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.5849,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5610149849894248,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.5618,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.6012118231180207,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.5895,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.5419774303083599,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.573,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.5834261509604727,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.5778,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.7442040505426215,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.543,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.5608360766339122,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.5766,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.5336694824144989,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.5658,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.48279377714006544,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.5753,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.5618559399208866,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.5757,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.49276617185991417,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.5594,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.5290184171674746,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.5927,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.6493735458647051,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.5455,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.4423538773141171,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5299,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.6506574398565668,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5196,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.5308939058437365,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5071,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.4741060943727034,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5079,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.5301385807615921,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5214,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.5371163813474019,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5297,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.49366152722571016,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5358,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.5024180978508569,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5241,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.4413231762649568,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5356,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.5258732918135207,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5368,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.5223147241934437,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5296,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.5375163480586942,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5147,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.582035444410161,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5245,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.5273595031002124,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.52,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.5611040621506325,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5263,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.548844449505319,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5366,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.5665812542711354,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5376,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.5642241486837152,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5258,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.485652767695931,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5082,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.5912941410204501,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.516,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.5834593170051962,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5185,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.7529229690612559,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5346,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.5601262480269764,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5057,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.6127290942864844,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5269,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.6907736138558027,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5238,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.46215322389010405,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5216,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.45226082965529246,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5042,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.5577890818378725,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5135,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.4667100875496768,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5215,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.47189287444040934,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5201,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.5213048767315231,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5082,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.4690185523118329,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5182,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.4412497589313711,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5181,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.5086205290508107,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5183,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.4706183995427797,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5235,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.4901334845785556,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5204,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.48472784672918673,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.524,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.45223106285564524,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5076,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.44361583062084975,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5414,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.46308709300269746,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5285,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.4781435961643036,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5182,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.402661684976389,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.521,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.3991704131707959,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5161,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.4597804399552961,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.522,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.3998577316485106,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5225,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.394153842452445,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5217,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.44691715916260105,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5297,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.38876905706867065,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.521,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.4894078310002364,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5107,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.5180639846592631,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5119,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.5439381666497349,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5093,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.4039275469248173,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5353,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.5473589377921226,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5256,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.4288356844679214,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5203,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.4735106390652167,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5123,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.40896717561034196,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5305,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.4410349762800099,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5319,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.5079250744281052,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.532,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.44076579739864324,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5264,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.37537943201235374,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5128,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.41237393987607446,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5275,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.363316500815416,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5244,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.425014925953578,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.518,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.43114263100455175,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5105,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.45819532316668937,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5171,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.40316436551027807,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5227,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.4863454548632279,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.519,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.43462699394574655,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5137,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.46135322544580204,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5185,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.33319442450047326,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5185,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.4582132133550959,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.5226,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.41660984473804563,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5321,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.3912135991639056,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.517,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.40600208187727943,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5182,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.41997997222874994,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5122,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.42949084628202583,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5133,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.5034292336854888,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5303,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.44777770771112657,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5183,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.4494952214390578,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5185,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.49373792181046083,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5077,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.5740923873817594,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5167,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.4778117533448908,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5191,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.46826018573367906,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5364,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.4606123536252319,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.5144,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.45593332486800875,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5152,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.3901729913076873,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5213,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.48771665937652264,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5249,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.44519677339922975,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5156,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.4296208176814055,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5084,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.45291997700127745,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5334,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.399787583123142,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5167,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.4483593769988753,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.5332,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.3874809786905489,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5359,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.5258964348001262,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5409,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.4889572635732261,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5225,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.5096080115309859,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5253,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.49021356254439913,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5038,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.4053936578491569,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5176,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.5788385392566552,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5262,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.4528226806227029,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5229,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.5074027331392565,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5327,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.5270169604176579,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5184,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.49445816695164035,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5232,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.4313958437153428,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5134,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.5386701134199456,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5213,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.5210694071798342,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5021,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.4228120964364825,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5281,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.49172696495619783,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5099,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.5276846004544801,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5236,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.4281751179474435,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5179,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.5214554496574264,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5119,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.5088371127407566,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5321,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.4439144326648592,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5272,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.5048886733777868,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5191,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.4211009848580178,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5198,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.5029434137213677,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5134,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.5786679650174797,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.51,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.3938606402798377,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5304,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.446610595175933,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5227,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.452942002201142,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5198,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.40853570263672806,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5229,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.4513239944610049,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5311,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.43452583558388747,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5243,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.4430162289886615,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5363,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.3707738936616716,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5121,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.42315188731383985,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5257,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.34629458790436457,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.508,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.4334101912477126,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.4996,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.3699230458055487,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5349,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.3660546146935754,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5135,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.4508123994464019,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5254,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.3874395683904648,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5391,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.3582486450891201,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5207,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.47566465379884365,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.4952,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.41938712152980134,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5205,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.39823312640533093,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5129,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.44814767983870524,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.521,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.33702774576110694,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5139,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.397338155928512,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.529,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.33112851148205924,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5201,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.4238706508028565,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5236,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.3542582408967708,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5177,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.4163319016211282,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5111,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.38278127040267035,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5281,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.4377978580750787,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.513,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.5041381317638495,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.515,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4148522318060153,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5175,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.4097453751915311,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5266,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.42784557656201727,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.5105,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.4079253527764946,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5154,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.4241088902248316,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5133,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.4065789100718314,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.522,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.3901987875294354,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.52,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.48373321850382406,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5247,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.46674112523116973,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5332,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.5838334281426452,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5233,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.3730617314558177,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5286,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.5306473318863888,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5171,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.454424966803546,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5093,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.5025719422864541,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5257,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.39635847363491095,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.529,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.48220979672765535,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5087,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.489077714159541,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5156,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.39274172071680935,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.4895,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.4771489071598694,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5007,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.3925381311291163,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5245,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.46009897150960916,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.508,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.4181750545399989,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5289,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.43741013614131335,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5186,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.41537702741358823,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5213,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.4055845074683972,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5154,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.42771034754417875,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5263,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.5071929368898419,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5051,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.41057639370197313,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5184,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.49141691367401624,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5201,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.4504089664731233,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5085,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.4469135463069614,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5075,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.47272061703466406,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5304,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.3872908213100151,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5046,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.441658393276213,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5134,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.4134891864789312,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.5101,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.42860509674713754,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5105,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.42861264638496643,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5233,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.41938809846522807,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5129,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.40504155271311815,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.519,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.36008871765396144,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5207,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.38405913235333194,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5022,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.39068144727326637,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.4998,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.3783640007773494,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5173,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.38153769889977307,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5185,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.4404057238387961,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5158,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.39649148726864747,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5131,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.45859771537497007,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5125,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.45394512042615004,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5246,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.4161528606986056,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5172,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.43118385646456925,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5087,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.3914589229373055,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.5203,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.4151560090524253,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.5009,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.4283393945471666,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5184,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.39917796417050416,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5206,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.39237157284292695,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5224,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.37925889289768333,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5068,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.42385173096809065,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5145,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.32893011582058707,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5119,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.39496175573086156,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.517,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.3584901386136566,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5202,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.341333954103923,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5087,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.3711416637568774,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5257,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.35799789773599194,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5079,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.4390189689593501,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5092,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.37624466734988254,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.4938,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.4178568304997194,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5315,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.3468470478657575,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5054,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.3886147906102647,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.5203,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.3200729661799345,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5009,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.38484625912446807,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5242,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.3528088225884382,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5119,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.40662566133729344,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5127,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.36000311489130254,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5269,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.35543092446189706,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5059,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.34954827539499805,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.4886,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.3598518621349006,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.4856,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.3686302598607585,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5189,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.3244270185915664,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.5173,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.4087429721342674,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.516,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.29288334240425207,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.516,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.38530990484345173,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5161,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.34650737783144897,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5056,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.3279940644326692,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.513,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.3592904690946271,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.4974,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.3761431038093163,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5206,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.3287720756831729,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.5055,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.3119195226736228,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5079,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.33691096226960565,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5271,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.3401838873595811,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5107,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.3587443096845678,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5129,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.3131507088356801,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5187,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.30299229662434785,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5106,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.3259840865409233,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5197,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.3278235977703117,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5155,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.34281746100542637,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.5177,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.36227884264976495,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5208,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.37120831152510214,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5113,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.45527511004553023,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5104,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.4828732167144543,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5228,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.40473057210273183,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.5058,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.4767843158260735,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4504,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.5519199177621151,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.4536,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.5572677864939767,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.4716,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.4697023992976325,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.4577,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.48762547911258736,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.4487,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.4502243009116827,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.4392,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.4312074666050013,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4489,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.4642049553517843,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.4444,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.38539606465874127,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4493,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.4483364803818987,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.4468,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.3654005215682684,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.4451,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.41206362487922893,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.4486,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.43005756060099815,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.4562,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.4018889836532493,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.419,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.4221530765312797,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.4328,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.41353975940665627,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.4381,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.374453063082544,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.4391,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.4651352929287388,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.4602,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.33924246750000353,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.4423,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.40818547250910336,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.4378,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.3833907794970141,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4397,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.36740546023409926,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.4593,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.3865996932766727,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.4407,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.5280439876657741,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.436,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.3784086119885681,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.4569,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.500461106358661,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4531,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.3611105905144991,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.4477,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.5278843145794785,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.4443,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.390964591964194,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.4446,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.4811442820971912,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.4455,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.34992693790463697,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.4531,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.44153272070380195,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.4509,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.4284777844201153,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.4584,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.42367010120467874,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.4218,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.38791147726937925,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.4489,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.42113770666588446,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.4577,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.3784507496523702,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.4457,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.39069024143400555,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4536,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.3429655904324977,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4431,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.37704936920177284,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.4509,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.36119679054926995,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.441,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.4484845209178578,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.459,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.38591075201862757,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.4395,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.43794347871891637,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.4525,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.43247375894133805,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4388,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.3449698210963669,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.4436,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.41216966473150396,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.4371,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.32744529064832734,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.4613,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.44897620433987623,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.4347,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.3927169778280724,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.4616,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.37824659263446614,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.4581,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.41623765228032256,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.4574,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.3151785249319252,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4486,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.4733928871687785,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.4476,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.312092141383167,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.4488,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.42477273172058383,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.4332,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.3413547641460081,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.4515,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.361831016092675,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.445,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.31557333833751083,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.4354,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.33814283189125616,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.4631,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.36565443773639944,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.4441,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.33287988708639277,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.4653,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.3325746729227563,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.4374,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.3492744534129122,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.4509,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.3052661974566435,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.4643,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.3865057509130595,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.4264,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.3945368431013577,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.44,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.31551503861804187,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.4567,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.36423383027684203,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4538,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.30873955447340107,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.4504,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.3443564522755062,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.4516,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.3040564527422456,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.4438,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.35828903043870486,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.4343,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.337944818302541,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.4494,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.3284969987971173,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.4475,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.4500444313571369,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.4471,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.30415839770589387,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.445,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.3651806077223294,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4289,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.3202114478300557,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.4537,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.3513470243059724,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.4203,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.36695743786906126,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.4674,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.31417881327618535,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.4605,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.3419788510351848,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.4422,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.36308662176674605,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.4528,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.2944403061742877,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.4631,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.34718521710004113,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.449,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.31995301235623297,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.436,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.36181491937825255,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4382,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.34914671895140376,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.4476,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.37529691496420603,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.4373,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.33389980537723735,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.4584,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.38034285306085447,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.4476,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.3444916969603286,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4254,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.33470349186845866,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.4554,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.3953680359833383,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.4556,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.32093431548417445,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.4435,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.38520416226090015,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.4527,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.3564438071792211,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.4437,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.3586405712470757,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4638,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.34021873961843463,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.4523,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.34243436119294735,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4287,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.35903765024043943,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.4468,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.3295831605131409,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.4568,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.39498900803915843,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.459,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.3340647031942189,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4392,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.35161448958934816,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.4394,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.31441471790788317,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4537,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.37981175450062726,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.451,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.3119850731413586,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.4482,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.3409126294783463,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.4377,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.3104066141283855,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4451,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.31686806834400416,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.4615,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.29137913471811716,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.4487,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.31390784254196746,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.4429,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.3149315009698974,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4466,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.328987570705141,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4519,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.33940340559009685,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4458,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.32568287523225375,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.4443,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.31748934259541506,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.4631,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.3267243921864814,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.4354,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.32804063619758783,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4338,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.322312900866131,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.4475,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.33888943234380836,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.4572,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.3850718403020001,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.444,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.33833296063284257,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.4182,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.36623916065140233,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.4678,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.3300432816688732,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.4399,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.3431851101720186,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.4342,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.315511476750774,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4488,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.3622345348888472,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.4364,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.3556779750328227,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4423,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.3290385341868557,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.4497,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.34234886351085553,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.4546,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.3661772552817619,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4452,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.3302876400389088,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.4631,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.35607183137338205,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.4437,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.32296804492252684,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.435,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3486852435686598,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.4381,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.3048065141183146,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.4563,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.36754841548600015,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4288,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.334320418588358,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.4455,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.33417706208457554,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4294,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.3716756027925991,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4496,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.2904481991356466,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.4452,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.40354032539489454,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4456,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.3147334546615363,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.4435,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.3769327246572636,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.4609,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.34078653194924025,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.4466,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.337213140832756,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.4482,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.3474797880174338,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.4443,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.30885768335086367,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4551,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.3645351556343355,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4519,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.34980565014809956,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.4626,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.31935353532998767,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.4434,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.37715411951232014,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.4578,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.32855890977976776,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.4424,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.3300255170766426,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.4572,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.36374759667618556,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.4495,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.3316958260097736,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.441,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.4173835563793416,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.4476,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.3477811124326278,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.4286,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.3461689461672754,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4568,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.3552720550474046,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.4207,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.3234077159424189,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.4439,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.3756239751256919,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.4515,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.3312294055316703,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4386,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.36136315931881213,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.4456,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.32853022362251433,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.4499,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.32216401127444305,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.457,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.321970912549318,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.4494,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.3178606444911143,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4559,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.322239146095412,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.4481,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.3367645503917186,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.4415,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.3109509730974705,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.4467,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.3134121092196203,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.4308,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.32321003428507455,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.435,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.34742370959869734,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4573,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.2979171658191354,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.4671,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.4063993781736134,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.4641,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.28500185247845383,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.447,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.3822427451210225,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.4632,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.305936989540333,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.4454,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.3230890577406607,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.438,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.3503003400644132,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.4592,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.3016076300900134,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.443,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.3159080355867219,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4414,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.32150720934263965,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.4472,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.3289732050505392,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.4704,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.3138555067959173,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4419,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.34204047328932374,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.448,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.3169916334052176,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.4363,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.3202883918041449,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.446,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.29338737977518353,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.4482,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.30309639561598245,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.4535,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.2879926758394045,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4377,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.3006233838934973,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4529,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.3067477764758271,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.4445,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.419793819329586,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.4505,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.279386659658663,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.4398,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.3075169558544236,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.4585,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.29350157540678723,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4427,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.2928047855397848,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.44,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.269269607552608,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.452,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.3248266697353287,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.4444,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.2766510056016294,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.4469,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.33251846268050284,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.44,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.2832733273157472,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.4551,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.3207568623814138,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.4524,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.29282954843753906,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.4609,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.29976312459307564,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.4633,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.287223831443758,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.4344,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.29777662666576316,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.4375,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.26841966537938267,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.4567,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.294405977647673,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.4645,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.31086306928157265,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.4435,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.2852974961512716,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.4453,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3138694929925375,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.451,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.29593942956130487,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.4282,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.2973415303085732,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4375,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.2957836583232251,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.4457,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.28125667565032436,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.4597,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.28333381323616286,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.4456,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.281996966102348,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.429,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.2729470653653364,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.4356,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.2756876592473894,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.4489,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.28021349720993366,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.4581,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.29538567460407195,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.4517,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.27024575907031456,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.4515,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.2892921249409058,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.4393,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.2920960384171734,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.444,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.2840220922135301,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.4517,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.28593578886884363,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.4451,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.29647635099186953,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.44,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.3243666760086749,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.4686,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.3108492773579036,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.46,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.25975277651145096,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.4399,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.2886611779751916,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.4533,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.3360165253610191,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.4305,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.3175266324153926,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.4363,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.29655914448246695,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.4427,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.28000006351148543,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.4581,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.3334961188520257,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4417,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.2767888900504906,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.4521,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.27323479768682307,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.4457,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.26996680552785557,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4354,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.2754187684551931,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.4211,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.31852572365783055,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.416,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.38005331618509147,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.3955,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.4131061200418659,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.3771,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.416931699773504,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.3676,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.4553660180434236,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.3671,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.38033930252458553,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.3845,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.33991301965300963,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.3766,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.36231076804984175,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.3843,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.3114004033861117,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.3763,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.34940946218113356,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.3704,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.33180214381219164,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.3817,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.3508339599549359,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.3657,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.3824086391927659,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.3859,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.340362883370116,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.3823,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.35468353901156807,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.3904,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.3437006016950763,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.3891,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.32725103920984805,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.3842,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.34740766483546265,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.39,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.30327423572950457,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.3779,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.2954269271268559,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.3821,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.3311783172163711,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.3709,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.31480995896716774,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.3861,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.33243654868904693,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.3646,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.3178565909634108,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.3724,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.2904469219426958,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.351,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.3171603198437099,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.3727,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.2799709838408766,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.3578,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.28617748960392786,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.3815,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.30006438228888604,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.3826,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.26921221833783315,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.3813,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.27323497537683644,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.3981,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.2823789951310968,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.3884,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.2732539686397762,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.3671,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.28338317748580544,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.3927,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.2838263321677898,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.3848,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.28112534786866583,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.3679,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.25188813078631095,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.3818,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.2968444904284662,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.4026,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.28635183536945963,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.3849,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.2869350020682886,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.3888,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.28609757297782107,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.3896,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.28966971756420584,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.3594,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.2905158634566682,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.3954,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.2845857115722114,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.3807,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.27896106878722154,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.3838,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.2794883939514435,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.4044,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.2936128263140425,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.3759,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.3154667063572609,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.3968,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.3166011861422139,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.3698,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.3017256545920382,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.3799,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.2955910789298858,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.3864,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.2832459883754405,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.4079,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.3255498432378702,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.3742,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.30892801504571094,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.3658,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.3525423885277233,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.3693,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.37172506652710735,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.3827,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.3044504505413011,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.3741,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.36821342672743995,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.3734,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.30464196902791657,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.3789,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.30833365405676794,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.3928,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.3242382749960778,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.3821,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.3122579655685792,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.374,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.3004219136896339,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.3676,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.3190758875431315,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.3925,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.2848241444976758,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.3692,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.3103393507643563,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.3922,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.32314358706908264,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.3801,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.2883507439465821,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.3772,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.3068795620823446,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.3846,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.32636476703944567,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.379,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.28141267873762027,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.3821,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.3106204261859702,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.3722,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.2979516355366347,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.3953,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.2921108021331715,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.3855,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.2662010272437067,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.3947,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.2531553092622918,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.387,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.28323360904869,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.3732,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.28423093281378786,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.3836,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.27669086106480845,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.3528,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.2931035952834546,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3879,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.2726827397277724,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.3901,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.27663469615131625,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.3883,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.28356075502612493,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.3762,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.26883401787257116,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.382,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.28532207162537176,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.3949,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.2836346002165896,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.3759,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.3294685057712913,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.3742,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.2835304519000043,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.3748,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.3220542707161319,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.4033,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.27777627922082376,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.377,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.31176534566941,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.3727,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.29143462915773627,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.3785,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.27323652013590827,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.3769,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.2841614465028908,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.3883,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.2828616992033762,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.3882,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.29888036367129867,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.3913,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.2573946437799912,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.3788,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.3054227066688168,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.3909,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.2920794371729224,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.3937,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.2667788050660287,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.3835,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.3102134890070007,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.3795,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.29313072388654415,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.391,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.26601461046105607,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.3924,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.2717036359200317,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.3658,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.27237717288513613,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.3874,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.2786105776012927,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.3714,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.27032968847559063,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.3762,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.27248869752162364,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.3799,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.305386212024308,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.3851,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.2565307115074045,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.3854,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.27727818543952754,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.3746,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.2645645043425613,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.3873,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.2677674606619267,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.3763,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.2799743849299804,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.3813,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.30078761975981483,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.356,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.2742547388370501,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.3908,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.29243980627991345,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.3657,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.27340234489753584,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.3668,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.28088123290504,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.3765,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.2825996698328971,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.3753,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.2822523067083223,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.3945,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.2685391920106073,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.3917,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.2897603772337803,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.3669,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.2775334857111468,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.3748,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.2767062550766906,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.3758,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.30159953324381794,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.3828,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.28144618094805207,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.3757,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.271286012541177,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.3771,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.29246565824343207,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.3825,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.3078827767507482,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.3773,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.2686048280299685,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.3811,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.3048103906160557,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.3834,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.30552004436615826,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.3795,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.28297174987759655,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.3776,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.2593822161600278,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.3795,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.284637751608038,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.3841,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.28075511291912025,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.3845,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.2855709389734554,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.3815,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.26137737459474225,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.3724,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.2610723998117865,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.39,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.2956954649090208,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.4084,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.26898382846718866,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.3757,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.2858543182568307,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.3691,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.28076467208767514,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.3951,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.2812700923764033,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.3982,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.2670400384136469,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.3757,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.2917014313615195,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.3713,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.2571892933028238,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.3796,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.2588097921099526,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.3666,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.2641110574913277,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.3947,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.2733723898856052,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.3675,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.2514203633129013,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.3963,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.25799916788833294,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.378,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.24630472263120007,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.3613,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.27470374224805394,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.3706,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.24324710265660004,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.3714,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.25276814838233863,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.372,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.25106757952110714,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.4003,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.26111156957529347,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.3775,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.2457162192283767,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.3827,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.24865013184928386,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.3871,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.257681980993,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.367,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.2361195340575047,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.3957,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.2689698678312118,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.3685,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.2522904473594589,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.3835,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.2533371139111199,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.3749,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.24973834108482554,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.3912,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.25871935202024804,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.3789,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.25246455848757815,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.3828,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.2516132577343639,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.3743,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.24494708865439735,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.3715,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.24905740174358906,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.3884,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.27617180344640985,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.3711,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.2527978731833567,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.3849,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.27085454885139276,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.3874,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.26191908752824417,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.386,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.2414720383490476,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.3816,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.24964486339003789,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.3817,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.25942768170125746,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.4013,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.25953706746890587,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.3836,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.2490133290718883,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.3626,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.2612535022194941,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.3768,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.25677566210615277,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.358,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.2543742567523976,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.3753,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.23967810174266477,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.3911,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.25685315975060724,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.3851,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.2542006035042682,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.3605,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.24666383989458093,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.3687,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.25844588184637973,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.3741,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.2628346841964937,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.3665,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.24025708870673082,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.3803,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.2590800033692381,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.3922,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.25279705579071454,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.3725,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.2512106771709809,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.3699,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.24580734415218633,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.3908,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.26908727742275546,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.3611,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.2594246225981509,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.3715,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.2769588403412091,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.3839,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.2592663816514475,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.3855,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.41972090005246715,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.3775,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.23642512193168908,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.3724,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.24173247175941026,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.3739,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.2397170818083481,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.3933,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.2453085841531161,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.3677,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.26148729690854555,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.372,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.25147187843676433,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.3691,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.2423068278289834,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.3808,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.247616051152239,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.3754,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.23996328512472367,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.3895,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.2381461840962323,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.3867,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.2573640004086483,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.3719,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.24633856431250192,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.3855,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.23982817212990454,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.3801,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.2455932594038904,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.3823,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.25243515702602776,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.3648,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.2265137714115501,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.4027,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.25200677535544264,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.3727,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.23749243215730423,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.3971,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.23975978496666256,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.3686,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.23773007311913613,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.3694,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.23211789208151826,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.3959,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.23766441768515867,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.3749,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.23412888259508255,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.3783,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.2615677896117995,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.3783,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.23560736807426244,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.378,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.24076602212346734,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.3881,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.2188048785493527,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.4016,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.24379874183299013,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.3648,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.24816420268914924,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.3938,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.23018992847295627,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.3998,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.2475141175082742,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.379,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.24032016563339523,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.3907,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.254220648316377,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.377,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.25385204147267004,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.3582,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.23470263197861593,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.3666,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.24203434243930172,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.3584,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.2313243666079395,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.3895,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.2562023937371923,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.3657,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.25156513812515574,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.3609,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.2479150644116485,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.4037,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.2564700832993442,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.3822,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.24357550492063637,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.3807,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.24032909064388536,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.3833,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.23582202511504138,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.3847,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.24513984138989403,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.3787,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 1.445614411572288,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.3923,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.2402170794991362,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.3942,
+      "step": 987
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.4294291658898469,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.3249,
+      "step": 988
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.34977191786260836,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.3271,
+      "step": 989
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.2830002161871063,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.3344,
+      "step": 990
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.44768074499917404,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.3428,
+      "step": 991
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.45438257869584475,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.3427,
+      "step": 992
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.3217794099286494,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.319,
+      "step": 993
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.356679769142266,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.3454,
+      "step": 994
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.4186111877670899,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.3326,
+      "step": 995
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.3491477673477894,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.3062,
+      "step": 996
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.2672168069759586,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.3537,
+      "step": 997
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.31796969260019164,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.3565,
+      "step": 998
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.35146942337481946,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.3524,
+      "step": 999
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.35245106210620025,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.3031,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.27743287079391776,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.328,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.2916780214797246,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.3411,
+      "step": 1002
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.2977682271189226,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.3391,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.33542288268896764,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.3157,
+      "step": 1004
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.27435022700152917,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.3318,
+      "step": 1005
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.2676863761834669,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.3555,
+      "step": 1006
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.2750056840385429,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.3266,
+      "step": 1007
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.45277022567793956,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.3336,
+      "step": 1008
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.26035899994586603,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.3453,
+      "step": 1009
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.26313628214289575,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.3284,
+      "step": 1010
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.23942692216713488,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.3509,
+      "step": 1011
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.271330539825594,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.3203,
+      "step": 1012
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.26266593331605925,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.3366,
+      "step": 1013
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.2669482032139464,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.3215,
+      "step": 1014
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.26982888077741773,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.3113,
+      "step": 1015
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.24983045214085384,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.3477,
+      "step": 1016
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.26177749150017554,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.3064,
+      "step": 1017
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.23526455980192726,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.3235,
+      "step": 1018
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.2483117668632775,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.3199,
+      "step": 1019
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.24993248891395992,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.3405,
+      "step": 1020
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.2437501768651391,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.3086,
+      "step": 1021
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.250138287195647,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.3349,
+      "step": 1022
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.23953708983974106,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.3265,
+      "step": 1023
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.2391991069785116,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.3409,
+      "step": 1024
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.2522363889692994,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.3263,
+      "step": 1025
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.2514851895034626,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.3192,
+      "step": 1026
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.24411791875390149,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.3333,
+      "step": 1027
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.255832712461286,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.3243,
+      "step": 1028
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.24454854904092474,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.3191,
+      "step": 1029
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.2294100546587492,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.3499,
+      "step": 1030
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.2384350230964805,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.3364,
+      "step": 1031
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.23264808262357056,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.3247,
+      "step": 1032
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.23602041120065886,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.3371,
+      "step": 1033
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.22898728932713558,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.3423,
+      "step": 1034
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.21266950561032463,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.3463,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.24487706034010095,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.3122,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.2359780232720692,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.3365,
+      "step": 1037
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.24976863107784067,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.3079,
+      "step": 1038
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.27512436036812193,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.3054,
+      "step": 1039
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.22739520266250518,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.354,
+      "step": 1040
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.2200249223063524,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.3457,
+      "step": 1041
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.2267753274602659,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.3309,
+      "step": 1042
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.23070480409129748,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.3377,
+      "step": 1043
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.23471794880511876,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.3254,
+      "step": 1044
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.23541941949815776,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.3376,
+      "step": 1045
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.21997685269616168,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.3481,
+      "step": 1046
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.2260046073908498,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.342,
+      "step": 1047
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.23114466114389073,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.3288,
+      "step": 1048
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.22415267514329756,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.3393,
+      "step": 1049
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.2231234766399456,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.3367,
+      "step": 1050
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.2580642437834052,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.3496,
+      "step": 1051
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.2282995665330048,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.3287,
+      "step": 1052
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.2268360315026251,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.3411,
+      "step": 1053
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.25416873517120325,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.3085,
+      "step": 1054
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.23597470739005627,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.3407,
+      "step": 1055
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.23080873145349362,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.3303,
+      "step": 1056
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 2.1714664451654766,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.351,
+      "step": 1057
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.23840004662731581,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.3081,
+      "step": 1058
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.23211423794817912,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.3446,
+      "step": 1059
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.23835021705413298,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.3177,
+      "step": 1060
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.23162660272478233,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.3417,
+      "step": 1061
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.24508319779831564,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.322,
+      "step": 1062
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.23648856590178582,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.3438,
+      "step": 1063
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.23386445166383676,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.3387,
+      "step": 1064
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.22499369987841533,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.3327,
+      "step": 1065
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.23273477300392612,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.3264,
+      "step": 1066
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.22333815259489292,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.3397,
+      "step": 1067
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.22392340875671848,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.3191,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.21970092382850323,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.356,
+      "step": 1069
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.22798651528674543,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.3405,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.22783148939799835,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.3279,
+      "step": 1071
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.20596208742359048,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.3643,
+      "step": 1072
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.23543790601401066,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.3297,
+      "step": 1073
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.22977848606538862,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.3206,
+      "step": 1074
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.22905042521737598,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.3103,
+      "step": 1075
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.2159615239842529,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.334,
+      "step": 1076
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.23526242063995884,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.3213,
+      "step": 1077
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.22479018151560054,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.3197,
+      "step": 1078
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.22874482971091933,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.3323,
+      "step": 1079
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.2291789935994844,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.3299,
+      "step": 1080
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.22597029202137103,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.3298,
+      "step": 1081
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.24266481283840508,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.3363,
+      "step": 1082
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.22308566733551166,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.3622,
+      "step": 1083
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.2245137973313299,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.3297,
+      "step": 1084
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.22590901247264825,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.3397,
+      "step": 1085
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.23886586351168498,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.3216,
+      "step": 1086
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.2370428729621412,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.3421,
+      "step": 1087
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.23427131821852143,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.3157,
+      "step": 1088
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.2171919506934338,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.3671,
+      "step": 1089
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.22247234807563215,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.3477,
+      "step": 1090
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.23669312163531256,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.32,
+      "step": 1091
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.20879867008938832,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.3701,
+      "step": 1092
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.22366501396409283,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.3372,
+      "step": 1093
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.22186514690547327,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.3364,
+      "step": 1094
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.22748965260834292,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.3233,
+      "step": 1095
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.2187570676919551,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.3497,
+      "step": 1096
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.21240725227642318,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.327,
+      "step": 1097
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.22091972016776323,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.3243,
+      "step": 1098
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.2477777140765659,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.31,
+      "step": 1099
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.22975976076303672,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.3103,
+      "step": 1100
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.21850152652541846,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.353,
+      "step": 1101
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.23503355776409735,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.3095,
+      "step": 1102
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.22656757147967665,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.3364,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.2820546146780969,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.3294,
+      "step": 1104
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.23448098525635958,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.3294,
+      "step": 1105
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.21873187298392566,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.3461,
+      "step": 1106
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.22542266389439986,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.3425,
+      "step": 1107
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.2126199955223394,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.3489,
+      "step": 1108
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.21890940097168193,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.3308,
+      "step": 1109
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.23358939104485033,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.3275,
+      "step": 1110
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.22378027808178746,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.3394,
+      "step": 1111
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.22218792189691106,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.3474,
+      "step": 1112
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.22944385073903836,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.3501,
+      "step": 1113
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.233051550204023,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.3271,
+      "step": 1114
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.259156385395774,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.3262,
+      "step": 1115
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.22604366518669597,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.3229,
+      "step": 1116
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.235512980292762,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.3094,
+      "step": 1117
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.2333590880137163,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.3411,
+      "step": 1118
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.21601957487252088,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.3213,
+      "step": 1119
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.23030975161230158,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.3345,
+      "step": 1120
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.22295212237597986,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.3392,
+      "step": 1121
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.21256977609764083,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.356,
+      "step": 1122
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.2220894690020251,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.3346,
+      "step": 1123
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.21143573638277818,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.3495,
+      "step": 1124
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.22089243511208173,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.3318,
+      "step": 1125
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.20969356088019123,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.3624,
+      "step": 1126
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.24459690976085574,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.3135,
+      "step": 1127
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.22000839623566282,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.3215,
+      "step": 1128
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.21940475195804407,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.3442,
+      "step": 1129
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.22320873467626115,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.3204,
+      "step": 1130
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.2302041010667337,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.2999,
+      "step": 1131
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.2113364858192355,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.3631,
+      "step": 1132
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.21158678711977472,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.3299,
+      "step": 1133
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.2229991914091403,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.3163,
+      "step": 1134
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.22472867358071502,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.3324,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.218528133714536,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.3414,
+      "step": 1136
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.22108671119942563,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.3373,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.21862652175767835,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.3425,
+      "step": 1138
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.23932981417309407,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.3288,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.2321262987112453,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.3247,
+      "step": 1140
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.2254425017987875,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.3374,
+      "step": 1141
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.22261463685705862,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.3258,
+      "step": 1142
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.20609754492764917,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.3424,
+      "step": 1143
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.2266347424218776,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.3308,
+      "step": 1144
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.2126245358190361,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.3465,
+      "step": 1145
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.22008432528802258,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.3328,
+      "step": 1146
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.23216376528463684,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.3186,
+      "step": 1147
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.21771852417754142,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.333,
+      "step": 1148
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.2219681614106277,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.3286,
+      "step": 1149
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.21548922427692255,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.3502,
+      "step": 1150
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.2118961276092372,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.3464,
+      "step": 1151
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.23411955971635354,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.3131,
+      "step": 1152
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.21502667836968814,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.3503,
+      "step": 1153
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.2314528832543848,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.3103,
+      "step": 1154
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.21687030975320717,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.3361,
+      "step": 1155
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.19972755099821263,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.3661,
+      "step": 1156
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.21981886186564825,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.3247,
+      "step": 1157
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.22492706437734358,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.3139,
+      "step": 1158
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.2363956038393975,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.3139,
+      "step": 1159
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.21396193740176642,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.3257,
+      "step": 1160
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.23010713916868422,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.3186,
+      "step": 1161
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.2342499152712685,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.331,
+      "step": 1162
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.21283148032175211,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.3163,
+      "step": 1163
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.22406952689003332,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.3297,
+      "step": 1164
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.2125694024242771,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.3231,
+      "step": 1165
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.2172596654229647,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.3305,
+      "step": 1166
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.21670744741798675,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.3338,
+      "step": 1167
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.21959203530275315,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.3428,
+      "step": 1168
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.21379753417525268,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.3455,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.20201394438084208,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.3588,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.21580173496630772,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.3394,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.21266502850368613,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.3425,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.2246965492236211,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.315,
+      "step": 1173
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.22386735477922654,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.3119,
+      "step": 1174
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.21095290542945358,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.3333,
+      "step": 1175
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.21871273551345222,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.3405,
+      "step": 1176
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.22738404462204914,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.3236,
+      "step": 1177
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.22444216327401162,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.3361,
+      "step": 1178
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.23200300549249267,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.3086,
+      "step": 1179
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.2242812815275193,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.3285,
+      "step": 1180
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.22540168913606667,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.3322,
+      "step": 1181
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.22631574906831303,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.3304,
+      "step": 1182
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.22205464808739758,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.3293,
+      "step": 1183
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.20995420612078455,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.3321,
+      "step": 1184
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.21165665166366254,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.3367,
+      "step": 1185
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.22054308773125003,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.3364,
+      "step": 1186
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.228424021246436,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.3352,
+      "step": 1187
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.22563509951870842,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.3206,
+      "step": 1188
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.2310893247412691,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.338,
+      "step": 1189
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.21103884616619567,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.345,
+      "step": 1190
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.21675057844825277,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.3443,
+      "step": 1191
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.2147015438528445,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.3469,
+      "step": 1192
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.22060690717359158,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.3297,
+      "step": 1193
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.22294021113133142,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.3427,
+      "step": 1194
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.21005300445256206,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.3425,
+      "step": 1195
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.2217957719246325,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.3219,
+      "step": 1196
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.22469355969656413,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.3199,
+      "step": 1197
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.22516768646402022,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.3365,
+      "step": 1198
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.22985387921499362,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.2913,
+      "step": 1199
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.21722607082964557,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.3212,
+      "step": 1200
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.21953127940513106,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.3331,
+      "step": 1201
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.22750485422550096,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.3272,
+      "step": 1202
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.22109267948158404,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.3149,
+      "step": 1203
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.2340676918618681,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.3104,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.21916118533426088,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.3166,
+      "step": 1205
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.21789651647669472,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.3272,
+      "step": 1206
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.22313007230325288,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.325,
+      "step": 1207
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.20966768933768593,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.3486,
+      "step": 1208
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.23280454072398074,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.3281,
+      "step": 1209
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.21717974150839672,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.3465,
+      "step": 1210
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.21662377092053964,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.3453,
+      "step": 1211
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.22047513204133365,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.336,
+      "step": 1212
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.21725694751057606,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.3421,
+      "step": 1213
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.21453636960297212,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.3382,
+      "step": 1214
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.22266527730266222,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.349,
+      "step": 1215
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.21854917257087417,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.322,
+      "step": 1216
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.2115573652881817,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.3428,
+      "step": 1217
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.23540665250211174,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.3276,
+      "step": 1218
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.20907562928074955,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.3412,
+      "step": 1219
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.23089247346019032,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.3222,
+      "step": 1220
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2196848143361643,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.3378,
+      "step": 1221
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.20924251090953133,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.3582,
+      "step": 1222
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.22710444300202995,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.3433,
+      "step": 1223
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.21105831340788767,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.3476,
+      "step": 1224
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.20721385819123417,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.3591,
+      "step": 1225
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.20192107641414905,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.3551,
+      "step": 1226
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.2080046861420482,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.351,
+      "step": 1227
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.21978401430660074,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.3382,
+      "step": 1228
+    },
+    {
+      "epoch": 4.992405063291139,
+      "grad_norm": 0.22603848894108178,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.3235,
+      "step": 1229
+    },
+    {
+      "epoch": 4.996455696202531,
+      "grad_norm": 0.21164771704455798,
+      "learning_rate": 0.0,
+      "loss": 0.3503,
+      "step": 1230
+    },
+    {
+      "epoch": 4.996455696202531,
+      "step": 1230,
+      "total_flos": 4.793704475332608e+18,
+      "train_loss": 0.06581524976385318,
+      "train_runtime": 10776.104,
+      "train_samples_per_second": 14.662,
+      "train_steps_per_second": 0.114
+    }
+  ],
+  "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": 4.793704475332608e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}