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": 4.982278481012658,
+  "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": 9.019757106412337,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.2725,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 9.219796226235124,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.2913,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 8.980767744945082,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.2858,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 8.835705248839716,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.2659,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 8.618443959087188,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.2733,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 8.294400955645248,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.2514,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 6.955273701291936,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.2284,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 6.22460029271415,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.1741,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.9942858556228136,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.1545,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.3432795009309317,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.1059,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.159318055969329,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.1167,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 5.760781520473386,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.1186,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 6.027518202771571,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.1032,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 5.811054486443033,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.1025,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 5.819412362823791,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.0683,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 5.0822920325202245,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.06,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 4.6281873838834295,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 1.0395,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 3.7401594136792125,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.999,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.5810712576569155,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.9995,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.742970134021055,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.9648,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.827159137354703,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.9835,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 2.0704375055677033,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.9526,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.8100995562017737,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.9325,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.5622851813908973,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.9365,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.3314996645600818,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.937,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.108311098620708,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.9143,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.0921836292672693,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.9122,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0538377271174268,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.9111,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.043624470948413,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8778,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 1.0544949474634098,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.8903,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8615112346791216,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.8646,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.9849359925725489,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.9091,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.9833967800077915,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.8435,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8382399190725907,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.8821,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.752631652343354,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.8721,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 1.0098281013277384,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.8523,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.8224519767868826,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.8604,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7546212452973846,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.8483,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.8857246695086332,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.8349,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7908159846449831,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.8493,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.7949942185018504,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8482,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.7020954741447472,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.8358,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.8438346735814912,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.8479,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.6414414412984621,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.817,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.8148479962667289,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.8272,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.7505897549981646,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.848,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7387014092073203,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.811,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.8537098559891633,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.8215,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.8394073356820698,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.8392,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7748846232540272,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.8172,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.6788851463118774,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.798,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7948168035285728,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.8402,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.76153185610127,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.8051,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.7744226513576732,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7948,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8578152683533041,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.8265,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 1.0090064373421064,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.7927,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.7659987828553099,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.8297,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.9580077577318935,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.809,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7207881384117872,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7972,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 1.02585291336667,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.8043,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.9304233449900655,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.7832,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 1.1926167995021522,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7944,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.8964098002792374,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.8269,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 1.0502182734501582,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.8005,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 1.1743261110527845,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.8015,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 1.0143038529923016,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.7833,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.0714721579458923,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.7827,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.074629833330477,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.7687,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.9690749436076751,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.774,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.7019481246472652,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.7898,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.0753868000012443,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7848,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 1.181166888512971,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.8062,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.8251095303151881,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.7961,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.879457347105681,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.8247,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.9764967616182906,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.7895,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.3806661807556582,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.8086,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.9941192656171867,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.801,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.157732125256111,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.7975,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1080038494093236,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.8189,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.9046863469988589,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7816,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.1190700305172696,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.7706,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.1588092402163943,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7831,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9762646204190102,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7708,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.3859147914355814,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7791,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.0504115381144847,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.7808,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.637291079929808,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7672,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.9435881733331717,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7736,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.7181354283628274,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.7537,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.3368264761347302,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.776,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.4379166662998237,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.7898,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.0718125224070751,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7749,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.1490193697630278,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.7775,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.2596102507653513,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.7962,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.1299117952064488,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7894,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.2289389223188938,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.7651,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.1608988872617583,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.7774,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.3818106202493021,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7632,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.9424934674592154,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.7862,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.4809928587786438,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.7871,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.1132585740250416,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.765,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.0752463495855211,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.7729,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.1055758090056196,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.7721,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.2176128657125223,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7619,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.1555284049574968,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7829,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.8154460417777794,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7751,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.1090754461306818,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7898,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.9171404785983326,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.7691,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.0243338235869908,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.7838,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.0865759032823001,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.7846,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.1419181150894415,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.8118,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.4083050290115098,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7826,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.9913734037004154,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7999,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.8422058653442275,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7706,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.1782827578639916,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.7696,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 2.1005547385835195,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.7739,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.584687760970992,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.7697,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.7232443955783523,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.7686,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.776323264539928,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.8047,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.492272348171322,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7825,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.3199934404657772,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.7693,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.296513635345143,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7971,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.1170774804600245,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.7671,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.0194624713070206,
+      "learning_rate": 4e-05,
+      "loss": 0.7713,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.5203754803929383,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7852,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.1793736128924166,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7719,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.6914237627486666,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7838,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.3442623099889675,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7571,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.6064936056914356,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7872,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.307773191928674,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7645,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.3730555499261536,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.764,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.124381406954777,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.7745,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.8140455213689416,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7698,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.071734572019169,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.7805,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.1031283044223525,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7641,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.2197275672643528,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7815,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.7897313307362803,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.7701,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.04559345712199,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.7564,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.8470361922307452,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.7623,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.9537171614566642,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.7598,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.7588877391810296,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7412,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.815836657866736,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.7797,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.8267769739101135,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7796,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.0086829417069376,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.7429,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.000495155424249,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.7526,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.919452166497955,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7565,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.8468532839185925,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.7679,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.9663141479357292,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7605,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.118727931982597,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7704,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.8839548326989405,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.7605,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.8851671446089404,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.7498,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.6455647336344585,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7499,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8223697147849998,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.7589,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.6902574129190943,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.7773,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.6942646136123302,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.7556,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.8084153252364819,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.7655,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.8513114962974072,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.7589,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.9489385467048663,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.7628,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0043995383304554,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.7747,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.6280111044518665,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7609,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.000506291198081,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.7507,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.8992679764287944,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.7445,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.7895750950028615,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.7628,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.0138594523327233,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.7813,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.8568151101749951,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.743,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.9601617342857897,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.777,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.7679110434261708,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.7535,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.8175972701683017,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.7607,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.8684630195095675,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.7608,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.9236167252975299,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.7459,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.0342910964054957,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.7606,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.89951234247247,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.7526,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.8454415300226107,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.7658,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.7089131803721079,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7368,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.6853854921513423,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.7531,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.6426133492658036,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.749,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7820618810924622,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7703,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.6235592257123975,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7496,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.7454855821908696,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.7677,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.653172795380942,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.7453,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7394258651817436,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.7575,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.7537952850808585,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.761,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.5718495928261353,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.7425,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.6414941229245142,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.7502,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.5585550818078232,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.7499,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6870428205671292,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.7688,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.8217296829265177,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.7396,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.8133602419994829,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.7442,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.6389509167339915,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.7655,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.866113734299406,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.7449,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.7468995721982269,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.7451,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.7915056368630372,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.7593,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.6424001800658535,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.7433,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.7693196256407843,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7408,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6013460204652211,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.7651,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.8605848939138017,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.7579,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.6493708928621392,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.7609,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.7673224650338957,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7681,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6419719087742707,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.7469,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.723794149223967,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.7415,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.6176403678211824,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7325,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.7016400355319499,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7463,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.7148428844364477,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.7563,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.6593731191359051,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.744,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.6775044486386971,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7689,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.6796654459016516,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7555,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6628266334516907,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.7442,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.8180236353883293,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.7331,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.692624565932649,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.7543,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.756984451882208,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.7472,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.7304904703604739,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7279,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.7278543202947529,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.7559,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.7374575934412068,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.7458,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.620355907779384,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.7437,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.7874884521930113,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.7712,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.8126424930069741,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.7356,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6463027281619157,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.712,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.7255520705290247,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.761,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.6756216109396374,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.7602,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.6221681316665898,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.7653,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.5842853381896861,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.7416,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.5906562054973156,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.7236,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.71757210374475,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.7326,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.6489305601021949,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7404,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6795481934966514,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.7517,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6529009770767054,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.7542,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.7903259494936948,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.7607,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.6466541096923974,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.7506,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.660602755780437,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.7157,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.7015131888004535,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.7599,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.560999438455394,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.7525,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.754832061944134,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.7231,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.818075919528731,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.7359,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.6339718523205655,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7292,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.7898425290857857,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.7748,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.6737391661437699,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.7612,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.7707054868837905,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.7327,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6318069025883917,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.7254,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.7019944294454183,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.737,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.5516925911340534,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.7273,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.6348456746272994,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.725,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.779804096165004,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.7268,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.5404469973731032,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.7572,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.6352473538857868,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.7496,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6203717640102037,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.718,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.5925486279546124,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.7367,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.6467527019764909,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.744,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.5471562156653079,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.7098,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.6816456388871805,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.6822,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.8721071748770719,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.6832,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.8037956046160548,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.6841,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.9210259978233091,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.6698,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.7030059576259358,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.6746,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.8369657365241461,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.6643,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.9431977437831075,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.6701,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.7457729348932631,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.6772,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.5953233682025599,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.6592,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.7860836758586592,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.7074,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.8592506340278201,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.6921,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.6182169588979572,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.6612,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.69319525199141,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.6643,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.7727405470319341,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.6767,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.6007852453427427,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.6682,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.6359501503849925,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.6643,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.6806038042876731,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.667,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.5937739767016129,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.6695,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.5590661646381482,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.6843,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.6575271380054047,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.6821,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.5841566311749377,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.6435,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.594595105652257,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.6722,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.5875256214283292,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.6806,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.5848104243439709,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.6677,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.5049173230082556,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.6776,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.5715365095180156,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.6746,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.568369818282661,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.6741,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.477483434942644,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.6592,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.6242995361342556,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.6686,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.6244210483709962,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.6606,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.5978281252795549,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.6713,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.7058418325862672,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.6879,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.6791285606619155,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.6717,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.5175252511372262,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.6673,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.6520146319220845,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.6822,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.517227949832193,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.6856,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.6789420857652525,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.6685,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.5033897169153999,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.6854,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.6095491331985133,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.6509,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.5474569458488029,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.6472,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.5074102312199306,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.6864,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.545104882638792,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.6794,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.4904103888968233,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.69,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.5120108950121701,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.6639,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.49675789675524124,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.6914,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.5604945088285855,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.6972,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.5813592146120259,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.6948,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.5504821010665161,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.6694,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.6999577974715756,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.6541,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.5498895677998419,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.6635,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.6459699019953724,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.6639,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.5766197548321209,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.6976,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.5837124834936831,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.6759,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.5861016033510625,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.6831,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.5298819707329516,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.6618,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.5972260483297076,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.6815,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.6027828761902067,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.6547,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.5497088868406804,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.6742,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.5452737845759379,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.671,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.573022767472074,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.6823,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.518503865716218,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.6675,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.6068369607784067,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.6735,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.46571628293996076,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.6709,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.4894057336711059,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.6815,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.45920779349569785,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.679,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.5216676187181264,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.6883,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.5640128699892951,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.6791,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.6086575799805767,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.6658,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.48827322228803555,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.6942,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.71020548610296,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.6727,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.4758157341966051,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.6695,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.6146032653124883,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.6834,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.5820322475219245,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.6596,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.4965546097364921,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.6988,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.49461554844679195,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.6698,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.47695225861261853,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.6688,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.5140304871122909,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.6658,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.48965055350924014,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.6779,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.6090361063776032,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.6646,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.6162882738244135,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.6756,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.5767312329971237,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.6776,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.5316085022328716,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.6679,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.5693876745855591,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.6706,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.562114661403973,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.6908,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.4528897304505485,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.668,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.5216987184300398,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.6818,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.48619004054028636,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.665,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.452684226013556,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.6562,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.4892523415849213,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.674,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.47897279936288445,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.683,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.5629808189872401,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.6598,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.43664106400012975,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.6623,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.5818667993240939,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.6843,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.5223218313240268,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.6901,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.5600233919462972,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.6769,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.5177259298808304,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.6599,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.46211986977489783,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.6678,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.5158357823100501,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.6639,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.4593698293219815,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.6861,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.5479360619071317,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.688,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.5007575039707488,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.6735,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.5559163238107043,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.6854,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.6614012052035517,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.6958,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.5257689023918134,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.6696,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.5364171619461426,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.6601,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.6682192442416576,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.6722,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.5556163920518622,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.6732,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.5189965379540226,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.6818,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.619243451870871,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.6778,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.5784829004183895,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.6672,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.5686140754386682,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.6997,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.6300997100936683,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.6654,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.5786081527823683,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.6547,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.5053528232281657,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.6527,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.5085672743911435,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.6921,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.5094575900810197,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.6617,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.4753261101139938,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.6849,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.4385555762749458,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.657,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.46748498211992245,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.66,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.44614311198906714,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.6889,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.5528250140089087,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.6719,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.5040142809535876,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.6613,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.49840776313093693,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.6937,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.5068415832684647,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.6829,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.44153656806525954,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.6776,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.4830499447853505,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.6819,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.5314034938824662,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.6706,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.548356574415684,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.6994,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.41728606129030693,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.6789,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.49571879680555464,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.651,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.4809864079356877,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.6748,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.5289804523753772,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.6605,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.5465946512966854,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.69,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.510204180147717,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.6803,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.5864705378231024,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.6811,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.6454680887741356,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.6645,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.5626096738711234,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.6831,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.5965565660970823,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.6693,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6975367597392814,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.6836,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.46313442317581166,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.6481,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.6646167051575029,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.688,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.5722122732254664,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.6838,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.5610488218553815,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.6964,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.5119631704477333,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.6575,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.687689119206693,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.6731,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.5023094380423784,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.7121,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.6024157671704569,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.6746,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4980614939326208,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.6846,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.532664762720472,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.6839,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.5980225355945433,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.6606,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.47522050016285666,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.6419,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.5347516745076074,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.6788,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.5012734686065533,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.6805,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.6106275887578038,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.6627,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.44425042364658074,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.6726,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.5365593686534104,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.7024,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.5161689334249295,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.6715,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.42023507607331506,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.6667,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.5500189898033637,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.7125,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.40054167813482555,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.6985,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.5605997115485089,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.6675,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.542674292082776,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.6595,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.5561356167155616,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.6716,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.49494362794456015,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.6736,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.5791635525790162,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.6885,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.44492558857995895,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.6715,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.5656752469638877,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.6668,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.4856642246656815,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.6645,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.522455656718542,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.6777,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.5010867667270198,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.6873,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5502938017494251,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.6811,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.4437959673769928,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.6795,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.5638594166642605,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.6599,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.452567102629356,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.661,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.5379705276672124,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.6499,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.46497717035023356,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.6894,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.46682109878961575,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.6424,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.47392877622248103,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.6633,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.5378895392057438,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.693,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.5060783650524531,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.668,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.47762152949000797,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.6541,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.47801231341205075,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.6694,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.5111474898965669,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.7047,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.5022223859554276,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.6669,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.45897900372230216,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.6732,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.5540162789391505,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.6859,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.45615757386445016,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.6508,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.5217205474151894,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.6573,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.48360407338817146,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.6823,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.45095982996102474,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.6544,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.45400195490630174,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.6669,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.4931046762682696,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.6665,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.39495292681886385,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.6893,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.4958000952341407,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.658,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.4397454039249349,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.6843,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.40577248538344957,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.6641,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.48067188428621366,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.6722,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.43579335063591357,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.6632,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.42483396797412193,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.6917,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.3949986008170488,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.6727,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.4662353446838353,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.6996,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.42014598502012757,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.669,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.4463607510622935,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.6722,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.45711313614658866,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.6748,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.4957311935592208,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.6669,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.49789217566981847,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.7101,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.41834766057822603,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.7044,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.5679633452883872,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.6526,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.42003826601735433,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.6777,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.5395920473105834,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.6846,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.461495811033505,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.6706,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.5752589858784037,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.6801,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.5043567002148526,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.661,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.5087598444395204,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.6702,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.4830694930187933,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.6792,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.45053400478206357,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.6731,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.46420488088714773,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.6664,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.4740943381912056,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.7069,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.46824104363135083,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.6728,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.4646688294206701,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.6887,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.49657890544118116,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.6728,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.5688037750342789,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.6595,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.4382535620600342,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.6845,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.6401316314945226,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.6642,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.5615361860777813,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.6605,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.4754137535844874,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.6755,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5568671850503497,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.6705,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.48161668393197066,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.674,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.5984413641028509,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.6686,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.5241428953133942,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.6707,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.44987660652540323,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.6804,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.6326919581145094,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.6594,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.39840686492305777,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.6963,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.6155697366618424,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.6804,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.42772461118170063,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.6741,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.544859127839233,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.7021,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.5678097629316172,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.6671,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.475957935926885,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.6438,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.4719721687797036,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.6761,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.45827475964472064,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.6493,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.5191485278863748,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.6574,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.4440849756552791,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.6732,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.5172532407635831,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.657,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.5683512024073453,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.6864,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.511675846949312,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6569,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.5291197568265716,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.6905,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.608172418524861,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.6539,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.6540185913386055,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.5595,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 1.5323834650174748,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.5684,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.7288577359756874,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.5711,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.5759699984321329,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.5604,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.6902220780102654,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.5804,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.6592150186058959,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.5557,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.5635149273495738,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.5804,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.6611844415433634,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.5684,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.548335381116385,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.5694,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.6823637997573101,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.5748,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.5276789024001405,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.5775,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.74948187127555,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.582,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.6311453631290602,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.5698,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.5954076057227602,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.5491,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.5206868945395754,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.5745,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.5831101207533882,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.5759,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6104820269784369,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.5692,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.5976458436067414,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.56,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.5452636756488134,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.5753,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.6685953237072266,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.5526,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.503916820686828,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.5641,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.7373962604215967,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.5662,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.602414896640201,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.5761,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.5384983178635243,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.5803,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.502185159894607,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.5699,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.4740046087494235,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.5816,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.45177268193431064,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.5674,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.48176480180004555,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.5748,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.4534478588383651,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.565,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.5049786856111124,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.5661,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.40323005777135285,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.5692,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.45140684736911596,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.5841,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.4541835801345641,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.5653,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.4698126206056321,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.5818,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.43772150915040975,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.5633,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.45974275894452493,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.594,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.4414443954741843,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.5627,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.4603137275574115,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.5875,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.45884954270382483,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.5742,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.4443368594912161,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.5627,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.4152702286212179,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.5907,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.4553415794954687,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.5908,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.4189850140691195,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.5569,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.43587128592269564,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.5556,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.42676658365700426,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.553,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.5394000747401573,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.5791,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.4290570212829462,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.5715,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.5227601813866625,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.5649,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.438250847872474,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.5568,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.48193179070777376,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.5667,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.4705987637984828,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.5688,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.44937956662540046,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.5763,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.43725555770732155,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.5616,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.44610283327107314,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.5928,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.3958459611078714,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.5662,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.4742484017786384,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.5637,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.4799300723061663,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.5827,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.4818020502746599,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.5653,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.4362168718418826,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.5631,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.5075027384614913,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.5985,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.46320047597891834,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.5942,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.43897348753414783,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.583,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.48248904367866374,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.5478,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.4023486845563039,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.6004,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.5720062862864921,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.5829,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.39115700635497275,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.5874,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.47418432769902114,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.5658,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.44847845739631714,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.5642,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.3913706696675215,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.5574,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.42986545887439537,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.5646,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.4113780215243145,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.5657,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.5043080314037901,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.5647,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.4184393273172197,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.5824,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.4590456217528273,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.5766,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.42778826757763094,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.546,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.4890154641677426,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.5725,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.3759013704690246,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.5578,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.5059587737329206,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.5778,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.45970113099721355,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.5791,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.4420624931930029,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.5637,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.47057767195716804,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.5745,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.42926172911477284,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.5696,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.44965982137698585,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.5604,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.4067172584803679,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.5719,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.4536564743978484,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.5693,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.4295122070463586,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.5714,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.4447755276902208,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.6053,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.38894802834368064,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.5665,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.450602216061664,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.5674,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.39130106915434504,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.5692,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.3927545516587104,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.5664,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.4116579780749965,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.5479,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.38995990371412487,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.5728,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.395400710182503,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.5794,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.37947212325758567,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.5709,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.3932728393729705,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.5732,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.38534531973102165,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.5961,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.361730394551962,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.5693,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.38093084912344005,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.5694,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.3785904888591309,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.5781,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.39196690396036316,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.5696,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.40372746512828217,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.5636,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.41619371373880537,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.5734,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.4061087146025876,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.5755,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.46586330827838707,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.5679,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.4442220312062564,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.5848,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.42839206238065186,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.5918,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.4432375400466288,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.5565,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.4089684113065393,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.5871,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.41128059778919146,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.5785,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.4094150020586968,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.5965,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.4007050776572626,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.5765,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.44010595523779056,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.5901,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.39231633193945475,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.5769,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.39848158208443424,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.5536,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.42137311921146875,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.5656,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.4581027373231048,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.5869,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.4039122152903104,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.565,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.47122241223174877,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.5671,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.44057115465858576,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.5644,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.4739186355407507,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.5899,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.43894372034095064,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.5679,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.47067413789734586,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.5904,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.443624044677815,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.5951,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.43088485806393473,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.5888,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.46174485525395503,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.5527,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.4299038852394687,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.5738,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.4697263139771469,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.5756,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.39196022945286657,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.5746,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.413017107786103,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.5623,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.4189421344576086,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.5867,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.4245279006604203,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.5871,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.39927817966634427,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.5656,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.44295766611640264,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.5654,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.43731191178027784,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.5824,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.4274660832415936,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.5678,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.3874209383175522,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.5759,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.4748846058269989,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.6018,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.4916637338597482,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.5804,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.4433971212649965,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.5708,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.46613853732510985,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.5679,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.45010223931765386,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.5784,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.4055799789289061,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.5653,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.4334731455623539,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.5592,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.3981584570924678,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.5808,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.40626044213782486,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.5759,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.4813792833102193,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.5937,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.3935890685612145,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.5809,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.3895043994042726,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.5823,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.38766666798677285,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.5882,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.3957081520209535,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.5665,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.38666587899805976,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.5901,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.4146047400576535,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.5786,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.37286041578450146,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.5875,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.39377408239366024,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.5728,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.4083850781830669,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.5733,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.40175486895294504,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.5958,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.35335311797082464,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.5546,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.4240936136257704,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.5643,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.41031157391329287,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.5846,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.37292293579115193,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.5808,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.3832838105175271,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.5662,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.3895752136903769,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.5647,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.38825166616425416,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.5738,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.431954839925427,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.5765,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.4298234753964369,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.594,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.4296297868515827,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.5931,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.46311947029617956,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.5712,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.42722976781418515,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.5694,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.3801176309786892,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.5589,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.3961478681740114,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.5782,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.42020720293880054,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.571,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.3947127996074993,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.5828,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.42366713422301894,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.571,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.4066189636319373,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.5632,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.4109847482877412,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.5658,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.41424705912958837,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.5833,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.4538974055408235,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.5855,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.3868942713348484,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.5723,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.39272765368500523,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.5801,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.4071022732580726,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.5535,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.36380384047644676,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.586,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.43211318480256705,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.5935,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.3809364584811487,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.5802,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.40358389977383696,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.57,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.3890874092187956,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.5845,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.3979297723289207,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.5721,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.3978011302854492,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.5868,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.3746528858145195,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.5608,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.42903175279858724,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.5838,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.38654437382495926,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.5649,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.404288905878619,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.5835,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.4633234000025514,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.5588,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.42915640601901395,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.575,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.3717020371867386,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.5785,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.476611711897961,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.5819,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.35924510826979494,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.5772,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.41865077687514,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.5584,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.3998527717745231,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.5803,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.3693220186895755,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.5655,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.4078488097842462,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.5877,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.3907682080299805,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.5721,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.4177731907887972,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.5774,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.3955836472346365,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.573,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.41831791166241716,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.5755,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.38953293834703545,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.5818,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.4169711665293022,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.5976,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.3617944724678994,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.5601,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.3995581074644841,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.5906,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.37483972738182497,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.5773,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.38611159696340336,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.5704,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.34490569165685364,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.5652,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.40087993399849736,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.5532,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.373006438637467,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.5607,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.40767365606070316,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.574,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.38641725777894903,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.5768,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.34138332298700685,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.5793,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.35632462928109393,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.5595,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.39104225329664355,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.5777,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.33889190604785324,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.5702,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.3825943925139079,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.557,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.36328251098487774,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.5782,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.3585076473866333,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.5834,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.3620837986690496,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.5957,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.36537939469200353,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.5537,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.38805879212628014,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.583,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.364947809309992,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.5791,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.38985662912866986,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.5854,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.3566759830215219,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.5773,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.3647303210225628,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.5826,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.35599613070140607,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.56,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.3967817950760457,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.5769,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.3662834436852198,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.5573,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.4252768191117911,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.5772,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.38729923248706255,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.5634,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.4056802248577033,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.5908,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.3775437984099464,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.5655,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.3710239076872086,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.5863,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.3781681735244666,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.567,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.41231423710962756,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.5793,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.383858363765344,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.5623,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.3777079584802492,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.5664,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.38303789880992706,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.5537,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.3762698475149607,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.5645,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.3601255493029612,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.5693,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.4098897300513248,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.5637,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.42581864677505504,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.5277,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.5807609934895693,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.4561,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.7497197870003277,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.4596,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.647359750714791,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.4577,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.8045132536886852,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.4542,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.7381263939732362,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.4582,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.518796678126342,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.4501,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.6122396496055299,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.4431,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.6252750748414262,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.4662,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.5214987509431108,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.4351,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.6025411938938965,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.4417,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.4451394286089756,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.4641,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.6523846319984302,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.4487,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.46672770886257603,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.4426,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.5014674835215647,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.4345,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.4959038876389297,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.4369,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.4572260239425412,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.4471,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.46062113203492255,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.4458,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.47388308448063976,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.4503,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.42467581423377665,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.4339,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.43552155405244963,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.4514,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.4421453535600298,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.4629,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.4198380433812562,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.4291,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.47086723215237825,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.4473,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.4521159878446533,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.4517,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.45464859579017175,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.4459,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.4396201367077948,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.4337,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.45526558304701853,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.4339,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.44693199283069407,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.459,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.4481055198546284,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.4664,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.4249178591826331,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.4423,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.4332499867666868,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.4472,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.40945963276578623,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.4506,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.4263714495427584,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.4419,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.42669875713704475,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.4434,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.41271982934960183,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.4764,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.46006498686699737,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.4384,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.42704583642930155,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.4629,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.39909696654983967,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.4629,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.48236176763681227,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.4664,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.46725846042658714,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.4529,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.4378058393873011,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.4549,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.4694068886939854,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.4488,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.4344575275179412,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.4483,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.4603789961000926,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.4409,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.4561378462561444,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.4477,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.4734618681349037,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.4614,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.41761671532519473,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.4547,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.47849384869682987,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.4568,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.40324771299332113,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.4329,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.4529629419109575,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.447,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.3924022697494419,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.4493,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.45722363917422754,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.4625,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.39388444209909546,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.4523,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.4612075515264238,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.4491,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.4079295914914598,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.4573,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.43802541021348357,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.4762,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.4316758147821095,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.4618,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.4056783653421317,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.4583,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.42179765303599387,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.4573,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.39165801163011893,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.4477,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.40814946884128334,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.4445,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.40752788464624534,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.45,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.39337483140692964,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.4551,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.43691455108172317,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.4589,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.395259211183542,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.4557,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.4183596787536176,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.4541,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.3967346349768188,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.4577,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.3758944131825608,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.4408,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.40471729672294965,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.4532,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.3965309589519926,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.4519,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.37937758472204697,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.462,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.3972304624799626,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.4525,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.41859762056860106,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.4517,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.4004667452090876,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.4571,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.38855998740273806,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.4446,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.40497650490446196,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.458,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.4306616187341871,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.4663,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.41032581009421304,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.4443,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.42302967821407633,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.4452,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.39552379604552995,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.4511,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.40121069096311446,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.4554,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.3932466546235397,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.4307,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.41632383534138784,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.4552,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.38868194943013573,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.4528,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.4188478509493282,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.4592,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.4070497471540291,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.4734,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.4022827450769274,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.4645,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.40131039127533674,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.4442,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.40213130579257494,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.4636,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.39249724908847045,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.437,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.4092422818175595,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.4599,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.37162655404106326,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.4408,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.3932285250803531,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.4545,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.39025694584026693,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.4545,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.38463553383168425,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.4541,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.37721638437668986,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.4522,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.3863712776270255,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.4617,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.3841459937664575,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.4446,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.3653581987669319,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.4471,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.3715515329094348,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.4418,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.38832201253773546,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.4548,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.3803939864549916,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.4523,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.3668168814187104,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.4488,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.38802396199506645,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.4586,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.36980676738899143,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.4345,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.3636777276455012,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.4508,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.39110956354332754,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.4565,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.3853769043781521,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.4498,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.3935086926017729,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.4698,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.4023199543851552,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.4513,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.4109781090396557,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.4644,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.3825829986757174,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.4414,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.39960047959527245,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.4464,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.39843530078101336,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.4774,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.389366963557674,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.4681,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.36483771140634785,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.451,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.3901348914295112,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.4542,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.3854363540710687,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.4567,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.3759642302670971,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.4685,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.39145436868678557,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.4448,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.37785910807504847,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.4496,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.3883988232563627,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.4473,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.3840039615229006,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.4569,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.37190271060830166,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.458,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.3854589017087581,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.4514,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.36479702093651717,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.4563,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.3702679106768453,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.4487,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.37590491684642885,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.4509,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.36094837136100594,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.4627,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.3770783314099437,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.4484,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.3838479757236588,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.4562,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.3727229069970339,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.4402,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.3618604655796233,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.4475,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.362900426222159,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.4423,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.3722304353844517,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.4499,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.36908406773829217,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.447,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.37318343741538496,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.4571,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.40155667939414713,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.4485,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.373499484240249,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.4551,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.36665542696879283,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.4472,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.3835588217904844,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.4673,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.38105531523183706,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.4582,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.38172318869565824,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.4635,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.36925643597269564,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.4457,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.39519358965888995,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.4485,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.3715699466463917,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.4532,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.3737901672291277,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.4646,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.39593515102054727,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.4625,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.3999443861341451,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.4402,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.36031090261769877,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.4495,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.35723289556408117,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.4506,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.36655665520936187,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.4515,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.3601826450730418,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.4529,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.3481867777319978,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.4556,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.3643176983394164,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.46,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.3760995941004457,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.4576,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.36024000590989896,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.446,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.36491302349945115,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.4669,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.3769509225010582,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.4541,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.3612256898375487,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.4446,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.38236803299310296,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.4445,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.37434343380651675,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.4502,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.3607468459202837,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.4537,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.38246939907170496,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.4513,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.3564865312851963,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.4586,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.35791819216878623,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.4471,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.35608576796897323,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.4462,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.345048162339605,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.46,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.36151253926101795,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.4466,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.36103049518525165,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.452,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.3520396371431307,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.4515,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.3491987564353251,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.4458,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.3590024852448456,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.4541,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.3492139360006747,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.4563,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.35198646179213067,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.4305,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.3584232549353952,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.4578,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.35803317114706673,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.463,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.35088244993492224,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.4475,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.36812816489474537,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.4505,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.3584676319965831,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.4535,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.3627928527608116,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.4526,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.3654365252914518,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.4627,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.38794043547069706,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.474,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.3672388828660255,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.4406,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.3825054948744728,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.4538,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.34265290331923437,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.4454,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.3574209896735425,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.4532,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.3569744570730435,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.4527,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.3563813887392998,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.4549,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.3730099031080738,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.4379,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.35430159364231184,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.4555,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.36730190630917975,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.4594,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.36394633357339223,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.4516,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.34165302780565737,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.4534,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.35707171440052177,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.4483,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.36183990670634325,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.4354,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.3499533806109604,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.4423,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.3607678032235268,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.4541,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.35625153413301414,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.4308,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.36318508911307673,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.4321,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.37070862618986444,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.4473,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.3649229732644494,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.4358,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.36022285302943574,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.4448,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.34713653504956715,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.4457,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.39859160053508846,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.4488,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.3623390905944738,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.4626,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.36604525707422403,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.4538,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.35051826176229406,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.4476,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.3556448931841828,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.435,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.3615242434696281,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.4423,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.3575741948798601,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.4494,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.3466180172979913,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.462,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.35612625985737106,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.4496,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.35902007277862974,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.4481,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.3736098051035649,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.4502,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.3483952908683482,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.4424,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.35232246629273883,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.4491,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.36140522893007654,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.4478,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.36131226125960114,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.4532,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.38202124409561045,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.4521,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.38010762859048647,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.4529,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.34657904101297965,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.4452,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.34944304955262323,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.4449,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.38360421967547426,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.4613,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.35847400313829997,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.4378,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.3519931171505103,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.4693,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.35004395971305413,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.4443,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.35793313535384397,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.4459,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.35305372532052476,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.4569,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.33524752713665457,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.4411,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.35909050875227594,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.4579,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.3638638081260243,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.4491,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.35317674959730233,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.4649,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.3636085763402221,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.4682,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.3454647641603134,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.4503,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.3488466910649181,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.4562,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.3459483289681123,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.4421,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.35018459819355263,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.4353,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.3451780923412394,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.4301,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.3589197515884094,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.4647,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.34988304137396586,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.4412,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.35045398264874006,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.4462,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.3374895162491545,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.4455,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.3809474383243772,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.4425,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.3683618150474189,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.4614,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.35167511503602567,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.4396,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.4903815172908818,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.404,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.6833829618112155,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.3586,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.5164273881693892,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.3485,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.4710651305430315,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.3574,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.7405973271144953,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.3477,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.8774549608556687,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.3532,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.5949246451183112,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.3523,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.5695407452257409,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.3603,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.6977624789758499,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.3564,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.6280555483480919,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.3513,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.5129904893771134,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.3497,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.4932155550873851,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.3445,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.48684477826510286,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.3464,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.5404824685285746,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.3464,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.5283373600476179,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.3339,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.471755920040289,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.3483,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.44523865411967395,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.3619,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.42289709060071873,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.3338,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.46639743300132,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.3587,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.4699015855839437,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.3459,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.4342676275140877,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.3531,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.4078703298415023,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.3457,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.404180379801703,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.3455,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.44055946072669755,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.3451,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.42508415026336877,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.3458,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.41340213407949855,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.3627,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.38060832788106597,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.3508,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.4047829684018124,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.3559,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.42049405214372243,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.3589,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.40151787655434285,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.3494,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.397931204249719,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.3547,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.3874150499860169,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.3587,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.3979950308240945,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.3546,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.3956228195010122,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.3474,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.3995517880534548,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.3508,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.3912981635594149,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.3459,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.38429201246221295,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.3396,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.3913644249116605,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.3473,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.3893174690198067,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.3585,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.40108023038283047,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.3545,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.3748434780032013,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.3545,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.36836115645941314,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.3347,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.36751766587612705,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.3458,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.3761288890219984,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.3527,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.37539707855032634,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.3392,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.3701713710379611,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.3408,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.37245287732169907,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.3545,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.3779889659603261,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.3457,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.3785137879052379,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.3489,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.37728993460255006,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.3418,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.3843435492318358,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.3465,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.3682637214762445,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.3487,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.3619130766455223,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.3445,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.37246662612499015,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.3459,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.3765817006254543,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.3409,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.3852102164294755,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.3593,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.38810666309032343,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.3389,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.37015900740367286,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.3532,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.3717683373795862,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.3469,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.3764176019757569,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.3478,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.3854168179309447,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.3485,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.3760641687723411,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.3372,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.37049980776906355,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.3352,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.379387210827923,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.3493,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.36996190374254945,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.3598,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.3696100408629272,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.3291,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.3792541124764722,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.3299,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.38421906118853677,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.3486,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.3885699895493243,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.3463,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.37605886573577113,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.3462,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.38056238834873224,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.341,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.36434354475854885,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.3473,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.3802748579444638,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.3365,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.3690186117727422,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.3348,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.3730976964972425,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.3356,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.36452167634400867,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.3339,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.37372183168893885,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.3433,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.3813526700006076,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.3586,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.37514413537813535,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.3529,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.38749745809533825,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.3487,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.3725778009267143,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.3622,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.373294498999819,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.3456,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.3680769257021167,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.3399,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.37164048886082846,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.3492,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.37032353326183465,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.3388,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.3662658137545875,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.343,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.3846332780131944,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.3503,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.3669131856130684,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.3463,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.3697193282555427,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.3479,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.3730782817048451,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.3476,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.3659052084020209,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.3435,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.37536872729592535,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.3452,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.38382658407733344,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.3525,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.3739601210853269,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.3512,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.3711487475655448,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.3417,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.367597509830736,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.3413,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.372701225603454,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.3488,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.3723069822955331,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.3497,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.3770749524735257,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.3534,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.36354868081798575,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.3375,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.3705481815651156,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.3574,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.37638884059917466,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.3607,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.3685224687884592,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.3495,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.3682923843207206,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.3402,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.3749680924041334,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.3414,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.36642643632386707,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.3445,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.3686038061666725,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.3443,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.3634587702685678,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.3454,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.3718799180739304,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.3655,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.3722803857637191,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.3536,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.35925922502047103,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.3509,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.36723176676309305,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.3459,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.3797592516518572,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.3506,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.3678472419203584,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.3638,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.3741788665118279,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.3482,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.3641103836758709,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.3461,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.36581474628933514,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.3577,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.37156271435191696,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.3449,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.3767402433110905,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.352,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.3547340106311124,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.3334,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.3768227795282951,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.3387,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.3658362740817512,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.3422,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.3698000327490867,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.3383,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.360814154378306,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.3386,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.3713146315917481,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.3299,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.36166910042080586,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.3481,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.37998227907785453,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.362,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.37484633929705025,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.3581,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.37227464860896003,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.3438,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.37192771263983815,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.3482,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.3571689119000214,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.3375,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.3616974202981606,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.3349,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.37274635055705746,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.338,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.36296373843351853,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.3464,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.36214878585557037,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.3486,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.3665387942274832,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.3312,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.36363816351416106,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.3499,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.3594379981146388,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.338,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.3702666008427339,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.3507,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.3712760305877203,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.3467,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.3683229478079645,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.3445,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.359499062016618,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.3418,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.36344962056483077,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.3297,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.375170197251079,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.3478,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.36342245457706523,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.3464,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.36672057698045235,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.3628,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.36584178456480604,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.3489,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.3660807400473438,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.3407,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.37152091458048114,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.347,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.36007420534936946,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.3394,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.3625102171944161,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.3369,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.36765749484600324,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.3347,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.3726157676205438,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.3418,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.36205446969682403,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.3548,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.35972361522373403,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.332,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.37012309982353,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.3491,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.3661798953652562,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.3509,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.35600849738819135,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.3267,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.37130751970276565,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.3541,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.3637419512051319,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.3444,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.3628193701119143,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.352,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.3678489740990106,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.356,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.3733371595668351,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.3599,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.367589385331941,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.3525,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.36418593124290766,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.3389,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.3836503012606182,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.353,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.36465012306298333,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.3368,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.3609040275893938,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.3527,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.3573909712852158,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.3314,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.37599924638740223,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.3538,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.3630611850845974,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.3438,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.36811809622154706,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.3501,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.36639969153258606,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.3472,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.36734805216041505,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.3477,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.36415629854927495,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.3503,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.35885879285936984,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.3413,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.36469172927176857,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.3378,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.35886287954922647,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.3454,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.3649982395744445,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.356,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.3597522384904767,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.3445,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.36645121450542956,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.3518,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.35655746454969073,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.3388,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.35676039393455833,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.3332,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.36494598888130214,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.3433,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.36261843591907816,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.3424,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.37095225115523633,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.3519,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.3607674728730146,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.3483,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.3600090341158943,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.3372,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.3653306755946259,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.3398,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.3580101909379665,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.3469,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.3578332092630076,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.3251,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.3684958345875703,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.3287,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.36240821156436964,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.34,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.3707132237683786,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.3501,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.3616269281271386,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.3492,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.36509980412070897,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.3518,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.3566921624100369,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.3447,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.365206989710881,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.3549,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.3597527368333769,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.3469,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.36911081742401974,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.3487,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.3591401562657883,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.3381,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.3588159909946332,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.3342,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.3699291672468204,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.3568,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.3634150073224162,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.3387,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.35406864193337456,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.3351,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.36236807760942363,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.3522,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.3595736465470377,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.3481,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.36231102903659174,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.34,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.3676992110546578,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.3513,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.3640149703029678,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.3485,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.36207054378484665,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.3343,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.3718975866166661,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.3468,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.36852480979447916,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.3579,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.36055278151084313,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.3379,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.3627968270593616,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.3469,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.36041353108900887,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.341,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.35843851338953153,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.3377,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.36603401047106104,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.3433,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.3528539555301012,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.329,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.37058821436549316,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.3377,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.36044846031548966,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.3432,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.35996330743769195,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.3338,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.3524772822746245,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.3351,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.35569049679940984,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.3322,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.3685384881088144,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.3415,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.3604143469281269,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.3491,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.3671184659205779,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.3527,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.36038025784191985,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.3373,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.36062376775537697,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.3462,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.36019931253797005,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.3469,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.3588033734904503,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.3357,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.3701006103408785,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.353,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.36170705786760826,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.3357,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.36393704796278065,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.3459,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.3651735079855879,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.3467,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.3702115707088083,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.3424,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.3610517558766831,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.3325,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.3573879008657575,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.3375,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.3677715488301095,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.3436,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.37449998675755203,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.3373,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.3652283985231186,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.346,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.36144782562608857,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.3391,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.3597388598678272,
+      "learning_rate": 0.0,
+      "loss": 0.3465,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 1.899139087207424e+18,
+      "train_loss": 0.5717073486345571,
+      "train_runtime": 13070.0674,
+      "train_samples_per_second": 12.089,
+      "train_steps_per_second": 0.094
+    }
+  ],
+  "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": 1.899139087207424e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}