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": 5.886713285535662,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.9738,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.321873246573988,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.0241,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.030745703064844,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.9869,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 5.969861427357096,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.9799,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 5.639863844073304,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.9945,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.227011250930781,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.9658,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 3.9685046298856546,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.924,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 3.701266162827651,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.918,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.4860039816485497,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.9031,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.276346461741084,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.88,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.2926337061679,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.8941,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.1806871293125445,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.8575,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 3.4068394689599093,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.8505,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 2.937113839992561,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.8422,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 2.4864553286993583,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.8349,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 1.9925256760517118,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.7874,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.1633270411599086,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.7895,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 1.8968940023088041,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.7921,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.3163328729853012,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.7763,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.14107447072635,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.7581,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.1579867665650907,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.7376,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.2216017939364199,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.747,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.098172458681128,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.7395,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 0.9202589455336603,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.7309,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.8164050928001017,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.7298,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.9913760975931764,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.7147,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.8397583571947503,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.7131,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.7255899676220274,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.7116,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.7159885625871358,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.6942,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8803735686074097,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7095,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.751557699048861,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.695,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.6602041935669695,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.6817,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.7755794048749889,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.672,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.829184174477792,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.691,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.694350603191771,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.6771,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.6369751866355542,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.6736,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.7778575582055787,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.6868,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.9293132509396508,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.6894,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6362009986855001,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.6825,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6263966097598189,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.6681,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.6913844933872928,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6545,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.6248911152092731,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.6638,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.5907099266951296,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.6711,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.5690560416347097,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.6631,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.693206643126378,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.6611,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.5285569492963472,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.6619,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.576714641861238,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.6565,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.5473834083304255,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.6524,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7136089514098295,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.6572,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.5901506476253272,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.6507,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.5687873011419605,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.6496,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7945970039879262,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.6547,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.6138680958330435,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.6591,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6525356927156145,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.6461,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.7370220180928954,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.6625,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.5285048516480975,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.6402,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.8685016775535046,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.6416,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.6435318906304228,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.6491,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.855747625576641,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.637,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.7922967780729577,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6483,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.799626606001735,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6508,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7361868303539818,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.645,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.7579821950378939,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6474,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.713765218437119,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.6371,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7549990042910213,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.6367,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.6615995097844751,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.6456,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.5724924354198369,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.651,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.7858477341635914,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.6263,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.9279783394102742,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.6418,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.7242075777321937,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6256,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.7174276376297689,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.6277,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.7566901620044731,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.6137,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.6451250343264461,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.626,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.8968692247489776,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.6422,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.1386705268862403,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.6384,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.1963382478032738,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.6371,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.0808009246662957,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.6245,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.7665028715920762,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.6061,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7205794181924197,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.6318,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.9814823333455319,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.64,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.1003306920701008,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6432,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.2546279054671168,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6316,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9643774031683621,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.6397,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.080390691335537,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.6443,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.9731026691415814,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.6299,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.0425471480912056,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.6215,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.08040119224057,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.6444,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.7438658001335307,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6364,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.0659355565278477,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.6215,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.8833912037199415,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6308,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.2728443145556187,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.636,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.6892519872781763,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.6158,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.156586690111508,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6384,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.2389327396539433,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.6315,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.748749729151802,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6257,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.2568480816390646,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.623,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.8390249899167179,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.6245,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.4591245457402908,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.6342,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.9310767698371853,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.6179,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.1286952833009463,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.6268,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.6844968198584374,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.6303,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.8154968267690889,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.6204,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.2767153742517734,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.638,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.1866767231746425,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.6269,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.9272561425486815,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6234,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.0886462439069062,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.6366,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.9459453359121199,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.624,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.2803751038191122,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.6105,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.085674326109735,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6123,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.404408508896452,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.6251,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.030872773341314,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.6208,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.7576959597217354,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.6303,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.000001168484571,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6049,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.541469789352896,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6368,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.4443359771748627,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.6246,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.2038928911322444,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.611,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.653523049820692,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6123,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.1239827752283014,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6237,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 2.0554655578599035,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.6323,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.658079692663371,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.6379,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.6844731301156086,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.6406,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.343477842621306,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6318,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.4410672347509104,
+      "learning_rate": 4e-05,
+      "loss": 0.6216,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.0251451960961018,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.6289,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.4991306597242837,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.6406,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.125151540059418,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.6185,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.3839913288336527,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.6246,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.0208414650140982,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.6163,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.3593230409261141,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.6112,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.0920290510476713,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.6205,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.1768190908712963,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.6152,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.1757747886394048,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.6395,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.352234228309438,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6187,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.2543783013569396,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.6161,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.9657296952617,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.6229,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.7652941500221218,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.6098,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.9035211923468522,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.6234,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 2.1789181167945695,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.6068,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.4649794850938271,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.6169,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 2.089771091126668,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6159,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.4239210523208097,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6198,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 2.218588418957195,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.6203,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.4750933721274817,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.617,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 2.0377525033937083,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.6101,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.8168478564144643,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.6202,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 2.112322503202226,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.6124,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.708403509780768,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.6234,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 2.290727691547323,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.6303,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.9690764710851107,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.6271,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.5147500933701057,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6041,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.4829374163322455,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.6128,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 1.7714332939349116,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6146,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.2405487157239516,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.6234,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 2.147680545741031,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6137,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.7392686983705166,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.6078,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 2.11068766036767,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.6166,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 2.204189120285359,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.6329,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.216759192152491,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.6194,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.3842968838586462,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.6216,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.1743581440419888,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.6186,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.2580269857118314,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.6153,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.9640208478310514,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.61,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.2635327417783213,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.6084,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 1.170982377314407,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.613,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.0196138523691463,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6105,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 1.4853573849603257,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.606,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.7982012660895725,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.6338,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 1.435767732244566,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.608,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.7928581694496983,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6079,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.201598078558288,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.6248,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.7251412825623611,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6133,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.1416596941022128,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6219,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.7400980951100917,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.6047,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.071601857847586,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.6008,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.9221307650306441,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.61,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.6570072112620979,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.6093,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.7366337537823385,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.6004,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.6584612561308628,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6122,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.698263825927111,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.6004,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7328174653026961,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.605,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.693320527610826,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.6079,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.8587615090851715,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.6078,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.7630651201695444,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.6146,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.7229543319530447,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6135,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.9617531474889677,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6089,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.9606574525970599,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.6184,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 1.0113080030266974,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.6088,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.8936267995852759,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.6111,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.9671523836574117,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.5966,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.8832882698269758,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6153,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.8389414211656023,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.6086,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.9569381115517818,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.5956,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 1.2260184045588216,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.5935,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.8801884508359951,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6098,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 1.4485800556507744,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.6151,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.7260000256853286,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6214,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 1.300737193256244,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.6073,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6539602733072436,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.6047,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.9883659778727418,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.6029,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.9372904902701531,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.6187,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.6993606390296205,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.5972,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.7996908704721808,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.6099,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.7945254746034353,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6176,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 1.2216839772198356,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.6078,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.5779995145041708,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6015,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 1.130846291471605,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6208,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.7841478226572511,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6056,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.9188505497770951,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.6117,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 1.1088765345351557,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6093,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.5252712146436849,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.5925,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.8986295807878718,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.5926,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5869073368204478,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.601,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.8764516531658003,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.6046,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.8868223987099823,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.6062,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.7740134725383282,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.6052,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6645244326218527,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.6077,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.7330671108511021,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.6078,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.883583352462517,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.6171,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.6822440948959668,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.5965,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.8428213087724521,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.6079,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.9093536583568387,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.5943,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.7672133824347744,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.6,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.8133787474369356,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6115,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.7593935139099012,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.6205,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.9190618343832762,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.604,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.6382959096473445,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.5921,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 1.0576373657560338,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.6007,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.9674323404478641,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.601,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.8678442142790646,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.6009,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 1.3194018077783458,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.59,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.6149788318455869,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.6237,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.9552465643303157,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.6059,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.9478156164434826,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.607,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 1.0722679325437006,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.5982,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.7784764073807204,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.6283,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.8493453931862156,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.6027,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1430794829404527,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.6052,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.833187665984086,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.599,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.882570690216721,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.6023,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.7507312227072489,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.6105,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.7541402756311706,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6003,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.9959233282510728,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.5984,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 1.181161655340514,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.5997,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.8898241931283064,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.6103,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.739690060331508,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.5895,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 1.0923562994545262,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6029,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.8513262752115228,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.6778,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.8720314877095489,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.5218,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.929564578802369,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5527,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 1.0560687829083248,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.519,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 1.093728209789614,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5195,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.7547113439699064,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5282,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.7700471138316024,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5161,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 1.1077989331270541,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5297,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.8109862090688862,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5356,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.6343905978958119,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5423,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 1.0061837667040223,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.538,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.8842661917797092,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.52,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.6137901368434956,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.4961,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.7411458942006218,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5578,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.9569976723287263,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5104,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.9372331718078606,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5155,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.5767309272784537,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5487,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.8992732888790219,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5181,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 1.1171856372616362,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5176,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.6513589028863577,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.511,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.8957427347867017,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5414,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.9451106935185146,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5318,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.8201581199534183,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.503,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.8065773010343129,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5177,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 1.0395731304989284,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5388,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 1.1263231667945555,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5164,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.5952805638941888,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5073,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 1.0145777100887554,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5481,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 1.1252690674587018,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5165,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.85538292418627,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5377,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.8073753353668225,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5241,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.8300851247121933,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5053,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.7001877966603064,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5189,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.7737329309568359,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5302,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.6379314848499399,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5193,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.6808118048116634,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.53,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.6569347588461063,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5209,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.8329513191784674,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5106,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.8949662362906731,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5391,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.7565957707232507,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5108,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.6241971736079042,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5145,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.6511289589622153,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5262,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.6698227614022944,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5233,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.5607110004786175,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5418,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.5277271230295384,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5191,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.586772191241864,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5294,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.5247871445170764,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5405,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.5817963054346201,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5106,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.5748099928651549,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5366,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.5555663474035389,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5209,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.6205880468771289,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5533,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.8521316145711896,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5057,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.9378067077613051,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5404,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.839050822211001,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5282,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.8394350534258326,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.535,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 1.2125437189301778,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5122,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.6770005504809133,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5166,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 1.1391560536929652,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5357,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.6553274522181346,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.525,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.7703295167268338,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5189,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.629828761712078,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5384,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.6470752227750499,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5222,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.7294736352723901,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5283,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 1.0216154682186331,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5159,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.5334233127207627,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5236,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.9824291190061076,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5285,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.6832824168993187,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5197,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.5459329217727772,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5222,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.5334951054856925,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5294,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7379442023146914,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5097,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.7001754331257742,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.523,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.6054049464608923,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5185,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.6775299662972235,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.5467,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.6806512329981531,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5125,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.5991814989717603,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5427,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.7632482000497032,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5218,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.6732132704710265,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5311,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.6888639422608139,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5455,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.809070422051105,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.4939,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.6526581563847923,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5143,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.512487300425761,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5227,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.6159431879792052,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5235,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.5197269939877349,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5274,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.7030538198547035,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5244,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.5976036430781447,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5331,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.4648905157619526,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.506,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.5631931266172467,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.528,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.5455425249578164,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5317,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.4556805288873786,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5145,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.4989251437167731,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5159,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.5339918852994691,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.543,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.5165182930977955,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5424,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.5365383411488943,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5061,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.5558432161433515,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.5255,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.5775856140571393,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5233,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.4935520111096479,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5109,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.6887560461564713,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5402,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.5123541717092784,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5085,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.5489084085120589,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5353,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.56525831523951,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5338,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.6231342394107532,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5451,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.5336415348997835,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5379,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.585533258929108,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5211,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.5890216431705804,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5039,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.6077437305693687,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5269,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.724726220682385,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5404,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.8094457817866517,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5278,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.750365128160768,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5131,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.6608534442900289,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5322,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.80712042688547,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5343,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.8411364772396679,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5253,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.5934170699683144,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5011,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.8242004501756653,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5285,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.8589465305442091,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5313,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.6200510264851975,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5526,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.7047720140532272,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5077,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.49662868664720916,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5348,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.7217197124149429,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5417,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.6489949880644144,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5133,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.7914156434892553,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5363,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.8715370517138579,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5294,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.9286518036793622,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5352,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 1.1093226357357493,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5216,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.7110616293001466,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5325,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.7758750800225914,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5304,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.7398877420204036,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5156,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.8276879583427771,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5173,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.5669069373932424,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5223,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.822529976886152,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5222,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.5078244979702431,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5197,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.67118188761622,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5421,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.545527825497799,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5458,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.7355475383516541,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5353,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.49178702508749234,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.4983,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.6215632568302426,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.521,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.5485378393240492,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5063,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.5768779223045821,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5437,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.6903696870712761,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5134,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.5673028027527057,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5321,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6434110838202574,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5065,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.5938673533214994,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5207,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.5238054177207452,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5347,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.5320432492891867,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5189,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.5759624156062807,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5223,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.5973655868866852,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5123,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.6024054245648018,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5447,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.6279425962127956,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5268,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.5493941034707197,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.4923,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6878473247750029,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5428,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.7399693316025017,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5247,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.5222775252715142,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.5403,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.7082658804271255,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.537,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.5536913333763057,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5025,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.5283434875274261,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5276,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.6028103391350018,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5062,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.7773620540020483,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5174,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.7184775425923093,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5261,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.5356833707466541,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5131,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.4958998220151337,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.4944,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.5042024859378739,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5285,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.5114359893586338,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5136,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.5072395876369009,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5271,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.5485339425564689,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.535,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.4772757525343688,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.4994,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.6210213755584676,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5612,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.6167219610007528,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.4976,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.7186910311827747,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5192,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.5904160618024878,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.514,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.7304666028417668,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5607,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.5907808157431252,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5225,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.6169115614412225,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.545,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5377880149214835,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.502,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.5122077113997268,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5541,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.6030200639683535,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.4987,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.536671081446229,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5183,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.7488743188639275,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5129,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.732243010965671,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5253,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.6445142309741994,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5087,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.6423231602964388,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5407,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.5461247030543995,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5125,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.6529961332225183,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.53,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.4721287201394258,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5164,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.5338981922083464,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.5087,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.5330494485568628,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5271,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.485067376012994,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5271,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.5309571324760306,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5144,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.49834695134607837,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5099,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.4807468135047916,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5097,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.5341485159705616,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5494,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.46619689544618986,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5108,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.519786892665976,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5306,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.4630045254767178,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.4987,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.5927681349925237,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5374,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.5944667295314437,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5378,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.5256580322827111,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5217,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.5347077283062237,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5383,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.48814174866377563,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5484,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.5738962902330796,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5339,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.5009560292086122,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.4729,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.5222417459090016,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.5182,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.6238219375433973,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5478,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.5062019601440908,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5149,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.5273822746787235,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5225,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.44955170607590916,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5031,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.5604906443010427,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5364,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.530625904948473,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5278,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.5138364562141364,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5015,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.6128048144426699,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5414,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.466955825298351,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5234,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.5808109206720673,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5326,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.576299871987708,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5308,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.5043705233223946,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.544,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.5675734936100325,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5035,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.5429900658334382,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5369,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.4860314977470431,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.523,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.5489792135926282,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.535,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.7000799705938923,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5305,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.49514990990010543,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5087,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.7034885777163787,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5267,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.4229250184027541,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5211,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.7286739684181357,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5237,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.5152519385710176,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.508,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.6159227722086239,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5133,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6307202511704183,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.516,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.5470668687425446,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5385,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.5889691399568805,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.53,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.6597967074190344,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5234,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7252152199842423,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5395,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.5368890661566711,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5116,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.6935051202309179,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.531,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.6367746999371215,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5184,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.5678936455610583,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.519,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.7780111383123156,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5462,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.5459203581658129,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.5226,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.8131099472198624,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5225,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.8490901973355692,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5041,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.7071847987743051,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5577,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 1.0570784304863279,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5134,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.5944748519504713,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.502,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.8439555519271933,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5334,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.6683592029351415,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5273,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.577343910642344,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5265,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.5914464233976963,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.5007,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.5449252264606232,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5248,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.5594232254230531,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5161,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.5578901925237421,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5096,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.8311109481620531,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5496,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.5812294385846675,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.5563,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.7462452159470773,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4094,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 1.313696462865701,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.4204,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.7151966619730175,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.4115,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.6058496603457089,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.4077,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.6645536887140224,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3914,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.5894592950256627,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.4095,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.6680897627736958,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4054,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.7711097320738187,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.4202,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.5061393881240831,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4042,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.8364562441851499,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.4288,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.5244196551879208,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3809,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.799860779789649,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.4142,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.4828774182194531,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3927,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.639298979715285,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.4136,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.617390668912843,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.4136,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.531659117491725,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.38,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6219680300704936,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.408,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.5314381381378128,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.4273,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.635130809810372,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3873,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.501610431207283,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3915,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.7575052524673342,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4337,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.7903837194589236,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3898,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.5507240375385452,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.42,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.7125701825169786,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.4056,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.5669227936395504,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.4145,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.46037663036768633,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4057,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.5673887123277307,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3782,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.5272834793710841,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.394,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.5851717528542905,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.4236,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.43662995362405305,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.366,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5526023049478579,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.4135,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.46179066155009724,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.4013,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.6103983631108844,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3987,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.5036496895775389,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.4094,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.5308162815064688,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.4018,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.4888113370269735,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3988,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.4677180534616612,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.4285,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.4382775514135411,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4083,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.48232463334697795,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3941,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.430709564493224,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.4371,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.4331998783636931,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3793,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.45745212146298053,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.4125,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.40392866982611386,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.4029,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.42001954716437007,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.401,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.432357737961763,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4115,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.48486867220443475,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.4094,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.4172825151497744,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.417,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.4476809016104859,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.4019,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.47701315845406006,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.4275,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.43423227173984186,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3949,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.44528512955682487,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.4176,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.48828050339756157,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.4229,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.44434790937336516,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4159,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.5630443242083172,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.4173,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.4311321935607088,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.4021,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.46195832536754833,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.394,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.46910713240585944,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.407,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.4294291748923378,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.4033,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5343528222407853,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.4239,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.4332535060497227,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.4065,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.55550498889403,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.4006,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.4147722130030719,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.4107,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.47835316843055714,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.4075,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.4404231772646878,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.4186,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.43607958809913716,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3975,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.5239686109036781,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.4218,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5535501737659442,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.4111,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.4715426541459667,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.4077,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.6111277843416129,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4139,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.40863419029411524,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.3862,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.6140887765944696,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.4313,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.45656408813597027,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3916,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.49124957951326736,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.4405,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.4816389096314662,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.4071,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.4369434364673721,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.386,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.4160394088073019,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.4302,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.44256806812686683,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3943,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.483744995967692,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4135,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.43697845554293263,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3898,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.5616101356797926,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.4274,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.3964829559094195,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3924,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.5056491752257644,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.4317,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.43970341763020654,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3965,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.40801761392585145,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.4066,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.43054059161581565,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3792,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.44517591732969475,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3983,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.4583539064676636,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.42,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5211020641834352,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4003,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.44172568177255406,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.4023,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.5718414351355844,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.4193,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.4090289738266864,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.4212,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.5843493196840029,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.4124,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.5156055032629983,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4029,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.4514342024677713,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.4065,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.45493877955328177,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.396,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.4373230815895015,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.423,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.45483780106672694,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.4232,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.43628684551404834,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.4043,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.504966673516326,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4188,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.4543331559355677,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3979,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.4410235742793327,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4189,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.4148513673263069,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3762,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.5148717376096094,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.4175,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.4251298285674877,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4194,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.5760423513599693,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4072,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.3873029476430978,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3993,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.524508554064902,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4369,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.4045542971332209,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.4044,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.48939316731225396,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3794,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.39206344914772023,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.4072,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.48886235016690494,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4105,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.47401305422700624,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.4315,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.4855114294347693,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.4045,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.42646330594761805,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.4372,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.37375412139744413,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3784,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.4086432626275759,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4126,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.4164312074784168,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4065,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.43695430281256814,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3939,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.3970999175642912,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.4212,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.37542608403757916,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3958,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.3858516843667893,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4266,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.4071831179966569,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.4213,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.35135730465548204,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3908,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.44232581248399916,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.4185,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.42378638111505534,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.408,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.4784403422710913,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.4152,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.39661981409822716,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.426,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.4038570493667419,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.4,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.41753480372911195,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4076,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5247660771714999,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3977,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.3562020008715709,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4066,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.4369077919255907,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.4223,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.41549232404499736,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.4051,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.41875054080360014,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4177,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.40290532556366104,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3979,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.4333144523562928,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.4419,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.38189476353533514,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3933,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.4267749305867019,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.4187,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.46970017853179624,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.414,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.4699215023348641,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4087,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.4092694416744476,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.402,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.49589685911236886,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4037,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.37898836927867685,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3963,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.4663996099563458,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3909,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.43228615125466674,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4338,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.4107238595344976,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.4019,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.41754880470593597,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3965,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.4474720972925264,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.4358,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.4161589896549677,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.4052,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.4724537775734016,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.4016,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.39888793983681015,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4215,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.49739759053163674,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4281,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.4359806477274325,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.4073,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.42583430094679586,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.4123,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.44165242480183,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.4136,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.4604600947374686,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.415,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.35202184947039733,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3932,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.47878419775388714,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.4207,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.4442887167958696,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.4269,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.4007829438909601,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3927,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.4353693570727875,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.4141,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.44386533261636374,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4164,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.3945332873495858,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3944,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.48262743203366965,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.4125,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.3615299792106722,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3991,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.49364775793025467,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4398,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.3767085619696233,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.4139,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.3856903473352313,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.4155,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.3581948908558457,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3993,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.3983399276455987,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.4085,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.42220619310366264,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4019,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.4583799886764314,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3981,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.4078775580223975,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3911,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.4568835673649418,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.4167,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.4211881708924004,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.4253,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.4286096621103251,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3776,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.540449657015209,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4458,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.5569095101014125,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3871,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.4930544266241417,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3915,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.47209371512577997,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.4284,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.49497345745177246,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3999,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.4224445372726848,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.4003,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.6254821809704775,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.4245,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.4288411693203483,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3944,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.5110260786627349,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.4214,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.39242949449084935,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4018,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.4921043911869658,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.4205,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.4489589668959947,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.4007,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.4326934558570849,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4085,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.47442262357656306,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.4058,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.5095366872091875,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.432,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.5518364377359785,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.4078,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.423405120695189,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.4007,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.5016901214316082,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.4263,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.3936830611835646,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4128,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.49467287884391614,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4306,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.3773735022743772,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.384,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.42277467660250073,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.4222,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.4213306586645914,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.4157,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.3622496326153911,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3881,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.36763571789886457,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4059,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.3609514224473155,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.4263,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.3782436792228902,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4225,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.33798070667006513,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3935,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.3773716600631473,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.4128,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.37536259512499404,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.4232,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.40091761657744074,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.411,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.38503766543449325,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3796,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.38477570678113177,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.4185,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.4572550427010871,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.4169,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.41040209555268586,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.4079,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.4516343399027948,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.4044,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.41968415035103723,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.4151,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.44527733969957645,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.4068,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.41593057349360124,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.4168,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.4323319273928237,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3942,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4307907717867101,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.427,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.4462842925503964,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3897,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.46096119183680273,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4252,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.3872164299393199,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.3924,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.46635377601954564,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.4024,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.3725984420933388,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.4228,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.6254520591004588,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4146,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.3874825844896869,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.407,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.5522284035117087,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.4056,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.43848879755104886,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.4078,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.4206888390890596,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.4239,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.4599741091540947,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3863,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.3780071871653579,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.4195,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.4605863079405191,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3839,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.4412710235856544,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.448,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.3843650492069743,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.4,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.42018565635385413,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.4234,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.3849424375514065,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.4002,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.41010988569103723,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.4098,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.3950894737785367,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.4103,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.42823671803244834,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3985,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.3764194776471229,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.4054,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.42064874999377944,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.433,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.3692571474466556,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.4085,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.34943441685663595,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3806,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.48313612172327597,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4133,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.40606855968409056,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.4235,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.4520981245071089,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.4014,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.40054498718442977,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4055,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.3887118683780621,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.4085,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.5531426656744965,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.4219,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.5197791215873427,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.289,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.43798907414818855,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2974,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.6162923029929174,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2942,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.6866527810033114,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.3141,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.4679956075790141,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2926,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.5375016452605005,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.3036,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.5332640638015611,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2931,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.5362902189351479,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.3076,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.4507900618136926,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2919,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.4787909628858425,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2892,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.5262344986357452,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2977,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.43187688879623753,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.3139,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.46206754624325075,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.283,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.4813754987090218,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.3032,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.42318241847238697,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2796,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.4613368354647796,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2941,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.45916313274838894,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.3041,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.4482142734977057,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.3016,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.48207671852236034,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.3069,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.4042230529192369,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2879,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.4386018113019707,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.3212,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.3528308181914062,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.283,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.41652576212731757,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.2866,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.3663792077662654,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.3163,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.38691019763787854,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.285,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.35897847187232185,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2892,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.4147928002059536,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.2857,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.34579695551660566,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2885,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.39915658505832463,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.3264,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.3493962745802246,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2911,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.358622449821768,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2874,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.3657732697546043,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2979,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.357159701758474,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2939,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.3728093921161006,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2928,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.32821056919990366,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2697,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.3688334158348578,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.291,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.3583174249137979,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2951,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.37664724168648545,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2852,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.34678917580096313,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.296,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.379406431977096,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2991,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.3625382533381951,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2774,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.3420771818965431,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.3017,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.38497207977479264,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2964,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.3477811999165168,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2937,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.35002657246268193,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.28,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.33453187968218123,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2879,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.33740392385789525,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2936,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.3469508273680211,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.2998,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.3457112217416669,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.3084,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.3311116539490288,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2911,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.33450821865185626,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2708,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.3344552926485109,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.3148,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.32122628959783706,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.2933,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.3511152639880121,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.3016,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.35830439233442024,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.306,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.34448854042088,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.2833,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.3482391226933635,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.3131,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.35260880159921437,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2921,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.3361214093574521,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.308,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.3369248939771357,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2828,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.3320025079391118,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2937,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.3415098816134867,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.3009,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.329139155438083,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2868,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.3450794678756165,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.3087,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.32579332343712797,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2844,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.3473612327442836,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.304,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.3185100889641007,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.2851,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.3316578625076207,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2975,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.3228184357302236,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2952,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.3333075511124877,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2951,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.33041496723142555,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.2945,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.3376937486523229,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2803,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.33108059086792707,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2942,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.3464932129450284,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.301,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.33586047964299603,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.2944,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.342966232277813,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2917,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.323560529937408,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2927,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.3648047566750104,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2986,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.3424146104204526,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.3535506081889873,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.3088,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.34982043157584075,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2859,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.43257163668406834,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.3035,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.33458539897939266,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2831,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.42393772378693106,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.3094,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.36173184653176915,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.3071,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.33128092629927486,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.297,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.4019259448845247,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.3051,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.3346448986970725,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2786,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.32810745384578804,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2898,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.366137149019249,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.3266,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.33775503505997007,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2842,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.3488892611569788,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.303,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.35502241893558634,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2966,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.34881892331877956,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2947,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.3382360608647854,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2915,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.3682993912999933,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.3064,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.30856185313945783,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2713,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.3194450716584849,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2785,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.36771521830272075,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2877,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.35455481347765666,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.3069,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.3194799314992769,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.3047,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.32439460212504595,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2737,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.3494809424210844,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.3024,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.3305559125700532,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.282,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.9795864861719776,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.3057,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.34075314368153226,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.3004,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.33899122282774863,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.3068,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.3485267866376398,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2739,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.36430483183662493,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2953,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.33664572367837997,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.3058,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 2.150189924454877,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.304,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.39367071430648026,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2851,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.3688669299653049,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2926,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.3361097567632177,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2994,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.45738942139265326,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.3285,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.37394472161047754,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.3057,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.36135960066373485,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2892,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.3918110752701107,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.2843,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.36724622659333983,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2976,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.32912529852996086,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.3056,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.3365396813371767,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2894,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.3656158227264251,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.2884,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.3106967043507718,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.309,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.3252107394901483,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2906,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.3396289271381184,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.287,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.34507797174650984,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2981,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.32642513436646875,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.303,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.3187957079015488,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.2822,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.34236993527524723,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2925,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.40288730566778536,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.3262,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.3063676358178623,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2799,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.32337823732708054,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2848,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.3242772946294327,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2908,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.32926213243732066,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.3001,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.3482199739930344,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.3016,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.33442537302491615,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.3073,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.400325691206051,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2773,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.33397086719556884,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2995,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.3172376596961087,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2937,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.34126266727645327,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.2769,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.3453787533801933,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.307,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.31460197248317395,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2804,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.3461409606873946,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.3235,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.3100185944932292,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2753,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.35210259043548325,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2809,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.3117958263580086,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.289,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.31788755001799984,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.3059,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.34310646375100007,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.289,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.33783586031044777,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2962,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.31503418957588636,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2963,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.3434341039720113,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2762,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.34488872968522466,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2989,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.3262570146688901,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2839,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.3316824260767909,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.2933,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.3101947395717167,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.2862,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.324925343021131,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.2974,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.30704285752180527,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2791,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.3451055286512017,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.3033,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.323464686437729,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2914,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.33255746445004947,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.3078,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.3357629000914737,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2981,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.3243694461463503,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.3039,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.2972672992694924,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2957,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.31935430161960027,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.3017,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.3057067338692804,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.3053,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.3119507881113558,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2993,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.28941415757130545,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.2797,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.3481540287412393,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.3125,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.3007448777869155,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2908,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.33474157612449706,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.3064,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.29804596977733167,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2776,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.32528536499563154,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.3043,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.32811737173134514,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.301,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.3014028351206588,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2635,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.32204181905708845,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.304,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.3142576758104505,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2754,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.34204544484052807,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.3059,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.31483440844875255,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.2983,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.3156126946436185,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2919,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.3181749228217184,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2881,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.3088307353247372,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2776,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.32452889657830764,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.3029,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.3327491806916009,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2904,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.3209208260603253,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.3018,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.3266827380138409,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.31,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.35379923207998065,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.3073,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.3198711133622165,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2877,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.307621028356193,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.294,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.3258163680845935,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2887,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.350424826080834,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.3072,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.3097439999344894,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2925,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.3169720601766559,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2903,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.31438997740720825,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.296,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.32628901405304217,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.3007,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.3094981809088742,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2918,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.3237073530169104,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.3015,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.3210410347341958,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2767,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.322793033020717,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2995,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.3216419247564139,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.292,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.32619401602264686,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.3031,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.3371006028464025,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.301,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.3096953733669968,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2988,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.3069971478935675,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2716,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.32308561587665147,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2994,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.3074119016364817,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.3066,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.32712364613182493,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.297,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3032052598821819,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.284,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.3082974857591568,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.2879,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.2962748992378511,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.2872,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.312202497600207,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2993,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.299312382995159,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2911,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.312912753603818,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.3034,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.3244856814425031,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.276,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.311280601085759,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2859,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.3584157624401271,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.3253,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.29725423743889895,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.2816,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.3012252279673916,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.295,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.3116338794814847,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2967,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.307222860162489,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2901,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.28863447382974505,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2692,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.3045134527089955,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.3041,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.3213729617884046,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2858,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.3276949056122546,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2928,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.29562131932874286,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2921,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.29955322449699556,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.295,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.31785985948301104,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.3041,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.3199065246260761,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.3088,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.3096213851681966,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.3103,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.30617012671857846,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.288,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.29183709487983733,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2808,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.2958408771818692,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.2969,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.31518950925051953,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.314,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.306641755469727,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.3061,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.2973802613056903,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2756,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.30343386753700324,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.3076,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.2830514204172296,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2779,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.2966306334674176,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.3014,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.30199870667972356,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2875,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.2991100639549581,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2837,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.29897927466729274,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2956,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.29612077716492474,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2942,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.29523083998114263,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2762,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.2990430548958996,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.3,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.2963976600755227,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.2968,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.3025727172497094,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.3046,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.3323707561509058,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.336,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.36256488256424063,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.2641,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.5175331289751448,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.2209,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.3990393440474967,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.206,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 7.609669216764191,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.2206,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.3339612531526637,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.2082,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.4009263561496693,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.2236,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.4287899779556298,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.2094,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.4488798687457987,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.2211,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.45408362712859923,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.2272,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.3648637598857922,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.2266,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.339155237936767,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.2069,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.37318420808737385,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.2238,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 1.1576816393064382,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.2237,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.3761138085341825,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.223,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.3313970642962272,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.2055,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.3259673479681569,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.2215,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.3308753046580029,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.2172,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.3327887901255327,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.2131,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.3342797352251244,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.235,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.31474677137431734,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.2096,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.34019396201106644,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.232,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.32480243369888917,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.2251,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.3030441097280852,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.2189,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.2999664800338915,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.2107,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.3200568154415722,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.2324,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.8989365551492398,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.2168,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.31256836138447086,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.2286,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.303354305225255,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.2194,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.2951702951559406,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.2136,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.29294694335317173,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.2059,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.30630960099124155,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.2259,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.29101037651796474,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.2058,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.31712102517237495,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.2295,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.29745084341040456,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.2102,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.3044571011691903,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.2343,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.30197315217365955,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.2177,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.28884885966167234,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.2046,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.3034431888460805,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.2269,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.28004889073702727,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.2102,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.27930597070433655,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.2077,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.2851900844320746,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.2199,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.28535904527862804,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.2165,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.29152453517779475,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.2196,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.3013379696994648,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.219,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.2835767925611197,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.2041,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.29356766557776404,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.2333,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.2863702661850028,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.2278,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.27938095483947295,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.2097,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.2793663837149951,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.2117,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.2990831393489933,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.2174,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.28421062758361426,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.2284,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.27818764593131684,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.2045,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.3757063796310124,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.2111,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.29494641731799615,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.2183,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.2832437556538324,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.2069,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.2671607635809148,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1962,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.3074570884660119,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.2364,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.2871919423023196,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.2095,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.29790468982658036,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.2203,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.2857107045043343,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.2117,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.28313173341349473,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.2093,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.28183062783805496,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.2229,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.2726359628205287,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.2175,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.29248392734577494,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.2343,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.29939090894682924,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.231,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.26706844786648565,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1911,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.2779433796356154,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.2211,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.27379462216188094,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.2097,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.28753072190542184,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.242,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.26784946731480375,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.2143,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.2889122237637015,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.2174,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.278416153785049,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.2089,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.27765531371232616,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.2191,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.28712158999640947,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.2142,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.29502217683256066,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.2199,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.2835217299462184,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.2191,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.29798129909533094,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.2298,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.2617655102681857,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1925,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.2754960899822647,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.2049,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.2855819123731452,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.2304,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.27622792927379597,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.2153,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.29729389589587696,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.2303,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.286611423367797,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.216,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.28424772939225534,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.2053,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.2932597285739013,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.2282,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.286134402333139,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.2199,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.2709934197911048,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.2154,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.2747192434841256,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.21,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.5913132993629683,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.2047,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.29775926124676205,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.232,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.26417017769230405,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1852,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.29413434133388716,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.2433,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.280902522939481,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.2007,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.27885205693426995,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.2077,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.280046825476282,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.223,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.27031983554886113,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.2124,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.28487681384985375,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.2202,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.28619155197194013,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.2267,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.2792678371353543,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.2094,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.28387085419802005,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.2237,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.27928115693849115,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.2175,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.27640780547670957,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.2171,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.2768229036844447,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.2143,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.286396581392847,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.2208,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.2716428752930402,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.2,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.29299745356965123,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.2469,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.28669348395133104,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.2217,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.28438242151482507,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.2274,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.2866671968382761,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.2092,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.29240403272677423,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.2233,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.277407397591182,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.2034,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.29866781159069106,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.2342,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.25810214729121833,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1857,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.28482454790727474,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.2231,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.27836982693005663,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.2265,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.27822713703398777,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.2117,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.29796656196274984,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.225,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.2778732012594435,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.2015,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.274958360559583,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.2229,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.2773414613340353,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.2217,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.27666641109003726,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.2189,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.27188145912474626,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.2067,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.28031590731992995,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.2185,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.2673119964535637,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1984,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.27357501616594315,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.2132,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.27807687797425373,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.2113,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.2951414746745389,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.2359,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.2639948804396606,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1822,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.2980011495239741,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.2434,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.26895763813783474,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.2151,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.2885328590730185,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.2284,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.28425320065874005,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.2121,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.2717259683987803,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.2069,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.27948727368159515,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.2278,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.27876626483360084,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.2096,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.28821549978581734,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.2161,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.2864311630523999,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.2251,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.28691597969840194,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.2138,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.2797645593586703,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.2339,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.2866303975649323,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.2095,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.2795890509262594,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.2204,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.26947990934721255,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.2118,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.2739197607069856,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.2296,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.2716556259272006,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.2154,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.28492319088544765,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.2327,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.2756295457122563,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.2155,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.2751460818379593,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.2143,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.2870245627283275,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.2303,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.27184839755408546,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.2215,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.2698217341944367,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.2112,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.28465367729057267,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.2269,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.2828691250285533,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.2125,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.2765255955853625,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.2258,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.26950944879703925,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.2128,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.2733749109566616,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.2201,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.27233859271634564,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.2037,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.2804502821686114,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.2226,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.27976755837120243,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.222,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.2896746412903016,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.2297,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.2850900013780694,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.2189,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.27961978839073565,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.2154,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.2612987704982986,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.2171,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.2943257378736707,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.2346,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.2662874611315876,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.2074,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.2789863303864319,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.2078,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.2928925140219756,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.2234,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.26780435885864406,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1972,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.2934463407139921,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.2309,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.2613946149221887,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1903,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.28427775317071924,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.2202,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.28381290716521357,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.229,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.27176530817180367,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.2159,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.27392278602571984,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.2103,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.27647849540207325,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.2176,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.2751083614233168,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.217,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.27207411732684295,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.2058,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.300328883754641,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.2423,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.26874376676978645,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.2079,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.2728566374540701,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.212,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.28556192272707687,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.2325,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.2749591626028843,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.2232,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.27667138864348345,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.2232,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.27605140918071114,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.2188,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.273424565062339,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.2116,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.2831846716781306,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.2244,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.2723603065157497,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.2159,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.26469080707198406,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.2218,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.2924609207218796,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.226,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.2753465059885192,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1999,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.2860297762891096,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.2369,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.2677040932181543,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1885,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.2865266732345786,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.2306,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.26961545863445274,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.2084,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.2715484325372613,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.2233,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.28329964425221205,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.2279,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.26630442960000733,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.2071,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.2799367364963802,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.2287,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.27862315316372305,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.2188,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.27154839913233103,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.2001,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.2815778493295658,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.2207,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.2775636741600175,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.2278,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.28757672858347233,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.2119,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.27113724074367224,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1976,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.27196761182385465,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.206,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.29001266036505935,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.2383,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.2695830290619848,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.2217,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.2776421879954989,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.2173,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.27849393956184987,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.2271,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.26912841491899425,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.2158,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.2731850941859809,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.2208,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.2859965083742554,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.2233,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.2812322338590815,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.2065,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.2786599681241473,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.2126,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.27895190331389497,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.2053,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.2923448009709105,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.226,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.2702354602389292,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.216,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.27428179725298485,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.2138,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.2854066635811131,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.226,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.265311009071038,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1933,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.2770315745321496,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.2157,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.28705666760472903,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.2321,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.26294199824914183,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.2005,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.281296465498341,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.2139,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.28007753078498465,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.2051,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.27895205490461933,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.2079,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.2785993796359204,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.2153,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.2882266368061207,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.2333,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.26716619851803197,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.2113,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.26199695635446196,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.2179,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.27751433260395375,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.2156,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.2606035489121528,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1888,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.28323035957807935,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.2337,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.26530273743782823,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.2095,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.2787195374673456,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.2228,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.27941844082619316,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.2216,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.25416337501221153,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.2028,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.2731942185636907,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.2192,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.2853345656748549,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.2281,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.2694185948896589,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.2137,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.26401633818405884,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.202,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.2739349448405763,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.211,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.28201833514255037,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.2148,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.26705861855301255,
+      "learning_rate": 0.0,
+      "loss": 0.2213,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 2.940653575933526e+18,
+      "train_loss": 0.41976423211456315,
+      "train_runtime": 126354.7473,
+      "train_samples_per_second": 1.25,
+      "train_steps_per_second": 0.01
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.940653575933526e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}