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": 6.8687622078476,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.1406,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.875698028651337,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.1623,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.744757178234629,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.1439,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.586992899871902,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.1395,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.181343165561811,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.1321,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 6.154586625137341,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.1279,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.639672764511178,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.0872,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.1232937090937645,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.0475,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.5719641825959534,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.0439,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.4778523730673006,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.0482,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.388977593022676,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.013,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.525427744103924,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.0001,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 3.9718718997933777,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.0103,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.6527572707919242,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.0065,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.1236143728168204,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.988,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.079331415780548,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.9473,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.6189320038918593,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.9576,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.2382505925607177,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.9257,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.7537068383916377,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.9233,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.3467490990952813,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.8981,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.2971062285989259,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.9004,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.3407601903233208,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.8939,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.1198201896648998,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8705,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.1529241269467851,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.881,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.079970681393599,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.8602,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.0143314106095895,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.8533,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.8928080222325984,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.8619,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.8134489089194391,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8498,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.002168152609542,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8434,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.792897771982383,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.8363,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.771630691005481,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.838,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.8704374540269969,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.8356,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.7259028449581668,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.8289,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8029571914720867,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.8096,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.6865152945449787,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.8279,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.7052067267187395,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.8027,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.8096197816614421,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.8107,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7217819008629692,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.8079,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.8172503219808794,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.822,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7044463673425091,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.7949,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.6737070889971355,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7986,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.7427283466081188,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.7894,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.6823145479921202,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.7967,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.6751214328417854,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.7969,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7767225435851748,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.7989,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6481120692832806,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.7985,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.6831706026669436,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.7999,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.7339237690764029,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.7741,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.6816155199571687,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.7887,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.6455309172499111,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7871,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.5844805175685869,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7736,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7996470394555786,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7929,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.7590309092383378,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.7754,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6961246538157719,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7823,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8382661494964694,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.7825,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.792775716673812,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.7919,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.7369287092683247,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.7559,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8988300703716864,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.7859,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7756585155941518,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7732,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.8284666567606935,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.7835,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.868364096464385,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.7855,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.9406286693059769,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7705,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 1.009867265463389,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.7698,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.8095218250187868,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.7832,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.8590033334682721,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.7703,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.9069563241443295,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.7867,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.7352118028296086,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.755,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.027602455838582,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.7672,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.7980599800273827,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.7699,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.9548925798849495,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.7572,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.092817502009092,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7643,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.751329518316308,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.7663,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.130402499095903,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.7497,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.8388161036642898,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7663,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.832694370762362,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.7639,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.8712489430681052,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.7431,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.046284958775405,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7638,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.1237497648735588,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.7565,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0771965665178302,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.7352,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.899598603576447,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7483,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.1444144871201303,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.7735,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.108025458882081,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7655,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9322674979550254,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7764,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.1694015989473,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7498,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.8078324532982756,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.7528,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.8548120073985004,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7599,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.1010030423717296,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7678,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.1948374136940458,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.7776,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.95207033210115,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.7525,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.7717892389994407,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.7509,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.7955411735673154,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.758,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.870071076154811,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.7647,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.04755896040898,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.762,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.790524095516262,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7599,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.3552756431906534,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.7654,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.812844204680522,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.7475,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.6750569732576188,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7518,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.8402509835263294,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.761,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.6692752601521526,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.7535,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.7809816894430215,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7504,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.7214821107179872,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.7584,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.681965832303908,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.7634,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.5025617797766138,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7691,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.3737602352987615,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7371,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.459079811245921,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7438,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.2380307779456676,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7358,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.2462572850443387,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.7426,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.0119938686292214,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.755,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.2811639616689583,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.7311,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.2258644874376223,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.7553,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.2870800068598016,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7495,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.2727277482658668,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7558,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.036135594889175,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7338,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.238883789570068,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.7504,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.8196953617346853,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.7446,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.617399816323621,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.754,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.4798542410444893,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.7598,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.0777393132657296,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.7528,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.5095814592428463,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7303,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.4010448268782814,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.7427,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.2282480713833468,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7561,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.078515008081659,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.7514,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 2.022878880583316,
+      "learning_rate": 4e-05,
+      "loss": 0.749,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.0012181323523495,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7411,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 2.6516729150691716,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7518,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 2.213644231045089,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7574,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.8460234142317147,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7531,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.5832537498916155,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7528,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 2.04481500652646,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7516,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.226964248364912,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.7459,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.919311668378772,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.7514,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.4980501650340143,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7347,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.4622976928979754,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.7475,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.3244805619947697,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7618,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.176018865975149,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7255,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.2521687603943557,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.7489,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.442416913400624,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.7451,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.2362031536788298,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.7435,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.0655857502199981,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.7416,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.2504310644073175,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7336,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.5723445041991244,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.7512,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.9678923014937586,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7419,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.854871977457656,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.7608,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.313837604228333,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.7516,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.614675621777834,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7274,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 1.471530649062178,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.7487,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.224514732311356,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7536,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.50000750172017,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7341,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.0300006892212459,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.7399,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.7227425265259495,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.7397,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.2311826764965743,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7361,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 2.0536793001570217,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.7434,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.9152207726373829,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.7538,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 1.2483274686295713,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.7438,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.8008333730019301,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.7267,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 1.2193907180234178,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.7377,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 1.54933499532176,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.7335,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4467720842249212,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.7335,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.3347393762182205,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7444,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.575140996705505,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.7411,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.2771278925323866,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.7389,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 1.390127022604749,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.7399,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.2108558751376728,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.72,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 1.0574958066685687,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.7341,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.3559616727405053,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.7289,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.9693194205918477,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.732,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.498848962518501,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.7268,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.9455083632299968,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.7255,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 1.597605069100217,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.7338,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.1222531430015008,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.7329,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.5059955601855786,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.7444,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.2376069689077536,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.7319,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 1.1351014427213462,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7285,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.3102904235185826,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.7331,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.0841819118497993,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.7422,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 1.3129116433929453,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7258,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 1.28409245252643,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7481,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.9562758819259704,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.7123,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 1.2873243665548901,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.7336,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 1.0070272828336628,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.7149,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 1.5541938046504016,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.7277,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 1.2666795003980893,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.7392,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 1.6075068000101211,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.7269,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 1.2185579136731717,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.7341,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 1.4893399236102172,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.7344,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 1.0387264300918446,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.7337,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 1.5133306835937144,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.7571,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 1.2719127345098633,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.727,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 1.3163022913706017,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.7428,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 1.1116405354680179,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.7282,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 1.3433196674306547,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.7223,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.9527193049888313,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.7158,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 1.4233306242511778,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7428,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 1.0672879450616355,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.7388,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 1.2263466435650356,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.735,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 1.0840930337279242,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.7214,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 1.0088030892840922,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7318,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 1.430487430995343,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.7277,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 1.0690155736265616,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.7321,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 1.0555025859376712,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7289,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 1.1418194223404536,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7271,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 1.0917009723895896,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.7292,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 1.2330990862139544,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.7213,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.894842293193897,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7163,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.8555975873913475,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7342,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.8890595824483587,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.7238,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 1.0467100653447203,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.7257,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 1.0772270182704036,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.7103,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 1.186787153930016,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.7398,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.8077336579358031,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7389,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.9866365616977814,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.706,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 1.3953116192821002,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.7293,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.899457562462761,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.7128,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 1.3906509511005902,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.7389,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 1.062248683140957,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.7358,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 1.1257426673488213,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.7315,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 1.0026268129789973,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.7358,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 1.0230427497503225,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.7276,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.8451867506758521,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.7146,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.9377206393470606,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.7348,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.9295489698360601,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.7231,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.7982674774648139,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.7201,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 1.0072188934403903,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7375,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.7987323255993874,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.7112,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.9590951799738628,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.7246,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.7204197845096844,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.7276,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.7782503306733489,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.7244,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.8555612017227723,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.7363,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.9576562911166931,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.7397,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.9154790770134565,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.71,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 1.5423811813657633,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.7291,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.712301356401536,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.7221,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 1.6579585456585493,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7266,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 1.012427341942415,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.7199,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 1.451881964087576,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.7306,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 1.1070197969716602,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.7314,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6797131171417938,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.7318,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 1.3660832498310227,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.7229,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 1.5329902385942307,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.733,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 1.379957040486996,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.7331,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 1.5167962968185413,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.7203,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 1.2771792771140469,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.7372,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 1.1917644103015144,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.7269,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 1.5643603749411537,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.7273,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 1.1729786650448166,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.7332,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 1.7681998843361681,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.7283,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 1.6209597260794875,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.8082,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 2.043426565635896,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.6203,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 1.5012584493872922,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.6474,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 2.04142279543475,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.6329,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 1.6575232959636605,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.6449,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 1.8096797468441863,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.6345,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 1.2872686982482504,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.608,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 1.6884917017279941,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.6498,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 1.2151090911954545,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.6332,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 1.8147472004723388,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.6128,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 1.259760728288371,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.6298,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 1.7803423464554478,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.6179,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 1.3033780137846354,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.6149,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 1.9002108032558096,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.6317,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 1.457184148032288,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.6046,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 1.8320014169263528,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.6427,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 1.7729912767196219,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.6282,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 1.2920403286238735,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5963,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 1.5364513834915563,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.654,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 1.3056781097402603,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.6116,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 1.0867346627906374,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.6126,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 1.2674219563121858,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.6194,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.862128985376276,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.6082,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 1.5276378181387766,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.6217,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.8003914934833943,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5964,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 1.6225178548222514,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.6126,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.9999606744240442,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.6146,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 2.0188110298463546,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.6499,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 1.6707857117528517,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5952,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 1.4745665137765034,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.6191,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 1.4107778772511173,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.6294,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 1.3254385113520077,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.6526,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 1.2346183557395363,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.6151,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 1.2367453690909793,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.6101,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 1.1973867714951612,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.6278,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 1.4119654468043497,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.6257,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 1.239617393779475,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.6198,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 1.193512060732952,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.6228,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.9691471013970925,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5993,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 1.2853904775199108,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.6237,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.9424976618389264,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.621,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 1.2170975948396243,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.6168,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 1.0210641504497788,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.6298,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 1.253830912757928,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.6098,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 1.0674676979385374,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.6049,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 1.1390398826868564,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.6286,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 1.028222195871677,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.6167,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 1.1409419723998315,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.6476,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.9086516340330834,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.6309,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.8601614182575011,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5866,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.9044974759789712,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.644,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.8643583112423455,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.6081,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.8940443758034448,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.6577,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.809643978105034,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.6244,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 1.0201559350402245,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.6008,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.7686539244177931,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.6398,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.7278042782621967,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.6041,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.7619890013372236,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.6376,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.7510426824947001,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.632,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.8693309265061525,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.6005,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.7387305035876407,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.6244,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.9033377413961056,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.6274,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.8492864412379462,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.6122,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.6911025411509512,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.6027,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.9607174423448418,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.6215,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.8809852769711956,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.6426,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.8623512285975552,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5958,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.6658265799284057,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.635,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.9404475200631305,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.6258,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9676807244010274,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.6602,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.7326064900628065,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.6071,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.7307695859787758,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.6147,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.6536689620464249,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.638,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.8498579256858642,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.6362,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.7195065641122985,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.6123,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.6600831577405002,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.6286,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.6870307379427152,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.6376,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.7278513524877642,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.621,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.8311844133392398,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.6082,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.7093266277550659,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.6297,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.7308471726606613,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.6307,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.7042324736580363,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.6378,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.678614963838566,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.6169,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.7715902203829221,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.6288,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.653957636516294,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.6431,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.6861955774229557,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.6217,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.7205482542778261,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.6348,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.6169645878239386,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.6524,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.6525702401803524,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.6167,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.7063919066709906,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.621,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.7560554481495189,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.6369,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.7651134419103339,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.6227,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.7500504162813403,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.607,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.6957291478334912,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.6291,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.7042771024797256,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.6431,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.6046438480143771,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.6215,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.7851366733274082,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.6607,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.7988450162519513,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.6149,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.6471980824591153,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.615,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.782839634970135,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.6736,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.6773891714745932,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5845,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 1.0707579795515083,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.6425,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.652997198405234,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.652,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.6512473094241407,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.6143,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.7074918575935768,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.6228,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.6743265189892563,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5964,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.5820986330954813,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.6325,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.6937971485707538,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5843,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.680397022489482,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.6487,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.7987526569910179,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.6098,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.5501931401426525,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.6237,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.7929829701599729,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.6176,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.630717638887243,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.6439,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.8076150573959905,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.6322,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.7013764336178118,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5995,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.5665662316090163,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.6218,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.6924084211059488,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.6126,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.6182342437571515,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.628,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.7389572478501386,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.6182,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.7124721785325538,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.6298,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.6168592040313798,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.6184,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.717831845284264,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5965,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.6762965820043155,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.6335,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.7461330155887776,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.6218,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.6967631202054173,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.63,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.6710609709051232,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.6246,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.6959302182770472,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.6371,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6115034551875388,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.6095,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.6321899958738292,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.6283,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.5631775902855923,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.6223,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.741256424071349,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.6376,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.6718819289324409,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.6295,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.7308609201722851,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.635,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.6557114539519824,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.6144,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.6909143705882782,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.6258,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.7258564581648054,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.6439,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.7060087462628789,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.6077,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.7546971170640903,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.6637,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.6230995479137006,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.6171,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.7868187324149968,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.606,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.6201783139642016,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.6487,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.6678755486938481,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5998,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.7808508564028946,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.6457,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.7447870367556021,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.6167,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.8093372993844724,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.6267,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.9009399211295329,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.6219,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.5904128160464305,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.6251,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.7984293567282584,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.6138,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7921229576736425,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.6203,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.7508184396684892,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.6377,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.6201739833017117,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.6305,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.6193697163789285,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.6154,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6803894330508151,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.6326,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.5909198250747347,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.6332,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.6635533492674651,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5976,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.6950663997050877,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.6614,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.6286416306699675,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.6127,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.6234092082097806,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.6116,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.7130402449434409,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.6479,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.6958631370993248,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.6326,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.7953179383303437,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.6178,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.615237867110919,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5893,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.7132310244975348,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.6458,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.7173702189781721,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.6143,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.7683200789708653,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.6142,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.7501662256563523,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5964,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.7361543347996655,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.615,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.647486783794424,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.6537,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.6445813880387442,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.6521,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.6311459713235759,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.6216,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.67706850242624,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.6201,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.6581836819203051,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.636,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.6812530561418914,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.6277,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.6463821952512617,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5933,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.8198780900963063,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.6534,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.5698330594192814,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.6136,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.8686546231876359,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.6135,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.6224532107379159,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.6466,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.7858244971681287,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.6282,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.6916823904725923,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5999,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.6867989449156129,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.6277,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.6593658112891684,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.6344,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.6234768461135287,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.6065,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.6036917576820751,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.6174,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.7343574964992997,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.6247,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.684836026389699,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.6392,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.6296460391826447,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.6058,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.5740585849909189,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.636,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.7258461626293405,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.6085,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.5315482802626037,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.6289,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.54700586792142,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.6117,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.6092909397639192,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.6255,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.5377192482937316,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.6259,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.5922816950649065,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.6306,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.5521168215790228,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.606,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.6573246652030229,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.6311,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.5809603213508177,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.6301,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.6053654248348213,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.6145,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.6079875176673685,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.6327,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.605284078502411,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.6237,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.5948552017302666,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.6078,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.6077359231995317,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.6302,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.6750013969471507,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.6157,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.5514756126680876,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5973,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.6748320500940588,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.616,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.5781442198718152,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.6577,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.7364889344652393,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5938,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.666531026582619,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.64,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.7270112078288132,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.6224,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.7820182402205331,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.642,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.6477271098241825,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.615,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.6563013997086226,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.6099,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.7437713821922106,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.6152,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.569908114491798,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.6161,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.7307189636233745,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.6272,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.6165615824039103,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.6149,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.5602370307860784,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.6013,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.646674315816007,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.6336,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.6100407795006492,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5921,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.6483380349761854,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.6561,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.5722701640167085,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.6045,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.679542295248065,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.6485,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.5475470653707951,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5903,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6356291065335491,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.644,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.7235555563061987,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.6155,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.7682164247250379,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.6324,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.5376366841224186,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.6183,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7129184717602354,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.6285,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.6065686029130755,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.6194,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.5770172114283263,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.6002,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.61549805560564,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.641,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.5728078729964974,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.6088,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.6632205227073912,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5936,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.7273643802854002,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.6719,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5526547855318737,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5841,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.6117475353029149,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.6535,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.5991954259632036,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.6196,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.647686236293547,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.6162,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.6610003155026665,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.6079,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.6135581757425104,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.6411,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.7123247177989603,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.6282,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.6530988400827179,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.6028,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.6261973924516571,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.6183,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.7093112909347986,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.6461,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.6267112438845442,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.6385,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.5384378010016241,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5833,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.6306968760045565,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.6509,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.7268470971195287,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.6663,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.7440273060656031,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4631,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 1.6773948167564432,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.5076,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.9159152391625036,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.476,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.8401297825028093,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.4966,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.7642441603669344,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.45,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.6956669370480666,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.4861,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.8005031875835144,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.458,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.9880377126055971,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.4936,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.7594264530782454,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4387,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.8983597923348147,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.4955,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.9390772123668318,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.4571,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.6724639630568005,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.473,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.6010068342380249,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.4753,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.6619414584263509,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.4752,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.6434258712597504,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.4752,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.6902481642815096,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.4693,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.5682100796765608,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.4635,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.7668187342046021,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.4767,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.711782984918525,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.4705,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.6255662624682575,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.4781,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.517372351846347,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4388,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.6515474059543772,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.4886,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.5318315306414825,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.472,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.6154420240195683,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.4789,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.5647559444231344,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.4562,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.5126772805483671,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4672,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.5904335249673386,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.4834,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.5749690059307748,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.4621,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.5254504093482992,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.4458,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.7010286759325002,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.4981,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5561609685199405,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.4804,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.5507023021960585,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.4639,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.5836293723137802,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.4924,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.5217774182911503,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.4561,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.6716591441565423,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.4771,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.5936989043723072,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.4813,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.6480677699087897,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.4825,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.5455637650840389,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4912,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.5889910324230855,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4252,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.5263910842254562,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.4974,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.5452521420795825,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.4394,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.5826727542853928,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.4906,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.5900107605872923,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.4713,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.5681551790448773,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.4639,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.7289281962597357,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4716,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.47406331428088383,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.4787,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.704499531020544,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.4722,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.6027878468357962,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.492,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.6542391625165151,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.4574,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.5752570498934968,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.46,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.6156813419036404,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.4835,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.6824348966254278,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.5023,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.5608928817548093,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4764,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.589420639025176,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.4511,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.5925115931555792,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.4763,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.5964606362171314,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.4919,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.6097542433305945,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.4662,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.5579846540147992,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.4892,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5424783627329681,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.4777,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.5446047727676292,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.4815,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.4959716160781475,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.4763,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.6192649003134177,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.4605,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.47365687547913954,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.4837,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.528293591335959,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.4882,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.4479134798364416,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.4405,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.5942217501590081,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.4742,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5555927386206367,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.489,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.48422219221282764,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.4838,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5275447434412965,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4643,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.47236570007446166,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.48,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5090072008391846,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.5042,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.49946507382149874,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.5014,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.44209557746846045,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.4384,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.5115415790023857,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.4997,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.4516089040521762,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.4587,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.4922794561427412,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.492,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.4935426359517961,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.4522,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.4820471312890303,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4751,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.6202768830252986,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.4892,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.5098883697877897,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.4418,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5429728880044657,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.5294,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.5872043005327583,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.4497,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.5450186713275919,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.4908,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.656187840793124,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.4594,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.5494657713253233,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.4962,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.6370628554616523,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.457,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.5259205884529244,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.501,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.6161259710464513,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4805,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.48907047840167006,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.4386,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.698255296364926,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.4884,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.4780741216605289,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.4562,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.6853825784481029,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.5147,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.5954310701242117,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4663,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.4989918884024488,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.479,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.5279794294188489,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.4732,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5329201409906026,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.5016,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.5285712000758012,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.4725,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.52370640626569,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.4733,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.6272680548982275,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4943,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.49214836941020795,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.4734,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.5299224863367986,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4888,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.5013625101037925,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.4721,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.5331313760854723,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.4582,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.5003464211527908,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4884,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.5051334852055483,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4646,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.49820386533086874,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.492,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.48766981465004533,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4656,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.4657809731715379,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.4653,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.5736825595622292,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.4693,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.5040737751738267,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.4844,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.5937763618674027,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4954,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.5111057523874123,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.4572,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5591567727485077,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.4826,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.49146303299639155,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.4923,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.5503763722657067,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4574,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.4773921937707085,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4587,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.5212309111000907,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4882,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.5269933109742779,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.4683,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.5820785052615237,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.4737,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.483816626282755,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.4736,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.5342742782890187,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4865,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.48248626438046205,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.4716,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.6100695314318116,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.4976,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5323640672139601,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.466,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.6498197595014858,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.4815,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.5215662525344946,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.4617,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.637798920746485,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.4945,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5368508474591979,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.4673,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.6751967124860406,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4955,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.6625732407808006,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.4754,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.5301141404037569,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4979,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.6025515986160633,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.4612,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.5357379361935614,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.4946,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.6625678039264505,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4619,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.5570389676918481,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.4727,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.6154088510027452,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.4761,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.5018397983558552,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.4844,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.5153483520037482,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.46,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.4953450376989121,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.5116,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.516190758013711,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4566,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.43301839649627477,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.4686,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.5355866595876333,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4898,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.4905096613371513,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4763,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.4870154530990457,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.4672,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.5017565487671001,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4998,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.4531311762795351,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.4639,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.4936862653411411,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.4557,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.48307473187702077,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.5014,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.5141445284587141,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.4645,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.4660369901608912,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.4932,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.513334919528763,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4866,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.4768262944546153,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4764,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.5391228501457609,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.4819,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.5596266767735707,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.4762,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.4762803344054926,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.4698,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.5099053543833149,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.477,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.48437848497973246,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.4757,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.44031422482287463,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.4768,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.46946940272398,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.4613,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.44273986515158026,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.4689,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.4494682254556167,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.4664,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.4635917639585946,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4944,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.43731720923929024,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.4782,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.4627507995016706,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.4704,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.5353171875598489,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.4907,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.4631787950160399,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4584,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5874165575345667,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.4699,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.4593785218418008,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.4645,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.6364216787074549,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.4906,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.47083382997416534,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.479,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.6858322232386626,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4653,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.4741506673330165,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.5024,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.542039148231069,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.4869,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.45380526905446206,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.4537,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.6427579446713222,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.5037,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.41838967037539626,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.4621,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.6029227258897398,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4693,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.44274848923997706,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.4809,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.5868490982416191,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.4654,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.40335519066030107,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.4506,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.5429893574825975,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.4931,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.44827398427775966,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.4846,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.44652659708217074,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.4816,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.5042060940271615,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.4779,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.41968091119469964,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.4632,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.47951908557400696,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4937,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.4069872348431596,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.486,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.4094064006372344,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.449,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.45544159692957037,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4966,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.43935172750013013,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.4826,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.40825699627028794,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.4563,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.42389201923155495,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.4845,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.4511059041787327,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.4641,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.41787808689758493,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.4772,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.4397380490324069,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4771,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.4710271576539968,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.464,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.4541746618031532,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.4901,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.45621776813649706,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.4581,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.4560509311845956,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.4922,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.41805588768018376,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.444,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.5235314296502315,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4954,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.45713055135078706,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.5029,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.4437367811758572,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4645,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.46315194633754914,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.4785,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.44402872080652583,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.4961,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.4276019417625408,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.4725,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.4923003585261359,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.4895,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.43433763945117454,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.4364,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.5504147970400258,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.4783,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.520560584782079,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.4937,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.5259401566165804,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.4715,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.47980954787287794,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.4633,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.5170136324913399,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.4724,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.45945291150325757,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.486,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.4781036656712252,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.4666,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.4600365970623364,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.4872,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.46602322996458434,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.4644,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.415185565854889,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.4647,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.47250487603022545,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4806,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.44882276409958094,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.4935,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.42014322495875883,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.482,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.4149299986637883,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.4785,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.4342380607033508,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4717,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.4617900001425357,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.4882,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.4937440919342697,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.4787,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.4388029680081619,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.5043,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.4521335167539767,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.4462,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.5479636890466983,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.5051,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.4240412169860886,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.4677,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.5304467191483712,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.4494,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.5197821415927032,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.4779,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.47528352760046483,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.4816,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.4455068075257178,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.4683,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.46446586058554407,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.4905,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.454612835492385,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.443,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.4995114003599476,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.5008,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.4470441588836381,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.4606,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.6137409686853761,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.506,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.4486289392886061,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.4547,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.5465838306950165,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.4598,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.555399763140617,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.5089,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.43811823396194416,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4837,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.5431222844221656,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.4762,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.4539064908281744,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.4818,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.45990860075750273,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4736,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.4837120238100096,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.4767,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.6460457282880057,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.4796,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.6936736284033208,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.3435,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.5608030337658461,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.3507,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.9121462441509166,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.3227,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.7883977922982885,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.3558,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.5839173229847632,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.3154,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.7459031209168319,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.36,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.6177580630915233,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.3131,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.6051483347864791,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.3458,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.533683593867446,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.327,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.6326484664092497,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.32,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.5671972135313502,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.3345,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.48296893086880033,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.3328,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.5365586860968201,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.3225,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.49432158845001695,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.343,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.4466407893930846,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.3246,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.47934459390046474,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.3164,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.46459601966074776,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.3397,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.4983176261451967,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.327,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.4254817013254277,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.3146,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.4652265036956184,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.3625,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.4629019969127904,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.3271,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.4695986786634738,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.3426,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.4215804426814082,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.3123,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.46469018956382036,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.347,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.48698767371437823,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.347,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.3951006153804747,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.3274,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.472365722683787,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.3308,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.4084922839738175,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.3129,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.4234335737841835,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.3457,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.4119936327529754,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.295,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.474180719569328,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.3531,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.43452066375693726,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.3416,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.43490150032613145,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.327,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.4152479782038131,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.307,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.41969305368155624,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.3297,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.3921360341532933,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.317,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.427351686972493,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.3325,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.379473498255368,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.3085,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.385441835767695,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.3326,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.449750695268769,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.3506,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.40609178101273324,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.3115,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.43087115447050534,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.3298,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.4083940084173462,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.3387,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.4301878022856456,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.3541,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.3716617440614672,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.312,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.4038948661355984,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.324,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.43011750991168746,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.3557,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.3947428080898899,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.325,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.40331561066673854,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.3131,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.4355981442838235,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.3406,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.39690694735593984,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.3456,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.44114384408124513,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.3326,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.3929636695853254,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.3317,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.3874707546389014,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.3154,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.44774706761142025,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.3634,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.3964095613570416,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.3036,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.4143957840692558,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.3183,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.4311516648650033,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.3531,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.4590755361375514,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.3455,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.3749226827288375,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.313,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.4198685048238739,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.3361,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.46988281746891075,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.3441,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.40062760257708685,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.3163,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.4473995479872143,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.3489,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.39219430550355366,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.3273,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.4254993450730925,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.3363,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.4168616064491047,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.3448,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.39095809225084327,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2974,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.4330899165181446,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.3591,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.3851006792387339,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.3252,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.40182661451016943,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.3043,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.4596961508993044,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.3493,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.39900926534506126,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.3348,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.39983547464628527,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.3202,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.39227052562670073,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.346,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.39457129837959287,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.3386,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.38025273939533055,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.3101,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.38021841686932223,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.328,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.4059064570597599,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3374,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.391953446071023,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.3356,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.37395649221923943,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.3295,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.4008753930513971,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.3375,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.374352098842071,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.321,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.39226833761083674,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.3417,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.40424177087186747,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.3212,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.35917728173534313,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.3194,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.3820283824593075,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.3219,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.3921312494571161,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.3367,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.37426156993506143,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.3403,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.37610601679667993,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.3315,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.3891596829262661,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.327,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.37104898922896895,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.3337,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.4175102194506245,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.3372,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.4154721880390203,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.3417,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.3894848247009197,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.3189,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.37161461525681083,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.3158,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.39158771813494975,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.3548,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.3820202964816249,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.3557,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.3702014477576964,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.3121,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.3787723467320406,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.3153,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.40792137396546707,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.3344,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.3862295277139137,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.3597,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.3854632600317161,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.3441,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.3782009113318332,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.3204,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.3662692593277316,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.3177,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.4062399876810598,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.3284,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.3943670133970271,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.3426,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.37327725914146004,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.3188,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.42104832945049064,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.3264,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.40180073590558857,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.3514,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.38204070925388106,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.3196,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.43522602823269557,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.3473,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.37585183487569945,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.3002,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.39602749527299524,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.3565,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.3856937314194611,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.3323,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.3813278263351092,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.3208,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.40696724422103964,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.342,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.3898351922535973,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.3155,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.38630993713503786,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.3272,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.35889917022672413,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.3138,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.409368159639302,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.3341,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.3500953274913299,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.3069,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.3909212992249595,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.3482,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.35857573553227806,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.3103,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.38998677163912865,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.344,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.3929978827216047,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.3284,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.4043885922559963,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.3294,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.38062050053140234,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.3416,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.35887914673615484,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.3186,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.3657483437002058,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.3211,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.38755036057596165,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.3405,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.405943742697129,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.3443,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.3761176817242856,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.3151,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.36857234046392234,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.3423,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.4012006220750209,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.337,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.36548178308252605,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.3095,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.3798306397682827,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.335,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.4001516236819065,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.3644,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.37903340877724473,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.3323,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.35854896950839493,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.286,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.38503516963006956,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.3498,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.37822986960271,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.3279,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.3571883068674647,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.3097,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.3744167518997576,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.3306,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.38652800336415244,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.3464,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.3600785687881973,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.3242,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.35475382176650955,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.3068,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.4145287223301772,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.3632,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.36356013454093244,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.323,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.38615714658990974,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.3276,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.39932760742486734,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2966,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.38703904723900784,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.328,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.42592945159602924,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.3607,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.4013664361804948,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.3284,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.3754968834749157,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.301,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.4050872558163307,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.3345,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.37488088891082166,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.3334,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.4026289431481999,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.328,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.3916998001176197,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.3208,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.38007150707628884,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.336,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.3977605444795489,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.3343,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.3950854470485989,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.3219,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.41658567398009977,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.3406,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.37224344483244,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.3369,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.38452520443091903,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.3164,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.45296101558125407,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.353,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.3724294648715434,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.3057,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.4128667032903224,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.3343,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.3924050564158303,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.3227,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.39661429665828496,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.3293,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.36296592105951875,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.3416,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.36114907661548784,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.3039,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.3779331523792722,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.333,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.41386483945422364,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.3507,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.3587054838389761,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.309,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.380311035036182,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.3283,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.39476868609721055,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.3499,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.3841985204090124,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.3235,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.3747736683900608,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.3285,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.4078854531286399,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.3129,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.389490493586613,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.3256,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.4005570263830722,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.3208,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.38442260243175425,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.3415,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.3631270880668593,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.3324,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.36610133667881706,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.3262,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.35479230893133756,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.3216,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.3593107172381478,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.3405,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.36300143374934635,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.3159,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.37146408879863324,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.3457,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.36859684261029435,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.3188,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.3737952037428108,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.3285,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.3716450289023985,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.3359,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.3684065720168852,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.3354,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.35253800767099014,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.3271,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.37377536764628017,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.3261,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.33459134735694546,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2966,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.3698789861397702,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.3374,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.3932224895242007,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.318,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.3646634671809195,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.326,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.3772693855847504,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.3365,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.35248122596270676,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.3232,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.3614596839682345,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.3167,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.35607978645439126,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.3222,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.3528811573727096,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.3371,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.3393130116901208,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.3173,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.3612936506978004,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.3427,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.34166792425726106,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.3126,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.35636377919333456,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.3408,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.3747324954754841,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.3212,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.379459836829519,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.3456,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.34280491103661104,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.3217,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.3680208479456666,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.3383,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.3795703245668642,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.3322,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.3499220947049023,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.3046,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.36493592384006934,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.339,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.373526371524666,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.342,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.357321983277808,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.322,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.35090354297492843,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.3206,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.3888038628126265,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.3334,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.3798651314975039,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.3338,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.33652026827020337,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.3116,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.37535222552190517,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.3312,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.35002293213953706,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.3054,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.37752211826855947,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.3463,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.36830407708520974,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.3427,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.3358101623742109,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.302,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.4515398013176596,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.3459,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.36129572361360823,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.3227,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.36616073568565866,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.3443,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.3323297241198494,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.3039,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.357291546373255,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.3224,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.4004402361595311,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.3567,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.3758055883127765,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.3261,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.3527078407275042,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.3266,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.3525988958157503,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.3034,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.35786140779700853,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.3367,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.34059508239243863,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.3163,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.3489701576669115,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.3315,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.3416520196424154,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.3194,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.3590464915503749,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.3491,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.3588172250431156,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.3363,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.35344460101710223,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.325,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.3527068886477966,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.3504,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.38171120828330224,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.3312,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.36622864932668664,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.3181,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.3731469769715752,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.3314,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.47266583229603326,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.2998,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.6549736430744462,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.2567,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.46611508740284047,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.2112,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.3699298935805668,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.2266,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.44814752142030656,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.2365,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.5529344445304829,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.2391,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.6846366430604947,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.2375,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.5442427289901451,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.2375,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.43898976910330784,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.2431,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.41184576613974916,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.2348,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.4112601713719824,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.2182,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.47144071677166915,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.2302,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.44826497910576213,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.2429,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.4244593888087723,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.2255,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.3761159485079617,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.2227,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.4082305974263138,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.2457,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.3846212223482011,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.2271,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.40705071361310247,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.2634,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.3735462249852269,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.2169,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.38661096227775055,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.2405,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.38654321815889975,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.2376,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.3579104273569476,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.223,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.39291821184427095,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.2652,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.3519734440089084,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.2353,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.361963803628788,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.2333,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.35418409882992413,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.2261,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.3553403760830664,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.2123,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.37687519904841515,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.2304,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.37018622878061,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.2256,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.369805060123977,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.2228,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.3627531246292421,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.2258,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.3633872277323009,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.2357,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.36724243892548936,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.2299,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.3358245078512779,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.2274,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.34575893759593623,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.2304,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.36378989389938315,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.2605,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.3134798664585143,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1891,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.3457774249156719,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.2487,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.3408824964495063,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.22,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.3476923013126343,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.2411,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.33853816317546404,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.2103,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.3521214169127072,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.2397,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.3573704534189763,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.2391,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.3429710478443334,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.2392,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.3504929776212029,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.252,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.3314192907364034,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.2317,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.3366674324839135,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.2466,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.309991517702345,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.2079,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.3520610216670174,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.2419,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.3337757149341867,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.2197,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.3546288368987571,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.2562,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.3361861347533665,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.2179,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.35505931769490723,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.2351,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.3367597333627204,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.2338,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.33786048520570705,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.2366,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.34738281913874725,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.2337,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.3542455070889648,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.2333,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.3363386040570138,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.2432,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.32297473302354746,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.2252,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.33822028509397395,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.239,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.3338759306986732,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.231,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.340545450218266,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.2347,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.32570721587516477,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.2249,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.35310444728739016,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.2456,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.3410474751665048,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.2405,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.3318386335487195,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.2285,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.3329233126331173,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.2296,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.34981994274579076,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.2263,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.33333659439118535,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.2378,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.35479805461861436,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.2494,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.33068237603745504,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.2219,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.3362142428976966,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.2426,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.3543572957873194,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.251,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.329863862035154,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.2213,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.35009305511714717,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.2337,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.3064436993626813,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.2072,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.32311539042545545,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.2253,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.33857261235403485,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.2517,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.3291743753170019,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.2256,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.349991935914628,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.2371,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.3379263847269115,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.2336,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.3356175897069439,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.2491,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.317567885713126,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.206,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.3384754384789875,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.2111,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.35904680330472105,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.2564,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.3472494383588616,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.2217,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.3416181388197876,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.2167,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.34190298031889826,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.22,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.33964501826286747,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.2168,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.34046970227660917,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.234,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.3411120438416351,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.2297,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.332463311373053,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.2174,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.3433760777710707,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.2343,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.35103311264555137,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.2259,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.32579794046628036,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.2092,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.3439507059758399,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.2444,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.3487530742075529,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.2507,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.3358209606039147,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.2333,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.3383115420353886,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.2345,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.32275160031118905,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.2121,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.33093805538803717,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.2389,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.33074205224170344,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.2396,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.32431876997906406,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.2205,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.3367845690883795,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.2445,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.3356444216081217,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.2337,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.33969551485549165,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.2356,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.33664277819888966,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.2224,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.36562777707792454,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.2336,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.3404117649608594,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.2453,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.3425253628959528,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.2329,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.33481008725722017,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.2292,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.3347069993011057,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.2243,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.4076996602067296,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.237,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.3430884476468083,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.2506,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.327453655963697,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.2081,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.3391996374056282,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.2364,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.3432280546964972,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.2428,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.31917027614880145,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.2119,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.3304290231974731,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.2392,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.32473948280282794,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.219,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.34577931108312365,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.2558,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.3392844283417272,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.2299,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.32367018256868046,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.2144,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.34473659454279443,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.2334,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.3315378179239808,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.2209,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.3321350939267872,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.2397,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.32725608399262135,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.232,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.33576539424672747,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.2361,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.3388735685431849,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.2454,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.3281982778930743,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.2272,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.3342644068406303,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.2466,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.31381176494694285,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.2189,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.3342949589920847,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.2418,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.31933980227665015,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.2161,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.3280640178911102,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.2238,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.3622002864657316,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.2548,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.3314193366810418,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.2301,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.3468387928362536,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.2549,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.329850697221897,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.2225,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.32425659658661515,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.2399,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.32229639300764584,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.2431,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.33179893333434424,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.2281,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.3539164377090325,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.2493,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.3309339606921283,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.2235,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.33156995738371936,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.2314,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.319370061895989,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.2175,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.37260127777109864,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.2631,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.3529918350326705,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.2314,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.3419288829198312,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.2295,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.32880636545041875,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.2152,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.3524959630089287,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.249,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.3192520893743383,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.2213,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.32994485401119544,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.2479,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.3232555127449444,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.2234,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.3250828944060748,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.2378,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.32938999308929706,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.2519,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.33196990894183154,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.2199,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.33540340345278974,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.2341,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.3171884660548241,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.2245,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.33227455803299033,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.2349,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.3332784696970675,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.2339,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.33311460918722696,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.2218,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.3375160852274976,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.235,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.3179226571352801,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.2159,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.3356124798250683,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.2256,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.35108260231987115,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.2406,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.32969180343294635,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.2255,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.3348836124726308,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.2387,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.331610687493752,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.2268,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.32032524396441303,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.2176,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.3390427480919875,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.2492,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.31601027494141387,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.2215,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.33856111953381307,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.2418,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.33446579847092067,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.2336,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.32776352411541276,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.237,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.32107448781692166,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.2158,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.3299655429917047,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.2168,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.3422048685674382,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.2379,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.3263055693015261,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.2404,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.32185145244241364,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.2274,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.334544692871507,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.2331,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.31901546343803705,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.2085,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.3565431210269172,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.2562,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.32037080995651546,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.2091,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.32858166878735906,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.2248,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.3488890927402801,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.2237,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.3424275147405652,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.2094,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.3310169532428046,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.2273,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.33318226388842287,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.2457,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.31665252411102734,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.2042,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.3503255500374727,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.2459,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.3326952854343642,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.2302,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.31880019579480284,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.2005,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.33776152627735706,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.2232,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.37391840379278624,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.2485,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.342455426612999,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.2262,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.3400264500121113,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.2409,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.3316250684053993,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.2369,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.3307099239704301,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.2358,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.3482655645075621,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.2449,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.3411206750789066,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.2336,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.32301468983390613,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.2171,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.3320288738672142,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.2384,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.32579009933144853,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.2267,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.3379070815304803,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.2469,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.322186255090506,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.2377,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.3154598531298778,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.2272,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.3235889150140109,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.2361,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.3308838014973892,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.2424,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.31109159889161,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.217,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.33653402852365616,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.2401,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.32911061303028183,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.242,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.3219624760819913,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.2286,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.3402083897951386,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.2534,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.3253928225307232,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.2169,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.3355344072457448,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.2343,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.3293084493338922,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.2361,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.3361315260416007,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.2332,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.34463860608348545,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.2408,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.3276179572814691,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.2288,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.3315702123741426,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.2432,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.3250125448927904,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.2234,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.3071197330939568,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.205,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.3517143742996699,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.2524,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.3133055332728603,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.2139,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.33085718544719145,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.2319,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.34491100466167446,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.2413,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.3289729984874956,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.2256,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.31047284414496834,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.2067,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.35124995083682975,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.2706,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.3211475163343523,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.2379,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.3163720485565034,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.2211,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.3196165495007088,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.2322,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.3359366901275587,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.2474,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.3360458450775309,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.2483,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.3224330696609965,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.2209,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.3313498930262363,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.2424,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.3324509942614369,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.2164,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.3448560802782917,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.2446,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.32486190710277296,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.2151,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.35010381102036836,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.2392,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.33993208881511,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.2406,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.3273132379458591,
+      "learning_rate": 0.0,
+      "loss": 0.2225,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 2.91005441675939e+18,
+      "train_loss": 0.4889029575436096,
+      "train_runtime": 129066.8604,
+      "train_samples_per_second": 1.224,
+      "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": 2.91005441675939e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}