diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.988354430379747,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 7.422906561943123,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.2141,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 7.702888358358759,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.2357,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 7.607040860643221,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.2461,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 7.368306134061331,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.2423,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.835078152122324,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.2161,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 6.964397144414775,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.2081,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.316132364065958,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.1579,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.902913832761833,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.1375,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.8130839931248963,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.105,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.512064563287814,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.1053,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.4091881399732715,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.078,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.587635225934844,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.056,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.663001246290184,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.0499,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.678858853520033,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.049,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 4.196552632559953,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.0313,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.609117178689184,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.0009,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.49195564706361,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.9526,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.3443739095029534,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.961,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.056600049341981,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.9652,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.7788110626793354,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.9458,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.4861669241773643,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.9298,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.4802502151062025,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.9313,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.1797367969985448,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.9178,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.4184267468389586,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.9123,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.4184303224691999,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.9036,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.2086969336957454,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.9191,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.0087276201987916,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.9021,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.9773808141313599,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8691,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.1103695693947813,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8894,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.9183563363924115,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.8722,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.9667435376132645,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.8524,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.9245785516847066,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.8753,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8228697066781137,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.8518,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8787161788080151,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.8651,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.9178918792326902,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.8627,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.8373099967464986,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.8704,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.7796349718689348,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.8206,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7560600729668382,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.858,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.7627670424481396,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.8474,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.8268066426777285,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.8395,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.7595301729080337,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8487,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.8417818273672617,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.8421,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.7976175841169139,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.8311,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.7696880212380378,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.8497,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6898602703067724,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.8334,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.810695991603064,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.8438,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7391928307250119,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.8288,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.6748818730595351,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.8093,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7811665210420796,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.8566,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.732755622053617,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.8284,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.8585507610033377,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.8107,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7358462743858749,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.8267,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.8022519655901112,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.843,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.7562825833602647,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.8258,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8632169241640572,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.8296,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.9179202335837002,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.8229,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.7608870719537989,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.8329,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.7775910352288569,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.8144,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7414206933345299,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7995,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.7655383297288205,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.7837,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.7532460303456554,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.7946,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.9256377542780551,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.8127,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.7196432080359312,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.803,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.8728025641411254,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.8044,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7400126388766031,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.7909,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.8181555218738844,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.8068,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.0498432767171755,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.8314,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.73241682442642,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.8011,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8795722420292786,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.812,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8102584963963866,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.8047,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.8180615169255179,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.8269,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.854820045430093,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.8397,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.875528325514032,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.8341,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.7934833973760674,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7861,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.7763395839585001,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.8028,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.8555078992177862,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.8068,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.7215838558182447,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.8076,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.8174661307882858,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.8113,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7614165127784678,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.7687,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.7885266268596463,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7917,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.8456884079737658,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.8012,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.7971956825552432,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.799,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.1557842001990846,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.8001,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.123076808690034,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.8149,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.0543019347242801,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.8085,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.8779657466430091,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7914,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.2729004331242888,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7816,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.8925099084785951,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.7918,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.2159415348075073,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.7964,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.0116141720733391,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.7875,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.9026179965722234,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7807,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.9809839062467299,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.793,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.9466365916632071,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.8108,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.9460021467161306,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7651,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.8833555766943111,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.8155,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.012002209311638,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.7853,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.060561338839618,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7852,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.8860585760320311,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.8108,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.9595852827379758,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.7827,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.28979162299127,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7976,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.8714387770934588,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.8225,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.3556669058484596,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.806,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.0977615680011834,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7806,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.9226512643855256,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.8086,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.2031171503790807,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7742,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.2557771324364213,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7874,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.9689594750916495,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.8126,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.4706623463800523,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.8109,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.8442187507905325,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.7915,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.2938172224007785,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.8037,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.0348391280941664,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.8218,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.9569388985783428,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.8015,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.0385705413077693,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7858,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.0387012191650757,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.8042,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.930207803964864,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.786,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.0181152938397893,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.8067,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.3017526691238335,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.7991,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.862566338230393,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.7596,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.1859778348906305,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7926,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.9932495309097827,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.7772,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.8678606129628308,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7707,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.8296400799488185,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.7983,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.8863707866165301,
+      "learning_rate": 4e-05,
+      "loss": 0.7706,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.8553876669029679,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7974,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.8843499972869348,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7913,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.9892378501213175,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7864,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.9912561793138861,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7754,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.9263742129567436,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7629,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.8268197036763643,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7951,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.8592295743902952,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.7906,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.9621600073872334,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.8049,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.1297015662857217,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7685,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.8735600782330988,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.8046,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.7233820504807518,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7948,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.818252224384426,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7906,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.9787719832988007,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.7912,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.9139191062229017,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.7954,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.0772996585449184,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.7798,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.8588684709869089,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.7964,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.8914726219010631,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7645,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.8110703720704451,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.7953,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.9611800312155091,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7926,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.9935332398185337,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.777,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.1024193201998653,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.779,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.046566760232304,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7795,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 1.0635012613872763,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.8053,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.2263884881103142,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.781,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.7694287056700693,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7639,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.1441098535779395,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.8103,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.9434160524357237,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.8101,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.9337220850324455,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7559,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8853507657698833,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.774,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.7077940732549152,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.7824,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.8393545420679678,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.7619,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.7642725341733266,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.771,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.7519049499283067,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.794,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.7683869549787148,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.784,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6992025011007348,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.7642,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.7451991318497846,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7949,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.764693063050947,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.7667,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.8438261411772313,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.779,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.7682940519598314,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.7573,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.8222597274506226,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.7837,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.872493059401796,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.7986,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.8108178366597225,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.773,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.793295865751314,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.7758,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.9103072972882941,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.7596,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.6884027408000097,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.768,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.8156105079151512,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.7629,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.6974240086655125,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.7564,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.7591354705171401,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.7609,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.7478197619429846,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.7515,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.7077792412404943,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7776,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.8983402675887933,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.7572,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.7866377465100618,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.7905,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7451315686818718,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7725,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.9334599597198369,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7755,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.7728056048891712,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.7703,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.7336121614442274,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.7503,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.951832998291399,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.8043,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.6912552094292868,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.7821,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.8690509517904712,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.7897,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.9156594679864376,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.7732,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.7083690181019243,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.7704,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.8684600228878294,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.8071,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.8286735924788008,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.765,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.644803591081563,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.7555,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.9603718753018231,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.757,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.7806024113349918,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.784,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.7735964923620416,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.7426,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6798183683781928,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.768,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.7188994378218536,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.7691,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.8151940026207329,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7742,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6517888281494737,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.7557,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.7283875868643281,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.7575,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.6772420732803581,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.7839,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.6492949359036914,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7509,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.7320550136555497,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.759,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.6770590530250605,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.7474,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.6567170174134177,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7777,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.6638029692436324,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7502,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.7914206025693997,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.7582,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.6429109274873981,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.739,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.7200682249299569,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7831,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.7058877133045693,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.759,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6090259147185133,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.7432,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.6252393349067338,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.748,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.6519561604719805,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.7732,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.6609182790476402,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.7707,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.5774413768516462,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7617,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.677233674321569,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.7651,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.7021664078724246,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.7872,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.6966220014416574,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.7763,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.7455232261565756,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.757,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.6622847565350851,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.7368,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.7126018939621527,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.7548,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.7430956178475127,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.7588,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.7710560630378285,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.7782,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.7234817543857133,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.7255,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.7827501570663026,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.7493,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6938897150470469,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.7437,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.9124588976825316,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.7692,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.8712348374475667,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7389,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6444913400444665,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.7575,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.8368786104878179,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.7565,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.60517390937758,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.7628,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.78722958728958,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.7891,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.6181146423592847,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.7353,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.8132675101381872,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.7606,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.7476651654004239,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.7683,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.7403564794361924,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.7661,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.6938198154953802,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.7315,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.6418425074624672,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7471,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.6950931474260089,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.7724,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.6043752470613616,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.7475,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.6600694908165322,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.7541,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6355517988386554,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.7341,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.6466992416931774,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.7788,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.616373224166387,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.7262,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.6832061977148579,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.7616,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.6865507060588584,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.743,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.6713782002619655,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.751,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.7923260767618671,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.7491,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6471093210323358,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.7459,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.6751234936938955,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.7334,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.7131258616989125,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.7411,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.6759578201830426,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.6868,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.7499048657400524,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.5807,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.7738160367200365,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.6075,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.8950972640244442,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5806,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.7657095783391072,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5702,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.7292711247530594,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5608,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.7942307995927587,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5592,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.6241240174247079,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5792,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.8688781890845596,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5752,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.7096220851370832,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5792,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.7977956041566215,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5784,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.6959884327396273,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5759,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.7367100464798199,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5661,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.7848627693144904,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5714,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.7072970419275445,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5975,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.6418328465696972,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5506,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.6015495101887566,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5632,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.6881410670023462,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5738,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.5624936033298371,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5781,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.758894344885508,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5772,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.6155818673941443,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5476,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.600074583306017,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5783,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.6259998056379338,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5779,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.5968757428591893,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5636,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.649939492971231,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5695,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.6720168854239323,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5647,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.6022474904397626,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5569,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.7132796694935034,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5737,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.6564490040071934,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5539,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.7091843214977804,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5682,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.6787376233281452,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5776,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.8609369473056703,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5731,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.667043385771197,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5869,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.7137066255644852,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5873,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.7127007735925714,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5838,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.6571106308201223,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5643,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.8568720176734744,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5861,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.6708503911015374,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5658,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.7836658543569626,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5461,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8532075353454198,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5907,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.7282818181052777,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5863,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.9217433174077014,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5778,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.6032753539150266,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5679,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.8100159073206425,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5762,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.6615306533006793,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5801,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.7804651454175479,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5951,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.6524917872957402,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5702,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.800605448242946,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5734,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.6787064964890753,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5588,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.916962227478168,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5766,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.6448789674407905,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5878,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.8827828744890575,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5557,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.6785213736141649,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5896,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.7805908049035186,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.6043,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.6664009553598303,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5705,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.7717401826174713,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5856,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.6579488870124272,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5779,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.7527715597564992,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5771,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.6938214425462718,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5882,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.7488652517536876,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5747,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.6705798350169715,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5737,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.6997497214722849,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5787,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.6642603677726596,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5897,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.6691108585609711,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5604,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.6103940169709108,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5574,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.6316914785168021,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5638,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.6070084072716453,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5631,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.6742147434592348,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5837,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.6745199552678681,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5753,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.6521424754171411,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5945,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.5639285750616503,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5759,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.5784549604819715,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5566,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.6442341102911217,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.5785,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.6170745754076958,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5629,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.6025358473488737,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5979,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.7436227966464957,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.597,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.5711626070157921,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5783,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.7417487582659472,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5889,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.570380387735818,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5759,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.6814839069468793,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.6195,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.6014401731883051,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5858,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.6413910704447554,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5729,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.6725699833205581,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5755,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.61904187353618,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.564,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.6231720937723725,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5876,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.5945455207914749,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.583,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.6364864856215301,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5719,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.5765481305427157,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5396,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.6662296772640425,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5618,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.5963768195778902,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5911,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.6842804509915968,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5732,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.6403270062944663,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5782,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.6318229034469919,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5694,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.6794444617056468,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.5752,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.5986120749070791,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5565,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.6836940546297935,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5918,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.5645968672064384,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.6016,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.6280044265143101,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5826,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.6947768577268058,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5892,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.5351535542554965,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5789,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.7411448502333975,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5818,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.6248401248869119,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5729,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.7417540114342677,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5699,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.6448802866966831,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5691,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.6279941291510439,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5659,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.5794644009200559,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.577,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.6942530693871237,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5852,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.5530349301794408,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5648,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.6710732567235467,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5686,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.616614591965033,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5884,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.6344521815583951,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5796,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.6162449920656154,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5998,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.6812111295474964,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5884,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.6254483852767588,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5893,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.8009503462713765,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5754,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.6113734914406898,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.603,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.7790677757476502,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5793,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.6988719233477392,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.57,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.646891255831299,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5734,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.6909252507829963,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.6055,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.6391148480903689,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5793,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.6728992685927987,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5743,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.5820091632422281,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5836,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.5756224300064866,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5966,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.5678120061309038,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5789,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.5596316425914099,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5875,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.6000215348267849,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5695,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.5744870840098932,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5832,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.6699984736398222,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5717,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.5541715637633771,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5805,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.6564708312163328,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5976,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.6414941749135319,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.6008,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.6158831051880793,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.597,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.6174332657604278,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5568,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.6079905877071572,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5831,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.5754954056631324,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5922,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.6135591573447864,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5933,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.5889988966198758,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5908,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.5972905312079683,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5902,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.6528488222098003,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5818,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.519902491861363,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5616,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.6707317543494025,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5995,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.5651907315963569,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5858,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.5846349811286324,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5992,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.5403034463455374,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5842,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.6144338913042274,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5949,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.6064424138008089,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5882,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.6061579048797464,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5778,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.5776339337944888,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5744,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.6277627404899003,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.584,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.5820928364160212,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.6118,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.5518006844115444,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5735,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.6101933027626153,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.589,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.5424504494918261,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5606,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.6071719422491065,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5784,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.592262315360551,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5947,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.6268260638216024,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5892,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.6014083493496852,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.6242,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.6308707910406182,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5919,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.600921654537833,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5773,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.576719898839359,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5813,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.5567574780190918,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.576,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.5943149241707116,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.5736,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.5537818799649329,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.6097,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.6146829967678756,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5887,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.6074892822320576,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5743,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.5576585875935829,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5753,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.5568723468674603,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5684,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.5536228549283083,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5868,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.5475328245239587,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5776,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.6035241144346931,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5853,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.5428448334197303,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5973,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.6139069861722561,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5738,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.5598812048614366,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5765,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.5318400403036803,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5764,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.6279921179984012,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5947,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.5561702081603254,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5722,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.5667794855145091,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5567,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.6129724881434941,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5907,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.5663375603869027,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5504,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.5608350068394448,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.588,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.5932994690073904,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.6,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.6172928488767868,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.5473,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.5645464941401216,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5803,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.5665019798201438,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5835,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.5925231822379149,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5755,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.5746500245330605,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5789,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.6422262241511842,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5704,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.5976567214413147,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5825,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.5886954649653618,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5638,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.5795346475208982,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5894,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.6702230359984103,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5955,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.5746764286918472,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5582,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.6339683374247717,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5925,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.6517259101945397,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5842,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.5869203030884732,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5729,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.6121741552720062,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5662,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6034469837635289,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5832,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.6180480523200633,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.5804,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.6304083821074268,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.5949,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.6102293821075309,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5897,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.6353230782013719,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5495,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.5963632025686288,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5727,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.6543498162413061,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5774,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.6284153728360771,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5772,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.586293885074202,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5901,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.713520591902104,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5754,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.5840341960408489,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5844,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.6360936349426374,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5893,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.5941270891435562,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5865,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.6044753070298362,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.583,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.606033855344946,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5814,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.6159455680400073,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5768,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.53314880150208,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5797,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.6591551332422841,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5734,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.5440434542127797,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.5775,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.6241767618104105,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5695,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.5316605850513698,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5766,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.6146152562696269,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5802,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.5532658782564565,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5649,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.5684209223275096,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5799,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.5984287296029598,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5821,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.5814505415697852,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5608,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.5542013914246428,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.538,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.6934461216729433,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5885,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.5451843914784754,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.5696,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.6576316098175797,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5869,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.5695950218087054,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5952,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.5391382935816151,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5696,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.6173794618012214,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5684,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.5079133380973178,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.6015,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.6283768561241894,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5662,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.5793752000124497,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5868,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.5826447836009198,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.5866,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.6406419446251226,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5684,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.5899544214205517,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5852,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.6567482412929407,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5919,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.5722091289565628,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5562,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.6239606193268612,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5762,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.5237730566600719,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5529,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.5769330819024582,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5845,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.6364123661400036,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5802,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.5412287934129294,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.5941,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.6203020253083017,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5966,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.5547048949685004,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.584,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.5806137191389419,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5859,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.5627079943683599,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5727,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 1.1467451951323526,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3897,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.598736426609169,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3589,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 1.4290103252398918,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.3744,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.9141542755693245,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3638,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.6889418639210251,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3533,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.6777803538419283,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3453,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.7325735151469293,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3535,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.6164258253867978,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.3367,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.7609747260671386,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.331,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.585362617131824,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.3539,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.8320314395570567,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3491,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.5294981411362225,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3557,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.7451925021803085,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3581,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.6395532682640372,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3392,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.6454652954705963,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3517,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.5848955297477944,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.368,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.7059539554936636,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3552,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.5701994712141664,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3397,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.6395835996994635,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.349,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.6032834304254867,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3378,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.5743810929951932,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3432,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.5621717365930037,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3234,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.5582135902871759,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3565,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.5531656411368777,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3741,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.5238847427453437,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3265,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.5409163656755601,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3315,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.523411349827576,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3439,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.5812737335505982,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3177,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.5627276844771222,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.3564,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.5782781507056104,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3473,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.5706715225347717,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.328,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.5307302754137471,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3709,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.5842943541277226,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3479,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.5106620910936857,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3535,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.5583006398343057,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3526,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.5860635653312907,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3575,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.5556223547043253,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3625,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.6225035577744756,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3627,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.5435280666444197,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3673,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.5543615916823053,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3689,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.5369893746135964,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3511,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.5833027236737165,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3365,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.5352705607107424,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3449,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.6059463466092381,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3506,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.5228348083166242,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3373,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.5737822571547578,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.3391,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.507082687797643,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3551,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.5577807686000903,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3645,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.5671162729839097,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3402,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.5350562983608491,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.3409,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.5319043520200435,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3333,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.5380593719804613,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3662,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.5703602780963183,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3449,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.524834293079008,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.3294,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.5847484023196674,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3547,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.5099785477886943,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3588,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.5612979602857949,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3554,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.5223208645593906,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3469,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.5168552995541092,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3355,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.5116667388282751,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3433,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.5052923172740602,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3697,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.48093216577218983,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3549,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.5295154785527093,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.326,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.5249591684089627,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3523,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.49862213408669004,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3559,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.5085336774274728,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.347,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.5927833429976849,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3405,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.5101745672344602,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.366,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.5735962742699685,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3729,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.5377830389679108,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3607,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.5289463878885912,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.3584,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.5084586539129058,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3582,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.53168078733453,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3446,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.5207708682866058,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3439,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.5372784678300352,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3476,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.5231722601650937,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3539,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.51962312986107,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3502,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.546876581545464,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3679,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.5128464595353217,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3515,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.5485527662245728,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3477,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.5091846637663624,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3607,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.5438790378641924,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3295,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.524392266390704,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.354,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.5146198363808534,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3571,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.5196023243690108,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3573,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.5487660419843632,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.343,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.5289840886984594,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3618,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.4869146209573304,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3513,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.5732951426952306,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3569,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.5416757621033429,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3698,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.5575676372246374,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3584,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.5298245486874035,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3616,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.524733160111688,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3311,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.5246342477808621,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3551,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.5386174194059576,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.3397,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.5248187686482414,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.364,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.518403200351898,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3544,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.5049855068835349,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3547,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.5258787561456004,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.351,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.5171439371642681,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3398,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.5414675810773167,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3533,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.507698074828279,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.3562,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.5475534225179196,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3576,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.5468905885974245,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3594,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.5183213551075895,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3505,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.5974591763473639,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3253,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.5311196456485385,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3619,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.5337416605392257,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3697,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.5517098964288216,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.353,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.5220536095512881,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3276,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.548176066657299,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3584,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.5013237103678199,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3515,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.5656531817458654,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3451,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.549417323898851,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3343,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.5169857366333398,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3542,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.5656729642859899,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3489,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.5160428181767366,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3725,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.5256426571917004,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3526,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.5298895472354657,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3565,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.5092459400886593,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3579,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.5515892967783338,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3688,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.5233958887507767,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3605,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.5156448388000928,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3686,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.5560977136859876,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3357,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.5128683105820622,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3533,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.5842938508109689,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3735,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.5153960811014598,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3636,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.5307476734117271,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3297,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.5354432560832244,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3305,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.5329298536579019,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3569,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.5131483533082781,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3469,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.4931851660231088,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.341,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.5070344832279179,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3459,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.5150359672995605,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3589,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.5128951801286649,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3625,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.5432456147970126,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.37,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.5020331016688848,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3439,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.5119427228718203,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3512,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.5292958398750576,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3543,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.5066780376661576,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3412,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.531718078979558,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3587,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.5122868421077256,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3647,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.5164253050547973,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.363,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.5022349114684149,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.346,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.5103599468182322,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3556,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.47879574996129315,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.3382,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.5256457856197226,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3728,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.5210533258128821,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3641,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.5088983627833931,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.376,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.49894383122524977,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3668,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.505808311253387,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3688,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.5444525354414488,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3622,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.48825466210379204,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3688,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.5426170460066809,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3517,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.5085891929958759,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3655,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.5101875525975765,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3464,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.5189872604787611,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3616,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.5543185139858463,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3265,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.49817183920609237,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3525,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.49049506520471076,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3508,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.4961057824642537,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3544,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.4948648975363461,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3688,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.5009677435471476,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3557,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.5062193636036492,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3597,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.5037754534062563,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3467,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.4919285483562402,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3732,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.5139717667618598,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3564,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.5227693211649667,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3705,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.5028551945597166,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3559,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.4957579430930181,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3364,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.5142133921405685,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3507,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.5001030931865029,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.363,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.5275576239759346,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3514,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.49851548137227575,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3318,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.5247116439977663,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3587,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.5185200991170165,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3404,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.5200453659664607,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3488,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.5041265521402949,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3577,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.5450444035513459,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3533,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.5324728761917483,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3619,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.5113212731410814,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3403,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.5452428130278169,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3593,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.5366425260304866,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3237,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.5281178947505906,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3654,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.542379830630766,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3445,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.5122090254770407,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3797,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.5155164680057729,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3376,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.5101651454080011,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3494,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.49167748603583405,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3599,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.49648064021098104,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3614,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.5291452285486734,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.373,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.5086641189829165,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.3539,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.5104518943903652,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3486,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.516282210941752,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.353,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.4925255024789146,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3378,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.49280526872648117,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3359,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.532973794319545,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3469,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.5123955488266961,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3499,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.4972365224713364,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3656,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.5116762396081894,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3417,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.5075495811783169,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3566,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.4879520070659885,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3518,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.48712003298754586,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3455,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.5355560564619778,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3442,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.5233529574763391,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3439,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.4964479071739173,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.3567,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.5095092759864654,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3412,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.4956451840934938,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3559,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.4987209190231309,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3539,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.4798379430516264,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3314,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.4895060618025084,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3447,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.49665027048741794,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.351,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.4830410522208456,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3497,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.4994974791098523,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.356,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.4879731254136405,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.351,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.5331734893982305,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.3647,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.4926817990072902,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3658,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.5018186689336128,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.3491,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.5083997033685884,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3527,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.4783065620138451,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3606,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.5159465047052381,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.3464,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.5229619525240099,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.356,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.5006770354264682,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3597,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.48997274763251697,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.342,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.4811330772646758,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3371,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.5060946792118203,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3477,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.4849921513652448,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3366,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.4668270034186336,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3598,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.501091815599208,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3525,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.46994232691005056,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3566,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.4767696566021661,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3352,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.504642786508532,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3623,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.46414793732392495,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3421,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.48254720988938593,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3595,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.48204855223358667,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.338,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.4725135320040492,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.346,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.49064330057836847,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.343,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.5067972254387637,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3605,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.4972840756706302,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3408,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.49578095650107123,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3551,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.5056173725936419,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3625,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.4784220796800086,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3278,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.4825968889057289,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3593,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.5162483542819932,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3322,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.5102419816436795,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3599,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.5168829454646838,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3494,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.5662622019110052,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.276,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.591313280379996,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.1953,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.44290538380345884,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1788,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.544854142605612,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1671,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.7554161569060809,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2024,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.6675362855185404,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1945,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.5210226455498518,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.192,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.47788157417854316,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1768,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.5505960811315909,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1966,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.5420970882975261,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.1881,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.502872385932334,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1947,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.4502226942966714,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1874,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.4741700217872557,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1896,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.5170821728085679,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1881,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.49386047197199556,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1988,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.4834574902694184,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1786,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.44768383614220164,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1811,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.41993652198160514,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1882,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.44455919590874327,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1924,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.4324650972309064,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1961,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.42312426948366366,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.188,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.4283705138285704,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1906,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.43866681463783164,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.176,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.43812470078753163,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1929,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.43948369964762235,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1937,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.4361360763648269,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1794,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.42092339927211986,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.186,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.4180585745841289,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1757,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.4080496311735775,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1885,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.42247520286971996,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1733,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.4119605458621346,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1843,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.43427306481438704,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.176,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.4027480085994738,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1908,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.4193463277928552,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2072,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.4042729507992835,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1892,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.42343789191296277,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1906,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.41168601149960643,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.189,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.4103703273186835,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1895,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.41173623815950405,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1714,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.4139487182034766,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.156,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.43584902189323804,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1903,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.4087079078662876,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1892,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.4065405585418902,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1946,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.4198227191563816,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1791,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.4094954753223793,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1792,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.4060290605634232,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1837,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.41415872832660017,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1705,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.40557470000658796,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.1821,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.4150994695594411,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1806,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.4206949275835078,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1907,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.40778779290502376,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1752,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.4070441146935695,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.183,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.4133886208312309,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1786,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.4045450501972083,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1789,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.4086578648925976,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1756,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.4108150281668105,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2013,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.4149629483605243,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.179,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.40092227050799045,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1775,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.4067476663630985,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.163,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.3961714615045738,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1923,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.4043031930057585,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1759,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.4131633554057088,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1907,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.41796544523545925,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1701,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.40311044018762904,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2009,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.4098979532382286,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.181,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.4196120473550571,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1762,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.4253201551645483,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1779,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.4265982803900859,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.188,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.40787381200031253,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1857,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.40463018342670853,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1668,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.40476543857184494,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1772,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.43063314605808906,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1833,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.41005682850773406,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1779,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.41161570850939855,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1821,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.40525731405966053,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1712,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.3905100270901123,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1793,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.4121652428903168,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1899,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.4030166903267624,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.195,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.40646742230867583,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1851,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.41961006536877465,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1735,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.41755220005525995,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1713,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.43067128675324096,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2005,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.39154285241133374,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1812,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.4181725155055304,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1741,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.39891562908613065,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1796,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.4061482936444006,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1473,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.4035688108538141,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.184,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.4197345524013537,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1888,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.40292294646410654,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.1851,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.4166981569917705,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.201,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.42295716726823973,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1862,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.4134836497860457,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1894,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.40806184596549566,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1841,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.4100187296175628,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1965,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.399698754545325,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1927,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.41658978492643206,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.18,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.40167358903277595,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1808,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.39757608938792666,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.1957,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.4096715436060928,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1809,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.4181583193437588,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1596,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.3883676226794139,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2036,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.40981613803575595,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1702,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.3932778077299231,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1728,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.39550131081349227,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1921,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.40829794421881943,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.176,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.42446313655295353,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1733,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.40746125050205984,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1799,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.41679742562840333,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2071,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.4002372705143844,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.1713,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.4038920269017868,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1954,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.3974499449639607,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1654,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.4099661487723157,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1965,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.4092770096200827,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1803,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.40270730881311523,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2006,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.41497517855526916,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.178,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.4265562416286503,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1664,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.4001068332218262,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1921,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.401596094807994,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1789,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.4211757141618712,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1898,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.41476209270543274,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1948,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.403456117622833,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1817,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.42747026362702084,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1689,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.41185251386214045,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1835,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.4010277827091964,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1865,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.4125882968452589,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1702,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.42255741458453466,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1679,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.3959243433524398,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.1728,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.40824715918795607,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1897,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.4023838090519276,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1596,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.404566734343757,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1836,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.4017549816715845,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2029,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.38946363500028813,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1907,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.4001183059713056,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1811,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.39571459226438466,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1775,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.3994023758740196,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1967,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.4100990904342074,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1687,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.3989299276787435,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.1879,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.4156426332289341,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1771,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.41613953576934715,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1712,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.4036256829793108,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.181,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.4037047543201808,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1654,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.4091873345494449,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1833,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.3889151418507826,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1815,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.3833507510229414,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1808,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.39416121713177527,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.158,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.40598299853923875,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1763,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.40639483145026994,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1752,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.39579228130451066,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1819,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.39859420590879974,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1917,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.4186268615775318,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1911,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.40151393809800556,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1737,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.40517474970626133,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1946,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.3978953746616613,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1789,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.40043102896400906,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1756,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.41062220482908196,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1947,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.41438255501517085,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1619,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.39405983168517356,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1813,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.414151176363047,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1745,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.4041231612197135,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1834,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.39494069313483277,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.1685,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.3984240481879914,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1917,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.3921281222163095,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1889,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.40667886781947105,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1695,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.4039507217049874,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1883,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.40895893753482954,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1721,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.3945991509880496,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1812,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.41695522007995567,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1685,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.4114451010761483,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1756,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.4082536105534582,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.173,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.4008788542787994,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1608,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.3939131683463835,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1734,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.4007540758965614,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1564,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.39584341775993465,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1864,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.400951562044233,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1627,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.39229124017761474,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1917,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.41741114700762194,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1714,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.4086615440025275,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1784,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.4028844051100741,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1869,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.40985539378313235,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1655,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.40417511245941,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.1517,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.40368705428471785,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1827,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.40942805525984016,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1904,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.39804778623599696,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1634,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.40704934648229973,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1881,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.3951961845204826,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1865,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.39397926743792056,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1682,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.38684806197585014,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1657,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.38184485350890657,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1846,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.3913136874180568,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1864,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.3940454045344097,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1655,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.394718537971424,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1567,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.4039064180538771,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1679,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.39103956831409514,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1796,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.38116927387026345,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.195,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.4006286318964666,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1988,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.39342110551312015,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1637,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.38529156389645525,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1745,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.41276448116523873,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1835,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.39315811098571113,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1828,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.4017979527522645,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1798,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.3898765745565289,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.207,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.3898844416265708,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1725,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.39297700685030346,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1677,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.4005862147539681,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1764,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.4147416497927672,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1897,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.40275643927334703,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.174,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.3946541915307498,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1792,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.38908506880670085,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.17,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.3894206309782253,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1684,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.386048656668763,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1797,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.39925853497503977,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.185,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.38738465939273226,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1892,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.3949392326150649,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.1885,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.4040812071091604,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1682,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.39859306983166665,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1789,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.4102692913188641,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.19,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.38410191984112413,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1874,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.38777984483457595,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1818,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.3923175381778711,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1685,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.3857523429222143,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1704,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.3886777555579344,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1817,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.3823123569461595,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1801,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.38783140284883677,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1757,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.37879805611499157,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2158,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.40465190077925745,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1748,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.39464049240322635,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1834,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.3900591124106923,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1948,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.3972539423413394,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1849,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.4054539323667461,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.166,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.37951889406767225,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1842,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.385942993073014,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.183,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.37519635235014187,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1667,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.383358544111056,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1695,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.3957850990419129,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1878,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.4004521065414957,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.172,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.38511005570085743,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1881,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.38163090929203963,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1935,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.4016917785149485,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1973,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.38849856851504333,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1923,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.38734023890889896,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1804,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.39213743718647803,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1783,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.3926732879614748,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1771,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.3822658194779312,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1647,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.40017105776467715,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1573,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.3913758666605072,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.186,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.390529399350253,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1861,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.387873552161067,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1794,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.44838519256246356,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.1179,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.3939183534401255,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.0982,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.35016046653281024,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1372,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.3019998239639578,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1376,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.28113319398014824,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1049,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.29259512926401193,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1048,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.3143129094278038,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.0963,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.35084536381592474,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1036,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.40795232446394236,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1165,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.4090398154896428,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.1155,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.3768095297670197,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1141,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.31934987843065016,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1044,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.29748740705322996,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.103,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.30018918160745417,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.0964,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.291819601692486,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.0992,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.2978775055410628,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1142,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.32193967449335664,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1058,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.30076784271055523,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1007,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.3106205598652892,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.1174,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.29447409431822275,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1106,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.28231366143529274,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1049,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.2926613311303082,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.105,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.2898071577554333,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1144,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.30626538355071986,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1174,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.29041061096673626,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.0959,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.31484508807358774,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.111,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.3131288323885708,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1112,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.2983392906648746,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1119,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.30162977284048914,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1053,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.29638225701506016,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1154,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.2842395116270297,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1148,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.2923955075913173,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1067,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.28093816337032695,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1069,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.2895027241975469,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1061,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.27904072942285335,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.0994,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.28655173867538286,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.0927,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.28073775387361993,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.0906,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.27760847030615876,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1169,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.27138598404622943,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.085,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.2920131374462114,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1289,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.2853166042151861,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.0977,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.29591268748296573,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1272,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.29205488852334516,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.0909,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.30161652607991146,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1064,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.27863977926731925,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.0969,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.28484262150810025,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1117,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.26400435153201113,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.0807,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.28828600187407016,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1295,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.27560046920297987,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1027,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.29013660312928413,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1105,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.2756969928420552,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0989,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.26529832463316827,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1019,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.2702030145560112,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1118,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.2817739035183043,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.0845,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.27948923961455685,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.103,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.27211874884069803,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.0899,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.2696293940948361,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.0927,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.2824162643862469,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1027,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.29009031611323605,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.125,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.28870196513788027,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.0958,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.27840868625141046,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1275,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.28797576558436117,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.0948,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.28819245251614795,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1076,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.2759113011894281,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1157,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.2821713149142146,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1032,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.2736318897910862,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1105,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.2825033034055862,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.12,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.2865428863953339,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1068,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.27656289803254186,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.0939,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.28000488037986326,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.0944,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.29843936083585854,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1045,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.28390383088255255,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1002,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.2683530306312209,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1088,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.2952909496720311,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.0944,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.28899124609356014,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1026,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.2795649660478316,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1008,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.28244375128784205,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.0938,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.2728658614455728,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1044,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.28584902637046306,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1148,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.27346186279377516,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.092,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.2841510750079851,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1071,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.2767963772838676,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.0985,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.2989847558810521,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0938,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.28840249870998597,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1129,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.27948463456333483,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1117,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.2838376112679734,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1023,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.2783566080989638,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1056,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.2682834125311438,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0945,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.2974042105202281,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1197,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.2886314002098791,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.0996,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.27948146002077523,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1011,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.2702490729196437,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1123,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.2854573174160886,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1106,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.29398267773516046,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.1357,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.28893347710892503,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1227,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.2831470229908277,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1077,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.32691668340708,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1044,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.27235200462069226,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1057,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.2835051078228697,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1013,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.2914864273931296,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1169,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.27873942614942643,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1166,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.2836702038743559,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1114,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.27579096122126356,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.0878,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.2768992427499534,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1023,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.2797546516658488,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.12,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.42703040325789804,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1108,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.3078236078569076,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1077,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.2722486533113106,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.0997,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.2855291378851576,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1363,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.2771648197230272,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1029,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.27741155615486296,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1243,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.2870921503454232,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.107,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.29179872327992146,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1151,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.27571922014534644,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.099,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.28019678084771765,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.1016,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.28093930509385734,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.0923,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.2687381578522314,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1101,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.275091184312658,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.0895,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.2814519735878001,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.095,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.28494847063205264,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1033,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.2894463505885869,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1134,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.280487729104497,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.0976,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.28244287324936507,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1051,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.27957055983741863,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.0871,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.2822610821761621,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1196,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.2856597859961196,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1085,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.29064942570236535,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1185,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.2883253668740008,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.0865,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.28193070038597073,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.095,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.2815286503531397,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1099,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.2800009350916547,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.0898,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.28213137055022,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.125,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.2846828907772675,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1055,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.2722560101864204,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.116,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.28206499728906004,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1041,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.2766426145515931,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1064,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.28898751746623863,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1195,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.28292482083683695,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1025,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.2846471448175624,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1245,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.2855028157059443,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0968,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.27074130048177236,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.0955,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.2962110822495446,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1138,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.2981202282032781,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1014,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.28956167826946166,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1107,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.27659626341815613,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1085,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.27976046692850187,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0837,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.2659432931979348,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.0944,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.28088565682498046,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.0932,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.28879311973555327,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1097,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.26994629763091244,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.0849,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.27987436325614384,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1086,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.2874591290105128,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1151,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.28655893585983,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.108,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.27899165545553917,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1079,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.2816191538213561,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1116,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.26180467030889476,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.101,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.27017337788580603,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.093,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.28326785529905313,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1165,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.2590274109827604,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.0825,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.28642010412918595,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1121,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.2848386676793521,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1292,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.27305246208921624,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.0786,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.29530345044301565,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1242,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.27728457755005465,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1096,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.27377885136934166,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.0978,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.287727553202241,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.0952,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.28518026691643544,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.0978,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.27222627447297065,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1042,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.27473179384192653,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.0931,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.2793974769662782,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1017,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.2760962842334961,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1022,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.2733933440143945,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.1098,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.27447959292280316,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1106,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.2737092459417993,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.0872,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.2682762494777621,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1114,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.2734848145878755,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.0915,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.28473699914824874,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.0985,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.2862293498666996,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1143,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.2733780153929358,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.0923,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.2755185331284402,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1012,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.2806492902137361,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.104,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.27981454058813965,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.0991,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.28805667429354,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1011,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.2805822234971483,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1044,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.2898514456403742,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1116,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.28401927855325637,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.0983,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.27993276752877705,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.0977,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.2714989501442365,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0912,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.2738250698481966,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1092,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.2864194007077226,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1079,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.2865066214329713,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1116,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.2730122802422818,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1122,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.2819738729393376,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1048,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.2822750176168729,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.0989,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.2737875937204234,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.0911,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.2877793960693684,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1095,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.2987700401893648,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.1151,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.2735178363047845,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1065,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.30747782015344927,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.104,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.28357347351079415,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1079,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.28930137486931695,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1179,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.2871553287021005,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1022,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.27996836582839774,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1079,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.2782027408284261,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1037,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.2713371394378539,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.28253056723941306,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1136,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.29248511879903333,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1061,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.26449437454905494,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.0978,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.2978456296540937,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1212,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.27941323050349304,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1088,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.2768885918781328,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1055,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.2732692431060451,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.0941,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.28049971528372064,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1103,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.28373743090759745,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1005,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.2830099239382214,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.0874,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.2817019400580277,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1151,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.3022828414678997,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1135,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.278003827844547,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1028,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.2766730309633947,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1001,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.2764584143484117,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.0947,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.2884769683214301,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1157,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.2831440336574793,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1012,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.28225464621440044,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1061,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.2760948390512447,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.0908,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.28637230448033324,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1177,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.30001346011925983,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1051,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.2880357072179983,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1198,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.28638810887046584,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.0962,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.28777615740353507,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.0919,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.26472556575656997,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.0995,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.2680754882921424,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1129,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.27273777842879465,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1151,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.2881782378321563,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1221,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.28983603669957686,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1151,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.28750970234802037,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.102,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.28405632558903404,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1078,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2791151923495753,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1114,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.28616669525748795,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1065,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.2766672009427938,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1043,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.27781062778516236,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.0976,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.2873230852811925,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1135,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.2737390398011529,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1141,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.2769167312599856,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1098,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.2881077892743404,
+      "learning_rate": 0.0,
+      "loss": 0.0965,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 1.3118518348470026e+18,
+      "train_loss": 0.4075780885672666,
+      "train_runtime": 43187.9684,
+      "train_samples_per_second": 3.658,
+      "train_steps_per_second": 0.028
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3118518348470026e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}