diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.982278481012658,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 5.911349107877243,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.1952,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 5.932879666164354,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.2155,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 5.92441185998976,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.1982,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 5.990646821124674,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.24,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 5.876644724017003,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.2201,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.353110250343048,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.1901,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.399970160405909,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.2113,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 3.849489497403001,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.1209,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.4450297343285325,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.1078,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.31465435840476,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.1284,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.1560226566965417,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.1528,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.129939811938016,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.0983,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 3.1933590697199254,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.0655,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 2.9155304844708576,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.0631,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 2.527251908606466,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.0826,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 1.9590087370716645,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.036,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 1.9001264450940216,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.9892,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 1.695733333290717,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.9882,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.6586498176332252,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.9365,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.3905616985485747,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.9446,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.1571291215329313,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.9125,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.05685003007498,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.9305,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.3507879027485863,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.9412,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.3848850560792798,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.9311,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.2389584025947282,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.8984,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.940966639102869,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.9173,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.007614546553969,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.9267,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0368298378497953,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8834,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.944275492368014,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8762,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8421180545697943,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.8632,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8839646629140581,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.8779,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.7921115656966778,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.8861,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.7803042541153036,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.8819,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.740797457713017,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.858,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.7520612612143182,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.8526,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.7245869395931376,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.8718,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.7454266242590425,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.8621,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7541012874642258,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.8407,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6866340554253384,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.8805,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6994156605522887,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.8375,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.8087930577359883,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8326,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.6465611199171433,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.8309,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.6803235418059077,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.8339,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.7874898872794668,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.8601,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.696867088214512,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.8229,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6749697921107919,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.8021,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7936965075494307,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.8365,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.785550146700562,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.8535,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.74711665723246,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.8446,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.6908821445002004,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.8325,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7097258639748224,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.8311,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7096172161094633,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.8302,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.784185827876358,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.8341,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6644942795017084,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.8346,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.747263950581024,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.8136,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7049577816857541,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.8025,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.7606958686062485,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.8185,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.6405708325890455,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.818,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.872463470160665,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.831,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.6965550277984938,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.8172,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.8289346198195373,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.7807,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7325082396798934,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.8109,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.9144034116990944,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.7948,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.7575129708790727,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.7842,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7556328861086589,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.8101,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.7553918325376433,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.8199,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.8897804979321724,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.8244,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.9951515743204266,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.8043,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.6870761079461248,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.8059,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 1.0097000989489255,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.8273,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.7836683190393479,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7861,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 1.0896663070660786,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.8098,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.8504694324238093,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.7714,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.1053752927834413,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.837,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.7650563696829097,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.7831,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.1134762539439764,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.7634,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.8958629817628895,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7801,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.9367358157768769,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.825,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1149356061465916,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.8281,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.8305945364058006,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7831,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.0534228934994334,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.7861,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.8197273574589239,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.785,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.8929948429955872,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7933,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.7651322767568661,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.8139,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.0228007132652852,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.8212,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.8021908685081163,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.8315,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.0590626663287181,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.8089,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.7692340364998896,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.8021,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.9773966100492726,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.774,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.8679428812297443,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.8229,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.0543621078648076,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.8138,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.7273116077835361,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.7449,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.0317234955214118,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.8064,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.971501575183046,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.796,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.8119627629945337,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.8416,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.1878895350913057,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.775,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.8000412557863799,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.804,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.8351421997568635,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.8242,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.8128873643656386,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.7872,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.8258421318570428,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7814,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.9579937313479742,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.7823,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.8286051214928695,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.7985,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.8543588426711834,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7834,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.8559745670257315,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7765,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.7676065751964928,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7509,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.0515115317386825,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.797,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.7742475017624065,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.8385,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.0478996314522693,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.7669,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.8164010428525539,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.8201,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.8658870447569352,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.7902,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.8603906153808364,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.8056,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.8454255830647611,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7934,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.8222775462982067,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.8012,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.792066962401852,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.8067,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.7789560774099055,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.8195,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.1497924711967131,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.8199,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.7340529447409657,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.7602,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.1008401794077558,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.7626,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.8705496968278656,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.8275,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.8799760312834677,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.8213,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.8522430130150677,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7759,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.7322300755037081,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.8011,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.8952546401565277,
+      "learning_rate": 4e-05,
+      "loss": 0.7398,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.9697870919601368,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.8362,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.8116822399104134,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7768,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.0533293570461872,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7935,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.0055298897289344,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.8345,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.8546922030753635,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7901,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.9936402082193629,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7776,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.0400872588468657,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.8086,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.7741897975838393,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.807,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.1151673031941662,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.779,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.8846752975276744,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.7745,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.8291430842921892,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7665,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.3014841557038692,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7759,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.7704401761981993,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.8232,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.9567705458965501,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.8218,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.9199766980900497,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.8122,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.8984940927763728,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.8306,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.752382672574233,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7903,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.9597300182486667,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.7989,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.8524609000770883,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.8004,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.802887488069474,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.7892,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.9535435002142715,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.7538,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.0947285242218714,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7637,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.8076551511998495,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.7861,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.7826900396570292,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.8258,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.831315048191772,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7948,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.79935521554649,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.8062,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.8375419782692419,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.7922,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.9113137051001848,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7776,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.6992725088675615,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.8258,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.876880421396542,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.8117,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.7449046117126098,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.8128,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.7634820778514588,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.7817,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.7320629574583174,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.8166,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.8658541156060348,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.785,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8772152559447359,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.8052,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.0331116175590438,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7653,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.8189163380576091,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.815,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.0394135594036276,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.8334,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.8051286576013484,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.8014,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.2424907072721838,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.7888,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.8448011896686858,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.8111,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.3577892696242733,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.7868,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.861986154016911,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.77,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.4281613202456989,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.8139,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 1.027777402613741,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.802,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 1.2172117131486848,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.8194,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.9513125305207866,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.8028,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.1130330416349528,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.788,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.8629897706660146,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.8097,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 1.0007791331187414,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7886,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.991145588247539,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.7959,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.7243469061117734,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.7785,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 1.0289092928527488,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7898,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.7176360224448203,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7852,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 1.0609173340040523,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.751,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.8810670035531808,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.784,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7938315707513133,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.8113,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.8061940665783219,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.8023,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.9741718269109972,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.7706,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 1.050454382113595,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.8026,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.9005123142588245,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.8094,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.8317685573134472,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.7566,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.8902128205858185,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.7497,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.7660556914491132,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.8549,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.7845711132352827,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.8006,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.6942074303465905,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.8048,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.676341053060021,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.7868,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.7337784456734684,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.8458,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.694636402673126,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.7543,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.8519782624924656,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7681,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6597648393898234,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.7886,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.7127841242963262,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.7718,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.7023703760562304,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.7997,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.7272638774349355,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7467,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.7217398560860341,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.7783,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.6644412052654356,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.8271,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.8256947419983265,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7805,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.6938118059781863,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7621,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.8679266088264159,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.7606,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.7971703747135689,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.7838,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.611158952751061,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7794,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.707248859491207,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7757,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6090100191206436,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.7452,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.7133423630508406,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.7954,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.7881734818690332,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.7664,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.8731549309766675,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.7636,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.8183847546617082,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7811,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.7862292802935588,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.7995,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.8029171977386159,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.7669,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.836589012970665,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.7387,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.7692990376026879,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.7757,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.7924431743677928,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.7913,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.7745753097887749,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.7532,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.72774516522752,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.7916,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.7657535576451227,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.8104,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.6844100777121547,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.7945,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.8006410618051576,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.8227,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6205293929170017,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.7639,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.8074285513304484,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.7698,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.6783104092216579,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7288,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.8661530537764257,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.7778,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6651570152335079,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.7691,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.7223689969530322,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.7801,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.7280533772174048,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.8075,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.7548829047110128,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.7637,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.6908722664737806,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.8103,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.6621512877698958,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.7959,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.6741214996030168,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.7578,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.637962758301598,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.7839,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.5787814617999634,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7744,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.6428611912951809,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.7727,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.6521864074168835,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.7773,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.748265227867317,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.7862,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7154990529913513,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.8216,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.6751012194032581,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.7881,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.7342853452490749,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.7889,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.7556623256194607,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.765,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.7628703888646837,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.8265,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.7319703440651463,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.7943,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.7421194155640197,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.7314,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.7536692898471273,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.773,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.6821460183411494,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.7881,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.679035901655673,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.7818,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.7510902088884673,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.8901,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.7705778132174754,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.6301,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.8049190537342664,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.6162,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.8384086960606655,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.6615,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.8622920772080463,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.6494,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.9258711647579994,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.6224,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.7520688525149218,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5763,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.8269426486781531,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.682,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.8653248718087412,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.6123,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.756484029817264,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.6321,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.7609245554608721,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.6103,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.8992249054147975,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.6288,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.7359908497121891,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.6105,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.756678904269953,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.6456,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.68417930100335,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.6046,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.6643088611641691,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.6525,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.6326526137189697,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.592,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.7677330673239964,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.6099,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.7206662369578765,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.6427,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.7256301424890368,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5744,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.8437390643880228,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.6097,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.7105373466078398,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.6409,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.6602346700091581,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.6318,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.6747948214788924,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.589,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.8030988329753774,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.638,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.6186590752032033,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5996,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.8650746642408986,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.6783,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.6127598579763573,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5666,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.6927700082782016,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.587,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.7079357446950084,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.6506,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.5889537013128078,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5792,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.7569382567629807,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.6235,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.9210893935291109,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.657,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.6102318374122997,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5979,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.6753260291066613,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5825,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.7328902963418659,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.6449,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.7454873874265894,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.6128,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.661252992642923,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.6431,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.6990902668360423,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5964,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.6931783013254048,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5716,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.7963290197062688,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.6062,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.6502009486362453,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.6488,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.6364908846669588,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5998,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.6584166319714647,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.6114,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.6276048695908606,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5786,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.6649304960200125,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.6278,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.6820965384170361,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.6267,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.6650642690253378,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5919,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.5950861746879271,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.6083,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.6662344433884583,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5883,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.6783778372644386,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.6493,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.6483141617598359,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.6013,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.6028845930594864,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.6112,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.6767936396110088,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.6323,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.5554382604693996,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5923,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.8571554827259303,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5831,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.6554054168983362,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.6695,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.7030248424711922,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.6029,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.6013508954303972,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5949,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.6490516732805707,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.6375,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.6568096326977396,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.6266,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.6860530305766259,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.6081,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.7062831726749773,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5975,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.6516685744473459,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.6099,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.8046471445879211,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.6247,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.670130479410025,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.6042,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.8097812714638303,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.6346,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.6778074615658999,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.6449,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.6776935054017306,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.6125,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6519421907417527,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.6537,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.6690106553787741,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5961,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.6296286564447195,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.6073,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.6708435854135021,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.6337,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.7554067300149826,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.585,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.6602745868151292,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.6526,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.7882890625974472,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.6164,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.5985476213773421,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.6008,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.739597106824148,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.6582,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.6868129930014946,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.6215,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.6235655121458028,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.6267,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.6392363551970746,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.6322,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.6412608682664244,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.6068,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.5811340764913041,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.6005,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.6574299181135196,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.6395,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.5546893114135103,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5536,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.7005883422931257,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.6317,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.6184995107803999,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.6388,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.6031707741063155,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.6077,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.6801362492032821,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.6028,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.816570601829903,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.6149,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.701399133382796,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.6419,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.8525004265321319,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.6318,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.8498526375046962,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.6043,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.6902513893073611,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.6079,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 1.0641909241286616,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.6225,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.668094956250389,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.6211,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.8099933050086029,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.6301,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.7285990013375508,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.6518,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.7833357942185337,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.6457,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.6727009653937923,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.6296,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.6487783825402778,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.6126,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.7132393390285112,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.6465,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.5607449838869152,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5931,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.8097150275524085,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.6436,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.5741362270649183,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5897,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.6315562286767517,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.6283,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.6771112349616202,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.6354,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.6300451226409386,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.6218,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.7769148554699374,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.6056,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.7433056505971002,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5978,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.7046193153493635,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.6149,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.6221357812402802,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5978,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.6869230020629721,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.61,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.5795874285893969,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.6583,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.613968178821049,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.6081,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.723568673180714,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5974,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.6188525102828321,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.6145,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.7508747183273365,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.6505,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.6232982943080587,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.6534,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.6914166419388095,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.6111,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.6192097496995117,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.6082,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.5348281350440572,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.6305,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.6461245600778388,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.6218,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.6297145868498049,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5941,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.6183043711733204,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5832,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.6950762332144292,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.6674,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.5777269873217972,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5712,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6612126442096017,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.6502,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.6088890510631714,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.6507,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.6180586812639303,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.6064,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.6490542614543741,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.6217,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.6229241484700092,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.6405,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.7005518442381004,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.6432,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.5462039962646452,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.6211,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.6876380472720047,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.6004,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.6330279632394057,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.6319,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.7395473460242206,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.6076,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.6073251740680411,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.6205,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.6731154144026046,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.6262,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6365971551559625,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.6186,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.6199268944427807,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.6015,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.7303878715785794,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.6537,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.7011744550912398,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.594,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.6697041085684845,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.6419,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.8483987680524742,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.6427,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.5838423133097708,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.6522,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.7099364351516575,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.6069,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.7248446926493185,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5928,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7359483301023946,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.6256,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.645825826554865,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.6178,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.6942583012527732,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.6525,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.5844140291677369,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5974,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6037101987574003,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5924,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.5809636271883194,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.6147,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.6523709836038515,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.6187,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.6541762788083555,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.6553,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.6489442162609477,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5953,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.6972236344300279,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.6883,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.6513347002957456,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.6153,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.6528359774142238,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5763,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.6381692967469893,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.6158,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.7988163077387223,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.6414,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.6586840779782565,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.6381,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.6872598058714526,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.6364,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.7019750624977229,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.6273,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.6897599321737895,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.6691,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.6273228770324777,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5765,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.6188335397388346,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.6003,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.7171009129855184,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.648,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.5903405154854325,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.6442,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.6728713920994887,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.6202,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5608366052100587,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5683,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.6846587464557798,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.6106,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.5810282753732989,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5787,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.6726750366055704,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.6097,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.5998324944658392,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.6445,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.6327869919152033,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.6231,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.6547872922681908,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5705,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.6336643639317704,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.6212,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.7313880590231266,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.6728,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.6525177564043597,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.6058,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.6699323929827276,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.647,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.6656463772649631,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.636,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.6947010755361812,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.6348,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.6312928724277762,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.6077,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.6276836637117998,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.6168,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.6729171671349553,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.6306,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.5928882976531602,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.6214,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.6302648559556353,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.6024,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.6029102059559185,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.648,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.5929456506725395,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.6469,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.7302499674936984,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.6353,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.6767090260834007,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.6101,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.6326686002104274,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.6179,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.7895500638140943,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.6529,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.619714782011584,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.6244,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.8150135936492813,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.6016,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.8056262936856612,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.634,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.6926875084356027,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.6021,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.933511448249615,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.6018,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.8292691482814444,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.6744,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.6741959456624466,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.637,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.656334472813528,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.6213,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.6802834133054343,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.6231,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.7088982779067079,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5976,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.6927228219958538,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.6703,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.5494232903267938,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5951,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.6487369819844014,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.642,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.5882624149853174,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.653,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.683490025598892,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.589,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.6199102179915539,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.6137,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.6444725882030607,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.6158,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.6610837374219127,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5697,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.6326985039477088,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.6318,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.5352114543983713,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5841,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.708287557569178,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.6226,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.5831346119164322,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.6048,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.6553789678233447,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.6753,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.7033656341014796,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.61,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.5722987693517009,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5971,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.645888588064559,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.6101,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.6131927419221173,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.6067,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.5946125145607014,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.6581,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6499763781038285,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.6288,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.5366645089337143,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5695,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.661467928425322,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.6135,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.586746260545769,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.6143,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6825909793426048,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.6277,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.5685589096212788,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.6569,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.5938060820089538,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.6103,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.6360164615765838,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.6281,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.6201373805604672,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.6124,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.536070297497024,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.6089,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.7538521627050988,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.6581,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5253413584962258,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5926,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.6666777590821967,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.6111,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.5934176939972035,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5912,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.5585784971374723,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.6063,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.6492030801399954,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.628,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.5846265413724384,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.6457,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.572504875566286,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.624,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.6243768105519168,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5958,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.6081808847153175,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.6335,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.6170426171288413,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5942,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.5592407613450462,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.6275,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.733440834962309,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6198,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.6033839945148659,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.6032,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.7365545430474986,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.6604,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.8134779068456764,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4104,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 1.143064897489079,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.4142,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.7672633206145915,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3891,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.9272200905909858,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.4906,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.6901750016060436,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3797,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.8384088352331504,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.4349,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.6748402131871521,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4027,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.620445547481274,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3807,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.7135588967943366,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4235,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.735039684889619,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.4519,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.6736185891597961,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.4031,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.7261716023432039,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.4062,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.6486911466064279,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3984,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.7262521300510448,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.4286,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.6263009820446667,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.4091,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.6858098885429561,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.4465,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6016926676522211,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3958,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.6314729737957672,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.4171,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.5178206041855338,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3751,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.6397340137138173,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.4109,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.5756819897668275,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4022,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.5651796161717068,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3905,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.6118797129370122,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.4339,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.5940067806168899,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.4128,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.523038241541841,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3884,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.5679935192627836,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4301,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.5555138557072326,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.4032,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.6188411037518908,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.446,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.5352148362851606,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.4281,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.5829769353683039,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.4272,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5163085802788568,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3886,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.6339105080220919,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.4283,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.5553987765034081,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3877,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.6190939959409746,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.4326,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.555590240779408,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3695,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.6000115863325964,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.4281,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.5374451346224902,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3781,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.5661709304283356,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4122,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.5965712892153815,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4411,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.49469083393432317,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3554,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.5312565554803539,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.4263,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.5161160010843614,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.4224,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.5086666893408621,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.4177,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.5183434900507473,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3848,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.6166260644931453,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4582,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.5432725975332365,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.427,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.5870618614696029,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.4079,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.5737917592882913,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3705,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.5829507949452502,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.4227,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.4926236253452163,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3608,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.5804104828247548,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.4141,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.5065820060362554,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3862,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.6559873322714755,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4339,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.5332793217175388,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.4338,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.6074452222833079,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.4287,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.46660075353615404,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3766,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.5001122052423655,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.392,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.5208236863375184,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.4314,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.4944156831427026,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3768,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.5499764888373989,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.4345,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.5331526293567236,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.4123,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.629625334637813,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.4402,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.5690961900965849,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3844,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.6564749949997635,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.4286,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.5743020654925322,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.4071,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.533030723562519,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3923,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5972002686921191,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.4426,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.5450367523770023,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.4104,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5339488099312735,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4153,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.5496743263408388,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.44,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5348915210289839,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.4245,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.5285420610816122,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3945,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5294812287208459,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.4358,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.5976370165486449,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.394,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.49919044171953125,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.4113,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.5405817416424304,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.4139,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.4861788021106842,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3817,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.5454885945468942,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4725,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.5082604375314912,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3917,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.541305233782534,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.421,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5622050492443087,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.4504,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.5092854549009826,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3567,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.5102245749806387,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.4301,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.5712006243897882,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.442,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.5240302986527484,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.422,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.49993444002356763,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.4097,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.536549880869419,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3997,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5279060592695372,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4525,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.5026879881112578,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.4018,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.48255524726296506,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.4211,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.5220096531560234,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.4286,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.5799652615557479,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.4217,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.5152224972493584,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.5458861795048068,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.4299,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.488985507449563,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3917,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5119488716282949,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.4188,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.5408837371323314,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.4193,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.49490111778419765,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.3949,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.5232205356327125,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4251,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.5186653542122266,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3994,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.5219547454829678,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4213,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.5201658771488353,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.4189,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.5118883738988748,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.442,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.5565922984549527,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4258,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.46882319008560086,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3626,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.5859997832467789,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.4356,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.5392048773221307,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4064,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.5536568178993265,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.4546,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.5514943871571197,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.4433,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.5114799838580844,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.4333,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.5171862088503761,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4078,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.5007007548684758,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3955,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5488119325498954,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.4627,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.5003157444387085,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3998,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.5579378045903332,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4163,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.5359218801354032,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4182,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.5131148013094297,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4009,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.5296441185915259,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.4061,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.539309524902357,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.4417,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5200626356010721,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.4298,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.49419149668879375,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4159,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.46667663280398874,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3975,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.5737829165197941,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.4349,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5409102432816283,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.4443,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.4929977299845013,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.4049,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.649112205267412,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.4399,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.5173476799134272,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.421,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5565514302572133,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.4234,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.5228361823893994,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4174,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5852092660289774,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.4433,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.471648440013087,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4223,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.5231138640082084,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3931,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.5442474398976133,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.4412,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.5623522544217104,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4142,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.594594489859165,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.4451,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.5814220267648719,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.4031,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.5299728931858083,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3941,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.5700686078223307,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.4229,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.5312644447090885,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3998,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.5673582381369815,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.458,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5858579622770653,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.4621,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.4527056496289146,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.3579,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.5526134407713108,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4418,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.5355085861861225,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.4056,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.5025413176062186,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4041,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.5552969646834052,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.4198,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.5598183842315616,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.4424,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.5078786660744198,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3953,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.4737097801105897,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.397,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.5425443994449175,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.4196,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.5011059111182591,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4417,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.49196817707435353,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.4847443761028937,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3926,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.538629263736856,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.4219,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.5194489368329018,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.4216,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.5368542343595832,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.4338,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.6041296212675368,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.4032,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.5135589946585969,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3902,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.5939192250034927,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.4479,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.514413342593351,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3965,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.5378818147334692,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.4265,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.5786595147944332,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4273,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.5482527784153487,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.4313,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.5319788047065682,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3792,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.5264402258749366,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.4328,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.539456162147961,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4219,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.4971249762532765,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.4191,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.5181691348068654,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3999,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5011083699856588,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.4161,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.496253520768842,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3914,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.4871318907196125,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4052,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.5452436226851554,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.4573,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.45124390374392276,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3802,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.514830703633679,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.417,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.5267771048360981,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3961,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.5338204491650221,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.4441,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.4877788893021744,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4186,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.4765740527768756,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.4184,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.5391843343836331,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.4679,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.5481296728267692,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3822,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.48353485631758764,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.4191,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.5118683744421533,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3905,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.5362953894218593,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.4209,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.48208984549904893,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.4196,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.5299664172063132,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.391,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.553668333845682,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4496,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.5038849534950921,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3841,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.5408453584480094,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.4064,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.5617358131629594,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4336,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.5881712325232057,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.4107,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.5434400943548842,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.4429,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.49614026731921307,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.4013,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.5395686067381272,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.406,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.476889910855722,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3979,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.5287888148220135,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4188,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.516103583884159,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4262,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.5163325767188366,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3843,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.5288189694019781,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3912,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.5531142005391482,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.4293,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.582937542050109,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.4438,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.5336924498010392,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4375,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.4964400920508137,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3611,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.6207621254887205,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4539,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.5172097398870653,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.4064,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.5208574287880444,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.372,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.5553563595472188,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.4571,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.5362801962247794,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.4416,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.5043074935299112,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3999,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.5347590342168678,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.4042,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.5183060559141445,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.4392,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.5259511747794424,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.4305,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.5308150880312807,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.4029,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.488293258000833,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.3785,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.4885054099696677,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.4207,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.5495688286364683,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.452,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.468912796948434,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.385,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5051848318469881,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.4089,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.5077583194049955,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.407,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.5184576124519298,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4168,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.5112719530619944,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.4481,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.47618161552546456,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3682,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.5278967396909993,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.4495,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.5242228810704334,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4547,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.4745209867932393,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3967,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.5031883191479056,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.4097,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.48403160366678805,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.411,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.47630814177678915,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.409,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.5171303219302887,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.4374,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.4987725583705368,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.387,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.509347999308542,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.4248,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.5347016998229156,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.4416,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.4683155835588839,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.4092,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.5789403743736475,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.4456,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.4448480054070412,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3775,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.5396924841754495,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.4173,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.5213950299923608,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.4202,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.5105503719991319,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.4137,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.5215673636563759,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.4199,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.4735542995508707,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3564,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.5240804552195562,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.4578,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.5234424203024152,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.4359,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.45162000873673314,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4006,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.47363602724241644,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.4154,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.4991761883150242,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.4171,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.4942206447412886,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4176,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.48816994183324974,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.3972,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.6151391894880769,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.4358,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.6508777264092559,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.2253,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.5136481443407273,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2551,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.6790244570162128,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2457,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.9271979678551159,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2762,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.5393147865328538,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2117,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.6619935154445342,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.2566,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.6080394709684988,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2521,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.5728730823061365,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2789,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.5529514423787775,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2348,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.6109595129415588,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2878,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.4803220639691369,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2286,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.5225972477888224,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.2471,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.5759965988428132,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.2708,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.4543257034994118,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2297,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.5361210724264404,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2381,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.49604973356191534,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2744,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.4724228744432251,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.263,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.43272699258861236,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.212,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.48969228212593807,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2435,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.44016696474940775,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2467,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.5251575528026733,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2555,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.4559453934883764,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.2333,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.4650435204159985,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.2442,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.4432450083613612,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2292,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.4863324347110487,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.2662,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.42660509785651646,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2436,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.4395303676726666,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.2207,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.5041299647584281,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.251,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.4512744474394398,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2564,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.4448065314321254,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2638,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.4285858692096865,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2359,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.4390965713133496,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.224,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.46437948861374956,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2415,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.42472426485495085,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2215,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.46034139402691654,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2599,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.5094153282902664,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2456,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.4469380163293908,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2384,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.45375322935235624,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2692,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.42412446955342853,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2356,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.45934913148461776,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2355,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.448980053503998,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2356,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.4194341244981012,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1958,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.4951820568265651,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2555,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.4347232630912599,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.207,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.47497740778157005,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2445,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.4764788096524863,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2595,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.4549698855193567,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2548,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.47177780074180886,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.2393,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.4520198715623824,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2227,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.4690597759786336,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2582,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.48342719574353477,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2397,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.4682010712314471,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2273,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.5162955986474339,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.254,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.43080359516113215,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2057,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.483949632017652,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2547,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.46680953249729107,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.2516,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.4631039297928389,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.2312,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.44331903527936184,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2361,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.4473447758832798,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2405,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.4850954587991175,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2789,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.47774118671888904,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2596,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.4671848406115577,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2408,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.43163657789416143,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2207,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.4572812515868304,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2362,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.48678961018632344,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2492,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.47348972927336425,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.2466,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.4313404929220938,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.2307,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.44515774460917845,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.249,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.4806025787921906,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2502,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.46171005275444094,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2392,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.43289391565343965,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.2496,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.5289236923701588,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2721,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.4327801409177784,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2173,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.46019598255682037,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2545,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.4865958169487888,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.2484,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.4381028346600458,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.226,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.4814591174154417,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2775,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.4626923086024724,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2342,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.44546167598181585,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2291,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.4467859015073777,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2265,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.44873349474960944,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2429,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.42307665801957395,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.2054,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.4793915339608981,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2847,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.4687140902327769,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2595,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.42557688836096186,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2261,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.4432251545516164,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2332,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.46273324413126965,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2554,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.45763694555469203,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2631,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.4662022737053176,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2469,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.4233431627834491,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.207,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.48476592690363274,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2752,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.4324913718063471,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2209,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.47008812012102974,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.243,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.40102178750464734,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.232,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.4305940871577585,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2387,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.44440312612377797,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.247,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.4754391154668054,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2489,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.42867550428873735,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2077,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.4249829279784103,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2175,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.5176086899421064,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2845,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.3992759843598576,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.18,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.47565775643376734,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2836,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.41231679790259834,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2175,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.44111950097272995,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.2196,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.4662217172735319,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.2481,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.44346282138861187,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2644,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.4608145495455611,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.2482,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.4068247948966404,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2121,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.4377409650184086,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2379,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.4466125307177132,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.2348,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.46412521380395844,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.2664,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.44495058002642496,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2301,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.4205217807941496,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2059,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.43587938326642534,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2791,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.43364047998202787,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2409,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.4503632251043996,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2328,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.4481693820674268,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2275,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.42299060535801836,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.2384,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.490514031592233,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2484,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.4529895979571549,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2448,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.42094371310879636,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2091,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.4483882367891757,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.2387,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.4405940166302681,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2578,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.4377843605577025,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2404,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.4098274621807884,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2198,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.4663236754566236,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.24,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.4614167236839473,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.256,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.44493317530130494,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.2341,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.4427651369430613,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2194,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.4902446125794476,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.264,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.40746705894329427,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2126,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.41655980636433176,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2398,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.45344040774571287,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2301,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.46559006714568196,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2373,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.4400913020538884,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.246,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.45208645376260675,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2571,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.4437765419959321,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2405,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.42535205267858023,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2488,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.43492805342791746,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2568,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.43864703684991213,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.2442,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.4091639955378538,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2255,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.467531594172494,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2837,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.4069343830458571,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2199,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.442909095046006,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2347,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.46871051665028823,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2626,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.42184834582550723,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2082,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.4322583727906099,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2631,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.44651161873719214,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.226,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.4496483594765702,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2382,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.4564249249585013,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2791,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.3930149283485265,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2139,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.44416974846042806,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2643,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.42305095169709395,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2585,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.40194163386078813,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.2439,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.44714412122375474,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.2764,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.4061245867801932,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.2435,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.4265696593449345,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2551,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.43370884702686435,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2277,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.4312307477925504,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2485,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.430479849799944,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2317,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.46742672835305854,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2721,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.4331601547447723,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.2345,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.4150230443965993,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2416,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.4155331527819223,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.2609,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.4168886746715146,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.2448,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.42561807617172426,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2841,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.4418857055954794,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.2473,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.43533680233872235,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.2461,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.4484662228029884,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2504,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.4071531541296904,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2187,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.4149428641539358,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2066,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.44688967393754003,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.244,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.4504037522101513,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2574,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.45076481271859076,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.236,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.4190146795508807,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1909,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.4596136490315033,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2575,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.4118354325950842,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2415,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.42960016956685465,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.2329,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.4315026897048354,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2298,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.44171817297068994,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2573,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.38853617042273575,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.228,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.4041542780560289,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2227,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.45670090196401475,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2761,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.40231731818137434,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2288,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.46566466787850885,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.269,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.40922001512896994,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.2329,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.4206480170542346,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2333,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.43329674929052575,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.2354,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.4573640551912311,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2313,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.4043513499169192,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.2023,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.4698761561650147,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2731,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.41971670478931494,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2178,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.4232964619352708,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.2306,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.43149133572204595,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2685,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.41276695263789415,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2252,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.4262235108073172,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2394,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.407461247155133,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2317,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.4413815173398027,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2469,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.42663072722076584,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2231,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.41449389004073495,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.2387,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.4436426668024802,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.2529,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.42079907000684397,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2153,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.4647905145017502,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2685,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.4230365855095609,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2202,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.397557921476997,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.2273,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.43429543007556365,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.2522,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.44457413930087203,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.227,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.4301099754300153,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.2374,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.44305885642314435,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.2331,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.4873342616169279,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.266,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.40591410089920665,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2221,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.40747362769396184,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.2406,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.43356396719478457,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.2431,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.4368393681114531,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2519,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.3911243513978015,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2286,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.44710176970188564,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.2749,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.3506779809074581,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1894,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.44815526043657006,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2752,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.3936274724734199,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2058,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.4711758399296064,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2789,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.375353326084472,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2058,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.41475284179474653,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.247,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.4239877236671172,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2186,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.45011280289320527,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2533,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.42179946607360497,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.2227,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.42405644550592114,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.2361,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.42655606321257145,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.2405,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.42166919357710025,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.2301,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.4047082036160849,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2257,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.4148021910135289,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2316,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.45610964775340673,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.2696,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.4108120570093928,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.2459,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.42386114723913515,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2498,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.40112172881977876,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2295,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.40293753346403544,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.2175,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.4154993006585934,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2205,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.41902564159362526,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.2086,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.4273958048850727,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2371,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.4171974739450479,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2499,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.4205864082666021,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2218,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.4652806676056634,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2962,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.4172513708731911,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2232,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.4197887309981835,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2348,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.42761927652189813,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.2499,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.4181938786185078,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2364,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.5508503142665926,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2823,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.44693346269089856,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.193,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.6351310799350247,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1601,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.5284529283378487,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1395,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.4865192633412016,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1643,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.3209179169405968,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1125,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.4139454148634718,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1627,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.4530467250842292,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1485,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.5877926619023143,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1492,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.6586899888477941,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.169,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.5210497396352782,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1404,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.4187220329515822,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1407,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.4043383844987144,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1378,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.41089763438383975,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1666,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.36747913641882757,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1492,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.42395167577273474,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.159,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.3807009005633751,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1266,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.39133222467831946,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1375,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.3682267668987938,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.1318,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.4123655183374743,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1654,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.3580405466295013,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1389,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.3865841361298139,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1609,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.358660406577196,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1234,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.40175575227356386,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1496,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.352720409268914,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1149,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.4182423520484852,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1533,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.40249108068551503,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1522,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.3785090256816242,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1277,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.3684622179054394,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1295,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.38609136347559975,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1637,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.3649015205856066,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1449,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.3566704386658606,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.135,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.3698752703058411,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1463,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.34994684772854856,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1249,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.35484492346195284,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.1395,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.4070903795494035,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.199,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.3886245453790563,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1772,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.3248430668675373,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1214,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.38055917124030164,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1482,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.31815924718442734,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1119,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.38636040166073016,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1708,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.3763036616413443,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1447,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.37242144483129885,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1508,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.37425155263331666,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1455,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.34183748925698,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1269,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.3700944405247611,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1635,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.3717710477868046,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1656,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.34822505214183347,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1602,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.35553266434416536,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1433,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.34525879669228565,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1343,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.3769859689822653,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1581,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.35381561002390166,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1539,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.3307671581922552,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1161,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.3598545619573882,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1435,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.3652673135410258,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.1636,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.3390213849989495,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1227,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.3566955141832025,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1465,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.4143664219891953,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1911,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.33442358209103534,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1278,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.3510208145327155,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1405,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.35746378605950985,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1328,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.38400196891097355,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1469,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.3632589163077723,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1381,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.3236002083750188,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1145,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.3907240268972469,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1664,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.334214694105105,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1392,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.354948101752173,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1476,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.35370379321654827,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1425,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.3410118960596152,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1398,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.32989500855115905,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.1261,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.36277449708344955,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1563,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.38322150235126795,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1662,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.32009453537284766,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1151,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.3535669651949472,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1561,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.36105681772476717,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1524,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.3610255173365187,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1189,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.3552892442345394,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1276,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.3392910967175238,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1338,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.3638539878625947,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1612,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.3395613491217787,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1498,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.36096106047074517,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1594,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.35623097822099103,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1303,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.363997808191153,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.1522,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.35718230690169206,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1445,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.3639029686836542,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1361,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.37104892187976574,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1573,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.3266478534222791,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1313,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.3836816883624995,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1414,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.35538176876116806,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1273,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.339513489103794,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1464,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.35507524606690427,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1733,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.3379342042269702,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1344,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.3697065667485444,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1478,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.3345383473697448,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.1389,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.34445548181589913,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1322,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.37944322556486376,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1674,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.3448115725189526,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1265,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.3616978761386189,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1452,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.3376667454335928,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1352,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.3559556188255116,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.147,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.3446353509584049,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1367,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.347666962915864,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1433,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.3740111443410206,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1512,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.34451619745605333,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1263,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.3749848145542104,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1405,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.3549262989252211,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1235,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.3555333546215091,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1337,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.35020747562089266,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1336,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.3522938934175831,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1273,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.3761595077984321,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1795,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.33512635209503794,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1328,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.3648765927124718,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1517,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.337609385682456,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1428,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.37742466841728367,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1666,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.36517191702239593,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.1527,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.3222087014943062,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1244,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.3727554836773707,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1494,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.3533574272468748,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1579,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.3380494412105825,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1386,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.36665134470122024,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1723,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.3641016830807905,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1582,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.35052775392790325,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.1452,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.3599172889715528,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1385,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.36138343939706324,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.168,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.3164830833766231,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1249,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.3807655988349806,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1625,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.3676886245366678,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.145,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.35275630930473756,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1352,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.36121944667589034,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.15,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.3379134852117626,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1232,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.3800539464302786,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1677,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.3438956242200092,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1333,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.3618197346404032,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1424,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.36362262948639623,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1481,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.3422461343118305,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.129,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.3765065913156621,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.168,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.34248580323514455,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1432,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.3395746747834063,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.13,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.3693052847138753,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1497,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.34589669182233973,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.1543,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.36149101670886324,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1452,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.3502834060407509,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1312,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.37848384011735986,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1492,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.3585534706403727,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1323,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.35042205623100614,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1238,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.385935095209931,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1739,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.3476183921541209,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1343,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.3814540095879333,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1585,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.34864530411904593,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1346,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.340919114574007,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1437,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.35055699902675425,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1369,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.3295657097667827,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1357,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.3484394415132693,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1417,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.3678740051649495,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1536,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.37539534585019774,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1654,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.35444390975053247,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1395,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.3550078063738749,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1431,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.34720629403921227,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1405,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.367160923809964,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1425,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.34564017835449834,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1216,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.359873550692203,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1517,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.36712541919264996,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1612,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.34383086329187507,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1439,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.36599261847493325,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1806,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.34497742147153454,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1377,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.36647119039316817,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1515,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.3755934466447417,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.158,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.3557481849799859,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.133,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.36217992541672855,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1302,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.3546422233934235,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1364,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.3523870516746194,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1317,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.37151594331881915,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.1546,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.35222163589905015,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1526,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.34245007998267984,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.1404,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.3278304315187293,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1209,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.3594705077161495,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1506,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.3543163061782126,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.1375,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.33886542286992977,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1221,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.3592197570343253,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.1421,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.35257619972892845,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1442,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.3567696953332879,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1347,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.3526982016798877,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1374,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.3319641533736129,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1271,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.3648196736974118,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1576,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.36670736366660067,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1662,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.3864652966547624,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1717,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.34504180657617717,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1379,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.3499249003538114,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.1473,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.3674156015805493,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1438,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.3578678427940451,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1377,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.37232036479656727,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1334,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.36941854247730366,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1578,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.33741420747783807,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1361,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.34962282908094505,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.135,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.3606408332344441,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1277,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.35596968095791004,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1355,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.3752549540952727,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.146,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.32875958799374266,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1045,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.3869655074975881,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1683,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.3692214883605167,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1616,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.33517300090861474,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1393,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.3380642714139381,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1572,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.3637077844493927,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1473,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.34177908284329,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1295,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.3840933936348438,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1832,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.3737060201789634,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1616,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.3085916925815364,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1162,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.36356038219334585,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1601,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.34880735481492114,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1515,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.3069257893889818,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1161,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.38469370679494735,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1779,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.31564896099116513,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1158,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.3567401517897065,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1451,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.3704104955642555,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1613,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.33989684484546234,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1281,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.36866531925979673,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1427,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.3265974358721464,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1211,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.35772717675892285,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1237,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.36898341516139854,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1556,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.3829042466600712,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1547,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.3372635158697827,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1111,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.3311599434931517,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1283,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.3694923246066739,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1527,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.3628462337644469,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1379,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.3316200267337368,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1315,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.3560847467160964,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.155,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.34272463389320973,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1413,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.3483895191341102,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.141,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.34833223014152254,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1328,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.33460772336758937,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1254,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.3806470629386757,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1616,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.3455919799137918,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1321,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.33094271319963214,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1329,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.3742395948752932,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1686,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.35736850322472585,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1415,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.34449627524620074,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1406,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.35172272508418634,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1335,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.3517438889934267,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1332,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.3560299459989392,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1465,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.35246076512383634,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1523,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.3450065318681402,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1482,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.3621343713551511,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1379,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.3400582039929114,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1253,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.35937144079689215,
+      "learning_rate": 0.0,
+      "loss": 0.1536,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 1.4024469661131735e+18,
+      "train_loss": 0.4511362501579087,
+      "train_runtime": 72052.3383,
+      "train_samples_per_second": 2.193,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4024469661131735e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}