diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.988354430379747,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 9.522793865885452,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.4338,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 9.954968266958081,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.4454,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 9.543963517680899,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.413,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 9.487663089908683,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.4325,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 8.602463651608858,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.3926,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 8.31951259174041,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.394,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 6.700230260113951,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.3322,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 6.110582103536716,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.3321,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.6864007288947276,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.2765,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.454948454301878,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.2826,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.1878317626962174,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.2349,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.4291997901248426,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.2332,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.275743224437562,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.2076,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.8456428089798647,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.1712,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.000766386916943,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.146,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.4038061561132382,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.1055,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.925718352603025,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 1.089,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.6062810414571915,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 1.1078,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.157339586130706,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 1.0841,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.6965637598446193,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 1.0491,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.4296046534350582,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 1.0345,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.3200270652685662,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.9981,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.5389307051076684,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 1.0537,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.4303579961055366,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.9907,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.1070446323881071,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 1.0359,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1535774898081708,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.996,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.1403413546150567,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 1.0316,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.139097720325244,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.9758,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.0708965405664412,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.9844,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 1.049236197398954,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 1.0025,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 1.1273188660151372,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.9896,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 1.0147608218481579,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.9597,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 1.0434505787473727,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.9664,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.935254845673156,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.9668,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 1.0908366307136146,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.9756,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.9187293034193043,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.9526,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 1.0603606975279183,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.9683,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 1.0083393514266663,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.9622,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.9111486037027149,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.9312,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 1.0289648230479582,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.9519,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 1.0007982026847386,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.9326,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.9790274031811389,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.9537,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.9638573042595157,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.9689,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 1.1250763073507437,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.9154,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 1.0539435148326521,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.9482,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.9444600474335155,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.9245,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 1.1024588087799267,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.9558,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.9680019181076955,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.9341,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 1.014977719586587,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.9404,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 1.0459724988619497,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.9386,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 1.0006353739450624,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.9508,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.8911857862555916,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.9564,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.970791622047499,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.9028,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 1.0209165528473947,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.9261,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8841118308161461,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.9565,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 1.1789914552526835,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.9553,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.9518421841251649,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.8914,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8929711047284821,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.9185,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 1.1034612170745561,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.9274,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.9574723314386061,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.9389,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 1.0316790021582332,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.9229,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 1.0655629649378127,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.9202,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.912648048249862,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.9054,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 1.0804041218797467,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.9499,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 1.1076927038120798,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.9172,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 1.0372826002320472,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.9569,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.0201524850159749,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.9529,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.9218887072600757,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.8888,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.0461807482179044,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.9341,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.9868434596309931,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.9523,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.0115002956084602,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.9426,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.939660188430469,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.9082,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.1073455066431488,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.915,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.095321115153,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.965,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.0969651959871614,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.9354,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.0587435193578933,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.897,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.2253953373122013,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.9224,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.9547272729570855,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.9176,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8985487133724208,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.8836,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.9001754755868541,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.9166,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.9369662061999107,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.9258,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.9488679154151252,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.9556,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9734709764289644,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.9104,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.911500138900667,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.8865,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.9190891471119207,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.9285,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.01028305188183,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.9293,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.9947667033813364,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.9256,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.0446354871788694,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.9214,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.0042539412481857,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.908,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.0196420781361686,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.9236,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.1065772904147513,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.9489,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.1894781798961729,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.9426,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.9594273383031484,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.9109,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.1387580439111,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.9053,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.9472662703033254,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.9238,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.0422119531546734,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.9162,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.956121965911761,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.9537,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.8975114987920162,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.9427,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.9249250750605167,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.9267,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.9634333334326542,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.9232,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.081919501503727,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.9455,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.0136686034736644,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.9264,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.0594066598210228,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.9412,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.088852597768854,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.9499,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.1055713846545554,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.9406,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.2089304018943394,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.9316,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.1223048082885914,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.9364,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.0290597123907403,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.9162,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.1002613669882022,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.9269,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.9852706041296636,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.9229,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.9883583055504231,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.9215,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.88122666970455,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.8968,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.9888665391732487,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.886,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.1468871255542914,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.9422,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.8984036631540263,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.9376,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.080033254064317,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.9071,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.8324094055696049,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.9303,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.0314486726085883,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.9393,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.9302744118803173,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.9192,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.9187856130769025,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.9362,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.9380750494469879,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.9539,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.9061738829528418,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.9219,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.9422790933499654,
+      "learning_rate": 4e-05,
+      "loss": 0.9224,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.9203848026591358,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.9517,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.031275966105792,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.9155,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.9328924931219537,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.9314,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.001152865916292,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.9342,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.0669351885341787,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.9423,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.915456899861613,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.919,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.0671020406167115,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.9153,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.0050580350815828,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.9564,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.034283750647708,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.9461,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.9108421893227445,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.9282,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.9006007061193769,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.9082,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.9173331335809053,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.9395,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.8597448538567423,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.9599,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.9173592934858192,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.9369,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.9169662365544147,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.9362,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.9175758767705322,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.9312,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.9771505484899282,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.9084,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.8578272657816676,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.9136,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.2335976925071264,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.9452,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.8885001457493569,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.9087,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.093549967663553,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.9265,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.8950732295008974,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.9248,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.8820168792772981,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.9422,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.9850124003656275,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.9486,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.8765803967945436,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.9452,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.0284222494901285,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.9372,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.8847525849663535,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.9618,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.8614764008783684,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.9202,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8048401901406104,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.9107,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.8264142480050107,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.9588,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.8129796423565366,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.9157,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.8301726635564046,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.9084,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.8895025622237099,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.9314,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.8169344495749649,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.9347,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8851630776959942,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.9282,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.8345084297306354,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.9493,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.8340493722601956,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.9543,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.8406526301299251,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.9561,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.8083984203510459,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.9328,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.8818691666157499,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.9079,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.8764653370875145,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.9106,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.8845919918862333,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.9467,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.9329758619293751,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.9051,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.8080368047448246,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.9352,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 1.0124209920272431,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.9354,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.757680565504912,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.9286,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.026745530735861,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.8981,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.7701482384793056,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.9222,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.1877523886257446,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.9646,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.8150574906897952,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.9101,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.0092171364815725,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.9288,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.8747356944636427,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.9657,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.9487991679782363,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.93,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.9428266307517416,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.9475,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.8643223631817586,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.9282,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.8210661089397101,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.9091,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7890948574106366,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.9301,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.8045023449860054,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.9171,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.8612512719242098,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.9318,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.8008236080651226,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.9273,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.8953294188512502,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.8992,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.8844432299376778,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.917,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 1.0032198613725882,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.9329,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.8556580722538887,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.9173,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.8009107698045406,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.9042,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.7915897080613359,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.9235,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.7971692485349358,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.8983,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.7438609212976683,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.9008,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.7691153668636966,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.9096,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.7394338432297067,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.9191,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.7583675448356398,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.9245,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.8117353637452638,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.9264,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.8288223630879854,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.9449,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.8373842162355399,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.9103,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.7922809270129146,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.9211,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.7979271362496668,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.9391,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.7511214034048337,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.9238,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.8306807813617255,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.8962,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.8138026604197592,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.9385,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.818235671852126,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.9013,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.7860642409801607,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.9149,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.7480015202108641,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.9204,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.8779141128010278,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.928,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.7961401668962209,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.9034,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 1.0366219226571858,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.9127,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.7920670174201848,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.9201,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 1.0168401628952797,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.9141,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.9425341880446795,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.9322,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.856952410270497,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.9386,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.7276329627305252,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.9316,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.8708191611460626,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.9005,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.7899312859468076,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.9254,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.9504043160787543,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.9476,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.9098478978006941,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.9277,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.9144100926607531,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.9358,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.957002281605761,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.9401,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.8335853255659449,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.8802,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.9768662259109788,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.9058,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.8312970522417972,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.9245,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.8066852939420116,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.916,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 1.0083776605795152,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.9277,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.8102818687084192,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.9494,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.8839312325854993,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.9484,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.8963693211892929,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.925,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.7766665722495955,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.9227,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.8614786152369522,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.9483,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.7744192056031474,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.9043,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.8147909084863193,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.9167,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.7800139742685406,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.9442,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.7659870016496725,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.9209,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.716795987871206,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.9081,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.7093750345826135,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.9015,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.773260832927041,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.8984,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7289062135382959,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.9358,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.7611606158002228,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.9449,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.7272276084244625,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.9293,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.7588604692942633,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.896,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.865726588222784,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.9267,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.7682323690342857,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.9045,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.8175216477397975,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.9122,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.7653803653061967,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.9399,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.7322776562525632,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.8973,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.7454335824306456,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.9023,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.8501999859205769,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.7831,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.944511568113067,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.6847,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.8334502017579511,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.6613,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 1.06056270406261,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.6524,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 1.0038205623386387,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.6463,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.8713751116672873,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.6295,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.9505132440299355,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.6504,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.8337249349693735,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.6885,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.8078400181870596,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.6379,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.8896193033316735,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.6324,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.8662228798730747,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.6328,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 62.17563569746746,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.6899,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 1.3338143283563009,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.6535,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 1.2406136601986024,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.6899,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 267.633193012905,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.8138,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 4.324494703736174,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.6335,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 1.6637380208598374,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.6694,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 1.3139841625355388,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.6476,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 1.146172481488102,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.6591,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 1.1207096063857174,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.6403,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 1.023988777225022,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.6372,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 1.3683762141048152,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.6334,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 1.1865026357466744,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.6683,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 1.0130327038445206,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.6393,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 1.2587261136370436,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.6232,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.9933450166325258,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.6279,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 1.0425935436524212,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.6428,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 1.0241548983376014,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.6328,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.9908966452247979,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.6348,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 1.0097514088533817,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.6507,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.8898879424222903,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5956,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.7937609078496091,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.6572,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.8714202086473534,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.6005,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.898535989022503,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.6551,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.8017259643575496,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.6371,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.8086312632475107,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.6442,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.8499636837014157,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.648,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.8490827893712323,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.6179,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.7962044410821669,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.6243,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9958710223430679,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.6492,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.7314238460651716,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.6233,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.8820171343401375,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.6202,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.7665209426779894,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.6197,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.8611594717260614,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.632,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.770126456421651,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.6324,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.8021127805050593,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.6421,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.8370299516058306,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.6391,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.7665964598483105,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.6289,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.8362278717777547,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.6233,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.7966584760906683,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.63,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.8387745338034974,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.6612,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.7870773906970957,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.6302,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.8128987476467681,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.6284,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.7682197099324307,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.6787,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.8297302300864595,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.6461,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.7880244534811804,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.6628,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.7604878060227019,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.6179,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.7860995865859981,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.6447,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.7766316431705708,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.6335,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.8281341921742452,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.6514,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.7849924281011561,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.6405,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.8475454592487781,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.6229,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.7588026368536304,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.6521,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.808302597505416,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.6401,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.7916240989477626,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.6262,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.7916148617012052,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.6428,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.7306017435118396,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.6183,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.7757083146678371,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.6584,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.7936292048096044,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.6462,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.7391591228717714,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.6589,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.8262311469936556,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.6161,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.7435020971149764,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.624,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.7377819693676946,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.6413,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.8166707090963052,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.6219,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.718782158033764,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.6315,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.7258989857060822,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.6554,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.82580313740461,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.648,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.7458451382344617,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.6409,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.7420153619491308,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.6285,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.7718930410989907,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.6595,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.8122260908746037,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.643,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.7516567973126825,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.6253,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.7090807593961819,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.6272,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.8127456671218983,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.6518,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.7841236279793588,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.6391,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.6734983976645627,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.6355,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.8173156826304518,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.6679,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.7031361896218662,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.6057,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.8436649565689535,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.6472,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.753038643223926,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.655,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.7451176234347322,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.6704,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.7421339197682764,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.6443,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.9107074513475483,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.6295,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.7600444393963146,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.6308,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.8580340786676304,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.6137,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.7618412959297477,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.6497,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.7936454762993559,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.6444,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.7768655516899606,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.6141,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.8210133376573814,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.6305,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.7968832456974997,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.667,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.8299967776725332,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.6672,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.778509575524894,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.6455,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.747242965959519,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.6352,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.9553166278951968,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.6547,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.747437053695602,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.6571,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.811098602178545,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.6148,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.7496167347663787,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.6469,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.7173723174024283,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.6183,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.8702723447728832,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.6483,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.7467937070901819,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.625,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.7557072384753494,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.6521,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.8425979160799905,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.6584,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.7864807620390492,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.6439,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.8569158162286893,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.6585,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.8236115931977565,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.6412,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.8029634622025597,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.6513,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.8189320167906478,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.6785,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.7977508853632214,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.6554,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7678239618149032,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.6515,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.8493211142538988,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.6465,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.8213465867985656,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.6315,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.7710812945160903,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.6587,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.774161640878061,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.6487,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.7433293504531457,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.6478,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.823263916533596,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.6559,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.7296143084861777,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.6665,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.7554564064687426,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.6377,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.7673983496921464,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.6555,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.7556721935147896,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.6576,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.6946345883965251,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.6525,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.7141787198730505,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.6559,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.7273775695833182,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.6707,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.6958770931247655,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.6671,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.7205826783403556,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.6305,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.734930406239389,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.6555,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.7179009150171346,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.6254,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.7652050596398978,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.6655,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.7607420249125364,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.6424,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.7037905415122755,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.6544,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.735457438092124,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.6157,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.7087695634194621,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.6185,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.7533970784898371,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.6696,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.7286349535959497,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.6426,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.7074699841158865,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.6652,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.8595038213399285,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.6265,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.7177737013053062,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.6565,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.8947796124246589,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.642,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.8033971100520003,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.6261,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.7929098755747078,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.655,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.74918095119173,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.6686,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.7791972139061412,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.6825,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.7948267914202203,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.6417,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.6628775304262262,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.6667,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.7734443769212686,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.6648,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.7639443942417388,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.667,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.7399280543616883,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.6399,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.7598363702170844,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.6577,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.7437900066317883,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.6625,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.7115192458249039,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.6552,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.742939633726292,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.6332,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.7600374026308377,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.6524,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.7378203500282309,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.6363,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.7292875703044858,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.6711,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.7347744342942394,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.6595,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.6946133456206068,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.6537,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.70806356429155,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.6594,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.7320109807740782,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.6577,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.7362810523172428,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.6326,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.7425529125828465,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.6334,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.755763120766639,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.6516,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.7243940904211422,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.6448,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.8049414068041266,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.6403,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.6998030880752673,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.6375,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.7136520483746102,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.637,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.7144956628129325,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.6515,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.7519771975655811,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.6419,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.7636664916936928,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.6445,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.7088714895147368,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.669,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.7781569823916603,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.6692,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.7200524174265219,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.6191,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.7502442732723971,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.6554,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.7321532058787398,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.6635,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.6952121972654067,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.6346,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.6983361155229868,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.6123,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.7309792302233554,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.6742,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.6816613125624857,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.6343,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.7413673043849408,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.6397,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.7313960408070231,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.6437,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.7257320470819733,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.6631,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.7802741788058793,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.643,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.7091555454327128,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.6637,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.7578712450992019,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.6686,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.7811762289225941,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.6515,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.6971633477446083,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.6514,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.708207335780122,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.6488,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.7352712354425601,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.6453,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.7197424944525731,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.6324,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7247286258066894,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.6345,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.7760386881203201,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.6591,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.7197225861590772,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.6595,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.7684840828621503,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.6361,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.7424476369395686,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.6541,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.7632405654214232,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.684,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.6982538899713869,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.6431,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.7272656046537773,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.67,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.7379527543753095,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.648,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.7647361635120961,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.6522,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.7413732320667049,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.6709,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.7896186541989624,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.6575,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.7588817489521367,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.6607,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.7976457800385535,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.6461,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.750620386671621,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.6481,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.724210773290399,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.6427,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.6870947599969207,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.6796,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.7845795929926744,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.6508,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.7062775396313856,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.6578,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.7182202776635912,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.6664,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.7660971636909207,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.6496,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.6761088901877347,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.6817,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.7682500761098141,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.6419,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.7189707278941833,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.6404,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.7093103451136347,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.6393,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.6985515135641872,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.6237,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.7015424217223998,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.6218,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.7561047351987388,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.642,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.6511119438982005,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.621,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.714661812264617,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.6525,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.7096120934558399,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.6553,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.7071021826998148,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.6562,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.7465308374657943,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.6341,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.6898042387979246,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.6532,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.6769974585681948,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.6317,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.78128422512674,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.648,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.7149985232653898,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.6655,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.736665668461665,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.6854,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.6838701593653577,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.6448,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.6966357114123533,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.6281,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.6833392677731197,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.665,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.6920355032823817,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.6779,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.6538991544774605,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.6353,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.6626424081745341,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.647,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.6779744929050447,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.637,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.7222060828753742,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.6649,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.7085078384608667,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.6195,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.6858737410189053,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.6617,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.7091764975896896,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6487,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.6932277780325159,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.6439,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 1.0730683599232758,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3892,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.6975382287835833,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3894,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 1.4099499950137095,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.3662,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 1.0873209625330638,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3416,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.7374335565110527,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3791,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.8630973697578118,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3649,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.8348747185056536,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3701,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.6757535847740499,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.3683,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.7687576588575437,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.367,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.81497605197791,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.3553,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.7671984551444712,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3569,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.7413841793011706,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3344,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.6895483662822144,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3631,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.7776522039032633,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3875,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.7070655584586982,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3538,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.7849426828440247,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3675,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.6751945720748289,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3575,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.7287210289072203,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3753,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.7049736595022338,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3485,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.6649660776759415,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3484,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.6834852966402408,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3471,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.6764814696951097,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3187,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.6937529614964689,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3796,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.709079981539317,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3381,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.6816701071071045,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3345,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.7365004613790581,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3597,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.669256222830748,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3301,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.691736643930418,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3707,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.7033126270793311,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.3459,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.6500166503110882,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3588,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.6602010727231714,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.3636,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.6763294120652601,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3617,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.6493882747422526,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3649,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.6420604505310624,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3495,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.6307539444057021,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3304,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.6487226795821414,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3843,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.6661877354443125,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3474,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.6328489888987204,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3651,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.7016705029360296,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3324,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.6233812069584074,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3488,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.7341446812430328,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3731,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.6457795754995119,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3816,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.6909793091878407,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3539,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.7241962959392354,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3783,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.6444033657638327,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3607,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.6828104538671392,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.34,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.643874730796033,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3346,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.6300946222691465,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3386,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.6715293951122352,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3584,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.6612551928989853,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.3631,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.6354622493067359,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3392,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.6472928232516163,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.356,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.6347095921204686,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3515,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.6455604149401292,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.3652,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.6407172581456356,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3607,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.6582408469134832,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3453,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.6390838844244807,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3572,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.635057575844083,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3433,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.6353880898163591,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3415,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.6499456168106446,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3379,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.6955438320144688,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3369,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.6205043539159136,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3583,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.6631044505080422,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.3423,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.7029268632172808,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3698,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.6262097751268731,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3561,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.6696595494991935,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3564,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.6493738176219362,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3648,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.6822476803098048,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3708,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.6538482898712497,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3613,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.667754210910952,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3484,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.6668654670283753,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.3669,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.663943495804708,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3383,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.7086326162063326,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3566,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.6231040152305926,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.371,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.6836612246105522,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.358,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.6550565545534498,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3285,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.6512421170621308,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.348,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.6926628149530414,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3466,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.6282471577122621,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3364,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.677288544528382,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3391,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.6276334253676391,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3579,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.6845896137150201,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3599,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.6453468028815765,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3774,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.6803683817974641,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3592,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.6491860054175945,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3627,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.6281441971747822,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3627,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.6876363775140354,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3477,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.6675215767184872,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3261,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.6144652871468264,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3508,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.6572599712189671,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.364,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.6461936256921507,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3532,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.6266428523785978,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3805,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.6550690400754096,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3614,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.6525997747009422,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3556,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.6270310271760353,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.3804,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.6593332323193604,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3649,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.6602200803228495,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3715,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.6419129993840214,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.343,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.648311775722004,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.363,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.650722417843675,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3516,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.6332024376045101,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3466,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.6388626822231003,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.373,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.6228913024040206,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3581,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.6079884914806039,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3597,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.6474069969824097,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3428,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.6347092985211719,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3443,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.6183076437592029,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.368,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.6546266434366506,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.319,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.660732019102172,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3529,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.6385792611861122,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3949,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.6343420110799095,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3551,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.6612128985027377,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3653,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.6363206472029987,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3544,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.6166526050365734,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3784,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.630669680946759,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3518,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.6494277552872839,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3494,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.6837858737213001,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3883,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.6357284861044946,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3468,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.6544090164588285,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3509,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.6968662565764505,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3784,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.6410931496237158,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3472,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.6465904942855569,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3583,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.6386415033171421,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3586,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.632112782481879,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3572,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.644784110789381,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3686,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.6257188851731668,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3592,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.6488080981008636,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3615,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.6685198161577339,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3623,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.6348094306272378,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3725,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.6532804155026938,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3496,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.647091980702466,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3533,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.6542744893338424,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.3348,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.6196670527430439,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3614,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.6293459086477332,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3726,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.6702007441194561,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3632,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.6932023899105668,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3614,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.627689557520693,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3594,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.6378872979754437,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.352,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.6765948112592299,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3558,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.6745433328552332,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3576,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.6284639632989457,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3812,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.6492125446340727,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3223,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.660108682624202,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.3808,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.6725598271713734,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3679,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.6027859546029404,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.365,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.6465603722598038,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.3549,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.6514960399821316,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3354,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.6162471992367229,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3275,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.6505766690303292,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3823,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.6422617043022083,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3394,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.6465164230750736,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3809,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.6412323447351103,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3663,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.6597067590064862,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3539,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.6394919875211622,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3641,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.6790368803005002,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3309,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.6328680861246224,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3458,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.6670840080027968,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.363,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.6483486480958935,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3873,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.6577020981840485,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3567,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.6372597208588325,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3547,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.6717077943049699,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3646,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.5880236286639944,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3608,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.6295534087378519,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3539,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.6421177609038778,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.341,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.6053761542259822,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3449,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.6374393038896184,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3651,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.6310624887468448,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3735,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.6375185867561088,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3744,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.6356090727010483,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3542,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.6264349685203249,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3577,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.6155827660388802,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3424,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.6395714949714628,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.371,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.6273241089037924,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3403,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.6270781887101271,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3483,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.6136901174329159,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3534,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.6517329134625399,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3516,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.6652729348302977,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3735,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.6367992696342348,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3459,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.6286902588453687,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3697,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.6523588730699148,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3601,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.6327160994241645,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3721,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.6300771170331333,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.351,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.6488201859184666,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3442,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.674941192544858,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3574,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.6280330307601654,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3559,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.6314624016994835,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.368,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.6548751829953806,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3403,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.6128953865404532,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3868,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.6100936387032646,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3363,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.6575765085273739,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3327,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.6343799078164272,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.3804,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.6467902060591467,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.3686,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.6388429287350491,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3551,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.6424770919856617,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3329,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.6346901685471357,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3601,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.622352430296181,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3452,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.6315213995042669,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3721,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.6683427978733583,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3415,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.6334098604703104,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3377,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.6499142271621416,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3553,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.6383540841609799,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3862,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.6717337755075155,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3391,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.6348666048487007,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3693,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.607738364466111,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3733,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.6625536196407977,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3487,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.671499966852791,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.3528,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.6191905283393901,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3827,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.6693175204577833,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3472,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.6663876821729235,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3528,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.6291896560952048,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3563,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.6242661814621269,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3583,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.6438806946759316,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3616,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.6259636309855531,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3692,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.6327671642663485,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.3693,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.6455750250524719,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3483,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.6171740321947186,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.3614,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.6508633034396681,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3502,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.6570754283405816,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.3671,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.6125082423695619,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3682,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.662702015210232,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3877,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.6489785071746814,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.363,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.6093231642159194,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3603,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.6316102707416197,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3654,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.6038505652671566,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3588,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.598307575301431,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3433,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.609452124097441,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3598,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.6188274421253823,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3575,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.638332147114857,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3878,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.6070826871594311,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3584,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.6300084528474881,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3502,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.6174249486348022,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3324,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.6252229641124424,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3621,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.6415782402824644,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3544,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.6112548512235375,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3218,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.6197622667499262,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3797,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.5989344705897799,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3533,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.6025968124101745,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3242,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.5881968809680417,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3832,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.6133771196521159,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3447,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.6030885811786886,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3744,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.6314831365194564,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3708,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.6125727856011062,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3409,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.5981848078803037,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3747,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.6051487598993202,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3569,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.6297457452351652,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3512,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.6255425724225495,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3482,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.6440410794444794,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.2733,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.6599586981669444,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.204,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.5040443883888212,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1955,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.5335016891317016,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1682,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.7626752997444614,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2108,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.7441659501453414,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1773,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.6556052883824863,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1802,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.5582774105012456,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1712,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.5798543170309648,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1706,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.6010349127036706,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.1726,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.6143827777285712,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2182,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.4983255294851266,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1837,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.47581186848741264,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1873,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.516386745044613,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1756,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.5680499759467013,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1648,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.5635221955459114,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1797,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.5199313140662964,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1797,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.5027698412493431,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1904,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.49958081381795577,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1733,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.48381617675101785,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1714,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.47320245804624045,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1818,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.5047890674484371,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1669,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.47660350340078544,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1903,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.467539475513793,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1709,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.49306609522033157,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.18,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.5374885109104625,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1734,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.5039895604866972,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1726,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.5275499059549316,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1968,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.5144624112563518,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1734,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.47523632484486944,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1618,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.47602389807119877,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1875,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.473766595313342,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1695,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.4989134463036094,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1845,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.48209497640511506,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.184,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.4725916740511468,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1973,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.4924552196523981,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1677,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.48819803844136445,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1585,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.4916487713578794,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1698,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.48729674362725595,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1771,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.5002940946668902,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1634,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.50277716080224,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1839,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.48282014972879256,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1734,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.4851051858159114,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.159,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.5153441454338761,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1522,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.47668585312380996,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1745,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.4781694746618437,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1941,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.49625828532314425,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1661,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.511812141674877,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.178,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.46996703733306483,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1426,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.4972789210711588,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1766,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.497835806376275,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1821,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.48312162522102187,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.1681,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.4907034400913169,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1895,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.4836477087832854,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1635,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.4676707484954717,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1767,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.49922119281110705,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1807,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.48570860481378564,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.1796,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.49863547323958446,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1938,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.48110230613267485,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.181,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.4939069866273379,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1634,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.49593152113174843,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1718,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.485383102132259,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1927,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.4819759859629802,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1794,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.4861453295659967,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.164,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.47432280171933583,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.179,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.48279196598390456,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1828,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.49670966757945356,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1698,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.4944443665027258,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1774,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.48495210687208634,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1685,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.47813958259968437,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1803,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.47848303821120375,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1918,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.4864087661192066,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1788,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.4945460625210856,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1717,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.4988651982082074,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.172,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.48501725092136266,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1599,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.47490201709300583,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1816,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.475920463208382,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1541,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.4628216581897864,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1759,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.48374679581745583,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1885,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.5073299445203405,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1719,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.4921073582739204,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1635,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.4765135184335875,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1976,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.4664554051141389,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1697,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.4768450729161539,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1811,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.46808793374048385,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1701,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.4853676167574442,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1362,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.476049844642011,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2031,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.5034070478661351,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1807,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.4732332326081297,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.1814,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.4832153237913954,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1832,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.5011857041841332,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1654,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.48620270274131844,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1728,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.4815516123500849,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1766,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.4747678722609156,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1544,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.46534954010754553,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1879,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.46793060552806137,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.1644,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.48139298333390323,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1653,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.47702743167229433,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.1608,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.49297168920303597,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1758,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.4672555655204312,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1674,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.4660187068551667,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1504,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.48525963006099826,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1582,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.485402628276436,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1898,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.4912528175563379,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.159,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.47244769397716174,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1637,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.49156636302691104,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1797,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.4919451737263143,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1547,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.46871353494345247,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1719,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.487955603548837,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.1854,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.4863107418330933,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1868,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.49110474843043794,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.174,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.48488892828600383,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1627,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.5062625959102857,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1601,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.5004171507167379,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1703,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.4865656297671864,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.171,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.4886127807360618,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1816,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.487295621284465,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1491,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.46694947857238245,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1765,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.471594819489572,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1749,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.48089939799057013,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1553,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.46389545512419894,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1701,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.4930418887232616,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1985,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.48210545617399275,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1423,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.48736213583003457,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1589,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.4871750873190747,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1726,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.5012304421959867,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1698,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.4924729253680271,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.172,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.47440595705722843,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1755,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.47046461968769726,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1625,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.45020743579006023,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1703,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.4896786837033695,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1788,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.4737847886121939,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1641,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.47933272585165193,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1733,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.47593071372643597,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1703,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.45866677127517513,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1866,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.4802320732173607,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1744,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.47825852880072656,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.1601,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.5099480761231938,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1744,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.4863074659350509,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1912,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.492617176743573,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1556,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.502183266622763,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1634,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.4782393988709792,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1631,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.4775339613008044,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1893,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.47538470316920023,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1934,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.4660599393186103,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.178,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.48525966855986663,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.191,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.44580748724096103,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1543,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.4638914316785568,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1554,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.47754531701248343,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1346,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.49421786849030375,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1746,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.4798430427854997,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1724,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.4707134772494206,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1698,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.47197067823235966,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1884,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.4666547101456753,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1623,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.4880360920086553,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1542,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.47135730835828427,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1729,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.4764109007186281,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1504,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.4776686293663782,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1808,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.4803061370480953,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1759,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.47976214581564053,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.1793,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.4807801476540099,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1627,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.45499557224666637,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1683,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.49079664783252724,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1795,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.491825754315027,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.182,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.47916715567146,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1634,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.4586970215461167,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.173,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.46693586330896514,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1633,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.46750168772298367,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1668,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.47246410059535965,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1402,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.47984521562726534,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1647,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.48653513079461363,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1737,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.46577253895077786,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1742,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.4746521259167058,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1474,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.4664866531623272,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1779,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.4774245201097941,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1696,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.47285669593301716,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1531,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.48018854556160645,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1693,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.46112755962579255,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1739,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.4612263575943562,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1491,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.4619336640374711,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.1701,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.4589650475949778,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1774,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.4557197369854214,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1852,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.4693854115207263,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1553,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.463748755897837,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1868,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.45119151882290165,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1797,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.4879518041580158,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1744,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.46927547606121,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1495,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.4718539157126119,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1505,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.5056424051734884,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1696,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.47993347150412685,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1484,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.4771456460971675,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.148,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.4730068280266553,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1776,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.49038039295562214,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1763,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.45721764263078923,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1459,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.4576099737575005,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1797,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.448285173270871,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1414,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.46162623362610133,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1682,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.48261195886193675,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1924,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.4756115159916952,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1969,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.4738652492800104,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1652,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.4671725271449489,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.182,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.4767002099027314,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1676,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.4435792758039588,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1571,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.48434308547689675,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1414,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.4643234508912412,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1733,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.4654914659078776,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1633,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.4639683558893644,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1649,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.44436316590116487,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1692,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.47467677770208094,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1439,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.47223991763500134,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1615,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.48649411600636977,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1876,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.4634403782963731,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1918,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.46091636059949015,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.1827,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.4769967115183238,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1951,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.462473746716882,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1585,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.47758279998305814,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1719,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.4748729675566282,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1791,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.474853858629187,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1678,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.46144408019671773,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1641,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.47492561155427576,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.152,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.4746036837885673,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.188,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.4519633404080866,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.155,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.4501903851264762,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1548,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.4537063219314308,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1731,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.47854963876995416,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1797,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.45363114460727066,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1746,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.44706614763603425,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1529,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.450323036476662,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1912,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.46514072634590653,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1709,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.4694208313550146,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1461,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.465721373031072,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1821,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.45562927343179804,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1836,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.4639900507478122,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1615,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.4845960479196958,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1525,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.46179126188856584,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1598,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.45404966340466185,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1617,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.4486639645407799,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1595,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.4734619663149631,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1844,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.44702447793707306,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1618,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.46677061994160657,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1722,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.4440519105568029,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1685,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.4735298772738196,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.169,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.48225736467911895,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1623,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.4533519957400664,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1628,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.4693560972710053,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1695,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.4690591512567515,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1722,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.45405012562437996,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1662,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.4610170953401511,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.0843,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.3873440716376881,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.0923,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.3474056017257876,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1036,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.3249471947848641,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1017,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.3122985019133751,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.0997,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.3258570119907939,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1028,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.3400354522678034,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1068,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.33474653053044306,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.0965,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.41524335074188323,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1117,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.4096046319430426,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.104,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.4191537023023533,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1048,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.42535361793744325,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1079,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.33970986373619505,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.0786,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.3469142976465309,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1045,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.32518934361273105,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.0895,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.3301556878964722,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1063,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.3295312810287854,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.0897,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.3480737983046676,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1086,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.350928494196185,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.0949,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.28945026223096604,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.0953,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.3133170522848053,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.0981,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.3045881168931987,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1119,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.31873028719212076,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1039,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.30292210036702677,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1136,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.31194308284463473,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1109,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.3223591034976283,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1033,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.3365818841369853,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1113,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.3258117987680397,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1078,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.33313279497351206,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1085,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.3114532649760953,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.0905,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.3179340713132016,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1139,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.3117569402326828,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1012,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.31608502826340346,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1116,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.2824414027528644,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.0887,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.30070805064526385,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.0914,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.3410163991224106,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1232,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.33610903991680036,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1073,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.3106081886551314,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1028,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.30464660082179645,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.103,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.2929013211264937,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1032,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.29903704495736194,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.0834,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.3061092121492965,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.0905,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.29594711471893176,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.0803,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.29703111288030015,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1145,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.29738532125747297,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.0905,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.32912509330949363,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1191,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.3045661389386556,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1089,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.33057916218755073,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.102,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.3033189983471951,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.0962,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.31406084842239396,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1129,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.30067380443338737,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0887,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.30413345909416584,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.106,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.2975127509891277,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.0898,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.3025172660771735,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.0881,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.31444431999975964,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.0833,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.30592602656305645,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1152,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.30044684979252606,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.0978,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.3078181383737722,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.104,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.30234545609600755,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.0817,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.3136760427927726,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.0874,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.2939085919588017,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.0969,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.31092522452686255,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.0983,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.3155515055632975,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.0921,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.29638737358827977,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1104,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.304925165830509,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.0932,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.28824434668793164,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.084,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.3074502678979281,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.0918,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.3174750018185149,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1084,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.31639592341379896,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.101,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.2794158439379412,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.0929,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.32031847137991826,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.0983,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.3070787117795409,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1098,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.3048408712171956,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1158,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.30391377889453897,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.0958,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.3330442351227201,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.0837,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.3091657964810012,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1229,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.30214561266001944,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.0952,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.2916972322652815,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1059,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.3017122170212275,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.0796,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.29863292070971975,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.0878,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.295874110315499,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.0774,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.29663034046816583,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1124,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.30848038727405985,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0852,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.3319669527545349,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.0993,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.3061526406159574,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.0992,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.29859704490226063,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1171,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.3116083600338662,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.103,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.29489908590025754,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0884,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.3203478416616815,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1151,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.3121608514405902,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.0959,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.32029592498370973,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.097,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.31696639196313375,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1088,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.31925778220717316,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.0995,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.3172518872885201,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.0951,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.30706915202712215,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.0723,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.33059146777103643,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.0948,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.30308205549498274,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.0743,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.30284377191945727,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.0967,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.32479097339775603,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.0903,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.2782393631210405,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.0847,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.29778376222360753,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.086,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.30213701558518496,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.114,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.2916638235187322,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.0713,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.3155596076693685,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.099,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.31765673554681845,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1029,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.29527882639825265,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.0787,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.31295195431681067,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1002,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.2961163285289017,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.0835,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.3108115417751281,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1003,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.31601879738488,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.0963,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.31813770055924706,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.0879,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.3140487038571509,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1044,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.3037413393751065,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.0866,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.32976036650033397,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1535,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.31821202178413815,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.0901,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.3093136880359081,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1024,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.30292851537969134,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1058,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.31499060503250415,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.081,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.30885257774950864,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1084,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.32227615403323745,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1213,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.3086430520147436,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.0956,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.3083601404688028,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.1113,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.2918586882228044,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.0733,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.2967431500649152,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.0987,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.28726413167831316,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.0993,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.35209113948398507,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.0898,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.2764199132350892,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.0757,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.30626665369802525,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.0997,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.3128373983081162,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.0989,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.29810437155461705,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.0863,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.3187463336096733,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1109,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.3136927195339577,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.0828,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.3019946095212555,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1154,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.3121189936029247,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.113,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.332799292364458,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1067,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.2829692230487866,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.0852,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.30710512535369555,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1022,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.29788123593438537,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.0931,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.3115110835788177,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1127,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.31985334356311457,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0891,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.31391822601992964,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1046,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.30783224068363635,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.0995,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.27646283542323485,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.0949,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.3201964496842425,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.119,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.31294449406306296,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.0987,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.3045017703420194,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0805,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.31719732874751355,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.118,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.307106427318849,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1098,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.304728379555118,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.0902,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.32068342235416675,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.0893,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.2826740208666823,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.0625,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.30470172017292946,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.0826,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.3196354380554341,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1158,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.30122734278103375,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.0847,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.3157171565363846,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.096,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.2942123386911957,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.0943,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.28816362545892726,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.0788,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.30452548432192356,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.095,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.30240206410573844,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.0899,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.2885413259227978,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.0868,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.31039310054588315,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1227,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.3086480341693738,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.0937,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.30490080570774436,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.102,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.31064117259919366,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.0972,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.31370005582040045,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1012,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.2834329165165301,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.0794,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.3092921102556254,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.095,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.28830595725783675,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.0855,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.3103242528848358,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.0918,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.32063241816923677,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.0958,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.31276594008301506,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1033,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.3357858350098859,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.131,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.30604559366575074,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.0942,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.2885928582936559,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.0798,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.29726272956041083,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.0788,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.30806172237951096,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1039,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.3092620234764918,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.1041,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.32244202496760366,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1026,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.31510713665121165,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.1063,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.30840709210948675,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.0795,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.2981142457646376,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.104,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.3018581450819205,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.0806,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.3065752253057783,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1087,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.2960220863442087,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.0821,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.3114045599764647,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.0857,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.3170817197611973,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.0979,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.3135910067067053,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1012,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.3029749446857747,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0877,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.297581783307933,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.0834,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.3161405690646216,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.0951,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.292947585994772,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1032,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.29193002799911427,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.0715,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.3124774545282383,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.0952,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.32296316117466045,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.0942,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.31814884847123226,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.0913,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.29558210652141725,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.0897,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.30432576822500035,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.0653,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.32123101702668455,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1212,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.31445162809079696,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1034,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.30349868812001113,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.09,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.3090497051449456,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1085,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.29865137200532477,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.104,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.3181044091882685,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.0888,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.30843546625958823,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.0907,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.29574055517644016,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.08,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.3257623985774419,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.0946,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.3174669409756023,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1029,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.295804689872343,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1069,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.3260200803507338,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1011,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.30133255729550007,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.108,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.31571846197585923,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1251,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.2990115402582325,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.0948,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.30133675403194116,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.0897,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.3038920026414039,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.0979,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.31902545442628205,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.0971,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.33205290336820315,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.101,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.30100662459951,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.0879,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.31096672240297074,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.0978,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.2991140804140959,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1019,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.32346777164908974,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1051,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.3177465936191514,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.0917,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.3097846908892458,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.0832,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.3101587825942074,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.0939,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.295282299784791,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.0705,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.32243584340615195,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.0779,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.2936322415222226,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.0829,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.3281839099402308,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1223,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.30109724543710437,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.0961,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.30327763557520226,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1002,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.3226047086191552,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1076,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.3038690007787612,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1124,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.31581861583110354,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.0982,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.30745512019791377,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.099,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.29900088261742086,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.0884,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.3133245839062767,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.0934,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.34586516492844044,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.0948,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.3154401522859063,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.0883,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.29259775488686024,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1241,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.31925635283541415,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.099,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.3122514246339603,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.105,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.29992339193520984,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.0968,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.3146288634672351,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.0971,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.327743541692515,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.0974,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.3022183255235741,
+      "learning_rate": 0.0,
+      "loss": 0.081,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 1.007640171539071e+18,
+      "train_loss": 0.44717758251278383,
+      "train_runtime": 34723.0943,
+      "train_samples_per_second": 4.55,
+      "train_steps_per_second": 0.035
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.007640171539071e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}