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": 8.58853613330509,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.1273,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 8.368071573947034,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.1046,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 8.597952753575159,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.1344,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 8.880403233569728,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.162,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 8.435039255574711,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.0956,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 7.834832910083181,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.1257,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 6.349848538322875,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.062,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 6.235458257814083,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.0379,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.5133909030038333,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.0067,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.0515647626068354,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9465,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.9346014434550147,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.9185,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.8365920899023425,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9179,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.004657017595472,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.9298,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.077900347373173,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.9322,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.388777897807061,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.8475,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.92812655002541,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.8506,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.6835133083168725,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.7977,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.4789193275357317,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.8382,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.035952543348585,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.803,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.5662151924688594,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.7759,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.607964982505537,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8364,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.6670787690207596,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.7713,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.5630249637232239,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.768,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.286587311155195,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.7522,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.0236566874180228,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.7507,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.9658318681539559,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.743,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.9923418030352095,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.7301,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.9905158245287717,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.733,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.9175611734258573,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.7196,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8523143579879601,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7306,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8747475490132902,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7041,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.9383571824718591,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.721,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8857534302189981,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.7232,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.828787271885409,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.7046,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.8241210099272175,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.7171,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.8524136962258234,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.6861,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.840210573224335,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.7052,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7508429427660394,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.685,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.9189785649148905,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.6862,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.8690635971792199,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.6766,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.730512888501541,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6783,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.7690989142619535,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.6921,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.8563126367257627,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.7185,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.7462259820964671,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.6605,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7503601548148805,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.67,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.9028501440387672,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.6871,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 2.1084031657758158,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.6972,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 1.9329674403941826,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.6879,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.9138362907680683,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.6923,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 223.27944296477165,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7082,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 1.7141369258182733,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.6949,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 1.2591689901845518,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.6752,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 2.041123272076082,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.6583,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 1.409791849683831,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.6675,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8086828170343617,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.6686,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 1.2167171344036574,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.6587,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.9383569757403981,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.666,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8208003484581715,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.6527,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.9206768929516913,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.6717,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.9931537143462605,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6752,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.9181304019967126,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6864,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 1.033163080975643,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.6702,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.8931455807754183,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6553,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.8886110702795539,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.6492,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 1.0594253896811485,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.6256,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.9609061574055509,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.6748,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.8329148873971765,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.6691,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 3.728139938135183,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.651,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.245608222184224,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.6718,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8890086949480798,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6376,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.1897765208766364,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.6333,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.975637631230067,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.6865,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.2489235363651663,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.6722,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.0876409084255114,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.6798,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.227051064610179,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.6939,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.0028496830241518,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.6826,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.4462083335694886,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.6855,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.0284214956583457,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.6491,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2210721019024273,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.6591,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.9322367880864245,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.6084,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.9385112268962418,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6432,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.8675177543809808,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6306,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.8416451540891292,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.6896,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.0116689036501123,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.6616,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.84649631261072,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.6538,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.8040136300434872,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.6378,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.91460410479814,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.6857,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.8729629774292391,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6482,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.7637035595653258,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.647,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.9287924131529468,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6579,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.741325155091244,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.6653,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.8529860013237602,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.6382,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.7381805416969137,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6781,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.9826088249621951,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7019,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.8934921385783008,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6853,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.8391849155162863,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.678,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.7602882142871198,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.6945,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.9146572543322372,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.6853,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.760274790184442,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.6405,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.8974763850574614,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.661,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.8325910155383442,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.6874,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.7667452184157418,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.671,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.8438704609117352,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.6722,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.8283192352619542,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.6649,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.8035380275586627,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6546,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.729653500288333,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.6468,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.8518927566064664,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.6475,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.8085942909689025,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.665,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.8070532988611295,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6372,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.8532600122231057,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.6534,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.8942998948504719,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.6498,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.002926592423171,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.641,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.9120098934741601,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6415,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.9380143472392178,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6304,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.81370674559914,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.6756,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.9413601243385854,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.64,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.9682874595989704,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6459,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.9122624840315027,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6511,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.8859293474733565,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.6421,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.9014957977716569,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.6534,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.8102266870023593,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.6581,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.8461114778865787,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.672,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.0514802148238334,
+      "learning_rate": 4e-05,
+      "loss": 0.7068,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.8235582821680756,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.6485,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.930951938626008,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.6537,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.9836904266778915,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.6598,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.7334904080084842,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.6566,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.8281283295153334,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.6476,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.8634073532322072,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.6556,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.6948289772630216,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.6644,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.7158549695641089,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.6513,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.9231193280513209,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.6615,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.6885833591696056,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6473,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.7913202708980839,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.6637,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.8275847998729401,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.6342,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.79213306959823,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.6252,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.7330677731184417,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.6818,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.7430835852367866,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.6655,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.6796042846198054,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.6628,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.7078800434232246,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6446,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.6319266998228185,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6975,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.7059288612609647,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.6562,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.6368487349068374,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.6451,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.6795708382360369,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.6792,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.6427060534781673,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.6208,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.612484670553353,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.6452,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.6827279418572139,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.6371,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.7293686755415079,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.6218,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.6772567518586076,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.6019,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.7336322193692398,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6409,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.6481536137755015,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.6513,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.6409700916301514,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6406,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.6034604109629728,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.6766,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.7298289642842747,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6638,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.6424334379066738,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.6889,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.6922418834977158,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.624,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.6369360664568334,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.6627,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6652319433442724,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.6402,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.5929446316458978,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.6569,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.6090744203796251,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.6364,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.6587614884369014,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.6153,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.6311263075231613,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.6638,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.594983804274016,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.6335,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.7094557090475045,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6247,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.6170993109084666,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6451,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.7543217686936601,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.6519,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.6088310390439176,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.6732,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.8693302595405114,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.6828,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.6468670872391211,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6323,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.7146576408382633,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.6652,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.6538312400667592,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.655,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.7985715879696673,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6689,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.6652569451990429,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.6417,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.7708037903587114,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.5948,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.6997389441108517,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.6356,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7266474475635475,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.6468,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.669232440830105,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.635,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.7460313088215234,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6336,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.8283205155776742,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.6116,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.6767156978184554,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.6574,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.7442504352780269,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.6762,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.7076704804300442,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.6425,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.6959995049272478,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.6178,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.6841032312293543,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6556,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.7218238575860992,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6206,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.7116172823939596,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.6339,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.8010078193408366,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.6367,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.6850174909245291,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.6425,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.7340847530437592,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.6349,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.6373317905346265,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6348,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6821987975420917,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.6295,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.6755585459718444,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.6296,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.7169065791642172,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.6764,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6099767297540594,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6588,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.6306646830701889,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.6272,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.5410236632721085,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6351,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.626874929851591,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.6548,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.572241801572449,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.617,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.6102936603714522,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.6444,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.5725643665420301,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.6373,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.6134228839803514,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.6327,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.7690110727440086,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.6735,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.5827960586670748,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6035,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.7632988959361899,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.6659,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.6570295107507819,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6431,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6202595720572815,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6436,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.7737731404105513,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6498,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.5759750524794048,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.6644,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.68004022938516,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6294,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.5580216104202523,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.684,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.5630820234681108,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.6374,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5516083430513145,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6151,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.5516966842024429,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.6243,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.6315392740852651,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.6256,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.5423692879834952,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.6255,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6221415728521917,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.6402,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.570722678194618,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.6476,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.6442518155887126,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.6378,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.6688457789282948,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.6527,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.7097861901618285,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.6304,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.588745189568131,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.6414,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.6449557732974901,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.652,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.6130195312140114,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6057,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6052129761410693,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.6392,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6266113129766155,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.6384,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.5982132842461046,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.6356,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.5441493784015916,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.6341,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.6203945563268402,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.6471,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.5407736144531398,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.6224,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.6352698374779214,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.6373,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.6689156354902173,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.645,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.6152720516229183,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.6419,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.6158130060803372,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.6634,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.6428144947551855,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.6188,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.6104597787598957,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.634,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.6426579856947144,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.62,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6646965306180128,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.6254,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.6250778843860042,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.6188,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.6472734903642414,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.6298,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.6391416736351171,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.6135,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.6341496301356775,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6122,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.6390536095794175,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.6317,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.7032182979768842,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.6303,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6351466598183807,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.5901,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.6420588987235302,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.6156,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.6760061648981666,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6397,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.6568042236831116,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.581,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.7105900577171068,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.4983,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.5972720775705576,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.4882,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.8103273795598421,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.4776,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.7169354575517243,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.4419,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.6903010634960418,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.465,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.752719357642927,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.4762,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.8035610014080837,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.4633,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.8128493767571628,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.456,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.8088802749209474,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.4974,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.9958986796932074,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.4746,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.6693603945814001,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.4589,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.8522511887616592,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.4836,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.7548442677258032,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.4725,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.6929221601771897,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.4676,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.8344260874265783,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.4476,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.6560320921178163,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.449,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.7657155345792684,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.46,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.6624957831116526,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.4713,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.7476099015196622,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.4471,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.735064117047132,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.4549,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.7070779767967079,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.4494,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.6929772348965088,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.4612,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.7760912340344637,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.4575,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.6130022191127001,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.4476,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.6792732458225226,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.4681,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.6749038812038228,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.4516,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.7113515446427943,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.4637,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.6747285483286852,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.4647,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.6273258606582179,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.459,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.6392632221418385,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.4586,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.700899100978642,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.4617,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.5906789389335682,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.4297,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.7118609880621364,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.4757,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.7281810033098065,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.4573,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.7101725454171898,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.4778,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.6619121216818693,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.4795,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.6539176247391364,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.4452,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.6122832439661751,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.4435,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6709800503099221,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.4639,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.6001445350820667,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.4849,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.6969475881320337,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.4727,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.5791190692337378,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.4519,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.6743830851607197,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.4735,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.632024245505915,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.492,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.6599185762332053,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.4814,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.5969848284144543,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.4665,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.6821147552775841,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.4811,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.5778980045399446,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.4764,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.6714295575649323,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.4639,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.5992429641211715,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.4588,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.6101732775664727,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.4819,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.6331165222771684,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.4708,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.5406517132696198,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5136,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.6277454120140366,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.462,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.6019745363141094,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.4904,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.5796301932086314,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.4404,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.6027236086167321,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.48,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.686110473415907,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.4618,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.5542810577159188,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.4766,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.6418540378670553,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.4614,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.5448402572903419,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.4582,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.6183040707059319,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.4816,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.5273891780405989,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.4826,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.5925383730631434,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.4327,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.5425732072699743,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.4537,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.5960339906961348,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.4614,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.5921826417694743,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.4748,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.5719186415968186,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.4606,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.6471159511254158,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.4629,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.5601658484431123,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.4655,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.6088972771839939,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.4415,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.5775116836322706,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.5047,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.610285428691966,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.4746,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.5536848258864214,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.4781,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.5698613770609317,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.4659,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.5979318046944179,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.4687,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.6361395192863373,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.4593,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.6226020221693606,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.4363,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.576827382390226,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.4481,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.6246554605791382,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.4618,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.5404512179252012,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.4637,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.6031831343477638,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.4765,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.5223572992566323,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.441,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.6155568215186296,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.4452,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.5242890891740736,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.4698,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.5726387270951608,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.4663,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.5724070000309497,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.4737,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.6004388275387859,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.4444,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.5551101411195459,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.4708,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.5953396780228186,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.4499,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.6339724798476168,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.4643,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.5940994929969913,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.464,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.6070097413991277,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.465,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.6391009617522212,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.4575,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.6729975053680617,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5022,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.5887720862687339,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.4842,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.6782499245222308,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.4867,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.5622751912397903,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.465,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.6864387190467428,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.4735,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.5434147187685592,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.4831,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.6291974744450396,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.4593,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.5987083419271627,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.4914,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.6517678609617189,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.4791,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.6250688893569823,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.4833,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.6262715588019816,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.4586,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.6941789375619095,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5065,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.6109699794192772,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.4494,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.751823618065008,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.4745,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.7071066672217827,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.4774,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.6419770825660985,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.4654,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.645757417638719,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.4679,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.577122369425139,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.4674,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.598269654765507,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.4967,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.6419373593685497,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.4533,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.5262188474767828,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.4921,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.6228273664203925,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.4729,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.5635280434442698,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.4783,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6502575823236246,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.4773,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.5505732216757251,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.4742,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.6161698256569267,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.4856,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.6046530398442644,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.459,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.6394016686681863,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.4732,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.530559184790106,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.4473,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.5396040913369808,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.4832,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.6012878071095595,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5111,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.5542419871925436,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.4768,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.5656739875154642,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.4741,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.5231831063204309,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.4488,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.5569326659042632,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.4653,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.49355399789498794,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.4689,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.5961196894788224,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.4708,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.5283542267254073,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.486,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.5765950329270525,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.4463,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.5447619248298353,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.4808,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.6055665372252346,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.4787,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.538214026025442,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.4788,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.6892187547056497,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.4607,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.5116307067971553,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.4609,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.7145160446908356,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.4503,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.5342533394061478,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.4794,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.7237650192722423,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.4851,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.528461859945327,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.4829,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.6551303081940585,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.4855,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.5531376961017279,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.466,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.5950760628025988,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.4683,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.5994283445257584,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.4677,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.7020084873335637,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.4526,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.5798626503532199,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.4874,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.6278018289604667,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.4691,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.6223749330433921,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.4849,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.6251268862008539,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.4708,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.5371156307348297,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.4732,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.5797308538204851,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.4604,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.554063402000413,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.4982,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.5405228394117974,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.4683,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.5941349576508678,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.4729,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.576290035386933,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.4762,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.5568667457213997,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.4308,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.6301124450154332,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.4762,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.5762111465931292,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.4649,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.6787530840296664,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.4542,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.5745188559433335,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.4685,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.6891697985571085,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.4919,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.546329244918577,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.502,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.7159405756759122,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.4763,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.5586896898827479,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.4812,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.6706764312613056,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.4817,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.5821688767162574,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.4772,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.6458293910932532,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.4416,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.605292142698827,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.4603,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.6437961207836644,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.4477,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.5834439499123111,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.4903,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.6649314686541042,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.4867,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.553131617828397,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.4805,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.6435649942653201,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.4728,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.5638461729842779,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.4792,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.6486845070122554,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.4985,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.5710730645176854,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.4592,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.5829988545401421,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.4674,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.6175568050256309,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.4863,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.5460209730563624,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.4729,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.5381724995131161,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.4405,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.5032721363878958,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.4745,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.5495265646319908,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.4627,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.5201262304335427,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.4725,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.6064136596806466,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.4561,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.5764673613797476,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.4513,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.6129771945051231,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.4932,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.5599891438980965,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.457,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.6064161255398303,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.479,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.5580839616719036,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.4821,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.6375158276142909,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.4736,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.5750347338804767,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.443,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.5465851286849523,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.4688,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.6259309163278164,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.4623,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.5210868354013399,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.4792,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.4958393354222492,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.4475,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.5562022640318252,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.4566,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.5244706243708722,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.4819,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.5275233239824495,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.4669,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.556081394399311,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.4627,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.5779840639432459,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.4625,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.5445208216066534,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.4637,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.6002531832303815,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.4917,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.5437797190507258,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.4714,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.5326831314815836,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.466,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.5566633726418017,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.478,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.5151566261699818,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.4874,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.5222104886616609,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.4697,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.5547098556758493,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.4625,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.5316161213911286,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.4807,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.545792435453774,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.4675,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.6091274654559886,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.4907,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.5407528226606386,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.4746,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.5691808614435953,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.4763,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.5012466042740676,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.4355,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.5785790834257599,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.4409,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.5617583992700564,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.484,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.5773192973787675,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.4403,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.5184640991631416,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.4612,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.5671699246545178,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.4881,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.5335934960953732,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.4677,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.5283010080965567,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.4511,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.5963915534961666,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.4822,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.5432676548262376,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.4778,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.5711716496596139,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.4709,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.6349293445880606,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.4653,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.5212100126931515,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.4739,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.5769629896672893,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.4546,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.6240826232064525,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.4744,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.5319320032223908,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.4796,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.5744121105320332,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.4603,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.5276806887559484,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.482,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.5495495765734068,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.468,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.5617557854594646,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.4723,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.5768957428067996,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.4819,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.5463579889574389,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.4738,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.5506970500790718,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.4895,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.5315595921736684,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.4806,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.5221308186404235,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.4687,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.575593104519141,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.4902,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.48953968149148996,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.4523,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.5411311851055279,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.4601,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.5526374942750824,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.4787,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.5916299110029125,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4926,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.5470556135326415,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.4406,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 0.8282446995098268,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.2924,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.586742833299635,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3088,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 1.089237125908081,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.2946,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.6360039881686604,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.2806,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.6064579829506439,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3022,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.661885885983905,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.2766,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.5596665024817902,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.2937,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.5052310052693685,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.2603,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.585721789518757,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.2917,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.5722439095704643,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.2836,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.5749537874887233,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3022,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.5128414566052684,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3029,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.5321853046297115,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.2682,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.5213985851318284,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.2668,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.4804463048762983,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.2773,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.5401133460987886,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.2961,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.5357134588514708,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3006,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.4981653521754497,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.29,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.5433826047400084,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.2794,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.5365111220596483,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.2838,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.5351484242620826,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.29,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.5201799946831618,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3131,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.5024830374627757,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2993,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.5188494931616093,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.312,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.4853032586266865,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.2772,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.5190783051447905,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.2866,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.49843045602257413,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.288,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.4666511645062999,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.2819,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.48420221528146723,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.2926,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.5120864609041871,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.2617,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.4786312625195475,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.2939,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.4863012239612095,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.2822,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.5007019967641706,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.271,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.5120537850000061,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.2816,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.4946348104044211,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.2676,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.5647047797582649,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.2882,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.5117602799935108,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.2783,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.5667203653212508,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.2846,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.4871507066038831,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.2717,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.4932112339562045,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3258,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.5195358473577913,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.2969,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.4832267131028698,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.2908,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.5116848686191477,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.2804,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.5026961591655797,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3077,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.485506794342039,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.281,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.4968490432138733,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.286,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.48849167763880097,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.2892,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.5096656057432343,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.2898,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.47498480393964504,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.27,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.5183478652224603,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.2926,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.4702977042251421,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.2802,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.5078565059097788,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.2918,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.49845765442440554,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.2776,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.4916425552281145,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.278,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.5057149031997401,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.2653,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.5285284373700571,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.2838,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.4670187153933858,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.2847,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.5039060535843118,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.2895,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.5405314768896156,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.284,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.47743738107779116,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.2743,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.5292656299922083,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.2926,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.509058994061866,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.2972,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.4761458901178783,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.3019,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.47326825691425684,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3018,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.520021649396261,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.2919,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.4699426349688776,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3054,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.4804812150877961,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.2896,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.48324063740461776,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.2722,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.46721839420980055,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3053,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.49388611571696417,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.286,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.48530174230943596,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.2793,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.473343885386964,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.2663,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.4987669209397457,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.2499,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.4921613968119873,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.2893,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.517925632061517,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.2763,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.5201191230544617,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.2881,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.5088486793899849,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.2551,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.46551410589142195,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.2954,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.5019908023010106,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.2797,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.4794213919456009,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.2796,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.46194899905070763,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.2685,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.4933181236536789,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.2916,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.484528604718971,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3053,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.4869464228534949,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.2865,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.4668155978218878,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.275,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.4846208616115795,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.2835,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.4695620332897581,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.2839,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.49369912347048794,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.2946,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.4805554309790309,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.2952,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.4532313901853835,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.2809,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.5176976444512263,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.2848,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.45335942026994785,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3142,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.4620432191220203,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3034,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.4905147999178627,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.2765,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.4689130084972855,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.299,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.47037445303431696,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.2665,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.4887640241227454,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.2833,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.48358871474785103,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.2842,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.49953808296038216,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.2677,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.510293183108694,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.2979,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.486442021456212,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.268,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.49956056845649294,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.2843,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.5291616102266176,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.293,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.5014164918808927,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.2803,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.49298129170714855,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.2957,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.5033135720115768,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.2995,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.4511962052065107,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3016,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.5278222587410356,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.2918,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.48292849796840237,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.2951,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.48614240531647307,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3051,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.4670497917385395,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.2852,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.4496642617457187,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3037,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.46608350038422874,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.2773,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.5006721007432772,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.2688,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.5279765988878166,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.2944,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.4849961600510434,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.2663,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.5148639027635645,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.2921,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.49117382407765525,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.2911,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.4715681144961897,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.2878,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.4780930409022001,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.2791,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.46430522228260857,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3028,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.4774334156045388,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2889,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.45090714469538246,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.2871,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.4849293870663762,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.2631,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.5107301729530503,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.2988,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.45342225794049673,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.2806,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.4680407000628561,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.2827,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.4430024724763211,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.2656,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.4531014355819758,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.2871,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.46420594347864985,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.275,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.4546854788320037,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.2807,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.47786654146349217,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.3055,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.45842300633351923,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.262,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.4663668439755189,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.2912,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.4610465279333058,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.2828,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.4467871271193587,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.2651,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.4783136360456997,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.2839,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.48004560324756534,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.2772,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.47078873037460905,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.2843,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.45944327838876814,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.2832,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.4646012425760017,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.2827,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.4530279930462132,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.2779,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.46454067023348655,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2648,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.459427589122945,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.2883,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.4632179243971944,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.2803,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.45613532988482763,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.2692,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.45449274749986396,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.2911,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.4749173104429681,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.2941,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.4637415864053045,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.2862,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.45926839080670073,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.2955,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.4505989734931718,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3145,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.45957190673843656,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3049,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.4971503313301464,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.2668,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.45367217907701285,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.2836,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.47151372440254286,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.287,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.4738930961053102,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.2755,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.47471790266872294,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3066,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.49908657032579345,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.2728,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.4754197340715637,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.2925,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.47284725471105127,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.2968,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.45481099599046665,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.2831,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.4529750929079318,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3006,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.43568597118831304,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.2866,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.4627112093254117,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.2659,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.4664453122967578,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.2684,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.44147320241442084,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.295,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.46847264220444257,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.2799,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.4815098757301856,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.2678,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.458303212030509,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.2729,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.4757793908164495,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.2744,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.4840637855100168,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3019,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.4724417603153787,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.2846,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.4625687891331373,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.2813,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.5003299108081725,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.2867,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.47865883948512217,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.2996,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.4763425238943243,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.2809,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.45344258069763826,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.2835,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.47253939885064744,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.2692,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.49743005576488536,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.2963,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.48453959974255384,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.265,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.46115871285292487,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.2948,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.49095773697020234,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.2724,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.4661971952976171,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.2704,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.4696016341587489,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.2881,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.4591896011678293,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.2895,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.49834797722500634,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.2873,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.45161631101436,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.2866,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.4498456627333069,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.2759,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.48367057225611343,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3009,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.44122061478032926,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.2978,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.45050765518256053,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.2794,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.47720347043548006,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.3128,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.44059633567291434,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.2933,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.46034972021487724,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.2711,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.46426428643992695,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.2844,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.4512643975558349,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.2732,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.469230884658159,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3044,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.45572102558143424,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.2796,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.43818454377622684,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.2905,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.43191649860559705,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.2892,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.47103749897123426,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.295,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.464308851908428,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.2989,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.47647997847441725,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.2719,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.4582570233815993,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.2775,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.4819385823967622,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.2852,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.4681642184902604,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.2889,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.43914872489189893,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.2872,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.4457309847980171,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.2835,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.4893194906436807,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.2734,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.43952998844778174,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.2867,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.4409984085396293,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.2807,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.47194720595934314,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.2697,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.46798668518709474,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.2737,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.45123225382369625,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.2856,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.4712976897291697,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.2971,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.4360233973733771,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.2821,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.49140338491298013,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.2767,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.49765354396430495,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.2913,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.47415549331421264,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.2776,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.4873662405170041,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.292,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.48403600916393574,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.2659,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.4521937910840975,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.2978,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.48017008606532047,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.2835,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.4675246057439373,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.2935,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.4490695143930646,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.2496,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.4634431830112357,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.2751,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.46319505895360324,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.2885,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.49142603674518254,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3006,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.4489761813133768,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.2719,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.44363364084336865,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.2929,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.46256362559151665,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.2867,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.47290527336750665,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.2964,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.442871307692817,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.2737,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.4330831299449121,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3016,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.46598828963446154,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.2724,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.45327446198589594,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.2652,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.42312896951086165,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.2757,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.45341820826813845,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3138,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.44300828848095125,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.2827,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.45804077774635554,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.2698,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.4390681570299565,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3021,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.46188419842970596,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.2855,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.4584940412683278,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.2654,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.4361312460183006,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.2728,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.4284074468202381,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2708,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.4687585946983475,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.2976,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.5031076989134806,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.2315,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.5125918381557849,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.1867,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.38988089588083,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1658,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.478462673569671,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1728,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.5743098787429364,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.1615,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.5856674183624859,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1761,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.44142422401657716,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1567,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.4439565328401261,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1491,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.4683987377201235,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1837,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.4638536981454729,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.1456,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.40527799577090545,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1587,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.39429158202304065,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1753,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.39578207767128126,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1375,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.45102184613181573,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1726,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.46966676341641483,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1689,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.39969456957685645,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1628,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.38328224817472406,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1481,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.39378031811775277,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.161,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.4072950695972632,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.166,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.4170116682507317,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1738,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.3993757463827332,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1737,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.38684837511324266,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1787,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.4053929235212583,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1815,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.4041417023237574,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1519,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.41860259599249133,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1798,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.38793161886193583,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1673,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.36964015939502404,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1611,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.3789850393750237,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1443,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.3675833742305326,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1625,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.3897926311087329,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1688,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.3740619639117436,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1374,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.352248342475438,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1711,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.3843493532735204,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1628,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.4292986143111384,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.1827,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.3663028634375464,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1597,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.3884011916314162,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1479,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.3792329181566654,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1676,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.3842778558902252,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1562,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.3648589017901102,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1608,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.37945945085762484,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1495,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.3680278033535353,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1692,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.36310435309332584,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1421,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.3782443073193828,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1793,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.39168500371079146,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1588,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.36664835702325543,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1789,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.38032772911303275,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1643,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.36441664722454126,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1649,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.370814871729749,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.154,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.37222939467251187,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1458,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.36309774390401717,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1671,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.3541116775851228,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1812,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.3670085035568729,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.145,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.3772421210722759,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1624,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.3572163908454282,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1528,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.36763412190074707,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1214,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.3877588192605807,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1549,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.36047138887643004,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.1472,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.3664481068616539,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1404,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.38442085507543555,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1662,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.3649059340331149,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1355,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.37494073201468164,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1704,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.36120084044432044,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1702,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.3612588651418685,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1776,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.3708723096532811,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.1561,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.3968362515215659,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.1564,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.3948960159801775,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1456,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.3694677639469941,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.175,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.3948135537001844,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1431,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.39299962931171106,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1382,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.36580552474837574,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1674,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.3756403738258354,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1496,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.38057793142736784,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1521,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.3894208861044804,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1603,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.3813146341130529,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.165,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.37800336326912864,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1585,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.36443792183439333,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1521,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.373214641898224,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1744,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.3974715895432565,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1657,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.3657306548254128,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1609,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.36129638794291996,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1608,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.37104122265498957,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1665,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.37006212945472844,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1797,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.37506826454204467,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1487,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.38152683176355207,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1232,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.3750608002264649,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1582,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.36502955416311267,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1608,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.35653133263433284,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1681,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.37052438029328927,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1607,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.3647215576780881,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.1545,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.3771575386722221,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1657,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.38256308780169723,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1602,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.38410629825153786,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1563,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.3544893086680657,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1535,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.35304520688543906,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1382,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.36398039036963514,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1486,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.3624828486422592,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.1557,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.3811001684155079,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1681,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.36884338116492615,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.154,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.35913236047165087,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1691,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.35739704909873193,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1682,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.35621011832137756,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1428,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.36025438832779877,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.15,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.38343996709126255,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1646,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.3670288105516792,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1455,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.35861897569322315,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1478,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.35829983100867024,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1443,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.3617489103289309,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1549,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.3671262026098243,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1688,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.36026050796327436,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.1564,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.36756992982498393,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1349,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.35917042355389256,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1398,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.3522686134827049,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1644,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.3640564537641408,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1719,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.35794656579055906,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1619,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.3802829648211374,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1879,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.3454630311248087,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1947,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.3541202348681889,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1518,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.3562639861432815,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1537,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.344044972228574,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1737,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.3636619134134887,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.149,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.3513221444626807,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1505,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.3559616252932611,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.152,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.36298773852763005,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1407,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.348101078553375,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1645,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.35417329796567876,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1571,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.37217930863137616,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1832,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.3793788766176517,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.1678,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.3702561948769752,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1636,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.35413952574115515,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1405,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.3723266808940605,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1547,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.39045384951013196,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1548,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.3853165370495703,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1415,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.36270516395821567,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1599,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.36484446537911336,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1579,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.3756519237985607,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1687,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.3596392742694853,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1487,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.3597406040326724,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.1731,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.3737855118057644,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1443,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.3723339670566699,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1555,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.3788972365522724,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1378,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.35993254811783537,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1534,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.3668974283172391,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1559,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.36225907562422965,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1338,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.3820222835054556,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1586,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.3561560432282962,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.1462,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.3595870272056075,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1621,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.34276021547242935,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1673,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.35455898347430786,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1373,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.3506483233500243,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1449,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.3744996227392854,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1724,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.38507737279992676,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1696,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.37110549988355196,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1519,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.3661742054521129,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1839,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.3705951789283633,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1688,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.37351363087503603,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1617,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.35047425045680036,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1697,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.3582003539333888,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.171,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.3804931572484399,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1625,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.36364437883252904,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1513,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.35048727357429066,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.1478,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.3517654410896592,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1533,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.35514377263378644,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1537,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.3755437086585866,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1593,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.3720562629647983,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.137,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.35259639662411413,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1486,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.34504335957585575,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1534,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.351004233460667,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1584,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.3692744197549358,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1939,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.3659300266729104,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1522,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.3436818690978363,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.151,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.36289381371668644,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1376,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.35051932615571724,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1445,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.3736214833657092,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1518,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.36828425352974786,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1457,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.35623675603632105,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1807,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.35534344250830324,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1504,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.3659427944668952,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1562,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.34367270252775184,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1547,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.36514484714962125,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1401,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.3605220816272273,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.1453,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.35688392753147213,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1613,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.3431025339087566,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1421,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.36932396473965884,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1717,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.37722212114501597,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1566,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.3571870596524292,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1496,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.3443491856859766,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1338,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.3370418863335358,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1678,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.3477504946253808,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.147,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.35460518191049606,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1488,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.35749899780899375,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1672,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.3461481647792584,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1462,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.3586557024002623,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1628,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.3496097878628896,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1776,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.36284569112174087,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1622,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.34942849127536757,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1468,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.34909411467139645,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1634,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.34215423843821224,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1501,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.3440941143126817,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1647,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.3687492425769216,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1586,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.36191621745306407,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1376,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.3531272487500724,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1777,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.34316487383308636,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.138,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.3467651863326572,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1529,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.35694332144400676,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1559,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.3524601277972167,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1687,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.3525931301747568,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1427,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.34175216330466884,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1493,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.3544287188819749,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.169,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.35860065378685946,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1537,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.35242128211958645,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1743,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.35708767437838057,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1838,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.34671111252880904,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1651,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.3568732839551748,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.1686,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.34764439453585533,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1579,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.33817596429762536,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1435,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.34349294989876317,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1634,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.36438823106620677,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1852,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.358249695722888,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1666,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.35847889061013116,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.159,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.3524110331284342,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1583,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.3606594056745689,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1398,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.350314798349193,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1513,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.34557454876084914,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1563,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.354687953477054,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1405,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.3349848260894284,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1449,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.34872916439292656,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1518,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.34170125512202043,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1548,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.35965293894058875,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1397,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.3394924043615626,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1372,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.35424690138835413,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1565,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.3685109923752025,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1572,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.3470461258098718,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1419,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.3446872526363923,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1653,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.3703211671942602,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.154,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.35509436219912727,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1487,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.3435696874440167,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1531,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.34683540944106517,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1559,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.3566613768164915,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1566,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.343714301523173,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1554,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.35554854384725304,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1623,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.355488990712807,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1711,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.3530826571788529,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1672,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.3494562983504821,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1453,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.34299831729246183,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1704,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.351078422131716,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1545,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.36869087525416455,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1394,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.34392190876255346,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1529,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.4021179680924069,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.1079,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.3744847267889842,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1117,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.2916114961939603,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1073,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.2624787535682735,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1271,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.23355194675531296,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.077,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.2800641248315272,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1031,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.3089744116383526,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1047,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.3394943201490061,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.0989,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.3457049396998426,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1019,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.3375255367357801,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.106,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.3245776370246612,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.0907,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.30542672169817336,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.086,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.2717942636248466,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1122,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.2641578631819137,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1041,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.2939707991433495,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.09,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.30670558938009046,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1051,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.2948117368054941,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1154,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.2823372434834665,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.0995,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.26533578351742765,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.1098,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.26906828001635724,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1025,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.27594812834356963,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1023,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.2741759124267442,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.0964,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.2623769374371785,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.113,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.26334804070608225,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.0983,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.27200936231838707,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1054,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.278617375513882,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1136,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.27889131109406873,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.116,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.2579390561126674,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1065,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.26101114557852023,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1036,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.2681048950767238,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1051,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.25984153394654136,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1068,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.271427605034457,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1138,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.26627502099029177,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.0855,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.2587224912808297,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.0947,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.2576053233330602,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.0942,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.263199878847666,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.0975,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.2638306091928673,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.099,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.2662390088078004,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.0926,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.2622781782480276,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1078,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.25004413938225123,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1025,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.24982855366969747,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1073,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.26227165006963066,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.0944,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.2663342177493356,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1023,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.24706304765237974,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1066,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.2651032302721895,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.0878,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.2771622683808167,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1213,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.2588818992587375,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1019,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.23765633689348364,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1013,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.2445058043472427,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.0982,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.2553829498257803,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1069,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.24767175802545785,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0945,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.23954185195529784,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.0911,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.258880583832178,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.0938,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.2534179303758477,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1072,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.2437836552630438,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.0931,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.26272550737583916,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.0947,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.2634425742405811,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.0897,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.24834430195975676,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1083,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.26983443994461165,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1023,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.2547222008918068,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1029,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.25673271335530945,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.0936,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.24577840655312316,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.0949,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.25155091681018493,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.0955,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.2549273921607578,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.0987,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.2622254371133505,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1139,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.2589178040714933,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1022,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.2567591872646608,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.0975,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.2632622446016957,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1041,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.25612068577112435,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1028,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.25503660719896004,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.134,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.25798761901027883,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1116,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.25442894652860587,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.107,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.26666949422157943,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.0973,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.2586379231815266,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1273,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.2612034050908108,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.0977,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.27087778702510706,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.0981,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.26199719669284516,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1022,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.26107685136199016,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1047,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.24944767781539898,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.0891,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.2617242943158962,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1036,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.2555875225729689,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.0869,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.2526252662812399,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.0839,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.2553669686763501,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0951,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.250339887779964,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.0969,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.2666820941716029,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1144,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.24442549123003898,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.0984,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.24402830319610283,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.0964,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.25268651985891183,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0856,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.24360273969559298,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1097,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.25831546145705575,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.0955,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.2638127665386737,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.0901,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.25907630899019685,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1085,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.24708284697644617,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.111,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.26081008194508604,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.0916,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.26316467041829006,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1175,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.255168531215649,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.0972,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.2604394764702707,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.0851,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.2598424340517516,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.0887,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.2661876785993073,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.0928,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.24771957139746428,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.0864,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.2641220557207087,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1098,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.24822335609257107,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1102,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.25892945373478565,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.0964,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.24317853466966222,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1048,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.2508189031319834,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.0905,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.24865740167333442,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1207,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.2658351770582443,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.109,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.2719859797968942,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.134,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.2542877631728742,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1038,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.24794392559342462,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.0885,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.24569484534937366,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1053,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.25767880219111583,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.0926,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.25620016005779994,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1037,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.25621506264282023,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.0778,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.24056862628261017,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.092,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.2475515983213626,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1001,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.24095891072122133,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.0838,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.24698914884391956,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1004,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.2596307629592867,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1021,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.2345180850618566,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1194,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.24730963895628563,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1123,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.2543085119863913,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.0839,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.2620189316856834,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.0983,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.25926097248819946,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1048,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.24022707769110202,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.0881,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.2466623621093566,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1033,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.25662205674468885,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.0982,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.2571750053541413,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1039,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.24531294836110526,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.0916,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.2583665823599055,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1148,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.25178326059147715,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1023,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.2545886171091869,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.0873,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.25548317633193446,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.0879,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.2437319878186728,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.0806,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.24602734689180428,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1114,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.25523838416589845,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1104,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.25426931563964433,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.0909,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.24559961947657588,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.0772,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.24461535153040512,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.0794,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.27365025183439157,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0971,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.24627697451731398,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.0862,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.2489521322424386,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.0977,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.24714627728779262,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.09,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.26279888467211393,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.0945,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.26824317157377514,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.0979,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.25542581939038356,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0935,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.24883708412971994,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.0796,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.25710476761745105,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.0966,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.24438706612587324,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.0853,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.27245673461520536,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1133,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.2552294413999096,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.0889,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.23853511636382932,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.0934,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.24326241624169584,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.09,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.2516344375549256,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.0905,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.2506858901617,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.0942,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.25626622072114336,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1068,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.3009573203923295,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.0978,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.27543718480980334,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.0826,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.2586071510103317,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1149,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.25126122312063076,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.0897,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.24671312396829712,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.0978,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.24640509784077308,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1012,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.25730997141266726,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1025,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.24260429279852508,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1059,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.24603836847896243,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.083,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.25563280058234983,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1045,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.26510725286782877,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1009,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.24380855318624153,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1046,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.25801395365175844,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1077,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.26656325999611247,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.101,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.24426663900856455,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.0794,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.24285738071414373,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.0903,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.25556106341906165,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.0973,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.24484016366378938,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.0892,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.2715932152281529,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.0851,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.2513271332078787,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.0842,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.24741847643568016,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.0946,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.24795368304095922,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1017,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.2543472146446525,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.103,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.2473823079277741,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.0825,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.2537959486612439,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.0961,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.23921190104350132,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.0879,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.2562652666219037,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1093,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.2565924905995063,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.0921,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.25394093221169745,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.0903,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.2569880714230178,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.083,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.26440680921075965,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.097,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.2538202162612028,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0989,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.27190094249017616,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1076,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.25952542235631526,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1086,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.24308841654628993,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.0972,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.24494247058286953,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1117,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.24099975435134047,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.089,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.24648931794503384,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.0898,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.2564205283178706,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.125,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.24632895945848624,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.0829,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.26009649089200837,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.0924,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.2566790040676526,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.0864,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.26770965328471696,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1012,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.250174116099337,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.0918,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.2607356580014589,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.0959,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.25594030133716,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.0991,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.2555081601086935,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.107,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.25608649257355487,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.0938,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.262096317511143,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.0908,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.2503046952604416,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.104,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.24232475749357482,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1114,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.2592032048034731,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1167,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.25336815795227646,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1007,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.24769224982056634,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.0995,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.2519940128831564,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.0921,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.2632050870240825,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1285,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.2726170780503193,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.0989,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.26843747082379144,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1143,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.2457685053426762,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1127,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.24759863189745812,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.0945,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.24690732337729876,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1029,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.2635340339571024,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.0874,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.24621403935868122,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.0921,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.24963559947715433,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.0869,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.25751150253261573,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1061,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.2560300982082225,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1017,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.26260422775773046,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.106,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.2418477278693503,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1012,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.2551801569506501,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1108,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.24961134203666566,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1019,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.25233570528316634,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1017,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.2586828315444483,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.102,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.24686766004280639,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1096,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.256212905572172,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.094,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.2596000936113895,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1023,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.2564166934765393,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1051,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.24465548595432057,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1034,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.2599461031441982,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.0995,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.24917482809285307,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1257,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.24591432981284425,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1056,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2586122022031135,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.096,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.24631173647687687,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.0809,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.2541821322431661,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.0965,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.2550562750239185,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1275,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.24835354393099385,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1048,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.2470963689876879,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.0981,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.2607381892486943,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1104,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.29172060834521485,
+      "learning_rate": 0.0,
+      "loss": 0.0843,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 1.2467177959037338e+18,
+      "train_loss": 0.33969098806865816,
+      "train_runtime": 44982.8499,
+      "train_samples_per_second": 3.512,
+      "train_steps_per_second": 0.027
+    }
+  ],
+  "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": 1.2467177959037338e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}