diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.982278481012658,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 7.401158452717806,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.1592,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 7.266607920753238,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.1384,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 7.143782491399172,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.1375,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.920411485562574,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.1303,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.770363170537323,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.1156,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 6.834217027617974,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.1457,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.3281620873396225,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.0877,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.653303470983566,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.0761,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.770120326747381,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.0194,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.541923846358149,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9853,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.621213951344658,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.0304,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.652270314470304,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9883,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.006797348533748,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.0212,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.3799076606795335,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.9612,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.0449864933882376,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9416,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.26454392144685,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.9319,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.4300472220446996,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.9105,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.2422659349990384,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.8726,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.9001649782114565,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8754,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.3492741869421254,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.866,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.3486314735623028,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8265,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.654480917049284,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.84,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.5341272523927747,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8244,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.3155358229327792,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.816,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.023588057732609,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.8262,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1017413036023416,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.8017,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.1811722356982624,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.7813,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0445678306104198,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8055,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.8367444067191703,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.7945,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.7466491514421049,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7707,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.9575880590144629,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7886,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.7330998312108333,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.7542,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8149650765180173,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.7875,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8876121212725425,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.7653,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.8471899801599694,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.7939,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.8485351498345748,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.7677,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.7462219798813616,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.7567,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.999758584303687,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.7575,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.8955003019770823,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.7536,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.778891213292504,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.7756,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.9583075313645376,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7497,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.9207632962458644,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.7647,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.7618232443022214,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.7519,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.7663927582953908,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.7473,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.8816436631390526,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.7544,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.8535695684761769,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.7508,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.6809279064982892,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.7364,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.8637584531405778,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.7372,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7006510136699226,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.7425,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7067233298650707,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7299,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7263929626211771,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7154,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.6756166769544646,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7371,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.8052800576817086,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.7477,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6671600286802184,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7175,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.6899861468791796,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.7182,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7388393190599716,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.7214,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.671083117137826,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.7498,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.6519203669047846,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.71,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7975862790464366,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7072,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.6591146885257743,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.7393,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.8749174467257491,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.7442,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7977348663026834,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7373,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.8207175505163375,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.729,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.8243166975792167,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.7176,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7911402139473752,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.7328,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.787026494240559,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.7139,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.754446013735502,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.7291,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.7760165263126005,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.7186,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8856379315295579,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.7365,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8541125873861473,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6939,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.1326902362681182,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7421,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.7665293954370076,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.7296,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.8219303303236518,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.6998,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.871673195567691,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7406,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.8808735171167452,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.7287,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.8214026904936216,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.7104,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.1623979554964414,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7436,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.9949391519214696,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.7175,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.918234968858693,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.705,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.1359859869317894,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.6994,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.8798086666420454,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.7037,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.9011083797166983,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7079,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9466788390066628,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7443,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.7551679462595378,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7016,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.8540667662888939,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.7002,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.9066706249077877,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7201,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.0754879921366105,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7158,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.9476563916705252,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.7266,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.4796638111546492,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.7044,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.8316031362997139,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.701,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.7820357950877492,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7044,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.1345149513602626,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.7241,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.9099109340963636,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.7052,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.53827575713497,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7061,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.775590763922866,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.71,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.5040415917556391,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.7071,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.5217650379099066,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7222,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.3196754364939547,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.7083,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.590504925947974,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.7003,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.1132595816091952,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7202,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.751247025173818,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.7201,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.4105867165880477,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.711,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.389802678230564,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.6832,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.3494881344016605,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7013,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.2775865012852199,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6927,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.0223926640055097,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7086,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.3633506431124736,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.7244,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.1668608456819802,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.7119,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.0540886997129468,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.7099,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.3969791855170182,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.7151,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.0315679500039883,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7032,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.53401233617438,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7177,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.9105675459545484,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6986,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.7644632463695529,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6912,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.0433454866563219,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.7151,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.469247490233606,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.7266,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.2911012992528454,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.743,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.4819128325580844,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6894,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.1141125306377768,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7014,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.497994886560217,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.7027,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.2423339010154042,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7042,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.1860032604900739,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.7161,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.323668717206232,
+      "learning_rate": 4e-05,
+      "loss": 0.7032,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.4922875114696037,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7046,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.3035933088239786,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.6917,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.294512551831444,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.6999,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.5080856437801469,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.6931,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.9790104449161361,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7186,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.8220700936549958,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7123,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.2518672818971737,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.6905,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.8124006589697785,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.7092,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.1109345342308676,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7091,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.4313931962325483,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.717,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.3724529528565053,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7056,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.04566910969818,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7017,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.300284196183284,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.703,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.1842212117465338,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.7133,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.0268034164618045,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.7202,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.121405303993681,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.7204,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.8834916763015997,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7148,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.0753778590659033,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6942,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.1878372898185565,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7098,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.9617901333274852,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.6999,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.0368452038258424,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.7056,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.8772104649939837,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.6669,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.9093829475430716,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.699,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.0573542272558476,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7181,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.054322047304787,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.6987,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.3139893564544898,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.7132,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.9278132862593048,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.69,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.4014502532062658,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.6996,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 1.0899707853289933,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6816,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.0929699916017055,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.6966,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 1.1214442306227976,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6941,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.0611143541129266,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.7206,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 1.0483381554766373,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.7299,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.9196191698612138,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.7165,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9958766968527022,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.7067,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.1638894666515947,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7278,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.8841449561822158,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.7034,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.2383727438895575,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.6938,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.9628839848926315,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.7326,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.413783794584143,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.702,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.9276302695672498,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6998,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.2890069776152313,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6958,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.9293481201302669,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.7169,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.2805547866704263,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.7142,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.8584936182609495,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.6979,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 1.2354799405995451,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6928,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.9385560182322905,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.6762,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.3083893395975195,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6818,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.129759811261824,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.7075,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.9593681823994321,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7031,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.9342301059114386,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.714,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.1197194292901833,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.7014,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7512730105062518,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7242,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 1.4151113584327333,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.722,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.7224032767820385,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6941,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 1.5790914743883464,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.6846,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.9868504186722136,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.6954,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 1.5328448809657933,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.6823,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 1.4055939432461986,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.718,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 1.192350877288389,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.6946,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 1.2304289122543812,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6915,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 1.108206203764334,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6944,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.9911336035540201,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.6919,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.6841498510719063,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.6758,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.9948646721616948,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.6958,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.8807136641623224,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.7018,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.8863835057023924,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6723,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.7899727612676242,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.695,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 1.0021492279747861,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.6971,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.906074132075834,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7064,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.9495720576511538,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6935,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 1.0872167292461106,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.6961,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.9450167914870186,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6965,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 1.1820908350966983,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7018,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.9154007224818752,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.6853,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 1.081146110705273,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.6789,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.9146864336251234,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.6964,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.9589877812540134,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.888100479827566,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.6865,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.9552888119745284,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6922,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.699818488086489,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.6894,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.9228745177921921,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7214,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.8736227961981742,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6752,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.8617308792886558,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6804,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.8072664599105871,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.6848,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.9479498825020978,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6982,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.9675689844014819,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7056,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.8596756324920217,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.6806,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 1.0750740412086255,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6854,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.8200198126495531,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.6919,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.8545166876498345,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.6906,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.6936741387135874,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.6941,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.9268336608371244,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.6639,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.8190117020452654,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.6964,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.75012639722542,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.7139,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.8498110013303078,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.6918,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.8990506805608882,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.6582,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.7804639877413257,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.7154,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 1.154266825558685,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.7037,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.8527505715424621,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7007,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 1.0677436318123297,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.6936,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.9850578081305938,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.6793,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.9170361426167907,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.6827,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 1.032886409762643,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.7012,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.9270560332084147,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.6774,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.7301641890173969,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.703,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.9546269336604359,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.6844,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.9887015857279297,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.7167,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.9309888197912674,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.6783,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 1.2765521704447242,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.6891,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.9769968763620233,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.7003,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 1.3433662730614089,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.7075,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 1.0041837447561741,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.6834,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.942717450614289,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.6952,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 1.1667775943507324,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.6886,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.7802358107747024,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.6706,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 1.276157046934311,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.7106,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.7168985930346717,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.7016,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 1.6548961950792915,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.6567,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 1.0197370705760327,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.6816,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 1.2370991242894012,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.687,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 1.208403577971206,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.697,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.7620024180659001,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.665,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 1.2721734174505748,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.7575,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.8395683226044319,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.551,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 1.1304268172378944,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.6005,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.9664843443472969,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5549,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 1.0063540398701554,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5343,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 1.0470576278837134,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.585,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.9765748686493293,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5223,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 1.1168593432478962,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.6045,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.835128616044665,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5645,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 1.2129681422897804,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5923,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.7309203838417588,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5437,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.9383820162454584,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5692,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.9037056515230105,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5524,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.7603842219546008,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5773,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.7991894378900329,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5571,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.6751735127128233,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5474,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.8637388246718543,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5521,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.7339412488587183,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5494,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.9508129659957059,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5849,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.8102683672200605,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5369,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.6486890159938167,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.574,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.7060444390678262,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5506,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.6367143562780563,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5442,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.8029355846743584,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5797,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.7629046448502752,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.569,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.7892443131448871,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5475,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.8456103521008009,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5688,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.8615157624214652,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5602,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.8686564449141154,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5412,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.8089784136278586,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5792,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.71454079494255,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.545,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.9679995291337343,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5809,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.6727026559888164,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5352,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.958092514840526,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5921,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.6877108909848757,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5428,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.6662760746788007,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5697,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.7370508208086644,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5622,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.8210756208130287,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5883,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.7556807578500114,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5521,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.695015790032603,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5447,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.7369433378703691,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5689,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.7549212484471463,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5312,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.7656896505279396,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.556,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.7579183422428215,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5179,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.7979388296904156,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5637,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.6973937703831271,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5968,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.7932424746426304,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5437,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.7014324148038777,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5818,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.7785208374239394,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5496,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.7692443539571028,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5459,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.6508311361880342,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5382,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.8329859222440673,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5809,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.877725589910606,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5989,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.7355998958124521,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5618,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.9066409275821774,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5743,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.8528930622412073,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5746,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.7033310151472028,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5802,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.8452804299855399,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5573,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.848470858727363,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5629,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.8362298747768162,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5527,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.6178747943797165,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5634,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.8618068804246464,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5395,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.869691905888505,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5935,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.742733735022853,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5537,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.8705721252364488,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5244,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.7168512721466515,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5993,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.7346609662594737,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5421,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.6267439053270316,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5726,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.6458408004408642,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5402,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7062782441510325,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5446,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.7284142074523903,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5725,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.668586171776339,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5709,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.8900960943007189,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.6107,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.8201927909201551,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5431,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.680442641850652,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5608,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.8009094159774036,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5484,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.7291511821831712,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5809,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.9358386151275478,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.579,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.999707063657424,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5635,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.5987949884172447,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5222,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 1.0826204754626916,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5983,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 1.0055274599124533,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5651,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.6941838587809247,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5498,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.806196371184415,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5363,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.7722997910643215,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.6266,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.6119059801184955,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.5238,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.7564883394243337,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5881,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.5980487571710528,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5395,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.7401853331814983,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5938,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.8089813937996736,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5522,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.64619066431582,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5823,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.6698836508199211,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5408,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.6449783178744393,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5369,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.7086686174799106,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.602,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.8397143332981191,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5664,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 1.061009447720249,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5806,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.6283235185357542,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5668,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.7924412168376195,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5609,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.6841155703792975,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5409,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.6359553624392408,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.603,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.7178315648109568,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.568,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.6721695623417914,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5544,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.7543898016275995,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.6175,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.7403753579038593,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5202,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.7150019310766572,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5762,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.7044738845844752,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5414,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.7545828399824194,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5831,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.9119251995506313,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5521,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.6921178800047978,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5715,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.8671202698855492,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5895,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.6341234406202726,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.553,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.6320059879899382,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5661,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.673170504459141,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.566,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.6099550964026601,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.614,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.6205935262098895,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5177,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.6541749623448165,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5855,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.5668859126601724,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5704,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.6836332179209257,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5613,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.5470905742445094,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5396,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.6510078299248986,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5401,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.6267544412316323,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5829,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.7639590354851425,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5528,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.693721378915266,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5907,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.6570879488486975,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5546,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.7269296449625829,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.6044,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.655672323331821,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5694,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.7599223872143167,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5405,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6482540424964973,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5821,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.7240067075007925,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5744,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.5933159850297323,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5288,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.6883464229914282,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.6075,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.7397363285994066,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5637,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.5707321956977788,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5393,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.6350484541909766,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5764,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.6247384696707757,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5768,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.6636750902475798,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5687,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.5481069591120683,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5373,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.5728964700858302,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.555,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.6761125788457333,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5417,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6557078359026843,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.591,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.6659245037635602,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.57,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.6622160914756388,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5485,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.6837517262021853,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5653,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.6785685460241594,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.61,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.7596655441116433,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.571,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.607393060037397,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5372,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.6619583111127041,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5621,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.7065507811256693,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5683,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6438788413945017,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5488,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.6882896615672858,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5599,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.7432317683924065,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.6142,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.6659136383572304,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5072,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6642466476525241,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5833,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.6739057495507992,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5791,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.7885965752945121,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5353,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.6445600188035829,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5757,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.7771936690139558,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5482,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.8640028102870341,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5776,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.8537696027955403,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5515,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.7611476074331381,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5675,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.8473726338981905,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5755,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.6735417690011596,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5455,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.6581610879133376,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.5562,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.6602925595792949,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5621,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.6412583045124893,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5704,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.6676910527410658,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5715,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.6396643940402758,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5547,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.5612430425360088,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5656,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.6568859077411137,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5417,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.6372370526275795,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5663,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.8220373540097621,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5514,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5332621002726037,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5648,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.7771494812120194,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5668,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.6364443532863169,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5411,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.8655077078939486,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5955,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.6110694648869929,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.522,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.7180752756306167,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5528,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.7117239198636521,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5823,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.6887311804548444,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5408,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.671496955250415,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5667,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.6039597560683883,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5467,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.6923259451308797,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5685,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.731493725167053,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.6056,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.5667352613222394,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5201,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.6218736312043432,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5875,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.5448621495647457,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5358,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.7219478372447425,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5723,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.7128298254787774,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5426,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.8232471685016599,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5857,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.7826665547124606,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5718,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.7082486315985711,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5411,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.8215098563689994,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5943,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.6298010947506244,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5257,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.7621463386718327,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5656,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.6980472340161767,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5853,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.722213820745583,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5871,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.6557375761774639,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5457,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.7025268053352106,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5747,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.5957530219103704,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.5335,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 1.8340092625895414,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.6798,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.9255023140125495,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5621,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.95774912658146,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5292,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.9169068991600323,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5806,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 1.1153979013664317,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5844,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.7312392048826039,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5525,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 1.0565805561357058,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5612,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.6710569267073712,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5742,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 1.1760443303343189,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.602,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.6568028644519838,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5384,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 1.1702493921020287,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5681,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.6605240634393518,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5576,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 1.2453960873542615,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.575,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.7441715073364837,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5871,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 1.0552981331448383,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.6152,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.663639857543049,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5739,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.9260972195981446,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.5287,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.6581753670779013,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5712,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.7844478493169159,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5615,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.613982593934506,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5164,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.7295276664005954,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.6169,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.7444819468294975,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5438,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.567489044794829,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5458,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.8183706159065386,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5731,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6405384976991315,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.5821,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.7513311891344178,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5308,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.6047348216639132,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.5448,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.765764329026889,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5733,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6289231377047968,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.568,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.597603616402993,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5742,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.717019877566111,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5736,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.5288227297986532,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5603,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.6905006560699237,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5719,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.6780362780909833,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5486,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.5476687688439644,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.5603,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5567196958523314,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5422,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.6966894375356012,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.6079,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.4917072080336899,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5289,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.6770680386530243,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5891,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.6935093672456719,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5295,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.6291291121329543,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.6083,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.508751586798676,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5311,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.6657501315449745,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5651,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.5737156238471024,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.5427,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.6237409059139108,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5266,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.6334329313415804,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.607,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.7093872746520126,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5719,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.6980473705901628,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.541,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.6833738661215869,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.6011,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.7625897089204398,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.369,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 1.4316359490023816,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.4303,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.6344535116913013,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3786,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.729388091747656,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3727,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.6348989485949029,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3792,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.8255670977929186,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3718,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.7030672379907864,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4046,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.6902905938358156,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3811,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.6172655020096192,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.3533,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.6571245851556204,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3846,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.6619275957750437,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3576,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.6357503976732041,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3761,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.6572269642842662,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.4179,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.5278648921699683,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3549,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.7399174739256162,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3701,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.5066852683101184,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3674,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6529373836082569,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.4092,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.6079718752134982,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3739,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.6529015631261516,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3818,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.6237788876818301,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3638,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.660567385719633,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3611,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.7710580332671384,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.4152,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.5745212739465361,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3963,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.5542262334508926,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3621,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.5358911467854194,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.374,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.5778411939076513,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4148,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.5633604583353168,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.4218,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.6825599340229619,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.4079,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.470216858967785,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3562,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.587831982073496,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.3777,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.6048467899369364,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3975,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 18.906570374154988,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3649,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.7649690712389136,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.36,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.5181180779069353,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3742,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.6175761090439359,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3979,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 9.697818585109998,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.4075,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 1.002173328135301,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.4419,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.6752391293669133,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3787,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.7963679824287346,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4055,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.7587840298682507,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.4189,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.6682445498490205,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3688,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.6404408255768336,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3776,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.6379772391456385,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.4032,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.6126767896486455,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3531,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.7310137854812867,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4258,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.726261171862778,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3857,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.7242826243541973,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3792,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.6325964754744694,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3707,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.6033024311549873,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.4103,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.715898179665916,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3702,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.5853741570508363,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3725,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.6732063501670721,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3769,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.6309004580833547,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4074,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.5993379151906123,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3499,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.5738909329258459,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3563,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.5888129920512724,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.4168,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.6374837689961024,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3832,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.5730770463846706,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3979,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.6250449639062321,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3708,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.4860872842159768,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.361,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.6326474074626004,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.4094,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.529996632401308,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.388,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.5910049891072909,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3624,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.537181288042941,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3938,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.585517892500599,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3892,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.623164916690793,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.4221,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5744904686538714,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.359,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.5533498196229047,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.39,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5642034522874753,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4161,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.6158840502708854,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.4017,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5434325275634072,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3814,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.5950864728801188,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3676,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5180052844129626,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.4266,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.5765951460997074,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3838,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.5309096366202485,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.4003,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.5353374278266003,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3741,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.47279459326442913,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3593,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.6018602111886175,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4502,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.4491145969897193,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3593,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.5320689841682394,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.385,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5009326617521691,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.4082,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.48717090153050036,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3854,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.5383045285351928,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.379,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.5438867679245184,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.4036,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.48212306020591084,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3519,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.5101945267641953,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.4112,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.4937455341576383,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3974,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5177835882033723,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3724,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.5380172618188144,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.4158,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.5248454300799789,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3653,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.5668506310379279,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.4176,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.5610689702407655,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3877,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.4969644200377474,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4039,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.5500505148967617,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.397,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.5568012328992419,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3675,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5522305480548202,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.4053,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.5843688987837679,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.4145,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.5358496219849378,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.3729,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.5928618042892273,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4418,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.5233473276677844,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.4028,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.5444249455588194,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.3621,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.651526851973198,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.4033,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.5549186067325691,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3625,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.6586986954308948,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3893,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.5382285490793254,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4205,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.5173659226428041,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3726,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.6553165174547289,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4468,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.4804653415527197,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3629,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.5861807893596804,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3902,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.5150260052427114,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3979,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.5883377211488728,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3695,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.5280852261537755,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.4083,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5172064612469686,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3795,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.524631751228396,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3704,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.51810442565464,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4016,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.5006146497135356,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3536,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.6341797218403746,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4259,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.45823111128889804,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3711,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.5960697723298098,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.392,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5562114267965992,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3954,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.5478512972927672,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4195,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.5641028176209816,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.4178,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.46359672285433035,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3348,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5260385156802954,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3955,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.504763574059862,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3941,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.49819297472397267,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3574,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.5042881679160549,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3753,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5461766835138714,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3778,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.5671257924198184,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4102,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.4786730911718309,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3457,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.6351124990968883,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4116,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.5558000927655399,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3861,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.5434906706395559,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.376,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.6261891894809128,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4408,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.5496608619422381,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3417,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.5197072281498729,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.4031,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.6079866854586449,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.4112,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.46778793440298916,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3746,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.5854272224205149,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3647,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.5903234340088108,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4351,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5587583876436365,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3656,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.6154330456413252,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.3624,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.5664160613456989,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4103,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.5142614596063944,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3875,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.5419985066814837,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4252,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.4895843215919776,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3552,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.5278281350175645,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3819,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.5494226379129381,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.4569,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.484884740962125,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3704,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.5846825494298528,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.4004,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.4800766006714834,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3602,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.6529333301902879,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4065,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.49921414093771715,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3824,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.5406569283945122,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.4074,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.5116949180890786,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3848,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.4838817910513246,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3741,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.6169410765767716,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.4426,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.4911538230691412,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.359,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.5364001960900358,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3955,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.497637531269674,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3568,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.5588209837259581,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.4309,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.4531927666258215,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3878,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.5954875095130111,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3872,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.4768716104271764,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3658,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.5457674692926724,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3947,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.47616783901653476,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3852,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5529163475376057,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.398,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.48729413281255063,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3749,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5104668140052949,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3876,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.5744499696488149,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3965,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.49077765706540777,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.3836,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.5315904311630649,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3985,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.508245863914563,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.4031,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.527627216720768,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.4008,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.45511124405668923,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3452,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.4441388791275465,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.389,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.5350337905240362,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4117,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.47893640165601153,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3787,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.5824241531864736,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3936,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.5067452300601742,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3974,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.5560527920944833,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3886,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.5105682733160449,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.4036,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.48830068190846604,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3869,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.50432641991232,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3604,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.4627920730648686,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3578,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.5472247902862012,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4053,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.5277349019620381,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.4129,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.4277131653462255,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.356,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.5063109367868638,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4177,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.5096017242141625,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.4198,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.4847578930678669,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.399,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.4874111142355298,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3774,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.4943005908792025,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3562,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.6445064646472194,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3828,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.5171386041812184,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4147,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.49010534450930593,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3451,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.47809602742465906,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3812,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.5141687652725255,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3854,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.5627956883294601,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3939,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.43696776224074996,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3698,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.619190116264203,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4503,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.4694459348511639,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3804,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.4859762730860158,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4088,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.5466475046799746,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3828,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.46354992281997565,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.3762,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.5960832267076766,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.4076,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.4867297485253524,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3542,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.5956619745983336,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.4213,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.4741143443535949,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3929,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.5859000777688876,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3894,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.4733642126829241,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3544,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.4803028022633944,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3925,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.551136193555408,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.4284,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.498425629238251,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3758,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.46315291181971696,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.3503,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.5292815467567473,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3924,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4623968629781034,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.3871,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.5298593469108978,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3642,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.48930728992043737,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.391,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.4555519637158697,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.3993,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.479458888747569,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3884,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.49662349510646164,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3848,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.5407207283284984,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4036,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.5231583170938777,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3861,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.4792366494824683,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3953,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.5048521060206903,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3851,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.5321585018384783,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3921,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.5097757685484077,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.4146,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.49486170082632475,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3711,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.4825108756113067,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3884,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.5145347072975675,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.4081,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.4628194898347834,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3878,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.48351791275961925,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3704,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.5183993517930766,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3863,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.4673240793828853,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3952,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.5009419412385478,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3508,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.4748434974010108,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3722,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.46343977818198395,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3853,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.5402336207119388,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.4386,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.44588204289653666,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3379,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.5022814285846076,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.4082,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.5228596598730983,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3869,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.5344645267817227,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.4107,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.4367749390872429,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3418,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 1.2550238582634006,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4873,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.5741465636562313,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.4176,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.5495900763119381,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.3614,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.6787304912253599,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.2436,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.5598494078540845,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2652,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.7382722314936719,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2357,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.6723971784825493,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1964,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.684740484606861,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2674,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.5662503483339659,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.2406,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.6303766266145193,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2367,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.6352637061263454,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2484,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.5166715427239629,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2257,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.5215470934683709,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2327,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.5474316250908031,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2413,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.9717944046384966,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.2299,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.5330829513498196,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.2548,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.4571842311337979,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2266,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.4917400831694865,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2113,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.5616393328526137,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2359,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.48538902414079915,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2362,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.5439806721257621,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.2849,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.455084581021439,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2208,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.4757764637376813,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2447,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.45507695847927543,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2029,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.4749676995127813,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.2423,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.46285924994120115,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.2551,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.4479257138470686,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2185,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.44903153298377535,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.2067,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.4521927297760767,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2426,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.4622352257099564,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.2384,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.44547249237860653,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.226,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.48764230104961337,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2346,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.4715182324886628,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2241,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.484833606471315,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2426,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.49194858329430796,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.229,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.4164576801365161,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1911,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.4708144926569082,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2376,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.4944698523864472,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2244,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.5208257333439653,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2524,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.44195414076435796,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2322,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.5232990942957609,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2245,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.45118633812133,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2295,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.4531765414979338,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2377,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.49986363130976236,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2555,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.39884761920831696,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.2093,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.48257552056410286,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2629,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.43413125011509845,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2189,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.49823580482409,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2335,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.43046842925541,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.211,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.42581495115694934,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2388,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.46740544171081594,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.2098,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.496768358907478,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2398,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.4684296770931861,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2325,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.42258804897844543,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.4975868252380762,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2558,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.471596956682885,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.2459,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.43477054697925255,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2094,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.44906157187787565,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2205,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.4799393378816408,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.2301,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.4556277108661895,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.2292,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.4464881007296542,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2146,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.4627683195408911,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2307,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.4666557113669408,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2668,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.4015268722610826,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1988,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.4856608946610803,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2594,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.39430408920056026,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2216,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.40752505216303364,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2225,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.47302381626571954,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2377,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.45117461618264704,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.2355,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.40912158717655117,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.2165,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.4896130752501134,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2499,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.43925934818382384,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2192,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.4560290442872339,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2351,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.4649848914740813,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.238,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.4378381247961777,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.203,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.4414941500522263,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2079,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.44756089724066067,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2422,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.43247646108580967,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.2386,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.44288851665198903,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2262,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.4808799150259693,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2274,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.4314470400506906,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2103,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.4935734217784459,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.244,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.45636670467471807,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2244,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.412800769281848,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2405,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.4688732934546179,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.2227,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.4137723408900003,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1994,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.48530710209293904,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.254,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.4098750802840489,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2143,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.4687546960564646,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2323,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.4228382692525434,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2323,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.4192090484058214,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2035,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.4225935601931842,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2585,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.430363110861996,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.222,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.4552904808786116,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2302,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.43820044547628806,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2318,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.41807369332019106,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2182,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.4552290313649582,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2249,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.469687557600756,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2517,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.4359751531692739,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.2401,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.4322962480591651,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2248,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.4844992739170976,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2595,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.41061321121080163,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1807,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.47940726282951956,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2316,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.49326002912353745,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.2533,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.4197905780850633,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2082,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.465179283492556,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2379,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.40682971394164125,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.2315,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.40629833794118586,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.2159,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.47566814640042454,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2292,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.44985213373709315,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.2255,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.4304395645128596,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2341,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.465021982593026,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2611,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.4437615995467993,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.2199,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.4202982485437777,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.2025,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.48497842471103436,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2449,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.4362606838695637,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2203,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.42590837377049534,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2314,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.43032853720782793,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2378,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.42734453266699535,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2383,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.4335610045994669,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2335,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.4336334438480036,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.247,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.45067647980453707,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2289,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.5463207586104946,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2196,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.4519316405890648,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2284,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.4314259374639776,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.2448,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.4184564176375397,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2212,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.4377853200624375,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2338,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.42665268015282637,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2166,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.4211843155377372,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2203,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.5069633975745441,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.2251,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.4266492726593593,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.2056,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.4263505213280563,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.214,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.4552887980305618,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.2506,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.4289908555329577,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2234,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.4133440162383229,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2124,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.47270205389815095,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2399,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.40243504688330944,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2136,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.42209988318720815,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2518,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.42489239873843576,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2435,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.3968710991754393,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2016,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.43138048302812154,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.222,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.41705091783009757,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2494,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.4211917781165425,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.2001,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.40082307469592526,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2077,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.4396250354170472,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2335,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.4349437702729493,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2521,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.4206102830203694,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2475,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.40820775813266147,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2223,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 1.136897055687294,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.3009,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.4250520913154719,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2288,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.3834880729663183,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.2074,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.4459003869926759,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2574,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.4227347619357332,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2278,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.3985743458662194,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2178,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.44423926934989544,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2242,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.4161160508733085,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2462,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.4017394466825577,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.2204,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.4253815249267584,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.2356,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.4227527949335672,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.2251,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.4346244075142884,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2415,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.41216948661696345,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2499,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.4332509100441981,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2077,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.42549265466852954,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2534,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.4093097665559732,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2169,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.40204347109598426,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.2192,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.4168622213999708,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2367,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.38091880174260806,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.2122,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.42143352696408926,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.2506,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.42396705072838614,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2138,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.40535605059992036,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.2263,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.4121351349639759,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.2229,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.40742797061190983,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2326,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.4127135055603271,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2316,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.4080472459863216,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2404,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.39126924161200954,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.214,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.4436901007108771,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2367,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.4417703380768508,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2398,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.39601551895849296,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2196,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.41387785085282924,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2451,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.4385478321839333,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2566,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.4374320796896465,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.2255,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.3981166292587963,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2133,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.40757077836085664,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2334,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.3971546846408769,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2031,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.4438493555693312,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2369,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.42309025401150335,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2489,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.40713042422789697,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2166,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.4518532248635652,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2578,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.40415146909081073,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.2138,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.43865612433445045,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2228,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.4280859663403088,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.2503,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.40003023840390245,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.22,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.4181546527966514,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.2269,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.4146170888600579,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2379,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.38499467476923416,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1948,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.407815552495595,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.2468,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.395061482566682,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2244,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.4115675680418675,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2504,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.409387353878462,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2148,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.4410705725754826,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2569,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.4159406745291562,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2267,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.4094619627247696,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2316,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.40177020656235907,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.208,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.4322401207356922,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.2402,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.3990418551214266,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2206,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.437093230777737,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2312,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.41352476015193806,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2253,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.3898646128512555,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.203,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.4246927942664762,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.2509,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3953396523241138,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2199,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.45718614589054235,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.2706,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.35829815008746435,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1785,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.41266568210611043,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.224,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.3748516122008536,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.201,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.4297108864110368,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.2493,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.43654125886937306,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.2131,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.3908801734280822,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.21,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.43871399811260753,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2528,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.4117928325229831,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.2206,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.4277601631151307,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.2115,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.41466439815095996,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2115,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.42267376766687104,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2393,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.40995479707647536,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2197,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.37985166171417134,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2151,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.37978038752920834,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2151,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.38737529427266637,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2133,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.3882860156476199,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2231,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.4233108924819289,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.276,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.3927187038674386,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.2256,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.38781498894982463,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.2031,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.4185245717860918,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.2548,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.42224870895821853,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2439,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.3833131917481219,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2212,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.3853784350557668,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.224,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.3971624143446534,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.2198,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.41555430870568244,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2237,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.4252833813332893,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2389,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.44617396479631216,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.2493,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.3740201540851861,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.216,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.4144576888250687,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.225,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.4439017609380389,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2319,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.4136962455478009,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2161,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.4030028946571779,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2476,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.372203331798868,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2152,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.40475978632097664,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2131,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.43507249879843574,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2356,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.40063207855490124,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.2258,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.38782834355928314,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2345,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.4018437804418386,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2302,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.4819262950905187,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1922,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.5492890991882737,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1306,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.5349458929628252,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1678,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.3829833045894242,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1434,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.34527128901758103,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1344,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.4080045763655604,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1365,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.5170360421310454,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1472,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.5791420219915356,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1298,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.6177971685305432,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.1538,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.5302053181773414,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1568,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.40366242129450997,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1268,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.43734650793641344,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1477,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.39502061975180514,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1298,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.4047000697917939,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1375,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.42169322220720745,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1273,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.4274998986478872,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1457,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.42082225934142836,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1413,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.3658016337269604,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.1281,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.3716248038786051,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1273,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.402582608302965,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1517,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.4028883466675116,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.15,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.4019083671089175,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.146,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.3894623691696887,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1393,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.37040865955368946,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1249,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.5650727465096888,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.2011,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.34957813214777705,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1195,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.3766018862746714,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1344,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.35984727276331346,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1327,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.3572665168075948,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1305,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.3824804190232189,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1376,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.34608470248278755,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1225,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.3638945109302099,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1435,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.3610557712930503,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1277,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.3595713423011219,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.1269,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.37162722230486966,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1369,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.3258877988084875,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1272,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.3766479010106916,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1597,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.34641702595913004,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1168,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.37878813201551437,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1439,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.35092879246078806,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1151,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.3893153019172687,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.141,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.3379964119667981,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1123,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.3406969860041264,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1232,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.37549486523766445,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1668,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.36877372251940926,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.144,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.33335219426560003,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1319,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.34048921120940756,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1257,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.3592634403945403,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1359,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.36477952310310935,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1437,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.33956562494856923,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1278,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.43762577069813147,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1796,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.3253907656855415,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1076,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.34311705923407354,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1288,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.3514327773914306,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.1292,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.35558795875222526,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1314,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.35997579671391505,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1417,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.3416798987323003,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1274,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.3288669273625737,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1331,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.3448109441458289,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1349,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.33667069848210673,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1219,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.36070477543209895,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1411,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.3271335862136915,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1173,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.38358057929766154,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1598,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.3231832911921822,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.0954,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.3742602150859742,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1358,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.3515082215488428,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1221,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.3811028779434799,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1341,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.33207056667480117,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1023,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.3305224348299434,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.115,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.34634738363251,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1266,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.3573473166990008,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1352,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.3838768212098692,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1525,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.35847024111630976,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1483,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.32555549563953945,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1157,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.35385528647988346,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1393,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.3580016341819149,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.126,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.3441929174746828,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.112,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.3745365601228441,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1379,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.3376554090862444,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1269,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.3704383391356872,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1631,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.34455316370620265,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1364,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.3445494066530414,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.1388,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.3494745756355801,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1295,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.36309554735072114,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1313,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.3488584110955264,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1173,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.3403826576279495,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1284,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.3715104450116425,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1561,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.35091801136063283,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.143,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.3566416874544265,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1406,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.3328647786169498,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1303,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.36335602040790144,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1537,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.34465619656747387,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1314,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.3851898776285608,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.1649,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.31757152718500076,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1136,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.33652003009292714,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1287,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.3451515184325092,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1437,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.3580482880116293,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1515,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.32282131893527827,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1174,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.36257440070211694,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1431,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.3415609678492482,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1266,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.35072600463188947,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1438,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.34563272873150086,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1502,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.3430315759564778,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1341,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.332955852913487,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1184,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.3591158555281369,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1295,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.36508764309752384,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1486,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.3628089121113883,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.133,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.35505737585716657,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1381,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.35335722171483813,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1258,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.35676382378148774,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1379,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.35642781808782914,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.13,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.3520757765703348,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1439,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.32571855310771625,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1256,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.35307324722125305,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.1387,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.35985824713883025,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1229,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.34587553792476083,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.121,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.35754338095260296,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1487,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.35374749711135695,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1246,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.33637461797777063,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1196,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.3489925810564436,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1363,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.3723551236346331,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.1451,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.3491230830744308,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1166,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.34251719702382316,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1164,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.3688094484695341,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1516,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.3480031797080519,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1408,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.31645024657386817,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1198,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.37831926596824766,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1704,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.3493359716834648,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1389,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.4360951556392713,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1365,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.3516557770220002,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1148,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.3578776447365938,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1368,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.34249945351885785,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.125,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.3458419492919962,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1312,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.3616772861604499,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1435,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.3573866117895114,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1474,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.3584172588775177,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1477,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.31049433941096355,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1128,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.34853871442523326,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1375,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.36179072031974685,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.1435,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.35771964134389495,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1332,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.3515240694803857,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1385,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.35906884000978107,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1389,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.33080282233276864,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1081,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.3726692516416002,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1418,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.3698893931140506,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1365,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.3489462836165819,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1268,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.36639985604568054,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1336,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.35571801080301696,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1385,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.3305014454639237,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1085,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.37365780212103583,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1432,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.3265846936907944,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1325,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.31604700180903467,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.124,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.3497891149920304,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1446,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.32474767511444563,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1203,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.36408624478828405,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1412,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.3621376901507701,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1243,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.33530556658682237,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1233,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.33965866531409844,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1218,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.33792977240537153,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1242,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.36706351688274963,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1454,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.33796253430478707,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1302,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.3270861844585207,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1261,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.32869648835392806,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1288,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.35869787488901106,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1377,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.3750641161737085,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1501,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.3557092752489101,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.133,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.3489265222972943,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1358,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.3718967854946084,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1598,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.33074612345394666,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1287,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.35069087183768627,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.138,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.326565935292905,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.121,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.36786373936499867,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1354,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.3288005098786737,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.1201,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.35455935012457274,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1279,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.33395890544288526,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1358,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.3349747781175772,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.1356,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.3646766792123617,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1427,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.33911126631200833,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.123,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.35580579619842223,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.143,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.3244547992936777,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1255,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.35043741369104797,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1243,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.34176492536792363,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1248,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.3578820946405256,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1328,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.33813105703361135,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1147,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.3593937548126975,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1381,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.3618551954370005,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1359,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.3243778860027688,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.1173,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.3441883437424383,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1301,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.3549464436160542,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1198,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.3661814576737508,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1482,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.35516719023203835,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1322,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.3342935426395882,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1374,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.3249786939585011,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1296,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.3485674367309886,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1383,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.3380193927191882,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1398,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.33817244260641827,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.148,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.3434977271037415,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1405,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.3742061955662875,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1632,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.31416043633210444,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1132,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.3561068119583198,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1498,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.3476739811191006,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1434,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.3344815807054645,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1282,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.35436512488101596,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1372,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.3233408101608997,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1197,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.3378428966250198,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1428,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.3505124452649678,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1325,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.36105561756076715,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1469,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.3292028001399058,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.12,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.3808350843241436,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1685,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.3296984003283629,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1245,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.3403997093051277,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1221,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.33695522610327466,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1301,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.33567161840186976,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1314,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.3439468653929013,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1474,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.3369715411245861,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1382,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.33731766605384383,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1396,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.3286969122756583,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1448,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.34557347026284696,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1326,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.3067382102806767,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.0997,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.37549252671378763,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1578,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.32263771224620036,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1256,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.33976940673140554,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1307,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.3387555380222364,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1392,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.33137984830505146,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.132,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.3553103908591017,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1426,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.3289145993315524,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1346,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.354305295761637,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.1438,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.3582213572744067,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1579,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.33432330157336126,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1367,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.3791201361357769,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.178,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.32984559283242926,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1308,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.34180982521701403,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1417,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.3278610115330221,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1268,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.35030549195891464,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1423,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.3469745201753895,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1276,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.3671951287950301,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1542,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.3337613714642892,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1272,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.35689846272522774,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1339,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.3594008327911265,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1382,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.3343435028083343,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1146,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.35530821604136115,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.16,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.32992249418261993,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1401,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.3501325839343482,
+      "learning_rate": 0.0,
+      "loss": 0.1459,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 1.489604730102481e+18,
+      "train_loss": 0.4123472108043791,
+      "train_runtime": 72459.5524,
+      "train_samples_per_second": 2.181,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.489604730102481e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}