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": 4.992911392405063,
+  "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": 6.565896938639643,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.0739,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.424877036189121,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.0304,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.675688024565499,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.0832,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.352567655471716,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.0432,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.1394666841542636,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.0445,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.872600490506451,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.0375,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.515773706171991,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.9992,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.200488354646107,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.0013,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.7539299966289255,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.9747,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.5384131132753165,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9538,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.3940329365895665,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.0117,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.7615522437070803,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9406,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.0566637383891235,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.9513,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.9336267477151248,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.9571,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.6469304032393213,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9393,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.792596920083004,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.8866,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.62344696439575,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.894,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.1098626475414073,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.8684,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.6178665581895086,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8719,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.1895765665603557,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.8529,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.0218687142914817,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8539,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.1847985232807718,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.8475,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.1277919285473021,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8133,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.0140428585227534,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.8151,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.9087338712275635,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.7938,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.9235267994779454,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.7922,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.8441016532415221,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.7916,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.9591795712065868,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.7853,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.804902855700884,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.7977,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.7975879463325245,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7747,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.7974681707822487,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7577,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.8343569527571383,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.7612,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.6694182326150714,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.7591,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8789379796682085,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.7603,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.6445767575441089,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.7405,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.6770247955397397,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.7398,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.6858280157354305,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.7456,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6403279583365179,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.7243,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6341403951861435,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.748,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.672016954235725,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.7496,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.6133279680326725,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7449,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.4807301217267841,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.7462,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.6426264779936044,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.7375,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.5349486950803585,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.7165,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6320496357729215,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.7315,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6111068775955827,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.7083,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.4911821136694014,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.7036,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.5749373815175803,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.7249,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.514168465484637,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.7134,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.5850932044942065,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7144,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.4904677239469707,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7287,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.5161204780147062,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7019,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.4749012557029868,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.6939,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.5739663936961527,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7005,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.5839546976448743,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.7156,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.45211125628125676,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.688,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.5440773051379888,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.6982,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.49066042383790304,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.7108,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.5404332535023633,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7035,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.5749934615563418,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6948,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.6250440749982804,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6895,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.5479702576267381,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.6909,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.5840045603517738,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6817,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.574793850525269,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.6924,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.6676237052563165,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.692,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.9619934541334264,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.6918,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.060898563576492,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.6791,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.904861265958466,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.7028,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.1853527291040817,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.6812,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.7857110585191401,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6866,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.9682602431491651,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.6885,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 1.4669981427432603,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.6984,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.633454232355496,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.6999,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.5971751443323539,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.6849,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.5966050905818251,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.6729,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.2296293856099136,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.6686,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.8070733998144802,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.6824,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.8445446872314081,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.6755,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9463490141931428,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.6931,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.27968689699765,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.6918,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.9643659417765228,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6769,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.0082010830767494,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6893,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9118520007011225,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.6797,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.8467329391054232,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.683,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.0281287680667526,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.669,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.4383431261117385,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.6747,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.0668810949475278,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.6786,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.3109536031574585,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6762,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.3902022527615097,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.652,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.0406067055206762,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6741,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.7199696920849745,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.6857,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.9729377453147651,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.6656,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 2.382657070863399,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6775,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 2.028553778397811,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.6873,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.3478551809928314,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6775,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.374453027501157,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.6735,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.2981715944174452,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.6808,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.068922695797907,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.665,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.1306548785963957,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.679,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.0620355030662811,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.6743,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.9440101070458233,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.6601,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.9872994254263004,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.6716,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.756761571024137,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7071,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.006814653029631,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.6734,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.0137736147627878,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6594,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.5249612663871746,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.6566,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.8150255853333629,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.6433,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.2585197395770062,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.6667,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.6565133256429025,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6759,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.007712038975302,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.6476,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.9229767494277894,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.6556,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.301454435867178,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.6657,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.7229560416105487,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6598,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.4396399477099036,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6713,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.2966439069361446,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.6661,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.0897664154775006,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.6636,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.227331278015716,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6611,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.9013064644008313,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6778,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.2327668990972602,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.6623,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.009995390725031,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.6718,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.122273097665521,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.6671,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.0498199056181385,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6571,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.2170239936489256,
+      "learning_rate": 4e-05,
+      "loss": 0.6619,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.0109905907140755,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.6451,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.9530381307445761,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.6567,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.0405593161569147,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.675,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.1189545066408981,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.6481,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.2534782225609422,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.6619,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.5144075297452226,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.6738,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.0761651131202516,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.673,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.1779176326124219,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.6562,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.3185457139977619,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.6581,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.019959758162389,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6619,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.3504252924134732,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.6587,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.722263604782836,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.6524,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.8938087297449309,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.6446,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.7993275874231304,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.664,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 2.374139086546867,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.6644,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.7230492616698447,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.653,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 2.0119475440335752,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6692,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.9389135348306736,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6535,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.4403349978255873,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.6748,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.530297008864757,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.6467,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.1291502250770709,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.6451,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.4895931863903717,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.6516,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.8277779348441653,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.645,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.400029223064251,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.6358,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.8151485747909567,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.6508,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.127917981409227,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.6461,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.8170674425202605,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6445,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.0560163118847938,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.6611,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8392257720896447,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.654,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.8421618110581768,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.6504,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.928022090820385,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6764,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.0583201526437274,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.6634,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.8148031316339325,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.6323,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.7237709529946899,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.6474,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6692226805131555,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.6368,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.6527067680990782,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.6373,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.6955730523626112,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.6483,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.9073792810634957,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.7331,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.7039249507523281,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.6445,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.7919618526474105,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.654,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.810962163186228,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6469,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.5942821513181452,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6352,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.6195308429522167,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.635,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.5239498374223093,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.6609,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.8155264405765932,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.6426,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.8064905763534658,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6496,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.795942110772038,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.6547,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.5159545723250966,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6193,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.6189667183317581,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6425,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.7584327560828357,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.6354,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.8269004642838677,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.6342,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.5983073457884152,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.6396,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7902187490402278,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.6478,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.8787042951906945,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.6399,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.7777724631014674,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6429,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.5027376283406825,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.6314,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.5876533354709571,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.6343,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.656008461780158,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.6337,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.627673955333263,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.6457,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.5869923804252781,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.6282,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.5979340937404183,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6365,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6717037876248872,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6419,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.6486136631874181,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.6325,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.4775678489789169,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.6521,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.4604126451601245,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.6359,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.49239548471778155,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.6456,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.5089155791039236,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6386,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.5429054262226821,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.6326,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.5161100801700521,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.6373,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.48923838383485607,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.6118,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6129248980503034,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6438,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.6625179518220522,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.6398,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.5054913981384187,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6351,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.5632102123455754,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.6235,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6864930362174487,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.6414,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.43503390222300553,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.6328,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.5128692063751873,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.6444,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.5808243467504917,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.6433,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.4536090142706016,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.6345,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.46078668530330275,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6368,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.4542042648346198,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.6337,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.6116190084199781,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6399,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.4686618053440677,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6337,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.5465812813619887,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6268,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.683566931403071,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.6194,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.8456912417524167,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6421,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.7005968590348535,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.6126,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.5752150912641875,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.6277,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.6196384020142428,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6275,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.6488415715968038,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.6309,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.6982459827435065,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.6337,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.5688983251630224,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.6343,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.4160966225198047,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.6239,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.524669697339945,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.6178,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.5152289480745182,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.6369,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.473699341767848,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.646,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.46064652637604775,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.6192,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.4242902276099433,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.6303,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.49401445340889244,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.6236,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.4825647028044612,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6107,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.4648926720618138,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.62,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.5764704044149312,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.6264,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.6926696416581806,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.645,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.5661594135986946,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.6381,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.3772105422739256,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.6181,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.48879146836297693,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.6301,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 1.1850082021436543,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.6246,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.44330537384560836,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.6257,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.5515900442430017,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.6292,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.6333829354802747,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.6223,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.625907461156176,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.628,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.5912875446488828,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.6388,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 2.534772848596236,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.6504,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7527199049628908,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.6302,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 1.195053650025741,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.6402,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.5980244768520538,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.6091,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.9985418414588404,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.6147,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 1.2747212246697324,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6365,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.9805542738778488,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.6365,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 1.2598365025741,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.6164,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.8484777556375804,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.6414,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 1.3333495679797724,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.6306,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.8805704782012573,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6336,
+      "step": 246
+    },
+    {
+      "epoch": 1.0015189873417722,
+      "grad_norm": 2.5265983356658173,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.6258,
+      "step": 247
+    },
+    {
+      "epoch": 1.0055696202531645,
+      "grad_norm": 2.2481644964234246,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.6036,
+      "step": 248
+    },
+    {
+      "epoch": 1.009620253164557,
+      "grad_norm": 0.6970741170778451,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5784,
+      "step": 249
+    },
+    {
+      "epoch": 1.0136708860759494,
+      "grad_norm": 1.7404018816795028,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.6056,
+      "step": 250
+    },
+    {
+      "epoch": 1.0177215189873419,
+      "grad_norm": 2.694243508496181,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.6644,
+      "step": 251
+    },
+    {
+      "epoch": 1.0217721518987342,
+      "grad_norm": 0.8196871020301907,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5919,
+      "step": 252
+    },
+    {
+      "epoch": 1.0258227848101267,
+      "grad_norm": 1.3237796521089902,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5998,
+      "step": 253
+    },
+    {
+      "epoch": 1.029873417721519,
+      "grad_norm": 1.3049962364368444,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5928,
+      "step": 254
+    },
+    {
+      "epoch": 1.0339240506329115,
+      "grad_norm": 0.5818349333370841,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5828,
+      "step": 255
+    },
+    {
+      "epoch": 1.0379746835443038,
+      "grad_norm": 0.97063078608405,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5959,
+      "step": 256
+    },
+    {
+      "epoch": 1.0420253164556963,
+      "grad_norm": 0.7713166513645306,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5953,
+      "step": 257
+    },
+    {
+      "epoch": 1.0460759493670886,
+      "grad_norm": 0.8942459730908692,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.6097,
+      "step": 258
+    },
+    {
+      "epoch": 1.050126582278481,
+      "grad_norm": 0.8941261968922476,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.6028,
+      "step": 259
+    },
+    {
+      "epoch": 1.0541772151898734,
+      "grad_norm": 0.7355014858586207,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5955,
+      "step": 260
+    },
+    {
+      "epoch": 1.058227848101266,
+      "grad_norm": 0.8129664540326133,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5837,
+      "step": 261
+    },
+    {
+      "epoch": 1.0622784810126582,
+      "grad_norm": 0.5650176145200243,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5856,
+      "step": 262
+    },
+    {
+      "epoch": 1.0663291139240507,
+      "grad_norm": 0.6848721166166323,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5911,
+      "step": 263
+    },
+    {
+      "epoch": 1.070379746835443,
+      "grad_norm": 0.5268100411296476,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5651,
+      "step": 264
+    },
+    {
+      "epoch": 1.0744303797468355,
+      "grad_norm": 0.5792215769193122,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.581,
+      "step": 265
+    },
+    {
+      "epoch": 1.0784810126582278,
+      "grad_norm": 0.9505730200739876,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5807,
+      "step": 266
+    },
+    {
+      "epoch": 1.0825316455696203,
+      "grad_norm": 0.5370718196680457,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5925,
+      "step": 267
+    },
+    {
+      "epoch": 1.0865822784810126,
+      "grad_norm": 0.5827574821201854,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5742,
+      "step": 268
+    },
+    {
+      "epoch": 1.0906329113924051,
+      "grad_norm": 0.5656627540660353,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5819,
+      "step": 269
+    },
+    {
+      "epoch": 1.0946835443037974,
+      "grad_norm": 1.0869401351883203,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.64,
+      "step": 270
+    },
+    {
+      "epoch": 1.09873417721519,
+      "grad_norm": 0.5212890436046597,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5982,
+      "step": 271
+    },
+    {
+      "epoch": 1.1027848101265822,
+      "grad_norm": 0.541285450616523,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5828,
+      "step": 272
+    },
+    {
+      "epoch": 1.1068354430379748,
+      "grad_norm": 0.6414610142018482,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.579,
+      "step": 273
+    },
+    {
+      "epoch": 1.110886075949367,
+      "grad_norm": 0.46427280043738534,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.598,
+      "step": 274
+    },
+    {
+      "epoch": 1.1149367088607596,
+      "grad_norm": 0.4966969238220982,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5885,
+      "step": 275
+    },
+    {
+      "epoch": 1.1189873417721519,
+      "grad_norm": 0.5338345494964832,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5929,
+      "step": 276
+    },
+    {
+      "epoch": 1.1230379746835444,
+      "grad_norm": 0.5701820414699905,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.587,
+      "step": 277
+    },
+    {
+      "epoch": 1.1270886075949367,
+      "grad_norm": 0.42826027474842443,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5812,
+      "step": 278
+    },
+    {
+      "epoch": 1.1311392405063292,
+      "grad_norm": 0.46087228282886944,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5863,
+      "step": 279
+    },
+    {
+      "epoch": 1.1351898734177215,
+      "grad_norm": 0.38588376992106554,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5707,
+      "step": 280
+    },
+    {
+      "epoch": 1.139240506329114,
+      "grad_norm": 0.38219530362553844,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5748,
+      "step": 281
+    },
+    {
+      "epoch": 1.1432911392405063,
+      "grad_norm": 0.39926936207110575,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5853,
+      "step": 282
+    },
+    {
+      "epoch": 1.1473417721518988,
+      "grad_norm": 0.39486826691757215,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5721,
+      "step": 283
+    },
+    {
+      "epoch": 1.151392405063291,
+      "grad_norm": 0.4530402706844666,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5753,
+      "step": 284
+    },
+    {
+      "epoch": 1.1554430379746836,
+      "grad_norm": 0.36848362424424375,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5666,
+      "step": 285
+    },
+    {
+      "epoch": 1.159493670886076,
+      "grad_norm": 0.3939714830086508,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.563,
+      "step": 286
+    },
+    {
+      "epoch": 1.1635443037974684,
+      "grad_norm": 0.4185048527791837,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5848,
+      "step": 287
+    },
+    {
+      "epoch": 1.1675949367088607,
+      "grad_norm": 0.3779768398992768,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5755,
+      "step": 288
+    },
+    {
+      "epoch": 1.1716455696202532,
+      "grad_norm": 0.397211278485165,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5723,
+      "step": 289
+    },
+    {
+      "epoch": 1.1756962025316455,
+      "grad_norm": 0.36036921811835726,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5873,
+      "step": 290
+    },
+    {
+      "epoch": 1.179746835443038,
+      "grad_norm": 0.3952858292291357,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5784,
+      "step": 291
+    },
+    {
+      "epoch": 1.1837974683544303,
+      "grad_norm": 0.3830737172540273,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5737,
+      "step": 292
+    },
+    {
+      "epoch": 1.1878481012658229,
+      "grad_norm": 0.3918236224719613,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5858,
+      "step": 293
+    },
+    {
+      "epoch": 1.1918987341772151,
+      "grad_norm": 0.3358760445223252,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5811,
+      "step": 294
+    },
+    {
+      "epoch": 1.1959493670886077,
+      "grad_norm": 0.364141189926486,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5742,
+      "step": 295
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6399048279620578,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.6369,
+      "step": 296
+    },
+    {
+      "epoch": 1.2040506329113925,
+      "grad_norm": 0.4026373729251476,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.568,
+      "step": 297
+    },
+    {
+      "epoch": 1.2081012658227848,
+      "grad_norm": 0.36953722405152084,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5781,
+      "step": 298
+    },
+    {
+      "epoch": 1.2121518987341773,
+      "grad_norm": 0.3621272665795014,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5828,
+      "step": 299
+    },
+    {
+      "epoch": 1.2162025316455696,
+      "grad_norm": 0.41254265012021796,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5724,
+      "step": 300
+    },
+    {
+      "epoch": 1.220253164556962,
+      "grad_norm": 0.45427768057765006,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5691,
+      "step": 301
+    },
+    {
+      "epoch": 1.2243037974683544,
+      "grad_norm": 0.49130353756910294,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5691,
+      "step": 302
+    },
+    {
+      "epoch": 1.228354430379747,
+      "grad_norm": 0.4382766111443482,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5799,
+      "step": 303
+    },
+    {
+      "epoch": 1.2324050632911392,
+      "grad_norm": 0.42072283804106375,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5735,
+      "step": 304
+    },
+    {
+      "epoch": 1.2364556962025317,
+      "grad_norm": 0.400473086382162,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5747,
+      "step": 305
+    },
+    {
+      "epoch": 1.240506329113924,
+      "grad_norm": 0.3609719734612689,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5629,
+      "step": 306
+    },
+    {
+      "epoch": 1.2445569620253165,
+      "grad_norm": 0.5073168238187583,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5756,
+      "step": 307
+    },
+    {
+      "epoch": 1.2486075949367088,
+      "grad_norm": 0.3809375487334342,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5758,
+      "step": 308
+    },
+    {
+      "epoch": 1.2526582278481013,
+      "grad_norm": 0.4932504038487977,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5683,
+      "step": 309
+    },
+    {
+      "epoch": 1.2567088607594936,
+      "grad_norm": 0.6909279206928767,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5757,
+      "step": 310
+    },
+    {
+      "epoch": 1.2607594936708861,
+      "grad_norm": 0.44384065428308855,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5869,
+      "step": 311
+    },
+    {
+      "epoch": 1.2648101265822784,
+      "grad_norm": 0.4808239799238354,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5727,
+      "step": 312
+    },
+    {
+      "epoch": 1.268860759493671,
+      "grad_norm": 0.4249471900763035,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5787,
+      "step": 313
+    },
+    {
+      "epoch": 1.2729113924050632,
+      "grad_norm": 0.5059898678828224,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5784,
+      "step": 314
+    },
+    {
+      "epoch": 1.2769620253164558,
+      "grad_norm": 0.36890927100868337,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5813,
+      "step": 315
+    },
+    {
+      "epoch": 1.281012658227848,
+      "grad_norm": 0.42318635646802805,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.577,
+      "step": 316
+    },
+    {
+      "epoch": 1.2850632911392406,
+      "grad_norm": 0.34241492575118926,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5809,
+      "step": 317
+    },
+    {
+      "epoch": 1.2891139240506329,
+      "grad_norm": 0.36325943216014706,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5711,
+      "step": 318
+    },
+    {
+      "epoch": 1.2931645569620254,
+      "grad_norm": 0.4921325337403487,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.5708,
+      "step": 319
+    },
+    {
+      "epoch": 1.2972151898734177,
+      "grad_norm": 0.3916245950553964,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5659,
+      "step": 320
+    },
+    {
+      "epoch": 1.3012658227848102,
+      "grad_norm": 0.3405700001935763,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5724,
+      "step": 321
+    },
+    {
+      "epoch": 1.3053164556962025,
+      "grad_norm": 0.46329867186771023,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5774,
+      "step": 322
+    },
+    {
+      "epoch": 1.309367088607595,
+      "grad_norm": 0.4367737028991729,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5752,
+      "step": 323
+    },
+    {
+      "epoch": 1.3134177215189873,
+      "grad_norm": 0.4386732527192421,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5946,
+      "step": 324
+    },
+    {
+      "epoch": 1.3174683544303798,
+      "grad_norm": 0.43535608173207146,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5856,
+      "step": 325
+    },
+    {
+      "epoch": 1.321518987341772,
+      "grad_norm": 0.3686485160669868,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.557,
+      "step": 326
+    },
+    {
+      "epoch": 1.3255696202531646,
+      "grad_norm": 0.39394713290083455,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5668,
+      "step": 327
+    },
+    {
+      "epoch": 1.329620253164557,
+      "grad_norm": 0.42799419674768996,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5666,
+      "step": 328
+    },
+    {
+      "epoch": 1.3336708860759494,
+      "grad_norm": 0.3633642263412647,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5679,
+      "step": 329
+    },
+    {
+      "epoch": 1.3377215189873417,
+      "grad_norm": 0.326582172597283,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5868,
+      "step": 330
+    },
+    {
+      "epoch": 1.3417721518987342,
+      "grad_norm": 0.41612279959130416,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5738,
+      "step": 331
+    },
+    {
+      "epoch": 1.3458227848101265,
+      "grad_norm": 0.43462062154179326,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.5774,
+      "step": 332
+    },
+    {
+      "epoch": 1.349873417721519,
+      "grad_norm": 0.40455081261645787,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5705,
+      "step": 333
+    },
+    {
+      "epoch": 1.3539240506329113,
+      "grad_norm": 0.3736869264225449,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5925,
+      "step": 334
+    },
+    {
+      "epoch": 1.3579746835443038,
+      "grad_norm": 0.4430975534899666,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5874,
+      "step": 335
+    },
+    {
+      "epoch": 1.3620253164556961,
+      "grad_norm": 1.0189664958190048,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5773,
+      "step": 336
+    },
+    {
+      "epoch": 1.3660759493670886,
+      "grad_norm": 0.48875718723687167,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5979,
+      "step": 337
+    },
+    {
+      "epoch": 1.370126582278481,
+      "grad_norm": 0.3463656250565052,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5808,
+      "step": 338
+    },
+    {
+      "epoch": 1.3741772151898735,
+      "grad_norm": 0.44137718323205477,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5766,
+      "step": 339
+    },
+    {
+      "epoch": 1.3782278481012658,
+      "grad_norm": 0.34985502503494387,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.576,
+      "step": 340
+    },
+    {
+      "epoch": 1.3822784810126583,
+      "grad_norm": 0.5821872820183871,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5824,
+      "step": 341
+    },
+    {
+      "epoch": 1.3863291139240506,
+      "grad_norm": 0.3730994503739556,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5705,
+      "step": 342
+    },
+    {
+      "epoch": 1.390379746835443,
+      "grad_norm": 0.3272432308051581,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5573,
+      "step": 343
+    },
+    {
+      "epoch": 1.3944303797468354,
+      "grad_norm": 0.7127634010482963,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5866,
+      "step": 344
+    },
+    {
+      "epoch": 1.3984810126582279,
+      "grad_norm": 0.377930933824937,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5852,
+      "step": 345
+    },
+    {
+      "epoch": 1.4025316455696202,
+      "grad_norm": 0.4335216318983424,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.566,
+      "step": 346
+    },
+    {
+      "epoch": 1.4065822784810127,
+      "grad_norm": 0.3993077602229474,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5674,
+      "step": 347
+    },
+    {
+      "epoch": 1.410632911392405,
+      "grad_norm": 0.3821932469446836,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5834,
+      "step": 348
+    },
+    {
+      "epoch": 1.4146835443037975,
+      "grad_norm": 0.8872312418284577,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.583,
+      "step": 349
+    },
+    {
+      "epoch": 1.4187341772151898,
+      "grad_norm": 0.39572585566474605,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5699,
+      "step": 350
+    },
+    {
+      "epoch": 1.4227848101265823,
+      "grad_norm": 0.4314834289193324,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5868,
+      "step": 351
+    },
+    {
+      "epoch": 1.4268354430379746,
+      "grad_norm": 0.4343292769030793,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5765,
+      "step": 352
+    },
+    {
+      "epoch": 1.4308860759493671,
+      "grad_norm": 0.3713590404033252,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5951,
+      "step": 353
+    },
+    {
+      "epoch": 1.4349367088607594,
+      "grad_norm": 0.44242807680152524,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.591,
+      "step": 354
+    },
+    {
+      "epoch": 1.438987341772152,
+      "grad_norm": 0.4696216029325405,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5658,
+      "step": 355
+    },
+    {
+      "epoch": 1.4430379746835442,
+      "grad_norm": 0.37964490771073856,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.575,
+      "step": 356
+    },
+    {
+      "epoch": 1.4470886075949367,
+      "grad_norm": 0.45873216441443215,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5727,
+      "step": 357
+    },
+    {
+      "epoch": 1.451139240506329,
+      "grad_norm": 0.4092841609229761,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5694,
+      "step": 358
+    },
+    {
+      "epoch": 1.4551898734177215,
+      "grad_norm": 0.3911420062015203,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5575,
+      "step": 359
+    },
+    {
+      "epoch": 1.459240506329114,
+      "grad_norm": 0.460331058224917,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5647,
+      "step": 360
+    },
+    {
+      "epoch": 1.4632911392405064,
+      "grad_norm": 0.4595744589630227,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5821,
+      "step": 361
+    },
+    {
+      "epoch": 1.4673417721518986,
+      "grad_norm": 0.44040955878200816,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5744,
+      "step": 362
+    },
+    {
+      "epoch": 1.4713924050632912,
+      "grad_norm": 0.49304288909033295,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5824,
+      "step": 363
+    },
+    {
+      "epoch": 1.4754430379746837,
+      "grad_norm": 0.4075531219411201,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5658,
+      "step": 364
+    },
+    {
+      "epoch": 1.479493670886076,
+      "grad_norm": 0.44950523013918753,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5743,
+      "step": 365
+    },
+    {
+      "epoch": 1.4835443037974683,
+      "grad_norm": 0.37867629804392505,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5843,
+      "step": 366
+    },
+    {
+      "epoch": 1.4875949367088608,
+      "grad_norm": 0.49227265709677204,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5806,
+      "step": 367
+    },
+    {
+      "epoch": 1.4916455696202533,
+      "grad_norm": 0.5195662266219353,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5625,
+      "step": 368
+    },
+    {
+      "epoch": 1.4956962025316456,
+      "grad_norm": 0.47773928505088464,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5994,
+      "step": 369
+    },
+    {
+      "epoch": 1.4997468354430379,
+      "grad_norm": 0.4317675790547549,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5618,
+      "step": 370
+    },
+    {
+      "epoch": 1.5037974683544304,
+      "grad_norm": 0.40857417542561797,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5801,
+      "step": 371
+    },
+    {
+      "epoch": 1.507848101265823,
+      "grad_norm": 0.42704847835020093,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5797,
+      "step": 372
+    },
+    {
+      "epoch": 1.5118987341772152,
+      "grad_norm": 0.3942534211311338,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5749,
+      "step": 373
+    },
+    {
+      "epoch": 1.5159493670886075,
+      "grad_norm": 0.47206623009153625,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5715,
+      "step": 374
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3770079874944735,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5685,
+      "step": 375
+    },
+    {
+      "epoch": 1.5240506329113925,
+      "grad_norm": 0.4665570968449033,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5616,
+      "step": 376
+    },
+    {
+      "epoch": 1.5281012658227848,
+      "grad_norm": 0.39058609635383545,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5695,
+      "step": 377
+    },
+    {
+      "epoch": 1.5321518987341771,
+      "grad_norm": 0.35747363087285233,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5708,
+      "step": 378
+    },
+    {
+      "epoch": 1.5362025316455696,
+      "grad_norm": 0.35049406460212873,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5726,
+      "step": 379
+    },
+    {
+      "epoch": 1.5402531645569622,
+      "grad_norm": 0.3777720756082823,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5726,
+      "step": 380
+    },
+    {
+      "epoch": 1.5443037974683544,
+      "grad_norm": 0.36077222637039236,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5745,
+      "step": 381
+    },
+    {
+      "epoch": 1.5483544303797467,
+      "grad_norm": 0.4860481746830461,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5934,
+      "step": 382
+    },
+    {
+      "epoch": 1.5524050632911393,
+      "grad_norm": 0.4961833237126473,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5742,
+      "step": 383
+    },
+    {
+      "epoch": 1.5564556962025318,
+      "grad_norm": 0.4320571894162534,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5619,
+      "step": 384
+    },
+    {
+      "epoch": 1.560506329113924,
+      "grad_norm": 0.3624282142614308,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5754,
+      "step": 385
+    },
+    {
+      "epoch": 1.5645569620253164,
+      "grad_norm": 0.3792892878372777,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5639,
+      "step": 386
+    },
+    {
+      "epoch": 1.5686075949367089,
+      "grad_norm": 0.351906254692954,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5462,
+      "step": 387
+    },
+    {
+      "epoch": 1.5726582278481014,
+      "grad_norm": 0.35547600407158053,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.568,
+      "step": 388
+    },
+    {
+      "epoch": 1.5767088607594937,
+      "grad_norm": 0.41300117675252423,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5708,
+      "step": 389
+    },
+    {
+      "epoch": 1.580759493670886,
+      "grad_norm": 0.31098878674793945,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5703,
+      "step": 390
+    },
+    {
+      "epoch": 1.5848101265822785,
+      "grad_norm": 0.4022886818835067,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5698,
+      "step": 391
+    },
+    {
+      "epoch": 1.588860759493671,
+      "grad_norm": 0.36074986564038686,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5828,
+      "step": 392
+    },
+    {
+      "epoch": 1.5929113924050633,
+      "grad_norm": 0.3637626692634795,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5683,
+      "step": 393
+    },
+    {
+      "epoch": 1.5969620253164556,
+      "grad_norm": 0.3797078648102642,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5699,
+      "step": 394
+    },
+    {
+      "epoch": 1.601012658227848,
+      "grad_norm": 0.3427214079175517,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5806,
+      "step": 395
+    },
+    {
+      "epoch": 1.6050632911392406,
+      "grad_norm": 0.5118782908938658,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5729,
+      "step": 396
+    },
+    {
+      "epoch": 1.609113924050633,
+      "grad_norm": 0.456414027269528,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.5763,
+      "step": 397
+    },
+    {
+      "epoch": 1.6131645569620252,
+      "grad_norm": 0.41139278799296636,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.568,
+      "step": 398
+    },
+    {
+      "epoch": 1.6172151898734177,
+      "grad_norm": 0.5029575176874077,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5744,
+      "step": 399
+    },
+    {
+      "epoch": 1.6212658227848102,
+      "grad_norm": 0.4134961342937195,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5876,
+      "step": 400
+    },
+    {
+      "epoch": 1.6253164556962025,
+      "grad_norm": 0.3805649206415535,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5703,
+      "step": 401
+    },
+    {
+      "epoch": 1.6293670886075948,
+      "grad_norm": 0.44832727737323136,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5797,
+      "step": 402
+    },
+    {
+      "epoch": 1.6334177215189873,
+      "grad_norm": 0.4908626913126581,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5792,
+      "step": 403
+    },
+    {
+      "epoch": 1.6374683544303799,
+      "grad_norm": 0.4490695781699458,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5724,
+      "step": 404
+    },
+    {
+      "epoch": 1.6415189873417722,
+      "grad_norm": 0.43657789176184786,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5722,
+      "step": 405
+    },
+    {
+      "epoch": 1.6455696202531644,
+      "grad_norm": 0.6198314733587138,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5751,
+      "step": 406
+    },
+    {
+      "epoch": 1.649620253164557,
+      "grad_norm": 0.43161346418693336,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5871,
+      "step": 407
+    },
+    {
+      "epoch": 1.6536708860759495,
+      "grad_norm": 0.39798012794448717,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5626,
+      "step": 408
+    },
+    {
+      "epoch": 1.6577215189873418,
+      "grad_norm": 0.39123559382475837,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.569,
+      "step": 409
+    },
+    {
+      "epoch": 1.661772151898734,
+      "grad_norm": 0.41276317292923453,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5609,
+      "step": 410
+    },
+    {
+      "epoch": 1.6658227848101266,
+      "grad_norm": 0.3346143871051394,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.574,
+      "step": 411
+    },
+    {
+      "epoch": 1.669873417721519,
+      "grad_norm": 0.4160857928414006,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5624,
+      "step": 412
+    },
+    {
+      "epoch": 1.6739240506329114,
+      "grad_norm": 0.31152974836808744,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.562,
+      "step": 413
+    },
+    {
+      "epoch": 1.6779746835443037,
+      "grad_norm": 0.4353926176559077,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5598,
+      "step": 414
+    },
+    {
+      "epoch": 1.6820253164556962,
+      "grad_norm": 0.36321846103373717,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5592,
+      "step": 415
+    },
+    {
+      "epoch": 1.6860759493670887,
+      "grad_norm": 0.3810605979860431,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5634,
+      "step": 416
+    },
+    {
+      "epoch": 1.690126582278481,
+      "grad_norm": 0.4146387232722059,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5697,
+      "step": 417
+    },
+    {
+      "epoch": 1.6941772151898733,
+      "grad_norm": 0.3676561380639018,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5652,
+      "step": 418
+    },
+    {
+      "epoch": 1.6982278481012658,
+      "grad_norm": 0.3259274662294342,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5863,
+      "step": 419
+    },
+    {
+      "epoch": 1.7022784810126583,
+      "grad_norm": 0.4085543042755284,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5619,
+      "step": 420
+    },
+    {
+      "epoch": 1.7063291139240506,
+      "grad_norm": 0.34158724652088956,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5683,
+      "step": 421
+    },
+    {
+      "epoch": 1.710379746835443,
+      "grad_norm": 0.38544286605443906,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5758,
+      "step": 422
+    },
+    {
+      "epoch": 1.7144303797468354,
+      "grad_norm": 0.3557036460172178,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5788,
+      "step": 423
+    },
+    {
+      "epoch": 1.718481012658228,
+      "grad_norm": 0.4851167897158188,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5618,
+      "step": 424
+    },
+    {
+      "epoch": 1.7225316455696202,
+      "grad_norm": 0.43205868808580067,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.578,
+      "step": 425
+    },
+    {
+      "epoch": 1.7265822784810125,
+      "grad_norm": 0.3778094456896258,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5721,
+      "step": 426
+    },
+    {
+      "epoch": 1.730632911392405,
+      "grad_norm": 0.4319649342234142,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5663,
+      "step": 427
+    },
+    {
+      "epoch": 1.7346835443037976,
+      "grad_norm": 0.4046007707000431,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5786,
+      "step": 428
+    },
+    {
+      "epoch": 1.7387341772151899,
+      "grad_norm": 0.3793042988828969,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.5715,
+      "step": 429
+    },
+    {
+      "epoch": 1.7427848101265822,
+      "grad_norm": 0.44082623138629345,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5722,
+      "step": 430
+    },
+    {
+      "epoch": 1.7468354430379747,
+      "grad_norm": 0.5250454659872341,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5673,
+      "step": 431
+    },
+    {
+      "epoch": 1.7508860759493672,
+      "grad_norm": 0.3752144502732557,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.571,
+      "step": 432
+    },
+    {
+      "epoch": 1.7549367088607595,
+      "grad_norm": 0.3863601291822436,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5757,
+      "step": 433
+    },
+    {
+      "epoch": 1.7589873417721518,
+      "grad_norm": 0.4994128077937738,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5646,
+      "step": 434
+    },
+    {
+      "epoch": 1.7630379746835443,
+      "grad_norm": 0.5414244421550419,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5688,
+      "step": 435
+    },
+    {
+      "epoch": 1.7670886075949368,
+      "grad_norm": 0.3336201510392659,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5571,
+      "step": 436
+    },
+    {
+      "epoch": 1.771139240506329,
+      "grad_norm": 0.5457006648416834,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5643,
+      "step": 437
+    },
+    {
+      "epoch": 1.7751898734177214,
+      "grad_norm": 0.5073752741090407,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5486,
+      "step": 438
+    },
+    {
+      "epoch": 1.779240506329114,
+      "grad_norm": 0.3698975676982019,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5679,
+      "step": 439
+    },
+    {
+      "epoch": 1.7832911392405064,
+      "grad_norm": 1.842136528384267,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.6044,
+      "step": 440
+    },
+    {
+      "epoch": 1.7873417721518987,
+      "grad_norm": 0.4471643135967781,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5833,
+      "step": 441
+    },
+    {
+      "epoch": 1.791392405063291,
+      "grad_norm": 0.5957429902192082,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5628,
+      "step": 442
+    },
+    {
+      "epoch": 1.7954430379746835,
+      "grad_norm": 0.3950060285975274,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5591,
+      "step": 443
+    },
+    {
+      "epoch": 1.799493670886076,
+      "grad_norm": 0.40596889504337186,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5668,
+      "step": 444
+    },
+    {
+      "epoch": 1.8035443037974683,
+      "grad_norm": 0.47968250839779675,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.5599,
+      "step": 445
+    },
+    {
+      "epoch": 1.8075949367088606,
+      "grad_norm": 0.3402847291309061,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.581,
+      "step": 446
+    },
+    {
+      "epoch": 1.8116455696202531,
+      "grad_norm": 0.4220764703829784,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5609,
+      "step": 447
+    },
+    {
+      "epoch": 1.8156962025316457,
+      "grad_norm": 0.44540374062912996,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5753,
+      "step": 448
+    },
+    {
+      "epoch": 1.819746835443038,
+      "grad_norm": 0.3719335051159528,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5748,
+      "step": 449
+    },
+    {
+      "epoch": 1.8237974683544302,
+      "grad_norm": 0.35686796975091434,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5725,
+      "step": 450
+    },
+    {
+      "epoch": 1.8278481012658228,
+      "grad_norm": 0.330773493124629,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5825,
+      "step": 451
+    },
+    {
+      "epoch": 1.8318987341772153,
+      "grad_norm": 0.3621578817835101,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5685,
+      "step": 452
+    },
+    {
+      "epoch": 1.8359493670886076,
+      "grad_norm": 0.3115280721229672,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5729,
+      "step": 453
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.3668137335328413,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5664,
+      "step": 454
+    },
+    {
+      "epoch": 1.8440506329113924,
+      "grad_norm": 0.32867430699224925,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5592,
+      "step": 455
+    },
+    {
+      "epoch": 1.8481012658227849,
+      "grad_norm": 0.42040112527832035,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.565,
+      "step": 456
+    },
+    {
+      "epoch": 1.8521518987341772,
+      "grad_norm": 0.35926991262489943,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5619,
+      "step": 457
+    },
+    {
+      "epoch": 1.8562025316455695,
+      "grad_norm": 0.3783787101058612,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5681,
+      "step": 458
+    },
+    {
+      "epoch": 1.860253164556962,
+      "grad_norm": 0.4555582302366678,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5464,
+      "step": 459
+    },
+    {
+      "epoch": 1.8643037974683545,
+      "grad_norm": 0.37017848126456443,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5672,
+      "step": 460
+    },
+    {
+      "epoch": 1.8683544303797468,
+      "grad_norm": 0.4535867353677617,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5599,
+      "step": 461
+    },
+    {
+      "epoch": 1.872405063291139,
+      "grad_norm": 0.3595247742589157,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.5608,
+      "step": 462
+    },
+    {
+      "epoch": 1.8764556962025316,
+      "grad_norm": 0.3909296354072627,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5668,
+      "step": 463
+    },
+    {
+      "epoch": 1.8805063291139241,
+      "grad_norm": 0.3850122150894832,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5626,
+      "step": 464
+    },
+    {
+      "epoch": 1.8845569620253164,
+      "grad_norm": 0.35847880610050675,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.553,
+      "step": 465
+    },
+    {
+      "epoch": 1.8886075949367087,
+      "grad_norm": 0.3890829355667952,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.574,
+      "step": 466
+    },
+    {
+      "epoch": 1.8926582278481012,
+      "grad_norm": 0.33215517124161265,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5612,
+      "step": 467
+    },
+    {
+      "epoch": 1.8967088607594937,
+      "grad_norm": 0.3833081590364025,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5708,
+      "step": 468
+    },
+    {
+      "epoch": 1.900759493670886,
+      "grad_norm": 0.3264950642708518,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.56,
+      "step": 469
+    },
+    {
+      "epoch": 1.9048101265822783,
+      "grad_norm": 0.36742296243500683,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.5678,
+      "step": 470
+    },
+    {
+      "epoch": 1.9088607594936708,
+      "grad_norm": 0.3511884013179289,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5674,
+      "step": 471
+    },
+    {
+      "epoch": 1.9129113924050634,
+      "grad_norm": 0.3391842662460048,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.5681,
+      "step": 472
+    },
+    {
+      "epoch": 1.9169620253164557,
+      "grad_norm": 0.35842435864496713,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5624,
+      "step": 473
+    },
+    {
+      "epoch": 1.921012658227848,
+      "grad_norm": 0.3146241383321096,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.577,
+      "step": 474
+    },
+    {
+      "epoch": 1.9250632911392405,
+      "grad_norm": 0.35950747694425195,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5713,
+      "step": 475
+    },
+    {
+      "epoch": 1.929113924050633,
+      "grad_norm": 0.36312961020617696,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5731,
+      "step": 476
+    },
+    {
+      "epoch": 1.9331645569620253,
+      "grad_norm": 0.35514776529298997,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5607,
+      "step": 477
+    },
+    {
+      "epoch": 1.9372151898734176,
+      "grad_norm": 0.37275444549435743,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5759,
+      "step": 478
+    },
+    {
+      "epoch": 1.94126582278481,
+      "grad_norm": 0.34464416643632506,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5707,
+      "step": 479
+    },
+    {
+      "epoch": 1.9453164556962026,
+      "grad_norm": 0.4189715418466446,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.563,
+      "step": 480
+    },
+    {
+      "epoch": 1.9493670886075949,
+      "grad_norm": 0.30929382102660324,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.569,
+      "step": 481
+    },
+    {
+      "epoch": 1.9534177215189872,
+      "grad_norm": 0.43575707681764747,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.566,
+      "step": 482
+    },
+    {
+      "epoch": 1.9574683544303797,
+      "grad_norm": 0.3170323202309556,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.584,
+      "step": 483
+    },
+    {
+      "epoch": 1.9615189873417722,
+      "grad_norm": 0.38924093100017326,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5708,
+      "step": 484
+    },
+    {
+      "epoch": 1.9655696202531645,
+      "grad_norm": 0.4397525102054439,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5714,
+      "step": 485
+    },
+    {
+      "epoch": 1.9696202531645568,
+      "grad_norm": 0.3152575842007319,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5589,
+      "step": 486
+    },
+    {
+      "epoch": 1.9736708860759493,
+      "grad_norm": 0.4142831268171398,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5762,
+      "step": 487
+    },
+    {
+      "epoch": 1.9777215189873418,
+      "grad_norm": 0.40787495979316857,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5728,
+      "step": 488
+    },
+    {
+      "epoch": 1.9817721518987341,
+      "grad_norm": 0.3976360352039801,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.5569,
+      "step": 489
+    },
+    {
+      "epoch": 1.9858227848101264,
+      "grad_norm": 0.36648115705359713,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5581,
+      "step": 490
+    },
+    {
+      "epoch": 1.989873417721519,
+      "grad_norm": 0.34707105229441365,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5571,
+      "step": 491
+    },
+    {
+      "epoch": 1.9939240506329114,
+      "grad_norm": 0.3742955549261726,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5678,
+      "step": 492
+    },
+    {
+      "epoch": 1.9979746835443037,
+      "grad_norm": 0.34251800878630034,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5635,
+      "step": 493
+    },
+    {
+      "epoch": 2.0030379746835445,
+      "grad_norm": 0.4362543607761948,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.5149,
+      "step": 494
+    },
+    {
+      "epoch": 2.0070886075949366,
+      "grad_norm": 0.6044338075146388,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.5255,
+      "step": 495
+    },
+    {
+      "epoch": 2.011139240506329,
+      "grad_norm": 0.6094678643782165,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.5172,
+      "step": 496
+    },
+    {
+      "epoch": 2.0151898734177216,
+      "grad_norm": 0.6310916311739054,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.5017,
+      "step": 497
+    },
+    {
+      "epoch": 2.019240506329114,
+      "grad_norm": 0.42733899096450423,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.5065,
+      "step": 498
+    },
+    {
+      "epoch": 2.023291139240506,
+      "grad_norm": 0.5091815911683367,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.5079,
+      "step": 499
+    },
+    {
+      "epoch": 2.0273417721518987,
+      "grad_norm": 0.4754496627102285,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.5154,
+      "step": 500
+    },
+    {
+      "epoch": 2.031392405063291,
+      "grad_norm": 0.502499817060006,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.5055,
+      "step": 501
+    },
+    {
+      "epoch": 2.0354430379746837,
+      "grad_norm": 0.4659285885424091,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.5084,
+      "step": 502
+    },
+    {
+      "epoch": 2.039493670886076,
+      "grad_norm": 0.4328871325999897,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.5187,
+      "step": 503
+    },
+    {
+      "epoch": 2.0435443037974683,
+      "grad_norm": 0.5159447474542375,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.5095,
+      "step": 504
+    },
+    {
+      "epoch": 2.047594936708861,
+      "grad_norm": 0.4276938469253526,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.5017,
+      "step": 505
+    },
+    {
+      "epoch": 2.0516455696202534,
+      "grad_norm": 0.4502542193506515,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.5041,
+      "step": 506
+    },
+    {
+      "epoch": 2.0556962025316454,
+      "grad_norm": 0.5427226532606654,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.5166,
+      "step": 507
+    },
+    {
+      "epoch": 2.059746835443038,
+      "grad_norm": 0.4180530805087877,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.5095,
+      "step": 508
+    },
+    {
+      "epoch": 2.0637974683544305,
+      "grad_norm": 0.4425293854834803,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.5105,
+      "step": 509
+    },
+    {
+      "epoch": 2.067848101265823,
+      "grad_norm": 0.38739108030297525,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.4914,
+      "step": 510
+    },
+    {
+      "epoch": 2.071898734177215,
+      "grad_norm": 0.4068360744955018,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.5003,
+      "step": 511
+    },
+    {
+      "epoch": 2.0759493670886076,
+      "grad_norm": 0.4405389725400748,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.5046,
+      "step": 512
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.44016508161766765,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.4963,
+      "step": 513
+    },
+    {
+      "epoch": 2.0840506329113926,
+      "grad_norm": 0.4586963782748912,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.5093,
+      "step": 514
+    },
+    {
+      "epoch": 2.0881012658227847,
+      "grad_norm": 0.3945673684778376,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4999,
+      "step": 515
+    },
+    {
+      "epoch": 2.092151898734177,
+      "grad_norm": 0.5590971818747368,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.5088,
+      "step": 516
+    },
+    {
+      "epoch": 2.0962025316455697,
+      "grad_norm": 0.37677115582428233,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.4979,
+      "step": 517
+    },
+    {
+      "epoch": 2.100253164556962,
+      "grad_norm": 0.3661284281355434,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.5064,
+      "step": 518
+    },
+    {
+      "epoch": 2.1043037974683543,
+      "grad_norm": 0.35469905076517183,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.5023,
+      "step": 519
+    },
+    {
+      "epoch": 2.108354430379747,
+      "grad_norm": 0.3455233071744697,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4994,
+      "step": 520
+    },
+    {
+      "epoch": 2.1124050632911393,
+      "grad_norm": 0.32008688016427805,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.5119,
+      "step": 521
+    },
+    {
+      "epoch": 2.116455696202532,
+      "grad_norm": 0.3472003510221478,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.4959,
+      "step": 522
+    },
+    {
+      "epoch": 2.120506329113924,
+      "grad_norm": 0.3335437194000592,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.5042,
+      "step": 523
+    },
+    {
+      "epoch": 2.1245569620253164,
+      "grad_norm": 0.3576435136514813,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.487,
+      "step": 524
+    },
+    {
+      "epoch": 2.128607594936709,
+      "grad_norm": 0.42173238761139303,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.5191,
+      "step": 525
+    },
+    {
+      "epoch": 2.1326582278481014,
+      "grad_norm": 0.331414119593145,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.4905,
+      "step": 526
+    },
+    {
+      "epoch": 2.1367088607594935,
+      "grad_norm": 0.4169111649501582,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.5093,
+      "step": 527
+    },
+    {
+      "epoch": 2.140759493670886,
+      "grad_norm": 0.4186849787866555,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.5114,
+      "step": 528
+    },
+    {
+      "epoch": 2.1448101265822785,
+      "grad_norm": 0.35590311834392824,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.53,
+      "step": 529
+    },
+    {
+      "epoch": 2.148860759493671,
+      "grad_norm": 0.31836013932504165,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.5142,
+      "step": 530
+    },
+    {
+      "epoch": 2.152911392405063,
+      "grad_norm": 0.3203491301601664,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.5161,
+      "step": 531
+    },
+    {
+      "epoch": 2.1569620253164556,
+      "grad_norm": 0.35194036117672606,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4979,
+      "step": 532
+    },
+    {
+      "epoch": 2.161012658227848,
+      "grad_norm": 0.316041489255002,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4999,
+      "step": 533
+    },
+    {
+      "epoch": 2.1650632911392407,
+      "grad_norm": 0.31946017561170714,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.5101,
+      "step": 534
+    },
+    {
+      "epoch": 2.1691139240506327,
+      "grad_norm": 0.32266144042244704,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.5132,
+      "step": 535
+    },
+    {
+      "epoch": 2.1731645569620253,
+      "grad_norm": 0.2755813254617997,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.5032,
+      "step": 536
+    },
+    {
+      "epoch": 2.1772151898734178,
+      "grad_norm": 0.3396157840034228,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.5092,
+      "step": 537
+    },
+    {
+      "epoch": 2.1812658227848103,
+      "grad_norm": 0.2978114487640706,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.4958,
+      "step": 538
+    },
+    {
+      "epoch": 2.1853164556962024,
+      "grad_norm": 0.32513598433424273,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4881,
+      "step": 539
+    },
+    {
+      "epoch": 2.189367088607595,
+      "grad_norm": 0.31526881797262024,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.5073,
+      "step": 540
+    },
+    {
+      "epoch": 2.1934177215189874,
+      "grad_norm": 0.37850167385502187,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.5023,
+      "step": 541
+    },
+    {
+      "epoch": 2.19746835443038,
+      "grad_norm": 0.42637945622650253,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.5183,
+      "step": 542
+    },
+    {
+      "epoch": 2.201518987341772,
+      "grad_norm": 0.3457837449268246,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.5061,
+      "step": 543
+    },
+    {
+      "epoch": 2.2055696202531645,
+      "grad_norm": 0.3324771563994227,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.5014,
+      "step": 544
+    },
+    {
+      "epoch": 2.209620253164557,
+      "grad_norm": 0.30678553433744526,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.5099,
+      "step": 545
+    },
+    {
+      "epoch": 2.2136708860759495,
+      "grad_norm": 0.3250000712940461,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.5063,
+      "step": 546
+    },
+    {
+      "epoch": 2.2177215189873416,
+      "grad_norm": 0.29525961487058333,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4979,
+      "step": 547
+    },
+    {
+      "epoch": 2.221772151898734,
+      "grad_norm": 0.3369261875433186,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.523,
+      "step": 548
+    },
+    {
+      "epoch": 2.2258227848101266,
+      "grad_norm": 0.3545224203764559,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.5104,
+      "step": 549
+    },
+    {
+      "epoch": 2.229873417721519,
+      "grad_norm": 0.3064960961789428,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.5046,
+      "step": 550
+    },
+    {
+      "epoch": 2.233924050632911,
+      "grad_norm": 0.34093939837164977,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.5161,
+      "step": 551
+    },
+    {
+      "epoch": 2.2379746835443037,
+      "grad_norm": 0.999526476089123,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.5112,
+      "step": 552
+    },
+    {
+      "epoch": 2.2420253164556962,
+      "grad_norm": 0.32974467219978054,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.5061,
+      "step": 553
+    },
+    {
+      "epoch": 2.2460759493670888,
+      "grad_norm": 0.3668415458336695,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.5161,
+      "step": 554
+    },
+    {
+      "epoch": 2.250126582278481,
+      "grad_norm": 0.3011506775998163,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.5038,
+      "step": 555
+    },
+    {
+      "epoch": 2.2541772151898734,
+      "grad_norm": 0.29239335199522387,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.5045,
+      "step": 556
+    },
+    {
+      "epoch": 2.258227848101266,
+      "grad_norm": 0.29023297758714983,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.5118,
+      "step": 557
+    },
+    {
+      "epoch": 2.2622784810126584,
+      "grad_norm": 0.3603401793168279,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.492,
+      "step": 558
+    },
+    {
+      "epoch": 2.2663291139240505,
+      "grad_norm": 0.29910310278560637,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.5056,
+      "step": 559
+    },
+    {
+      "epoch": 2.270379746835443,
+      "grad_norm": 0.36836357651622786,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.5022,
+      "step": 560
+    },
+    {
+      "epoch": 2.2744303797468355,
+      "grad_norm": 0.4156879554904488,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.5278,
+      "step": 561
+    },
+    {
+      "epoch": 2.278481012658228,
+      "grad_norm": 0.30541515154626997,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.5056,
+      "step": 562
+    },
+    {
+      "epoch": 2.28253164556962,
+      "grad_norm": 0.2898988116196288,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.5181,
+      "step": 563
+    },
+    {
+      "epoch": 2.2865822784810126,
+      "grad_norm": 0.3520995200184041,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.5012,
+      "step": 564
+    },
+    {
+      "epoch": 2.290632911392405,
+      "grad_norm": 0.27786924085437925,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.5179,
+      "step": 565
+    },
+    {
+      "epoch": 2.2946835443037976,
+      "grad_norm": 0.37059046267903256,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.4897,
+      "step": 566
+    },
+    {
+      "epoch": 2.2987341772151897,
+      "grad_norm": 0.26429510207637785,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.5162,
+      "step": 567
+    },
+    {
+      "epoch": 2.302784810126582,
+      "grad_norm": 0.37209784526692335,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.5065,
+      "step": 568
+    },
+    {
+      "epoch": 2.3068354430379747,
+      "grad_norm": 0.2888162601608562,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.5251,
+      "step": 569
+    },
+    {
+      "epoch": 2.3108860759493672,
+      "grad_norm": 0.2839004775207586,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.4858,
+      "step": 570
+    },
+    {
+      "epoch": 2.3149367088607593,
+      "grad_norm": 0.309464143222048,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.507,
+      "step": 571
+    },
+    {
+      "epoch": 2.318987341772152,
+      "grad_norm": 0.28009784174259006,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.5084,
+      "step": 572
+    },
+    {
+      "epoch": 2.3230379746835443,
+      "grad_norm": 0.32281507136606297,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.5191,
+      "step": 573
+    },
+    {
+      "epoch": 2.327088607594937,
+      "grad_norm": 0.2889899083597477,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.5092,
+      "step": 574
+    },
+    {
+      "epoch": 2.331139240506329,
+      "grad_norm": 0.30732774652812334,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.5004,
+      "step": 575
+    },
+    {
+      "epoch": 2.3351898734177214,
+      "grad_norm": 0.34099884408405,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.524,
+      "step": 576
+    },
+    {
+      "epoch": 2.339240506329114,
+      "grad_norm": 0.2938431487784586,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.5078,
+      "step": 577
+    },
+    {
+      "epoch": 2.3432911392405065,
+      "grad_norm": 0.33801125684080935,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.5028,
+      "step": 578
+    },
+    {
+      "epoch": 2.3473417721518985,
+      "grad_norm": 0.27898819541957964,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.5106,
+      "step": 579
+    },
+    {
+      "epoch": 2.351392405063291,
+      "grad_norm": 0.3062629867609308,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.5061,
+      "step": 580
+    },
+    {
+      "epoch": 2.3554430379746836,
+      "grad_norm": 0.3082425801295347,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.5134,
+      "step": 581
+    },
+    {
+      "epoch": 2.359493670886076,
+      "grad_norm": 0.31840731765876906,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.5076,
+      "step": 582
+    },
+    {
+      "epoch": 2.363544303797468,
+      "grad_norm": 0.433413219210841,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.5101,
+      "step": 583
+    },
+    {
+      "epoch": 2.3675949367088607,
+      "grad_norm": 0.309884010672882,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.4893,
+      "step": 584
+    },
+    {
+      "epoch": 2.371645569620253,
+      "grad_norm": 0.3374344624519763,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.5122,
+      "step": 585
+    },
+    {
+      "epoch": 2.3756962025316457,
+      "grad_norm": 0.331267612176851,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.518,
+      "step": 586
+    },
+    {
+      "epoch": 2.379746835443038,
+      "grad_norm": 0.9721809024937123,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.5195,
+      "step": 587
+    },
+    {
+      "epoch": 2.3837974683544303,
+      "grad_norm": 0.3409272216112119,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.5049,
+      "step": 588
+    },
+    {
+      "epoch": 2.387848101265823,
+      "grad_norm": 0.37332594679406456,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.5073,
+      "step": 589
+    },
+    {
+      "epoch": 2.3918987341772153,
+      "grad_norm": 0.33929043477258025,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.5046,
+      "step": 590
+    },
+    {
+      "epoch": 2.3959493670886074,
+      "grad_norm": 0.4650410901820957,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.5058,
+      "step": 591
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2986020442691154,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.5126,
+      "step": 592
+    },
+    {
+      "epoch": 2.4040506329113924,
+      "grad_norm": 0.45386360146811444,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.5118,
+      "step": 593
+    },
+    {
+      "epoch": 2.408101265822785,
+      "grad_norm": 0.8489673707365041,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.5047,
+      "step": 594
+    },
+    {
+      "epoch": 2.4121518987341775,
+      "grad_norm": 0.3567944348269758,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.5138,
+      "step": 595
+    },
+    {
+      "epoch": 2.4162025316455695,
+      "grad_norm": 0.3918411956084219,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.5022,
+      "step": 596
+    },
+    {
+      "epoch": 2.420253164556962,
+      "grad_norm": 0.3527432930207906,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.5029,
+      "step": 597
+    },
+    {
+      "epoch": 2.4243037974683546,
+      "grad_norm": 0.38997798913102655,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4911,
+      "step": 598
+    },
+    {
+      "epoch": 2.4283544303797466,
+      "grad_norm": 0.33674283825092577,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.5091,
+      "step": 599
+    },
+    {
+      "epoch": 2.432405063291139,
+      "grad_norm": 0.3948267859575767,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.4937,
+      "step": 600
+    },
+    {
+      "epoch": 2.4364556962025317,
+      "grad_norm": 0.2993867271495416,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.5046,
+      "step": 601
+    },
+    {
+      "epoch": 2.440506329113924,
+      "grad_norm": 0.2806429003936777,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.5006,
+      "step": 602
+    },
+    {
+      "epoch": 2.4445569620253167,
+      "grad_norm": 0.3432823367656082,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.5076,
+      "step": 603
+    },
+    {
+      "epoch": 2.4486075949367088,
+      "grad_norm": 0.27773351259389273,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.4963,
+      "step": 604
+    },
+    {
+      "epoch": 2.4526582278481013,
+      "grad_norm": 0.28400004028989323,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.5094,
+      "step": 605
+    },
+    {
+      "epoch": 2.456708860759494,
+      "grad_norm": 0.338456835148589,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.513,
+      "step": 606
+    },
+    {
+      "epoch": 2.460759493670886,
+      "grad_norm": 0.2743884488857805,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.4986,
+      "step": 607
+    },
+    {
+      "epoch": 2.4648101265822784,
+      "grad_norm": 0.28488857787414157,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.4915,
+      "step": 608
+    },
+    {
+      "epoch": 2.468860759493671,
+      "grad_norm": 0.29045367358523294,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.5122,
+      "step": 609
+    },
+    {
+      "epoch": 2.4729113924050634,
+      "grad_norm": 0.29380217034632755,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4944,
+      "step": 610
+    },
+    {
+      "epoch": 2.476962025316456,
+      "grad_norm": 0.3326448986585829,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.517,
+      "step": 611
+    },
+    {
+      "epoch": 2.481012658227848,
+      "grad_norm": 0.28414991036376325,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.4979,
+      "step": 612
+    },
+    {
+      "epoch": 2.4850632911392405,
+      "grad_norm": 0.37423908224394287,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.4939,
+      "step": 613
+    },
+    {
+      "epoch": 2.489113924050633,
+      "grad_norm": 0.28982086767179355,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.4982,
+      "step": 614
+    },
+    {
+      "epoch": 2.493164556962025,
+      "grad_norm": 0.2736574719663669,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4976,
+      "step": 615
+    },
+    {
+      "epoch": 2.4972151898734176,
+      "grad_norm": 0.32835595095044723,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.5076,
+      "step": 616
+    },
+    {
+      "epoch": 2.50126582278481,
+      "grad_norm": 0.2668794310915097,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.5026,
+      "step": 617
+    },
+    {
+      "epoch": 2.5053164556962026,
+      "grad_norm": 0.3203920518526405,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.5093,
+      "step": 618
+    },
+    {
+      "epoch": 2.509367088607595,
+      "grad_norm": 0.3180391411280136,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.5377,
+      "step": 619
+    },
+    {
+      "epoch": 2.5134177215189872,
+      "grad_norm": 0.33111819127611186,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.5024,
+      "step": 620
+    },
+    {
+      "epoch": 2.5174683544303798,
+      "grad_norm": 0.3492570846425375,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.5089,
+      "step": 621
+    },
+    {
+      "epoch": 2.5215189873417723,
+      "grad_norm": 0.2681811700428865,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.5119,
+      "step": 622
+    },
+    {
+      "epoch": 2.5255696202531643,
+      "grad_norm": 0.33221223068069633,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.5171,
+      "step": 623
+    },
+    {
+      "epoch": 2.529620253164557,
+      "grad_norm": 0.2420058706882021,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.5118,
+      "step": 624
+    },
+    {
+      "epoch": 2.5336708860759494,
+      "grad_norm": 0.31860573257044494,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.5087,
+      "step": 625
+    },
+    {
+      "epoch": 2.537721518987342,
+      "grad_norm": 0.2843774291495513,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.5027,
+      "step": 626
+    },
+    {
+      "epoch": 2.5417721518987344,
+      "grad_norm": 0.2885851507590268,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.5113,
+      "step": 627
+    },
+    {
+      "epoch": 2.5458227848101265,
+      "grad_norm": 0.26673090244533215,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.5042,
+      "step": 628
+    },
+    {
+      "epoch": 2.549873417721519,
+      "grad_norm": 0.2789874651777166,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.5089,
+      "step": 629
+    },
+    {
+      "epoch": 2.5539240506329115,
+      "grad_norm": 0.2895606835065868,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.5045,
+      "step": 630
+    },
+    {
+      "epoch": 2.5579746835443036,
+      "grad_norm": 0.30624937524683893,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.5109,
+      "step": 631
+    },
+    {
+      "epoch": 2.562025316455696,
+      "grad_norm": 0.2872117568262824,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.5116,
+      "step": 632
+    },
+    {
+      "epoch": 2.5660759493670886,
+      "grad_norm": 0.3142493509740597,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.4983,
+      "step": 633
+    },
+    {
+      "epoch": 2.570126582278481,
+      "grad_norm": 0.3076885612445364,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4918,
+      "step": 634
+    },
+    {
+      "epoch": 2.5741772151898736,
+      "grad_norm": 0.29182501704852243,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.4967,
+      "step": 635
+    },
+    {
+      "epoch": 2.5782278481012657,
+      "grad_norm": 0.2713468499783712,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4856,
+      "step": 636
+    },
+    {
+      "epoch": 2.5822784810126582,
+      "grad_norm": 0.2933560437629066,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.5078,
+      "step": 637
+    },
+    {
+      "epoch": 2.5863291139240507,
+      "grad_norm": 0.27312234099216515,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.5023,
+      "step": 638
+    },
+    {
+      "epoch": 2.590379746835443,
+      "grad_norm": 0.3239137266633085,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.5037,
+      "step": 639
+    },
+    {
+      "epoch": 2.5944303797468353,
+      "grad_norm": 0.29625127350734937,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.5056,
+      "step": 640
+    },
+    {
+      "epoch": 2.598481012658228,
+      "grad_norm": 0.2778476385678652,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.5149,
+      "step": 641
+    },
+    {
+      "epoch": 2.6025316455696204,
+      "grad_norm": 1.0382939353768676,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.5154,
+      "step": 642
+    },
+    {
+      "epoch": 2.606582278481013,
+      "grad_norm": 0.29732384955295177,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.5071,
+      "step": 643
+    },
+    {
+      "epoch": 2.610632911392405,
+      "grad_norm": 0.3642196612636565,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.5128,
+      "step": 644
+    },
+    {
+      "epoch": 2.6146835443037975,
+      "grad_norm": 0.38472714971044586,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.5038,
+      "step": 645
+    },
+    {
+      "epoch": 2.61873417721519,
+      "grad_norm": 0.3352925783664387,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4994,
+      "step": 646
+    },
+    {
+      "epoch": 2.622784810126582,
+      "grad_norm": 1.6804934027620009,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.559,
+      "step": 647
+    },
+    {
+      "epoch": 2.6268354430379746,
+      "grad_norm": 0.39093278045389956,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.5211,
+      "step": 648
+    },
+    {
+      "epoch": 2.630886075949367,
+      "grad_norm": 0.29063479876017695,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.507,
+      "step": 649
+    },
+    {
+      "epoch": 2.6349367088607596,
+      "grad_norm": 0.3606291514430408,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.5037,
+      "step": 650
+    },
+    {
+      "epoch": 2.638987341772152,
+      "grad_norm": 0.3382753238108862,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.5026,
+      "step": 651
+    },
+    {
+      "epoch": 2.643037974683544,
+      "grad_norm": 0.2940172288697098,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.5136,
+      "step": 652
+    },
+    {
+      "epoch": 2.6470886075949367,
+      "grad_norm": 0.34534821062628707,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.5085,
+      "step": 653
+    },
+    {
+      "epoch": 2.651139240506329,
+      "grad_norm": 0.28100703698032165,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.5133,
+      "step": 654
+    },
+    {
+      "epoch": 2.6551898734177213,
+      "grad_norm": 0.4486610323362489,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.5055,
+      "step": 655
+    },
+    {
+      "epoch": 2.659240506329114,
+      "grad_norm": 0.3257240650636601,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4849,
+      "step": 656
+    },
+    {
+      "epoch": 2.6632911392405063,
+      "grad_norm": 0.31412279331525406,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.5,
+      "step": 657
+    },
+    {
+      "epoch": 2.667341772151899,
+      "grad_norm": 0.31994061272647706,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.4975,
+      "step": 658
+    },
+    {
+      "epoch": 2.6713924050632913,
+      "grad_norm": 0.2839931566489129,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.5179,
+      "step": 659
+    },
+    {
+      "epoch": 2.6754430379746834,
+      "grad_norm": 0.3016023179954782,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4949,
+      "step": 660
+    },
+    {
+      "epoch": 2.679493670886076,
+      "grad_norm": 0.29850180257790276,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.5198,
+      "step": 661
+    },
+    {
+      "epoch": 2.6835443037974684,
+      "grad_norm": 0.3355756562453831,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.51,
+      "step": 662
+    },
+    {
+      "epoch": 2.6875949367088605,
+      "grad_norm": 0.3010081223893235,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.5031,
+      "step": 663
+    },
+    {
+      "epoch": 2.691645569620253,
+      "grad_norm": 0.29701508451506253,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.5085,
+      "step": 664
+    },
+    {
+      "epoch": 2.6956962025316455,
+      "grad_norm": 0.2949604570768768,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4975,
+      "step": 665
+    },
+    {
+      "epoch": 2.699746835443038,
+      "grad_norm": 0.2741732621236274,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.5146,
+      "step": 666
+    },
+    {
+      "epoch": 2.7037974683544306,
+      "grad_norm": 0.2795984973352661,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.5015,
+      "step": 667
+    },
+    {
+      "epoch": 2.7078481012658226,
+      "grad_norm": 0.2932523919974111,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.489,
+      "step": 668
+    },
+    {
+      "epoch": 2.711898734177215,
+      "grad_norm": 0.30701762189518805,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.5027,
+      "step": 669
+    },
+    {
+      "epoch": 2.7159493670886077,
+      "grad_norm": 0.3664436381983604,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.5173,
+      "step": 670
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.37218100023787803,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.514,
+      "step": 671
+    },
+    {
+      "epoch": 2.7240506329113923,
+      "grad_norm": 0.3005183642126687,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.507,
+      "step": 672
+    },
+    {
+      "epoch": 2.728101265822785,
+      "grad_norm": 0.33948353510315105,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.505,
+      "step": 673
+    },
+    {
+      "epoch": 2.7321518987341773,
+      "grad_norm": 0.3037107569259369,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.4903,
+      "step": 674
+    },
+    {
+      "epoch": 2.73620253164557,
+      "grad_norm": 0.3062561019648404,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.5129,
+      "step": 675
+    },
+    {
+      "epoch": 2.740253164556962,
+      "grad_norm": 0.38242934472108614,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.5051,
+      "step": 676
+    },
+    {
+      "epoch": 2.7443037974683544,
+      "grad_norm": 0.30682596520650823,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.5047,
+      "step": 677
+    },
+    {
+      "epoch": 2.748354430379747,
+      "grad_norm": 0.2832856546198741,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.5128,
+      "step": 678
+    },
+    {
+      "epoch": 2.752405063291139,
+      "grad_norm": 0.27914005249526913,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.4957,
+      "step": 679
+    },
+    {
+      "epoch": 2.7564556962025315,
+      "grad_norm": 0.2786977797008464,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.5086,
+      "step": 680
+    },
+    {
+      "epoch": 2.760506329113924,
+      "grad_norm": 0.2812342786299232,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.4963,
+      "step": 681
+    },
+    {
+      "epoch": 2.7645569620253165,
+      "grad_norm": 0.3023182602774889,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.4913,
+      "step": 682
+    },
+    {
+      "epoch": 2.768607594936709,
+      "grad_norm": 0.3024093320076416,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.5159,
+      "step": 683
+    },
+    {
+      "epoch": 2.772658227848101,
+      "grad_norm": 0.3061180834281883,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.51,
+      "step": 684
+    },
+    {
+      "epoch": 2.7767088607594936,
+      "grad_norm": 0.2967213374015336,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.5165,
+      "step": 685
+    },
+    {
+      "epoch": 2.780759493670886,
+      "grad_norm": 0.2797594541616186,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.5132,
+      "step": 686
+    },
+    {
+      "epoch": 2.7848101265822782,
+      "grad_norm": 0.29622626681564856,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.5135,
+      "step": 687
+    },
+    {
+      "epoch": 2.7888607594936707,
+      "grad_norm": 0.29217232072580024,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.5065,
+      "step": 688
+    },
+    {
+      "epoch": 2.7929113924050633,
+      "grad_norm": 0.2733708606960058,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.5117,
+      "step": 689
+    },
+    {
+      "epoch": 2.7969620253164558,
+      "grad_norm": 0.3147460990130967,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.5149,
+      "step": 690
+    },
+    {
+      "epoch": 2.8010126582278483,
+      "grad_norm": 0.2510551159832791,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.5116,
+      "step": 691
+    },
+    {
+      "epoch": 2.8050632911392404,
+      "grad_norm": 0.26420729485531735,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.5045,
+      "step": 692
+    },
+    {
+      "epoch": 2.809113924050633,
+      "grad_norm": 0.33359499470871423,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.5034,
+      "step": 693
+    },
+    {
+      "epoch": 2.8131645569620254,
+      "grad_norm": 0.26127104051150135,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.5113,
+      "step": 694
+    },
+    {
+      "epoch": 2.8172151898734175,
+      "grad_norm": 0.26965365618786763,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.5003,
+      "step": 695
+    },
+    {
+      "epoch": 2.82126582278481,
+      "grad_norm": 0.2898296916682902,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.5106,
+      "step": 696
+    },
+    {
+      "epoch": 2.8253164556962025,
+      "grad_norm": 0.26716898288319324,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4997,
+      "step": 697
+    },
+    {
+      "epoch": 2.829367088607595,
+      "grad_norm": 0.33580901490987125,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.5121,
+      "step": 698
+    },
+    {
+      "epoch": 2.8334177215189875,
+      "grad_norm": 0.23774930144481626,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.5055,
+      "step": 699
+    },
+    {
+      "epoch": 2.8374683544303796,
+      "grad_norm": 0.25824578623713296,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.5137,
+      "step": 700
+    },
+    {
+      "epoch": 2.841518987341772,
+      "grad_norm": 0.24826119905960592,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.5088,
+      "step": 701
+    },
+    {
+      "epoch": 2.8455696202531646,
+      "grad_norm": 0.28138739885303093,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.5091,
+      "step": 702
+    },
+    {
+      "epoch": 2.8496202531645567,
+      "grad_norm": 0.270293555468743,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.5165,
+      "step": 703
+    },
+    {
+      "epoch": 2.853670886075949,
+      "grad_norm": 0.4004094680574716,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.4946,
+      "step": 704
+    },
+    {
+      "epoch": 2.8577215189873417,
+      "grad_norm": 0.2377063266423008,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.499,
+      "step": 705
+    },
+    {
+      "epoch": 2.8617721518987342,
+      "grad_norm": 0.32249069739683334,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.519,
+      "step": 706
+    },
+    {
+      "epoch": 2.8658227848101268,
+      "grad_norm": 0.2622748114409681,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.5037,
+      "step": 707
+    },
+    {
+      "epoch": 2.869873417721519,
+      "grad_norm": 0.2795567148966245,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.5064,
+      "step": 708
+    },
+    {
+      "epoch": 2.8739240506329113,
+      "grad_norm": 0.2606518401008203,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.5141,
+      "step": 709
+    },
+    {
+      "epoch": 2.877974683544304,
+      "grad_norm": 0.2828063041743011,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.5213,
+      "step": 710
+    },
+    {
+      "epoch": 2.882025316455696,
+      "grad_norm": 0.2698562441303304,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.5027,
+      "step": 711
+    },
+    {
+      "epoch": 2.8860759493670884,
+      "grad_norm": 0.2550079356221659,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.5035,
+      "step": 712
+    },
+    {
+      "epoch": 2.890126582278481,
+      "grad_norm": 0.2635556630343899,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.5031,
+      "step": 713
+    },
+    {
+      "epoch": 2.8941772151898735,
+      "grad_norm": 0.29218630533881185,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.5067,
+      "step": 714
+    },
+    {
+      "epoch": 2.898227848101266,
+      "grad_norm": 0.24441023189111422,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.506,
+      "step": 715
+    },
+    {
+      "epoch": 2.902278481012658,
+      "grad_norm": 0.2703630327642198,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.5201,
+      "step": 716
+    },
+    {
+      "epoch": 2.9063291139240506,
+      "grad_norm": 0.28947337318152383,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.5057,
+      "step": 717
+    },
+    {
+      "epoch": 2.910379746835443,
+      "grad_norm": 1.2178768498556587,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.5533,
+      "step": 718
+    },
+    {
+      "epoch": 2.9144303797468356,
+      "grad_norm": 0.2995506829503013,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.5007,
+      "step": 719
+    },
+    {
+      "epoch": 2.918481012658228,
+      "grad_norm": 0.26203176344907514,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.5121,
+      "step": 720
+    },
+    {
+      "epoch": 2.92253164556962,
+      "grad_norm": 0.42316018259267496,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.5139,
+      "step": 721
+    },
+    {
+      "epoch": 2.9265822784810127,
+      "grad_norm": 0.3566059582349299,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.4989,
+      "step": 722
+    },
+    {
+      "epoch": 2.9306329113924052,
+      "grad_norm": 0.29124146890473973,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.5033,
+      "step": 723
+    },
+    {
+      "epoch": 2.9346835443037973,
+      "grad_norm": 0.3334143799980178,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.5027,
+      "step": 724
+    },
+    {
+      "epoch": 2.93873417721519,
+      "grad_norm": 0.27783302753308775,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.5107,
+      "step": 725
+    },
+    {
+      "epoch": 2.9427848101265823,
+      "grad_norm": 0.3473444837765498,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.5139,
+      "step": 726
+    },
+    {
+      "epoch": 2.946835443037975,
+      "grad_norm": 0.24737234729370994,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.502,
+      "step": 727
+    },
+    {
+      "epoch": 2.9508860759493674,
+      "grad_norm": 0.2916766346539096,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.4993,
+      "step": 728
+    },
+    {
+      "epoch": 2.9549367088607594,
+      "grad_norm": 0.2846705595440091,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.5128,
+      "step": 729
+    },
+    {
+      "epoch": 2.958987341772152,
+      "grad_norm": 0.26862201019889864,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.4911,
+      "step": 730
+    },
+    {
+      "epoch": 2.9630379746835445,
+      "grad_norm": 0.2720967053247452,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.5048,
+      "step": 731
+    },
+    {
+      "epoch": 2.9670886075949365,
+      "grad_norm": 0.2923310712244276,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.5096,
+      "step": 732
+    },
+    {
+      "epoch": 2.971139240506329,
+      "grad_norm": 0.26812197569847157,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.4983,
+      "step": 733
+    },
+    {
+      "epoch": 2.9751898734177216,
+      "grad_norm": 0.2956764334134365,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.511,
+      "step": 734
+    },
+    {
+      "epoch": 2.979240506329114,
+      "grad_norm": 0.2578439144208077,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.5051,
+      "step": 735
+    },
+    {
+      "epoch": 2.9832911392405066,
+      "grad_norm": 0.2962687933472538,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4983,
+      "step": 736
+    },
+    {
+      "epoch": 2.9873417721518987,
+      "grad_norm": 0.2618450146312952,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.5025,
+      "step": 737
+    },
+    {
+      "epoch": 2.991392405063291,
+      "grad_norm": 0.2986155683395708,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.5098,
+      "step": 738
+    },
+    {
+      "epoch": 2.9954430379746837,
+      "grad_norm": 0.27189423924183115,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.5045,
+      "step": 739
+    },
+    {
+      "epoch": 3.008101265822785,
+      "grad_norm": 0.4480462078259469,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.4455,
+      "step": 740
+    },
+    {
+      "epoch": 3.012151898734177,
+      "grad_norm": 0.2948988602091823,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.448,
+      "step": 741
+    },
+    {
+      "epoch": 3.0162025316455696,
+      "grad_norm": 0.5850443445119945,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.441,
+      "step": 742
+    },
+    {
+      "epoch": 3.020253164556962,
+      "grad_norm": 0.37690646069046735,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.4445,
+      "step": 743
+    },
+    {
+      "epoch": 3.0243037974683546,
+      "grad_norm": 0.411535817005082,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.4532,
+      "step": 744
+    },
+    {
+      "epoch": 3.0283544303797467,
+      "grad_norm": 0.3736115784927986,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.4464,
+      "step": 745
+    },
+    {
+      "epoch": 3.0324050632911392,
+      "grad_norm": 0.40085115167038027,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.4517,
+      "step": 746
+    },
+    {
+      "epoch": 3.0364556962025318,
+      "grad_norm": 0.3923020087422178,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.456,
+      "step": 747
+    },
+    {
+      "epoch": 3.0405063291139243,
+      "grad_norm": 0.4140926002430996,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.4592,
+      "step": 748
+    },
+    {
+      "epoch": 3.0445569620253163,
+      "grad_norm": 0.3346573211826702,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.4437,
+      "step": 749
+    },
+    {
+      "epoch": 3.048607594936709,
+      "grad_norm": 0.44256164162147354,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.4539,
+      "step": 750
+    },
+    {
+      "epoch": 3.0526582278481014,
+      "grad_norm": 0.34391866918648784,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.4482,
+      "step": 751
+    },
+    {
+      "epoch": 3.056708860759494,
+      "grad_norm": 0.4418765080788587,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.4331,
+      "step": 752
+    },
+    {
+      "epoch": 3.060759493670886,
+      "grad_norm": 0.389599409757935,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.4419,
+      "step": 753
+    },
+    {
+      "epoch": 3.0648101265822785,
+      "grad_norm": 0.35153763922883846,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.4445,
+      "step": 754
+    },
+    {
+      "epoch": 3.068860759493671,
+      "grad_norm": 0.4194834431533087,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.4354,
+      "step": 755
+    },
+    {
+      "epoch": 3.0729113924050635,
+      "grad_norm": 0.3617572854329102,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.4397,
+      "step": 756
+    },
+    {
+      "epoch": 3.0769620253164556,
+      "grad_norm": 0.37285406949562017,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.4674,
+      "step": 757
+    },
+    {
+      "epoch": 3.081012658227848,
+      "grad_norm": 0.3556184555337555,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.4337,
+      "step": 758
+    },
+    {
+      "epoch": 3.0850632911392406,
+      "grad_norm": 0.2955860619750152,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.4533,
+      "step": 759
+    },
+    {
+      "epoch": 3.089113924050633,
+      "grad_norm": 0.3851654576654011,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.4514,
+      "step": 760
+    },
+    {
+      "epoch": 3.093164556962025,
+      "grad_norm": 0.29445458166909577,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.4393,
+      "step": 761
+    },
+    {
+      "epoch": 3.0972151898734177,
+      "grad_norm": 0.32275087962698024,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.4469,
+      "step": 762
+    },
+    {
+      "epoch": 3.1012658227848102,
+      "grad_norm": 0.2846129893688357,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.4511,
+      "step": 763
+    },
+    {
+      "epoch": 3.1053164556962027,
+      "grad_norm": 0.3174150928038643,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.4423,
+      "step": 764
+    },
+    {
+      "epoch": 3.109367088607595,
+      "grad_norm": 0.2745155597743709,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.4596,
+      "step": 765
+    },
+    {
+      "epoch": 3.1134177215189873,
+      "grad_norm": 0.2949468858830593,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.4412,
+      "step": 766
+    },
+    {
+      "epoch": 3.11746835443038,
+      "grad_norm": 0.26635986929998134,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.4434,
+      "step": 767
+    },
+    {
+      "epoch": 3.1215189873417724,
+      "grad_norm": 3.9024923411316554,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.4763,
+      "step": 768
+    },
+    {
+      "epoch": 3.1255696202531644,
+      "grad_norm": 0.3570001426008588,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.4393,
+      "step": 769
+    },
+    {
+      "epoch": 3.129620253164557,
+      "grad_norm": 0.2519058863117581,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.453,
+      "step": 770
+    },
+    {
+      "epoch": 3.1336708860759495,
+      "grad_norm": 0.36146682605093167,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.4446,
+      "step": 771
+    },
+    {
+      "epoch": 3.137721518987342,
+      "grad_norm": 0.25301849905408996,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.4494,
+      "step": 772
+    },
+    {
+      "epoch": 3.141772151898734,
+      "grad_norm": 0.33066283387812356,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.4461,
+      "step": 773
+    },
+    {
+      "epoch": 3.1458227848101266,
+      "grad_norm": 0.3083280283960555,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.4475,
+      "step": 774
+    },
+    {
+      "epoch": 3.149873417721519,
+      "grad_norm": 0.29458631749354147,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.4539,
+      "step": 775
+    },
+    {
+      "epoch": 3.1539240506329116,
+      "grad_norm": 0.3192958256439885,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.4402,
+      "step": 776
+    },
+    {
+      "epoch": 3.1579746835443037,
+      "grad_norm": 0.2529527692710603,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.4415,
+      "step": 777
+    },
+    {
+      "epoch": 3.162025316455696,
+      "grad_norm": 0.28468829635749326,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.4618,
+      "step": 778
+    },
+    {
+      "epoch": 3.1660759493670887,
+      "grad_norm": 0.2816960244180696,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.4346,
+      "step": 779
+    },
+    {
+      "epoch": 3.170126582278481,
+      "grad_norm": 0.2748734179578441,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.4387,
+      "step": 780
+    },
+    {
+      "epoch": 3.1741772151898733,
+      "grad_norm": 0.2684885437452951,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.4309,
+      "step": 781
+    },
+    {
+      "epoch": 3.178227848101266,
+      "grad_norm": 0.2896179872344483,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.4505,
+      "step": 782
+    },
+    {
+      "epoch": 3.1822784810126583,
+      "grad_norm": 0.24752096145610797,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.4465,
+      "step": 783
+    },
+    {
+      "epoch": 3.186329113924051,
+      "grad_norm": 0.25796722192870225,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.4405,
+      "step": 784
+    },
+    {
+      "epoch": 3.190379746835443,
+      "grad_norm": 0.26483190540661855,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.4519,
+      "step": 785
+    },
+    {
+      "epoch": 3.1944303797468354,
+      "grad_norm": 0.26023574728845195,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.452,
+      "step": 786
+    },
+    {
+      "epoch": 3.198481012658228,
+      "grad_norm": 0.2894361315507193,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.4387,
+      "step": 787
+    },
+    {
+      "epoch": 3.2025316455696204,
+      "grad_norm": 0.23722568890535395,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.4483,
+      "step": 788
+    },
+    {
+      "epoch": 3.2065822784810125,
+      "grad_norm": 0.27603711179915474,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.4517,
+      "step": 789
+    },
+    {
+      "epoch": 3.210632911392405,
+      "grad_norm": 0.30752846596528577,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.4463,
+      "step": 790
+    },
+    {
+      "epoch": 3.2146835443037975,
+      "grad_norm": 0.23334341701426106,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.4467,
+      "step": 791
+    },
+    {
+      "epoch": 3.21873417721519,
+      "grad_norm": 0.32447748363594636,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.45,
+      "step": 792
+    },
+    {
+      "epoch": 3.222784810126582,
+      "grad_norm": 0.2690635491430007,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.447,
+      "step": 793
+    },
+    {
+      "epoch": 3.2268354430379746,
+      "grad_norm": 0.26988975088017486,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.4391,
+      "step": 794
+    },
+    {
+      "epoch": 3.230886075949367,
+      "grad_norm": 0.29515956410696,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.4574,
+      "step": 795
+    },
+    {
+      "epoch": 3.2349367088607597,
+      "grad_norm": 0.23898253831567903,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.4451,
+      "step": 796
+    },
+    {
+      "epoch": 3.2389873417721518,
+      "grad_norm": 0.2862788859016862,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.4642,
+      "step": 797
+    },
+    {
+      "epoch": 3.2430379746835443,
+      "grad_norm": 0.2682210163323285,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.4391,
+      "step": 798
+    },
+    {
+      "epoch": 3.247088607594937,
+      "grad_norm": 0.2740096986392261,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.4436,
+      "step": 799
+    },
+    {
+      "epoch": 3.2511392405063293,
+      "grad_norm": 0.272987103020129,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.4444,
+      "step": 800
+    },
+    {
+      "epoch": 3.2551898734177214,
+      "grad_norm": 0.2609251359543171,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.4532,
+      "step": 801
+    },
+    {
+      "epoch": 3.259240506329114,
+      "grad_norm": 0.2857062156041436,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.4362,
+      "step": 802
+    },
+    {
+      "epoch": 3.2632911392405064,
+      "grad_norm": 0.26781072641550147,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.4498,
+      "step": 803
+    },
+    {
+      "epoch": 3.267341772151899,
+      "grad_norm": 0.2988879474196617,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.4415,
+      "step": 804
+    },
+    {
+      "epoch": 3.271392405063291,
+      "grad_norm": 0.2551880645168497,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.4528,
+      "step": 805
+    },
+    {
+      "epoch": 3.2754430379746835,
+      "grad_norm": 0.2725509304335135,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.4446,
+      "step": 806
+    },
+    {
+      "epoch": 3.279493670886076,
+      "grad_norm": 0.28860753767449704,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.4467,
+      "step": 807
+    },
+    {
+      "epoch": 3.2835443037974685,
+      "grad_norm": 0.24835368663811058,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.4379,
+      "step": 808
+    },
+    {
+      "epoch": 3.2875949367088606,
+      "grad_norm": 0.24188395073520846,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.4477,
+      "step": 809
+    },
+    {
+      "epoch": 3.291645569620253,
+      "grad_norm": 0.24053812386589704,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.4502,
+      "step": 810
+    },
+    {
+      "epoch": 3.2956962025316456,
+      "grad_norm": 0.2517940224683123,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.4447,
+      "step": 811
+    },
+    {
+      "epoch": 3.299746835443038,
+      "grad_norm": 2.226880128609337,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.4597,
+      "step": 812
+    },
+    {
+      "epoch": 3.3037974683544302,
+      "grad_norm": 0.2647040862390811,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.4477,
+      "step": 813
+    },
+    {
+      "epoch": 3.3078481012658227,
+      "grad_norm": 0.24177747676376915,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.4546,
+      "step": 814
+    },
+    {
+      "epoch": 3.3118987341772153,
+      "grad_norm": 0.23438126592184597,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.4505,
+      "step": 815
+    },
+    {
+      "epoch": 3.3159493670886078,
+      "grad_norm": 0.2549665078136357,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.4495,
+      "step": 816
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.26213025519731087,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.4428,
+      "step": 817
+    },
+    {
+      "epoch": 3.3240506329113924,
+      "grad_norm": 0.2512791673526738,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.4431,
+      "step": 818
+    },
+    {
+      "epoch": 3.328101265822785,
+      "grad_norm": 0.24736045495283526,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.4616,
+      "step": 819
+    },
+    {
+      "epoch": 3.3321518987341774,
+      "grad_norm": 0.2724070508413059,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.4439,
+      "step": 820
+    },
+    {
+      "epoch": 3.3362025316455695,
+      "grad_norm": 0.24874681109841867,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.4466,
+      "step": 821
+    },
+    {
+      "epoch": 3.340253164556962,
+      "grad_norm": 0.2827229298471516,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.4453,
+      "step": 822
+    },
+    {
+      "epoch": 3.3443037974683545,
+      "grad_norm": 0.2745954551055071,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.4435,
+      "step": 823
+    },
+    {
+      "epoch": 3.348354430379747,
+      "grad_norm": 0.2523249361418178,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.461,
+      "step": 824
+    },
+    {
+      "epoch": 3.352405063291139,
+      "grad_norm": 0.26698312868456403,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.4532,
+      "step": 825
+    },
+    {
+      "epoch": 3.3564556962025316,
+      "grad_norm": 0.253023092418923,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.4533,
+      "step": 826
+    },
+    {
+      "epoch": 3.360506329113924,
+      "grad_norm": 0.30229793676303024,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.4675,
+      "step": 827
+    },
+    {
+      "epoch": 3.3645569620253166,
+      "grad_norm": 0.26276392140515453,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.428,
+      "step": 828
+    },
+    {
+      "epoch": 3.3686075949367087,
+      "grad_norm": 0.3081109354948334,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.4528,
+      "step": 829
+    },
+    {
+      "epoch": 3.372658227848101,
+      "grad_norm": 0.37218640233484546,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.4519,
+      "step": 830
+    },
+    {
+      "epoch": 3.3767088607594937,
+      "grad_norm": 0.27083842631305505,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.4495,
+      "step": 831
+    },
+    {
+      "epoch": 3.3807594936708862,
+      "grad_norm": 0.3079507867821389,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.4618,
+      "step": 832
+    },
+    {
+      "epoch": 3.3848101265822783,
+      "grad_norm": 0.3220071376342831,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.4546,
+      "step": 833
+    },
+    {
+      "epoch": 3.388860759493671,
+      "grad_norm": 0.3015492628183811,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.4395,
+      "step": 834
+    },
+    {
+      "epoch": 3.3929113924050633,
+      "grad_norm": 0.24249247331781212,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.4678,
+      "step": 835
+    },
+    {
+      "epoch": 3.396962025316456,
+      "grad_norm": 0.2840200763663764,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.436,
+      "step": 836
+    },
+    {
+      "epoch": 3.401012658227848,
+      "grad_norm": 0.26631631750450696,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.4344,
+      "step": 837
+    },
+    {
+      "epoch": 3.4050632911392404,
+      "grad_norm": 0.29143816210129225,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.4413,
+      "step": 838
+    },
+    {
+      "epoch": 3.409113924050633,
+      "grad_norm": 0.2816631895809057,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.4521,
+      "step": 839
+    },
+    {
+      "epoch": 3.4131645569620255,
+      "grad_norm": 0.22337083360077625,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.4357,
+      "step": 840
+    },
+    {
+      "epoch": 3.4172151898734175,
+      "grad_norm": 0.3392948090399526,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.4497,
+      "step": 841
+    },
+    {
+      "epoch": 3.42126582278481,
+      "grad_norm": 0.25969747032477064,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.4638,
+      "step": 842
+    },
+    {
+      "epoch": 3.4253164556962026,
+      "grad_norm": 0.2862919584469846,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.4534,
+      "step": 843
+    },
+    {
+      "epoch": 3.429367088607595,
+      "grad_norm": 0.30334437203138054,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.4363,
+      "step": 844
+    },
+    {
+      "epoch": 3.433417721518987,
+      "grad_norm": 0.2522226542497337,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.4411,
+      "step": 845
+    },
+    {
+      "epoch": 3.4374683544303797,
+      "grad_norm": 0.2815994171099675,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.4497,
+      "step": 846
+    },
+    {
+      "epoch": 3.441518987341772,
+      "grad_norm": 0.27197342153356907,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.4507,
+      "step": 847
+    },
+    {
+      "epoch": 3.4455696202531647,
+      "grad_norm": 0.2841837567830409,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.4296,
+      "step": 848
+    },
+    {
+      "epoch": 3.449620253164557,
+      "grad_norm": 0.2216782555517014,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.4486,
+      "step": 849
+    },
+    {
+      "epoch": 3.4536708860759493,
+      "grad_norm": 0.25716901697109323,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.4486,
+      "step": 850
+    },
+    {
+      "epoch": 3.457721518987342,
+      "grad_norm": 0.26514314143696105,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.4568,
+      "step": 851
+    },
+    {
+      "epoch": 3.4617721518987343,
+      "grad_norm": 0.2515763866947208,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.4387,
+      "step": 852
+    },
+    {
+      "epoch": 3.4658227848101264,
+      "grad_norm": 0.2454910705621161,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.4433,
+      "step": 853
+    },
+    {
+      "epoch": 3.469873417721519,
+      "grad_norm": 0.2527162010266985,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.4451,
+      "step": 854
+    },
+    {
+      "epoch": 3.4739240506329114,
+      "grad_norm": 0.24149295559415318,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.4415,
+      "step": 855
+    },
+    {
+      "epoch": 3.477974683544304,
+      "grad_norm": 0.24505161308120224,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.4391,
+      "step": 856
+    },
+    {
+      "epoch": 3.482025316455696,
+      "grad_norm": 0.22912696980809436,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.45,
+      "step": 857
+    },
+    {
+      "epoch": 3.4860759493670885,
+      "grad_norm": 0.25416343558753696,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.4556,
+      "step": 858
+    },
+    {
+      "epoch": 3.490126582278481,
+      "grad_norm": 0.2515696741367962,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.4573,
+      "step": 859
+    },
+    {
+      "epoch": 3.4941772151898736,
+      "grad_norm": 0.23157930162649745,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.4435,
+      "step": 860
+    },
+    {
+      "epoch": 3.4982278481012656,
+      "grad_norm": 0.22223191236002657,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.4456,
+      "step": 861
+    },
+    {
+      "epoch": 3.502278481012658,
+      "grad_norm": 0.2397879415844327,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.4423,
+      "step": 862
+    },
+    {
+      "epoch": 3.5063291139240507,
+      "grad_norm": 0.2489180025093492,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.4599,
+      "step": 863
+    },
+    {
+      "epoch": 3.510379746835443,
+      "grad_norm": 0.22590353001969093,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.4439,
+      "step": 864
+    },
+    {
+      "epoch": 3.5144303797468357,
+      "grad_norm": 0.2361955313702804,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.4466,
+      "step": 865
+    },
+    {
+      "epoch": 3.5184810126582278,
+      "grad_norm": 0.22210611264513594,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.4377,
+      "step": 866
+    },
+    {
+      "epoch": 3.5225316455696203,
+      "grad_norm": 0.22744615892987946,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.4626,
+      "step": 867
+    },
+    {
+      "epoch": 3.526582278481013,
+      "grad_norm": 0.2340469351934775,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.4458,
+      "step": 868
+    },
+    {
+      "epoch": 3.530632911392405,
+      "grad_norm": 0.22921031149353957,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.4503,
+      "step": 869
+    },
+    {
+      "epoch": 3.5346835443037974,
+      "grad_norm": 0.2345786490767904,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.4426,
+      "step": 870
+    },
+    {
+      "epoch": 3.53873417721519,
+      "grad_norm": 0.24888013491079877,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.4425,
+      "step": 871
+    },
+    {
+      "epoch": 3.5427848101265824,
+      "grad_norm": 0.23733618538630719,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.4443,
+      "step": 872
+    },
+    {
+      "epoch": 3.546835443037975,
+      "grad_norm": 0.22496742797445174,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.4434,
+      "step": 873
+    },
+    {
+      "epoch": 3.550886075949367,
+      "grad_norm": 0.27609391798249,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.4557,
+      "step": 874
+    },
+    {
+      "epoch": 3.5549367088607595,
+      "grad_norm": 0.22858604378540825,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.4471,
+      "step": 875
+    },
+    {
+      "epoch": 3.558987341772152,
+      "grad_norm": 0.22417600322898307,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.4344,
+      "step": 876
+    },
+    {
+      "epoch": 3.563037974683544,
+      "grad_norm": 0.24037256575518703,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.4449,
+      "step": 877
+    },
+    {
+      "epoch": 3.5670886075949366,
+      "grad_norm": 0.25305201712794245,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.4632,
+      "step": 878
+    },
+    {
+      "epoch": 3.571139240506329,
+      "grad_norm": 0.2215682843750075,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.4551,
+      "step": 879
+    },
+    {
+      "epoch": 3.5751898734177217,
+      "grad_norm": 0.23854250045351436,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.4516,
+      "step": 880
+    },
+    {
+      "epoch": 3.579240506329114,
+      "grad_norm": 0.226723674508071,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.4452,
+      "step": 881
+    },
+    {
+      "epoch": 3.5832911392405062,
+      "grad_norm": 0.25468929098778575,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.4544,
+      "step": 882
+    },
+    {
+      "epoch": 3.5873417721518988,
+      "grad_norm": 0.25374394294438435,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.4487,
+      "step": 883
+    },
+    {
+      "epoch": 3.5913924050632913,
+      "grad_norm": 0.22759570553222533,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.4413,
+      "step": 884
+    },
+    {
+      "epoch": 3.5954430379746833,
+      "grad_norm": 0.268226173248734,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.4485,
+      "step": 885
+    },
+    {
+      "epoch": 3.599493670886076,
+      "grad_norm": 0.23537054500964977,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.4432,
+      "step": 886
+    },
+    {
+      "epoch": 3.6035443037974684,
+      "grad_norm": 0.2142051016260504,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.448,
+      "step": 887
+    },
+    {
+      "epoch": 3.607594936708861,
+      "grad_norm": 0.24795470270948086,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.4422,
+      "step": 888
+    },
+    {
+      "epoch": 3.6116455696202534,
+      "grad_norm": 0.5342923385225277,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.464,
+      "step": 889
+    },
+    {
+      "epoch": 3.6156962025316455,
+      "grad_norm": 0.2413480866882507,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.4533,
+      "step": 890
+    },
+    {
+      "epoch": 3.619746835443038,
+      "grad_norm": 0.2537287965388932,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.4569,
+      "step": 891
+    },
+    {
+      "epoch": 3.6237974683544305,
+      "grad_norm": 0.25139056613126115,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.4348,
+      "step": 892
+    },
+    {
+      "epoch": 3.6278481012658226,
+      "grad_norm": 0.25598509267375175,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.4479,
+      "step": 893
+    },
+    {
+      "epoch": 3.631898734177215,
+      "grad_norm": 0.23801143148708834,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.4504,
+      "step": 894
+    },
+    {
+      "epoch": 3.6359493670886076,
+      "grad_norm": 0.2468857041739758,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.4498,
+      "step": 895
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.2364118922271143,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.4523,
+      "step": 896
+    },
+    {
+      "epoch": 3.6440506329113926,
+      "grad_norm": 0.24193704711965638,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.4455,
+      "step": 897
+    },
+    {
+      "epoch": 3.6481012658227847,
+      "grad_norm": 0.26198892925627765,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.4536,
+      "step": 898
+    },
+    {
+      "epoch": 3.6521518987341772,
+      "grad_norm": 0.22236305804444,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.4435,
+      "step": 899
+    },
+    {
+      "epoch": 3.6562025316455697,
+      "grad_norm": 0.2260602281779419,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.4561,
+      "step": 900
+    },
+    {
+      "epoch": 3.660253164556962,
+      "grad_norm": 0.22945659532711024,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.46,
+      "step": 901
+    },
+    {
+      "epoch": 3.6643037974683543,
+      "grad_norm": 0.21871301064341642,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.4466,
+      "step": 902
+    },
+    {
+      "epoch": 3.668354430379747,
+      "grad_norm": 0.2254517732551904,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.4348,
+      "step": 903
+    },
+    {
+      "epoch": 3.6724050632911394,
+      "grad_norm": 0.25381001695210226,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.4527,
+      "step": 904
+    },
+    {
+      "epoch": 3.676455696202532,
+      "grad_norm": 0.22907779040566473,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.4449,
+      "step": 905
+    },
+    {
+      "epoch": 3.680506329113924,
+      "grad_norm": 0.2418655489582189,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.4643,
+      "step": 906
+    },
+    {
+      "epoch": 3.6845569620253165,
+      "grad_norm": 0.2377352751705066,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.4667,
+      "step": 907
+    },
+    {
+      "epoch": 3.688607594936709,
+      "grad_norm": 0.24174064294084707,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.4583,
+      "step": 908
+    },
+    {
+      "epoch": 3.692658227848101,
+      "grad_norm": 0.2546052177369739,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.4366,
+      "step": 909
+    },
+    {
+      "epoch": 3.6967088607594936,
+      "grad_norm": 0.2285217037700982,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.4501,
+      "step": 910
+    },
+    {
+      "epoch": 3.700759493670886,
+      "grad_norm": 0.23001459168555385,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.4305,
+      "step": 911
+    },
+    {
+      "epoch": 3.7048101265822786,
+      "grad_norm": 0.23858650179108795,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.46,
+      "step": 912
+    },
+    {
+      "epoch": 3.708860759493671,
+      "grad_norm": 0.2227432347725773,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.4564,
+      "step": 913
+    },
+    {
+      "epoch": 3.712911392405063,
+      "grad_norm": 0.20953672534272721,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.4516,
+      "step": 914
+    },
+    {
+      "epoch": 3.7169620253164557,
+      "grad_norm": 0.22869256132794716,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.4437,
+      "step": 915
+    },
+    {
+      "epoch": 3.721012658227848,
+      "grad_norm": 0.24610419773558198,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.4394,
+      "step": 916
+    },
+    {
+      "epoch": 3.7250632911392403,
+      "grad_norm": 0.22243630350240703,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.4544,
+      "step": 917
+    },
+    {
+      "epoch": 3.729113924050633,
+      "grad_norm": 0.24396594579508624,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.4554,
+      "step": 918
+    },
+    {
+      "epoch": 3.7331645569620253,
+      "grad_norm": 0.2233596355008928,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.4596,
+      "step": 919
+    },
+    {
+      "epoch": 3.737215189873418,
+      "grad_norm": 0.2393959449095979,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.4243,
+      "step": 920
+    },
+    {
+      "epoch": 3.7412658227848103,
+      "grad_norm": 0.22245187631564373,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.452,
+      "step": 921
+    },
+    {
+      "epoch": 3.7453164556962024,
+      "grad_norm": 0.223403522703384,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.4504,
+      "step": 922
+    },
+    {
+      "epoch": 3.749367088607595,
+      "grad_norm": 0.2130012608944047,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.4589,
+      "step": 923
+    },
+    {
+      "epoch": 3.7534177215189874,
+      "grad_norm": 0.21288510841544123,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.4394,
+      "step": 924
+    },
+    {
+      "epoch": 3.7574683544303795,
+      "grad_norm": 0.2697782112494956,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.445,
+      "step": 925
+    },
+    {
+      "epoch": 3.761518987341772,
+      "grad_norm": 0.21382157161791107,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.4494,
+      "step": 926
+    },
+    {
+      "epoch": 3.7655696202531646,
+      "grad_norm": 0.22398970803140192,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.4353,
+      "step": 927
+    },
+    {
+      "epoch": 3.769620253164557,
+      "grad_norm": 0.21131930630123275,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.4596,
+      "step": 928
+    },
+    {
+      "epoch": 3.7736708860759496,
+      "grad_norm": 0.22417152537100674,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.4506,
+      "step": 929
+    },
+    {
+      "epoch": 3.7777215189873417,
+      "grad_norm": 0.2113275651541379,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.45,
+      "step": 930
+    },
+    {
+      "epoch": 3.781772151898734,
+      "grad_norm": 0.8529575841964336,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.4719,
+      "step": 931
+    },
+    {
+      "epoch": 3.7858227848101267,
+      "grad_norm": 0.2151269479584616,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.4339,
+      "step": 932
+    },
+    {
+      "epoch": 3.7898734177215188,
+      "grad_norm": 0.20803957717161944,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.4515,
+      "step": 933
+    },
+    {
+      "epoch": 3.7939240506329113,
+      "grad_norm": 0.23175305221136308,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.4397,
+      "step": 934
+    },
+    {
+      "epoch": 3.797974683544304,
+      "grad_norm": 0.23161130502118327,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.4414,
+      "step": 935
+    },
+    {
+      "epoch": 3.8020253164556963,
+      "grad_norm": 0.22337419434479508,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.4424,
+      "step": 936
+    },
+    {
+      "epoch": 3.806075949367089,
+      "grad_norm": 0.23319178895912746,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.4442,
+      "step": 937
+    },
+    {
+      "epoch": 3.810126582278481,
+      "grad_norm": 0.6170049189132342,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.4482,
+      "step": 938
+    },
+    {
+      "epoch": 3.8141772151898734,
+      "grad_norm": 0.20672579224235424,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.4438,
+      "step": 939
+    },
+    {
+      "epoch": 3.818227848101266,
+      "grad_norm": 0.23920739804333963,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.4576,
+      "step": 940
+    },
+    {
+      "epoch": 3.822278481012658,
+      "grad_norm": 0.2447337282034391,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.4415,
+      "step": 941
+    },
+    {
+      "epoch": 3.8263291139240505,
+      "grad_norm": 0.21470285332470745,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.4437,
+      "step": 942
+    },
+    {
+      "epoch": 3.830379746835443,
+      "grad_norm": 0.2217994448210214,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.4538,
+      "step": 943
+    },
+    {
+      "epoch": 3.8344303797468355,
+      "grad_norm": 0.23443749523337348,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.4423,
+      "step": 944
+    },
+    {
+      "epoch": 3.838481012658228,
+      "grad_norm": 0.2374445398141554,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.4556,
+      "step": 945
+    },
+    {
+      "epoch": 3.84253164556962,
+      "grad_norm": 0.21801224244705117,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.4489,
+      "step": 946
+    },
+    {
+      "epoch": 3.8465822784810126,
+      "grad_norm": 0.21699896641917796,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.4497,
+      "step": 947
+    },
+    {
+      "epoch": 3.850632911392405,
+      "grad_norm": 0.20788344234666484,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.4593,
+      "step": 948
+    },
+    {
+      "epoch": 3.8546835443037972,
+      "grad_norm": 0.21257145944352526,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.4359,
+      "step": 949
+    },
+    {
+      "epoch": 3.8587341772151897,
+      "grad_norm": 0.22901971510706456,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.4483,
+      "step": 950
+    },
+    {
+      "epoch": 3.8627848101265823,
+      "grad_norm": 0.21517154700746094,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.429,
+      "step": 951
+    },
+    {
+      "epoch": 3.8668354430379748,
+      "grad_norm": 0.23004092937900136,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.4138,
+      "step": 952
+    },
+    {
+      "epoch": 3.8708860759493673,
+      "grad_norm": 0.22605559472909653,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.4412,
+      "step": 953
+    },
+    {
+      "epoch": 3.8749367088607594,
+      "grad_norm": 0.2283233452177007,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.4523,
+      "step": 954
+    },
+    {
+      "epoch": 3.878987341772152,
+      "grad_norm": 0.21848194441852775,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.432,
+      "step": 955
+    },
+    {
+      "epoch": 3.8830379746835444,
+      "grad_norm": 0.21777171915400334,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.4438,
+      "step": 956
+    },
+    {
+      "epoch": 3.8870886075949365,
+      "grad_norm": 0.236211157361395,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.4537,
+      "step": 957
+    },
+    {
+      "epoch": 3.891139240506329,
+      "grad_norm": 0.23636039876049214,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.4498,
+      "step": 958
+    },
+    {
+      "epoch": 3.8951898734177215,
+      "grad_norm": 0.23417170184720734,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.4483,
+      "step": 959
+    },
+    {
+      "epoch": 3.899240506329114,
+      "grad_norm": 0.22080921954914626,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.4679,
+      "step": 960
+    },
+    {
+      "epoch": 3.9032911392405065,
+      "grad_norm": 0.22608126140570667,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.4427,
+      "step": 961
+    },
+    {
+      "epoch": 3.9073417721518986,
+      "grad_norm": 0.22125043574001615,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.4584,
+      "step": 962
+    },
+    {
+      "epoch": 3.911392405063291,
+      "grad_norm": 0.2154229094092577,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.4482,
+      "step": 963
+    },
+    {
+      "epoch": 3.9154430379746836,
+      "grad_norm": 0.2144872996289274,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.4595,
+      "step": 964
+    },
+    {
+      "epoch": 3.9194936708860757,
+      "grad_norm": 0.21686982382197117,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.4466,
+      "step": 965
+    },
+    {
+      "epoch": 3.923544303797468,
+      "grad_norm": 0.22961980312072844,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.4504,
+      "step": 966
+    },
+    {
+      "epoch": 3.9275949367088607,
+      "grad_norm": 0.2092791024157416,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.4435,
+      "step": 967
+    },
+    {
+      "epoch": 3.9316455696202532,
+      "grad_norm": 0.22353783551699197,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.4552,
+      "step": 968
+    },
+    {
+      "epoch": 3.9356962025316458,
+      "grad_norm": 0.21355801601007518,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.4331,
+      "step": 969
+    },
+    {
+      "epoch": 3.939746835443038,
+      "grad_norm": 0.2182688753308274,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.4389,
+      "step": 970
+    },
+    {
+      "epoch": 3.9437974683544303,
+      "grad_norm": 0.21407128965451894,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.4417,
+      "step": 971
+    },
+    {
+      "epoch": 3.947848101265823,
+      "grad_norm": 0.21897518749412,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.4567,
+      "step": 972
+    },
+    {
+      "epoch": 3.951898734177215,
+      "grad_norm": 0.20282833683806223,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.4581,
+      "step": 973
+    },
+    {
+      "epoch": 3.9559493670886074,
+      "grad_norm": 0.20086137430407427,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.4353,
+      "step": 974
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.21329418498097105,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.4517,
+      "step": 975
+    },
+    {
+      "epoch": 3.9640506329113925,
+      "grad_norm": 0.21709363524029268,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.4436,
+      "step": 976
+    },
+    {
+      "epoch": 3.968101265822785,
+      "grad_norm": 0.20489594684554546,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.4494,
+      "step": 977
+    },
+    {
+      "epoch": 3.972151898734177,
+      "grad_norm": 0.203016242345807,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.4523,
+      "step": 978
+    },
+    {
+      "epoch": 3.9762025316455696,
+      "grad_norm": 0.191967543318881,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.4509,
+      "step": 979
+    },
+    {
+      "epoch": 3.980253164556962,
+      "grad_norm": 0.21084258805680936,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.4488,
+      "step": 980
+    },
+    {
+      "epoch": 3.984303797468354,
+      "grad_norm": 0.202437011909258,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.4441,
+      "step": 981
+    },
+    {
+      "epoch": 3.9883544303797467,
+      "grad_norm": 0.21011902077835237,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.4264,
+      "step": 982
+    },
+    {
+      "epoch": 3.992405063291139,
+      "grad_norm": 0.20708427512884406,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.4539,
+      "step": 983
+    },
+    {
+      "epoch": 3.9964556962025317,
+      "grad_norm": 0.20333666798156108,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.4534,
+      "step": 984
+    },
+    {
+      "epoch": 4.000506329113924,
+      "grad_norm": 0.22218992212198657,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.4267,
+      "step": 985
+    },
+    {
+      "epoch": 4.004556962025316,
+      "grad_norm": 0.39968292538602546,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.4025,
+      "step": 986
+    },
+    {
+      "epoch": 4.008607594936709,
+      "grad_norm": 0.2771926101773036,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.3942,
+      "step": 987
+    },
+    {
+      "epoch": 4.012658227848101,
+      "grad_norm": 0.2971660642048338,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.4035,
+      "step": 988
+    },
+    {
+      "epoch": 4.016708860759493,
+      "grad_norm": 0.44422915371298055,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.4161,
+      "step": 989
+    },
+    {
+      "epoch": 4.020759493670886,
+      "grad_norm": 0.3306704307138634,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.3917,
+      "step": 990
+    },
+    {
+      "epoch": 4.024810126582278,
+      "grad_norm": 0.25659051159985014,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.414,
+      "step": 991
+    },
+    {
+      "epoch": 4.0288607594936705,
+      "grad_norm": 0.354026356823384,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.4306,
+      "step": 992
+    },
+    {
+      "epoch": 4.0329113924050635,
+      "grad_norm": 0.3833769762852071,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.4014,
+      "step": 993
+    },
+    {
+      "epoch": 4.0369620253164555,
+      "grad_norm": 0.26465634253817855,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.4076,
+      "step": 994
+    },
+    {
+      "epoch": 4.0410126582278485,
+      "grad_norm": 0.2387210183874982,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.4202,
+      "step": 995
+    },
+    {
+      "epoch": 4.045063291139241,
+      "grad_norm": 0.320700137650539,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.4045,
+      "step": 996
+    },
+    {
+      "epoch": 4.049113924050633,
+      "grad_norm": 0.31978988712437006,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.4063,
+      "step": 997
+    },
+    {
+      "epoch": 4.053164556962026,
+      "grad_norm": 0.26364898827145433,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.4149,
+      "step": 998
+    },
+    {
+      "epoch": 4.057215189873418,
+      "grad_norm": 0.21776925807999903,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.3982,
+      "step": 999
+    },
+    {
+      "epoch": 4.06126582278481,
+      "grad_norm": 0.24636623195282276,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.4191,
+      "step": 1000
+    },
+    {
+      "epoch": 4.065316455696203,
+      "grad_norm": 0.29169332637141365,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.4003,
+      "step": 1001
+    },
+    {
+      "epoch": 4.069367088607595,
+      "grad_norm": 0.2510059674699072,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.4118,
+      "step": 1002
+    },
+    {
+      "epoch": 4.073417721518988,
+      "grad_norm": 0.2204207439831788,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.4041,
+      "step": 1003
+    },
+    {
+      "epoch": 4.07746835443038,
+      "grad_norm": 0.24585067873094368,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.4044,
+      "step": 1004
+    },
+    {
+      "epoch": 4.081518987341772,
+      "grad_norm": 0.2397601962996456,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.4058,
+      "step": 1005
+    },
+    {
+      "epoch": 4.085569620253165,
+      "grad_norm": 0.2574102471181848,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.4087,
+      "step": 1006
+    },
+    {
+      "epoch": 4.089620253164557,
+      "grad_norm": 0.23599988408076225,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.4087,
+      "step": 1007
+    },
+    {
+      "epoch": 4.093670886075949,
+      "grad_norm": 0.22091197010493976,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.4242,
+      "step": 1008
+    },
+    {
+      "epoch": 4.097721518987342,
+      "grad_norm": 0.24898062662916065,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.401,
+      "step": 1009
+    },
+    {
+      "epoch": 4.101772151898734,
+      "grad_norm": 0.25378626946621113,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.3983,
+      "step": 1010
+    },
+    {
+      "epoch": 4.105822784810127,
+      "grad_norm": 0.22433504421496767,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.4146,
+      "step": 1011
+    },
+    {
+      "epoch": 4.109873417721519,
+      "grad_norm": 0.23713901751110952,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.4188,
+      "step": 1012
+    },
+    {
+      "epoch": 4.113924050632911,
+      "grad_norm": 0.21893544084768105,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.4087,
+      "step": 1013
+    },
+    {
+      "epoch": 4.117974683544304,
+      "grad_norm": 0.23256870395332865,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.4044,
+      "step": 1014
+    },
+    {
+      "epoch": 4.122025316455696,
+      "grad_norm": 0.2387361205305584,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.4029,
+      "step": 1015
+    },
+    {
+      "epoch": 4.126075949367088,
+      "grad_norm": 0.21070985612404913,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.3954,
+      "step": 1016
+    },
+    {
+      "epoch": 4.130126582278481,
+      "grad_norm": 0.2366806559314245,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.4169,
+      "step": 1017
+    },
+    {
+      "epoch": 4.134177215189873,
+      "grad_norm": 0.2342468242978536,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.4002,
+      "step": 1018
+    },
+    {
+      "epoch": 4.138227848101266,
+      "grad_norm": 0.22948624376573448,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.4049,
+      "step": 1019
+    },
+    {
+      "epoch": 4.142278481012658,
+      "grad_norm": 0.23078589376497644,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.4224,
+      "step": 1020
+    },
+    {
+      "epoch": 4.14632911392405,
+      "grad_norm": 0.21449286621947375,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.4089,
+      "step": 1021
+    },
+    {
+      "epoch": 4.150379746835443,
+      "grad_norm": 0.2370912362019178,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.4112,
+      "step": 1022
+    },
+    {
+      "epoch": 4.154430379746835,
+      "grad_norm": 0.21891111792515203,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.3981,
+      "step": 1023
+    },
+    {
+      "epoch": 4.1584810126582274,
+      "grad_norm": 0.25925301081768826,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.404,
+      "step": 1024
+    },
+    {
+      "epoch": 4.16253164556962,
+      "grad_norm": 0.21255797141231741,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.4153,
+      "step": 1025
+    },
+    {
+      "epoch": 4.1665822784810125,
+      "grad_norm": 0.21103839575315822,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.4106,
+      "step": 1026
+    },
+    {
+      "epoch": 4.170632911392405,
+      "grad_norm": 0.20778501937979213,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.4051,
+      "step": 1027
+    },
+    {
+      "epoch": 4.1746835443037975,
+      "grad_norm": 0.2241552673557978,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.3909,
+      "step": 1028
+    },
+    {
+      "epoch": 4.17873417721519,
+      "grad_norm": 0.20249382762460283,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.4073,
+      "step": 1029
+    },
+    {
+      "epoch": 4.1827848101265825,
+      "grad_norm": 0.2136935552622026,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.3919,
+      "step": 1030
+    },
+    {
+      "epoch": 4.186835443037975,
+      "grad_norm": 0.21337940682604117,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.418,
+      "step": 1031
+    },
+    {
+      "epoch": 4.190886075949367,
+      "grad_norm": 0.22055835347878183,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.4108,
+      "step": 1032
+    },
+    {
+      "epoch": 4.19493670886076,
+      "grad_norm": 0.22758801487764477,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.4055,
+      "step": 1033
+    },
+    {
+      "epoch": 4.198987341772152,
+      "grad_norm": 0.3624594180179317,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.4158,
+      "step": 1034
+    },
+    {
+      "epoch": 4.203037974683545,
+      "grad_norm": 0.20871345253760487,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.4064,
+      "step": 1035
+    },
+    {
+      "epoch": 4.207088607594937,
+      "grad_norm": 0.20380199223923667,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.427,
+      "step": 1036
+    },
+    {
+      "epoch": 4.211139240506329,
+      "grad_norm": 0.20952319637807804,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.4112,
+      "step": 1037
+    },
+    {
+      "epoch": 4.215189873417722,
+      "grad_norm": 0.22130972622472667,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.4056,
+      "step": 1038
+    },
+    {
+      "epoch": 4.219240506329114,
+      "grad_norm": 0.21280740863593872,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.3978,
+      "step": 1039
+    },
+    {
+      "epoch": 4.223291139240506,
+      "grad_norm": 0.21176582761576637,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.4001,
+      "step": 1040
+    },
+    {
+      "epoch": 4.227341772151899,
+      "grad_norm": 0.2150208972132154,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.4172,
+      "step": 1041
+    },
+    {
+      "epoch": 4.231392405063291,
+      "grad_norm": 0.2368063542275589,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.4067,
+      "step": 1042
+    },
+    {
+      "epoch": 4.235443037974684,
+      "grad_norm": 0.21460494789803117,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.405,
+      "step": 1043
+    },
+    {
+      "epoch": 4.239493670886076,
+      "grad_norm": 0.20967153926382215,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.4226,
+      "step": 1044
+    },
+    {
+      "epoch": 4.243544303797468,
+      "grad_norm": 0.21900332027450736,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.3921,
+      "step": 1045
+    },
+    {
+      "epoch": 4.247594936708861,
+      "grad_norm": 0.23239371135777678,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.407,
+      "step": 1046
+    },
+    {
+      "epoch": 4.251645569620253,
+      "grad_norm": 0.20160247221396504,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.419,
+      "step": 1047
+    },
+    {
+      "epoch": 4.255696202531645,
+      "grad_norm": 0.21141178170434677,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.4077,
+      "step": 1048
+    },
+    {
+      "epoch": 4.259746835443038,
+      "grad_norm": 0.21723079138319984,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.4064,
+      "step": 1049
+    },
+    {
+      "epoch": 4.26379746835443,
+      "grad_norm": 0.20257416501196068,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.4044,
+      "step": 1050
+    },
+    {
+      "epoch": 4.267848101265823,
+      "grad_norm": 0.20541315035589677,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.3908,
+      "step": 1051
+    },
+    {
+      "epoch": 4.271898734177215,
+      "grad_norm": 0.20184588804459785,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.4093,
+      "step": 1052
+    },
+    {
+      "epoch": 4.275949367088607,
+      "grad_norm": 0.21645415733723788,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.3987,
+      "step": 1053
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.20496248603072353,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.4134,
+      "step": 1054
+    },
+    {
+      "epoch": 4.284050632911392,
+      "grad_norm": 0.1925908208032047,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.4139,
+      "step": 1055
+    },
+    {
+      "epoch": 4.288101265822785,
+      "grad_norm": 0.20483259191127556,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.3866,
+      "step": 1056
+    },
+    {
+      "epoch": 4.292151898734177,
+      "grad_norm": 0.21024021394313674,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.3968,
+      "step": 1057
+    },
+    {
+      "epoch": 4.296202531645569,
+      "grad_norm": 0.2052231066564629,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.4043,
+      "step": 1058
+    },
+    {
+      "epoch": 4.300253164556962,
+      "grad_norm": 0.21341223330107098,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.3869,
+      "step": 1059
+    },
+    {
+      "epoch": 4.3043037974683545,
+      "grad_norm": 0.21548818863830582,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.4197,
+      "step": 1060
+    },
+    {
+      "epoch": 4.3083544303797465,
+      "grad_norm": 0.20269856952825668,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.3997,
+      "step": 1061
+    },
+    {
+      "epoch": 4.3124050632911395,
+      "grad_norm": 0.2004000202760375,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.4104,
+      "step": 1062
+    },
+    {
+      "epoch": 4.3164556962025316,
+      "grad_norm": 0.18937084470328602,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.4134,
+      "step": 1063
+    },
+    {
+      "epoch": 4.320506329113924,
+      "grad_norm": 0.19864639441663118,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.4085,
+      "step": 1064
+    },
+    {
+      "epoch": 4.324556962025317,
+      "grad_norm": 0.20983440818341706,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.4052,
+      "step": 1065
+    },
+    {
+      "epoch": 4.328607594936709,
+      "grad_norm": 0.2046796994952921,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.3998,
+      "step": 1066
+    },
+    {
+      "epoch": 4.332658227848102,
+      "grad_norm": 0.1949948575565398,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.4067,
+      "step": 1067
+    },
+    {
+      "epoch": 4.336708860759494,
+      "grad_norm": 0.1958311443069672,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.4175,
+      "step": 1068
+    },
+    {
+      "epoch": 4.340759493670886,
+      "grad_norm": 0.21302169795783985,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.4197,
+      "step": 1069
+    },
+    {
+      "epoch": 4.344810126582279,
+      "grad_norm": 0.20879585758156688,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.406,
+      "step": 1070
+    },
+    {
+      "epoch": 4.348860759493671,
+      "grad_norm": 0.20273942434392045,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.4166,
+      "step": 1071
+    },
+    {
+      "epoch": 4.352911392405064,
+      "grad_norm": 0.1986512450684521,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.4037,
+      "step": 1072
+    },
+    {
+      "epoch": 4.356962025316456,
+      "grad_norm": 0.19028994142065184,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.4215,
+      "step": 1073
+    },
+    {
+      "epoch": 4.361012658227848,
+      "grad_norm": 0.19673314520536642,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.4059,
+      "step": 1074
+    },
+    {
+      "epoch": 4.365063291139241,
+      "grad_norm": 0.1996437089826445,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.4079,
+      "step": 1075
+    },
+    {
+      "epoch": 4.369113924050633,
+      "grad_norm": 0.19868587592022566,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.4066,
+      "step": 1076
+    },
+    {
+      "epoch": 4.373164556962025,
+      "grad_norm": 0.19843659089121346,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.4201,
+      "step": 1077
+    },
+    {
+      "epoch": 4.377215189873418,
+      "grad_norm": 0.20075972462808211,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.4032,
+      "step": 1078
+    },
+    {
+      "epoch": 4.38126582278481,
+      "grad_norm": 0.2043812950964407,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.3985,
+      "step": 1079
+    },
+    {
+      "epoch": 4.385316455696202,
+      "grad_norm": 0.20164998196815076,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.3974,
+      "step": 1080
+    },
+    {
+      "epoch": 4.389367088607595,
+      "grad_norm": 0.20036682958281063,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.3995,
+      "step": 1081
+    },
+    {
+      "epoch": 4.393417721518987,
+      "grad_norm": 0.19781502592390007,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.4272,
+      "step": 1082
+    },
+    {
+      "epoch": 4.39746835443038,
+      "grad_norm": 0.19731198073507403,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.4047,
+      "step": 1083
+    },
+    {
+      "epoch": 4.401518987341772,
+      "grad_norm": 0.21095595427030997,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.4016,
+      "step": 1084
+    },
+    {
+      "epoch": 4.405569620253164,
+      "grad_norm": 0.21177090636246354,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.4071,
+      "step": 1085
+    },
+    {
+      "epoch": 4.409620253164557,
+      "grad_norm": 0.19368885640026698,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.4098,
+      "step": 1086
+    },
+    {
+      "epoch": 4.413670886075949,
+      "grad_norm": 0.19719948804618387,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.4033,
+      "step": 1087
+    },
+    {
+      "epoch": 4.417721518987342,
+      "grad_norm": 0.20065414409516802,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.4003,
+      "step": 1088
+    },
+    {
+      "epoch": 4.421772151898734,
+      "grad_norm": 0.31253618278740614,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.3849,
+      "step": 1089
+    },
+    {
+      "epoch": 4.425822784810126,
+      "grad_norm": 0.19842650687770316,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.397,
+      "step": 1090
+    },
+    {
+      "epoch": 4.429873417721519,
+      "grad_norm": 0.20287128683094288,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.4035,
+      "step": 1091
+    },
+    {
+      "epoch": 4.433924050632911,
+      "grad_norm": 0.2061143611217459,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.4356,
+      "step": 1092
+    },
+    {
+      "epoch": 4.4379746835443035,
+      "grad_norm": 0.1933633790537764,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.4243,
+      "step": 1093
+    },
+    {
+      "epoch": 4.442025316455696,
+      "grad_norm": 0.19650481841408374,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.4198,
+      "step": 1094
+    },
+    {
+      "epoch": 4.4460759493670885,
+      "grad_norm": 0.19901968265054434,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.4062,
+      "step": 1095
+    },
+    {
+      "epoch": 4.450126582278481,
+      "grad_norm": 0.19929823776817737,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.4102,
+      "step": 1096
+    },
+    {
+      "epoch": 4.4541772151898735,
+      "grad_norm": 0.20789276606331528,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.3947,
+      "step": 1097
+    },
+    {
+      "epoch": 4.458227848101266,
+      "grad_norm": 0.19640310783438553,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.4081,
+      "step": 1098
+    },
+    {
+      "epoch": 4.462278481012659,
+      "grad_norm": 0.20353642714714354,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.407,
+      "step": 1099
+    },
+    {
+      "epoch": 4.466329113924051,
+      "grad_norm": 0.1936980994774482,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.4243,
+      "step": 1100
+    },
+    {
+      "epoch": 4.470379746835443,
+      "grad_norm": 0.2013438080027856,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.4043,
+      "step": 1101
+    },
+    {
+      "epoch": 4.474430379746836,
+      "grad_norm": 0.19741984089922573,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.3979,
+      "step": 1102
+    },
+    {
+      "epoch": 4.478481012658228,
+      "grad_norm": 0.21278735025311674,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.4029,
+      "step": 1103
+    },
+    {
+      "epoch": 4.482531645569621,
+      "grad_norm": 0.19600032037701556,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.4159,
+      "step": 1104
+    },
+    {
+      "epoch": 4.486582278481013,
+      "grad_norm": 0.19048065693642574,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.3994,
+      "step": 1105
+    },
+    {
+      "epoch": 4.490632911392405,
+      "grad_norm": 0.19449058625454602,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.4177,
+      "step": 1106
+    },
+    {
+      "epoch": 4.494683544303798,
+      "grad_norm": 0.19707018061437773,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.4069,
+      "step": 1107
+    },
+    {
+      "epoch": 4.49873417721519,
+      "grad_norm": 0.2018138565282781,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.426,
+      "step": 1108
+    },
+    {
+      "epoch": 4.502784810126582,
+      "grad_norm": 0.19943849060186936,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.397,
+      "step": 1109
+    },
+    {
+      "epoch": 4.506835443037975,
+      "grad_norm": 0.19487864594432122,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.4166,
+      "step": 1110
+    },
+    {
+      "epoch": 4.510886075949367,
+      "grad_norm": 0.19441819871752783,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.4132,
+      "step": 1111
+    },
+    {
+      "epoch": 4.514936708860759,
+      "grad_norm": 0.19442294270755198,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.4133,
+      "step": 1112
+    },
+    {
+      "epoch": 4.518987341772152,
+      "grad_norm": 0.1917967537745069,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.3969,
+      "step": 1113
+    },
+    {
+      "epoch": 4.523037974683544,
+      "grad_norm": 0.2013911719872365,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.4081,
+      "step": 1114
+    },
+    {
+      "epoch": 4.527088607594937,
+      "grad_norm": 0.19131336543516486,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.427,
+      "step": 1115
+    },
+    {
+      "epoch": 4.531139240506329,
+      "grad_norm": 0.19685136376219184,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.4226,
+      "step": 1116
+    },
+    {
+      "epoch": 4.535189873417721,
+      "grad_norm": 0.19389834390549243,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.4123,
+      "step": 1117
+    },
+    {
+      "epoch": 4.539240506329114,
+      "grad_norm": 0.19699226662396452,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.413,
+      "step": 1118
+    },
+    {
+      "epoch": 4.543291139240506,
+      "grad_norm": 0.19782175385497514,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.4168,
+      "step": 1119
+    },
+    {
+      "epoch": 4.547341772151899,
+      "grad_norm": 0.2058143940421534,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.4011,
+      "step": 1120
+    },
+    {
+      "epoch": 4.551392405063291,
+      "grad_norm": 0.1892912621106266,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.432,
+      "step": 1121
+    },
+    {
+      "epoch": 4.555443037974683,
+      "grad_norm": 0.1911324355140839,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.4059,
+      "step": 1122
+    },
+    {
+      "epoch": 4.559493670886076,
+      "grad_norm": 0.20546617819948967,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.4041,
+      "step": 1123
+    },
+    {
+      "epoch": 4.563544303797468,
+      "grad_norm": 0.2379317276672048,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.4072,
+      "step": 1124
+    },
+    {
+      "epoch": 4.56759493670886,
+      "grad_norm": 0.19436980815844065,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.4214,
+      "step": 1125
+    },
+    {
+      "epoch": 4.571645569620253,
+      "grad_norm": 0.1909807429183004,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.3931,
+      "step": 1126
+    },
+    {
+      "epoch": 4.575696202531645,
+      "grad_norm": 0.19543970630120377,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.4157,
+      "step": 1127
+    },
+    {
+      "epoch": 4.5797468354430375,
+      "grad_norm": 0.19621320682917884,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.4219,
+      "step": 1128
+    },
+    {
+      "epoch": 4.5837974683544305,
+      "grad_norm": 0.19848108621617705,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.403,
+      "step": 1129
+    },
+    {
+      "epoch": 4.5878481012658225,
+      "grad_norm": 0.268144410225894,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.4049,
+      "step": 1130
+    },
+    {
+      "epoch": 4.5918987341772155,
+      "grad_norm": 0.19915661509101446,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.4026,
+      "step": 1131
+    },
+    {
+      "epoch": 4.595949367088608,
+      "grad_norm": 0.19005772015544425,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.3966,
+      "step": 1132
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.19663846136475643,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.4043,
+      "step": 1133
+    },
+    {
+      "epoch": 4.604050632911393,
+      "grad_norm": 0.1926782073253448,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.3922,
+      "step": 1134
+    },
+    {
+      "epoch": 4.608101265822785,
+      "grad_norm": 0.20224089494094388,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.4094,
+      "step": 1135
+    },
+    {
+      "epoch": 4.612151898734178,
+      "grad_norm": 0.2009618790535245,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.4145,
+      "step": 1136
+    },
+    {
+      "epoch": 4.61620253164557,
+      "grad_norm": 0.1917329577765777,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.4056,
+      "step": 1137
+    },
+    {
+      "epoch": 4.620253164556962,
+      "grad_norm": 0.1892481972939858,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.4068,
+      "step": 1138
+    },
+    {
+      "epoch": 4.624303797468355,
+      "grad_norm": 0.20532372251931647,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.3981,
+      "step": 1139
+    },
+    {
+      "epoch": 4.628354430379747,
+      "grad_norm": 0.1948336164757943,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.4019,
+      "step": 1140
+    },
+    {
+      "epoch": 4.632405063291139,
+      "grad_norm": 0.20192826013147716,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.3984,
+      "step": 1141
+    },
+    {
+      "epoch": 4.636455696202532,
+      "grad_norm": 0.18592417943669104,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.4224,
+      "step": 1142
+    },
+    {
+      "epoch": 4.640506329113924,
+      "grad_norm": 0.20037193323217048,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.3905,
+      "step": 1143
+    },
+    {
+      "epoch": 4.644556962025316,
+      "grad_norm": 0.1942735124759689,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.407,
+      "step": 1144
+    },
+    {
+      "epoch": 4.648607594936709,
+      "grad_norm": 0.19661159184744803,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.3975,
+      "step": 1145
+    },
+    {
+      "epoch": 4.652658227848101,
+      "grad_norm": 0.1955082344899654,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.4126,
+      "step": 1146
+    },
+    {
+      "epoch": 4.656708860759494,
+      "grad_norm": 0.2165645847003483,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.4162,
+      "step": 1147
+    },
+    {
+      "epoch": 4.660759493670886,
+      "grad_norm": 0.19152749637056807,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.4236,
+      "step": 1148
+    },
+    {
+      "epoch": 4.664810126582278,
+      "grad_norm": 0.1908724332393585,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.4166,
+      "step": 1149
+    },
+    {
+      "epoch": 4.668860759493671,
+      "grad_norm": 0.19387659183449066,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.4175,
+      "step": 1150
+    },
+    {
+      "epoch": 4.672911392405063,
+      "grad_norm": 0.19478543740739343,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.4034,
+      "step": 1151
+    },
+    {
+      "epoch": 4.676962025316456,
+      "grad_norm": 0.19726977762514972,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.3864,
+      "step": 1152
+    },
+    {
+      "epoch": 4.681012658227848,
+      "grad_norm": 0.1987206671686503,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.3998,
+      "step": 1153
+    },
+    {
+      "epoch": 4.68506329113924,
+      "grad_norm": 0.19712011652284303,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.3969,
+      "step": 1154
+    },
+    {
+      "epoch": 4.689113924050633,
+      "grad_norm": 0.2127136667685158,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.3848,
+      "step": 1155
+    },
+    {
+      "epoch": 4.693164556962025,
+      "grad_norm": 0.199708408412489,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.4074,
+      "step": 1156
+    },
+    {
+      "epoch": 4.697215189873417,
+      "grad_norm": 0.19985563863453643,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.4101,
+      "step": 1157
+    },
+    {
+      "epoch": 4.70126582278481,
+      "grad_norm": 0.1936775807336959,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.4163,
+      "step": 1158
+    },
+    {
+      "epoch": 4.705316455696202,
+      "grad_norm": 0.18804738712775498,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.4157,
+      "step": 1159
+    },
+    {
+      "epoch": 4.7093670886075945,
+      "grad_norm": 0.18953256319310097,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.4052,
+      "step": 1160
+    },
+    {
+      "epoch": 4.713417721518987,
+      "grad_norm": 0.20678604856908117,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.3979,
+      "step": 1161
+    },
+    {
+      "epoch": 4.7174683544303795,
+      "grad_norm": 0.19405685388328026,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.409,
+      "step": 1162
+    },
+    {
+      "epoch": 4.7215189873417724,
+      "grad_norm": 0.20433228419384958,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.4156,
+      "step": 1163
+    },
+    {
+      "epoch": 4.7255696202531645,
+      "grad_norm": 0.1971794748286958,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.4193,
+      "step": 1164
+    },
+    {
+      "epoch": 4.729620253164557,
+      "grad_norm": 0.1968983370573593,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.4008,
+      "step": 1165
+    },
+    {
+      "epoch": 4.7336708860759495,
+      "grad_norm": 0.20170768001612216,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.4093,
+      "step": 1166
+    },
+    {
+      "epoch": 4.737721518987342,
+      "grad_norm": 0.1993102731656133,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.4027,
+      "step": 1167
+    },
+    {
+      "epoch": 4.741772151898735,
+      "grad_norm": 0.19829749231068863,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.4069,
+      "step": 1168
+    },
+    {
+      "epoch": 4.745822784810127,
+      "grad_norm": 0.18917158830349837,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.4111,
+      "step": 1169
+    },
+    {
+      "epoch": 4.749873417721519,
+      "grad_norm": 0.1938693435821872,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.4136,
+      "step": 1170
+    },
+    {
+      "epoch": 4.753924050632912,
+      "grad_norm": 0.1968048922502675,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.4129,
+      "step": 1171
+    },
+    {
+      "epoch": 4.757974683544304,
+      "grad_norm": 0.1922001043529497,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.4177,
+      "step": 1172
+    },
+    {
+      "epoch": 4.762025316455696,
+      "grad_norm": 0.19570487848771065,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.4122,
+      "step": 1173
+    },
+    {
+      "epoch": 4.766075949367089,
+      "grad_norm": 0.1969134259230168,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.3921,
+      "step": 1174
+    },
+    {
+      "epoch": 4.770126582278481,
+      "grad_norm": 0.19102248268397867,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.4154,
+      "step": 1175
+    },
+    {
+      "epoch": 4.774177215189873,
+      "grad_norm": 0.2067022765139133,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.3966,
+      "step": 1176
+    },
+    {
+      "epoch": 4.778227848101266,
+      "grad_norm": 0.19471559086327075,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.4008,
+      "step": 1177
+    },
+    {
+      "epoch": 4.782278481012658,
+      "grad_norm": 0.19398652633304167,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.3979,
+      "step": 1178
+    },
+    {
+      "epoch": 4.786329113924051,
+      "grad_norm": 0.19522359290894015,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.4077,
+      "step": 1179
+    },
+    {
+      "epoch": 4.790379746835443,
+      "grad_norm": 0.189950203243214,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.4152,
+      "step": 1180
+    },
+    {
+      "epoch": 4.794430379746835,
+      "grad_norm": 0.19324423312288133,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.408,
+      "step": 1181
+    },
+    {
+      "epoch": 4.798481012658228,
+      "grad_norm": 0.19478245809441858,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.4047,
+      "step": 1182
+    },
+    {
+      "epoch": 4.80253164556962,
+      "grad_norm": 0.19130517220825186,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.4173,
+      "step": 1183
+    },
+    {
+      "epoch": 4.806582278481013,
+      "grad_norm": 0.18678710119920797,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.4121,
+      "step": 1184
+    },
+    {
+      "epoch": 4.810632911392405,
+      "grad_norm": 0.19784703077808952,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.3998,
+      "step": 1185
+    },
+    {
+      "epoch": 4.814683544303797,
+      "grad_norm": 0.19590286701770654,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.4198,
+      "step": 1186
+    },
+    {
+      "epoch": 4.81873417721519,
+      "grad_norm": 0.21039256128858974,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.4225,
+      "step": 1187
+    },
+    {
+      "epoch": 4.822784810126582,
+      "grad_norm": 0.19639123037185469,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.4146,
+      "step": 1188
+    },
+    {
+      "epoch": 4.826835443037975,
+      "grad_norm": 0.20134686389534417,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.4111,
+      "step": 1189
+    },
+    {
+      "epoch": 4.830886075949367,
+      "grad_norm": 0.19643064749663203,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.409,
+      "step": 1190
+    },
+    {
+      "epoch": 4.834936708860759,
+      "grad_norm": 0.19019150007159574,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.4018,
+      "step": 1191
+    },
+    {
+      "epoch": 4.838987341772152,
+      "grad_norm": 0.19554607621213563,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.4036,
+      "step": 1192
+    },
+    {
+      "epoch": 4.843037974683544,
+      "grad_norm": 0.2025314340224753,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.4102,
+      "step": 1193
+    },
+    {
+      "epoch": 4.847088607594936,
+      "grad_norm": 0.1924195773776015,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.408,
+      "step": 1194
+    },
+    {
+      "epoch": 4.851139240506329,
+      "grad_norm": 0.19321977931224246,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.406,
+      "step": 1195
+    },
+    {
+      "epoch": 4.8551898734177215,
+      "grad_norm": 0.1935975000057012,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.3981,
+      "step": 1196
+    },
+    {
+      "epoch": 4.8592405063291135,
+      "grad_norm": 0.19912951851897948,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.3933,
+      "step": 1197
+    },
+    {
+      "epoch": 4.8632911392405065,
+      "grad_norm": 0.1910697932318618,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.4247,
+      "step": 1198
+    },
+    {
+      "epoch": 4.867341772151899,
+      "grad_norm": 0.20903917111655343,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.3964,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8713924050632915,
+      "grad_norm": 0.1956168149430498,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.4085,
+      "step": 1200
+    },
+    {
+      "epoch": 4.875443037974684,
+      "grad_norm": 0.20056352395985239,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.4087,
+      "step": 1201
+    },
+    {
+      "epoch": 4.879493670886076,
+      "grad_norm": 0.19359013776617792,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.4145,
+      "step": 1202
+    },
+    {
+      "epoch": 4.883544303797469,
+      "grad_norm": 0.1963318238376584,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.4019,
+      "step": 1203
+    },
+    {
+      "epoch": 4.887594936708861,
+      "grad_norm": 0.20250618720140826,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.3965,
+      "step": 1204
+    },
+    {
+      "epoch": 4.891645569620254,
+      "grad_norm": 0.1965590931767785,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.3908,
+      "step": 1205
+    },
+    {
+      "epoch": 4.895696202531646,
+      "grad_norm": 0.19970576600769369,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.4085,
+      "step": 1206
+    },
+    {
+      "epoch": 4.899746835443038,
+      "grad_norm": 1.092825416188952,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.4312,
+      "step": 1207
+    },
+    {
+      "epoch": 4.903797468354431,
+      "grad_norm": 0.19376106812055138,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.3973,
+      "step": 1208
+    },
+    {
+      "epoch": 4.907848101265823,
+      "grad_norm": 0.18920875413162927,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.4112,
+      "step": 1209
+    },
+    {
+      "epoch": 4.911898734177215,
+      "grad_norm": 0.19662436039025016,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.4194,
+      "step": 1210
+    },
+    {
+      "epoch": 4.915949367088608,
+      "grad_norm": 0.19645951110220322,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.3887,
+      "step": 1211
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.19727738998156405,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.4116,
+      "step": 1212
+    },
+    {
+      "epoch": 4.924050632911392,
+      "grad_norm": 0.19089159131580866,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.4201,
+      "step": 1213
+    },
+    {
+      "epoch": 4.928101265822785,
+      "grad_norm": 0.1822956248629886,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.4066,
+      "step": 1214
+    },
+    {
+      "epoch": 4.932151898734177,
+      "grad_norm": 0.19270793327685123,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.4116,
+      "step": 1215
+    },
+    {
+      "epoch": 4.93620253164557,
+      "grad_norm": 0.19687328239892615,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.4079,
+      "step": 1216
+    },
+    {
+      "epoch": 4.940253164556962,
+      "grad_norm": 0.19527011814408435,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.3969,
+      "step": 1217
+    },
+    {
+      "epoch": 4.944303797468354,
+      "grad_norm": 0.19336700871817514,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.3963,
+      "step": 1218
+    },
+    {
+      "epoch": 4.948354430379747,
+      "grad_norm": 0.19640490017571324,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.4053,
+      "step": 1219
+    },
+    {
+      "epoch": 4.952405063291139,
+      "grad_norm": 0.19836171357971175,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.4083,
+      "step": 1220
+    },
+    {
+      "epoch": 4.956455696202532,
+      "grad_norm": 0.19325805439099555,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.4175,
+      "step": 1221
+    },
+    {
+      "epoch": 4.960506329113924,
+      "grad_norm": 0.19477291004744382,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.4032,
+      "step": 1222
+    },
+    {
+      "epoch": 4.964556962025316,
+      "grad_norm": 0.1886607271551426,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.403,
+      "step": 1223
+    },
+    {
+      "epoch": 4.968607594936709,
+      "grad_norm": 0.1905650260583666,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.4032,
+      "step": 1224
+    },
+    {
+      "epoch": 4.972658227848101,
+      "grad_norm": 0.19674728155365995,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.4214,
+      "step": 1225
+    },
+    {
+      "epoch": 4.976708860759493,
+      "grad_norm": 0.19091742819070942,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.4065,
+      "step": 1226
+    },
+    {
+      "epoch": 4.980759493670886,
+      "grad_norm": 0.18837404134781832,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.3883,
+      "step": 1227
+    },
+    {
+      "epoch": 4.984810126582278,
+      "grad_norm": 0.18964811102574436,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.4137,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9888607594936705,
+      "grad_norm": 0.18970199994451659,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.4111,
+      "step": 1229
+    },
+    {
+      "epoch": 4.992911392405063,
+      "grad_norm": 0.19440843930659504,
+      "learning_rate": 0.0,
+      "loss": 0.3859,
+      "step": 1230
+    },
+    {
+      "epoch": 4.992911392405063,
+      "step": 1230,
+      "total_flos": 6.085998278900974e+18,
+      "train_loss": 0.17071973751715527,
+      "train_runtime": 22522.1157,
+      "train_samples_per_second": 7.015,
+      "train_steps_per_second": 0.055
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.085998278900974e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}