diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.982278481012658,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 9.933298721627814,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.8178,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 9.897819497107257,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.8435,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 9.959729308466024,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.7978,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 9.841084825490139,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.7812,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 9.224569384269643,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.8007,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 9.201906017819061,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.8101,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 7.261037855311741,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.7217,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 6.9268303591835485,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.7235,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 4.1786676019541185,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.6538,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.725397871410673,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.6585,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.6144087219821617,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.6638,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 6.727856164507345,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.6482,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 6.709680231141028,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.6345,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 6.819781922830209,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.6107,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 6.358479763558506,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.6071,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 5.129696946803027,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.5502,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 4.88237126921362,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 1.567,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 3.831215973439533,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 1.5153,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.9896153971872512,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 1.5199,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 2.2856836758173706,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 1.4982,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 2.3458750821434498,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 1.4826,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 2.4390259172121223,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 1.452,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 2.263225931909868,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 1.436,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 2.0592425231154814,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 1.42,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.612433437091964,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 1.4258,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.4538904013821738,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 1.4295,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.7208305683466307,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 1.4155,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.546216554836619,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 1.392,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.172588758346927,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 1.3881,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 1.5110114910255077,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 1.3917,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 1.2674040313167458,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 1.372,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 1.210674013224536,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 1.3559,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 1.2178120633526373,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 1.3626,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 1.0797338608965623,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 1.3555,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 1.1261892745516537,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 1.3515,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 1.0158869748246238,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 1.3441,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.9613705931197408,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 1.3358,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 1.0289433733318691,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 1.3351,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.9941746051672029,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 1.3198,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.9520419096796571,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 1.3204,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 1.237549022985381,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.3302,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.9305621317614939,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 1.3284,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.9080151523547578,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 1.3291,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.9626446313031342,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 1.3398,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 1.0935273041231537,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 1.3405,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.9595553373527506,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 1.3209,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 1.4349868663398164,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 1.3307,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 1.1727813243603573,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 1.326,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 1.2924896058583515,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 1.3221,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 1.6075317808400273,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 1.3228,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 1.0160806494151875,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 1.3249,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 1.3038622421600723,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 1.284,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 1.059360155414228,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 1.308,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.8729632808800111,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 1.288,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 1.1910013804161288,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 1.3086,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.9170151877278664,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 1.2988,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 1.102495241219338,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 1.2864,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 1.1787584441957,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 1.3032,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 1.1504936938152774,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 1.2869,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 1.3136328739498286,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 1.2869,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 1.162137996087394,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 1.3099,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 1.0187637612897926,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 1.3104,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 1.1773984723061925,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 1.2662,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 1.2100642936417008,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 1.2942,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 1.26480500230016,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 1.289,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 1.1992282536437207,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 1.298,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.1777622297620634,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 1.2952,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.4576673753007698,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 1.2892,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.21752818913171,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 1.2798,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 1.0563711356552656,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 1.2976,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.2519312317261713,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 1.2802,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 1.1674995101915469,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 1.2796,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.186326379256793,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 1.3004,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.3708127779316692,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 1.2795,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.3114551879278658,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 1.2733,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.3416993718752204,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 1.2991,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.4054919906309085,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 1.2723,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.2089496609738961,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 1.2644,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2753877462881964,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 1.287,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.354655118278104,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 1.2776,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.4407877823509578,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 1.3079,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.2838233171245736,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.305,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.2470806178705933,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 1.2791,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.217609595123036,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 1.3132,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.8382600917892271,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 1.2862,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.3215285337246088,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 1.2685,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.9846372812135762,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 1.3023,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.5283058641876313,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 1.3161,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.756734267285129,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 1.3043,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.5679608196674877,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 1.2516,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.3923559238342174,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 1.2755,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.4238896639021854,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 1.2774,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.356362634598468,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 1.2634,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.3675453411632543,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 1.2822,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.4194627065729042,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 1.2756,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.274969732938223,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 1.2357,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.813267955079747,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 1.2664,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.3676405906942297,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 1.2779,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.8158323053910375,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 1.2827,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 2.328954218686755,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 1.2792,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.9064613924281169,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 1.2739,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 2.596307309778557,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 1.2727,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.7984053796736303,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 1.2756,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 2.507185622845171,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 1.2923,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.9203261035500043,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 1.2831,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 2.3331822897940224,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 1.2717,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.8664773360493212,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 1.2626,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 2.292086406818919,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 1.2847,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.9186721185797075,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 1.2609,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 2.1743759127244577,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 1.2284,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.873004446473688,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 1.2544,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.705936933825921,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 1.2678,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.9874611748970357,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 1.265,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.4192685631574702,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 1.2549,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 2.0386379577422122,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 1.28,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.3850994392717229,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 1.2808,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.7581988959901718,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 1.2713,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.4756711967934037,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 1.2656,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.6594867350878697,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 1.2441,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.2931965504555782,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 1.2627,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.4421681062077512,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 1.2589,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.3215358301769276,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 1.2422,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.4033176184020035,
+      "learning_rate": 4e-05,
+      "loss": 1.2563,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.2131679749893058,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 1.2392,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.2803537920940653,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 1.2599,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.597970235336043,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 1.2392,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.3492040119330886,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 1.2654,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.076342084234448,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 1.2511,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.899813264157206,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 1.2675,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.373827372090609,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 1.255,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 2.719124532284048,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 1.2804,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 2.435344686205741,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 1.2751,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.4879370516416515,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 1.2627,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.34864506211106,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 1.238,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.7047755779787932,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 1.2883,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.2748363587545162,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 1.2785,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.360745828233388,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 1.2438,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.3488259518134602,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 1.2827,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.1972255305729358,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 1.2701,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.2102473888114986,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 1.2541,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.1663169480091446,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 1.2621,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.4806161983909714,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 1.2593,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.164207043995882,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 1.276,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.496751937640438,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 1.2905,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.2440381877527282,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 1.2701,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 1.4853429120691295,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 1.2764,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.23447056784449,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 1.2244,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.5012986890025253,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 1.2537,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.3900807200729093,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 1.2273,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.0319429270350107,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 1.2336,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.2890670709747514,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 1.26,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.9504177397537688,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 1.2896,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.2073125952710062,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 1.266,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 1.2108206782251456,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 1.2812,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.370421326496685,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 1.2793,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 1.0617061947505173,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 1.2224,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.9864746726362693,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 1.2725,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3867783865601402,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 1.271,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.242400174400085,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 1.2688,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.619620311598154,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 1.2511,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.4364484447137902,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 1.249,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 1.771055800949035,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 1.238,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.3906554826582653,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 1.2396,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 1.6066635144357058,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 1.2528,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.2070793666484902,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 1.2808,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 1.4549875949987627,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 1.2714,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.141041753325671,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 1.2203,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.9476671118042549,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 1.2807,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 1.3453793183108158,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 1.2677,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.9933107203125915,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 1.2344,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.348743937930853,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 1.2579,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.025320445251145,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 1.2678,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 1.5449402833336423,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 1.2381,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.1397546230689823,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 1.2707,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.9800174349192665,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 1.2749,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 1.4654625468202886,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 1.2479,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 1.5649948657250898,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 1.2452,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 1.3706715679622867,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 1.2522,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 1.0407236047805566,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 1.2632,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 1.9113746282620594,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 1.2509,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 1.4033149511035359,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 1.2609,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 1.807747226468799,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 1.2613,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 1.5268291322133318,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 1.2703,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 1.588719886394012,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 1.242,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 1.1958866572498055,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 1.2543,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 1.3773625479004505,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 1.2437,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 1.5170765768483432,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 1.2384,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 1.1407162392852277,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 1.2485,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 1.5386512048407477,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 1.2679,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 1.2251730690225693,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 1.2536,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 1.408108724704713,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 1.2397,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 1.2197980084638553,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 1.2447,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 1.2746670933719755,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 1.2293,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 1.3512867233034471,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 1.2557,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 1.1329461056672794,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 1.254,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 1.2668346122283785,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 1.2623,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 1.2443396585478637,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 1.2528,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.9150808190583143,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 1.222,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 1.4263164131533803,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 1.2688,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.8411367083420277,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 1.2579,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 1.5906717000463564,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 1.2429,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.9530431447291341,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 1.2553,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 1.8358571422342145,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 1.2358,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 1.4585528468201707,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 1.2582,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 1.4275653096713663,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 1.2301,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 1.3658094777637684,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 1.2337,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 1.280353091067545,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 1.2316,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 1.1405926514661329,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 1.226,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 1.1306669999517491,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 1.2446,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 1.1016566466283921,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 1.2381,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 1.2199977346747586,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 1.2289,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 1.091176952654132,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 1.224,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 1.2094906668588903,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 1.2343,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 1.1052019513312572,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 1.254,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 1.151212277526975,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 1.2299,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 1.0187444896650277,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 1.2198,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 1.2024717094843038,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 1.2441,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 1.015676239980724,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 1.2403,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 1.040506184227416,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 1.2335,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 1.0732980215644987,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 1.2351,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 1.3117768548399273,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 1.25,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 1.0586212065587202,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 1.2384,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 1.2378747791210107,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 1.2466,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.9384588852195727,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 1.2305,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 1.0404840108399178,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 1.2446,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.9971829751292264,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 1.2365,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 1.0817489648840979,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 1.2292,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.9593454419239398,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 1.2326,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.9640531283593995,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 1.2556,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.8540057120715031,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 1.253,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 1.0964404624563135,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 1.2341,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.7998900875727888,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 1.2362,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 1.1273197844834193,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 1.2618,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.677859781960143,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 1.2647,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 1.0674003272288481,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 1.2466,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.7962551984225934,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 1.2417,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7064423932237379,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 1.2369,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.8933875773364179,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 1.2328,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.8826540039267996,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 1.2429,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.6704718607742554,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 1.2478,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.793481933029087,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 1.2312,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.8067636039561972,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 1.2203,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.8264781583647357,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 1.2327,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6661350310643096,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 1.2318,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.7734851936871938,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 1.2221,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.7214510416742959,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 1.2407,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.7423834707176196,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 1.2245,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.8383601022919843,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 1.1686,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.7752393811446099,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 1.1787,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 1.0555626500487545,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 1.1553,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.8644750555468655,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 1.1663,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.7903490707195843,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 1.1381,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.9052862072140597,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 1.1349,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 1.0553629610990418,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 1.1584,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 1.0378925889386388,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 1.1607,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.8873461532847762,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 1.1783,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.9147625205012248,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 1.1833,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.8773054664826107,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 1.143,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.983855951569538,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 1.1842,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 1.0066064367608125,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 1.1526,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 1.030816302493887,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 1.1598,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.6185350800183042,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 1.1376,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.784175503442257,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 1.1713,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.7244663857327224,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 1.1349,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.7027298354232484,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 1.185,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.7696537298890939,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 1.1591,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.7237553367385098,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 1.1515,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.7911600084883476,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 1.1705,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.979499300176695,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 1.1479,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.9091071735595879,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 1.1391,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.7947885444196131,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 1.1301,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.762571569929015,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 1.136,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.6959469248034448,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 1.1352,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.6016922739401732,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 1.148,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.6845312656594692,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 1.1539,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.6192567802478047,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 1.1717,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.6244099682043345,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 1.1468,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.781393599265035,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 1.1421,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.8145652656258217,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 1.1401,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.752360146007145,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 1.1634,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.8436105745370066,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 1.1427,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.821123052171786,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 1.1388,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.8433498945167828,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 1.1552,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.8131224800162512,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 1.1682,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.8852924842055218,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 1.1368,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.7732441498199645,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 1.1705,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.6668429237987028,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 1.1425,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.732408955517127,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 1.1504,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.7528788012879877,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 1.1236,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.6365757616045409,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 1.1255,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.8669354635089168,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 1.1472,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.810762853691914,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 1.1671,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.8215539623348017,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 1.1496,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 1.001540768276704,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 1.1196,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.8607157287008111,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 1.1462,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.8283940748533274,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 1.1633,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.798549350225124,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 1.1463,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 1.1481258101752376,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 1.1495,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.883973923902975,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 1.1623,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.6879847118579288,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 1.136,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.8637709924391279,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 1.1475,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.6702391522578585,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 1.1578,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.756186949373457,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 1.1916,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.7223244530105629,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 1.1683,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.6231641676280993,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 1.1566,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.6955808201773662,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 1.1796,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.7012468832695,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 1.1681,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.63016610156459,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 1.1667,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.7214578159176559,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 1.165,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.627002991751359,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 1.1635,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.6697395212411436,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 1.1859,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.7000380529685968,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 1.1387,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.6473325779455186,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 1.1262,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.6456559775390991,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 1.1448,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.6217278712396029,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 1.1608,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6016368996868794,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 1.1826,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.6784845387964498,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 1.1616,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.649227468182824,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 1.1251,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.5937696910287711,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 1.1463,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.5829389009078099,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 1.1536,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.7121087361584678,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 1.1565,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.5848203978143939,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 1.1528,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.7990113872738007,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 1.1668,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.6615969360285675,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 1.1789,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.740429977462607,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 1.1588,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.7712169641916816,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 1.1601,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.6623743607871191,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 1.1436,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.7112983437937824,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 1.1526,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.6054599920574315,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 1.1514,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.5898683468456464,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 1.1174,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.592255716063486,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 1.1712,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.759368211257049,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 1.145,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.9020925843152267,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 1.1578,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.9938801759919645,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 1.1487,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.6822022172751128,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 1.1654,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.9960011796091998,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 1.1788,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.9141639339449058,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 1.167,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.6811586836150645,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 1.156,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.7817160648787812,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 1.169,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.6482912094523019,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 1.1469,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.7357875186731013,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 1.1628,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.6620856393585671,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 1.1778,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.7643591924645242,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 1.1422,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.6753964890285318,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 1.1869,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.6687875402040061,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 1.1595,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.6807292748902014,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 1.1661,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.6674172603846734,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 1.1652,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.6689268597921838,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 1.1659,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.6746394490078387,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 1.1515,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.7444518127676196,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 1.1447,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.6535355941505684,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 1.171,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.6101329784272493,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 1.1597,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.7482485283544967,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 1.1436,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.647677565696097,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 1.1655,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.7440958911067717,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 1.1629,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.7960136581017142,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 1.1367,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.8331818708033708,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 1.1491,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.7053542335276067,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 1.1682,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.7205034913380446,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 1.1618,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.8627257144488724,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 1.1577,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.551923279327445,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 1.1377,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.747886445257357,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 1.1538,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.8239461665521868,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 1.184,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.7705707398326068,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 1.143,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.8342625943001156,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 1.1698,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.7300540770021413,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 1.1617,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.8955477162323741,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 1.2018,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.5913027821699496,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 1.1564,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.7539256346580357,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 1.1811,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.8496773341867427,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 1.1533,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.5760506640810407,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 1.1699,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.5646870504067393,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 1.165,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.7525565458584117,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 1.1641,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.7090679376936387,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 1.1684,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.5799881797883643,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 1.1422,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.7007658258573316,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 1.1718,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.6839958246819896,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 1.1646,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.5825558954705621,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 1.1616,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.732261656996349,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 1.1434,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.760095262171451,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 1.1722,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.5210055430159378,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 1.1589,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.7475749196188591,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 1.1673,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.6594452212366566,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 1.1779,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.5935043631965298,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 1.1674,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.6992857478493116,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 1.1961,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6063318005172927,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 1.1789,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.6363131764750006,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 1.1618,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.5291270432875703,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 1.1477,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.5473029937255857,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 1.1528,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.5804619624960515,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 1.1654,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.5345193640376179,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 1.1668,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.5321669903077104,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 1.1626,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.5700661840884022,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 1.138,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.4950685734713498,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 1.1582,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5533645002535346,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 1.1546,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.5246454919619364,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 1.1556,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.6980326018027233,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 1.1646,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.7159884361474195,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 1.1333,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6856935533848656,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 1.1462,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.6150648957622421,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 1.1592,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.6976201592051443,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 1.1626,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.672083586558063,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 1.1803,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.6440889804825302,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 1.155,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.6128962245701063,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 1.1503,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.6878083984717698,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 1.1619,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.6018751128944752,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 1.1764,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.5943705225984935,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 1.1462,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.808146214942223,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 1.1732,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.6218398479232518,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 1.1505,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.6361800664297104,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 1.1577,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.5346081235658657,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 1.1494,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.7240825366635217,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 1.145,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.5233574689181881,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 1.1373,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.7205979206799599,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 1.1864,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.7731753649396087,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 1.1565,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.6006782603286414,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 1.1406,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.6744408915835239,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 1.1336,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.6705273123156373,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 1.1438,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.6134641473880076,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 1.193,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.6936392964253701,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 1.1424,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.5294858735781733,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 1.1525,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.6365635284280515,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 1.1514,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.619519354522933,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 1.1525,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.5916238717661698,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 1.1851,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.6651132141169918,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 1.1257,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.7696282910433754,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 1.1363,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.6742260594846335,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 1.1505,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.6764110275957895,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 1.1575,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.5711191881723127,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 1.1787,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.7450416218884675,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 1.1507,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.5312116055737036,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 1.1447,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.7062451564323639,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 1.1712,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.5006868724799162,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 1.1254,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.7479047420884823,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 1.1599,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.5282897190852283,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 1.1604,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.6369812747161735,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 1.149,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.6836631059765329,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 1.1753,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.5624956993507891,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 1.1592,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.7420241041135692,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 1.1664,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.6359882701967161,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 1.1645,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.6263848697402968,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 1.1379,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.6462332661644951,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 1.1563,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.6882735134203297,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 1.1428,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.5077001134861462,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 1.1433,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.5920336399781427,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 1.1651,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.621209099050027,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 1.1525,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.6013074242299851,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 1.1638,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.5816963099391691,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 1.1399,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.6288816879723969,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 1.1232,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.5444342787208661,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 1.1444,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.588101277466446,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 1.1695,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.6699403701747185,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 1.1409,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.659982479744482,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 1.1231,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.666191927222366,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 1.1802,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.7567479723085003,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 1.1685,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.6970138336517401,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 1.1439,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.5754842225552638,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 1.1517,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.6245371628955033,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 1.1626,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.6354123984954438,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 1.1273,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.7466824319740564,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 1.1603,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.5799160300815901,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 1.1554,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.6842509393087578,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 1.1367,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.6103866231555901,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 1.1601,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.6329168017790784,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 1.148,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.5999221044258835,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 1.1717,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.5724615094336478,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 1.1371,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.7583895254452035,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 1.1445,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.5973397429661461,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 1.1683,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.7630028389627302,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 1.1684,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6602007306957379,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 1.1219,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.6983417866368752,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 1.1854,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.7174141219318854,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 1.1487,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.6556679814504864,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 1.1607,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6482308437040363,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 1.174,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.6474383812411197,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 1.1505,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.5735527611864775,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 1.1482,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.6119866453799896,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 1.1327,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.6677082200262521,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 1.1693,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.6834439231658457,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 1.1565,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.5452361050297119,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 1.1684,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5856953069100633,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 1.1719,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.5304289185360662,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 1.1636,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.5671630949874984,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 1.152,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.5281688026683441,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 1.1645,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.6178649432350114,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 1.1273,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.48743991020474353,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 1.1654,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.5103761656925145,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 1.1507,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.5600543832567956,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 1.1385,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.4784001145700006,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 1.131,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.5907135370866342,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 1.1654,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.5067774231546158,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 1.1562,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.5385332084918243,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.1582,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.5354799323614053,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 1.1476,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.5801635181454197,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 1.1345,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.6895035224297531,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 1.0331,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 1.2092320313673883,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 1.0262,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.7835402868546538,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 1.0096,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.7848931048544997,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 1.0536,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.7003899744892192,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 1.0021,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.8693436740092586,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.9979,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.9280926834223986,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.9974,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.6884498997135877,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 1.0316,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.7228154705924649,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 1.0244,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.9893516794468438,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.9767,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.6143624021765863,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.9979,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.849996589920869,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 1.0144,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.661618423664458,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 1.0203,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.6679045817431669,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 1.01,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.6661145201550285,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 1.0222,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.6387559488801365,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 1.0156,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6235425032168542,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 1.0281,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.6838910243587044,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 1.0026,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.5397016077246916,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 1.0188,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.696784129644162,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 1.0004,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.7006685341361082,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 1.0083,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.5936753928696896,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 1.0294,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.7012854125074445,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 1.0092,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.6033585773045091,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 1.0197,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.6766770211171466,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 1.0186,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.6648314771848939,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 1.0343,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.6719551472239573,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 1.0088,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.6003514076267086,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 1.0144,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.6670270325307559,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 1.0294,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.6078921085380056,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.992,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.693062430169994,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 1.0206,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.6481387873992199,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 1.0401,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.5641535422210479,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 1.0356,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.5311822158331811,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 1.0133,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.6313724261130622,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 1.0211,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.5755198026265722,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 1.0138,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.6343596980550483,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.9967,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.6677506920589293,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 1.0225,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.5670298710242881,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 1.0073,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.6701666402693496,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 1.0165,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.5840554230350358,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 1.0297,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.5462457155921613,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 1.0158,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.5948879552367591,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 1.0102,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.5709513962588751,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 1.0146,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.7105521352683684,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 1.005,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.5288298496430593,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 1.0039,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.6066533775739462,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 1.0463,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.4870070750405779,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 1.0132,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.5738040560093357,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 1.0178,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.5197347689950114,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.9993,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.5038843717132108,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 1.0041,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.66549808145693,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 1.0315,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.4871339346103557,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.9958,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.6139299428669691,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 1.0172,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.5637856188863154,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 1.0017,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.5383416212276905,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 1.0149,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.5595330744870007,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 1.0218,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.589950590347505,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 1.0334,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5715264406692745,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 1.0204,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.7535338185706086,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 1.0272,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.5379369784813346,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.9897,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.7116159618042617,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 1.0193,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.4679924378502827,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 1.0257,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.5637566920004994,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 1.0109,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.4485705339309698,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 1.0348,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.5334499230558688,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 1.0228,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5260005703080777,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 1.0199,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.520172274194265,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 1.0082,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5094704901558924,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 1.0314,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.5936032272916024,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 1.0369,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5108988909558086,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 1.0191,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.6587729506277713,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 1.0309,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5440398970206946,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 1.024,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.5578029007796318,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 1.0219,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.6903894779635844,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 1.0097,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.5415806936625543,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 1.0181,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.5242665184382573,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 1.0231,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.5590023235404082,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 1.021,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.5392259827932185,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 1.0194,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.5975528088229662,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 1.0297,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5079216009835216,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 1.0346,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.5219122239463513,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 1.0632,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.5139185306280853,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 1.0164,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.49193173960610537,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 1.0296,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.46691870069933034,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 1.0319,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.5523887632253133,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.9987,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.516921810911513,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 1.0275,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5900062953091885,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 1.0417,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.5612954061813461,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 1.0299,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.5381203488988708,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 1.0352,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.5333543476695336,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 1.0333,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.5273404908783355,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 1.0294,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.5245383463434636,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 1.0293,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.554681220598988,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 1.0303,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.5036996184253952,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 1.0401,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5229637027529087,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 1.0207,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.491384332679668,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 1.0406,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.5255224163281365,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 1.0229,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.5617986927899088,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 1.0617,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.4920215239649119,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 1.046,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.49492193986034355,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 1.014,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.5508439469870827,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 1.0333,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.44875249990939386,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 1.0272,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.5517701170144657,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 1.0136,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.4756375671739574,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 1.0344,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.6074358494848272,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 1.0136,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.45848388393615697,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.9962,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.5578498095919927,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 1.0033,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.447137712656289,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 1.0328,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.49673531889142347,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 1.0188,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.48139903004180523,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 1.0293,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.46525581296497814,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 1.0048,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5324924178702928,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 1.0101,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.49832008960752955,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 1.0176,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.47135857514656343,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 1.0128,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.48960084845830415,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 1.0605,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.4981175924190684,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 1.0373,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.6048582727269858,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 1.044,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.5078286734838149,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 1.0313,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5191819139070797,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 1.0536,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.5507895321263605,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 1.0121,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.5035978041036229,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 1.0136,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.5679857665814879,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 1.0064,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5503633480644989,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 1.0353,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.5375922050503924,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 1.0052,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.7486033469161537,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 1.0206,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.5257536054943663,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 1.0254,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5730511276583133,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 1.0299,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.5491278780233214,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 1.0185,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5462466952331015,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 1.0012,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.5429754616549164,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 1.0038,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.6026974029153855,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 1.0182,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.5313924945353171,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 1.0375,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.5429977017183559,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 1.0266,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.5524829264994087,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 1.0266,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.5264496946029424,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 1.0141,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.5143315844952089,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 1.0154,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.49100271539581425,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 1.0156,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.5371490440713882,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.9955,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.511997600605205,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 1.0402,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5618922745501259,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 1.0064,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.5574722139404477,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 1.025,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.5882415460148407,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 1.0159,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.49081413260467366,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 1.0296,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.5838467723866013,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 1.0728,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.6170912260429801,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 1.013,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.47248179185113337,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 1.0317,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.5681806819560992,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 1.0247,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.463708228533855,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 1.018,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.5325929881207099,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 1.0404,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.490401810929384,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 1.0289,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.4573573157201373,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 1.0195,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.4861001409780556,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 1.0064,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.4728049689992945,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 1.0197,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.4670732151790282,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 1.0012,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.44750593282068835,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 1.0091,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.540672614135796,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 1.0211,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.47553678903831514,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 1.0314,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.4687879062052804,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 1.015,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.4761124488619028,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 1.0146,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.4877113503656656,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 1.034,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.4630754894238843,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 1.0341,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.45848156601819184,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 1.0341,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.4966310123748713,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 1.0003,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.4554447671259762,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 1.0126,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.46875537996124655,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 1.0067,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.4976944808717196,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 1.0096,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.5446925688077884,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 1.0207,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5008286427817888,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 1.0227,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.5332582571428496,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 1.037,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.5416691768983506,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 1.0335,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.5041289288935831,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 1.0402,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.5307269797916867,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 1.0244,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.5026127996377834,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.9953,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.4553480580546981,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 1.0152,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.5097429721013352,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 1.0411,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.49531667952964054,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 1.0445,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.42789543896033205,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 1.0534,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.47993803280822295,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 1.0255,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.5429741328275045,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 1.0378,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.45009794669001796,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 1.0016,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.4671981915242967,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 1.0002,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.5252792905726302,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 1.0305,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.47379738750543554,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 1.0186,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.49891873613816495,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 1.0256,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.4907339228346661,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 1.0284,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.4491712581937517,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 1.0277,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.5146975000104856,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 1.0364,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.5481977786317666,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.9977,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.4144066106189655,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 1.0169,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.5897906663111407,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 1.0601,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.45683625437849484,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 1.0058,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.5445950406714416,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 1.0154,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.5119743541531611,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 1.0419,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.5246144317621387,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 1.0163,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.43377413208229887,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 1.0062,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.5331754347259767,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 1.0135,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.47589680989101923,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 1.0201,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.5517993141730592,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 1.0372,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.47765324629891404,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 1.0269,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.5416044897936797,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 1.016,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.46436913098439336,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 1.0334,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.49974978237854684,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 1.0144,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.45433943286804207,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 1.0318,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.4695881473857097,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 1.0252,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.42926994962072956,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 1.016,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.4643111773017718,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 1.0189,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.4518100051036514,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 1.0263,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.4437576751881414,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 1.0072,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.45800741999273964,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 1.0207,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.48620023401260565,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 1.0012,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.4582693383521256,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 1.0301,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.46997718822847556,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 1.0104,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.46927038684222133,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 1.0108,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.5041285787918498,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 1.0078,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.5399136546666372,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 1.0255,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.6383710075078183,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 1.0281,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.4923022559229843,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 1.0547,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.5470922862641557,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 1.0419,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.5745866326777408,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 1.0017,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.43578157644968657,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 1.0331,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.5724417179682006,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 1.0214,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.571811218243599,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 1.028,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.5456423227749969,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 1.0152,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.48897761642746446,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 1.0215,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.5419756217224223,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.9806,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.4847201915368628,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 1.0118,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.4847685304991527,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 1.0264,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.4756336198559765,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 1.0292,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.4605018370615728,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 1.0492,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.4995563884295711,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 1.0587,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.4629606397352756,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 1.0002,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.5121702323100864,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 1.0168,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.44509760128001585,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 1.0075,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.5066683404498322,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 1.046,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.4617198437054005,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 1.0159,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.4803137142559452,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 1.0125,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.437251076824209,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 1.0451,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.47531563978040436,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 1.0166,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.46586860207167996,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 1.0102,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.4527089648702027,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 1.0226,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.4437355051161407,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 1.0135,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.42633032133116083,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 1.0329,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.4054174841533925,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 1.0298,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.42404852229400325,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 1.0101,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.4846216035603012,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 1.0124,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.5100472648721559,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.9767,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.7940850340756402,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.8815,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.8472581636317164,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.8533,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 1.1196993014960313,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.8488,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 1.0153465184443184,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.8605,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.826462503480141,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.8859,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.6484184668157622,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.8576,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.725425202944048,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.8382,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.7616537997611557,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.8637,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.6462896283778906,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.8703,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.7220316010693504,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.8489,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.7114444699527576,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.8826,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.617933946293048,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.8443,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.6943428154692264,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.8747,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.6792956559209735,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.8328,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.5850260060911979,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.8476,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.6301254783379673,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.8517,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.6694008952730683,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.8703,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.5525784085342285,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.8381,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.5680610355560322,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.8354,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.5485095756448677,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.8611,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.5618432087341189,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.858,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.5486784099241084,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.846,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.6225896834499991,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.8513,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.5392828817736721,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.8523,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.6294064087077287,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.8513,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.6724661332113148,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.8731,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.5114666067189799,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.8471,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.6640261957507612,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.8505,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.6375009643138477,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.8713,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.5665719793299382,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.8645,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.6508042835076818,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.8615,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.523022008387731,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.8584,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.6046989386315796,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.8676,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.686740217615702,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.839,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.5166187782569124,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.8558,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.5577086776852163,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.8628,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.5877570910991192,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.8683,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.5285863222116387,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.841,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.49682731118680074,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.8238,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.547394026090636,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.8604,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.47926112406609717,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.8733,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.5199765954548031,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.8508,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.5373819617556451,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.8486,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.47779899464747505,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.87,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.5107792600372495,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.8462,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.6018675418247474,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.8396,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.48484303516460425,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.8454,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.5322482082345672,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.8616,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.5537949900707131,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.8483,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.51453721945,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.8685,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.5012382718314914,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.8544,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.5058488987870322,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.8697,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.49506017951103093,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.8515,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.4921062485871461,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.8704,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.4715942980907537,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.8624,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.5167271281630322,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.8746,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.4715008273775736,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.8768,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.4718589866200359,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.8407,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.4948416711430985,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.8523,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.4988185986406195,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.848,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.5168591169036743,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.8852,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.5108886392907334,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.8551,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.4826735949338651,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.8744,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.5475371964312847,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.8659,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.5402728331837098,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.8626,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.5374773441841417,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.8552,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.49334675206037387,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.8621,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.5811335818898145,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.8678,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.5221978329717215,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.8713,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.5545509486631972,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.8556,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.4851448552925207,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.8429,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.5203222948539638,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.877,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.5011749672391853,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.8444,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.49002170003142786,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.8722,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.45090799234723117,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.8629,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.5154154906832641,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.8466,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.5042249940801934,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.8543,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.47777314558949063,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.8413,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.5129928648171661,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.8662,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.4747399922839749,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.8659,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.5178435161207365,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.8407,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.441372049072934,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.8597,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.48696560180458887,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.8896,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.4793400452479297,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.8461,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.4498068326285848,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.8594,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.469888209566026,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.8536,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.47705766739937333,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.8691,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.45605148536097295,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.89,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.5470579389762384,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.8559,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.49464733549203643,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.8604,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.4682054358413757,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.8544,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.49773485385052296,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.8437,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.4638899094688478,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.8655,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.4892214322439687,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.8628,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.4720366516869082,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.8466,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.5050874497283874,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.8586,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.46039202617554503,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.8502,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.4741288489695432,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.8485,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.4544383608758889,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.863,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.44421783829640366,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.8601,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.474671674144896,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.869,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.4726014331049998,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.8612,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.42591144693962607,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.8615,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.49038026410144775,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.8623,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.4349745417581744,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.8327,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.4605217677295788,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.8629,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.4688154246012126,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.8612,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.442719788770277,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.8525,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.4353956061570409,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.8394,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.4666698629651009,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.8615,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.4896865629985745,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.865,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.45322888201532946,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.864,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.45956447961461894,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.8508,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.47279988072578844,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.8712,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.5077997030799645,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.8674,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.46873177403947935,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.8663,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.5353437167850071,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.8519,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.4543633512495401,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.8658,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.49264460762343015,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.8513,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.4951063230915622,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.8667,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.44099396444065886,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.8303,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.4650406046530908,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.8558,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.49381916430436745,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.8758,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.45152550937213043,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.8633,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.4704217587868533,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.8623,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.48724891824680505,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.8612,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.5182634549927265,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.8616,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.4841175626016344,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.8682,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.44398176561843006,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.835,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.46032212217150315,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.8485,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.4853228588172091,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.8801,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.4667046031555135,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.8458,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.43356477955034306,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.8456,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.45852910904743077,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.8685,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.47269673230146425,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.8544,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.48028146837920654,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.8636,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.43948553608546737,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.881,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.441711083164455,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.8553,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.49809664632255973,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.8705,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.45473410569513895,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.8564,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.4503261802650196,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.8491,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.4855618703994967,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.8547,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.4293433445469158,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.8683,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.45460836785133596,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.8766,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.4472154302647609,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.8533,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.42658969102079963,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.8683,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.43476724912119574,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.8606,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.44895103957587856,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.8442,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.45367721227215985,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.8492,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.4245925418018429,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.8423,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.4487673139347265,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.862,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.45047147576212493,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.8492,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.4629569305659359,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.8708,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.4347177828032504,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.8612,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.44401055368542536,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.8444,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.4341929762403499,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.8687,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.4604145885247143,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.8822,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.4510126094907305,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.8725,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.43841890839465214,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.8668,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.43377494756106416,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.8536,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.4544241485280512,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.8682,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.4300511245757169,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.8672,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.45722018896945604,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.8613,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.4626337467017989,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.8757,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.4245502855792999,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.8506,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.44267397354146654,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.8754,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.43255853672787625,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.8582,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.4253317983644499,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.8765,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.4302880799649695,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.8344,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.4339420553039311,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.8611,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.43131845868563506,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.8578,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.41622739962031957,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.8538,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.4356736702915069,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.8751,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.44266438492323185,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.8465,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.4199543977807518,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.8475,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.42898294742075693,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.8579,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.45372597212590554,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.877,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.4782097675487259,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.8605,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.42667519173035895,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.845,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.4412596572308266,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.8782,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.43721685167453056,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.8581,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.4358381565410245,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.8539,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.46709876598481115,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.8577,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.4318497674876077,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.8579,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.45093121162418454,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.8559,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.43365208597287247,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.8722,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.4222340533854372,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.8546,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.4306142019218229,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.8507,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.434284943368926,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.8411,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.42871593278171216,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.8332,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.44833722344271615,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.8531,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.4306046713177489,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.8814,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.44848781499988855,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.8504,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.4243191033573401,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.8696,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.4199135048918373,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.8612,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.44378401886314917,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.8341,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.4321715154290951,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.8552,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.42871133064396516,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.8454,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.4689330655595317,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.859,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.4355530919401553,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.8592,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.4504301897111289,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.8643,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.42189873529982413,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.8608,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.4190743313607874,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.8394,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.4279058445407614,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.8795,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.4302420482429957,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.8399,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.4270955513795819,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.8529,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.4121961072304005,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.8742,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.43858761693119175,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.8696,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.4405054286279495,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.8412,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.40862724538485984,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.8525,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.4168952589362802,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.8478,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.44683118210279316,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.8535,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.45633524728567043,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.8519,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.4137304246484163,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.8294,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.40517432394566916,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.8496,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.4551332544227311,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.8448,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.4426683058120676,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.848,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.42814054952957725,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.853,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.41995701596474777,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.8654,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.4328372730710024,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.8401,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.44836473702417,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.8789,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.4299508183685661,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.8619,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.45508823108559676,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.8745,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.43349126432492535,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.8634,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.43620502823452584,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.8627,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.4628623135193572,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.8466,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.4272519685837642,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.8697,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.4116999538454061,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.8436,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.43896133333935833,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.8419,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.4476967228913993,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.8416,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.4436884835655507,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.8708,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.41944909267345304,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.8455,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.40816168274947523,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.8562,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.41313944786583956,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.8581,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.42754059642545406,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.8379,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.40913005879554865,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.8534,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.42831682529661047,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.8485,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.44262886273491836,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.8627,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.4181257483904811,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.8524,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.4143983358723759,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.8497,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.41457448251945356,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.8479,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.435093984728811,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.8334,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.4377403617922161,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.8482,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.4267208530940564,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.8599,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.41038153079985584,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.8468,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.4112766723167426,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.8539,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.6611584306513932,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.7993,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.9583794298794673,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.7376,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.7321579122000444,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.7253,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.6579865184281549,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.729,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 1.242467236037362,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.7224,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 1.374220968163499,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.7133,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.9083786058044742,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.687,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.7665226288021144,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.7111,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.856742534984717,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.7122,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.8446093533763239,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.7157,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.7094851506069519,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.7046,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.6268606686154433,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.7106,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.6228247669725995,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.7191,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.7015812584466826,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.6865,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.7440031416535453,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.7307,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.7114773976875085,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.7175,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.5646695877084114,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.6985,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.5839802640498584,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.7109,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.5858288553551011,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.7088,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.6339025783641914,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.7034,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.6140790256864238,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.7163,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.5302804531972476,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.6978,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.5509354719553615,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.7225,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.568709054740914,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.717,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.5678990116502838,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.7111,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.5766307451681424,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.7017,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.5198608496598639,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.6903,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.5237210704007911,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.7144,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.5359247596022375,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.6991,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.5670110358718077,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.7179,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.5106317976449242,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.6753,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.5401787224612814,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.7288,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.5245119467123522,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.7068,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.5142973450050567,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.7094,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.5131404690899716,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.7076,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.531942802173234,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.7448,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.5279130126767428,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.7356,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.5011034322426758,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.6951,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.5026886527334059,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.7134,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.4913151025852923,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.7357,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.5026351317757181,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.7074,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.48611841031088965,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.7066,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.4937676282167012,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.7286,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.5036775199331502,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.7049,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.49405263756675777,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.707,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.5006939255026086,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.7147,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.4955527623578856,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.737,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.497062742431914,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.7176,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.4779180034777856,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.6945,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.47533393312681765,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.7194,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.4974020993572035,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.7425,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.48921927697542045,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.7071,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.49335771658976024,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.7048,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.4895980642763978,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.7248,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.48651937740697204,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.7115,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.49298515978144175,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.729,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.5032463421888747,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.6966,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.48271388422857114,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.7047,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.4806323264074567,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.7367,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.46960717467375185,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.7066,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.48355217119956184,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.7045,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.4716432708881365,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.6977,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.4725574809379581,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.7112,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.4830757886917016,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.7068,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.47597731087467043,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.7022,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.49040803044903375,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.6878,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.5076461544121226,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.7212,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.46926411286721886,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.6904,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.4733786178392552,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.7361,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.48162073110167114,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.7054,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.47645181853574586,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.7001,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.4789362085427117,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.7117,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.4739338115898973,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.702,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.4728264985543456,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.7066,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.47542459012982796,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.6996,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.4599096799650364,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.6847,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.47548168392070383,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.707,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.48435088860204367,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.7232,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.49442286877589997,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.6955,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.48550904933855227,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.7266,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.4812781698591079,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.7101,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.46675464468913747,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.72,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.48611655178398033,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.7009,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.48927445323678875,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.7024,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.4735972033414075,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.7059,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.477159951185616,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.7227,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.45951262445016283,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.6971,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.4748680227045038,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.6988,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.45746116176155566,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.6899,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.4645666415922422,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.7002,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.4652083972993901,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.7177,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.4770108084680748,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.6891,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.48176768318279933,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.7006,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.47212763192411145,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.7096,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.4689129888733726,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.6986,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.4840616913988068,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.7237,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.47527393630851306,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.7074,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.47714534374109047,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.6895,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.48925435923932126,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.7065,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.45948962182272235,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.691,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.4660688299187154,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.6935,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.47675794206183325,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.7134,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.47828233072302745,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.7009,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.46162951423460646,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.7021,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.48133978900255964,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.693,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.49142779488745536,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.7162,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.47246042591365944,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.7105,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.4721199576885026,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.7223,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.4722969873436003,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.7193,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.46643506045552124,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.7021,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.4696312108335307,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.6973,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.46089568195877834,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.7087,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.4678116872473261,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.7229,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.45860861181236184,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.7194,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.4686949121074428,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.7132,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.47235307394573145,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.6999,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.4658385082734779,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.7309,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.4698441371042859,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.7091,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.4723970253406769,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.6987,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.472892648043723,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.7087,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.4650378124491558,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.7061,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.46548755841832035,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.6844,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.46968843649622716,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.7102,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.4618826252665963,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.6856,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.47673085477356547,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.7081,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.46503650852728906,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.718,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.4782096702364812,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.7222,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.4751131078647094,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.7348,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.4700023534908212,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.7158,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.4582613269690092,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.6901,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.4795307530459541,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.7035,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.46917252196577597,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.7111,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.4768017142268406,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.7123,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.4602520702460931,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.6991,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.45491265014353127,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.7077,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.46805995251792165,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.6942,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.47102629639855165,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.712,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.4606921863716593,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.7005,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.47159463083269837,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.7125,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.46653697035758235,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.6809,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.45984003008181273,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.698,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.47508456329565657,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.7191,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.46884346133679355,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.7022,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.4741204135737283,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.6959,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.46159415365960194,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.7011,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.4647851577622568,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.7086,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.46909601422078095,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.7196,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.47378320439885285,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.7233,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.4670281449888087,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.6916,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.468118557929843,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.6979,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.4557014633707797,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.7017,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.47171535529063435,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.7083,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.4710862827629821,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.7336,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.47310489980271214,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.7248,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.462440180107535,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.71,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.47650528083724153,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.7152,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.4663516529158096,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.7208,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.46923847987254025,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.7118,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.4656918173865396,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.6998,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.4542669393840388,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.7196,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.465328690030621,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.7186,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.4596199034954166,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.709,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.46205374841501895,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.7079,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.4597931242942473,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.7037,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.47431235279394923,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.7268,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.4622505740616268,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.6881,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.46061954957770435,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.7126,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.4630447338638172,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.6973,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.4588114501106474,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.7021,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.46019262827718294,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.7121,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.46305099118640536,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.7073,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.4621325349031714,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.71,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.4658496649342986,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.6886,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.46216732772858105,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.6892,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.4554300388636575,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.695,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.46674816950612363,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.7045,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.4632920066521023,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.6981,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.5784518087058715,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.7119,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.46060896568713544,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.7026,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.4634375075621392,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.6847,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.46709061342816877,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.686,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.46592879749631144,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.7241,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.46392860128183516,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.7128,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.46759426172925156,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.7049,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.46287866388708293,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.7065,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.46086104488780216,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.7068,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.4677580888780486,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.6714,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.4773580051540583,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.7073,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.4671885101235506,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.7346,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.4834070358252625,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.7071,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.4660824574416042,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.6831,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.4799221334363283,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.7131,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.4629456904000208,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.6919,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.4680967917390402,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.7074,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.46200443816457737,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.7238,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.461111139652867,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.7066,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.4606576337140547,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.7028,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.4571636066044727,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.7149,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.4567010754856929,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.7149,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.4622601385178996,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.7175,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.4587546617654815,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.7187,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.4667702556420012,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.6966,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.4599501619745314,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.7119,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.46612847518901457,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.7172,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.4660602617446201,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.7223,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.46509600894492625,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.7254,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.4590671348062732,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.7212,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.4664343263030432,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.7155,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.4592604405574057,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.7137,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.4576097655122136,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.703,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.4659712553498694,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.6932,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.4627341817635609,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.7063,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.4672227961895586,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.7171,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.46056636100330667,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.7139,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.4594705702046612,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.7122,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.46344855757470726,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.7212,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.4582634148107596,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.7025,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.464827815867559,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.718,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.45907820109757347,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.7111,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.46204180024544145,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.6927,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.4590570358367952,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.703,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.4649979453970776,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.6942,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.4585177500740186,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.6991,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.45988955561723205,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.6931,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.46160509375120545,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.6957,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.45770905649703353,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.689,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.4615418488797541,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.7012,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.46173248429569286,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.7246,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.4619400354151063,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.7158,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.46386561139423577,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.7154,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.4611610621816007,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.7137,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.45992161464926373,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.7095,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.45308759591341885,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.7081,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.6123635344569365,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.7113,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.47072404686160174,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.6879,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.46638974962549407,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.7078,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.46140620537082006,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.7028,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.46919808962795967,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.6982,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.46170609712707994,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.7185,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.45836213476764615,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.6838,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.46897800618641133,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.7145,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.455994247804533,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.6984,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.4603004127990118,
+      "learning_rate": 0.0,
+      "loss": 0.7094,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 2.1316152429779026e+18,
+      "train_loss": 1.01131284905643,
+      "train_runtime": 15678.0836,
+      "train_samples_per_second": 10.078,
+      "train_steps_per_second": 0.078
+    }
+  ],
+  "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": 2.1316152429779026e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}