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": 4.994430379746835,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 6.22599875273059,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.0623,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.5598847813605055,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.105,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.565084258211005,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.1151,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.254271432188064,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.0412,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.3468346367943145,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.1342,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.864084995254926,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.0801,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.809462061071857,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.0554,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.6766943244353865,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.0306,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.3354942855489624,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.9443,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.168903852551838,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9559,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.1080851042595268,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.9516,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.283826366556956,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9467,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.653432217611224,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.9642,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.518358275032542,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.9679,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 4.137953116514839,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9267,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 3.0147823048265945,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.9267,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.734306641265883,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.8658,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.558078164227987,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.8929,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.3472917745901514,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8718,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.9017679816152662,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.8466,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.756012290599556,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8568,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.8094941011766317,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.8505,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.9490264505755415,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8301,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.696890803725592,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.8238,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.5007719772598427,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.8383,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.25433458953558,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.8119,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.223077315754893,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.8132,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.1169395401875812,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8081,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.1769596259305428,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8142,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.9578494793270255,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.8066,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.911412266825585,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7898,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.945628220963479,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.786,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8402978691520325,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.7979,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8475444622228041,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.7699,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.7676721255360393,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.7627,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.7637887196644552,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.7533,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.7530196656445277,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.8085,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6771973599762365,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.7754,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6647559741141915,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.7782,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7469309879680991,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.7509,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.721742979892065,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7524,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.6163485306708881,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.7597,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.6990818815479554,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.7609,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.7188678443444191,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.7507,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7415785341966821,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.7411,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6824970703356978,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.7635,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.8337493924916543,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.7611,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.8221708902286977,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.7754,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7892230812773534,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.7351,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7417020322035166,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7527,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.8175151500535462,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7435,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7272026580117908,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7512,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.6543393788203817,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.7356,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6845178037416481,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7374,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.640063993178633,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.7343,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.5802205999967748,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.7387,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.74904243284132,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.7326,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.704465283904925,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.7663,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7316728229740714,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7308,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.7571556146760426,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.719,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.6620370639779196,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.7412,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.68503070057442,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7152,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.732576310670641,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.7353,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.6791417866025076,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.7011,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7074983476711499,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.713,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.7188397729746981,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.707,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.6596647217336008,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.7265,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.8434329312329251,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.7071,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.7514092489532948,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.7151,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8337538567583284,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.7276,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.843125697408138,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7254,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8707158471507829,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.7285,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.7189774932269192,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.7343,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.6783726025956076,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7257,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.9174676917059343,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.735,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.741649384557656,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.7242,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.6935448689542012,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7051,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.6933641982996406,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.7358,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6513070096458032,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.7019,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.8394540754973744,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7418,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.709350676417348,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.7227,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.7195782093893198,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7159,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.8872664472201129,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7073,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.7266170274439567,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7175,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.304035014905291,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.7699,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.6600186853423643,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7099,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.2097513895733545,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7264,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.7544605583261244,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.7353,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.0501950137950595,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.7279,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.9094076974008092,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.7122,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.8192492031894368,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7593,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.710087199657759,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.7064,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.8211883410220698,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.7177,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.8239728559689253,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7125,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.6936927736613792,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6979,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.8928604106840103,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.727,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.8034535726312871,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7233,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.779568090920758,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.7371,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.8299959291694796,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.7018,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.8889199839238504,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.722,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.7563964741489447,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.7206,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.2957795383255253,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.7007,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.8656173661244458,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.677,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.9636961906753858,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.694,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.1735725708912008,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7375,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.6824814042500119,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.6888,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.7678678176569966,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.6995,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.8587888126804407,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.6849,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.9371856880001034,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.7122,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.655941719487077,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.7069,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.7985593606983035,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7055,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.8991175408979234,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.735,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.072704271651066,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7272,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.8225436777901519,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6852,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.6850145595311731,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.7012,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.9871766507214141,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.7569,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.0904989728539114,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.751,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.739823034218386,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.7452,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.7741191242523082,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7125,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.6952880479758923,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.6744,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.656173463639302,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7023,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.7830346554468607,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6786,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.7177472311284697,
+      "learning_rate": 4e-05,
+      "loss": 0.7222,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.8487296972318362,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7398,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.8061055093139083,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7286,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.7736926080344221,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7018,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.6451872460825371,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7223,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.834072778817587,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7147,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.7757378153931814,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7265,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.776641918228933,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.7167,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.7629110358723558,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.7204,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.7124148499016172,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7423,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.6873336175804019,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6981,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.6866428158587848,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.6769,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.6827888908963669,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7278,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.7936585749439673,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.6986,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.6950984912109116,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.6993,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.6334236013967993,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.7041,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.8007211492541381,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.6983,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.699491416749459,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7133,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.6938295684590193,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.7251,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.8397377253183963,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7003,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.6187055292192599,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.6975,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.7725655975312627,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.7078,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.6231977748888494,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7032,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.9124254036042626,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.7121,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.5862968763361801,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7089,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.8680545055033086,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7063,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.7169699814502852,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.7053,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.5850750681493677,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6911,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.7269282710097525,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7187,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8407960333219302,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.717,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.8691870384661381,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.6846,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.554446042183467,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6843,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.7578092755132997,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.7037,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.8543908453495949,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.719,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.6096943370801213,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.7267,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.70017294987985,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.7193,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.5696214615742113,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.6967,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.6206548443212551,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.7089,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.5104970681308961,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.6962,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.609606106682744,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.7239,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.5414259368779898,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.6839,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.7222434316186463,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6957,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.5683046081097822,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.696,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.6038097815453862,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.6837,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.5914753693443165,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.7109,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.6695689436461733,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.7221,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.6693922273227682,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6975,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.6070101591337917,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.7152,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.7181837996110827,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6882,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.6976662303030912,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6961,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.7611103641419915,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.6925,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.0174230365604118,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.7137,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.7071973772161825,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.7167,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7733231033557196,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.6979,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.754926021907597,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7019,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.6251438492342895,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.7069,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.7617231295535153,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.687,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7412772668963452,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.7187,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.7094196649121698,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.7048,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.6733488274516681,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.672,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.7994169356026142,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.6891,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.6733228423514274,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6835,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6750238383740669,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6888,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.8290957696309738,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.709,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.7296982080819082,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.7083,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.6869862060931498,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.6885,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.6215773067098205,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.6791,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.8042320872846576,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6729,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.5411501085789768,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.6814,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.6631142181539554,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.6949,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.5502859856180827,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.6765,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.5567385937683927,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6957,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.652990536614473,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.704,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.6373064007777782,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6785,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.689588037540927,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.6851,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.7002683686759191,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.6905,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.5612723353952912,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.7066,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.6044206344965569,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.6875,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.5812875856179368,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.6905,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.5200252039782841,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.6889,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.5402379586730123,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6793,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.49815187439050596,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.694,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.6355195046913974,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6803,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.5556972815709361,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6783,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.6491582100445652,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6877,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.5797409371791026,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.7023,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.5777670931624581,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6988,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.6555701844733911,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7001,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.5942610468313054,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.6965,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.6003172988424741,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6811,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.7355822108195439,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.6878,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.527689658182083,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.7051,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.6999882331616134,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.6995,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.7073071353903464,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.703,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.6433363555553467,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.67,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.7342063925562021,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.693,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.5566250912640663,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.6976,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.7146815590179182,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.6822,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.5614261487541229,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.6777,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.5947323480095654,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.6986,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.5235870252425812,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6772,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.5280724328909607,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.6915,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6199396189047617,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.6815,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.5130038199821404,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.6847,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.6604928827304238,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.6933,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.4844668661650776,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.6609,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.693882509477124,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.6844,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.527071139895751,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.6856,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.6516562990872043,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.6859,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.5756882370576104,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.6944,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.575995833443124,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.6798,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.651513098302593,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.6806,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.5491117806911815,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.6413,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.6034146045315155,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.6527,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.48139118696649486,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.6671,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.5785494433011105,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.661,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.507194808648046,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.6871,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.5756410441699856,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.7042,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.490235151376117,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6822,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.5460602343178522,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.7164,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.4589736221742644,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.6692,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6449816889103409,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.6667,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.5708087701739483,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.6738,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.6059421450485811,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6799,
+      "step": 246
+    },
+    {
+      "epoch": 1.0035443037974683,
+      "grad_norm": 0.7423131370652276,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.5794,
+      "step": 247
+    },
+    {
+      "epoch": 1.0075949367088608,
+      "grad_norm": 0.5666059802728226,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.569,
+      "step": 248
+    },
+    {
+      "epoch": 1.011645569620253,
+      "grad_norm": 0.6833768360981736,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5229,
+      "step": 249
+    },
+    {
+      "epoch": 1.0156962025316456,
+      "grad_norm": 0.8056279936156647,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5728,
+      "step": 250
+    },
+    {
+      "epoch": 1.019746835443038,
+      "grad_norm": 0.5887954632373861,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5834,
+      "step": 251
+    },
+    {
+      "epoch": 1.0237974683544304,
+      "grad_norm": 0.7354538220724992,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5591,
+      "step": 252
+    },
+    {
+      "epoch": 1.0278481012658227,
+      "grad_norm": 0.5635644863396279,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5416,
+      "step": 253
+    },
+    {
+      "epoch": 1.0318987341772152,
+      "grad_norm": 0.5516432158400189,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5354,
+      "step": 254
+    },
+    {
+      "epoch": 1.0359493670886075,
+      "grad_norm": 0.6262634645981777,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.54,
+      "step": 255
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7638177527367713,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5291,
+      "step": 256
+    },
+    {
+      "epoch": 1.0440506329113923,
+      "grad_norm": 0.5623511950969552,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5374,
+      "step": 257
+    },
+    {
+      "epoch": 1.0481012658227848,
+      "grad_norm": 0.7184543964251036,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5537,
+      "step": 258
+    },
+    {
+      "epoch": 1.0521518987341771,
+      "grad_norm": 0.5745270373314391,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5649,
+      "step": 259
+    },
+    {
+      "epoch": 1.0562025316455697,
+      "grad_norm": 0.7076326946638312,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5584,
+      "step": 260
+    },
+    {
+      "epoch": 1.060253164556962,
+      "grad_norm": 0.5498684068743064,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5446,
+      "step": 261
+    },
+    {
+      "epoch": 1.0643037974683545,
+      "grad_norm": 0.6847535377303859,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5484,
+      "step": 262
+    },
+    {
+      "epoch": 1.0683544303797468,
+      "grad_norm": 0.5966896971963485,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5373,
+      "step": 263
+    },
+    {
+      "epoch": 1.0724050632911393,
+      "grad_norm": 0.6598901899348616,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5605,
+      "step": 264
+    },
+    {
+      "epoch": 1.0764556962025316,
+      "grad_norm": 0.5787899647536958,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5541,
+      "step": 265
+    },
+    {
+      "epoch": 1.080506329113924,
+      "grad_norm": 0.785749084000776,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5232,
+      "step": 266
+    },
+    {
+      "epoch": 1.0845569620253164,
+      "grad_norm": 0.5800149305214698,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5372,
+      "step": 267
+    },
+    {
+      "epoch": 1.0886075949367089,
+      "grad_norm": 0.6642058353270507,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.523,
+      "step": 268
+    },
+    {
+      "epoch": 1.0926582278481012,
+      "grad_norm": 0.6550412614458886,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5421,
+      "step": 269
+    },
+    {
+      "epoch": 1.0967088607594937,
+      "grad_norm": 0.6536573129180672,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5361,
+      "step": 270
+    },
+    {
+      "epoch": 1.100759493670886,
+      "grad_norm": 0.7573688054012108,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5322,
+      "step": 271
+    },
+    {
+      "epoch": 1.1048101265822785,
+      "grad_norm": 0.6167264054553406,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5178,
+      "step": 272
+    },
+    {
+      "epoch": 1.1088607594936708,
+      "grad_norm": 0.6456556654892814,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5384,
+      "step": 273
+    },
+    {
+      "epoch": 1.1129113924050633,
+      "grad_norm": 0.712473638770278,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5445,
+      "step": 274
+    },
+    {
+      "epoch": 1.1169620253164556,
+      "grad_norm": 0.6537418462282745,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5463,
+      "step": 275
+    },
+    {
+      "epoch": 1.1210126582278481,
+      "grad_norm": 0.7185167177728711,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5399,
+      "step": 276
+    },
+    {
+      "epoch": 1.1250632911392404,
+      "grad_norm": 0.5669136998476453,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5325,
+      "step": 277
+    },
+    {
+      "epoch": 1.129113924050633,
+      "grad_norm": 0.69774345658869,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5682,
+      "step": 278
+    },
+    {
+      "epoch": 1.1331645569620252,
+      "grad_norm": 0.5673260070514585,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5425,
+      "step": 279
+    },
+    {
+      "epoch": 1.1372151898734177,
+      "grad_norm": 0.6600435266627458,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5784,
+      "step": 280
+    },
+    {
+      "epoch": 1.14126582278481,
+      "grad_norm": 0.5900343873219048,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5619,
+      "step": 281
+    },
+    {
+      "epoch": 1.1453164556962026,
+      "grad_norm": 0.6177769287714238,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5433,
+      "step": 282
+    },
+    {
+      "epoch": 1.1493670886075948,
+      "grad_norm": 0.5250296860035064,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5539,
+      "step": 283
+    },
+    {
+      "epoch": 1.1534177215189874,
+      "grad_norm": 0.6940565414101302,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.556,
+      "step": 284
+    },
+    {
+      "epoch": 1.1574683544303797,
+      "grad_norm": 0.603685322656871,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5257,
+      "step": 285
+    },
+    {
+      "epoch": 1.1615189873417722,
+      "grad_norm": 0.6105664902492919,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5278,
+      "step": 286
+    },
+    {
+      "epoch": 1.1655696202531645,
+      "grad_norm": 0.6337032627475189,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5704,
+      "step": 287
+    },
+    {
+      "epoch": 1.169620253164557,
+      "grad_norm": 0.7758670920319001,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5633,
+      "step": 288
+    },
+    {
+      "epoch": 1.1736708860759493,
+      "grad_norm": 0.6966855387453308,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5468,
+      "step": 289
+    },
+    {
+      "epoch": 1.1777215189873418,
+      "grad_norm": 0.6460889490410895,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5365,
+      "step": 290
+    },
+    {
+      "epoch": 1.181772151898734,
+      "grad_norm": 0.7399637875273538,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5327,
+      "step": 291
+    },
+    {
+      "epoch": 1.1858227848101266,
+      "grad_norm": 0.693004080166953,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5542,
+      "step": 292
+    },
+    {
+      "epoch": 1.189873417721519,
+      "grad_norm": 0.708360370585347,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5647,
+      "step": 293
+    },
+    {
+      "epoch": 1.1939240506329114,
+      "grad_norm": 0.6419607380230467,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5404,
+      "step": 294
+    },
+    {
+      "epoch": 1.1979746835443037,
+      "grad_norm": 0.6343717315060352,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5684,
+      "step": 295
+    },
+    {
+      "epoch": 1.2020253164556962,
+      "grad_norm": 0.5815905295428346,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5342,
+      "step": 296
+    },
+    {
+      "epoch": 1.2060759493670887,
+      "grad_norm": 0.545568946555618,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5274,
+      "step": 297
+    },
+    {
+      "epoch": 1.210126582278481,
+      "grad_norm": 0.5747583374880652,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5527,
+      "step": 298
+    },
+    {
+      "epoch": 1.2141772151898733,
+      "grad_norm": 0.5780094906663592,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5484,
+      "step": 299
+    },
+    {
+      "epoch": 1.2182278481012658,
+      "grad_norm": 0.5375921583652787,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5643,
+      "step": 300
+    },
+    {
+      "epoch": 1.2222784810126583,
+      "grad_norm": 0.6149654856971485,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5132,
+      "step": 301
+    },
+    {
+      "epoch": 1.2263291139240506,
+      "grad_norm": 0.6298424309440511,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5523,
+      "step": 302
+    },
+    {
+      "epoch": 1.230379746835443,
+      "grad_norm": 0.6074320641855933,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5735,
+      "step": 303
+    },
+    {
+      "epoch": 1.2344303797468354,
+      "grad_norm": 0.6094365931357523,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5524,
+      "step": 304
+    },
+    {
+      "epoch": 1.238481012658228,
+      "grad_norm": 0.5428099116641716,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5277,
+      "step": 305
+    },
+    {
+      "epoch": 1.2425316455696203,
+      "grad_norm": 0.6475816049877877,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5407,
+      "step": 306
+    },
+    {
+      "epoch": 1.2465822784810126,
+      "grad_norm": 0.6065394243608251,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5686,
+      "step": 307
+    },
+    {
+      "epoch": 1.250632911392405,
+      "grad_norm": 0.588501970545366,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5431,
+      "step": 308
+    },
+    {
+      "epoch": 1.2546835443037976,
+      "grad_norm": 0.5645335640262126,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5285,
+      "step": 309
+    },
+    {
+      "epoch": 1.2587341772151899,
+      "grad_norm": 0.5219389373936031,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5413,
+      "step": 310
+    },
+    {
+      "epoch": 1.2627848101265822,
+      "grad_norm": 0.5871900709549536,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5649,
+      "step": 311
+    },
+    {
+      "epoch": 1.2668354430379747,
+      "grad_norm": 0.5938143238266838,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5626,
+      "step": 312
+    },
+    {
+      "epoch": 1.2708860759493672,
+      "grad_norm": 0.5555713592613682,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5316,
+      "step": 313
+    },
+    {
+      "epoch": 1.2749367088607595,
+      "grad_norm": 0.6452133289264463,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5347,
+      "step": 314
+    },
+    {
+      "epoch": 1.2789873417721518,
+      "grad_norm": 0.5733820038385815,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5484,
+      "step": 315
+    },
+    {
+      "epoch": 1.2830379746835443,
+      "grad_norm": 0.6505665656922345,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5602,
+      "step": 316
+    },
+    {
+      "epoch": 1.2870886075949368,
+      "grad_norm": 0.6285745451113557,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5374,
+      "step": 317
+    },
+    {
+      "epoch": 1.2911392405063291,
+      "grad_norm": 0.6111932643607194,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5701,
+      "step": 318
+    },
+    {
+      "epoch": 1.2951898734177214,
+      "grad_norm": 0.7252643398521333,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.5488,
+      "step": 319
+    },
+    {
+      "epoch": 1.299240506329114,
+      "grad_norm": 0.5394711215749944,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5706,
+      "step": 320
+    },
+    {
+      "epoch": 1.3032911392405064,
+      "grad_norm": 0.612396940671032,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5618,
+      "step": 321
+    },
+    {
+      "epoch": 1.3073417721518987,
+      "grad_norm": 0.5163236417104146,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5462,
+      "step": 322
+    },
+    {
+      "epoch": 1.311392405063291,
+      "grad_norm": 0.6407826501767444,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5496,
+      "step": 323
+    },
+    {
+      "epoch": 1.3154430379746835,
+      "grad_norm": 0.6318493219182842,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5419,
+      "step": 324
+    },
+    {
+      "epoch": 1.319493670886076,
+      "grad_norm": 0.528787167752284,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5403,
+      "step": 325
+    },
+    {
+      "epoch": 1.3235443037974683,
+      "grad_norm": 0.6124683608338031,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5476,
+      "step": 326
+    },
+    {
+      "epoch": 1.3275949367088606,
+      "grad_norm": 0.5993456225074266,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5431,
+      "step": 327
+    },
+    {
+      "epoch": 1.3316455696202532,
+      "grad_norm": 0.6063917665134133,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5445,
+      "step": 328
+    },
+    {
+      "epoch": 1.3356962025316457,
+      "grad_norm": 0.4875250715043547,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5329,
+      "step": 329
+    },
+    {
+      "epoch": 1.339746835443038,
+      "grad_norm": 0.6258773075173554,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5627,
+      "step": 330
+    },
+    {
+      "epoch": 1.3437974683544303,
+      "grad_norm": 0.5327337943153759,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5655,
+      "step": 331
+    },
+    {
+      "epoch": 1.3478481012658228,
+      "grad_norm": 0.5744142182382432,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.5333,
+      "step": 332
+    },
+    {
+      "epoch": 1.3518987341772153,
+      "grad_norm": 0.5826638036863402,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5491,
+      "step": 333
+    },
+    {
+      "epoch": 1.3559493670886076,
+      "grad_norm": 0.5563438499811334,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5738,
+      "step": 334
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.5857105426673532,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5384,
+      "step": 335
+    },
+    {
+      "epoch": 1.3640506329113924,
+      "grad_norm": 0.5522438174672775,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5414,
+      "step": 336
+    },
+    {
+      "epoch": 1.368101265822785,
+      "grad_norm": 0.6305289146850055,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5392,
+      "step": 337
+    },
+    {
+      "epoch": 1.3721518987341772,
+      "grad_norm": 0.6653086052577376,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5662,
+      "step": 338
+    },
+    {
+      "epoch": 1.3762025316455695,
+      "grad_norm": 0.5357116440300191,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5422,
+      "step": 339
+    },
+    {
+      "epoch": 1.380253164556962,
+      "grad_norm": 0.6471778612556419,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.563,
+      "step": 340
+    },
+    {
+      "epoch": 1.3843037974683545,
+      "grad_norm": 0.5282651340258878,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5558,
+      "step": 341
+    },
+    {
+      "epoch": 1.3883544303797468,
+      "grad_norm": 0.6577277253274287,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.562,
+      "step": 342
+    },
+    {
+      "epoch": 1.3924050632911391,
+      "grad_norm": 0.5359354755471737,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5445,
+      "step": 343
+    },
+    {
+      "epoch": 1.3964556962025316,
+      "grad_norm": 0.6377153091406164,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.542,
+      "step": 344
+    },
+    {
+      "epoch": 1.4005063291139241,
+      "grad_norm": 0.5563333606679879,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5476,
+      "step": 345
+    },
+    {
+      "epoch": 1.4045569620253164,
+      "grad_norm": 0.5764291521949273,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5822,
+      "step": 346
+    },
+    {
+      "epoch": 1.4086075949367087,
+      "grad_norm": 0.6281044480628484,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5629,
+      "step": 347
+    },
+    {
+      "epoch": 1.4126582278481012,
+      "grad_norm": 0.6103188113028086,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5393,
+      "step": 348
+    },
+    {
+      "epoch": 1.4167088607594938,
+      "grad_norm": 0.6991312772950096,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5602,
+      "step": 349
+    },
+    {
+      "epoch": 1.420759493670886,
+      "grad_norm": 0.5561353311707176,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5538,
+      "step": 350
+    },
+    {
+      "epoch": 1.4248101265822783,
+      "grad_norm": 0.6139367869057414,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5402,
+      "step": 351
+    },
+    {
+      "epoch": 1.4288607594936709,
+      "grad_norm": 0.5528837436601336,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5385,
+      "step": 352
+    },
+    {
+      "epoch": 1.4329113924050634,
+      "grad_norm": 0.6686422746305377,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5361,
+      "step": 353
+    },
+    {
+      "epoch": 1.4369620253164557,
+      "grad_norm": 0.582740663482285,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5688,
+      "step": 354
+    },
+    {
+      "epoch": 1.441012658227848,
+      "grad_norm": 0.5518342073904965,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5678,
+      "step": 355
+    },
+    {
+      "epoch": 1.4450632911392405,
+      "grad_norm": 0.5989245355321469,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5658,
+      "step": 356
+    },
+    {
+      "epoch": 1.449113924050633,
+      "grad_norm": 0.5509259134919747,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5351,
+      "step": 357
+    },
+    {
+      "epoch": 1.4531645569620253,
+      "grad_norm": 0.5696901859855683,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5597,
+      "step": 358
+    },
+    {
+      "epoch": 1.4572151898734178,
+      "grad_norm": 0.590293829933058,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5296,
+      "step": 359
+    },
+    {
+      "epoch": 1.46126582278481,
+      "grad_norm": 0.5959670830360397,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5267,
+      "step": 360
+    },
+    {
+      "epoch": 1.4653164556962026,
+      "grad_norm": 0.6609966502842927,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5703,
+      "step": 361
+    },
+    {
+      "epoch": 1.469367088607595,
+      "grad_norm": 0.5884503196278549,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5552,
+      "step": 362
+    },
+    {
+      "epoch": 1.4734177215189874,
+      "grad_norm": 0.5101509659305372,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5316,
+      "step": 363
+    },
+    {
+      "epoch": 1.4774683544303797,
+      "grad_norm": 0.6470110195789914,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5628,
+      "step": 364
+    },
+    {
+      "epoch": 1.4815189873417722,
+      "grad_norm": 0.5356385105994236,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5564,
+      "step": 365
+    },
+    {
+      "epoch": 1.4855696202531645,
+      "grad_norm": 0.629761717864451,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5561,
+      "step": 366
+    },
+    {
+      "epoch": 1.489620253164557,
+      "grad_norm": 0.5603744895514128,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5597,
+      "step": 367
+    },
+    {
+      "epoch": 1.4936708860759493,
+      "grad_norm": 0.5187793739609968,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5533,
+      "step": 368
+    },
+    {
+      "epoch": 1.4977215189873418,
+      "grad_norm": 0.5803695870577301,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5344,
+      "step": 369
+    },
+    {
+      "epoch": 1.5017721518987341,
+      "grad_norm": 0.5600491439568547,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5574,
+      "step": 370
+    },
+    {
+      "epoch": 1.5058227848101264,
+      "grad_norm": 0.6042671768107654,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.572,
+      "step": 371
+    },
+    {
+      "epoch": 1.509873417721519,
+      "grad_norm": 0.5577162419339625,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5446,
+      "step": 372
+    },
+    {
+      "epoch": 1.5139240506329115,
+      "grad_norm": 0.5538398849103182,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5769,
+      "step": 373
+    },
+    {
+      "epoch": 1.5179746835443038,
+      "grad_norm": 0.5947912338314855,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5665,
+      "step": 374
+    },
+    {
+      "epoch": 1.522025316455696,
+      "grad_norm": 0.5735163577617419,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5741,
+      "step": 375
+    },
+    {
+      "epoch": 1.5260759493670886,
+      "grad_norm": 0.599250113619715,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5377,
+      "step": 376
+    },
+    {
+      "epoch": 1.530126582278481,
+      "grad_norm": 0.4917065560217992,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5629,
+      "step": 377
+    },
+    {
+      "epoch": 1.5341772151898734,
+      "grad_norm": 0.5292323761445168,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5472,
+      "step": 378
+    },
+    {
+      "epoch": 1.5382278481012657,
+      "grad_norm": 0.46450567265307163,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5066,
+      "step": 379
+    },
+    {
+      "epoch": 1.5422784810126582,
+      "grad_norm": 0.5722550653093055,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5425,
+      "step": 380
+    },
+    {
+      "epoch": 1.5463291139240507,
+      "grad_norm": 0.599018956326515,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.528,
+      "step": 381
+    },
+    {
+      "epoch": 1.550379746835443,
+      "grad_norm": 0.5551910602250585,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5519,
+      "step": 382
+    },
+    {
+      "epoch": 1.5544303797468353,
+      "grad_norm": 0.5561900541080198,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5528,
+      "step": 383
+    },
+    {
+      "epoch": 1.5584810126582278,
+      "grad_norm": 0.5106267693261455,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5769,
+      "step": 384
+    },
+    {
+      "epoch": 1.5625316455696203,
+      "grad_norm": 0.5634288490644561,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5146,
+      "step": 385
+    },
+    {
+      "epoch": 1.5665822784810126,
+      "grad_norm": 0.5130166879953415,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5435,
+      "step": 386
+    },
+    {
+      "epoch": 1.570632911392405,
+      "grad_norm": 0.4925603260723388,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5482,
+      "step": 387
+    },
+    {
+      "epoch": 1.5746835443037974,
+      "grad_norm": 0.5196822594795975,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5336,
+      "step": 388
+    },
+    {
+      "epoch": 1.57873417721519,
+      "grad_norm": 0.5684798986944776,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5656,
+      "step": 389
+    },
+    {
+      "epoch": 1.5827848101265822,
+      "grad_norm": 0.4766071214483685,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5165,
+      "step": 390
+    },
+    {
+      "epoch": 1.5868354430379745,
+      "grad_norm": 0.4723460388209602,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5563,
+      "step": 391
+    },
+    {
+      "epoch": 1.590886075949367,
+      "grad_norm": 0.5319524301416224,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5494,
+      "step": 392
+    },
+    {
+      "epoch": 1.5949367088607596,
+      "grad_norm": 0.48943021149506666,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5667,
+      "step": 393
+    },
+    {
+      "epoch": 1.5989873417721518,
+      "grad_norm": 0.5192355634385646,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5463,
+      "step": 394
+    },
+    {
+      "epoch": 1.6030379746835441,
+      "grad_norm": 0.5053609222279576,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5476,
+      "step": 395
+    },
+    {
+      "epoch": 1.6070886075949367,
+      "grad_norm": 0.5712289044194957,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5522,
+      "step": 396
+    },
+    {
+      "epoch": 1.6111392405063292,
+      "grad_norm": 0.5663427923626153,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.541,
+      "step": 397
+    },
+    {
+      "epoch": 1.6151898734177215,
+      "grad_norm": 0.5819911823277947,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5414,
+      "step": 398
+    },
+    {
+      "epoch": 1.6192405063291138,
+      "grad_norm": 0.5899952154993683,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5237,
+      "step": 399
+    },
+    {
+      "epoch": 1.6232911392405063,
+      "grad_norm": 0.5244305788066796,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5414,
+      "step": 400
+    },
+    {
+      "epoch": 1.6273417721518988,
+      "grad_norm": 0.5896040410008375,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.577,
+      "step": 401
+    },
+    {
+      "epoch": 1.631392405063291,
+      "grad_norm": 0.5278494907177776,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5576,
+      "step": 402
+    },
+    {
+      "epoch": 1.6354430379746834,
+      "grad_norm": 0.4955037640205377,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5569,
+      "step": 403
+    },
+    {
+      "epoch": 1.639493670886076,
+      "grad_norm": 0.49596704451927887,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5581,
+      "step": 404
+    },
+    {
+      "epoch": 1.6435443037974684,
+      "grad_norm": 0.5296292510928717,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5574,
+      "step": 405
+    },
+    {
+      "epoch": 1.647594936708861,
+      "grad_norm": 0.5609978501780742,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5463,
+      "step": 406
+    },
+    {
+      "epoch": 1.6516455696202532,
+      "grad_norm": 0.5155309679863851,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5329,
+      "step": 407
+    },
+    {
+      "epoch": 1.6556962025316455,
+      "grad_norm": 0.5762678510027369,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5224,
+      "step": 408
+    },
+    {
+      "epoch": 1.659746835443038,
+      "grad_norm": 0.4759247413221198,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.5263,
+      "step": 409
+    },
+    {
+      "epoch": 1.6637974683544305,
+      "grad_norm": 0.4897276428681564,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5351,
+      "step": 410
+    },
+    {
+      "epoch": 1.6678481012658228,
+      "grad_norm": 0.4972173168991221,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5442,
+      "step": 411
+    },
+    {
+      "epoch": 1.6718987341772151,
+      "grad_norm": 0.5214610436000963,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5382,
+      "step": 412
+    },
+    {
+      "epoch": 1.6759493670886076,
+      "grad_norm": 0.4784630953992785,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5639,
+      "step": 413
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.5088928931462287,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5531,
+      "step": 414
+    },
+    {
+      "epoch": 1.6840506329113925,
+      "grad_norm": 0.5126228642505679,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5701,
+      "step": 415
+    },
+    {
+      "epoch": 1.6881012658227847,
+      "grad_norm": 0.5096497593021799,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5416,
+      "step": 416
+    },
+    {
+      "epoch": 1.6921518987341773,
+      "grad_norm": 0.5695743925022497,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5473,
+      "step": 417
+    },
+    {
+      "epoch": 1.6962025316455698,
+      "grad_norm": 0.47554764408449196,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5541,
+      "step": 418
+    },
+    {
+      "epoch": 1.700253164556962,
+      "grad_norm": 0.5340043532061052,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5375,
+      "step": 419
+    },
+    {
+      "epoch": 1.7043037974683544,
+      "grad_norm": 0.5204398030378468,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5454,
+      "step": 420
+    },
+    {
+      "epoch": 1.7083544303797469,
+      "grad_norm": 0.4811333656896627,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5462,
+      "step": 421
+    },
+    {
+      "epoch": 1.7124050632911394,
+      "grad_norm": 0.47158490322543145,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.557,
+      "step": 422
+    },
+    {
+      "epoch": 1.7164556962025317,
+      "grad_norm": 0.4882549405961056,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5515,
+      "step": 423
+    },
+    {
+      "epoch": 1.720506329113924,
+      "grad_norm": 0.5545298804565051,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5493,
+      "step": 424
+    },
+    {
+      "epoch": 1.7245569620253165,
+      "grad_norm": 0.48142843667684365,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5372,
+      "step": 425
+    },
+    {
+      "epoch": 1.728607594936709,
+      "grad_norm": 0.5027286999389239,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5522,
+      "step": 426
+    },
+    {
+      "epoch": 1.7326582278481013,
+      "grad_norm": 0.5148010661810222,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.569,
+      "step": 427
+    },
+    {
+      "epoch": 1.7367088607594936,
+      "grad_norm": 0.5323602739724823,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.553,
+      "step": 428
+    },
+    {
+      "epoch": 1.7407594936708861,
+      "grad_norm": 0.5271041447791031,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.567,
+      "step": 429
+    },
+    {
+      "epoch": 1.7448101265822786,
+      "grad_norm": 0.8430237952810636,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5414,
+      "step": 430
+    },
+    {
+      "epoch": 1.748860759493671,
+      "grad_norm": 0.5361456717166091,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.575,
+      "step": 431
+    },
+    {
+      "epoch": 1.7529113924050632,
+      "grad_norm": 0.9549988066403682,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.57,
+      "step": 432
+    },
+    {
+      "epoch": 1.7569620253164557,
+      "grad_norm": 0.5310042795303653,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5451,
+      "step": 433
+    },
+    {
+      "epoch": 1.7610126582278482,
+      "grad_norm": 0.5465178569649939,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5544,
+      "step": 434
+    },
+    {
+      "epoch": 1.7650632911392405,
+      "grad_norm": 0.5227218073483648,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5561,
+      "step": 435
+    },
+    {
+      "epoch": 1.7691139240506328,
+      "grad_norm": 0.5306817540771638,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5263,
+      "step": 436
+    },
+    {
+      "epoch": 1.7731645569620254,
+      "grad_norm": 0.5601463718613631,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5732,
+      "step": 437
+    },
+    {
+      "epoch": 1.7772151898734179,
+      "grad_norm": 0.6260789665852018,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5472,
+      "step": 438
+    },
+    {
+      "epoch": 1.7812658227848102,
+      "grad_norm": 0.585834761120176,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5751,
+      "step": 439
+    },
+    {
+      "epoch": 1.7853164556962025,
+      "grad_norm": 0.5749529897541679,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5535,
+      "step": 440
+    },
+    {
+      "epoch": 1.789367088607595,
+      "grad_norm": 0.5526335086175432,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5487,
+      "step": 441
+    },
+    {
+      "epoch": 1.7934177215189875,
+      "grad_norm": 0.5344677603345381,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5631,
+      "step": 442
+    },
+    {
+      "epoch": 1.7974683544303798,
+      "grad_norm": 0.4845397334593785,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5281,
+      "step": 443
+    },
+    {
+      "epoch": 1.801518987341772,
+      "grad_norm": 0.5087425910841364,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5564,
+      "step": 444
+    },
+    {
+      "epoch": 1.8055696202531646,
+      "grad_norm": 0.5137792748732339,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.5371,
+      "step": 445
+    },
+    {
+      "epoch": 1.809620253164557,
+      "grad_norm": 0.43753358494164224,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.545,
+      "step": 446
+    },
+    {
+      "epoch": 1.8136708860759494,
+      "grad_norm": 0.4947521507737175,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5428,
+      "step": 447
+    },
+    {
+      "epoch": 1.8177215189873417,
+      "grad_norm": 0.5031008142068853,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5444,
+      "step": 448
+    },
+    {
+      "epoch": 1.8217721518987342,
+      "grad_norm": 0.528554945837947,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5335,
+      "step": 449
+    },
+    {
+      "epoch": 1.8258227848101267,
+      "grad_norm": 0.483977988704386,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5384,
+      "step": 450
+    },
+    {
+      "epoch": 1.829873417721519,
+      "grad_norm": 0.5246556907243801,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5216,
+      "step": 451
+    },
+    {
+      "epoch": 1.8339240506329113,
+      "grad_norm": 0.5226456694153269,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5435,
+      "step": 452
+    },
+    {
+      "epoch": 1.8379746835443038,
+      "grad_norm": 0.499218188516578,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5546,
+      "step": 453
+    },
+    {
+      "epoch": 1.8420253164556963,
+      "grad_norm": 0.48813259888930244,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5586,
+      "step": 454
+    },
+    {
+      "epoch": 1.8460759493670886,
+      "grad_norm": 0.570658531813821,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5525,
+      "step": 455
+    },
+    {
+      "epoch": 1.850126582278481,
+      "grad_norm": 0.516630533498001,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5297,
+      "step": 456
+    },
+    {
+      "epoch": 1.8541772151898734,
+      "grad_norm": 0.5071607678219332,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5759,
+      "step": 457
+    },
+    {
+      "epoch": 1.858227848101266,
+      "grad_norm": 0.5357174339749406,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5615,
+      "step": 458
+    },
+    {
+      "epoch": 1.8622784810126582,
+      "grad_norm": 0.5204529686812428,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5748,
+      "step": 459
+    },
+    {
+      "epoch": 1.8663291139240505,
+      "grad_norm": 0.5781006319283905,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5296,
+      "step": 460
+    },
+    {
+      "epoch": 1.870379746835443,
+      "grad_norm": 0.49894251108505977,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5339,
+      "step": 461
+    },
+    {
+      "epoch": 1.8744303797468356,
+      "grad_norm": 0.4993735773223357,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.511,
+      "step": 462
+    },
+    {
+      "epoch": 1.8784810126582279,
+      "grad_norm": 0.49588058997735746,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5317,
+      "step": 463
+    },
+    {
+      "epoch": 1.8825316455696202,
+      "grad_norm": 1.9662137291141786,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5466,
+      "step": 464
+    },
+    {
+      "epoch": 1.8865822784810127,
+      "grad_norm": 0.5863721597051955,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5424,
+      "step": 465
+    },
+    {
+      "epoch": 1.8906329113924052,
+      "grad_norm": 0.4709576950703224,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.564,
+      "step": 466
+    },
+    {
+      "epoch": 1.8946835443037975,
+      "grad_norm": 0.4939612075111702,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5293,
+      "step": 467
+    },
+    {
+      "epoch": 1.8987341772151898,
+      "grad_norm": 0.6186166887825804,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5345,
+      "step": 468
+    },
+    {
+      "epoch": 1.9027848101265823,
+      "grad_norm": 0.51339701744342,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5537,
+      "step": 469
+    },
+    {
+      "epoch": 1.9068354430379748,
+      "grad_norm": 0.5920848483893704,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.5564,
+      "step": 470
+    },
+    {
+      "epoch": 1.910886075949367,
+      "grad_norm": 0.49862041246111305,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5257,
+      "step": 471
+    },
+    {
+      "epoch": 1.9149367088607594,
+      "grad_norm": 0.49443471393181526,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.5642,
+      "step": 472
+    },
+    {
+      "epoch": 1.918987341772152,
+      "grad_norm": 0.48980229685054705,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5297,
+      "step": 473
+    },
+    {
+      "epoch": 1.9230379746835444,
+      "grad_norm": 0.5297570875364437,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5621,
+      "step": 474
+    },
+    {
+      "epoch": 1.9270886075949367,
+      "grad_norm": 0.47062601173487706,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5326,
+      "step": 475
+    },
+    {
+      "epoch": 1.931139240506329,
+      "grad_norm": 0.4914784036331714,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5464,
+      "step": 476
+    },
+    {
+      "epoch": 1.9351898734177215,
+      "grad_norm": 0.44674269047475595,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5699,
+      "step": 477
+    },
+    {
+      "epoch": 1.939240506329114,
+      "grad_norm": 0.5014539766204374,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5303,
+      "step": 478
+    },
+    {
+      "epoch": 1.9432911392405063,
+      "grad_norm": 0.4655526175256599,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5482,
+      "step": 479
+    },
+    {
+      "epoch": 1.9473417721518986,
+      "grad_norm": 0.5415828207456148,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.5483,
+      "step": 480
+    },
+    {
+      "epoch": 1.9513924050632911,
+      "grad_norm": 0.4467217163452547,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5389,
+      "step": 481
+    },
+    {
+      "epoch": 1.9554430379746837,
+      "grad_norm": 0.5012635720275918,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5507,
+      "step": 482
+    },
+    {
+      "epoch": 1.959493670886076,
+      "grad_norm": 2.619224040738405,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5672,
+      "step": 483
+    },
+    {
+      "epoch": 1.9635443037974682,
+      "grad_norm": 0.5821372302583113,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5762,
+      "step": 484
+    },
+    {
+      "epoch": 1.9675949367088608,
+      "grad_norm": 0.42477445069482667,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5612,
+      "step": 485
+    },
+    {
+      "epoch": 1.9716455696202533,
+      "grad_norm": 1.0697333570047327,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5623,
+      "step": 486
+    },
+    {
+      "epoch": 1.9756962025316456,
+      "grad_norm": 0.5295668480828337,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.565,
+      "step": 487
+    },
+    {
+      "epoch": 1.9797468354430379,
+      "grad_norm": 0.4905052300961051,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5488,
+      "step": 488
+    },
+    {
+      "epoch": 1.9837974683544304,
+      "grad_norm": 0.5747141969104295,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.538,
+      "step": 489
+    },
+    {
+      "epoch": 1.987848101265823,
+      "grad_norm": 0.4507809555961371,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5473,
+      "step": 490
+    },
+    {
+      "epoch": 1.9918987341772152,
+      "grad_norm": 0.5847091649530249,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5366,
+      "step": 491
+    },
+    {
+      "epoch": 1.9959493670886075,
+      "grad_norm": 0.5016291627832887,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5522,
+      "step": 492
+    },
+    {
+      "epoch": 2.0030379746835445,
+      "grad_norm": 0.7123522410615091,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.4244,
+      "step": 493
+    },
+    {
+      "epoch": 2.0070886075949366,
+      "grad_norm": 0.5890854862140222,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3458,
+      "step": 494
+    },
+    {
+      "epoch": 2.011139240506329,
+      "grad_norm": 1.8349062454458818,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4,
+      "step": 495
+    },
+    {
+      "epoch": 2.0151898734177216,
+      "grad_norm": 0.5853756394217018,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.375,
+      "step": 496
+    },
+    {
+      "epoch": 2.019240506329114,
+      "grad_norm": 0.6446046973168144,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3918,
+      "step": 497
+    },
+    {
+      "epoch": 2.023291139240506,
+      "grad_norm": 0.6701048395458993,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3742,
+      "step": 498
+    },
+    {
+      "epoch": 2.0273417721518987,
+      "grad_norm": 0.5661436846614132,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3754,
+      "step": 499
+    },
+    {
+      "epoch": 2.031392405063291,
+      "grad_norm": 0.6318359271747416,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3876,
+      "step": 500
+    },
+    {
+      "epoch": 2.0354430379746837,
+      "grad_norm": 0.6073717414295051,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.3746,
+      "step": 501
+    },
+    {
+      "epoch": 2.039493670886076,
+      "grad_norm": 0.6980962372905243,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.4118,
+      "step": 502
+    },
+    {
+      "epoch": 2.0435443037974683,
+      "grad_norm": 0.5318485128419436,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.3885,
+      "step": 503
+    },
+    {
+      "epoch": 2.047594936708861,
+      "grad_norm": 0.642990777315413,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3613,
+      "step": 504
+    },
+    {
+      "epoch": 2.0516455696202534,
+      "grad_norm": 0.5390571854377426,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3652,
+      "step": 505
+    },
+    {
+      "epoch": 2.0556962025316454,
+      "grad_norm": 0.5460444099749192,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3671,
+      "step": 506
+    },
+    {
+      "epoch": 2.059746835443038,
+      "grad_norm": 0.6182976425646696,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3653,
+      "step": 507
+    },
+    {
+      "epoch": 2.0637974683544305,
+      "grad_norm": 0.609870429823148,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3706,
+      "step": 508
+    },
+    {
+      "epoch": 2.067848101265823,
+      "grad_norm": 0.523814508378828,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3598,
+      "step": 509
+    },
+    {
+      "epoch": 2.071898734177215,
+      "grad_norm": 0.539834628732631,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3653,
+      "step": 510
+    },
+    {
+      "epoch": 2.0759493670886076,
+      "grad_norm": 0.5363532429468977,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3729,
+      "step": 511
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.5556318697060987,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3884,
+      "step": 512
+    },
+    {
+      "epoch": 2.0840506329113926,
+      "grad_norm": 0.5062504464721904,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.391,
+      "step": 513
+    },
+    {
+      "epoch": 2.0881012658227847,
+      "grad_norm": 0.5759170833356876,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.383,
+      "step": 514
+    },
+    {
+      "epoch": 2.092151898734177,
+      "grad_norm": 0.4507194600845694,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3808,
+      "step": 515
+    },
+    {
+      "epoch": 2.0962025316455697,
+      "grad_norm": 0.5469838260232879,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3948,
+      "step": 516
+    },
+    {
+      "epoch": 2.100253164556962,
+      "grad_norm": 0.45552763258096124,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3607,
+      "step": 517
+    },
+    {
+      "epoch": 2.1043037974683543,
+      "grad_norm": 0.506428810630819,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3735,
+      "step": 518
+    },
+    {
+      "epoch": 2.108354430379747,
+      "grad_norm": 0.4871382273602299,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3446,
+      "step": 519
+    },
+    {
+      "epoch": 2.1124050632911393,
+      "grad_norm": 0.5307678285746118,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3767,
+      "step": 520
+    },
+    {
+      "epoch": 2.116455696202532,
+      "grad_norm": 0.43816640831422793,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.345,
+      "step": 521
+    },
+    {
+      "epoch": 2.120506329113924,
+      "grad_norm": 0.4881413334283607,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.393,
+      "step": 522
+    },
+    {
+      "epoch": 2.1245569620253164,
+      "grad_norm": 0.48317432897573975,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3649,
+      "step": 523
+    },
+    {
+      "epoch": 2.128607594936709,
+      "grad_norm": 0.4915895820494067,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.404,
+      "step": 524
+    },
+    {
+      "epoch": 2.1326582278481014,
+      "grad_norm": 0.5054459286360102,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3892,
+      "step": 525
+    },
+    {
+      "epoch": 2.1367088607594935,
+      "grad_norm": 0.4169348576786,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3633,
+      "step": 526
+    },
+    {
+      "epoch": 2.140759493670886,
+      "grad_norm": 0.5095200949376132,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3946,
+      "step": 527
+    },
+    {
+      "epoch": 2.1448101265822785,
+      "grad_norm": 0.4623242302114887,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.393,
+      "step": 528
+    },
+    {
+      "epoch": 2.148860759493671,
+      "grad_norm": 0.48943092385186787,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.402,
+      "step": 529
+    },
+    {
+      "epoch": 2.152911392405063,
+      "grad_norm": 0.44072914046623013,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.4123,
+      "step": 530
+    },
+    {
+      "epoch": 2.1569620253164556,
+      "grad_norm": 0.48558776489574246,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3916,
+      "step": 531
+    },
+    {
+      "epoch": 2.161012658227848,
+      "grad_norm": 0.41708194262991666,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4087,
+      "step": 532
+    },
+    {
+      "epoch": 2.1650632911392407,
+      "grad_norm": 0.54495345077615,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3777,
+      "step": 533
+    },
+    {
+      "epoch": 2.1691139240506327,
+      "grad_norm": 0.4180215653600075,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3634,
+      "step": 534
+    },
+    {
+      "epoch": 2.1731645569620253,
+      "grad_norm": 0.47964998710101336,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.402,
+      "step": 535
+    },
+    {
+      "epoch": 2.1772151898734178,
+      "grad_norm": 0.4375096870892239,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3529,
+      "step": 536
+    },
+    {
+      "epoch": 2.1812658227848103,
+      "grad_norm": 0.48324982019225343,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3797,
+      "step": 537
+    },
+    {
+      "epoch": 2.1853164556962024,
+      "grad_norm": 0.424777857138626,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3816,
+      "step": 538
+    },
+    {
+      "epoch": 2.189367088607595,
+      "grad_norm": 0.5121802470265813,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.385,
+      "step": 539
+    },
+    {
+      "epoch": 2.1934177215189874,
+      "grad_norm": 0.43321632643737606,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3852,
+      "step": 540
+    },
+    {
+      "epoch": 2.19746835443038,
+      "grad_norm": 0.473359224628763,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3666,
+      "step": 541
+    },
+    {
+      "epoch": 2.201518987341772,
+      "grad_norm": 0.4340282391357051,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3725,
+      "step": 542
+    },
+    {
+      "epoch": 2.2055696202531645,
+      "grad_norm": 0.4480098756888857,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.3994,
+      "step": 543
+    },
+    {
+      "epoch": 2.209620253164557,
+      "grad_norm": 0.4475828660998836,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3934,
+      "step": 544
+    },
+    {
+      "epoch": 2.2136708860759495,
+      "grad_norm": 0.4551473095323218,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.394,
+      "step": 545
+    },
+    {
+      "epoch": 2.2177215189873416,
+      "grad_norm": 0.4465508739489327,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3661,
+      "step": 546
+    },
+    {
+      "epoch": 2.221772151898734,
+      "grad_norm": 0.4523719829841633,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.3785,
+      "step": 547
+    },
+    {
+      "epoch": 2.2258227848101266,
+      "grad_norm": 0.4243457753369533,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3811,
+      "step": 548
+    },
+    {
+      "epoch": 2.229873417721519,
+      "grad_norm": 0.4259505942041634,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3889,
+      "step": 549
+    },
+    {
+      "epoch": 2.233924050632911,
+      "grad_norm": 0.41445929927695935,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3943,
+      "step": 550
+    },
+    {
+      "epoch": 2.2379746835443037,
+      "grad_norm": 0.4238407078192064,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3801,
+      "step": 551
+    },
+    {
+      "epoch": 2.2420253164556962,
+      "grad_norm": 0.43792743651275917,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3801,
+      "step": 552
+    },
+    {
+      "epoch": 2.2460759493670888,
+      "grad_norm": 0.44050623540045813,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3678,
+      "step": 553
+    },
+    {
+      "epoch": 2.250126582278481,
+      "grad_norm": 0.4207788000321435,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3823,
+      "step": 554
+    },
+    {
+      "epoch": 2.2541772151898734,
+      "grad_norm": 0.4615252912956907,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3998,
+      "step": 555
+    },
+    {
+      "epoch": 2.258227848101266,
+      "grad_norm": 0.4084034219949688,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.3578,
+      "step": 556
+    },
+    {
+      "epoch": 2.2622784810126584,
+      "grad_norm": 0.40791713235305277,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.39,
+      "step": 557
+    },
+    {
+      "epoch": 2.2663291139240505,
+      "grad_norm": 0.42896133413352944,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3805,
+      "step": 558
+    },
+    {
+      "epoch": 2.270379746835443,
+      "grad_norm": 0.43994467276990823,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3816,
+      "step": 559
+    },
+    {
+      "epoch": 2.2744303797468355,
+      "grad_norm": 0.4468065519221976,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3709,
+      "step": 560
+    },
+    {
+      "epoch": 2.278481012658228,
+      "grad_norm": 0.4222194481998099,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3961,
+      "step": 561
+    },
+    {
+      "epoch": 2.28253164556962,
+      "grad_norm": 0.41652508186935794,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.376,
+      "step": 562
+    },
+    {
+      "epoch": 2.2865822784810126,
+      "grad_norm": 0.44384679219024636,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.369,
+      "step": 563
+    },
+    {
+      "epoch": 2.290632911392405,
+      "grad_norm": 0.4111729731666362,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.4034,
+      "step": 564
+    },
+    {
+      "epoch": 2.2946835443037976,
+      "grad_norm": 0.412157398504703,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3951,
+      "step": 565
+    },
+    {
+      "epoch": 2.2987341772151897,
+      "grad_norm": 0.4535085116140931,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3944,
+      "step": 566
+    },
+    {
+      "epoch": 2.302784810126582,
+      "grad_norm": 0.416403743178123,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3961,
+      "step": 567
+    },
+    {
+      "epoch": 2.3068354430379747,
+      "grad_norm": 0.397013816519887,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3815,
+      "step": 568
+    },
+    {
+      "epoch": 2.3108860759493672,
+      "grad_norm": 0.4428318983159883,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3677,
+      "step": 569
+    },
+    {
+      "epoch": 2.3149367088607593,
+      "grad_norm": 0.4164240478980906,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3755,
+      "step": 570
+    },
+    {
+      "epoch": 2.318987341772152,
+      "grad_norm": 0.4625885362040827,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3932,
+      "step": 571
+    },
+    {
+      "epoch": 2.3230379746835443,
+      "grad_norm": 0.43228463208607515,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3802,
+      "step": 572
+    },
+    {
+      "epoch": 2.327088607594937,
+      "grad_norm": 0.4516417937483259,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3786,
+      "step": 573
+    },
+    {
+      "epoch": 2.331139240506329,
+      "grad_norm": 0.5301435294252334,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3756,
+      "step": 574
+    },
+    {
+      "epoch": 2.3351898734177214,
+      "grad_norm": 0.4507271360405122,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3893,
+      "step": 575
+    },
+    {
+      "epoch": 2.339240506329114,
+      "grad_norm": 0.5583474387358044,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.4123,
+      "step": 576
+    },
+    {
+      "epoch": 2.3432911392405065,
+      "grad_norm": 0.44480279161698877,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3742,
+      "step": 577
+    },
+    {
+      "epoch": 2.3473417721518985,
+      "grad_norm": 0.4786439894429779,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3823,
+      "step": 578
+    },
+    {
+      "epoch": 2.351392405063291,
+      "grad_norm": 0.5755227813903568,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3938,
+      "step": 579
+    },
+    {
+      "epoch": 2.3554430379746836,
+      "grad_norm": 0.42863552160611557,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3902,
+      "step": 580
+    },
+    {
+      "epoch": 2.359493670886076,
+      "grad_norm": 0.5605517168810585,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3695,
+      "step": 581
+    },
+    {
+      "epoch": 2.363544303797468,
+      "grad_norm": 0.4689272812941543,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4116,
+      "step": 582
+    },
+    {
+      "epoch": 2.3675949367088607,
+      "grad_norm": 0.4695190980671398,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3852,
+      "step": 583
+    },
+    {
+      "epoch": 2.371645569620253,
+      "grad_norm": 0.5099668572709174,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3975,
+      "step": 584
+    },
+    {
+      "epoch": 2.3756962025316457,
+      "grad_norm": 0.437633356134844,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3857,
+      "step": 585
+    },
+    {
+      "epoch": 2.379746835443038,
+      "grad_norm": 0.46542470609639414,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3527,
+      "step": 586
+    },
+    {
+      "epoch": 2.3837974683544303,
+      "grad_norm": 0.47523854807550436,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3991,
+      "step": 587
+    },
+    {
+      "epoch": 2.387848101265823,
+      "grad_norm": 0.4384220349630565,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.4076,
+      "step": 588
+    },
+    {
+      "epoch": 2.3918987341772153,
+      "grad_norm": 0.44473949975888005,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3793,
+      "step": 589
+    },
+    {
+      "epoch": 2.3959493670886074,
+      "grad_norm": 0.41747697302993586,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3915,
+      "step": 590
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.44489955306132206,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3762,
+      "step": 591
+    },
+    {
+      "epoch": 2.4040506329113924,
+      "grad_norm": 0.42809006057397936,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.3653,
+      "step": 592
+    },
+    {
+      "epoch": 2.408101265822785,
+      "grad_norm": 0.4420209966186644,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3942,
+      "step": 593
+    },
+    {
+      "epoch": 2.4121518987341775,
+      "grad_norm": 0.4204223814808839,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3953,
+      "step": 594
+    },
+    {
+      "epoch": 2.4162025316455695,
+      "grad_norm": 0.4355471652382909,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4026,
+      "step": 595
+    },
+    {
+      "epoch": 2.420253164556962,
+      "grad_norm": 0.4198527241267566,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3804,
+      "step": 596
+    },
+    {
+      "epoch": 2.4243037974683546,
+      "grad_norm": 0.4565180569496214,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3818,
+      "step": 597
+    },
+    {
+      "epoch": 2.4283544303797466,
+      "grad_norm": 0.45299278821382816,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3784,
+      "step": 598
+    },
+    {
+      "epoch": 2.432405063291139,
+      "grad_norm": 0.44026616871293256,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3923,
+      "step": 599
+    },
+    {
+      "epoch": 2.4364556962025317,
+      "grad_norm": 0.41603968968036886,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3961,
+      "step": 600
+    },
+    {
+      "epoch": 2.440506329113924,
+      "grad_norm": 0.4684903780613742,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.378,
+      "step": 601
+    },
+    {
+      "epoch": 2.4445569620253167,
+      "grad_norm": 0.45055656766792257,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3781,
+      "step": 602
+    },
+    {
+      "epoch": 2.4486075949367088,
+      "grad_norm": 0.4156572665434236,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3907,
+      "step": 603
+    },
+    {
+      "epoch": 2.4526582278481013,
+      "grad_norm": 0.4159003162491722,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3871,
+      "step": 604
+    },
+    {
+      "epoch": 2.456708860759494,
+      "grad_norm": 0.4471120655649881,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3922,
+      "step": 605
+    },
+    {
+      "epoch": 2.460759493670886,
+      "grad_norm": 0.4403705684439131,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3797,
+      "step": 606
+    },
+    {
+      "epoch": 2.4648101265822784,
+      "grad_norm": 0.4661623810129075,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3823,
+      "step": 607
+    },
+    {
+      "epoch": 2.468860759493671,
+      "grad_norm": 0.42795913891497284,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3601,
+      "step": 608
+    },
+    {
+      "epoch": 2.4729113924050634,
+      "grad_norm": 0.43064230536936254,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3877,
+      "step": 609
+    },
+    {
+      "epoch": 2.476962025316456,
+      "grad_norm": 0.45118073910756673,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3858,
+      "step": 610
+    },
+    {
+      "epoch": 2.481012658227848,
+      "grad_norm": 0.4538695120318161,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.375,
+      "step": 611
+    },
+    {
+      "epoch": 2.4850632911392405,
+      "grad_norm": 0.5068223721972698,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3574,
+      "step": 612
+    },
+    {
+      "epoch": 2.489113924050633,
+      "grad_norm": 0.4793665868911935,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3915,
+      "step": 613
+    },
+    {
+      "epoch": 2.493164556962025,
+      "grad_norm": 0.4407257551118661,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3974,
+      "step": 614
+    },
+    {
+      "epoch": 2.4972151898734176,
+      "grad_norm": 0.5713933327994071,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3921,
+      "step": 615
+    },
+    {
+      "epoch": 2.50126582278481,
+      "grad_norm": 0.4364351450939994,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3796,
+      "step": 616
+    },
+    {
+      "epoch": 2.5053164556962026,
+      "grad_norm": 0.4858906975959655,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3963,
+      "step": 617
+    },
+    {
+      "epoch": 2.509367088607595,
+      "grad_norm": 0.4885002129107575,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.4117,
+      "step": 618
+    },
+    {
+      "epoch": 2.5134177215189872,
+      "grad_norm": 0.4290536651374336,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3783,
+      "step": 619
+    },
+    {
+      "epoch": 2.5174683544303798,
+      "grad_norm": 0.5043090666905959,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3933,
+      "step": 620
+    },
+    {
+      "epoch": 2.5215189873417723,
+      "grad_norm": 0.5190666786692868,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3985,
+      "step": 621
+    },
+    {
+      "epoch": 2.5255696202531643,
+      "grad_norm": 0.4413729818331779,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3726,
+      "step": 622
+    },
+    {
+      "epoch": 2.529620253164557,
+      "grad_norm": 0.4655646307250244,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3937,
+      "step": 623
+    },
+    {
+      "epoch": 2.5336708860759494,
+      "grad_norm": 0.48845815690229677,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3895,
+      "step": 624
+    },
+    {
+      "epoch": 2.537721518987342,
+      "grad_norm": 0.45454470166075517,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.3964,
+      "step": 625
+    },
+    {
+      "epoch": 2.5417721518987344,
+      "grad_norm": 0.5475854557565052,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3825,
+      "step": 626
+    },
+    {
+      "epoch": 2.5458227848101265,
+      "grad_norm": 0.39494688433121594,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3884,
+      "step": 627
+    },
+    {
+      "epoch": 2.549873417721519,
+      "grad_norm": 0.4913285752383977,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3779,
+      "step": 628
+    },
+    {
+      "epoch": 2.5539240506329115,
+      "grad_norm": 0.5166101434046626,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3953,
+      "step": 629
+    },
+    {
+      "epoch": 2.5579746835443036,
+      "grad_norm": 0.42911628387629264,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3718,
+      "step": 630
+    },
+    {
+      "epoch": 2.562025316455696,
+      "grad_norm": 0.4140400608020258,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.4081,
+      "step": 631
+    },
+    {
+      "epoch": 2.5660759493670886,
+      "grad_norm": 0.44890835163195625,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3901,
+      "step": 632
+    },
+    {
+      "epoch": 2.570126582278481,
+      "grad_norm": 0.3898547228426782,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3792,
+      "step": 633
+    },
+    {
+      "epoch": 2.5741772151898736,
+      "grad_norm": 0.42154464133770564,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4045,
+      "step": 634
+    },
+    {
+      "epoch": 2.5782278481012657,
+      "grad_norm": 0.41496169026965507,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3751,
+      "step": 635
+    },
+    {
+      "epoch": 2.5822784810126582,
+      "grad_norm": 0.4122453107260126,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.384,
+      "step": 636
+    },
+    {
+      "epoch": 2.5863291139240507,
+      "grad_norm": 0.4418000053800445,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3885,
+      "step": 637
+    },
+    {
+      "epoch": 2.590379746835443,
+      "grad_norm": 0.4353061271508657,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3965,
+      "step": 638
+    },
+    {
+      "epoch": 2.5944303797468353,
+      "grad_norm": 0.42527035607933417,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.3858,
+      "step": 639
+    },
+    {
+      "epoch": 2.598481012658228,
+      "grad_norm": 0.41506627107843136,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3904,
+      "step": 640
+    },
+    {
+      "epoch": 2.6025316455696204,
+      "grad_norm": 0.4026292785247221,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3964,
+      "step": 641
+    },
+    {
+      "epoch": 2.606582278481013,
+      "grad_norm": 0.43077423831206263,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.4028,
+      "step": 642
+    },
+    {
+      "epoch": 2.610632911392405,
+      "grad_norm": 0.4235679405055007,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3789,
+      "step": 643
+    },
+    {
+      "epoch": 2.6146835443037975,
+      "grad_norm": 0.43431599889302197,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3947,
+      "step": 644
+    },
+    {
+      "epoch": 2.61873417721519,
+      "grad_norm": 0.379748856558887,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3901,
+      "step": 645
+    },
+    {
+      "epoch": 2.622784810126582,
+      "grad_norm": 0.4244040268095676,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3862,
+      "step": 646
+    },
+    {
+      "epoch": 2.6268354430379746,
+      "grad_norm": 0.416160620613638,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3675,
+      "step": 647
+    },
+    {
+      "epoch": 2.630886075949367,
+      "grad_norm": 0.4311113036887535,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3697,
+      "step": 648
+    },
+    {
+      "epoch": 2.6349367088607596,
+      "grad_norm": 0.4783641019622,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3812,
+      "step": 649
+    },
+    {
+      "epoch": 2.638987341772152,
+      "grad_norm": 0.43437308243071493,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3887,
+      "step": 650
+    },
+    {
+      "epoch": 2.643037974683544,
+      "grad_norm": 0.47302328265013094,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3866,
+      "step": 651
+    },
+    {
+      "epoch": 2.6470886075949367,
+      "grad_norm": 0.43159412796278246,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3956,
+      "step": 652
+    },
+    {
+      "epoch": 2.651139240506329,
+      "grad_norm": 0.4255473653775355,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3657,
+      "step": 653
+    },
+    {
+      "epoch": 2.6551898734177213,
+      "grad_norm": 0.44477269375799566,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.409,
+      "step": 654
+    },
+    {
+      "epoch": 2.659240506329114,
+      "grad_norm": 0.41620125208113823,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.369,
+      "step": 655
+    },
+    {
+      "epoch": 2.6632911392405063,
+      "grad_norm": 0.44135760836721055,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3795,
+      "step": 656
+    },
+    {
+      "epoch": 2.667341772151899,
+      "grad_norm": 0.4247590324304213,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3919,
+      "step": 657
+    },
+    {
+      "epoch": 2.6713924050632913,
+      "grad_norm": 0.4451955003626694,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3677,
+      "step": 658
+    },
+    {
+      "epoch": 2.6754430379746834,
+      "grad_norm": 0.41097073543377005,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3725,
+      "step": 659
+    },
+    {
+      "epoch": 2.679493670886076,
+      "grad_norm": 0.4490981896848652,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3942,
+      "step": 660
+    },
+    {
+      "epoch": 2.6835443037974684,
+      "grad_norm": 0.45953677079344674,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3934,
+      "step": 661
+    },
+    {
+      "epoch": 2.6875949367088605,
+      "grad_norm": 0.4114569763507636,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.4124,
+      "step": 662
+    },
+    {
+      "epoch": 2.691645569620253,
+      "grad_norm": 0.4327278313281332,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3885,
+      "step": 663
+    },
+    {
+      "epoch": 2.6956962025316455,
+      "grad_norm": 0.3998228563484913,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3675,
+      "step": 664
+    },
+    {
+      "epoch": 2.699746835443038,
+      "grad_norm": 0.4149492408818599,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4043,
+      "step": 665
+    },
+    {
+      "epoch": 2.7037974683544306,
+      "grad_norm": 0.40422830334942933,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3861,
+      "step": 666
+    },
+    {
+      "epoch": 2.7078481012658226,
+      "grad_norm": 0.4008094997622014,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3886,
+      "step": 667
+    },
+    {
+      "epoch": 2.711898734177215,
+      "grad_norm": 0.4343032440602679,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3689,
+      "step": 668
+    },
+    {
+      "epoch": 2.7159493670886077,
+      "grad_norm": 0.41983621240376534,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.375,
+      "step": 669
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.4023012382958717,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3749,
+      "step": 670
+    },
+    {
+      "epoch": 2.7240506329113923,
+      "grad_norm": 0.39767532397208727,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3927,
+      "step": 671
+    },
+    {
+      "epoch": 2.728101265822785,
+      "grad_norm": 0.4035940419110842,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3885,
+      "step": 672
+    },
+    {
+      "epoch": 2.7321518987341773,
+      "grad_norm": 0.3891760196523704,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3918,
+      "step": 673
+    },
+    {
+      "epoch": 2.73620253164557,
+      "grad_norm": 0.4201024361293722,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3991,
+      "step": 674
+    },
+    {
+      "epoch": 2.740253164556962,
+      "grad_norm": 0.4165945183968349,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3971,
+      "step": 675
+    },
+    {
+      "epoch": 2.7443037974683544,
+      "grad_norm": 0.39853812102740405,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3826,
+      "step": 676
+    },
+    {
+      "epoch": 2.748354430379747,
+      "grad_norm": 0.4243273478211013,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3892,
+      "step": 677
+    },
+    {
+      "epoch": 2.752405063291139,
+      "grad_norm": 0.42386508923504346,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3892,
+      "step": 678
+    },
+    {
+      "epoch": 2.7564556962025315,
+      "grad_norm": 0.39677726866007673,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3954,
+      "step": 679
+    },
+    {
+      "epoch": 2.760506329113924,
+      "grad_norm": 0.3928233344248664,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3766,
+      "step": 680
+    },
+    {
+      "epoch": 2.7645569620253165,
+      "grad_norm": 0.40840022407928483,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3483,
+      "step": 681
+    },
+    {
+      "epoch": 2.768607594936709,
+      "grad_norm": 0.4235708173415745,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3766,
+      "step": 682
+    },
+    {
+      "epoch": 2.772658227848101,
+      "grad_norm": 0.42196781710225134,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3832,
+      "step": 683
+    },
+    {
+      "epoch": 2.7767088607594936,
+      "grad_norm": 0.41088938084178167,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.3802,
+      "step": 684
+    },
+    {
+      "epoch": 2.780759493670886,
+      "grad_norm": 0.41090310102521393,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.3658,
+      "step": 685
+    },
+    {
+      "epoch": 2.7848101265822782,
+      "grad_norm": 0.41661422287815014,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3785,
+      "step": 686
+    },
+    {
+      "epoch": 2.7888607594936707,
+      "grad_norm": 0.4194220522755447,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3675,
+      "step": 687
+    },
+    {
+      "epoch": 2.7929113924050633,
+      "grad_norm": 0.4090496660008446,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3777,
+      "step": 688
+    },
+    {
+      "epoch": 2.7969620253164558,
+      "grad_norm": 0.4278660656546428,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3984,
+      "step": 689
+    },
+    {
+      "epoch": 2.8010126582278483,
+      "grad_norm": 0.4129824336998422,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4042,
+      "step": 690
+    },
+    {
+      "epoch": 2.8050632911392404,
+      "grad_norm": 0.401646580863828,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3992,
+      "step": 691
+    },
+    {
+      "epoch": 2.809113924050633,
+      "grad_norm": 0.410014755708853,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3732,
+      "step": 692
+    },
+    {
+      "epoch": 2.8131645569620254,
+      "grad_norm": 0.40773472363117347,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3931,
+      "step": 693
+    },
+    {
+      "epoch": 2.8172151898734175,
+      "grad_norm": 0.43709060706111086,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3938,
+      "step": 694
+    },
+    {
+      "epoch": 2.82126582278481,
+      "grad_norm": 0.40254779592180634,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3861,
+      "step": 695
+    },
+    {
+      "epoch": 2.8253164556962025,
+      "grad_norm": 0.4308076419976859,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3781,
+      "step": 696
+    },
+    {
+      "epoch": 2.829367088607595,
+      "grad_norm": 0.41650232896095246,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3866,
+      "step": 697
+    },
+    {
+      "epoch": 2.8334177215189875,
+      "grad_norm": 0.4277653167758347,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3805,
+      "step": 698
+    },
+    {
+      "epoch": 2.8374683544303796,
+      "grad_norm": 0.39922172038686143,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.3602,
+      "step": 699
+    },
+    {
+      "epoch": 2.841518987341772,
+      "grad_norm": 0.4334210307073013,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3738,
+      "step": 700
+    },
+    {
+      "epoch": 2.8455696202531646,
+      "grad_norm": 0.40740709930590935,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3685,
+      "step": 701
+    },
+    {
+      "epoch": 2.8496202531645567,
+      "grad_norm": 0.41349667489728903,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3838,
+      "step": 702
+    },
+    {
+      "epoch": 2.853670886075949,
+      "grad_norm": 0.3986186413902107,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3897,
+      "step": 703
+    },
+    {
+      "epoch": 2.8577215189873417,
+      "grad_norm": 0.41013648908672695,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3615,
+      "step": 704
+    },
+    {
+      "epoch": 2.8617721518987342,
+      "grad_norm": 0.4117579557566084,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3972,
+      "step": 705
+    },
+    {
+      "epoch": 2.8658227848101268,
+      "grad_norm": 0.3957359915421058,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3812,
+      "step": 706
+    },
+    {
+      "epoch": 2.869873417721519,
+      "grad_norm": 0.41426061260698094,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.3639,
+      "step": 707
+    },
+    {
+      "epoch": 2.8739240506329113,
+      "grad_norm": 0.38925596637701726,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3967,
+      "step": 708
+    },
+    {
+      "epoch": 2.877974683544304,
+      "grad_norm": 0.4199958627692816,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.3792,
+      "step": 709
+    },
+    {
+      "epoch": 2.882025316455696,
+      "grad_norm": 0.38563501233449005,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3948,
+      "step": 710
+    },
+    {
+      "epoch": 2.8860759493670884,
+      "grad_norm": 0.38312726227410016,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.3968,
+      "step": 711
+    },
+    {
+      "epoch": 2.890126582278481,
+      "grad_norm": 0.404887547598452,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.374,
+      "step": 712
+    },
+    {
+      "epoch": 2.8941772151898735,
+      "grad_norm": 0.4117862502835243,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.396,
+      "step": 713
+    },
+    {
+      "epoch": 2.898227848101266,
+      "grad_norm": 0.4070141606684349,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.3799,
+      "step": 714
+    },
+    {
+      "epoch": 2.902278481012658,
+      "grad_norm": 0.405441718497378,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.4031,
+      "step": 715
+    },
+    {
+      "epoch": 2.9063291139240506,
+      "grad_norm": 0.40143220324374623,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3889,
+      "step": 716
+    },
+    {
+      "epoch": 2.910379746835443,
+      "grad_norm": 0.3855711599072202,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3855,
+      "step": 717
+    },
+    {
+      "epoch": 2.9144303797468356,
+      "grad_norm": 0.41626617182577286,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3688,
+      "step": 718
+    },
+    {
+      "epoch": 2.918481012658228,
+      "grad_norm": 0.405760741522798,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3658,
+      "step": 719
+    },
+    {
+      "epoch": 2.92253164556962,
+      "grad_norm": 0.4018684011235456,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3902,
+      "step": 720
+    },
+    {
+      "epoch": 2.9265822784810127,
+      "grad_norm": 0.4096286957805344,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3913,
+      "step": 721
+    },
+    {
+      "epoch": 2.9306329113924052,
+      "grad_norm": 0.893563933556425,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3877,
+      "step": 722
+    },
+    {
+      "epoch": 2.9346835443037973,
+      "grad_norm": 0.42407634871323835,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3883,
+      "step": 723
+    },
+    {
+      "epoch": 2.93873417721519,
+      "grad_norm": 0.3963467834371664,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3515,
+      "step": 724
+    },
+    {
+      "epoch": 2.9427848101265823,
+      "grad_norm": 0.427838729891535,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3637,
+      "step": 725
+    },
+    {
+      "epoch": 2.946835443037975,
+      "grad_norm": 0.40826065964617136,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3836,
+      "step": 726
+    },
+    {
+      "epoch": 2.9508860759493674,
+      "grad_norm": 0.41386503920804757,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3942,
+      "step": 727
+    },
+    {
+      "epoch": 2.9549367088607594,
+      "grad_norm": 0.401027434675032,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3725,
+      "step": 728
+    },
+    {
+      "epoch": 2.958987341772152,
+      "grad_norm": 0.44623779369757505,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.377,
+      "step": 729
+    },
+    {
+      "epoch": 2.9630379746835445,
+      "grad_norm": 0.4055273005516531,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.4016,
+      "step": 730
+    },
+    {
+      "epoch": 2.9670886075949365,
+      "grad_norm": 0.41673218129866,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3975,
+      "step": 731
+    },
+    {
+      "epoch": 2.971139240506329,
+      "grad_norm": 0.4523298977574976,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3771,
+      "step": 732
+    },
+    {
+      "epoch": 2.9751898734177216,
+      "grad_norm": 0.3966625714888095,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3851,
+      "step": 733
+    },
+    {
+      "epoch": 2.979240506329114,
+      "grad_norm": 0.4701092509098863,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3764,
+      "step": 734
+    },
+    {
+      "epoch": 2.9832911392405066,
+      "grad_norm": 0.46887103149214004,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3712,
+      "step": 735
+    },
+    {
+      "epoch": 2.9873417721518987,
+      "grad_norm": 0.39232876480653717,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3734,
+      "step": 736
+    },
+    {
+      "epoch": 2.991392405063291,
+      "grad_norm": 0.4387016441903268,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3844,
+      "step": 737
+    },
+    {
+      "epoch": 2.9954430379746837,
+      "grad_norm": 0.40047511639438993,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3925,
+      "step": 738
+    },
+    {
+      "epoch": 3.0025316455696203,
+      "grad_norm": 0.49395503734026375,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3444,
+      "step": 739
+    },
+    {
+      "epoch": 3.006582278481013,
+      "grad_norm": 0.4822901048820036,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.2834,
+      "step": 740
+    },
+    {
+      "epoch": 3.010632911392405,
+      "grad_norm": 0.4342731546933268,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.262,
+      "step": 741
+    },
+    {
+      "epoch": 3.0146835443037974,
+      "grad_norm": 0.7022005034461969,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.2802,
+      "step": 742
+    },
+    {
+      "epoch": 3.01873417721519,
+      "grad_norm": 0.5289648016290172,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2635,
+      "step": 743
+    },
+    {
+      "epoch": 3.0227848101265824,
+      "grad_norm": 0.4918288916101244,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2313,
+      "step": 744
+    },
+    {
+      "epoch": 3.0268354430379745,
+      "grad_norm": 0.5106252078783823,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2485,
+      "step": 745
+    },
+    {
+      "epoch": 3.030886075949367,
+      "grad_norm": 0.500266159471987,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2737,
+      "step": 746
+    },
+    {
+      "epoch": 3.0349367088607595,
+      "grad_norm": 0.40703072258955336,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.2592,
+      "step": 747
+    },
+    {
+      "epoch": 3.038987341772152,
+      "grad_norm": 0.4174011728864301,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2711,
+      "step": 748
+    },
+    {
+      "epoch": 3.043037974683544,
+      "grad_norm": 0.49450168958564406,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2557,
+      "step": 749
+    },
+    {
+      "epoch": 3.0470886075949366,
+      "grad_norm": 0.408463590324864,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2755,
+      "step": 750
+    },
+    {
+      "epoch": 3.051139240506329,
+      "grad_norm": 0.3742351711638953,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.253,
+      "step": 751
+    },
+    {
+      "epoch": 3.0551898734177216,
+      "grad_norm": 0.3825194431003038,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2404,
+      "step": 752
+    },
+    {
+      "epoch": 3.0592405063291137,
+      "grad_norm": 0.426056653068137,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.2646,
+      "step": 753
+    },
+    {
+      "epoch": 3.0632911392405062,
+      "grad_norm": 0.4504499233189166,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.2675,
+      "step": 754
+    },
+    {
+      "epoch": 3.0673417721518987,
+      "grad_norm": 0.4117028901557912,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2866,
+      "step": 755
+    },
+    {
+      "epoch": 3.0713924050632913,
+      "grad_norm": 0.37564837034201265,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2694,
+      "step": 756
+    },
+    {
+      "epoch": 3.0754430379746838,
+      "grad_norm": 0.36485289741581745,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2786,
+      "step": 757
+    },
+    {
+      "epoch": 3.079493670886076,
+      "grad_norm": 0.3919949534051962,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2693,
+      "step": 758
+    },
+    {
+      "epoch": 3.0835443037974684,
+      "grad_norm": 0.36625941473969637,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.2643,
+      "step": 759
+    },
+    {
+      "epoch": 3.087594936708861,
+      "grad_norm": 0.35872790378380826,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2681,
+      "step": 760
+    },
+    {
+      "epoch": 3.091645569620253,
+      "grad_norm": 0.37392108324359197,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2494,
+      "step": 761
+    },
+    {
+      "epoch": 3.0956962025316455,
+      "grad_norm": 0.38957782489525195,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2447,
+      "step": 762
+    },
+    {
+      "epoch": 3.099746835443038,
+      "grad_norm": 0.3543304736559721,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.2689,
+      "step": 763
+    },
+    {
+      "epoch": 3.1037974683544305,
+      "grad_norm": 0.3891177662806512,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.2315,
+      "step": 764
+    },
+    {
+      "epoch": 3.107848101265823,
+      "grad_norm": 0.38404751226449,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2543,
+      "step": 765
+    },
+    {
+      "epoch": 3.111898734177215,
+      "grad_norm": 0.3654752790334944,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.2486,
+      "step": 766
+    },
+    {
+      "epoch": 3.1159493670886076,
+      "grad_norm": 0.3596767866591725,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2616,
+      "step": 767
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.36738247235290833,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.2503,
+      "step": 768
+    },
+    {
+      "epoch": 3.124050632911392,
+      "grad_norm": 0.3585109938717279,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2422,
+      "step": 769
+    },
+    {
+      "epoch": 3.1281012658227847,
+      "grad_norm": 0.343220138050211,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2538,
+      "step": 770
+    },
+    {
+      "epoch": 3.132151898734177,
+      "grad_norm": 0.36535584880432215,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2596,
+      "step": 771
+    },
+    {
+      "epoch": 3.1362025316455697,
+      "grad_norm": 0.35077234778104593,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2702,
+      "step": 772
+    },
+    {
+      "epoch": 3.1402531645569622,
+      "grad_norm": 0.3545635100818043,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2335,
+      "step": 773
+    },
+    {
+      "epoch": 3.1443037974683543,
+      "grad_norm": 0.37017545865699053,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2391,
+      "step": 774
+    },
+    {
+      "epoch": 3.148354430379747,
+      "grad_norm": 0.350297914847873,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2721,
+      "step": 775
+    },
+    {
+      "epoch": 3.1524050632911393,
+      "grad_norm": 0.3788777540962781,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2606,
+      "step": 776
+    },
+    {
+      "epoch": 3.1564556962025314,
+      "grad_norm": 0.34355191336336405,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2733,
+      "step": 777
+    },
+    {
+      "epoch": 3.160506329113924,
+      "grad_norm": 0.3773252630420873,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2659,
+      "step": 778
+    },
+    {
+      "epoch": 3.1645569620253164,
+      "grad_norm": 0.3525973795616757,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2353,
+      "step": 779
+    },
+    {
+      "epoch": 3.168607594936709,
+      "grad_norm": 0.3696536363188408,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2657,
+      "step": 780
+    },
+    {
+      "epoch": 3.1726582278481015,
+      "grad_norm": 0.37656826563060775,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2629,
+      "step": 781
+    },
+    {
+      "epoch": 3.1767088607594935,
+      "grad_norm": 0.332911920412832,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2612,
+      "step": 782
+    },
+    {
+      "epoch": 3.180759493670886,
+      "grad_norm": 0.3561935501905511,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.245,
+      "step": 783
+    },
+    {
+      "epoch": 3.1848101265822786,
+      "grad_norm": 0.3606880079072412,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2775,
+      "step": 784
+    },
+    {
+      "epoch": 3.188860759493671,
+      "grad_norm": 0.3353549902842139,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2758,
+      "step": 785
+    },
+    {
+      "epoch": 3.192911392405063,
+      "grad_norm": 0.3404321278718524,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2583,
+      "step": 786
+    },
+    {
+      "epoch": 3.1969620253164557,
+      "grad_norm": 0.3589901885392696,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2594,
+      "step": 787
+    },
+    {
+      "epoch": 3.201012658227848,
+      "grad_norm": 0.3407841551979717,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2654,
+      "step": 788
+    },
+    {
+      "epoch": 3.2050632911392407,
+      "grad_norm": 0.3365511386483695,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.281,
+      "step": 789
+    },
+    {
+      "epoch": 3.209113924050633,
+      "grad_norm": 0.34439105714765517,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2762,
+      "step": 790
+    },
+    {
+      "epoch": 3.2131645569620253,
+      "grad_norm": 0.35030074568607933,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2517,
+      "step": 791
+    },
+    {
+      "epoch": 3.217215189873418,
+      "grad_norm": 0.3667370656508157,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2386,
+      "step": 792
+    },
+    {
+      "epoch": 3.2212658227848103,
+      "grad_norm": 0.35283337828642236,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2522,
+      "step": 793
+    },
+    {
+      "epoch": 3.2253164556962024,
+      "grad_norm": 0.3717623503834315,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.2511,
+      "step": 794
+    },
+    {
+      "epoch": 3.229367088607595,
+      "grad_norm": 0.3335904128474494,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2701,
+      "step": 795
+    },
+    {
+      "epoch": 3.2334177215189874,
+      "grad_norm": 0.34675034129323856,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2633,
+      "step": 796
+    },
+    {
+      "epoch": 3.23746835443038,
+      "grad_norm": 0.35956975892047405,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.2258,
+      "step": 797
+    },
+    {
+      "epoch": 3.241518987341772,
+      "grad_norm": 0.3389911086243913,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.2375,
+      "step": 798
+    },
+    {
+      "epoch": 3.2455696202531645,
+      "grad_norm": 0.34317369555841754,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2638,
+      "step": 799
+    },
+    {
+      "epoch": 3.249620253164557,
+      "grad_norm": 0.33284345884766414,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2546,
+      "step": 800
+    },
+    {
+      "epoch": 3.253670886075949,
+      "grad_norm": 0.3371336849207521,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2544,
+      "step": 801
+    },
+    {
+      "epoch": 3.2577215189873416,
+      "grad_norm": 0.36039249646454397,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2477,
+      "step": 802
+    },
+    {
+      "epoch": 3.261772151898734,
+      "grad_norm": 0.34452888331480713,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2514,
+      "step": 803
+    },
+    {
+      "epoch": 3.2658227848101267,
+      "grad_norm": 0.3348044285076572,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2573,
+      "step": 804
+    },
+    {
+      "epoch": 3.269873417721519,
+      "grad_norm": 0.34493029098901096,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2616,
+      "step": 805
+    },
+    {
+      "epoch": 3.2739240506329113,
+      "grad_norm": 0.35567406638422866,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2331,
+      "step": 806
+    },
+    {
+      "epoch": 3.2779746835443038,
+      "grad_norm": 0.3495471911271008,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.2303,
+      "step": 807
+    },
+    {
+      "epoch": 3.2820253164556963,
+      "grad_norm": 0.32448136782668674,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.2494,
+      "step": 808
+    },
+    {
+      "epoch": 3.286075949367089,
+      "grad_norm": 0.371579368336639,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2522,
+      "step": 809
+    },
+    {
+      "epoch": 3.290126582278481,
+      "grad_norm": 0.3465164340640093,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2329,
+      "step": 810
+    },
+    {
+      "epoch": 3.2941772151898734,
+      "grad_norm": 0.34644894015221306,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2755,
+      "step": 811
+    },
+    {
+      "epoch": 3.298227848101266,
+      "grad_norm": 0.36604184876432444,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.253,
+      "step": 812
+    },
+    {
+      "epoch": 3.3022784810126584,
+      "grad_norm": 0.3289990193434461,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.267,
+      "step": 813
+    },
+    {
+      "epoch": 3.3063291139240505,
+      "grad_norm": 0.35342635057824817,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2137,
+      "step": 814
+    },
+    {
+      "epoch": 3.310379746835443,
+      "grad_norm": 0.3619044124624644,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2414,
+      "step": 815
+    },
+    {
+      "epoch": 3.3144303797468355,
+      "grad_norm": 0.36276066234661164,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.2791,
+      "step": 816
+    },
+    {
+      "epoch": 3.318481012658228,
+      "grad_norm": 0.3374903778987933,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2738,
+      "step": 817
+    },
+    {
+      "epoch": 3.32253164556962,
+      "grad_norm": 0.3352450822203068,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2399,
+      "step": 818
+    },
+    {
+      "epoch": 3.3265822784810126,
+      "grad_norm": 0.34863189354003005,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2721,
+      "step": 819
+    },
+    {
+      "epoch": 3.330632911392405,
+      "grad_norm": 0.3417748049869419,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.241,
+      "step": 820
+    },
+    {
+      "epoch": 3.3346835443037977,
+      "grad_norm": 0.3316476412988616,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2808,
+      "step": 821
+    },
+    {
+      "epoch": 3.3387341772151897,
+      "grad_norm": 0.3413513034224004,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2461,
+      "step": 822
+    },
+    {
+      "epoch": 3.3427848101265822,
+      "grad_norm": 0.34499951000348583,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.2755,
+      "step": 823
+    },
+    {
+      "epoch": 3.3468354430379748,
+      "grad_norm": 0.3408025391849509,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2548,
+      "step": 824
+    },
+    {
+      "epoch": 3.3508860759493673,
+      "grad_norm": 0.3516115486725499,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2441,
+      "step": 825
+    },
+    {
+      "epoch": 3.3549367088607593,
+      "grad_norm": 0.3602732857100681,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2635,
+      "step": 826
+    },
+    {
+      "epoch": 3.358987341772152,
+      "grad_norm": 0.336780846419326,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2605,
+      "step": 827
+    },
+    {
+      "epoch": 3.3630379746835444,
+      "grad_norm": 0.3563267209627132,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2748,
+      "step": 828
+    },
+    {
+      "epoch": 3.367088607594937,
+      "grad_norm": 0.36469364585704217,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2478,
+      "step": 829
+    },
+    {
+      "epoch": 3.371139240506329,
+      "grad_norm": 0.3482251430414893,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2512,
+      "step": 830
+    },
+    {
+      "epoch": 3.3751898734177215,
+      "grad_norm": 0.3481830238473514,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.262,
+      "step": 831
+    },
+    {
+      "epoch": 3.379240506329114,
+      "grad_norm": 0.3794946727403097,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2554,
+      "step": 832
+    },
+    {
+      "epoch": 3.3832911392405065,
+      "grad_norm": 0.3591814868751582,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2704,
+      "step": 833
+    },
+    {
+      "epoch": 3.3873417721518986,
+      "grad_norm": 0.33189340579277027,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2479,
+      "step": 834
+    },
+    {
+      "epoch": 3.391392405063291,
+      "grad_norm": 0.3497983743885353,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.259,
+      "step": 835
+    },
+    {
+      "epoch": 3.3954430379746836,
+      "grad_norm": 0.3422686631035466,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2682,
+      "step": 836
+    },
+    {
+      "epoch": 3.399493670886076,
+      "grad_norm": 0.3256347374696348,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.2466,
+      "step": 837
+    },
+    {
+      "epoch": 3.403544303797468,
+      "grad_norm": 0.32718608174078245,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2599,
+      "step": 838
+    },
+    {
+      "epoch": 3.4075949367088607,
+      "grad_norm": 0.37936887826245447,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2532,
+      "step": 839
+    },
+    {
+      "epoch": 3.4116455696202532,
+      "grad_norm": 0.3391062527398742,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2699,
+      "step": 840
+    },
+    {
+      "epoch": 3.4156962025316457,
+      "grad_norm": 0.3217951145304219,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2635,
+      "step": 841
+    },
+    {
+      "epoch": 3.419746835443038,
+      "grad_norm": 0.3581342458454149,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.2574,
+      "step": 842
+    },
+    {
+      "epoch": 3.4237974683544303,
+      "grad_norm": 0.3342486362367731,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2633,
+      "step": 843
+    },
+    {
+      "epoch": 3.427848101265823,
+      "grad_norm": 0.3398896856218047,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2384,
+      "step": 844
+    },
+    {
+      "epoch": 3.4318987341772154,
+      "grad_norm": 0.3220721517939167,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.2479,
+      "step": 845
+    },
+    {
+      "epoch": 3.4359493670886074,
+      "grad_norm": 0.34589013405393876,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.2531,
+      "step": 846
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.32397291801079137,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2807,
+      "step": 847
+    },
+    {
+      "epoch": 3.4440506329113925,
+      "grad_norm": 0.34401841089591173,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.246,
+      "step": 848
+    },
+    {
+      "epoch": 3.448101265822785,
+      "grad_norm": 0.35190912723759094,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2472,
+      "step": 849
+    },
+    {
+      "epoch": 3.452151898734177,
+      "grad_norm": 0.34381549736498734,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2515,
+      "step": 850
+    },
+    {
+      "epoch": 3.4562025316455696,
+      "grad_norm": 0.33829604358049914,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.2517,
+      "step": 851
+    },
+    {
+      "epoch": 3.460253164556962,
+      "grad_norm": 0.33496490548863045,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.2442,
+      "step": 852
+    },
+    {
+      "epoch": 3.4643037974683546,
+      "grad_norm": 0.35096082040602544,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2702,
+      "step": 853
+    },
+    {
+      "epoch": 3.4683544303797467,
+      "grad_norm": 0.3227410551744003,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2598,
+      "step": 854
+    },
+    {
+      "epoch": 3.472405063291139,
+      "grad_norm": 0.35100622760897915,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2596,
+      "step": 855
+    },
+    {
+      "epoch": 3.4764556962025317,
+      "grad_norm": 0.35442512790304953,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2466,
+      "step": 856
+    },
+    {
+      "epoch": 3.480506329113924,
+      "grad_norm": 0.3470334630255758,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2469,
+      "step": 857
+    },
+    {
+      "epoch": 3.4845569620253163,
+      "grad_norm": 0.3574060913888215,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.26,
+      "step": 858
+    },
+    {
+      "epoch": 3.488607594936709,
+      "grad_norm": 0.354174339024617,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.2231,
+      "step": 859
+    },
+    {
+      "epoch": 3.4926582278481013,
+      "grad_norm": 0.4329089243386059,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.246,
+      "step": 860
+    },
+    {
+      "epoch": 3.496708860759494,
+      "grad_norm": 0.3515814445068424,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2519,
+      "step": 861
+    },
+    {
+      "epoch": 3.500759493670886,
+      "grad_norm": 0.38755587229275906,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2243,
+      "step": 862
+    },
+    {
+      "epoch": 3.5048101265822784,
+      "grad_norm": 0.3520589323849171,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.2652,
+      "step": 863
+    },
+    {
+      "epoch": 3.508860759493671,
+      "grad_norm": 0.3383905214571411,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2504,
+      "step": 864
+    },
+    {
+      "epoch": 3.5129113924050634,
+      "grad_norm": 0.33752633748477007,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2541,
+      "step": 865
+    },
+    {
+      "epoch": 3.516962025316456,
+      "grad_norm": 0.3373379525323659,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.259,
+      "step": 866
+    },
+    {
+      "epoch": 3.521012658227848,
+      "grad_norm": 0.3369148579818177,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2574,
+      "step": 867
+    },
+    {
+      "epoch": 3.5250632911392406,
+      "grad_norm": 0.360086316597193,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.2645,
+      "step": 868
+    },
+    {
+      "epoch": 3.529113924050633,
+      "grad_norm": 0.3299279777723768,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.2521,
+      "step": 869
+    },
+    {
+      "epoch": 3.533164556962025,
+      "grad_norm": 0.3439028022451909,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2717,
+      "step": 870
+    },
+    {
+      "epoch": 3.5372151898734177,
+      "grad_norm": 0.35300065897931,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.2309,
+      "step": 871
+    },
+    {
+      "epoch": 3.54126582278481,
+      "grad_norm": 0.33012201619033044,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2456,
+      "step": 872
+    },
+    {
+      "epoch": 3.5453164556962027,
+      "grad_norm": 0.3371876028061706,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2409,
+      "step": 873
+    },
+    {
+      "epoch": 3.549367088607595,
+      "grad_norm": 0.3373973655200395,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2535,
+      "step": 874
+    },
+    {
+      "epoch": 3.5534177215189873,
+      "grad_norm": 0.323468587343594,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2524,
+      "step": 875
+    },
+    {
+      "epoch": 3.55746835443038,
+      "grad_norm": 0.33043890146736005,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2578,
+      "step": 876
+    },
+    {
+      "epoch": 3.5615189873417723,
+      "grad_norm": 0.3481878781292326,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2586,
+      "step": 877
+    },
+    {
+      "epoch": 3.5655696202531644,
+      "grad_norm": 0.33590589866616666,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2564,
+      "step": 878
+    },
+    {
+      "epoch": 3.569620253164557,
+      "grad_norm": 0.33946869757299014,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2621,
+      "step": 879
+    },
+    {
+      "epoch": 3.5736708860759494,
+      "grad_norm": 0.35015244741305596,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2534,
+      "step": 880
+    },
+    {
+      "epoch": 3.577721518987342,
+      "grad_norm": 0.32637835273538207,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.25,
+      "step": 881
+    },
+    {
+      "epoch": 3.5817721518987344,
+      "grad_norm": 0.3405412931893513,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2767,
+      "step": 882
+    },
+    {
+      "epoch": 3.5858227848101265,
+      "grad_norm": 0.3393859522302317,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2641,
+      "step": 883
+    },
+    {
+      "epoch": 3.589873417721519,
+      "grad_norm": 0.3406274530942591,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2442,
+      "step": 884
+    },
+    {
+      "epoch": 3.5939240506329115,
+      "grad_norm": 0.34179262942135913,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2525,
+      "step": 885
+    },
+    {
+      "epoch": 3.5979746835443036,
+      "grad_norm": 0.32372541493564483,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2666,
+      "step": 886
+    },
+    {
+      "epoch": 3.602025316455696,
+      "grad_norm": 0.3143813285855106,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2644,
+      "step": 887
+    },
+    {
+      "epoch": 3.6060759493670886,
+      "grad_norm": 0.32405333622738913,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.245,
+      "step": 888
+    },
+    {
+      "epoch": 3.610126582278481,
+      "grad_norm": 0.3222211551275423,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.2468,
+      "step": 889
+    },
+    {
+      "epoch": 3.6141772151898737,
+      "grad_norm": 0.3265332559318164,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2668,
+      "step": 890
+    },
+    {
+      "epoch": 3.6182278481012657,
+      "grad_norm": 0.34969128520151743,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2689,
+      "step": 891
+    },
+    {
+      "epoch": 3.6222784810126583,
+      "grad_norm": 0.3415702724595714,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2296,
+      "step": 892
+    },
+    {
+      "epoch": 3.6263291139240508,
+      "grad_norm": 0.3282469292045241,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2653,
+      "step": 893
+    },
+    {
+      "epoch": 3.630379746835443,
+      "grad_norm": 0.33500671161821965,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2644,
+      "step": 894
+    },
+    {
+      "epoch": 3.6344303797468354,
+      "grad_norm": 0.3378829823951131,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.2616,
+      "step": 895
+    },
+    {
+      "epoch": 3.638481012658228,
+      "grad_norm": 0.31062059387359564,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.2489,
+      "step": 896
+    },
+    {
+      "epoch": 3.6425316455696204,
+      "grad_norm": 0.3308458661737303,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.2443,
+      "step": 897
+    },
+    {
+      "epoch": 3.646582278481013,
+      "grad_norm": 0.32796070736507194,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2626,
+      "step": 898
+    },
+    {
+      "epoch": 3.650632911392405,
+      "grad_norm": 0.32255246578583147,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2651,
+      "step": 899
+    },
+    {
+      "epoch": 3.6546835443037975,
+      "grad_norm": 0.3369453874319817,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2639,
+      "step": 900
+    },
+    {
+      "epoch": 3.65873417721519,
+      "grad_norm": 0.34612892134565687,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2644,
+      "step": 901
+    },
+    {
+      "epoch": 3.662784810126582,
+      "grad_norm": 0.3351630038636742,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2697,
+      "step": 902
+    },
+    {
+      "epoch": 3.6668354430379746,
+      "grad_norm": 3.7217693042149245,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.2747,
+      "step": 903
+    },
+    {
+      "epoch": 3.670886075949367,
+      "grad_norm": 0.3410248218039902,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2509,
+      "step": 904
+    },
+    {
+      "epoch": 3.6749367088607596,
+      "grad_norm": 0.3374580657916512,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.2708,
+      "step": 905
+    },
+    {
+      "epoch": 3.678987341772152,
+      "grad_norm": 0.3521821803723178,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.2394,
+      "step": 906
+    },
+    {
+      "epoch": 3.683037974683544,
+      "grad_norm": 0.3438771298864118,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.254,
+      "step": 907
+    },
+    {
+      "epoch": 3.6870886075949367,
+      "grad_norm": 0.6013219936682868,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.2533,
+      "step": 908
+    },
+    {
+      "epoch": 3.6911392405063292,
+      "grad_norm": 0.32971025036836044,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.25,
+      "step": 909
+    },
+    {
+      "epoch": 3.6951898734177213,
+      "grad_norm": 0.34011741923916905,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2588,
+      "step": 910
+    },
+    {
+      "epoch": 3.699240506329114,
+      "grad_norm": 0.3939804663422689,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2479,
+      "step": 911
+    },
+    {
+      "epoch": 3.7032911392405063,
+      "grad_norm": 3.2430797923751533,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.267,
+      "step": 912
+    },
+    {
+      "epoch": 3.707341772151899,
+      "grad_norm": 0.34510851751440824,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.2536,
+      "step": 913
+    },
+    {
+      "epoch": 3.7113924050632914,
+      "grad_norm": 0.3569642821333488,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2788,
+      "step": 914
+    },
+    {
+      "epoch": 3.7154430379746834,
+      "grad_norm": 0.32935189286076216,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2597,
+      "step": 915
+    },
+    {
+      "epoch": 3.719493670886076,
+      "grad_norm": 0.323682931428987,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2682,
+      "step": 916
+    },
+    {
+      "epoch": 3.7235443037974685,
+      "grad_norm": 0.32212301276243127,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2865,
+      "step": 917
+    },
+    {
+      "epoch": 3.7275949367088606,
+      "grad_norm": 0.35208444388605264,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2765,
+      "step": 918
+    },
+    {
+      "epoch": 3.731645569620253,
+      "grad_norm": 0.34943552647634935,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.262,
+      "step": 919
+    },
+    {
+      "epoch": 3.7356962025316456,
+      "grad_norm": 0.3355270124065066,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2586,
+      "step": 920
+    },
+    {
+      "epoch": 3.739746835443038,
+      "grad_norm": 0.3228249416067482,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2456,
+      "step": 921
+    },
+    {
+      "epoch": 3.7437974683544306,
+      "grad_norm": 0.35470230877224157,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2247,
+      "step": 922
+    },
+    {
+      "epoch": 3.7478481012658227,
+      "grad_norm": 0.348839481411762,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2579,
+      "step": 923
+    },
+    {
+      "epoch": 3.751898734177215,
+      "grad_norm": 2.5770204346272623,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.3377,
+      "step": 924
+    },
+    {
+      "epoch": 3.7559493670886077,
+      "grad_norm": 0.31299975263374985,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2819,
+      "step": 925
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.34423213830090876,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2362,
+      "step": 926
+    },
+    {
+      "epoch": 3.7640506329113923,
+      "grad_norm": 0.34019979174858067,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.2607,
+      "step": 927
+    },
+    {
+      "epoch": 3.768101265822785,
+      "grad_norm": 0.36372615358440563,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2406,
+      "step": 928
+    },
+    {
+      "epoch": 3.7721518987341773,
+      "grad_norm": 0.3291608392067456,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.2498,
+      "step": 929
+    },
+    {
+      "epoch": 3.77620253164557,
+      "grad_norm": 0.33038640037389394,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.242,
+      "step": 930
+    },
+    {
+      "epoch": 3.780253164556962,
+      "grad_norm": 0.3426346257838029,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.2362,
+      "step": 931
+    },
+    {
+      "epoch": 3.7843037974683544,
+      "grad_norm": 0.32929270548056533,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2503,
+      "step": 932
+    },
+    {
+      "epoch": 3.788354430379747,
+      "grad_norm": 0.33064209174668013,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2518,
+      "step": 933
+    },
+    {
+      "epoch": 3.792405063291139,
+      "grad_norm": 0.3522344938447979,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.2311,
+      "step": 934
+    },
+    {
+      "epoch": 3.7964556962025315,
+      "grad_norm": 0.30791659451764386,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2616,
+      "step": 935
+    },
+    {
+      "epoch": 3.800506329113924,
+      "grad_norm": 0.32156946941791936,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2518,
+      "step": 936
+    },
+    {
+      "epoch": 3.8045569620253166,
+      "grad_norm": 0.31308001460852963,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2735,
+      "step": 937
+    },
+    {
+      "epoch": 3.808607594936709,
+      "grad_norm": 0.3507205811116923,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2639,
+      "step": 938
+    },
+    {
+      "epoch": 3.812658227848101,
+      "grad_norm": 0.31838139387059616,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2614,
+      "step": 939
+    },
+    {
+      "epoch": 3.8167088607594937,
+      "grad_norm": 0.32193314065628575,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2714,
+      "step": 940
+    },
+    {
+      "epoch": 3.820759493670886,
+      "grad_norm": 0.32818215914039606,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.2392,
+      "step": 941
+    },
+    {
+      "epoch": 3.8248101265822783,
+      "grad_norm": 0.3232861605396304,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.2601,
+      "step": 942
+    },
+    {
+      "epoch": 3.8288607594936708,
+      "grad_norm": 0.3340616624739708,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2618,
+      "step": 943
+    },
+    {
+      "epoch": 3.8329113924050633,
+      "grad_norm": 0.3383621937161854,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2703,
+      "step": 944
+    },
+    {
+      "epoch": 3.836962025316456,
+      "grad_norm": 0.3167963126083905,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2613,
+      "step": 945
+    },
+    {
+      "epoch": 3.8410126582278483,
+      "grad_norm": 0.3315825930949492,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.24,
+      "step": 946
+    },
+    {
+      "epoch": 3.8450632911392404,
+      "grad_norm": 0.3302414562667117,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.2572,
+      "step": 947
+    },
+    {
+      "epoch": 3.849113924050633,
+      "grad_norm": 0.31098382637088484,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2645,
+      "step": 948
+    },
+    {
+      "epoch": 3.8531645569620254,
+      "grad_norm": 0.3277882914732334,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.2502,
+      "step": 949
+    },
+    {
+      "epoch": 3.8572151898734175,
+      "grad_norm": 0.3189579519256551,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.2485,
+      "step": 950
+    },
+    {
+      "epoch": 3.86126582278481,
+      "grad_norm": 0.3132274804453835,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2584,
+      "step": 951
+    },
+    {
+      "epoch": 3.8653164556962025,
+      "grad_norm": 0.3260235880382673,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2512,
+      "step": 952
+    },
+    {
+      "epoch": 3.869367088607595,
+      "grad_norm": 0.330721189802237,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.2555,
+      "step": 953
+    },
+    {
+      "epoch": 3.8734177215189876,
+      "grad_norm": 0.33337973980968644,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.2574,
+      "step": 954
+    },
+    {
+      "epoch": 3.8774683544303796,
+      "grad_norm": 0.3419724239822643,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2454,
+      "step": 955
+    },
+    {
+      "epoch": 3.881518987341772,
+      "grad_norm": 0.33353020688928675,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2396,
+      "step": 956
+    },
+    {
+      "epoch": 3.8855696202531647,
+      "grad_norm": 0.3400368721604568,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.2545,
+      "step": 957
+    },
+    {
+      "epoch": 3.8896202531645567,
+      "grad_norm": 0.3226360685043356,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.2405,
+      "step": 958
+    },
+    {
+      "epoch": 3.8936708860759492,
+      "grad_norm": 0.33615559375692555,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2665,
+      "step": 959
+    },
+    {
+      "epoch": 3.8977215189873418,
+      "grad_norm": 0.34321726951357745,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2575,
+      "step": 960
+    },
+    {
+      "epoch": 3.9017721518987343,
+      "grad_norm": 0.31737951628537947,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2612,
+      "step": 961
+    },
+    {
+      "epoch": 3.905822784810127,
+      "grad_norm": 0.3201957779869363,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2503,
+      "step": 962
+    },
+    {
+      "epoch": 3.909873417721519,
+      "grad_norm": 0.3252940434318872,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2527,
+      "step": 963
+    },
+    {
+      "epoch": 3.9139240506329114,
+      "grad_norm": 0.3296594494925174,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2577,
+      "step": 964
+    },
+    {
+      "epoch": 3.917974683544304,
+      "grad_norm": 0.31993680641589883,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2426,
+      "step": 965
+    },
+    {
+      "epoch": 3.922025316455696,
+      "grad_norm": 0.3224039310466981,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.2376,
+      "step": 966
+    },
+    {
+      "epoch": 3.9260759493670885,
+      "grad_norm": 0.3148253878899576,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.2413,
+      "step": 967
+    },
+    {
+      "epoch": 3.930126582278481,
+      "grad_norm": 0.32847899541370795,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.2519,
+      "step": 968
+    },
+    {
+      "epoch": 3.9341772151898735,
+      "grad_norm": 0.3589815988329803,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.2453,
+      "step": 969
+    },
+    {
+      "epoch": 3.938227848101266,
+      "grad_norm": 0.32326830675985374,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2599,
+      "step": 970
+    },
+    {
+      "epoch": 3.942278481012658,
+      "grad_norm": 0.33701910501749005,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2625,
+      "step": 971
+    },
+    {
+      "epoch": 3.9463291139240506,
+      "grad_norm": 0.3258689405225757,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.2487,
+      "step": 972
+    },
+    {
+      "epoch": 3.950379746835443,
+      "grad_norm": 0.3348389855440537,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.237,
+      "step": 973
+    },
+    {
+      "epoch": 3.954430379746835,
+      "grad_norm": 0.32264886612043686,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2717,
+      "step": 974
+    },
+    {
+      "epoch": 3.9584810126582277,
+      "grad_norm": 0.31526303879803114,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2398,
+      "step": 975
+    },
+    {
+      "epoch": 3.9625316455696202,
+      "grad_norm": 0.32362146152895976,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.2459,
+      "step": 976
+    },
+    {
+      "epoch": 3.9665822784810127,
+      "grad_norm": 0.32395817878733973,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2471,
+      "step": 977
+    },
+    {
+      "epoch": 3.9706329113924053,
+      "grad_norm": 0.3263980259783227,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.2832,
+      "step": 978
+    },
+    {
+      "epoch": 3.9746835443037973,
+      "grad_norm": 0.31974756880187233,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2469,
+      "step": 979
+    },
+    {
+      "epoch": 3.97873417721519,
+      "grad_norm": 0.32976879160314715,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2849,
+      "step": 980
+    },
+    {
+      "epoch": 3.9827848101265824,
+      "grad_norm": 0.3219929261146135,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2607,
+      "step": 981
+    },
+    {
+      "epoch": 3.9868354430379744,
+      "grad_norm": 0.3385151154652636,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2329,
+      "step": 982
+    },
+    {
+      "epoch": 3.990886075949367,
+      "grad_norm": 0.31434981916697036,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2533,
+      "step": 983
+    },
+    {
+      "epoch": 3.9949367088607595,
+      "grad_norm": 0.31751669772506885,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2408,
+      "step": 984
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.3507687688243209,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.2233,
+      "step": 985
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.41029957276757034,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1937,
+      "step": 986
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.3891243263844287,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.1929,
+      "step": 987
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.2638205258165373,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1868,
+      "step": 988
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.31584745551077587,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1909,
+      "step": 989
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.42201552309772794,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1864,
+      "step": 990
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.4658707289722327,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1836,
+      "step": 991
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.40646952112203927,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1987,
+      "step": 992
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.31852570586391765,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1876,
+      "step": 993
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.2958117016267485,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1859,
+      "step": 994
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.32702343567025066,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1946,
+      "step": 995
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.3334630317994951,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.2131,
+      "step": 996
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.32432189078536483,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.185,
+      "step": 997
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.29999211247434043,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1905,
+      "step": 998
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.2735455310698422,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.2006,
+      "step": 999
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.28629293755020874,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1949,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.5213998007516057,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1989,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.3192848172787333,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1791,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.32112547115484796,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.2208,
+      "step": 1003
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.289449344006025,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1971,
+      "step": 1004
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.2736498288303326,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.2008,
+      "step": 1005
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.28625847193506554,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1815,
+      "step": 1006
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.2826039597135102,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1954,
+      "step": 1007
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.2960585501329785,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1925,
+      "step": 1008
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.2814627759400984,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1979,
+      "step": 1009
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.2603042864936993,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1823,
+      "step": 1010
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.28395336838877167,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1972,
+      "step": 1011
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.2856959062686879,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1763,
+      "step": 1012
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.2758588929774589,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1898,
+      "step": 1013
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.27995370588872415,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1785,
+      "step": 1014
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.2958631653782012,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.2219,
+      "step": 1015
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.2658794409503579,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1866,
+      "step": 1016
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.27435293582323345,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.2027,
+      "step": 1017
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.27727226734287047,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1829,
+      "step": 1018
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.26151851185444736,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1741,
+      "step": 1019
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.2569856257389745,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1657,
+      "step": 1020
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.26806869486856716,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.2053,
+      "step": 1021
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.27293429745700987,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.2026,
+      "step": 1022
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.2563150281455751,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1745,
+      "step": 1023
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.26975752575717704,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.2031,
+      "step": 1024
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.2656975651578527,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1853,
+      "step": 1025
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.2686734151613357,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1897,
+      "step": 1026
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.2772681009744244,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1803,
+      "step": 1027
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.2688309144421408,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.196,
+      "step": 1028
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.24982828656899486,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1872,
+      "step": 1029
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.2597288142380288,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.2178,
+      "step": 1030
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.26926464754690593,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1729,
+      "step": 1031
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.24566374468470803,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1973,
+      "step": 1032
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.2664707328207486,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1817,
+      "step": 1033
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.2657098223914025,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1843,
+      "step": 1034
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.2655555029355679,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.197,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.25124079989647125,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.187,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.2674249796885765,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1878,
+      "step": 1037
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.2604722598246572,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1938,
+      "step": 1038
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.27029788918511044,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1947,
+      "step": 1039
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.2618686024589062,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.2077,
+      "step": 1040
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.2706653956838585,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.17,
+      "step": 1041
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.2559864281975217,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.198,
+      "step": 1042
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.2499987270006716,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1803,
+      "step": 1043
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.2630917387201679,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1937,
+      "step": 1044
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.25306518077970525,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1821,
+      "step": 1045
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.2628260685103581,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.2078,
+      "step": 1046
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.26104401982725006,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1865,
+      "step": 1047
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.26743513712075845,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1985,
+      "step": 1048
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.25753524637928904,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.2087,
+      "step": 1049
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.2671668725346025,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.2027,
+      "step": 1050
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.2583393850646206,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1748,
+      "step": 1051
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.26213149838689975,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1805,
+      "step": 1052
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.2662403953472665,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1821,
+      "step": 1053
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.26640883243463037,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.193,
+      "step": 1054
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.2701236928633287,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1818,
+      "step": 1055
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.2709840762189434,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.2046,
+      "step": 1056
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.257669628713909,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1626,
+      "step": 1057
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.2672629845531242,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.2011,
+      "step": 1058
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.2599394512235911,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1948,
+      "step": 1059
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.27313818458552463,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.2098,
+      "step": 1060
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.2589450298793582,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1824,
+      "step": 1061
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.2565658635506086,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1699,
+      "step": 1062
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.253909782149014,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1952,
+      "step": 1063
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.2592703781344579,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1859,
+      "step": 1064
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.25630866319250345,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.2119,
+      "step": 1065
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.2643159690117824,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.2041,
+      "step": 1066
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.2771801740229088,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.2201,
+      "step": 1067
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.2720455272141513,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1877,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.2547175849076973,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.1923,
+      "step": 1069
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.27486776486126935,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1754,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.2638952115557761,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1639,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.26133451836134397,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1753,
+      "step": 1072
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.25098850033221315,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1875,
+      "step": 1073
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.2663294926695448,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1983,
+      "step": 1074
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.2553073295065841,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1965,
+      "step": 1075
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.27141949566269885,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1935,
+      "step": 1076
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.24928830079240705,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1826,
+      "step": 1077
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.25096807072475114,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1967,
+      "step": 1078
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.2591208575111559,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.2152,
+      "step": 1079
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.2600722794424445,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.1545,
+      "step": 1080
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.2569824436053389,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1968,
+      "step": 1081
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.26186678173092076,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1783,
+      "step": 1082
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.2551508694118239,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1675,
+      "step": 1083
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.26851450630981627,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1794,
+      "step": 1084
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.2548729241855416,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.221,
+      "step": 1085
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.2730078425717428,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.2008,
+      "step": 1086
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.2663212013551046,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1841,
+      "step": 1087
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.26655165144975024,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.2033,
+      "step": 1088
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.26329631797879194,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1841,
+      "step": 1089
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.2516679222793993,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.186,
+      "step": 1090
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.25546224317993604,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1973,
+      "step": 1091
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.261862165971091,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1773,
+      "step": 1092
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.24637549782869145,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1681,
+      "step": 1093
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.2640514155705031,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.2105,
+      "step": 1094
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.27375877199144305,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1672,
+      "step": 1095
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.2635721407515113,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1787,
+      "step": 1096
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.2638433936166717,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.2157,
+      "step": 1097
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.26037870674204144,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.2016,
+      "step": 1098
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.2727244185386274,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1977,
+      "step": 1099
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.25891217978658765,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.2036,
+      "step": 1100
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.2522929733118263,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.1805,
+      "step": 1101
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.2552832308784635,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1797,
+      "step": 1102
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.24865781944848736,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1921,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.25585379162508765,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1719,
+      "step": 1104
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.25961164816791377,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1906,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.2725933479458815,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1849,
+      "step": 1106
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.2576437878916549,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1941,
+      "step": 1107
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.2470374747973991,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.1756,
+      "step": 1108
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.26474075257368407,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.2165,
+      "step": 1109
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.27149131932156045,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.2092,
+      "step": 1110
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.2588916849021906,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.2003,
+      "step": 1111
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.2635589239367197,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1997,
+      "step": 1112
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.25760003918434915,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1935,
+      "step": 1113
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.26476518481824524,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1816,
+      "step": 1114
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.25845897864973827,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1969,
+      "step": 1115
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.2624078074666615,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.2148,
+      "step": 1116
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.2562799619254189,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1743,
+      "step": 1117
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.24947625004765422,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1985,
+      "step": 1118
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.26459719014103306,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.167,
+      "step": 1119
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.26368937495051026,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1926,
+      "step": 1120
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.26061566933138497,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1993,
+      "step": 1121
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.2676424091039148,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1936,
+      "step": 1122
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.2555727262458985,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1646,
+      "step": 1123
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.2605459175195548,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1835,
+      "step": 1124
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.2549301119955827,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1787,
+      "step": 1125
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.2594895756140398,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.1756,
+      "step": 1126
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.2606635300603076,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1588,
+      "step": 1127
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.24706792304457212,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1823,
+      "step": 1128
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.2710713392474118,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.2109,
+      "step": 1129
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.2695778702823276,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.2099,
+      "step": 1130
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.2612179978997489,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1964,
+      "step": 1131
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.26102125872645,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1922,
+      "step": 1132
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.3228108561472618,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1901,
+      "step": 1133
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.26975635167664347,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.192,
+      "step": 1134
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.2534036017396075,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1946,
+      "step": 1135
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.2621867854838689,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1833,
+      "step": 1136
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.25564971053864793,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1844,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.2643208507578772,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1873,
+      "step": 1138
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.25504577529213895,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1907,
+      "step": 1139
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.2596725273222212,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1863,
+      "step": 1140
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.2554706725098294,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1671,
+      "step": 1141
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.24745129530331864,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1804,
+      "step": 1142
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.2710783703542428,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.184,
+      "step": 1143
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.25653172090352744,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1984,
+      "step": 1144
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.2626844656026387,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1895,
+      "step": 1145
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.2645270758462679,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.2188,
+      "step": 1146
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.26150338566843334,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1945,
+      "step": 1147
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.2544277583241751,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1926,
+      "step": 1148
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.2557675872660858,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1914,
+      "step": 1149
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.2572351694951515,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1873,
+      "step": 1150
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.2520383384516057,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1842,
+      "step": 1151
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.2607555965423969,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.2053,
+      "step": 1152
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.261191140391286,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1656,
+      "step": 1153
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.2561473276658763,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1802,
+      "step": 1154
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.24868523854483005,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.2078,
+      "step": 1155
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.25126189923195097,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1929,
+      "step": 1156
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.25180163516536813,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.2097,
+      "step": 1157
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.2604152875397546,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.1646,
+      "step": 1158
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.24499267157080276,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.176,
+      "step": 1159
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.25361812796784533,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.192,
+      "step": 1160
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.26414471318277727,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.213,
+      "step": 1161
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.25638486158700313,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.2054,
+      "step": 1162
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.2623943108088503,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.187,
+      "step": 1163
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.2641564647315532,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.202,
+      "step": 1164
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.25946206614240686,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.1855,
+      "step": 1165
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.2675371020798864,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1893,
+      "step": 1166
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.2704681757232546,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1709,
+      "step": 1167
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.24964169267326913,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1867,
+      "step": 1168
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.26598272523842303,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.2048,
+      "step": 1169
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.24231873822724279,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.2013,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.26798232958384716,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1858,
+      "step": 1171
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.2520497973893246,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1908,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.2630303107910168,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1908,
+      "step": 1173
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.27280594464713676,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.176,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.26373217194261483,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1944,
+      "step": 1175
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.2639553106723838,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1864,
+      "step": 1176
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.2563499420923002,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1825,
+      "step": 1177
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.2552263361535951,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1947,
+      "step": 1178
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.2600586015354575,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1849,
+      "step": 1179
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.25547529506365285,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1638,
+      "step": 1180
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.2485509722956513,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1757,
+      "step": 1181
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.2581035972617905,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.2102,
+      "step": 1182
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.25399762327076747,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.1865,
+      "step": 1183
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.26473112773372653,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1547,
+      "step": 1184
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.2650656520026236,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.2173,
+      "step": 1185
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.2563996122500032,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1999,
+      "step": 1186
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.2648694368676355,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1942,
+      "step": 1187
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.26725905439188774,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1778,
+      "step": 1188
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.25789880300711643,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1952,
+      "step": 1189
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.2555092092679847,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1955,
+      "step": 1190
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.25773930390112904,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1742,
+      "step": 1191
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.25398443292248024,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1932,
+      "step": 1192
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.26603743587980855,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1923,
+      "step": 1193
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.2581808115997865,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1953,
+      "step": 1194
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.25615857821660254,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.19,
+      "step": 1195
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.2647186083184127,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1904,
+      "step": 1196
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.25718000568448646,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1772,
+      "step": 1197
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.26994160347661206,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1876,
+      "step": 1198
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.2526599075686546,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1636,
+      "step": 1199
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.2570336900359519,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1991,
+      "step": 1200
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.2645493115419525,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.215,
+      "step": 1201
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.2571381986380164,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1846,
+      "step": 1202
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.26045257433489055,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1957,
+      "step": 1203
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.2540611848079979,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.2007,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.270950812317858,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.2125,
+      "step": 1205
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.25620072716625697,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1866,
+      "step": 1206
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.2662570009332228,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1862,
+      "step": 1207
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.2507269363685567,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.2139,
+      "step": 1208
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.2651518892992907,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.207,
+      "step": 1209
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.2601441018654287,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1822,
+      "step": 1210
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.26151855898667714,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1983,
+      "step": 1211
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.27511374374326986,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1864,
+      "step": 1212
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.26598334240847815,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1822,
+      "step": 1213
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.2558693808912695,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.1969,
+      "step": 1214
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.25546970821408593,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1923,
+      "step": 1215
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.25942678014516635,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1783,
+      "step": 1216
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.26120334239022813,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1737,
+      "step": 1217
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.26653092242765614,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.2176,
+      "step": 1218
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.25272431922128347,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1873,
+      "step": 1219
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.2672501487791775,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1925,
+      "step": 1220
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.25974992151652443,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1793,
+      "step": 1221
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.264138015254269,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.196,
+      "step": 1222
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.24920088401737464,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.213,
+      "step": 1223
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.27691750623544564,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1735,
+      "step": 1224
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.252108228098281,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.2212,
+      "step": 1225
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.26184080803765647,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1502,
+      "step": 1226
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.2503553710925597,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1654,
+      "step": 1227
+    },
+    {
+      "epoch": 4.98632911392405,
+      "grad_norm": 0.26689403437552556,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.2169,
+      "step": 1228
+    },
+    {
+      "epoch": 4.990379746835443,
+      "grad_norm": 0.26304445408558136,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.2022,
+      "step": 1229
+    },
+    {
+      "epoch": 4.994430379746835,
+      "grad_norm": 0.26805757562300464,
+      "learning_rate": 0.0,
+      "loss": 0.2092,
+      "step": 1230
+    },
+    {
+      "epoch": 4.994430379746835,
+      "step": 1230,
+      "total_flos": 2.4701266730271375e+18,
+      "train_loss": 0.42356404370166423,
+      "train_runtime": 42145.5969,
+      "train_samples_per_second": 3.747,
+      "train_steps_per_second": 0.029
+    }
+  ],
+  "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": 2.4701266730271375e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}