diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.988354430379747,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 6.738137424432949,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.9668,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.704480183106894,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 0.9715,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.649852168234561,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.9555,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.586384339989349,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.965,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.009424153534045,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.9484,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.851554206902821,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.9727,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.220419241345155,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.8843,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 3.977587799836835,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.8994,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.9500264035888115,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.8499,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.74529493974799,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.8951,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.2388546719948255,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.8459,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.605664435522847,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.881,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.42657753494484,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.8385,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.1547475177727335,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.8076,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.774279488756445,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.8323,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.148514458280813,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.7746,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.089569715105327,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.8118,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 1.8128797101669885,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.7438,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.3929441683392823,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.7345,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.3435509653560478,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.717,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.1518944179113615,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.7285,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.037800881018344,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.6999,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.0406570785553142,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.715,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.1054781909533145,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.6935,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.0808832018750802,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.6748,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.8120786833078844,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.6848,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.8022044098431059,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.6677,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.8202570499156832,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.6676,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.9122335465172019,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.6656,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8128208709238237,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.6648,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.6857485662725991,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.68,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.6743229086446373,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.6533,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8179942317888155,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.6633,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.7923267760833331,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.6475,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.580942452996161,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.6614,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.7405932422874135,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.6458,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.6935200212548842,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.6487,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6151757390966838,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.6444,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.593954913303251,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.6422,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6049660429457115,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.6232,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.6202598276442614,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6469,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.5252543366810118,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.6273,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.5990662753459397,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.6385,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.5555298700793329,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.6335,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.5397270311831047,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.6529,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.547448981025657,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.6366,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.5465929255940167,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.6169,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.5968501964145153,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.6544,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.5649668464239493,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.6249,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.5862494161295905,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.6553,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.5585290861876406,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.6364,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.5656095503736616,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.6427,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.7090104270042087,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.6337,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.5220301052114755,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.6381,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.6632143725654617,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.6222,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.5650459842153049,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.6087,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.6353465575731497,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.6412,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.5298639288059516,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.6029,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.6408073698377098,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.6314,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.6663725445327395,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6222,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.619667014775608,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6148,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.6522160972568782,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.6211,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.6588528929532699,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6293,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.6085291038558257,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.6069,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7361879054546313,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.6299,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.5114808132070221,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.5976,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.6556791255027764,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.5968,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.6120042447557519,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.6091,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.6364528311090728,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.595,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.545332925458142,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6195,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.6446808684081377,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.605,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.526482946232368,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.5987,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.6843137029671412,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.61,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.5313501957668688,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.5983,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.6229037916787954,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.6045,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.7587650172852438,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.6049,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.5235923200776036,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.6048,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.6116443836101014,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.6048,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.57070110102058,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.6319,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.6622733709006261,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.5958,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.6061409289778379,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6077,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.5608443801246532,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6189,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.6407790595134794,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.6122,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.5013268060202937,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.6147,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.673211258614305,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.6224,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.6500998849186943,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.6219,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.6881382375633367,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.5877,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.6538399166277753,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6127,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.6673693527332356,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.6064,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.6435865622653156,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6227,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.6061801625589748,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.5998,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.7004898619528622,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.6088,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.6329491215213612,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.5816,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.6291953434236165,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.6135,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.5646486186711192,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6254,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.5588405585306142,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.6181,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.6090569622349508,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.6241,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.6592102214276453,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.6159,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.5840309246985822,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.5796,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.7467970959605532,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.5922,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.8194736760940756,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.6133,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.0144578977657448,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.6216,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.6775048187699853,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.6182,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.6113068331939774,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.6052,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.8456062724729609,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6112,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.0825740906837211,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.6077,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.6541651225804224,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.609,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.8519757827412257,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.5937,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.9211472979410629,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6461,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.6137722565175325,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.5939,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.656857990479076,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.6096,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.7852666926770623,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.6143,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.6580716052028316,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6154,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.5947979326827149,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6113,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.7770493177926533,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.6013,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.7024616318584296,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.6151,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.6346037172305418,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6277,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.5744007160048461,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6188,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.6725047890493829,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.5887,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.6852570006283522,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.5932,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.6710149270584966,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.6097,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.6636621367931815,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6143,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.710755017071099,
+      "learning_rate": 4e-05,
+      "loss": 0.6269,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.6275540909964789,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.6113,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.6901366809135769,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.6073,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.7346298849301377,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.6011,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.9800665767350758,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.5966,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.5884804993063429,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.5944,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.7508414192755807,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.6028,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.6871060770932548,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.5926,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.5916655495006354,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.5751,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.7826157784275275,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.6117,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.6642512515630379,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6148,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.7617489932819406,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.5945,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.6189617390006067,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.6049,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.6031115478369041,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.5958,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.7201004473334499,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.5927,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.5379927300985372,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.5965,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.6563389837085589,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.6243,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.6753157024054424,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6146,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.6877767099181129,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6296,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.5752605559701961,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.5883,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.8374236313404673,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.6307,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.9151232703658515,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.604,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.6536304616997746,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.5889,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.770032852859139,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.5813,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.6242371711219942,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.6197,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.7500252297814209,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.6094,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.6493090127303927,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.6057,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.6453297509560996,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6181,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.5307546520149223,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.5879,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.7430731696636991,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6191,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.5322885618585647,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.5736,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.6884472258792911,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.5956,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.5263316895719893,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.5843,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.5908621809303349,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.6174,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.5307657018899118,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.6233,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5581620129178141,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.6286,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.6852498281327223,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.5937,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.4888967534305476,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.6164,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.6899908678833494,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.607,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.7245805460275944,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.6324,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.5788193119252569,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.6169,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.48773449638781474,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.5661,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.5580113881533004,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6074,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.5349589030169938,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.6152,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.49175258866167015,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.5573,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.5079252854704773,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.5933,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.498785004422796,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.622,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.4755082809002897,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.6062,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.5348528734662116,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6117,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.5547833146001612,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6146,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.5885419275139901,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.5892,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.4995308120697694,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.6069,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.5491403011308585,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.632,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.5286876157633286,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.5971,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.5345013170992413,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.6115,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.5568082783726575,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6034,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.5254289375831895,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.5741,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.5545690505018157,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.581,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.5051485704407143,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.59,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.5277848667754391,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.6101,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.5084676229680698,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.6131,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.60984490654569,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6105,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.5591147215011563,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6062,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.5010226640977942,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.5988,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.48361277255168245,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.5968,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.5186851571985629,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.6003,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.5478305860906268,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.6009,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.4956522694001798,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6016,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6015328330076325,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.5872,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.584511705426499,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.5964,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.574769932895746,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.6301,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.7597426887703222,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6443,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.5802595370489321,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.5915,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.6741671606982922,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6108,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.6779656992456008,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.6172,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.46610432599374396,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.602,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.5860503271918165,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.582,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.47741833706033293,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.5862,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.5916555935168274,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.6036,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.5030102582206382,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.5816,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.5102149560705204,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.5922,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.5148338451792978,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.5757,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.46217041373978324,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.5835,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.4976071018127545,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.5806,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.4732137221621373,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.5848,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.5081718159431623,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.5901,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.6346184408951029,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6291,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.5018102279464927,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.6144,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.501579611506324,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.5979,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5629385199921488,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6334,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.58033578996432,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.5984,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.44027330863922365,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.5732,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.594747973179513,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.5971,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.4853521947125498,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.6289,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.44834791621734155,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.585,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.4589296718851956,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.5856,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.5304748376249275,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.6053,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.465515468585629,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.5771,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.5187602764975598,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.6214,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.5157609367399433,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.5738,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.4947264924900894,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6205,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.5657917813269095,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.5876,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.5197020692561348,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.5653,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.48195701413116193,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.5868,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.5024574769925578,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.5717,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.4567493541715026,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.5657,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.5450795603433926,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.6141,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.45535282712309183,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.5804,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.49484589406920376,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.5788,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.554836307601814,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.5863,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.4213365910438387,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.5718,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.5864085685294956,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.6015,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.45369661093829344,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.5977,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.5420358730818624,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.6019,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4556418754035017,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.574,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.5059351259331573,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.6248,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.44513647511365634,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.6076,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.4854613196415651,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.5855,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.42021487057433615,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.5791,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.4765056042972292,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.5641,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.4677621916597309,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.584,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.4897677782905455,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.5817,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.54823939819082,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.5918,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.46854387807150577,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.5586,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.5633863557720329,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.5219,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.5188609809682287,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.4431,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.5547705682261194,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.4446,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.7445849040323272,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.4646,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.714040220560507,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.4564,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.75956906899657,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.447,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.5331684410653749,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.4363,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.8099532384256963,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.4531,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.768728346767181,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.4798,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.7586718198939353,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.4393,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.5074153304577635,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.4413,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.8203666231165918,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.443,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.5644041730740964,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.4438,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.6421064685223954,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.4586,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.6920875272272764,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.4329,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.5470370442846337,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.4407,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.6373835166651246,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.443,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.7033504585266319,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.4427,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.4856772162082972,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.4465,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.7306608438137587,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.4424,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.5840008775376311,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.44,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.5263219910705594,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.4573,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.6172437077847818,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.4459,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.48889895571295977,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.4604,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.5352382594845302,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.4537,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.6234904958325614,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.4307,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.45936000688763257,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.4453,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.6956214833856849,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.4577,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.5619224841540308,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.4453,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.4564171928605993,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.4287,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.502058528731428,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.4099,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.4310338912861414,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.4334,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.49592284721689966,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.4435,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.474191544758235,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.4626,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.4478609087781942,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.4482,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.4966344590039205,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.4534,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.502151197830293,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.4377,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.5390038357940732,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.4536,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.48948266370451937,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.4254,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5793884819174924,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.4661,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.5630897459045852,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.4675,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.49622049641019533,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.4491,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.5108146999686072,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.4332,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.5825240746219692,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.4564,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.48447503726621205,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.467,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.5193482826349367,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.4557,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.4853480064064051,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.4684,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.5151775463593634,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.4309,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.42011065690869204,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.4415,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.5144041584733956,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.4429,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.4636271891458896,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.4385,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.50328932470835,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.432,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.4656628257324611,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.4531,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.44976410243264225,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.4463,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.5198217282151804,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.4573,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.4568922256249722,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.4603,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.5631822047739654,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.4388,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.44137016357884423,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.4807,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.5229933917790297,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.4381,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.47522144522921916,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.4568,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.5124681465336983,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.433,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.4400194876893733,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.4575,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.511099663546347,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.4463,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.4794562174128532,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.4628,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.5003092145524097,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.4499,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.5559576602021349,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.4506,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.5379625062679073,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.4578,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.5268942142243712,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.4596,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.6285710810473106,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.4648,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.5072507376263062,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.4616,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.6338686214760938,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.4542,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.6082728042768922,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.4469,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.47337758511090083,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.446,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.5787152705955461,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.4547,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.4414592539103977,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.4574,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.5300335852391126,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.4682,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.4550622068242876,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.4563,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.5786825903558119,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.4416,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.5270953149436474,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.4439,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.486133170243594,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.45,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.5122278579874243,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.4508,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.5008354188235864,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.431,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.4593932528816152,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.4335,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.4311811629427631,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.4428,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.4697963768853903,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.4599,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.47566450638451785,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.4615,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.43231101368824704,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.4619,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.5406729548176163,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.4355,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.49014672381817576,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.4629,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.573910962527949,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.4599,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.4852929494727567,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.4543,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.49502879101106434,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.4629,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.4714402271438486,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.4251,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.4800007691991582,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.4596,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.4629135179733209,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.4623,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.45008604944508745,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.4714,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.4850469389371198,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.4497,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.5692926276664114,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.4678,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.5939729042005103,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.4489,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.49412739463752553,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.4702,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.5306905007129946,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.4542,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.4707680382867705,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.4585,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.4877057568878019,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.4622,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.4762358792917313,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.4602,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.5109987134418685,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.4535,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.491271194085416,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.4429,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.4967172590213714,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.4437,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.5327104559400961,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.4652,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.49324808623388516,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.4644,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.6047695627817109,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.4664,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.42114739613079893,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.435,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.506209229858558,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.4432,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.5238171396310222,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.4602,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.49871844135254173,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.4653,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.5309438670903763,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.4479,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.4539900027066011,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.4593,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.4767470613263526,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.4595,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.4269770880856819,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.4601,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.45107684957366373,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.4334,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.4636441270482069,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.4537,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.43440612282715996,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.449,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.4517059782322685,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.454,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.4429472843587689,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.452,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.46940435265045627,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.4519,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.4623462098056322,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.4598,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.4765572294797722,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.4691,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.4680130805601337,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.4609,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.5114083808862041,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.4531,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.4756799865874779,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.4824,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.5243241690660272,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.4598,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.49529062534122525,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.4636,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.4439672604767949,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.4542,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.506578381776089,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.4648,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.4295681418092351,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.4382,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.4663286363124973,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.4628,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.44597237772341775,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.4665,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.4855517491083876,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.4562,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.4384263959094648,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.4526,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.41392434443482745,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.4676,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.5195043906093884,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.4605,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.41103984129677196,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.4453,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.5209008372111822,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.4811,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.4215880796155846,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.4547,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.4653220020236176,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.4569,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.4638887660948843,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.4433,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.43326948880282945,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.4662,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.4535814544804793,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.4539,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.45924505641389524,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.4614,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.45541308514596684,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.474,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.4428179209505052,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.4541,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.43482329180374724,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.4546,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.44876333541280805,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.4596,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.43019113540814125,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.4434,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.44856499894666524,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.4435,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.43820750413743054,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.4503,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.47391853497725966,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.4338,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.43003980024059846,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.4648,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.4874708733022434,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.4488,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.43812083582408085,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.4679,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.43930099331225564,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.452,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.47214406664877,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.4468,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.43919344919148295,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.4496,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.44603543789737365,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.459,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.4404855779437762,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.4459,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.42436801220420267,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.4626,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.44749319863306847,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.4569,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.4952571555558035,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.4491,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.46748168797632306,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.4655,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.5727876468946981,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.465,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.4347978718167872,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.4592,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.4435845503073326,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.455,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.4158224580776832,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.4472,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.4478745541443723,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.4494,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.4129900649103557,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.4415,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.4831878934303196,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.4621,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.42428741570828865,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.4583,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.4570411194195891,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.4411,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.42553195113640907,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.4595,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.45539003711904885,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.4583,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.4351079798004603,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.4576,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.4219817705375046,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.4557,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.5272735002301088,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.4408,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.4737049647975755,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.4589,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.49473794488911804,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.4439,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.4643682283016952,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.4704,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.4357168056810496,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.4618,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.4491691603154195,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.4591,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.4681970573954517,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.4456,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.4394205977094379,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.4663,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.43439412551988177,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.454,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.4477708968259312,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.4395,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.43845935059967445,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.4523,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.446495679262483,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.4748,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.4661263167468186,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.455,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.42408841809510944,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.4641,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.45567877440852483,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.464,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.4468376555201501,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.4609,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.4718183132512513,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.438,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.42337369145963594,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.4377,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.47856531600688323,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.4579,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.4360984135190318,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.4471,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.4442290277389359,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.4444,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.4131124126315819,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.4647,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.44953520073882774,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.4685,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.40941125401912676,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.472,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.4735092993184471,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.452,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.4247818089797212,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.4585,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.48078805828342847,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.4688,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.44447642491894224,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.465,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.4714152855182472,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.4549,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.45894251730168717,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.4558,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.43950829711740313,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.4556,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.4889459407202916,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.4405,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.4393466663256847,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.44,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.5086927007474396,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.454,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.47419912012984017,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.4734,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.4309235410855153,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.4628,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.4575207430224126,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.4539,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.475994213750644,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.4588,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.46076457077095717,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.442,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.497514459401866,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.4511,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.42541323729098723,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.4593,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.4720109625021154,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.4606,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.4446298839503674,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.4458,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.47428217959295255,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.4617,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.40940714090223096,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.4336,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.5233991800769894,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.4497,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.42471419218443895,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.4607,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.4966401604773226,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.4868,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.4878366198257101,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.4623,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.4660042280347889,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.4389,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.4910785438015133,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.4576,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.4448322951417005,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.4701,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.46291023850748136,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.4617,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.46636419055788425,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.4679,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.4841778524316119,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.4672,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.49723537633003567,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.4583,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.4599579440129499,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.4393,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.5177015884252069,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.4562,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.44692843897407725,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.4557,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.44270875705277035,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.4575,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.49249455920515695,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.4499,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.42727915940760763,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.4418,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.46514600813093215,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.442,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.42257263128823197,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.4449,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.4389879095470197,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4503,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.4058092089418687,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.4477,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 0.887807355398491,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.2962,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.5435283196155324,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.2562,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 1.2945345590490893,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.2893,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.4997376817195279,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.2928,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.5761779473394475,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.2957,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.5649309049483469,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.2835,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.4895935777559413,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.2741,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.49822439691553094,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.2798,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.5264037679747599,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.2806,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.5538597683460899,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.2828,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.4520760213597788,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.2803,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.516746700536684,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.2893,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.4831625416909708,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.2727,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.495872060604196,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.2598,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.49543171634415145,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.2692,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.5334201360289501,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.2765,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.4580726546548506,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3007,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.4907436560405742,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.2896,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.4891987363387425,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.2675,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.4352949162515945,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.2835,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.47206776155641983,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.278,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.447534846472364,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.2593,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.4635585789510846,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2864,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.43967947059215085,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.2755,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.4796508649232373,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.2776,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.4353981468378608,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.2804,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.4789607000400417,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.2768,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.4205319131862905,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.2662,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.5079200179667785,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.2799,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.4239292194872771,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.2668,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.474292982568178,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.2687,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.42285142080403637,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.2857,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.4361514628524347,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.2662,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.40444027007368194,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.267,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.448971202886092,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.2811,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.4324429195721114,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.2942,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.43164374881848705,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.2742,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.4438764784103316,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.2861,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.40688452884485815,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.2709,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.44569675704774303,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.2768,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.3861782204161773,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.2787,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.4614079771907007,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.2745,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.40003271299013593,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.2725,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.4488778262477719,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.2739,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.3982221741590615,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.2605,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.43445724618303316,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.2655,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.4270201394130976,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.2655,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.44471928583652215,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.2786,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.4249257805754625,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.2592,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.43147509211332147,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.2782,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.4344075266075954,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.2826,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.40429897542755927,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.2879,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.4398660392082459,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.2651,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.42978347317414284,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.2778,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.41666283964125467,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.2913,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.4148622352965865,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.275,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.434317473424833,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.2839,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.3911093220341875,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.306,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.4207321630245316,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.2755,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.41168108013654636,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.2905,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.4343241085247585,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.279,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.40338271789968805,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.2728,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.3975827052872322,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.2649,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.4105831250425762,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.2783,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.4061507831480528,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.2704,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.4063355465961293,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.2804,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.3843731689702759,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.2809,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.3977963423453418,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.2757,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.3918142735094154,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.2928,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.3962155221596916,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.2837,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.4198199336226785,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.2729,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.37623339777093245,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.2711,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.4381870071350417,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.2839,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.4026126196127328,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.2681,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.4406291843410084,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.2808,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.42866376564842745,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3008,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.38986558838555857,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.2769,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.41367913952458146,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.2732,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.4084019728968326,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.2694,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.39601891601335404,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.2715,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.3983730413847545,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.2842,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.40981757086008463,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.2949,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.41513635069567334,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.2823,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.4071400289965067,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.2846,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.4020843757812165,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.2795,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.4084385651124559,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.2791,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.4063170384560954,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3018,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.40595527735198056,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.2975,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.4068420895100194,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.2824,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.3853087531642031,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.2804,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.3862868276131623,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.2768,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.3962106060774979,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.2948,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.3917501597643043,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.2718,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.4018043890813915,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.2763,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.44110410789354415,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.2783,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.4145846990560126,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.2815,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.43812613400215145,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.2669,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.41430714374310484,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.27,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.42689922384406487,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.2796,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.47287978915914153,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.2844,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.39844538116392225,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3153,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.44483754302711725,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.2961,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.40001449905800607,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.2728,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.46357947665813315,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.2864,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.40902071556265907,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.2975,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.42490240484739783,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.2998,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.4226332265672028,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.2829,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.45133413038379155,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.2854,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.3951015144180344,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3031,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.4017401045660559,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.2812,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.42618206127904945,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.2972,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.3892992134970232,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.2982,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.46274140174745043,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.2912,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.38598921624590915,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.28,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.4103318324244433,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.2936,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.4162004139712585,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.2586,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.4167120438057965,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.2795,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.40361362159179043,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.2995,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.4118890214590572,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.2904,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.4144631673094275,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.2825,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.4638498124463217,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.2918,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.3997371056362086,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2729,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.47453298338556965,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.2929,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.38923737716869256,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.2788,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.4587734861427746,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.2728,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.40861894469379634,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.2684,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.4476299326617825,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.2847,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.43434532124331626,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.2773,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.41039608281124634,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.2789,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.402319200013175,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.2865,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.3879383062460831,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.2767,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.42328792463880227,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.2984,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.40294053633781485,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.2767,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.44530068731917544,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.2903,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.38326878980211054,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.2844,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.4066151883154577,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.2749,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.38338624299936197,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.2802,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.40052197974224457,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.2761,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.3868991397452875,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.2679,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.41442787181187635,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.2786,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.392637797774676,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.2988,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.3793664769959631,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.2809,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.3894588294010504,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2572,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.39916040109465367,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.2735,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.37138000654219194,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.2755,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.4310158435345107,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.2706,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.38475782773851575,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.2902,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.39215878171418994,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.2794,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.41005505355094884,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3018,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.3941953647251406,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.2645,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.38857431500571293,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.2871,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.40219787966246956,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.277,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.3895987835520362,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.2892,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.3723526514938595,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.2767,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.3974183760005829,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.2866,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.3831490839076185,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.2792,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.4161678280406475,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.2765,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.38505534951354214,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.2608,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.38919161854555384,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.2649,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.42551472791220846,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.2847,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.4493637319018401,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.2769,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.4129104895868485,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.2724,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.40501399280962863,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.288,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.38387580492089346,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.2783,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.421050220704885,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.2747,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.3935307811182046,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.2839,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.3994248683435304,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.2901,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.3879840049172871,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.2823,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.3974592604705227,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.2678,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.37657223130076417,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.2847,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 2.9293731698471692,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.2844,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.47184564972594567,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.2898,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.40562123530519484,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.2784,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.4089940174101844,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.2874,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.40615526052550305,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.2705,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.422078862714908,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.2826,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.41766632298099254,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.2988,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.41059839433785983,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.2762,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.4260039787619452,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.2837,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.40015145441369426,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.2848,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.4022405804821003,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.2806,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.4034152944951102,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.2871,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.4052117657519094,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.2721,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.38686698410747433,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.2921,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.4073552163549865,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.2774,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.40554192125804167,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.2965,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.39528770175166766,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.2592,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.40591364526839707,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.282,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.3873380794916649,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.2956,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.38998840618842867,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.2759,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.3892743812431699,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.2795,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.39374787503012065,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.2593,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.400191457265138,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.305,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.4016887107925414,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.2937,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.3894657499273909,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.2996,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.38447511860708417,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.2995,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.3757686004032544,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.2929,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.39580971699523954,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.298,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.3970232196307396,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.2905,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.386656078497716,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.275,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.37774544325016146,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.278,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.38603814451996676,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.2867,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.3968609585349975,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.302,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.3954445007228064,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.2754,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.38081920404110664,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.2884,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.3858584404747633,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.2654,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.3704639260666378,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.2852,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.38921933086655336,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.293,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.4270909038432567,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.2713,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.37250649068322894,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.2816,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.393740302005169,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.2835,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.4029789900794336,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.2684,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.37516533685767506,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.2692,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.37860778912657733,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.2801,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.39035614553581255,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.2781,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.4036138935014821,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.2897,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.37088265394242964,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.2825,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.3773492253576865,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.2933,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.38082138763442513,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.2832,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.3714812482429228,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.2878,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.3691780633215788,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.274,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.39326847549024857,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.2854,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.38636281166057007,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.2903,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.37145568560121217,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.2789,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.35609624961833913,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.2885,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.3839931163218711,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.2762,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.38413383321256717,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.2826,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.3683790196953599,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.2848,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.3795333717422509,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.2731,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.38027522884057113,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.2681,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.379915889072295,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.2825,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.38284617087613293,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.2602,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.3849443605937237,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.2735,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.38118196122394704,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.2812,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.3911637241302771,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.2905,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.3734866838442479,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.2751,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.38065735015857904,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.2675,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.41387259745290444,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.2926,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.38004420843882886,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.2669,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.373740722614481,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.2875,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.4005778730185085,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.2736,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.3684684178058191,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.2849,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.3653152320574722,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.2745,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.38147755155978824,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.282,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.3943466596169522,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2928,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.4062818156700276,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.2811,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.41702483695113596,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.2144,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.5208230254665859,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.1801,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.3581313985880267,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.165,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.4395398217335517,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1508,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.6624088223982547,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.1735,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.531586629136385,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1501,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.3829936707883009,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1696,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.45912826387171035,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1518,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.4671641838971554,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.165,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.376370827741618,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.1434,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.36084305316384363,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1574,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.39203365833931164,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1577,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.3895488379905902,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1466,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.39846688322965884,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1354,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.37721865741203525,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1545,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.33324899064536356,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1451,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.3492066408451327,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1542,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.3566816034374583,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1659,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.36508175328894676,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1527,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.35450374059136003,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1479,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.3665451683738164,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1504,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.3631537460742381,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1466,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.3701272364903994,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1451,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.3613756280656335,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1529,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.32742420712071185,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1464,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.3657674991022707,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1498,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.34449910061927125,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1489,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.3396554564168939,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1571,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.34144718574050364,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.176,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.34567960736705444,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1533,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.3385740125866981,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1499,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.36147132953308375,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1586,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.342823907727766,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1743,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.32999857819382683,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.1547,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.34168623507442575,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1505,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.35162269802183044,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1701,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.3128185817111032,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1643,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.3291473167272917,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1528,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.35125276570926045,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1662,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.33032720776636004,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1704,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.3459081544103809,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1525,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.3288639062250872,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1635,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.3317516004324739,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1703,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.3273875864944427,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1497,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.3491289268720269,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1638,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.33641481812870866,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1479,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.3228939815536499,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1751,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.3459761724298092,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.1657,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.3378081730403833,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1559,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.336956879468339,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1512,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.3319089717093352,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1475,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.3345569474740725,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.1682,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.33226703565172655,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1436,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.33204477479837075,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1236,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.3330547868654958,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1597,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.3359415582835304,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1449,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.34016798388268593,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.16,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.3348829326436335,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1546,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.3269851180389334,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1562,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.3295694107777284,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1475,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.330153436574161,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.164,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.3397556052969577,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.156,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.34211593836858256,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1659,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.3361348619631758,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.1471,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.34025704966440734,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.1451,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.3314681357863452,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1638,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.3376688597135208,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1659,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.33279953737677526,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.159,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.33258251834062164,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1586,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.3374241901443898,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1451,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.3422319264813296,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1502,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.33697301273370056,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1723,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.34709230649898043,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1525,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.3363221221090753,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1571,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.33253467564376354,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.135,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.3338429085744522,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1457,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.3221354782426449,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1439,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.321031992060786,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.137,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.33659120538072446,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1552,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.3325370776614276,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1696,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.318054327157221,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1453,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.3303501279466663,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.16,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.31659947960267604,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1499,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.3357945158114857,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1627,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.3308894408670392,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1382,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.32553244646722096,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1387,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.3426506925985158,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1561,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.33875721495857963,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1753,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.320096444917792,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.153,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.33222589494369476,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1527,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.331452361636876,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1827,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.33889888847752886,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1541,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.3222295648401658,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1587,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.3283827918110762,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1586,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.3388736823556234,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1672,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.33088376978421774,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.1545,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.33031506063457244,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1515,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.3417501633893918,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.1508,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.33611963169746134,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1567,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.31536759569321376,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1399,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.3385571975399883,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1647,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.35084062134264155,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1645,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.32118200208703873,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1447,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.3396076851798396,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1507,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.3437163897158861,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1512,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.334879775170315,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1418,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.3434324463333136,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1733,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.34752907368080965,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1565,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.33751208937620725,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.1504,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.3262943167641179,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1639,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.33191939685435606,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1536,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.33321545501274996,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1569,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.32499327190906185,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1534,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.33607073008638333,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1383,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.33439427204269273,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1481,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.3292194991473772,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1481,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.330039328460945,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1558,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.3547037671563254,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1567,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3428076328044179,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1614,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.3339264421847442,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1646,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.325535128895608,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1508,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.3355441093607341,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1746,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.3240077482768786,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1561,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.32403119758287646,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.163,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.3278880482428015,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1634,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.3516761266449727,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.146,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.33236098379618795,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.1469,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.32010312342187974,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.165,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.3253472028441089,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1528,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.3318605262115,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1478,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.3205790089521689,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1522,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.3211776311642773,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.139,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.3333626958028667,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1674,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.3139921919849614,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1437,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.33375320784571105,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1514,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.31954960123964576,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1487,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.3168513083366419,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.1472,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.3231969531272366,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1645,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.3201601685187134,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1516,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.3107517839234179,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1615,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.3240755904321847,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1463,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.3295288960355797,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1416,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.31673586555832733,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.15,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.33069209784879433,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.163,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.3219206775185867,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.161,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.3367685139414499,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.138,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.3338478628572284,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1559,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.3081018718195932,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1606,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.30363955042671853,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1471,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.32851319387053624,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1394,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.3253973951059937,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1539,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.32816819757843785,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.144,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.33142570255432513,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1467,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.32069601682572074,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1375,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.3248769631677263,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1624,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.3216932026652119,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1452,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.312089221923854,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1555,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.3277758855510551,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1558,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.3204895099558436,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1444,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.32196846909904664,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.1437,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.30685216053761266,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1532,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.32052275874136643,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1532,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.32966000638634874,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1593,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.31254563272090186,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1563,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.338348294882547,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1644,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.3253443105130793,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1595,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.3181426528494649,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1533,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.31763838970381586,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1561,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.32784107378183885,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1433,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.3204888814889805,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1495,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.3301524658013571,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1506,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.31575519419103204,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1537,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.32384239668693826,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1537,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.32161019476628494,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1542,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.32459087598351566,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1437,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.3238804086153072,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1459,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.3203992819021694,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1604,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.3178472180850627,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1578,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.3162742737926381,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1662,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.3144971514621422,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.1514,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.32776354350884246,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1591,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.32047903777566944,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1527,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.32638772307789227,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1493,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.3275383714288657,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1798,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.3112463269111321,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1616,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.31714327343502824,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1665,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.3141587620518263,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1515,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.3128344332285383,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1541,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.3231478675072037,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1541,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.32020053655977526,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1519,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.31593866499636036,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1452,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.3141193399401488,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1423,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.3181111353998613,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1609,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.3167461981948986,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1588,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.3175358983107682,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1384,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.31372041253256017,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.139,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.31848945027985176,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1376,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.31283354485917075,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1465,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.3284666732895116,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1391,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.321425371066553,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1644,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.3304505917292473,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1572,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.31069336593544844,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1404,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.31267142102348927,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1525,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.32689286569992754,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1556,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.32070661270642914,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.145,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.31703398092540014,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1608,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.31260159488328415,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1466,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.31539490471805615,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1547,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.31866834781677983,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1813,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.3044800991395673,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1463,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.3184302692361387,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1592,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.31928593291005253,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1632,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.3213322427858275,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.1592,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.3176953671619879,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1492,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.31488920246171764,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1446,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.31424209352110233,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1505,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.33108918791350317,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1359,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.31382631553110146,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1392,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.3172499210075426,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1282,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.4368791980179305,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1714,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.33007870936232886,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.153,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.31298268643301963,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1677,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.31329103759132,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1407,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.3192456742993008,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1369,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.30567346404327145,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1637,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.31701501065130744,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1435,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.3229064150851624,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1681,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.3168013589789412,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1466,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.31613237368175484,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1642,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.3100148205301194,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1581,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.3230934568058146,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1563,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.31467976260399083,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1588,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.31520797468882866,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1536,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.3119702134055112,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1369,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.3124377485837646,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.163,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.3155885332944191,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1619,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.30883246170672224,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.144,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.3146703999924596,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1422,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.33193658340558335,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1577,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.31992902281256563,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1553,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.31208880613017953,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1635,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.3064798583350837,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1558,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.3112614227984009,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1439,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.3096568728715481,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1423,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.32387575104652144,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1534,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.3127665221321124,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1412,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.3202222984314029,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1508,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.3926121997639457,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.0976,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.37909297675979675,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.0912,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.3182557104140122,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.0981,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.26486359485542693,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.0852,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.23556833426366727,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.0965,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.28710320472581974,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1004,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.3236154303718752,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.0906,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 1.4139435948670624,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.0775,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.3771864718142506,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.0898,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.3534639766508332,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.0957,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.3123023407261607,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.0891,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.28691265632673113,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1022,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.2580827683848673,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.0907,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.27862534676773915,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.0945,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.3149722944418918,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1014,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.3049779436681834,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.0949,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.28827132705363523,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.0898,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.2690793326775318,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.0981,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.23827820688182857,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.0872,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.2506531687328481,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1004,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.2705327199048512,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1011,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.27655059132281024,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.0936,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.26333406888431193,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.0804,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.26818861090663104,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.0918,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.2706837852779368,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.0947,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.2545855188261536,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.0953,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.2508179323640254,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.0875,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.2449526304921578,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.0982,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.233999844659075,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.0914,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.2623536952193745,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1045,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.2528562417230574,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.0982,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.2490462964483932,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.0981,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.23875068889368903,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.0957,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.24055936913684228,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.0933,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.23815149948042125,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.0774,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.24524698487562038,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.0909,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.2444982655358449,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.0819,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.25500753622090344,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1046,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.2618212143328462,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.0903,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.2458037396856661,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.093,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.24045577570669296,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.0916,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.24630914107101254,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.0962,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.23154935904439297,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.0873,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.23572274703578164,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.0872,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.23645936955050662,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.0924,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.24495488304948712,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1053,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.24397026424860574,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.0916,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.24618389894754203,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.0923,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.24518674467791307,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1032,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.2489009641252913,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1125,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.2402523279331498,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0825,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.23448870037701422,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.0896,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.2509842665568552,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.0999,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.24583205543106165,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.0905,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.2503118877931947,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.1109,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.2523486374166944,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.0834,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.25839494333535545,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.0982,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.2460562028350464,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1054,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.2390205179368803,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.0875,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.24053982232460516,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.0962,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.24951029091680613,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.0986,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.23863360939151623,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.0872,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.24270919633471966,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.0841,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.24498674633607812,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.0911,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.23676134056713014,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.087,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.2350906300248434,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.0885,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.24751333118027302,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.0976,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.24318917553805863,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.0879,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.24269390414228154,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1025,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.24164860351532744,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.1029,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.2453938374534618,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.0862,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.24735470046859806,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1043,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.24149144482682816,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.0929,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.234514885149423,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.0936,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.2513573739515705,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.0805,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.2382438541434963,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.0827,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.23883629821632754,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.0998,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.23038164805155517,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.077,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.23994549150399913,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.0906,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.23891438226193046,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.0918,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.24598341905040336,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1004,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.23063758486326735,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.0963,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.2399081849762779,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0885,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.24753789052571615,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.095,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.22900902915673513,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.0849,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.24036605736396544,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.09,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.2341128510875137,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.0894,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.23528518705635199,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0949,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.24923236956610167,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1047,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.2485414028168334,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1023,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.2479133624001341,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.09,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.23718143536609718,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.0845,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.2461060389302043,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.0971,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.2449348416600689,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.0875,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.23631571377662802,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1079,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.23988175887163707,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1056,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.2526315445282431,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1136,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.23746575943538656,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.0807,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.243475986068808,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.0756,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.23114415799129845,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.0757,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.24304268161618928,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.0847,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.23661339812190532,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.0935,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.24725777551951889,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1059,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.24249230223350732,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.0779,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.24340670288404195,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.105,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.2360006458177422,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.0977,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.24719289582547463,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.0895,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.2509000814554369,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1097,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.2406876512567959,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.0898,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.23060857309254582,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.0982,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.24576025680043662,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1044,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.2429286451631526,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.0983,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.24672809562448694,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.0908,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.2306309081910957,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.082,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.2381492433677817,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.0963,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.23637571800609902,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.096,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.23725033237894486,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.0981,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.24286902073925826,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1023,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.23527353686764307,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.096,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.24974072210415374,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1048,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.24081716084979632,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.0794,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.2389770005666609,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.0911,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.24929851803454167,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.0883,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.23709835351081548,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.0907,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.2395260774510407,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1066,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.2344646426471566,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.0884,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.2514847213181512,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.102,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.23576725743127733,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.092,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.2457410348709192,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1073,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.24623484550436014,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.086,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.2354860525868563,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.0805,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.24363497131103673,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.0939,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.2386347309996161,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.0909,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.23632878675900199,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1029,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.24531153302153935,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.0876,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.2402508111195689,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1007,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.23377146949886696,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.0862,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.24109016517397894,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1034,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.24542835722160722,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.0936,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.2489790567655161,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0809,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.2356683732580991,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.0983,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.2403048034498668,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.089,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.23808270065144865,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.097,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.25250159702819763,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.0865,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.24247673582560555,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.0812,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.23873280320709347,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0966,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.2355711319735643,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.0834,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.23986467257914115,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.0926,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.23213116972910225,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.0777,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.23516427807930454,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.0819,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.23088893586153994,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.087,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.23717208713222387,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.0764,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.24629286276165896,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.0937,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.23788116471896478,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1047,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.2363084094475843,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.0915,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.24533543316075768,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1005,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.24963139307440327,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1157,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.231915939184196,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.0977,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.23474494020459968,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.0812,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.23536417494713993,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.0981,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.23469862954370474,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.0897,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.2429676804619465,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.0985,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.2520980845714133,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1005,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.24695910195994164,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.0868,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.24247486525302908,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.0946,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.24522169736681962,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.0998,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.24909905734278656,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.0916,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.2283555856515814,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.0936,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.2394617819503974,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1051,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.23958315640300226,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.0881,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.23159134835906633,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.0942,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.233086408667481,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.0903,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.23134844591274784,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.0969,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.23892345891711947,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.0806,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.24425268882727924,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.098,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.23516578280397074,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.0944,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.2400701806810615,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.0969,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.2490906017836151,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1063,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.24624768621456938,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.0958,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.23730844346421232,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.0926,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.2447494296520216,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.096,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.22674748144847953,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.0851,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.23847833202571408,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.0945,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.23747025827648227,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.0834,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.24639585342977083,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.0865,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.24586120247683355,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.0934,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.23951108071149524,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.0958,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.2382858682720842,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0868,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.23857792812585896,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.0862,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.2409937593775192,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1017,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.24261859910330835,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1084,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.2394370486763488,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1088,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.24005227289891273,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.0846,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.24647140410187718,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.0929,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.2371708834097907,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.0775,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.2282177884604912,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.0844,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.23904138057202207,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.0893,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.2310262058949595,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.0819,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.24130721306501762,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.0935,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.2373989628109118,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.098,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.2373539828823543,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.0968,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.235732914536266,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.0864,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.24726603028069832,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1015,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.2387853842863171,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.0963,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.24401095131115463,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.084,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.24505271602909545,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1009,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.24268169491993416,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.0919,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.2467564044148994,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.0978,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.2489190604347198,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.097,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.23492753904787372,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1016,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.2432005253900947,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.0887,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.2317578197230689,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.0953,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.23455059114895876,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.0804,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.2390985135068817,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.0978,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.23919941828106348,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.0911,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.23391112062589797,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1127,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.22919313979560194,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.0856,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.24875766470740943,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1025,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.22208355238151883,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.0714,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.22974034774144683,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.0841,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.2329595437174183,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.0878,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.23694945983596744,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.0956,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.24015048862765845,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.0971,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.2322074929859014,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.097,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.24492300364321398,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.0889,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.23804301418162865,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.0921,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.24146883135570507,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.0938,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.233449658511918,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.0997,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.23853227463956378,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.0953,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.24462936291638604,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.095,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.2404871367602787,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.0879,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.2405192596087071,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.0932,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.2396638305399567,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.0965,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.23894960468793555,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.0912,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.2302405955285471,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.0752,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.23676700452246272,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.0982,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.22699830015791672,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.0854,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.25054501175180294,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.0943,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.24335299488879078,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.0992,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.24124604726442955,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.0821,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.24464883938017368,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1069,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.2392521111889513,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.0896,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.23607901347230847,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1039,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.24572560157276674,
+      "learning_rate": 0.0,
+      "loss": 0.0905,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 2.0151031491020718e+18,
+      "train_loss": 0.3231248075399942,
+      "train_runtime": 59615.8773,
+      "train_samples_per_second": 2.65,
+      "train_steps_per_second": 0.021
+    }
+  ],
+  "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": 2.0151031491020718e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}