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": 7.190901404481192,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.1692,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 7.051225638402173,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.1442,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.959607679118453,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.1135,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.692895434364896,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.1069,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.397078554675219,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.1049,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 6.574571057049284,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.1323,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.020295253441259,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.0449,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.6177388981952525,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.0436,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.499832899413368,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.9866,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.282490535353663,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9979,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.1324561936519553,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.0051,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.740506429279618,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9647,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 5.3350448217325095,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.0202,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 5.095096718919127,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.9594,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 4.721768939186621,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9656,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 3.3470449831162106,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.9179,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 3.0561612789391575,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.9366,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.7049915815157695,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.9135,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.2567805448057054,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8938,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.9737960498610374,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.8827,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.716880813414956,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8558,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.587599210286518,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.8344,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.7169036911608293,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8494,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.6664797897391332,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.8411,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.4331342518785164,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.8324,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.2079182277448255,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.8579,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.3179998929818766,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.8339,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.1500168698539697,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8394,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.9786854559880325,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8292,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 1.0505359999578043,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.8196,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 1.097415513547258,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.8026,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.9099145245589982,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.8039,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8822321683062831,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.8111,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.849027651961397,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.8142,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.8118166679329141,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.813,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.7260233924649556,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.818,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.823289339733132,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.7911,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6916524900454462,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.779,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6972094724044967,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.7978,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7519259091590217,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.785,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.677406835663151,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8178,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.6593565905383925,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.7712,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.7332676009163879,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.7861,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.7098514833794023,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.7579,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7016921081841316,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.7892,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6549242340177437,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.7813,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.6448090305801469,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.7405,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.6393999473612035,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.749,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.694833909282302,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.767,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.628034351690571,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7748,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7175115765450881,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7945,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.5820676425766123,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7512,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.636423018028593,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.7537,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6316378908182841,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7561,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.6552895402753788,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.7601,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.6937232969003874,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.7484,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.7037602766098012,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.7627,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8356397736722753,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.7767,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7380245978798398,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7746,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.8162769512868107,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.751,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.7316342705939795,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.765,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.8665544083806537,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7497,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.7196705311152518,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.7342,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.9398944164955052,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.7702,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7649893506255979,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.7659,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.706320339432766,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.7569,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.8769122964925999,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.7604,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.764834049223063,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.7562,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8074870677465491,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.7742,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8954714031237119,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.7565,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.8360386243065755,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7557,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.7249811828725226,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.7569,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.8158866236178446,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.7391,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.6863321748086537,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7549,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.663681300294841,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.7153,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.8992429805174783,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.777,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.6431840163457149,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7271,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.7651668651668866,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.7379,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8415808905258515,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.7881,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.7987284432599308,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7834,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.7392772784096134,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.7486,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.9124091739092063,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7541,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.817398644933799,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7305,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.8022667393875829,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7298,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.7013469219053348,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.739,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.8355351589730851,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7232,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.6651600765571432,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7496,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.8527881535927727,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.7434,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.6543554877202244,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.7303,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.7926686720576096,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.7207,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.783000357371692,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7598,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.8934341831487459,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.7461,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.7282186920022763,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.7541,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.9794319804325968,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7532,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.7774355025499959,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.7363,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.8382649593457463,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.7765,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.7426610591276028,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7544,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.8287339690918574,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.7641,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.7656695649043026,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.7245,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.7881197448714423,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7175,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.780186845771214,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.7359,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.7759538714714287,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.7142,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.8383885491416494,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7368,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.8033478824413708,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7605,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.7852851671336001,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7066,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.9235258773468378,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7429,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.667827271045595,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.7394,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.7470680672737354,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.73,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.870406579298031,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.7314,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.8014796793937949,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.7487,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.7325547697968101,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7638,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.7797211355036209,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7248,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.7625503985071023,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7727,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.9940826526589976,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.7369,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.7391119951273213,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.7097,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.8053575061637791,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.7286,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.1035244754479745,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.7421,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.0837075567566241,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.7092,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.8683751097507507,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7623,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.9070774147484988,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.7406,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.840137647434597,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7503,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.8777777318651654,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.7134,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.7943247481246558,
+      "learning_rate": 4e-05,
+      "loss": 0.707,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.7360216467472152,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7377,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.8899041748528635,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7189,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.0250649745392326,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7411,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.9796067595578145,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7528,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.8948607384212809,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7311,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.6561779816520469,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.756,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.7410174051331848,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.7312,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.739443763692555,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.7361,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.8739602192902958,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7282,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.6097888609874863,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.73,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.6511250874502622,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.722,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.6717804998972988,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7115,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.8978474010316462,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.7507,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.7652634950930033,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.7379,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.8341938405239581,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.7308,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.0130391915262815,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.7072,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.6388914275038264,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7502,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.7443454362771561,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.7616,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.9992115877949723,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7602,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.8918016769511364,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.7582,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.8276473020165002,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.7333,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.6821333924790406,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7605,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.8996435250099992,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.766,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.7639859267844576,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7122,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.726229797304396,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.726,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.7581627483688994,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.7616,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.771728654478235,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.7291,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.6167650669043562,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7245,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.7073903244613691,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.7668,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.6588688057653092,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.749,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.6799756217090517,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.7335,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.6065243948342005,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.7222,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.6500712744605657,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.7249,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.7357653397833449,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.74,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6794045634607658,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.7415,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.5939261186945629,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7406,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.6441179848324072,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.7066,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.626246960289174,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.7232,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.550071253832395,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.6986,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.7189689523974143,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.733,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.5803663818632857,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.7533,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.6397171126891444,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.7365,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.6294095061152363,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.7444,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.6067333734351715,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.7244,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.6042543764329238,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.7355,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.5920299644947065,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.7331,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.6995350448952782,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.7662,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.6993264151865388,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.7319,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.6117393167876553,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.7216,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.6396886154165112,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7149,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.7007358909717539,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.729,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.7789660869271491,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.7377,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.9333648230800345,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.727,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.6580991876635299,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7318,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 1.063063557404282,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.7601,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 1.184876260885402,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.7428,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7177734763384346,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.7316,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 1.0668309364461168,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.7334,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.6774333941067255,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.7082,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.6732114311155132,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.7223,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.7342083249705929,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.7321,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6154543622786308,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.7044,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.6956115408168532,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.7357,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.6619847557306027,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.7057,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.5878446802457633,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.7223,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.5991063363507843,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.7159,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.5790637337717411,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.7343,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6606855488040851,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.7565,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.6219888660130205,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.7287,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.686669966748809,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7178,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6178766941764763,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.7137,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.5816627227070387,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.7461,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.6125108913941943,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.7344,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.6556587962455535,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7006,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6586055490668822,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.7365,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.5870603028484443,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.7221,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.70136493190662,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7307,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.5507015544250633,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7339,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.7579411635031736,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.7133,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.6133079773881273,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.7085,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.5684718041507141,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7239,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.631928668901881,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7504,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6336564914475888,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6932,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.499946415677406,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6797,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.7107672329125726,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.7183,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.5547470634635769,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6964,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.6474238176361249,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.6957,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.7062744964067227,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.7229,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.6241259744564678,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.741,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.6515078221821883,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.7094,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.5394323606793987,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.7237,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.6095625425176738,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.7056,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6172731007495936,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.7392,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.5595711205920004,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.7014,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.6623168603638904,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.6945,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.5663884275832212,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.7401,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.6868503050802344,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.7147,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.5702299989803181,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.732,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.5590381566120586,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.7371,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.6055933898086717,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7122,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6416274980270033,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.7326,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.5476747846573933,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.7066,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.5877574220877326,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.6939,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.523562713012341,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.7212,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.5502327939266493,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.7123,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.4791777702716554,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.6999,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.5782403490916153,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.7169,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.5297107713225085,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.7271,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.5182417252749495,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.6967,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.5276897861941668,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7149,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.531723330349942,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.6711,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.5796042178013429,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.718,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.5390299263901398,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.7096,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6028598723102316,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.7095,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.5571689119541224,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.706,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.5539995784967549,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.7003,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.6837171640245477,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.676,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.5538951068344181,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6745,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.6729689252704022,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.7096,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.5444358076131028,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.7084,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6595309782085004,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.7055,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.5317078314207873,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.706,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.517911484006942,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6872,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.616542593091935,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.6667,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.5735637452580938,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.5731,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.6746694139642946,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5794,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.6242722065161886,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5591,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.7101909660989344,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5507,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.6639512935348818,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5624,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.6366666204975779,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.551,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.6921403875800817,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5762,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.6364373815002052,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5877,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 3.826681643677377,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5605,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.8116843150437565,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5667,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.6411935615619888,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5743,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.7582451162060722,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5886,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.8019448787411588,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.593,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.6506019577910462,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5478,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.6448267826726746,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5642,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.6238309385487583,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5537,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.7272129314519626,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5376,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.7098804415794944,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5858,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.7890239011796226,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5719,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.6942469768358287,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5788,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.612112747165591,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5175,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.840404086178476,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.579,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.6423078868165512,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5506,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.7883048319631082,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.554,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.7702251335288742,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.5555,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.7563274629821547,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5495,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.7013200843384998,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5581,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.660619352555484,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5612,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.7247187342084234,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5554,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.7999761391209388,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.533,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.6764357323937197,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5579,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.8380511497180624,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5485,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.7276480881073278,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5562,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.6866653240399239,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5454,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.8026062511833789,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5494,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.7171869596936482,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5468,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.9135293425830762,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5582,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.6786871245521915,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5652,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7381811723508628,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5893,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.7320949531670509,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5534,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.6650500483768336,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5363,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.6527041194118853,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5644,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.6400964618176189,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5851,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.5917570095987915,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5586,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.5417310974946382,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5726,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.5791574565324539,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5592,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.5466814394220549,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5483,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.5594521861126135,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5676,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.5620715711055998,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5663,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.5566730258039203,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5962,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.49323119915682934,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5401,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.6630058036392946,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5666,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.5699066453296573,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5855,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.5289703789161617,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.554,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.5391442054377428,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5585,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.535600282463096,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5535,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.5465666658215828,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5648,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.5089931797583095,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5799,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.6173030375394424,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5723,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.5768171198790069,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5576,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.570763339643696,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5337,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.5591765383015607,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5806,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.5439187614108945,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5828,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.5270026387338699,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5654,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.5092771864230822,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5469,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.5886422738491766,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5563,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.5595241852487232,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5502,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.577631632658033,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5658,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.6506103642695971,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5851,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.5287995565293417,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5617,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.6294587538519871,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5759,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.5724220296326376,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.56,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.5777957366408841,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5418,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.5458690114020456,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5737,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.5243788709608046,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.573,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.6261408512204654,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5593,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.5787289389136574,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5329,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.5640096076099105,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5505,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.5961263264470968,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5826,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.6214225477923013,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.558,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.5523716101822423,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5423,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.6134148059011193,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5527,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.5823552732169466,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5849,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.6452104876129356,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5688,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.5264755078806451,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.5543,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.6321584310396682,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5842,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.5485900270930143,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5719,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.5617665574135458,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5477,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.5213204158079547,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5511,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.5603127267172388,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5655,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.5581021763626922,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5411,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.5718793695862455,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5802,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.6200353610197415,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.5603,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.5054617341205304,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5537,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.5439147669540596,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5473,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.5369943131114929,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5817,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.4866353559680135,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5629,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.5591552134122262,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5704,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.5265482804404384,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5925,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.5476072629626323,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5717,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.5382885063335417,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5511,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.5985074024613601,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5741,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.5001811092312017,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.561,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.6266189290952527,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5582,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.5254904390591227,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5642,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.5275779903529928,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5716,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.5861315821238481,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5617,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.517133102539322,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5397,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.5953815150189339,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5652,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.5676101173048175,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5843,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.564835835505198,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5802,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.5989294456169791,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5722,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.5351861803440486,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5719,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.6850983865110504,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5639,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.5118352672122104,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5642,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.6002919030667598,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5407,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.4992891152870657,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5636,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5292049130954618,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5629,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.49384950398431143,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.6028,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.534743933754103,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5582,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.5282198732101683,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5547,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.5622449759377719,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5358,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.582083831617939,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5536,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.5059942159626537,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5483,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.6025292993496548,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5539,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.5210713078445697,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5548,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.6556948141258002,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5795,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.5068242288152303,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5589,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.561890742952319,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5615,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.5180651088043136,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5764,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.5558645037020938,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5637,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.556180177428061,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5788,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.620303106965505,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.575,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.5420809342060263,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5613,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.6852184942000812,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5453,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.5464797013446153,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.552,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.6644124623746328,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5576,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.6125871540632157,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5562,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.5655663220881555,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5518,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.7080046950141624,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5482,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.5636614125512427,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5761,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.6505951351223528,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5652,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.5632854421530793,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5714,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.6697577422240389,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5611,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.5411960908973954,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5616,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.6539927267453391,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5842,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.5416185040352602,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5781,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.5199909055823728,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5653,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.5711196945061209,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5822,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.6131783865438132,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.5729,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.5712811792776132,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5703,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.6200828658856946,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5658,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.49778296588628745,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5618,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.5657624910553918,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5815,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.5878450855274067,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5717,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.5334431736292033,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5678,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.584413094503241,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5521,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.5179200790990399,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.569,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.5309456217745878,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5603,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.6699877782689999,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5704,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.5569351181884489,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5879,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.6013735766912347,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.568,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.5907195647270309,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5699,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.5798999542704003,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5619,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.5603361495478024,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5546,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.4980258423648179,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5499,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.5360984424878351,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5585,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.5651102114619427,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5785,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.49093889408070185,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5683,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.603023516838559,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.6025,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.5127548587831127,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5451,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.5668363205935075,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5738,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.5465360912034068,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5799,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.542363003707889,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5655,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.59048689366975,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5598,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.5631754771276604,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5742,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.5284358515604691,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5623,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.6264832875549537,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5693,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.5441883760270967,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5731,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.5805645514103459,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5803,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.5295494793312746,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5758,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.584098328263877,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.5843,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.5889181866342664,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5704,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.621605450580845,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.568,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.5793831299607144,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5574,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.5813361257052053,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5567,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.6326583711008914,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5566,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.5761251924716608,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.55,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.5135174708949741,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5579,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.6993640341210392,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5716,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.604332692939568,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5701,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.5308453303224744,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5652,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.6220665741227448,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5613,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.5128332296581554,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5606,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.5215734426093811,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5699,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.5029939853844808,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5475,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5256021286621395,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5606,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.49297983052648237,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.582,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.533634014311471,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.5817,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.49422942930335784,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5671,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.4778814080287163,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.5511,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.4837307727328528,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.585,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.5500334254216247,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5414,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.5849289852654976,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5834,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.6851441005524191,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5687,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.5189215797336858,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5835,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.5417395855672873,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5636,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.5375128932183207,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5909,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.5234687199223308,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5589,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.5492798148376636,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5801,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.5216064665460479,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5755,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.5595995263539512,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.568,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.5787068210236129,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5733,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.5156361442059805,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5774,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.5517322274176978,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.5573,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.5116958522428543,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5793,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.4374843553859036,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5606,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.5107860606830668,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5664,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.5517011075394017,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5416,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.4938613418874763,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5786,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.6002097621485626,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5612,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.4668511065361712,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.539,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.5446809002765682,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.5441,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.513871314037061,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5623,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.6309549073142006,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.5892,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.5251738497408172,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5598,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.5356732350823828,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5913,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.49957411875354996,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5617,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.5053236582967193,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5813,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.5376632808172528,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5711,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.5264738873562244,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5604,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.5058250577427734,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5622,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.5619328434134006,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.5734,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.504117507554053,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5578,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.5344232205022844,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5444,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.542897169249038,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5779,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.5446527187364558,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5646,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.6563590681895782,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5495,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.458377203466753,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5748,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.589142467063088,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5517,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.5047439076367575,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5683,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.5219554397319108,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.5938,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.5610614696756229,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5584,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.4768747817360498,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5832,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.5524499227272227,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5619,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.48976163475619316,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5459,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 0.9637096513249046,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3982,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.6784088925714888,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3797,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 1.1211284469671667,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.3808,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.7341042896773003,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.372,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.6250716646523884,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3993,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.7041332087934312,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3668,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.5374319045681779,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3778,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.6747468633166239,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.3694,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.5796378706655496,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3716,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.7169252121484504,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.387,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.5923455179637341,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3712,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.7191855197186855,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.383,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.561015234512655,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.364,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.6218243233014099,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3915,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.5661357886088846,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3814,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.6613016694129545,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3741,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.5092402990824846,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3743,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.643867696108929,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3701,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.5821910795154506,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3919,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.5604263008463369,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3629,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.556627941515685,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3686,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.5487303276027243,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3788,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.5531435327878613,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3646,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.5119487618591579,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3601,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.47385108061587916,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3697,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.46636033552064693,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.395,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.5164568963586864,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3723,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.4632398487049904,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3476,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.5414993587489378,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.3882,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.5106417529490032,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3526,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.4935023766001845,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.4003,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.4962737363109778,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3764,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.5215470140009274,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3657,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.5091065918734566,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3809,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.5116740243363739,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3745,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.5809853542581201,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3774,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.5083997052542013,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3717,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.5684361811395124,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3907,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.49159889350931424,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3714,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.516541629153083,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.361,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.5264248985280915,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3501,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.4896135743972396,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3916,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.4948607532820462,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3651,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.5300389840191382,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3908,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.5422498632155157,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3878,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.4859533659199391,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.3828,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.5180874988994738,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3525,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.5600417286299402,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.37,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.4786697663526683,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3691,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.5855520748672085,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.3869,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.47973043035897395,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3703,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.4830445826890563,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3909,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.4944475263150758,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.382,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.494549166298531,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.3869,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.48317548669490096,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3779,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.5137299790307739,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.376,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.5263533697113645,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3876,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.5080565428094569,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3794,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.5209845321040182,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3779,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.5006422796540798,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3936,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.4871508684162012,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3809,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.5048571381659989,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3563,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.5045764775256539,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.3659,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.5489574168120245,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3799,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.5264355041290949,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3626,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.49818156806871744,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3747,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.4846642635688406,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.373,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.4933413311407337,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3801,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.48614236219612356,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3783,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.4626689981244063,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.367,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.49472063956912343,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.3874,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.49891347391466423,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3872,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.49162064191563726,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3763,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.535817745463568,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3786,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.48363116707760806,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3935,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.46621437282626493,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3711,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.45366947556239157,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3506,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.4982295832820097,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.369,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.505021500343266,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3549,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.5067020239638199,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.4017,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.48577969318615377,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3692,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.4993138735138936,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.4023,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.4713858767516867,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3559,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.49611353900642297,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3783,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.4926070009004814,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3695,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.5127146635260782,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3701,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.439900602010005,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.381,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.46014092848022314,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3647,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.5006932385954939,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3922,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.45555776102365086,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3744,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.4802799710743835,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3785,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.4399812834075029,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3731,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.5280483125931545,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3866,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.452866322187474,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3832,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.471727964356464,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.3771,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.45484382439749665,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.4068,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.45587172002295046,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3889,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.43428892518840406,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3704,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.4792483230198426,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.3912,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.4498413816011751,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3599,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.45810226932109926,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3824,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.47665839149519085,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4088,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.43950963113618496,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3813,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.459589931549135,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3941,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.44990611396238034,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3817,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.4866485269724401,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3885,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.4588947990038194,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3828,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.4861321898761674,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3895,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.46872877858564904,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3659,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.4639506113241653,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3833,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.45464259671732216,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3921,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.43674674600928076,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3794,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.46267841396224724,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.392,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.44731055538160264,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3631,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.4412543689825874,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3782,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.42747364308054003,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3492,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.4882090774966562,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4076,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.44100834795277016,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3721,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.4182795722919671,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.4076,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.44229573543816914,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3642,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.4539214140597022,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3598,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.4505472289539449,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.373,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.4459279184920543,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3813,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.43955258142756876,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3789,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.48273036512534023,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3764,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.4623190171915114,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3959,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.4297725872102259,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3804,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.47915984824744884,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3703,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.4481147850764211,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3914,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.4445224062951568,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3968,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.44106474960715497,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3759,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.4496087602520478,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.3874,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.4169412566432786,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3838,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.4393683746294493,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3985,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.4536234731271243,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3575,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.43249724371159254,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3803,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.4455456224923025,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3343,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.48693455029830657,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3768,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.4737319170337223,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3546,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.43844251345570706,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.4066,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.43387836754508985,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3778,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.45478159541473934,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3742,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.429227575536342,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.3671,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.44019252824460353,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3795,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.4307945280172593,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3873,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.4334408935524527,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4102,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.43041001926827577,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3908,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.4465353167225165,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3861,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.4610918424282069,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3835,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.45135454029604377,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3516,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.46353046285071625,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3878,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.4962048029254238,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3785,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.43484519494615126,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3813,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.4318658787472834,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3803,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.4525706830681805,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3906,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.4870941088441526,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3563,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.4431208612601278,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3607,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.44811316994325123,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.394,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.43132295851062724,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3727,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.4180888264034889,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3649,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.44388053368540653,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3828,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.4189860602301019,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3823,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.41351649179546596,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3647,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.43787112635667697,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3754,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.42432578579784197,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3815,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.4160717414871724,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3784,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.4337575403060488,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3848,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.4516733768234508,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3745,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.4137848502900846,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3717,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.44073957150031523,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3954,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.4237346187227689,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3879,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.43325847194937095,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.3723,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.4506545001455313,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.4073,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.4347006099588823,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3751,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.4345505164884983,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3805,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.4452345094901938,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3733,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.4175318294031625,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3744,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.45481399860615157,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3667,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.4505795724352439,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3747,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.47412129027342137,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3912,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.4350052243907834,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3907,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.4151741748886296,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3729,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.4678484830651789,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3851,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.41427139376165145,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3894,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.455668902633862,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3765,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.4536903122836467,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3776,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.42428346364618863,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3532,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.4882832960072155,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3896,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.42473995627750316,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3573,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.444775989082852,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3733,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.43376639236258485,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.3872,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.4309873891508825,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.401,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.43060396458198724,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3729,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.4074422641695221,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.4042,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.4622599057492962,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3548,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.4432671873312026,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3905,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.44388161350095556,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4002,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.42372986857950207,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3845,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.4531028269740001,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3663,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.46479959256672476,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3854,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.4458617973886558,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3689,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.4258610489316151,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3773,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.4215890754563039,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3795,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.43033738558357226,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3872,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.4358432443333096,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3813,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.41503243846559007,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.3807,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.43455856275717814,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3827,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.4376556953562068,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.381,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.4580047796383711,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3635,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.4256932949079572,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3734,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.4253751549408488,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3869,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.4264394596961722,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3593,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.43259058321159727,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3889,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.41393510476507794,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.375,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.4085456133995289,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3894,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.40762588158841584,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.365,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.42410295178513735,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3887,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.4333907735116901,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.376,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.41197919446286635,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3796,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.417992300147192,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3753,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.4237499490481243,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.384,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.4237777714755167,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3976,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.43233185344798275,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3657,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.436681855840781,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3807,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.4237179563310809,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3851,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.4234437140811551,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3783,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.42866920061086083,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.4017,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.42720649161521274,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3911,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.4445389561091741,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3648,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.4468194323832039,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3729,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.4252248926456682,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3912,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.41473146055442733,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3844,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.4404499221556875,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3623,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.4150515447859639,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3983,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.4282595764545946,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3729,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.4233086568343235,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3797,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.42749779588620845,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3715,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.4341048039222509,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.4073,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.4405692629220085,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3535,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.41819834768473063,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3645,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.4482313628058863,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.366,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.45555348584764255,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3763,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.4118242103030459,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3731,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.4198377285299964,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.367,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.454284285685915,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3641,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.3935147644678996,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3786,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.49827090041795863,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.3084,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.5570821521121857,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.2607,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.429446041878255,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.239,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.6863850206090812,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2554,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.6784381748678522,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2412,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.4780430219569316,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2628,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.5535168710898904,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2722,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.5130995015330254,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.257,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.4799694025465978,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2284,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.4104062031923791,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2481,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.4751658307608332,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2444,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.4930817371771081,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2331,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.42216540038906814,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2197,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.423960657645564,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.243,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.40838388079220367,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.2262,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.428046009764779,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2527,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.4612745385014847,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2525,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.3765483652362484,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2173,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.4059287430934035,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2577,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.43997178922860963,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.2465,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.36835490308381386,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.24,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.39442759628397567,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2585,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.3652795422889439,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2505,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.3698912972267679,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.2501,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.38244172986412495,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.245,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.36616223124654396,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2507,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.3660051274502978,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.2534,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.394140529317022,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2425,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.38088824982053454,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.2607,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.38202506096924077,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2439,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.37854017647136284,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.224,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.3786524568135752,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2188,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.3706936019045695,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2526,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.37029481830414046,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2082,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.3847368165382648,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2615,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.36486334217896205,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2509,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.3903652008149532,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2233,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.38566131267964304,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2379,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.36658630224169164,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.249,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.37201407269135556,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.246,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.3747484420571046,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2409,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.3710243007477183,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2167,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.3670259505783207,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2211,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.35972571227757055,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.238,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.36803586202105765,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2685,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.38740193641928755,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2255,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.3511641639517541,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2631,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.38777707845558546,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2291,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.3857991049857386,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2192,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.358426667909315,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.2459,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.35735821288829067,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2593,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.3882676679693686,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2277,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.35170335939794645,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2474,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.38078054002878947,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2566,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.3864337954337756,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.2277,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.3681622336990674,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2358,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.36892047444651627,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2499,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.3528121705483557,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.255,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.381680591367992,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.2448,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.3556303132481771,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2408,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.3502116470319604,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2561,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.36675412000222224,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2456,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.3533155671997567,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2715,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.3780210097972922,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2342,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.376689364400828,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2026,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.36953504816107097,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2204,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.3768747520061229,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2493,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.37109349995224783,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.2242,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.36411455274601,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.2252,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.3771874014619997,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2237,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.36208146515091877,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2412,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.36977123661210365,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2575,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.36759092563536394,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.2661,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.37062543222407396,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2037,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.36328712913902295,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2479,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.3627622436621377,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2337,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.37920378679420386,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.2269,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.35961396698889664,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2439,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.38551848660543353,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2316,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.3690741723292421,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2323,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.34716669573764974,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2292,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.3789257664735025,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2402,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.3510599250914695,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2587,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.35659092965831746,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.231,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.35904604514641647,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2201,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.3637628431596781,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2358,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.37560807050048783,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2406,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.38209008943166844,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2585,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.36220645456599354,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2431,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.38959651244596466,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2281,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.39037047594015756,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2405,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.35922969489390033,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.2236,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.3538140719378987,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2283,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.36402598636316025,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2393,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.35437169097639837,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2498,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.3604116768061497,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2362,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.377009154334376,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2211,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.34374054406045684,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.2281,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.3554532796519636,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.233,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.35764987100966716,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2219,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.3455758710619347,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2363,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.349284225187927,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2718,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.35980183063167676,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.2163,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.37094867506904483,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2331,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.3582364676165723,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2193,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.3729203123919673,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.2184,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.36148495442870365,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.2248,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.34899492813895694,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2271,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.3866207860041546,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.2338,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.35208718242690173,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2644,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.3688733339720294,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2639,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.39006940619178004,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.245,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.36208467910033526,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.2359,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.3585074137894943,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.232,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.34953078291827416,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2345,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.35119671542224906,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2305,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.34126078220923317,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2374,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.3602346415572256,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2517,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3628681707636662,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2221,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.3614850107235767,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.2548,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.35086595196033543,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2297,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.3464055571730444,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2509,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.3649434227014736,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2281,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.35418908098307056,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.2179,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.34620612744415824,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2446,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.3640229960698763,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2477,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.33831168874490586,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2365,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.3568429881108622,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2421,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.3426841687902959,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.2112,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.37676760434834033,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.2425,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.3491725043330073,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2343,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.33972042517973555,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.2327,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.35862015362549887,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2563,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.36887694303662694,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2481,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.3668330577163821,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2408,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.3629634501411095,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2416,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.3693454561691696,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2258,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.373389323716455,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2395,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.3439747632535296,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2452,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.3397306019977132,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2517,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.3693913417517541,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2329,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.363712603886341,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.23,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.336121707235696,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2411,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.3663818484701192,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2372,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.3537359830424942,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2256,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.3597961378945756,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2363,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.36728805904248596,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2516,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.3466095816583572,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2129,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.3712152735615179,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.241,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.35999645875697844,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.2548,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.35236787499382943,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2393,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.3527265847875911,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2383,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.35194448694034725,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2256,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.35652011175567366,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2254,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.35745218871946427,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2301,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.34988123025335327,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.2433,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.3611449461901992,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.2283,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.34481087209785677,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.2273,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.3585916898842242,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2299,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.3479775294882408,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2421,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.33398542895731176,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2277,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.34505947974399986,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2082,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.36199106304117956,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2342,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.3667399772816841,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.2346,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.3638523980170062,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2318,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.34698310868114474,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.2103,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.3478841033717775,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.2207,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.3558141420780378,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2478,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.35842568657508944,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.2212,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.3305307397998585,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.2338,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.34939598392627275,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2324,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.3342825066081085,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2399,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.3397052758060018,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2228,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.3568688329698101,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.2404,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.35631057186988235,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2248,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.34118608534376116,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2454,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.34470546029797716,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2688,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.3507330238439248,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2352,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.3456200597246797,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2564,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.34730397575718125,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.25,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.3459425109588164,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2291,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.3425306458152667,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2286,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.3469946966278309,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2352,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.34883864740216525,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2263,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.3434824724033804,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2099,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.3453823541966748,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2368,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.34795476082564203,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2253,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.3409587501704064,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.2529,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.3456507537266118,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.258,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.34136787067522867,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.236,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.3670288459370995,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2283,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.3510630501621811,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.2331,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.3467680405444758,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2472,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.35306347901924673,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2154,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.3594580981225422,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.2565,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.3361545736413372,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2059,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.3504317044612151,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2379,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.34350162172918586,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2459,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.3466120309300032,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2424,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.3440131188508643,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2345,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.3627048790885728,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2216,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.3221543157438691,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.2481,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.3419418445644442,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.2334,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.3544621594471216,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2305,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.33628262143393134,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2445,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.3393491946497233,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1919,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.33537485137577694,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.2129,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.36651532303584006,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.2182,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.3550475917497399,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2269,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.3312980988315202,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.2595,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.3307025436009807,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.2543,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.3518762551545779,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2346,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.3410730086064957,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2267,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.33778789243239876,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.2542,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.35766976840277154,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.2325,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.3494362397797124,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2326,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.354874005693334,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2528,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.34066932371021597,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.2434,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.33571325618613684,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.2289,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.34884257531957436,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2109,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.3413650524726005,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2546,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.3399442331098133,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2392,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.3278565372206573,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2467,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.34816661479408706,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2077,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.3450381801767028,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2424,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.3377844557144171,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2487,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.33194128606518714,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.236,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.34815731496376845,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.2116,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.3481636989407873,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.219,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.34751111032255205,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.22,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.35692818581573316,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.24,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.3340126635046371,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2333,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.33586100370625976,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.232,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.34541541371569817,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.2491,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.3343779856570918,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2508,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.3406883643147221,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2529,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.32988755437379325,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.2423,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.35751028622148373,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2161,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.35021645157033676,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.2371,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.33515209571073024,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2407,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.3316231007130163,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2496,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.338003431658423,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2444,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.3428176406264192,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2546,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.3290076454339016,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2347,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.34708990813086216,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2313,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.3343478708173886,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.2431,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.34148378086907427,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2405,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.49478389557046715,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.1751,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.39509479047480284,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1812,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.3676547794531944,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1877,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.2883728465025368,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1769,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.28947420759301495,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1633,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.4174396800510771,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.188,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.44838109760432593,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1789,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.47555320284164043,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1907,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.3667258278485744,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1712,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.30616434621313693,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.163,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.33198594809689547,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1512,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.3541823969811357,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1662,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.3499499485361766,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1849,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.32897253798471604,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1878,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.3238530112407101,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1931,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.2799072959470771,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1819,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.2888832214919998,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1565,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.32708276672021824,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1728,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.3144302386212357,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.1902,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.32864443031886986,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1828,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.298666669702849,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1582,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.29440440980317556,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1732,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.2814080643936291,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1677,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.2710758863516794,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1528,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.2773386502377593,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1542,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.2935569792091573,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1507,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.28974412053400395,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1775,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.2735130341936111,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1529,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.281770196640326,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1872,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.2698237082007911,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1652,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.2781860725906129,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1747,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.2724834789169913,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1716,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.27592262098143305,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1613,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.26712407777203434,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1675,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.2780202071998212,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.1751,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.26458256710660294,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1634,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.2670476635266229,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1645,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.2677198000143161,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1994,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.2662111846911306,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1537,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.2743163011108146,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1836,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.2627591441546112,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1662,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.271445765612377,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1796,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.27534757022435163,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1693,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.2772465609122766,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1734,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.2582779651719193,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1554,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.272094345356668,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1651,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.2840066811534647,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1857,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.2569797633268909,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1578,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.2649770818395828,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1695,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.26457098401397844,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1642,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.26751735766007256,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1595,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.2494693010391481,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1863,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.25996152544787887,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1611,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.2971659369087782,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1519,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.2615860731202491,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.1556,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.26402241677147503,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1485,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.27801093675350974,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1935,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.26344255737526695,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1779,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.27532996579081076,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1839,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.24946074283022668,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1457,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.26301539130124424,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1748,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.2617008144422121,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1849,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.26157492709381786,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1536,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.2696715604083065,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1558,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.26214254375650725,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1827,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.2637156419348116,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1626,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.2696072142336843,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1744,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.2688453424179305,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1732,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.26352793170325667,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1651,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.26107235134694434,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.1731,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.2576025404995936,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.169,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.2728589001925702,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1694,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.2647459864839984,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1709,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.2694704708337589,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1742,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.2552370051048809,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1747,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.27514248455056767,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.185,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.27720459274913495,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1641,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.2599187008039048,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1864,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.26832345504989197,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1716,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.2572351276377806,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1471,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.267637175730133,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1816,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.26845224014301966,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1582,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.2558701013350939,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.1487,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.2743645427006471,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1625,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.2557778251012477,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1702,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.2564936980643495,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1769,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.27439646339722296,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1658,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.2587362553466058,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1654,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.27250058301359037,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1905,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.2728440093802848,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1619,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.2704741143525539,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1539,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.2658244517468011,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1686,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.2698457274214847,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1813,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.2811662252506199,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.1754,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.2611876041562969,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1572,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.26742483844971904,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1633,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.2675926962794175,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1586,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.2748167226993404,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1315,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.2662054197790258,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1531,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.24811323160035825,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1545,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.269053135573093,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1599,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.2682565535309368,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1698,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.2593357931834693,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1506,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.2580106371093722,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1672,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.26361749263121653,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1764,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.2675018442300554,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1563,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.2776318793798391,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1654,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.25523257490718565,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1849,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.27340525376072977,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1636,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.2689654797302995,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1664,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.2657226120499649,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.187,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.25434846797552113,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1491,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.2623498019533576,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1755,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.26492407294706866,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.162,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.277354971275243,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.1674,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.260681902985479,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1714,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.2807714864442416,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1549,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.26980005508090027,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1893,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.2711074595547318,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1762,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.26174007859984516,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.17,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.2689308924072279,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1961,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.2558147355460892,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.1769,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.24899487830490946,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1436,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.2675193384344742,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1539,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.2659005148679619,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1637,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.26322920250742143,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1782,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.26446474420129645,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1651,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.25809794532112207,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.189,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.27074321194153256,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.159,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.26122252411392183,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1879,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.25633972812015343,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1736,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.2739831725778443,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1752,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.26258567476788414,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1505,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.2703493795995899,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1685,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.26205198464296187,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1478,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.26465934440646993,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1743,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.2651769973374547,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1863,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.26830554933937456,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1688,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.268011857362914,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1671,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.26764240633798625,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.1546,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.26443322721987833,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1691,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.2564004744779444,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1656,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.2707511529184889,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1442,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.26075933195458456,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1908,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.26406762360634406,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1613,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.25492006378860776,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1796,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.26208516007467964,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.2042,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.2840123638708133,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1754,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.2608478110556892,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1504,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.24623866608530517,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.187,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.27240157210081284,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1766,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.27323543806137557,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1734,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.269327567226204,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1504,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.2671861544947398,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1727,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.2494385183337245,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1468,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.26242942467417146,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1618,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.2624756784680058,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1459,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.2684685433403018,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1639,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.2554651749016498,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1604,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.27032446162801654,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1567,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.2541898547966036,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1875,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.2650892834328462,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1853,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.248931354205672,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1566,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.26386727905196217,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1545,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.26900910666242306,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1682,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.26123915624338995,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1648,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.26476602400472377,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1835,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.25086711775858517,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1644,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.2642564196871151,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1929,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.27411406062334387,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1709,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.2686532701282629,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1713,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.25793649824424214,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.1813,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.253281275669205,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1401,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.25638276879021243,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.1434,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.26755855258425587,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1481,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.27344425334121464,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1805,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.27173543955462975,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.1592,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.26543561105954205,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1621,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.2535096807318926,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.1533,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.263958240901493,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1996,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.2535013176474823,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1772,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.2659058133634098,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1794,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.2571448433915386,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1596,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.27104686416503093,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1665,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.26574394096558107,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1629,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.2624791783876033,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1458,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.26682033524691773,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1729,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.263983578237984,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.1617,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.2699068583378983,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1627,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.2585431118568067,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1591,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.261892821949762,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1561,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.2604876136985379,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1783,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.25078012548771633,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1558,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.2586097674578887,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1964,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.2737702741766769,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1968,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.26858207911690535,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1563,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.2573242839772949,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.1478,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.2660377221616584,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1333,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.25552961261654883,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1724,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.25754289686374404,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1334,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.2690096077155245,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1756,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.2674507565673869,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1774,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.2564121912323577,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1748,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.26509412351038975,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.2,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.2645249646445966,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1481,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.26981064386576065,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1687,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.26782116245749465,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1884,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.2514024102268895,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1578,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.2753429290093358,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1452,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.23701454113236958,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1582,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.26276166534114004,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1631,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.2684990721535729,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1679,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.25790993659023015,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1601,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.25603315530584453,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.174,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.25777661381204153,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.176,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.27077553694979056,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1673,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.2608617591444112,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1434,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.2554304354593244,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1755,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.25967802577506915,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.2016,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.2760692261055218,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1641,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.2639131073660296,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1607,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.2678188779459348,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1794,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.2563692107054463,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1602,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.2709785015743075,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1623,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.26099997141621345,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1885,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.2645757824631569,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1891,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.25947618098034825,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1704,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.26501637031177094,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.1808,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.2661891635099563,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1788,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.26040408618766353,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1603,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.2666521347929813,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1601,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.275122912083145,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1892,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.2698195550962356,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1776,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.26051283549308396,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1381,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.2731372263982851,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1821,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.26752330107722505,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1666,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2577065884920341,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1614,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.2736598953539788,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1587,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.26320274095288565,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1884,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.25841368892509026,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1765,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.2545444414757548,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1495,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.2565064312660701,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1926,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.25697630418562484,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.169,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.2610147404934558,
+      "learning_rate": 0.0,
+      "loss": 0.1747,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 2.1322399681253212e+18,
+      "train_loss": 0.42330787913586065,
+      "train_runtime": 80213.6281,
+      "train_samples_per_second": 1.97,
+      "train_steps_per_second": 0.015
+    }
+  ],
+  "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": 2.1322399681253212e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}