diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.982278481012658,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 6.662616662871593,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.9892,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.590321802308029,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 0.9984,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.529092064430264,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.9818,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.580326512665684,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.9997,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.154461219803104,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.9757,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.748477898895604,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.9533,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.596830177325959,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.9471,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.088749125391237,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.9232,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.7965332701970933,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.9115,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.5981242735498347,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.8991,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.436468930211653,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.8697,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.7199418731932794,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.8788,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.03631829809906,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.8803,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.506120838059873,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.8375,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.135041037492064,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.8349,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.2028157757701434,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.8058,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.2530647984386487,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.7989,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 1.9735441476885516,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.7731,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.3119082916871343,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.7553,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.0985130850653735,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.75,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.3980557253200991,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.7558,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.4200168533198263,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.7363,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.1521628322341657,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.7302,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.0609839650993058,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.7293,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.9931943991588859,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.7091,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.8365837745308097,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.6938,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.9479305662944937,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.6993,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.7892857692379727,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.6822,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.7134570590295328,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.698,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8283524315621241,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.6802,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.6836464903343163,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.6861,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.6256741704373667,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.674,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.6180080792760823,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.6674,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.5971856521608635,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.6901,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.5817199470212611,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.6644,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.5862353085955999,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.6501,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.5493701367512476,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.6401,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.5928652264695229,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.6466,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6175048659704155,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.6583,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.5267792105514205,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.649,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.6267276322300351,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6368,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.5879410782796199,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.6547,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.579119380632237,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.6487,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.5536357122149453,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.6386,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6440463480080174,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.6456,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.5969164979312146,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.642,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.6941442283175786,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.6213,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.5674403936287549,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.6416,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7613034435536075,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.6283,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.546098593165548,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.6398,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7004400083064217,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.6306,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.6938752321534223,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.622,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.7442891019697032,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.6167,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6532876477203335,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.6159,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.7615960804172206,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.6177,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.8125133785814551,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.6115,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.6955098795323231,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.6217,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.930153737786519,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.6145,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 1.2520679799982435,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.6212,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.6665072454549082,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6217,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 1.4021037545558581,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6164,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.6984869752541061,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.613,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 1.0421697494640707,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6018,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.9564331805089213,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.6016,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 1.2541164301507812,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.6146,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.6438802821469878,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.6071,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.4850075989011402,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.6154,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.6804490656752605,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.598,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.2360301882574185,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.6208,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.9845828033437889,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6072,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.8810894440419179,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.6191,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.6394558318437567,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.5894,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.7815861564158628,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.5917,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.6120364139406278,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.5973,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.8403973040811326,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.586,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.7760928908744718,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.5976,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.8231551189433023,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.599,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.8991426546061697,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.6015,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4038894297214197,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.5952,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.6989406427462892,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.595,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.089928812519747,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6037,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.163976712017112,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5905,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.1383686622400617,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.5955,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.8706227253909136,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.5983,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.017959869079347,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.591,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.2458070070866119,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.5901,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.0271175205218575,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.586,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.3278108265068314,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6116,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.1386736222961396,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.5827,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.9918784464069685,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.5899,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.4547387754811616,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.6056,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.400160798783366,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.5799,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.2330236840004383,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6095,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.583219541876609,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.5804,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.178011358001323,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.5898,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.6344780154952743,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.5995,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.3604645600541714,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.5824,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.4797109903673789,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.6084,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.258048160198841,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.571,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.0868972151093133,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.5873,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.1411094595852844,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.5939,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.0482017905773338,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.5985,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.0281445569107588,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.5682,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.3089774911414347,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.5954,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.9050933515353694,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.5915,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.109227275131008,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.5898,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.2088560380933804,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.5863,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.5168612749988678,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.5733,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.366020374689941,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.5916,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.8089722464384175,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.5851,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.1138172263941497,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.5989,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.9038833714951986,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.5922,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.6554626858885813,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.5806,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.938247949958785,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.5975,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 2.1147578133483353,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.5877,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.6843944012718022,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.5907,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.3708052329326457,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.5896,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.4111298333169702,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.5957,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.119310317389887,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.5937,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.3277418450239493,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.5962,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.3979105135782535,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.5891,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.0343732925783384,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.5829,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.4071079781053675,
+      "learning_rate": 4e-05,
+      "loss": 0.5868,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.117176790163501,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.5808,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.0918072064926827,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.5839,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.8923095006282145,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.5793,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.3286810915260028,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.5911,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.6836579005720632,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.5774,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.9886351400429372,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.5821,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.9776654727233999,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.5764,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.0434760113247248,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.5883,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.8461224650013642,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.5761,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.2558926949350244,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.5711,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.8014348886351269,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.5802,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.8159759166672971,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.5904,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.156123891730944,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.582,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.6288709417077698,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.579,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.7248624599026035,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.5803,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.8509851283710458,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.5849,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.7436866833790663,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.5789,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.7969894317606165,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.5741,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.9288341186854239,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.5672,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.7512118653579475,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.569,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.5742652173216287,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.58,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.6630136841432545,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.5755,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.7924001565977292,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.5687,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.6320146084626354,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.5904,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.7179039700716864,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.5748,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.8179640908853416,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.5815,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.6046327009880448,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.5712,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.9268691131523122,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.5716,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8212237541329644,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.5691,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.9718011040866575,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.5797,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.9504443084280677,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.5793,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.7973407236124045,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.5736,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.8166193281625769,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.5762,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.7993547202219162,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.5765,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8309210120777175,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.5659,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.5386549413883037,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.5652,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.8468752768809192,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.5744,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.7143246656798474,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.5829,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.6046708701594374,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.5838,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.490959684099286,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.5703,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.6100103535392198,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.5787,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.6260382159766842,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.5773,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.5160177925978527,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.5646,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.4717199618417172,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.5674,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.7596038761731944,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.5523,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.7658653336959556,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.5707,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.7996775665534328,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.5782,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.9695711089109967,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.5696,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.7602598016247715,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.5768,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.49838348681228256,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.5729,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.8132405121780625,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.5584,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.8754128232563635,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.5758,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.5987946942735778,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.5482,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.6275636435708395,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.5631,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.7363926699245539,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.5694,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.5446733185769901,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.5634,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.6537198725489846,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.5705,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.5612857591149255,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.5643,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.616684542302835,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.5615,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.8476139792872723,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.5653,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.8347599185798479,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.5733,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.7098307830515225,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.5586,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.7654219323712057,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.5717,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.7532912481290563,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.5654,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.757072854755886,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.5612,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.6375185536445552,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.5695,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.48735384759085293,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.5701,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.5148648510641181,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.553,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.5964766391485741,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.5579,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.5444909275348857,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.5808,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.48880616216736295,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.5568,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.4532540601513749,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.5651,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.5346708043227476,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.5555,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.4767961289208517,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.5657,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.5503076626769017,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.553,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.5712613189585695,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.5596,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.46624856028219686,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.5599,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.5099186373757741,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.5669,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.5508366985958343,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.5535,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.561361155837794,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.5511,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.6766365487908581,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.5591,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.6435125359545428,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.5649,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6739838573416004,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.5716,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.5277820259803565,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.5617,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.655839997861402,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.5673,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.734273808626907,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.5593,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.6031548553751666,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.5644,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.558627721799987,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.5727,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5523976030658423,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.5504,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.6731992859002756,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.5872,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.6462736774748615,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.5656,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.489426709528649,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.5619,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.5264280281471407,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.5598,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.5771991871379302,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.5639,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.6460064143964647,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.5654,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.5201049689441594,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.5424,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.4678662111953695,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.5723,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6154198321763803,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.5567,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.5923437735591786,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.559,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.543509565671562,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.5613,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6013461111564484,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.5606,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.5958691173979005,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.5624,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.575978049149666,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.5586,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.44408904699939195,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.5694,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.5228741080099514,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.5602,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.5742233396954913,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.5411,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.4393339488944027,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.5727,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.5257811470534609,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.5578,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.5873423252375396,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.5559,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.4449944158912531,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.5497,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.4589423278384248,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.5663,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.4264764457235144,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.5517,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.537374926363578,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.5689,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.45576825442730606,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.5511,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.5829105128145059,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.5577,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.47506276990780877,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.5594,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.4071433926942792,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.5643,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.4624710911887666,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.5479,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.41991304501258164,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.5491,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.44780528364855726,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.5577,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.5542682999369771,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.5351,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.6392803109588199,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.5537,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.4920322256282272,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.5557,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.7308564662072053,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.6301,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.9152635433649577,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.4986,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.5953743711982408,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.4897,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.6094128135915415,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5053,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 1.100666245808355,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5145,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.7718105818403568,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5051,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.6079250525919511,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5013,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.9602120940533156,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5303,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.9940591570847567,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.4961,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.6280205788463633,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5158,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.8159742100701881,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.4962,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.8721735819162872,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5184,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.6467384626149201,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.4962,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.5727965687682227,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.4926,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.707829885315636,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5141,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.46853668368247064,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.4999,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.546896614308456,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5085,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.57728891206334,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.509,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.5914316499637375,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5069,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.4346446410804504,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.4929,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.6132466536242743,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5166,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.5002745237900053,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5056,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.5701226808136606,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5007,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.5668931707931708,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.4787,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.4603502228622988,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5218,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.563947361723255,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.4722,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.43731651567184165,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5111,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.48158274811342366,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.4981,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.4684813105501885,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.4931,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.42710152917052363,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.4985,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.5237990631855314,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.4956,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.5408402944209869,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.4977,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.3984432483440159,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.4913,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.5522184298229761,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.524,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.4512677054321439,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.508,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.46736234183345093,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5165,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.44137525257057764,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.4923,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.3813254878289986,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.4907,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.4675724970479699,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.4928,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.4756196072201426,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.4976,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.5170049136906921,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5087,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.5177163584122241,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.4863,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.6110347763819207,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.4955,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.542994064981615,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5241,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.547294633377755,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.4828,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.5219941067063374,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.4959,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.4470075739071652,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5097,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.4463664214642283,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.4939,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.4654326704029975,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5045,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.5074600707108874,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.4808,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.5697860008705563,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5309,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.49972884023710484,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.4639,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.46459218843478267,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5112,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.48371392669386803,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5152,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.6123852077291384,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.4993,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.4407544243710736,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5098,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.441471958777221,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.4966,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.4670689044669711,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5064,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.35985742136020066,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.4823,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.43460334597521155,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5094,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.40175092987358696,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5102,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.43516303407040147,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.4983,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.4011307992190346,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5201,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.39873095219549987,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.5101,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.38512749607460284,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.4977,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.4340014982936869,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.4941,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.37355980707226877,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5095,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.3927393383060636,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5014,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.35140443283154915,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5085,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.4010225266358918,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.4969,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.3790447875727764,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.514,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.43401498544367784,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5087,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.40826732930401427,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.4992,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.3848732696068618,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.4911,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.4508512309810857,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5352,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.43197451764236744,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.4679,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.48468066082867445,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5189,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.4879767151448927,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5009,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.55249947771327,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.4955,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.3653782467892423,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5017,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.5272242519114112,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.479,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.4852407008839443,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5038,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.4515846762466498,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5117,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.3690513585327604,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.4915,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.45763064044389273,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.523,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.4889688678507937,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.498,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.40074051114408304,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.4861,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.42123698778682495,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5274,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.49506935121546386,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.4936,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.39780095086416717,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5203,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.5033249273706359,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5075,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.5650316993774965,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5035,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.5041529172359794,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.4949,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.46814997187392604,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.5204,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.6727389692809194,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.4865,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.4139896018936916,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5073,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.5222640001514327,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.4874,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.4596638042341951,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5002,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.5123775549722365,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5016,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.42158088809208877,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5094,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.415074886737434,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5036,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.488445267595726,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5083,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.49177514614504464,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5057,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.4208537963016274,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.509,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.5540759116462696,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.4848,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.447696091774962,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5124,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.49621806585898254,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.4874,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.5499030152100579,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.4985,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.5015917486861217,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5089,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.41935125590143596,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.4713,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.5544136256140294,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5117,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.42912514986764005,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.4884,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.44742973734594127,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.509,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.6064809277590381,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5146,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.44838026757027777,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.4976,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.40826587799267877,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.4842,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.5342576933353755,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5157,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.4080213188014965,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.4899,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.43979677393383115,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5061,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.3755069218624205,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5021,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.45524564819024277,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.4942,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.39125884034464997,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.4808,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.3785083221498995,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.4995,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.3919717399070924,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5035,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.3990344672841378,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.503,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.35961405604598057,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5103,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.39020151337133696,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.4986,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.35443698165735904,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.4834,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.3717990110627283,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5036,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.34454907668536244,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.4944,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.43080147396132,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.4976,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.3949728261408338,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5168,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.45491436947791475,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.4989,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.36899860626082387,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.4928,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.46551739843202156,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.4972,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.37540520643011366,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.492,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.3758078976164837,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5203,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.49364423810290514,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.4965,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.329780414569527,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.492,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.44885071158003903,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.4975,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.4181723142914954,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5053,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.39915557710566524,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5058,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.3996178901110703,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.4879,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.38397820366959823,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5056,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.40777760571223504,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5128,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.41889289382265144,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5148,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.3538131461474258,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.4971,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.46747261444905336,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5105,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4114354493344334,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.4729,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.44995835921477717,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5164,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.40882186340729876,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.4874,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.4426922526963698,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.4825,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.4822246888613928,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5159,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.49500506374418113,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.4938,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.46293586055421704,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5086,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.48715935595117604,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.4854,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.39662634680414194,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5089,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.5385907818622812,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5075,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.43885777719251684,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.4853,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.43303063438850536,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5069,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.4779893712694188,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.4917,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.39798743286802957,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5003,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.4134397676560219,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.5162,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.3732798454756735,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.4883,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.34177943482230133,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5046,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.41544139934809987,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5126,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.456365552550996,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.4937,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.3231464685470071,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.4844,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.4783806534864812,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5108,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.38615496008570577,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.506,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.41635564661405305,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.4879,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.39499934907847006,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.4955,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.36738245041155143,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.475,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.3891641535643607,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5188,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.41664269272289806,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.4995,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.38357952397918554,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.4954,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.3971161915444133,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.4918,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.46705207009819577,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5093,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.3909078296481515,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.4793,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.4280269521271874,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.5092,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.4756313941629605,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.4911,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.3956097509486246,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5069,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.468115304251085,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.505,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.4087832794036067,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.4792,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.47832889983780486,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5288,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.44469282366169055,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.4844,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.4229848785448364,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5076,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.43640784027153734,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5036,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.5158682593479009,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.518,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.3960240888870027,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.4961,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.5367258240301733,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5128,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.46483712572392377,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.4814,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.4288823935093076,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5121,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.5067719013864692,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.4991,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.44447287277046194,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5011,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.4189755161273104,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5017,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.48008103276160974,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.4948,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.4190798377193123,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5058,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.45222527422129905,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.4913,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.45178072952324916,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.4899,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.4018823688343576,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5211,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.3781240734987887,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.481,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.422347376371009,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.4875,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.36988817215642195,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.4944,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.48083738019668504,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.4879,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.3896743554274215,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5183,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.4663791580791228,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.4811,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.43790125912070654,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.4979,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.401484098787838,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.515,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.4948017393746451,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.4726,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.3652646519609087,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5097,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.5940260484404566,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.4851,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.4506952140763045,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5021,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.4537729812843372,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.4874,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.45582138627745933,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5158,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.4544690382883625,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.4934,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.42813517485191255,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5257,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.4619328847645678,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.4999,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.41886277382415954,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.4881,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.5156634088699437,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.4826,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.396068631645756,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.4912,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.5517694369382735,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5118,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.3839397397112393,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.4968,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.4360456441149106,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.4911,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.37862226658127834,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5045,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.40556635167153565,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.4927,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.37075498271986473,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.4933,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.422890015618768,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5089,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.36248420988033386,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.4765,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.4897455310710657,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5025,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.3907760155682443,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5055,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.5978995799751671,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.4887,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.4762612074065911,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5117,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.4348733108174687,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.4964,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5641411443596913,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.4976,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.369977376207787,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5104,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.5063492220682845,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.4898,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.42820469369016634,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.4802,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.35895094379620657,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5246,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.5223598690805838,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.4918,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.4390914016640701,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5109,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.4742302009892399,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.4812,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.4205255326507264,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.4986,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.3840420539447363,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5013,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.3520790374087347,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.4833,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.3098258950625639,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4798,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.3691612367689034,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5019,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.4151862174413598,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.4926,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.42949766583560384,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4378,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.4571825041705832,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.407,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.5394792360307351,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.4302,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.4850509993379135,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.4542,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.5332834477037787,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3859,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.46647610822258817,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.4116,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.5668609827549376,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4581,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.40418209293062224,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3993,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.475994177366702,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4266,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.42214283694012206,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.4182,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.44954730204385246,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.4096,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.4640028288511275,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.4274,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.3745260767540929,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3985,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.433452059670225,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.418,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.36556055365936424,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.4245,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.3991825970771593,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.4252,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.3864293203223652,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.4152,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.3833524472008076,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.4311,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.3451144267130315,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.4225,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.29915579297307165,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.4018,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.4178150931651441,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4196,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.3954946400514365,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.42,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.3566355159709355,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.4175,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.41556910381988166,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.421,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.36694353132359675,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.405,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.40116372318974974,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4318,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.3779805554276021,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.4113,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.4025098250839837,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.4239,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.41725416022546014,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.423,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.37193910834298843,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.4116,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.46093591735553513,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.4163,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.3383486197074377,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.4105,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.4524180985627073,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.4087,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.38346426078448287,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.4048,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.3705635799629847,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.4196,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.3443786707320739,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.424,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.39336097056753466,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3994,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.3755313530593026,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4327,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.31780452987145047,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4074,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.3194818614725666,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.4312,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.35742850452202923,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3944,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.34111969339416315,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.4378,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.38742047257472195,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.398,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.38050107546007,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.4367,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.33295168463902325,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4196,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.33156730877251384,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.4317,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.30100523594236156,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.4011,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.3435810715096179,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.4178,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.3065663535241235,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.4177,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.38722298599491606,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.4333,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.31380376932306303,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3945,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.3454827473302313,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.4257,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.33701961587472357,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4154,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.3317685493829687,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3952,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.37307005793624304,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.4428,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.31094126572720415,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.4064,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.35401256455069074,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.4243,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.36703874215579935,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.4277,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.38029137217650305,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.4093,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.3241589148441376,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.42,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.41346284212664114,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.432,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.3369684263337746,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.4038,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.37789305522305044,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.4021,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.41138855185220113,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.4351,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.3242277272698431,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.4298,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.43634315631309467,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.4332,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.3465163311927188,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3918,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.36699421426141676,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.4386,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.3780857486592971,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4371,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.33602361350084176,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.4129,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.3551083445957222,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.4189,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.37752377110678664,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.4239,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.41334285449399777,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.4093,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.4132555055502049,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.4312,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.40813604116164476,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.4001,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.38233857579790653,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.4082,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.36437690401641515,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.4103,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.3656916515173736,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4332,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.34400079301069564,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.4093,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.3487691069497065,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.4138,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.39824011635916123,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.4352,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.325649077541785,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.4325,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.37870068369267773,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.4266,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.3344235409565685,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.4042,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.42769550183740235,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.4255,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.3776417000675652,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.4222,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.49808761105242066,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.4152,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.3053382888653074,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4201,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.43333078366507777,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.4237,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.3691238011117673,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.423,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.3503929228376759,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.4143,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.35175281901806094,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.4125,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.33136707409797467,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4356,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.31057103659791835,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.4014,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.3554449959245304,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.4531,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.3316297566944148,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.4168,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.35127008698715173,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.4292,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.32352586287507473,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.4201,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.33253737589968924,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4372,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.3963552842507783,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.4344,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.30053650378270647,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4107,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.41078942176117345,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.4073,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.3344029056013768,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.4199,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.4391615154671474,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4417,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.32219326100858764,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4068,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.4239956446340334,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.4233,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.35596986735787783,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4431,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.33867078306697046,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.4012,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.3101666642908049,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.4062,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.3531891507172152,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.4309,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.301056575378034,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4171,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.34238792734596224,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.4346,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.30969050522440594,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.398,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.34976337174111716,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.4414,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.38405010750386664,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4052,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.3674556190097053,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4229,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.36806166162543924,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4138,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.41013966821167597,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.434,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.37605913917824213,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.4195,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.34469431224426395,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3948,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.34551128875009657,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4091,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.388241937070555,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.4094,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.3277675836281335,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.4089,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.35219986370213935,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.4129,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.352586929124702,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.4423,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.3259450957570288,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.44,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.401543857824389,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.386,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.3160472831018946,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.4283,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.39592204219791083,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.419,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.3051978164675245,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.4178,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.38902262975972784,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4194,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.3256347614740087,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.4238,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.34275562589373526,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.4313,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.3017537704274225,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4123,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.28971597144809025,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.4317,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.28533929015485254,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.4283,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.28414319817832523,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.4034,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.31373140462036303,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.4283,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.3157812465169007,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.4045,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.3123773657620353,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4296,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.3772801269157079,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.431,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.31714195801092226,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4093,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.34745764178671873,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4118,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.3537251349526658,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.4137,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.3157112353481535,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4293,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.3599647090361324,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.4192,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.3302280553161587,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.4215,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.3328655891889176,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.4154,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.3844068433465576,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.4215,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.3190849125474624,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.4204,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.32387804196020414,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4093,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.3308685144862466,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4457,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.3140857539728255,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.416,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.3602956258521807,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.416,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.33336686916080616,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.4345,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.3195632109884459,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.4049,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.32441279351256885,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.4107,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.3521633770242431,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.4413,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.30944661293240844,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.4033,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.3467225855298678,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.4288,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.34408369111125264,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.426,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.32255115338990065,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4124,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.3071210423147393,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.4302,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.31596602628609927,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.4004,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.3112573634166517,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.4264,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.32145286419622826,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4197,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.34247369784636544,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.4262,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.3287872156051601,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.442,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.2869153970093758,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.4104,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.3525790400022746,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.4357,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.3008317034440441,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4058,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.3466019944655964,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.4332,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.33425276871783566,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.4175,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.3667639480420127,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.4274,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.3408589931847111,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.4286,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.31706661718918516,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.406,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.3134482768477086,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4167,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.34086682321081446,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.4414,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.3188121899622525,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.4213,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.30013970041458116,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.4044,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.3422178454126141,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.444,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.3225376989863517,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.4203,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.3544503982052626,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.4113,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.3051110167569122,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.422,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.34234007842777453,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.4111,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.3251634251767516,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4261,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.32727629872361563,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.426,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.31324158530452045,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.4212,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.3028942045069718,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4101,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.3146563829047906,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.4371,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.2902241522809416,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.4228,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.3469366603008498,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.4257,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.2747364754958515,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.4014,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.31625062197131976,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.4044,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.3251256721381961,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4283,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.3467931962343039,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4386,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.3000876287233663,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.4177,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.31895593892375057,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.4334,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.3201248849481094,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.4276,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.2880143930427725,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.4288,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.35086715521508205,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4162,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.2854496031107259,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.438,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.28089849374433656,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4215,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.280999077847332,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.4123,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.2738193658867251,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.4226,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.2986265058872976,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.4251,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.27651117746483245,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.4056,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.3048080266124469,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.4405,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.303213573863703,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.4354,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.2917079647724254,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.4117,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.3027208359086882,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3913,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.278236689489087,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.4298,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.2923824043184262,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.4361,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.32385212846023703,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.4179,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.29943844164536104,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.4131,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.3608635253927078,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.417,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3366977076727583,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.4343,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.3241081927890642,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.4212,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.3294122959669625,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4184,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.3014148673298466,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.4249,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.32479729362636145,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.4115,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.3180553341882973,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.4361,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.31088776637258864,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4152,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.3061624836954432,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.4224,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.3256512044062037,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.4095,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.3043161268784425,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.4157,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.2898970567303764,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.4345,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.305666370326013,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.4261,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.32195837907668823,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.4125,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.3152565756802517,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.4121,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.2803936116798484,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.4532,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.3453050837857198,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.4123,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.27776672421661164,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.4188,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.30011214185750096,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.4188,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.2708449578400798,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.4136,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.2977834229084369,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.4118,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.28224646316603297,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.4241,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.28443628108981445,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.4053,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.29464008820123594,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.4665,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.3264451208020082,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3674,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.32438720228860474,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.4327,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.36143761365924953,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4433,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.345488696585721,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.4174,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.3357827563666393,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3939,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.35783456930786617,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4248,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.3122771777695455,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.4301,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.4187836841101248,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.445,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.4180177517752738,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.3241,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.4099001259957694,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.3674,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.576179883638116,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.3415,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.41839514526960464,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.3369,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.46404378541072955,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.3377,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.435897238631139,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.3611,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.39087592074352145,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.3322,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.46206692724368675,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.3487,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.3768744445484221,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.3246,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.47691459520287666,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.3074,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.3813855369868436,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.3533,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.407481916383,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.3331,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.35971013650915107,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.3309,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.3955249971429173,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.3474,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.4104854455484034,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.3382,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.34110150976675335,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.3249,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.3976783196150078,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.3308,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.37892304668387955,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.3556,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.36725848415357915,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.3283,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.34340957035594444,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.3304,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.34301754770354004,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.344,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.3349291604671934,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.3262,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.33164628170102356,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.343,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.3303719649641221,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.3316,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.33000753682149875,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.339,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.3233207609535081,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.3564,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.3003553751064546,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.3231,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.31252310814048706,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.3498,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.32694207497239497,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.3474,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.2943279153590369,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.3126,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.2899195455565239,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.3179,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.37513431339663006,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.3467,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.29990656704051577,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.3313,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.306202793774844,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.3187,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.3043374660692077,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.3386,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.2785839976131833,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.3393,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.2953462568955936,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.3456,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.2882326323653571,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.3245,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.3044351753505888,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.3406,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.3233163845164041,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.3542,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.30089632925496324,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.3244,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.28308290034905215,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.345,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.3066934492859621,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.3434,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.27799913806905063,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.3322,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.32111547831257625,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.345,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.26161517971921955,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.3297,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.2925288692174803,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.3393,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.29309915311442186,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.3398,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.2815008770492994,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.3412,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.26995642231297484,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.3206,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.2947007975673466,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.3531,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.3110158127025877,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.3402,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.30193005765024106,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.3431,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.30501322644931844,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.3419,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.30487798128705235,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.3401,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.27767503815474237,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.346,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.3075016067097669,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.3457,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.27658667290860134,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.3402,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.28721448801078153,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.3365,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.3154563678008809,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.3411,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.32207655262972457,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.3296,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.31822201181270765,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.3535,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.26731026620078757,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.3398,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.31767297356048896,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.343,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.26933169523483563,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.3342,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.2912237260006869,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.3264,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.3083295521674947,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.3604,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.2914189587612868,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.3479,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.30567931099300233,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.331,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.28024660155780806,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.3435,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.30579483868772417,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.322,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.32228824420707153,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.3551,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.27348523410622466,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.3393,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.32382317083600975,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.3524,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.29756735364123127,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.3169,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.2787886537058838,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.3367,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.32025003978004213,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.3587,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.2684920552007293,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.3234,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.28322024591229494,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3397,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.3093164243435819,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.3592,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.2814365183436773,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.326,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.2983114901618155,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.3434,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.2916865390769206,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.3198,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.2714155595701574,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.3244,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.3004331998546704,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.3358,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.30478750467547366,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.347,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.27009531091752753,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.3362,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.28538707127441354,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.3259,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.3360021328236316,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.3388,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.2801544493909381,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.3662,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.3268880079171038,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.3346,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.2841929594432747,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.3319,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.3047001915654386,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.3404,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.3140848839246428,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.348,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.2800740578504856,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.3369,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.3572223559813274,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.3452,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.2587956141909544,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.3165,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.33588204057641824,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.3446,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.2812583357275076,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.3251,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.2889629397986029,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.3384,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.2941709901161932,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.3247,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.30757610995179535,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.3533,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.26143756514699984,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.32,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.3104386795394737,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.3513,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.2932914183984801,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.3408,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.2788441135269714,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.318,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.2994248953222668,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.343,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.2815662882918097,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.3358,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.28575338853498145,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.353,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.27201815734896484,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.3297,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.26507047608340173,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.326,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.2790204116047137,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.3345,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.2831335889928299,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.3334,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.26755455341981776,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.3534,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.2693182623452552,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.3262,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.27186150013021776,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.3383,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.25488700456418273,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.3067,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.2880806675929237,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.3647,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.28209955818353066,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.3372,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.27576655951419743,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.3388,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.2872023098901688,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.3252,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.2831532212262706,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.3484,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.2796788968404012,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.3347,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.2779437695638498,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.35,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.2623369742032426,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.3179,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.28699545168893387,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.3467,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.2852040506076711,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.3467,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.26160283203163115,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.3204,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.28855156006852256,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.3572,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.2821885175162862,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.3417,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.2878707394709164,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.3448,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.28075149800634025,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.3497,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.2648061626370504,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.3216,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.2780060482526703,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.3394,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.2969755026856433,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.3368,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.28982886951614384,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.3608,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.27569504325513666,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.3311,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.2706540038726229,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.3338,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.2974620564347139,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.3453,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.28606265528223984,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.3355,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.2750711805062159,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.336,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.26085861091172396,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.3252,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.27487349299912517,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.3608,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.2631158001208826,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.338,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.27057171019056914,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.3236,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.30206396063109686,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.3641,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.277730028947812,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.3385,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.2626593378055619,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.3353,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.29061258316333627,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.3527,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.2775714871576441,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.3315,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.2726071565146685,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.3295,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.2795899219492836,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.3501,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.2714733207298279,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.3438,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.2637233071130505,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.3248,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.2646923951246073,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.3383,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.2717973209812442,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.3241,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.2813568886896034,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.3476,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.27516135368081013,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.3296,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.27240001971861216,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.3452,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.2553546633045854,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.3324,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.25969486539575226,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.3469,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.27502301373530214,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.3365,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.2727797022865313,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.3504,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.2649584354011078,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.3305,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.24678729450178177,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.311,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.3062206661988717,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.3694,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.2621371028573883,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.3243,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.2632817132143415,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.3299,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.2697084314720634,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.3452,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.25284967845691464,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.3416,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.2621539120033613,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.3392,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.27544940881108776,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.3374,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.2615395997368971,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.3244,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.27368162385769557,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.3389,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.2642010408121073,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.3325,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.26781274877619243,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.3424,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.2643279938850239,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.3311,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.2662462050250192,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.3333,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.2573710513457158,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.3222,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.2626134631399547,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.3531,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.25583351773802804,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.3331,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.272307162055375,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.3475,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.2585898206625426,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.334,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.24779746015442033,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.3197,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.2627535132033043,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.3571,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.2684780884518755,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.3305,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.26957846689022935,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.3413,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.26422078742288446,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.3473,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.2509736545152725,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.3185,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.27586978505966897,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.3473,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.28745304054597864,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.3383,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.24871245014845486,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.3222,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.27548655186408666,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.3407,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.28603030998544915,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.3251,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.2872560859198482,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.3577,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.2617953053238492,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.3097,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.3088190253188793,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.3426,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.2711410764375707,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.3304,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.2743939636363093,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.3303,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.2965311121508827,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.3521,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.2805675254532581,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.3396,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.27756718474990016,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.3385,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.25985501989018955,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.3165,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.27506198577408986,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.3326,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.26231820336646755,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.3434,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.2744075391114722,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.3578,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.2403915314695108,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2998,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.29229530601987735,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.3632,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.26938272913262806,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.3284,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.2562154191501439,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.3306,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.263084240807043,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.3301,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.2666812105361788,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.3314,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.27643265738614553,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.3569,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.279265233073344,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.3537,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.25405463717015553,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.3357,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.27098907952350076,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.3415,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.2579337517786352,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.3383,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.23613676324371777,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.3115,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.2833523467192545,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.3636,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.2566595569487224,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.338,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.2529460772552101,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.3413,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.2427976164415439,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.3198,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.2583520440502811,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.3277,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.24110482013638382,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.3315,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.24305680349352768,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.348,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.23749998477760398,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.3265,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.2721824101037565,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.3457,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.24315128780477294,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.3246,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.25547747134273313,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.3519,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.2569545010252269,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.3296,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.24251249593783025,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.3264,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.2555988933665625,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.3579,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.2281588188511257,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.3142,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.26574420286142125,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.3403,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.2685662345072967,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.3548,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.2448084490374798,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.3271,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.24721089039270494,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.3285,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.2446914203789275,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.3289,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.272924973347164,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.3614,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.24969765135496885,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.3231,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.23616868672786626,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.3291,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.2489045061303957,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.3473,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.25715921001140185,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.3379,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.23666151931815524,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.3306,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.2598629346142014,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.3425,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.2572929129352196,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.3403,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.34409277812331274,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.3538,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.3868495142325845,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.256,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.3649851450019248,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.2888,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.28298369731853784,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.2837,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.3930062670343488,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.2708,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.52381602405521,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.2838,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.42203188750487636,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.2679,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.33983988523847514,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.279,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.3422222580576081,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.2758,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.35317097965302396,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.2662,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.3843928050023469,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.2883,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.33080492674817813,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.2731,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.2917771462965878,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.2677,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.30989710448776264,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.2786,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.3325828495654465,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.2809,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.3308272903507982,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.2722,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.32118594803429185,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.2682,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.3069275241668295,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.2981,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.2788321702104338,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.2831,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.29213618271563446,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.2664,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.3270181894698904,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.2874,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.3036514798961664,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.2718,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.28463056143833104,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.2756,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.2741624845150095,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.2874,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.2737222212552246,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.2715,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.2891694021030767,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.2678,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.2771329733000229,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.2708,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.29288484206813786,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.2839,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.2794105556041134,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.2758,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.2731330352070759,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.2703,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.26063783626593123,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.2762,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.26643477924270215,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.2824,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.25400360697956686,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.2659,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.25929045174136417,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.2714,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.25654436222676646,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.2829,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.26786226072457897,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.2737,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.2497007376593628,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.2606,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.26729740348814296,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.2729,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.27416523502934814,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.2732,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.24247648655510465,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.2578,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.2557736926145904,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.2775,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.2755218521190632,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.2964,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.25209463507156626,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.2807,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.24882994961200122,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.2759,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.24094370964913916,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.2581,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.2589658331827684,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.2818,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.24943626243763883,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.2752,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.25877742900571893,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.2824,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.25034867792448995,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.2756,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.25169319378085414,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.2827,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.25772848776763485,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.2807,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.25652405917191295,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.2772,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.24717596892742025,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.2643,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.259186396655701,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.2828,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.24479874458358644,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.2552,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.26043507770525254,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.2819,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.260486524900932,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.28,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.2603386816468642,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.2733,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.26812207284330725,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.2788,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.24786847570091722,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.2664,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.24664745482121767,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.2822,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.24553229616949088,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.2839,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.24422076551173724,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.2664,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.24924412269831803,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.2749,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.24588459694634096,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.2745,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.2442122745149059,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.2682,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.26683487953034546,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.2952,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.24652689110155215,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.2618,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.25795580776569665,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.2786,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.2599852210892758,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.2947,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.24034517525614715,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.2569,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.2567197831535816,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.2927,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.24626097396769334,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.2694,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.26028180798670036,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.2968,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.2366019842993696,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.2604,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.249499268565977,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.2683,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.24840326729857828,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.2738,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.25084650636762407,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.2806,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.25330406431584546,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.2685,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.266128513558524,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.2861,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.24402560050274621,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.2626,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.2589809406823865,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.2832,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.2459412651903684,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.2736,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.251567734398992,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.2792,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.24817860335619554,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.2615,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.2611378952407828,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.2775,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.25608113010506034,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.2851,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.2443455971954493,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.2552,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.2556250205462524,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.2735,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.2603026578424663,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.281,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.25112978887760345,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.274,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.25112245194811816,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.2754,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.25600206764867,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.2712,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.26282120274056037,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.2914,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.23730478300474553,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.2618,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.2519051610645686,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.2739,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.2443463785958939,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.2732,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.24633241071510906,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.2727,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.2558737519231689,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.2796,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.25048028790152865,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.2734,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.23909895113491736,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.2608,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.2510441216566675,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.287,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.2386792287319763,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.274,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.24202343751595473,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.2666,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.2585047725452716,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.2851,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.24655755706867227,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.2732,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.2534816339596381,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.2832,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.23898067639769302,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.2675,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.2334488739054537,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.2638,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.24214303358759962,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.2753,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.24714916171703852,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.2891,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.24445862127860257,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.2769,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.2428345889414152,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.2667,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.2581772088907234,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.2956,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.23714812209999891,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.2539,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.25649663539615414,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.2756,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.2544403939142219,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.2808,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.23688725822152584,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.2669,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.24588531676462513,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.2693,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.2576212024886822,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.2863,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.2362827871091994,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.2778,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.24453978650685368,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.2646,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.24310932339175764,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.281,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.2593732710586703,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.2877,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.24366580068659185,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.2705,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.24474752207108944,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.2757,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.24077582209752088,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.2851,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.22881643163254828,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.2561,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.2581018077643161,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.2792,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.2602331978955641,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.2851,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.25113428336876004,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.2715,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.24635664610399885,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.2619,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.24123840283113698,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.2656,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.2417481786443334,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.2794,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.2382081813768977,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.2733,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.24528183763800893,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.2886,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.23819783533032035,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.2644,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.2524032429185898,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.2917,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.2441860022888359,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.2735,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.25271881481190894,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.2876,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.2320347741666031,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.252,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.24308291912671093,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.2527,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.25228396022877403,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.2712,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.25420098462295626,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.2892,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.24721295742502888,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.2762,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.24741880525898688,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.2824,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.24245984531467862,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.2684,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.23504173708312998,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.2612,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.23859957598025475,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.2705,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.24125566022134118,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.2774,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.24842681424142984,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.28,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.23923711665034633,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.2639,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.2486666144843296,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.2808,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.2449571545300206,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.2819,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.2340336559289449,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.2694,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.24781141056898862,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.2979,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.23524171962275434,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.2738,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.231506174243513,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.2751,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.2405370092297729,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.2888,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.22877386752377152,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.2663,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.2511531131094518,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.289,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.24768150333579933,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.2779,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.24542183293052183,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.265,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.23756471487618355,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.2601,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.24581476480357733,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.2922,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.24227959655059833,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.2691,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.24149062767243587,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.265,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.24677851932790118,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.275,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.24073080041340555,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.2673,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.24838328341017385,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.2892,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.24404324025021165,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.2702,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.24371377039345674,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.2806,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.23110589441255552,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.2674,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.2411997181586206,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.284,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.23510355657054016,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.2711,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.23549395128086936,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.2667,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.2408138262394178,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.2721,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.24840260964271113,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.2839,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.23084974363645067,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.2601,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.24189978058108075,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.2744,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.2424965351560545,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.279,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.2255598917270664,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.2599,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.24933413063334986,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.2799,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.25282243282108274,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.2965,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.237327461087362,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.2604,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.23657527175847917,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.2645,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.24167961650202527,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.2782,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.2531011959151395,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.2971,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.243278019898102,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.2649,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.25218050002087883,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.2921,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.23763039986876566,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.2561,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.2539843546789716,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.2898,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.22860519593439266,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.2652,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.23388538197126874,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.2632,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.23872092905824205,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.2753,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.2475221817042354,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.2788,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.24612021953181287,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.2796,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.23542703390257133,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.2751,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.24921953038714045,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.2817,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.23765722025634797,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.2765,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.24392317531078128,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.2842,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.23111611684402578,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.2566,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.2543654258368666,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.2991,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.23038371566277155,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.2489,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.2527420885001834,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.2983,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.23355157356831988,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.2539,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.23113626584624145,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.2751,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.23197042536259546,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.2671,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.24060909222221863,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.2798,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.23812728317589682,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.271,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.24409647308661095,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.294,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.23484013996554243,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.2579,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.2519742430057532,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.2944,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.23050194790563394,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.259,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.23299193824288766,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.2677,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.2419670434516861,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.2846,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.23623516955354798,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.2832,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.24285948198736196,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.282,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.23514496569565926,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.2583,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.24156731992624114,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.2768,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.2540787515294349,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.2775,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.24506553611035292,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.2779,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.2435812473303966,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.264,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.2365688578561421,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.2559,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.23848553407855877,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.2691,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.23843728679915208,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.2787,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.23746698779742198,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.2751,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.2421036327109141,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.2873,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.22541944697659358,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.2561,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.2511521898876772,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.2932,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.2573567968176219,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.2787,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.26102202216769954,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.28,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.23132932333352227,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.2627,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.24959325222645465,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.2926,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.23588152891199346,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.2532,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.24688059455954758,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.2838,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.2393379477510183,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.2729,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.2386430764857625,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.2784,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.24034922606331066,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.2743,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.2348682121637442,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.2563,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.25007812419020575,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.2807,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.23942726792210664,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.2757,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.24255946886977522,
+      "learning_rate": 0.0,
+      "loss": 0.291,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 3.489440295328678e+18,
+      "train_loss": 0.430346977032297,
+      "train_runtime": 84449.8491,
+      "train_samples_per_second": 1.871,
+      "train_steps_per_second": 0.015
+    }
+  ],
+  "logging_steps": 1.0,
+  "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": 3.489440295328678e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}